DRAFT
8 ENFORCEMENT WORKSHOP ON
PLANT INSPECTION AND
I EVALUATION PROCEDURES
VOLUME V
8 CONTROL EQUIPMENT OPERATION
I AND MAINTENANCE - ELECTROSTATIC
PRECIPITATORS
(0
c
O
35
o> U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF ENFORCEMENT
•5 OFFICE OF GENERAL ENFORCEMENT
WASHINGTON, D.C. 20460
/A/O/V
07-0O-7&-
-------
REFERENCE MATERIAL FOR TECHNICAL WORKSHOP
ON EVALUATION OF INDUSTRIAL AIR POLLUTION
CONTROL EQUIPMENT OPERATION AND MAINTENANCE
PRACTICES
Volume V
Operation and Maintenance of
Electrostatic Precipitators
Compiled by
PEDCo Environmental, Inc.
505 S. Duke Street
Durham, North Carolina 27701
Contract No. 68-01-4147
PN 3470-2-0
Prepared for
U.S. ENVIRONMENTAL PROTESTION AGENCY
Office of Enforcement
Division of Stationary Source Enforcement
Washington, D.C. 20460
May, 1979
-------
FOREWORD
The following document is a compilation of selected
technical information and publications on the evaluation of indus-
trial air pollution control equipment operation and maintenance
practices. The reference~manual is intended to be an instructional
aid for persons attending workshops sponsored by the U.S. Environ-
mental Protection Agency Regional Offices.
11
-------
TABLE OF CONTENTS
Page No.
Volume V: Operation and Maintenance of Electrostatic Precipitators
V-l. Air Pollution Control Equipment - Electrostatic
Precipitators. H.L. Engelbrecht; November 1978. 1-1
V-2. Selecting and Specifying Electrostatic Precipitators.
G.S. Schneider, T.I. Horzella, J. Cooper, and
P.J. Streigel. Chemical Engineering; May 26, 1975,
pp. 94-108 2-1
V-3. Electrostatic Precipitators in Industry. R.L. Bump.
Chemical Engineering, January 17, 1977, pp. 129-136 3-1
V-4. Electrostatic Precipitator Maintenance Survey. APCA TC-1
Particulate Committee. Journal of the Air Pollution
Control Association, Vol. 26, No. 11, November 1976,
pp. 1061-1064. 4-1
V-5. Maintenance Program and Procedures to Optimize
Electrostatic Precipitators. J. Katz. IEEE Transactions
on Industry Applications, Vol. I-A-11, No. 6, November/
December 1975, pp. 674-680. 5-1
V-6. Operational Monitoring and Maintenance of Industrial
Electrostatic Precipitators for Optimum Performance.
IEEE Conference Record, IAS Annual Meeting, October 11-
14, 1976, Chicago, Illinois (No. 76-CH 1122-1IA). 6-1
V-7. An Electrostatic Performance Model. J. McDonald and
L. Felix, Southern Research Institute. Prepared for
U.S. Environmental Protection Agency, IERL, Research
Triangle Park,N.C., EPA-600/8-77-020b, December 1977. 7-1
iii
-------
v-i.
AIR POLLUTION CONTROL EQUIPMENT
OPERATION AND MAINTENANCE COURSE
- ELECTROSTATIC PRECIPITATOR -
H. L. Engelbrecht
Wheelabrator - Frye, Inc.
1-1
-------
OPERATION AND MAINTENANCE COURSE
- ELECTROSTATIC PRECIPITATOR -
By H. L. Engelbrecht
1. Abstract
Increasing emphasis on air pollution control focuses the attention
of Control Agencies on the performance of the control equipment and
the system surrounding it. Continued performance at high collecting
efficiency levels requires adequate sizing, suitable design, and good
operation and maintenance procedures. Thus, it becomes extremely im-
portant to determine if the owner/operator of the air pollution con-
trol equipment maintains this equipment and follows these procedures.
The air pollution control equipment never operates by itself; i.e., it
is always influenced, restricted, and dependent on the process equipment
it serves. Thus, this equipment need also to be operated within given
parameters and under specific conditions. Operating and maintenance
procedures apply equally.
To understand the operation of an electrostatic precipitator, it is
essential to know some basic facts about precipitator theory; for ex-
ample, influences on the collecting efficiency by changes in process
conditions, fuels, etc.
Each precipitator consists of certain components necessary to perform
the tasks of charging, transporting, collecting, and removing of the
particulates to be collected. Various designs are in use, and the par-
ticularities of each design are reviewed.
1-2
-------
Common equipment failures are analyzed to examine major causes and to
obtain back-up information for preventive maintenance programs.
Although specific maintenance program requirements may vary from
process to process and from plant to plant, some basic steps and pro-
cedures are common to all. These include safety, inspection program,
and definition on inspection and maintenance responsibilities.
i
Normal precipitator operation includes proper procedures for start-up
and shut-down, as well as guidelines for trouble shooting.
In summary, the operation of an electrostatic precipitator requires a
planned program of operator's training, equipment know-how, preventive
maintenance, adequate spare parts inventory to maintain compliance with
the air pollution control emission codes and to prevent enforced limi-
tations of the plant's production schedule.
1-3
-------
TABLE OF CONTENTS
1. Abstract
2. List of Figures and Tables
3. Reasons for Good Operation and Maintenance Procedures
4. Electrostatic Precipitation Process
5. Design and Components
6. Installation
7. Operation and Performance
8. Inspection and Maintenance Surveys
9. Plant Inspection and Maintenance Program
10. Normal Precipitator Operation
11. Precipitator Inspection and Evaluation
12. Improving Precipitator Operation
13. Conclusion
14. References
15. Literature
-------
2. List of Figures and Tables
Fig. 1: Electrostatic Precipitator Collecting Efficiency as
a Function of Precipitator Specific Collecting Area
Fig. 2; Electrostatic Precipitator Classification Limited to
Conventional (Cottrell-Type Single Stage) Designs
Table 1: Comparison between American and European
Precipitator Design
Table 2: Design Factors Which Should Be Included in Precipitator
Designs Specifications and Evaluations
Fig. 3: Precipitator Unavailability
Table 3: Precipitator Start-up Checklist
Table 4: Precipitator Short-Time Shut-down Checklist
Table 5: Precipitator Shut-down Checklist
Fig. 4; Current-voltage Characteristics
Fig. 5: Changes in Voltage and Current Readings
Fig. 6: Optical Density Print-Out
Fig. 7: Relationship between Optical Density and Dust Concentration
Fig. 8: Electrical Power Levels of a Precipitator in an Arcing Mode
of Operation
1-5
-------
3. Reasons for Good Operation and Maintenance Procedures
Electrostatic Precipitators represent a major portion of the investment in
an industrial plant. Equipment life and performance ,have become essential
to the operation of the plant. Increasing emphasis on air pollution con-
trol focuses the attention of owners and operators on sustained performance
at a high level of collecting efficiency. This requires not only an ade-
quately sized precipitator of a suitable design, but equally important,
good operation and maintenance procedures.
There are several reasons for proper precipitator maintenance. The most
important are:
a. Continuously meeting present emission control codes
b. Prolonging precipitator life
c. Maintaining productivity of process unit served
by the precipitator
d. Reduction of operating expenses
e. Better public relations
4. Electrostatic Precipitation Process
Electrostatic precipitators are applied to collect particulate matter
after a variety of industrial processes in the power, rock products,
metallurgical, and chemical industries. The fundamentals of their opera-
tion have been extensively described in the literature. Two of the many
publications in this field are referenced (Ref. 1 and 2).
To understand the operation of an electrostatic precipitator, it is es-
sential to know some of the basic precipitator theory.
Electrostatic precipitators are sized on the basis of a formula which gives
an exponential relationship between the collecting efficiency, the size of
1-6
-------
a precipitator, the gas volume, and the precipitator rate parameter.
k
E = 1 - exp - (Wk A/Q) (1)
and £. = 1 - E (2)
with E = fractional collecting efficiency
^r _ = penetration
Wk = precipitation rate parameter, ft/sec.
O
Q = gas volume, ft /sec.
A = collecting surface area, ft^
k = exponent, usually.. 0.5
The ratio between collecting surface and gas volume is called the
"specific collecting surface area" (SCA) and expressed in sq. ft.
of collecting surface per 1000 cfm.
1/k
SCA = 1000 (ln£)
60 Wk
1/k
and Wk = 1000 (ln£^) (4)
60 SCA
with Wk = precipitation rate parameter, ft/sec.
2
SCA = specific collecting surface area, ft /1000 cfm
For a given application; for example, flue gas from a power boiler,
the collecting efficiency can be plotted as a function of the SCA
with Wk considered a constant (Fig. 1 - Line A). A different Wk
results in a different line; for example, Line B (Ref. 3).
1-7
-------
Experience factors considered in the selection of the precipitator size, i.e.,
Wk or SCA by the vendor include:
- particle size
- specific dust resistivity
- gas velocity distribution through
the precipitator
- gas analysis
- gas moisture content
- electrical sectionalization
- electrode design
- field height
- field length
- number of fields/bus sections
- electrical power supplies
5. Design and Components
Various designs of electrostatic precipitators are in use. A classification
according to the method of removing the precipitated dust from the collecting
surface is:
1. Dry-process precipitators:
dust removal by rapping and/or gravity.
2. Wet-process precipitators:
dust removal by water-sprays or overflow
systems and gravity.
A classification according to the shape of the collecting surface is:
3. Plate-type precipitators:
the collecting surfaces are plates.
4. Pipe-type precipitators:
the collecting surfaces are pipes.
A third classification can be made according to the direction of gas flow
inside of the precipitator:
5. Horizontal gas flow precipitator
6. Vertical gas flow precipitator
1-8
-------
A summary of various classifications as they apply to an industrial single-
stage electrostatic precipitator is shown in Fig. 2. One can assume that
over 95 percent of all precipitators in use are of the dry-process, plate-
type, horizontal gas flow variety. For this reason, the following comments
on precipitator components are based on this type.
An electrostatic precipitator has to perform specific operations to collect
partieulate matter from a gas stream.
I
o Charging of particles
o Transporting the charged particles
to the collecting surface
o Neutralizing the charged particles on
the collecting surface
o Removal of the particles for the
collecting surface to the hopper
of the precipitator
Each of these tasks requires certain components. For example, charging of
particles requires a discharge and high-voltage energizing system. Thus,
a precipitator consists of the following basic equipment:
o Discharge system
o Collecting surface
o Rapping systems
o High-voltage energizing system
o Precipitator casing
0 Ancillary equipment, such as dust handling systems
There are basic differences between precipitator designs originating from
the U.S. and those from Europe. A summary of these differences is given
in Table 1; other designs are also in common use, which differ from those.
1-9
-------
In general, the discharge system consists of small diameter wires spaced
equally between the collecting surface plates. The discharge system is
normally connected to the negative pole of the transformer/rectifier and
serves as a means to accelerate electrons, which, in turn, ionize gas
molecules. These travel towards the grounded collecting surface and charge
dust particles entering the space between the electrodes by attachment.
Thus, the dust particles migrate to and are precipitated on the collecting
surface. The collecting surfaces consist of vertical plates with ribs and/
or stiffeners. They are supported from the top and are mounted in parallel
rows, up to 50 ft. high. Spacing between plates is normally 9 to 12 in.
Rapping systems are provided for each electrode system to keep their sur-
faces free from accumulated dust and to remove the precipitated dust into
the dust hopper. Rapping systems act on one or more rows of plates at a
time; they are normally single or multiple impact rappers or hammers, im-
pacting at the top or bottom of the collecting surfaces. The discharge
system is normally rapped by single impact hammers, or rappers, or multiple
impact vibrators at the top or the center.
Electrical timers are provided to adjust the rapping frequency to the require-
ments of a specific application. Provisions are also made to adjust the rap-
ping intensity.
Electrical energizing systems for electrostatic precipitators have now
developed into systems including:
o protective equipment
o thyristors for AC voltage control
o reactors
1-10
-------
o transformer-silicons rectifier combinations
in oil-filled tanks, and the
o necessary electronic control circuitry to
control the thyristors to react to the
actual conditions inside of the precipitator
The precipitator casing provides the enclosure and the support for all of the
internal parts. It is designed for the requirements of the individual unit
with respect to gas temperature, negative, or positive pressure, wind, snow,
or seismic loads, etc. Adequate openings for inspection and maintenance should
be located ahead and after each electrical field, in each individual hopper,
and also in each support insulator housing.
Depending on the application and requirements of a specific precipitator in-
stallation, ancillary equipment is added to either increase the performance
or to protect both operating personnel and precipitator.
Such equipment may consist of additional indicating or recording instrumentation,
heating systems for support insulators and/or hoppers, additional rapping sys-
tems for the hopper walls or gas distribution systems, key interlock system to
prevent access to any hazardous area of the precipitator while the equipment
is energized.
Also, classified as ancillaries could be ventilating systems with or without
heaters for the insulator housing to prevent condensation. The dust discharge
system normally consists of an air lock at the hopper outlet and either a screw
conveyor, drag conveyor, pneumatic conveyor, or similar dust transportation
system. The air lock can be either a rotary valve, a single-or-double acting
flap valve, or a similar device discharging a specific amount of dust from
the hopper and preventing any in-or-out leakage of gas or air. The performance
1-11
-------
of each of these components is essential to the reliability and availability
of the electrostatic precipitator.
6. Installation
The installation of an electrostatic precipitator requires a careful review
of its future operating conditions; i.e., gas volume, dust load, and required
collecting efficiency. The precipitator manufacturer will also want many
more data; for example, on fuels, variation of load, gas, and dust analysis,
etc. Some of these data will be hard numbers; others will be expected op-
erating ranges.
Together, with other criteria, such as possible space limitations, structural,
mechanical, and electrical requirements, they form the basis for the specifi-
cation of the electrostatic precipitator.
A summary of design factors, which should be included in precipitator design
specifications and evaluations, is given by H. J. White (Table 2, Ref. 4).
After the selection process of the submitted proposals has been completed by
the vendor and a decision has been made, the vendor will design, fabricate,
and ship the precipitator components for assembly and construction in the
field. Fabrication and construction are as critical to the performance of
the electrostatic precipitator as is proper sizing. It is recommended that
the buyer familiarizes himself with the Quality Control procedures and the
structural and mechanical standards used by the supplier.
During the erection of the electrostatic precipitator, one subject is of
prime importance. This is the alignment between the two electrodes of the
1-12
-------
electrical field; i.e., between collecting surface plates and discharge
wires.
Many problems occurring later in the operation result from mis-alignment
of the electrodes. And many other problems not quite apparent during the
erection stage; for example, thermal stresses, will also result in align-
ment problems. Alignment problems can also be the result of lack of clear-
ances of moving parts inside of the precipitator.
After construction of the precipitator is complete, a thorough check-out
phase for all components is recommended. This check-out should be super-
vised by the manufacturer's representative to ensure that all equipment
and components are functioning properly. Only after this has been achieved
in a manner satisfactorily for both the purchaser and the vendor should
the precipitator be judged ready for service.
7. Operation and Performance
The operation and performance of an electrostatic precipitator; i.e., its
capability to achieve and maintain a required collecting efficiency and/or
dust residual level, depends on a number of factors, which are closely re-
lated to the specific advantages and disadvantages of the electrostatic
precipitation process.
Main advantages of this process are:
- high collecting efficiency even for small particle sizes
- low pressure drop
- low energy requirements
- adaptability to various types of effluents (wet, dry, corrosive)
- fully automatic operation
1-13
-------
There are also inherent disadvantages with this process which, if
occurring, can cause a severe reduction in the performance of the
precipitator. These disadvantages are either process or design and/or
equipment related and some major disadvantages are listed.
Process related
- sensitivity to process changes resulting in changes in
gas temperature, gas flow rate, gas analysis, dust load,
dust particle size, and dust analysis
- problems caused by dust build-up
- extensive arcing
- back corona
- corrosion
Design and/or equipment related
- dependency on good electrode alignment
- dependency on adequate power levels
- problems caused by dust build-ups
- problems caused by uneven gas velocity distribution, dust
distribution and temperature gradients
—reentrainment (hopper, rapper, saltation)
- sneakage
- breakage of electrodes
- failures of mechanical equipment, such as rappers, drives,
etc.
- failures of electrical equipment, such as HV transformer-
rectifiers and AV controls, rapper/heater controls
- failures or breakdown of insulators
- air inleakage through hoppers, precipitator shell, doors, etc,
- hopper pluggage
r- inadequate rapping intensity and/or frequency
Each of these problems will eventually manifest itself with a specific
malfunction which can be analyzed and corrected.
A summary of problems associated with electrostatic precipitators was
published by Szabo et al (Ref. 5).
1-14
-------
Many of these disadvantages or problems of the electrostatic precipitator
become apparent at the time of the initial start-up. Others may only
become apparent after weeks, months, or even years of operation. Quite
often, the problem is further complicated by the simultaneous occurrence
of several of these problems.
In general, problems associated with electrostatic precipitators need
to be analyzed, if they are process or design related and treated ac-
cordingly.
8. Inspection and Maintenance Surveys
Attempts have been made to analyze precipitator malfunctions, and, thus,
to predict failures. The principal factors are:
o design of the precipitator
o hardware and controls
o improper erection
o operating conditions
o and operating and maintenance practices
A survey by the Industrial Gas Cleaning Institute (IGCI) published
April 1971 (Ref. 6) concentrated on:
o maintenance
o actual vs. expected maintenance costs
o operating and maintenance instructions
o repair and replacement
o corrosion
o useful life of precipitator
o dust removal systems
The results of the survey showed that 59 percent of the respondents
reported daily routine inspections; weekly inspections were reported
by 9 percent and other (quarterly, yearly) inspections by 27 percent.
1-15
-------
Of the precipitators surveyed, 60 percent reported corrosion within
three years from start-up, 40 percent of these involved un-insulated
precipitators.
The expectancy of useful precipitator life, defined as the period in
which the precipitator is amortized, brought the following response:
o 5.5%
0 11%
0 11%
0 11%
o 33%
0 11%
indicated
indicated
indicated
indicated
indicated
indicated
5
10
12Js
16
20
30
years
years
years
years
years
years
o 16.5% indicated 35-40 years
Only 23 percent reported "known" precipitator maintenance costs. Manu-
facturer's maintenance instructions were considered adequate by 96 per-
cent of the respondents. Repairs and replacement problems in order of
severity were stated for:
o discharge electrode failure (68 percent)
o rapper malfunctions (40 percent)
o insulator failures (28 percent)
o dust build-ups causing shorts (28 percent)
o hopper plugging (24 percent)
o transformer-rectifier failures (20 percent)
It is interesting to note that, where dust removal systems were in 100
percent continuous operation, there were no frequent operational prob-
lems reported.
A similar study was performed by TVA and reported by J. Grecco (Ref. 7)
This study showed outages caused by failures of various components of
the precipitators.
1-16
-------
Figure 3 shows the major causes for unavailability of nine different
precipitators (A to I). The newest survey of four major precipitator
user industries was undertaken by the TC-1 Committee of the Air Pollu-
tion Control Association (APCA) in the latter part of 1974 (Ref. 8).
The study was intended to survey the user's degree of satisfaction
with this equipment from an operational and a maintenance viewpoint.
Component failures were grouped by frequency, such as frequent, infre-
quent, or very seldom. A summary of the results is presented.
COMPONENT FAILURE FREQUENCY (PERCENT)
Frequent Infrequent Very Seldom
o Discharge Electrodes 22.2% 45.3% 28.8%
o Collecting Surfaces 9.9% 15.2% 56.8%
o Rappers or Vibrators 22.2% 35.8% 33.3%
o Support Insulators 8.6% 42.0% 41.6%
o Dust Removal Systems 24.7% 35.8% 30.0%
A summary of the user's opinion on the magnitude of the problem caused
by the component failure is given:
COMPONENT MAGNITUDE OF PROBLEM
CAUSED BY FAILURE (PERCENT)
Major Minor No Problem
o Discharge Electrodes 22.6% 53.1% 21.0%
o Collecting Surfaces 17.3% 32.1% 45.3%
o Rappers or Vibrators 10.3% 53.1% 28.4%
o Support Insulators 9.5% 49.4% 37.9%
o Dust Removal Systems 24.7% 49.0% 17.3%
The causes for these component failures vary from installation to
installation. The following comments are based on the above APCA
survey.
1-17
-------
Discharge electrode failures are typically caused by either fatigue, corrosion,
or electrical arcing with the latter being the predominant cause. Corrosion
and fatigue failures ranked second and third. The major cause of collecting
surface failures is fatigue at the point of suspension; corrosion is the second.
Rapping systems using vibrators, either pneumatic or electric, seem to be a
higher maintenance item than the impulse type rapper. Support insulator fail-
ures are mostly caused by arc-overs due to accumulations of dust or moisture
on the surface of the insulator.
Dust removal has always been a major cause of precipitator malfunction; failures
are normally due to hopper plugging. Screw conveyor and dust valve problems
rank second and third.
9. Plant Inspection and Maintenance Program
Safety
It is obvious that high-voltage electricity can be extremely dangerous. There-
fore, all practicial safety measures must be observed even though the system
may already incorporate interlocks and other safety devices.
The system should never be adjusted with the high-voltage power on. The rapper
circuitry, which is independent of the high-voltage circuitry, is nonetheless
also dangerous and must be treated as such.
Spark-rate feedback signals are often taken from the primary of the high-voltage
supply and can be 400V. a.c. or more. Fuses on these lines should be removed
before maintenance or adjustment is attempted.
1-18
-------
Explosive gas mixtures could be created if air is introduced into some
systems. If necessary, the system should be purged with an inert gas
before introducing air. In all cases, a system should be purged with
fresh air before it is entered.
Insulator heaters, hopper heaters, and rapping motors should be shut-
off prior to entering the unit. Only authorized personnel should be
allowed inside the precipitator. The Key-Interlock System Procedure
has to be followed whenever inspection and/or maintenance requires that
the precipitator access doors be opened.
Inspection Program
An inspection program should be aimed at providing a minimum of inspec-
tion time and produce maximum results in preventive maintenance. But
too often internal inspections are either completely neglected or the
hot and dusty precipitator gets a rather time-consuming complete inspec-
tion. The following is an attempt to shorten the inspection time con-
siderably by providing a list of key points requiring partial inspection.
It should be noted, however, that the partial inspection may indicate
that a complete inspection is necessary.
Inspection of the electrostatic precipitator should be set up on the
following basis by the operator:
Daily (per shift)
Weekly
Quarterly
Annually
1-19
-------
The inspection instructions issued by the manufacturer should be followed.
Routine daily or weekly inspections should cover the following areas:
I. Control Centers
Precipitator Control Panels
Ancillary Control Panels
II. Precipitator
Dust Removal System
Gas Distribution Plate Rapper Drives
Collecting Plate Rapper Drives
Discharge Wire Rapper Drives
Transformer-Rectifiers
Key Interlocks
A more detailed program should be established and followed for quarterly and
annual inspections, including inspections inside of the electrostatic precipitator.
Such a program should include inspection of:
III. Dust Build-Up in Hoppers
IV. Corrosion
- around access doors
- box girders and penthouse areas
- precipitator housing
- sidewalls
- top and bottom corners
V. Rapping systems
- wear
VI. Gas Distribution Plates and Interior Baffles
VII. Discharge electrodes
- establish failure pattern, causes
VIII. Collecting Surfaces
IX. Insulators
1-20
-------
X. Electrical Instrumentation
- a-c voltage
- a-c current
- d-c voltage
- d-c current
- spark rate meters/indicators
XI. Final Check
As a check on the inspection on the work done, always allow
for time to close the precipitator properly and energize it
with the normal supply at the maximum possible input, follow-
ing the manufacturer's instructions.
The voltage on the transformer primary will usually be lower,
but the current is considerably higher than during gas load.
This check is a good indication as to whether or not a short
still exists before the unit is put back in operation. The
readings taken during the air-load test should be logged for
future reference. Data sheets should be prepared so that a
historical record of the precipitator's overall performance
can be established.
Inspection and Maintenance Responsibility
Precipitator inspection is performed in a scheduled program, and addi-
tionally, whenever the precipitator becomes available for inspection
during shut-down periods.
Maintenance is always performed as a continuous task covering components
accessible during normal operation (rappers, control panels, drives, etc.)
1-21
-------
and during scheduled or unscheduled shut-down periods equipment not
normally available for maintenance work.
The objective of inspection is to point out areas or components requiring
maintenance work, to repair existing faults or damages, or to point to
areas or components where such work may be required in the near future and
could be repaired now as preventive maintenance work rather than requiring
an additional shut-down of the equipment.
A preventive maintenance schedule should be established for each installa-
tion, detailing the precipitator parts to be checked and maintained daily,
weekly, monthly, quarterly, semi-annually, annually, and on a situational
basis.
All of this work requires skilled personnel to recognize deficiencies
during an inspection and to direct the necessary maintenance and/or
repair work.
10. Normal Precipitator Operation
Each electrostatic precipitator installation is different not only in de-
sign, but also in its application. Therefore, specific check lists for
start-up and shut-down of the equipment cannot be established. The follow-
ing Tables 3, 4, and 5 are only given as examples for steps to be taken
during these operations (Ref. 9).
11. Precipitator Inspection and Evaluation
Precipitator inspection and evaluation by Control Agency Personnel starts
normally with an unacceptable stack discharge characterized by a high
1-22
-------
opacity level. This condition necessitates a good look at the pre-
cipitator; i.e., both process and hardware. The first, and most of
the time best indication of the present state of operation of the
precipitator, is to look at the electrical controls and to try to
interpret the meter readings in the control cabinets.
The following general guides when reading these meters should help:
1. When the gas temperature increases, the voltage will increase,
and the current will decrease. Arcing can develop. When the
gas temperature decreases, the voltage will decrease, and the
current will increase.
2. When the moisture content of the gases increases for any given
condition, the current and voltage will also tend to increase
in value.
3. If reduced voltage exists because of a sparkover, a rise in
moisture may allow for an increase in the precipitator voltage
level.
4. An increase in the concentration of the particulate will tend
to elevate voltages and reduce current flows.
5. A decrease in the particle size will tend to raise voltage
while suppressing current flow.
6. A higher gas velocity through the precipitator will tend to
raise voltages and depress currents.
7. Air inleakage may cause sparkover in localized areas resulting
in reduced voltages.
8. A number of precipitator fields in series will show varying
readings with the voltage-current ratio decreasing in the
direction of gas flow.
1-23
-------
9. If a hopper fills with dust causing a short, the voltage will
be drastically reduced, and the current will increase.
10. If a discharge electrode breaks, violent arcing can be observed
with the meters swinging between zero and normal.
11. If a transformer-rectifier unit shorts, voltage will be zero
at a high current reading.
12. If a discharge system rapper fails, the discharge wires build
i
up with dust; the voltage increases to maintain the same
current level.
13. If a plate rapper fails, the voltage decreases to maintain a
current level under sparking conditions.
In addition, electrical readings taken at prior plant visits or by plant
personnel can be used to evaluate an eventual change in the operation of
the electrostatic precipitator. Cunningham (Ref. 11) published voltage-
current characteristics (Fig. 4) and information on changes in voltage
and current readings at a specific precipitator installation (Fig. 5).
Routine surveillance of the electrostatic precipitator, its connecting
ductwork and ancillary equipment, can provide additional input to the
general performance level and elimination of problem areas. For exam-
ple j hoppers of dry collectors must be emptied continuously or peri-
odically to prevent overfilling. Dust levels in a hopper may be deter-
mined by sounding, visual inspection (capped ports), or high dust level
indicators arranged to alert operating personnel visually or audibly
(Ref. 10).
At this point, all mechanical equipment inspection reports should be
reviewed for indications of pending problems not taken care of during
1-24
-------
past outages.
Changes in process conditions need to be evaluated to observe any
influences they could have on the precipitator performance.
Troubleshooting charts are normally provided by the precipitator
manufacturer. It is best to use these charts since they are
tailored to the specific precipitator design and application.
A general troubleshooting chart was published by Szabo et al
(Ref. 5).
12. Improving Precipitator Operation
In a general sense, any maintenance work done on an electrostatic
precipitator should improve its operation or at least maintain its
performance at a prior established level. But maintenance work can
also be classified as a means of incorporating improvements which
will either upgrade the performance, increase the lifetime, or re-
duce the operating cost of an existing precipitator.
Improvements in precipitator operation can possibly be achieved by:
- changes in the process served by the precipitator
- adding of equipment
- fine tuning
The results of changes in each of the three categories cannot be pre-
dicted entirely, but definite estimates can be given based on prior
performance.
1-25
-------
Changes in the Process
Changes in the process served by the precipitator could result in
more favorable conditions of gas volume, temperature, moisture,
dust load, or particle size. Because of the wide variety of possi-
ble precipitator applications, no qualification is attempted.
Adding of Equipment
Adding of new equipment and/or replacing of obsolete equipment can
very definitely improve the performance of the electrostatic precip-
itator. The obviously most effective addition would be increase the
installed collecting surface area by either adding an additional
precipitator section in parallel or in series. In series will
generally increase the collecting efficiency whereas in parallel
allows to handle a larger gas volume at the same or similar collect-
ing efficiency level.
In specific instances, components can be added which will provide
a more efficient and trouble-free operation of the precipitator.
These are, for example:
- Transformer-rectifiers and controls
- Rapping systems
- Rapper controls
- Gas distribution devices, such as turning vanes, perforated
plates, etc.
- Flow control devices, such as dividers, baffles, etc.
- Insulator vent systems
- Insulator heating systems
- Hopper heating systems
- Dust discharge valves
1-26
-------
Fine Tuning
Fine tuning of an electrostatic precipitator used to be done either by
visual observation of the stack or by optimizing the instrument readings
in the control cabinet. Neither method is too accurate, and in addition,
they are both extremely time consuming and, thus, impractical for actual
use. In addition, when the whole precipitator consisted of maybe two
or three fields in series and one or two sections in parallel (for a
total of two to six bus sections), a change in the electrical settings
or the rapping sequence was easily detectable. With today's precipi-
tators having sometimes six to eight fields in series and four to eight
sections in parallel (for a total of 24 to 64 bus sections), the effect
of changing the operation of a single bus section is hardly noticeable.
This leaves the operator only with the possibility to use real time
recording instrumentation such as transmissometers and oscilloscopes
with multi-channel recorders.
A transmissometer measures light absorption caused by the dust in the
light path with the absorption being a measure of its quantity. The
most prominent feature of this device is its capability to produce
an instantaneous recording of the dust flow; i.e., immediately recog-
nizing changes in operating conditions causing changes in precipitator
performance, rapping spikes, electrical shorts, etc. (Fig. 6).
In addition, a relationship between optical density and dust concentra-
tion can be established (Fig. 7, Ref. 11), which allows for instantaneous
evaluation of changes in precipitator performance caused by fine tuning
1-27
-------
of the precipitator; i.e., changing variables such as:
- rapping frequency
- rapping intensity
- electrical power input
- spark rates
- flue gas conditioning
- gas velocity
- gas and dust distribution
Each of these and other parameters will affect the precipitator per-
formance in a different way. The use of the transmissometer allows
to optimize these variables and to achieve the highest possible level
of collecting efficiency.
An oscilloscope is another valuable tool to fine-tune an electrostatic
precipitator by recording the current and voltage levels immediately
before and after sparking (Fig. 8).
The oscilloscope will reveal the proper function of the automatic
voltage control system and provide a real-time indication of the
voltage/current fed into an electrical bus section of the precipitator.
It is obvious that, in addition to the above recommended indicators
for the performance of the precipitator, the process related data;
for example, fuel rate, gas temperature, etc., need also to be scruti-
nized when evaluating the performance of the electrostatic precipitator.
1-28
-------
13. Conclusion
In conclusion, it can be said that the operation and maintenance of
the electrostatic precipitator has to be an item of constant atten-
tion to the plant operator, and that the burden of keeping precipi-
tator performance at a high level of collecting efficiency can be
eased by careful planning, inspecting, maintaining, and record
keeping.
This paper covers some of the problems that may face the maintenance
department... Whether these problems actually arise are subject to
many variables, including the thought and care given to reliability
features in the original design... Success or failure of any given
installation will often rest on process effects, frequent compre-
hensive internal inspections, and correction of repetitive maintenance
problems. This care can sometimes result in acceptable performance
from a marginal collector (Ref. 12).
1-29
-------
14. References
1. White, H. J., "Industrial Electrostatic Precipitation"
Addison-Wesley Publication Company
Reading, Massachusetts, 1963
2. Oglesby, Sabert, Jr., "A Manual of Electrostatic Precipitator
Technology"
Southern Research Institute
Birmingham, Alabama, 1970
3. Engelbrecht, H. L., "Hot or Cold Electrostatic Precipitators
for Fly Ash from Coal-fired Boilers"
APCA Western Pennsylvania Technical Meeting on Coal Utilization,
Pittsburgh, Pennsylvania, April 1976
4. White, H. J., "Electrostatic Precipitation of Fly Ash"
Journal of the Air Pollution Control Association
27(3) P. 206-217, March 1977
5. Szabo, M. F. et al., "Electrostatic Precipitator Malfunctions
in the Electric Utility Industry", PEDCO
Environmental, Inc.
Cincinnati, Ohio , 1977
6. Industrial Gas Cleaning Institute (IGCI), "Survey of Electrostatic
Precipitator Operating and Maintenance Costs", April 1971
7. Greece, J., "Electrostatic Precipitators - An Operator's View"
Specialty Conference: Design, Operation, and Maintenance of
High Efficiency Particulate Control Equipment, St. Louis,
Missouri, March 1973
8. APCA TC-1 Particulate Committee, "Electrostatic Precipitator
Maintenance Survey", Journal of the Air Pollution Control
Association, Pittsburgh, Pennsylvania, November 1976
9. Engelbrecht, H. L., "Plant Engineer's Guide to Electrostatic
Precipitator Inspection and Maintenance"
Plant Engineering, April 1976
10. Smith, E. M., "Preventive Maintenance Helps Prevent Pollution",
Pollution Engineering, March/April 1971
11. Cunningham, R. L., "Operational Monitoring and Maintenance of
Industrial Electrostatic Precipitators for Optimum Performance"
I.A.S. Annual Meeting, 1976
12. Katz, J., "Maintenance Program and Procedures to Optimize Elec-
trostatic Precipitators", IEEE Cement Industry Technical Con-
ference, Mexico City, Mexico, May 1974
13. Frenkel, D. I., "Tuning Electrostatic Precipitators",
Chemical Engineering, June 19, 1978
1-30
-------
15. Literature
The following is a brief sample of books, papers, and articles pub-
lished during the last few years on subject of precipitator main-
tenance. Numerous other publications are available for the worker
in this field. But most important of all, the information manuals,
etc., of the equipment manufacturer should be consulted, and train-
ing sessions should be requested from the manufacturer.
Cross, F. L. and Hesketh, H. E., "Handbook for the
Operation and Maintenance of Air Pollution Control
Equipment"
Technoihic Publication
Technomic Publishing Co., Inc.
Westport, Connecticut, 1975
Kester, Bruce E. (Editor)
"Design, Operation, and Maintenance of High Efficiency
Particulate Control Equipment"
Specialty Conference sponsored by the Greater St. Louis
Section and the Technical Council of the Air Pollution
Control Association, St. Louis, Missouri, March 1973
Industrial Gas Cleaning Institute, Inc.
•Publication No. E-P 1
"Terminology for Electrostatic Precipitators"
Approved February 1964, revised October 1967
and January 1973
Katz, J., "Maintenance Program and Procedures to
Optimize Electrostatic Precipitators"
IEEE Cement Industry Technical Conference
Mexico City, Mexico, May 1974
Published IEEE Transactions on Industry Applications
Vol. 1A-11 November/December 1975
Szabo, MF. et al., "Control of Fine Particulate from Coal-
Fired Utility Boilers" Paper No. 77-14.1,
70th Annual Meeting of the Air Pollution Control Association
(APCA), Toronto, Ontario, June 1977
Crynack, R. R., "A Review of the Electrical Energization
Equipment for Electrostatic Precipitators"
71st Annual Meeting of the Air Pollution Control Association
(APCA), Houston, Texas, June 1978
Frenkel, David I., "Tuning Electrostatic Precipitators",
Chemical Engineering, Pages 105-110, June 19, 1978
1-31
-------
Literature
Steele, C. Jay, "Corrosion Protection Strategy for Pollution
Control Equipment", Pollution Engineering, Pages 49-50,
March 1978
Schneider, G. G. et al, "Selecting and Specifying Electrostatic
Precipitators", Chemical Engineering, Pages 94-108, May 26, 1975
Lane, W. R. and Fletcher, H. R., "Introduction to Electrostatic
Precipitator Energization and Control"
71st Annual Meeting of the Air Pollution Control Association
(APCA) Houston, Texas, June 1978
Proceedings "Operation and Maintenance of Electrostatic Precipi-
tators" Specialty Conference sponsored by the Michigan Chapter
East Central Section Air Pollution Control Association
Detroit, Michigan, April 1978
1-32
-------
#
u
lu
u
u.
Uj
K
O
O
u
or
o
UJ
of
(L
99.5
99.0
200
300
COLLECTING
400
FIG. 1; ELECTROSTATIC PRECIPITATOR COLLECTING EFFICIENCY AS A FUNCTION
OF PRECIPITATOR SPECIFIC COLLECTING AREA (SCA).
Reference:
Engelbrecht, H. L., "Hot or Cold Electrostatic Precipitators for Fly Ash
from Coal-Fired Boilers"
APCA-Western Pennsylvania Technical Meeting on Coal Utilization
Pittsburgh, Pennsylvania, April 1976 (Reference 3).
1-33
-------
PROCESS:
DRY
WET
COLLECTING
SURFACE:
PLATE
PIPE
GAS FLOW: HORIZONTAL \ HORIZONTAL
VERTICAL VERTICAL VERTICAL
FIG. 2: ELECTROSTATIC PRECIPITATOR CLASSIFICATION LIMITED TO
CONVENTIONAL (COTTRELL-TYPE SINGLE STAGE) DESIGNS
1-34
-------
TABLE 1: COMPARISON BETWEEN AMERICAN AND EUROPEAN PRECIPITATOR DESIGN
COMPONENT
AMERICAN DESIGN
EUROPEAN DESIGN
1. Discharge System
2. Collecting Surfaces
Straight wires or discharge rods stretching from an
upper support frame to a lower guide frame, wires
held by weights, wire electrodes larger than the
height of the collecting surfaces. Upper frame
supported by two insulators.
Reinforced panels 6-9 ft. long in direction of gas
flow, 15 to 36 ft. high.
3. Rapping Systems
Magnetic or pneumatic rappers or vibrators strike
the upper support frames of the discharge system.
Magnetic or pneumatic rappers or vibrators strike
the supporting elements of the collecting surface
plates. Rapping impact always from the top.
CO
ui
Discharge wires of short length mount-
ed in wire frames, each frame support-
ed at both ends by structural ele-
ments, in turn, supported by four
insulators.
Collecting surfaces consist of roll-
formed panels approximately 18 in.
wide, 15 to 50 ft. long, supported
by a structural member from the top,
interlocked with each other, and tied
together at the bottom by a rapper
bar.
Each discharge frame rapped by a
single hammer, all hammers mounted on
a common shaft. Each row of collect-
ing surface panels rapped by a single
hammer, all hammers mounted on a
common shaft. All drives located on
the outside of the precipitator. Rap-
ping impact at or below the center
at the trailing edges of the discharge
frames and at the bottom trailing
edges of the collecting surfaces.
-------
TABLE 2
DESIGN FACTORS WHICH SHOULD BE INCLUDED IN PRECIPITATOR
DESIGN SPECIFICATIONS AND EVALUATIONS (REF. 4)
1. Corona electrodes: type and method of supporting.
2. Collecting electrodes: type, size, mounting, mechanical, and
aerodynamic properties.
3. Rectifier sets: ratings, automatic control system, number,
instrumentation, and monitoring provisions.
4. Rappers for corona and collecting electrodes: type, size, range
of frequency and intensity settings, number, and arrangement.
5. Hoppers: geometry, size, storage capacity for collected dust,
number, and location.
6. Hopper dust removal system: type, capacity, protection against
air inleakage and dust blow-back.
7. Heat insulation of shell and hoppers, and precipitator roof pro-
tection against weather.
8. Access doors to precipitator for ease of internal inspection and
repair.
9. Provisions for obtaining uniform, low-turbulence gas flow through
precipitator. This will usually require a high-quality gas flow
model study made by experienced people in accord with generally
accepted techniques, with full report to precipitator purchaser
before field construction.
10. Quality of field construction of precipitator, including adherence
to electrode spacing and rigidity requirements.
11. Warranties: performance guarantees, payment schedules, adequate
time allowance for performance tests, penalties for non-performance.
12. Support insulators for high-tension frames: type, number, reliability.
Air venting, if required.
13. Inlet and outlet gas duct arrangements.
14. Structure and foundation requirements.
1-36
-------
E § INSULATORS 1 IfLYASHB jl
23
22
21
20/
7
UJ
J
(0 6
_J
z 5
D
Z
u
U 4
(T
u
a.
3
2
1
_ DUD CONTROLS BID
-
^_
/
lyv
-
.
_
~ Z
™ —
z c
~ c
\ \
\ \
\ V
v \
3 E
: :
TT
j ~
: :
: =
: =
™ —
• —
\ 1
; =
— — 1
: ~
i ••
INTERNAL SHORTS C ElUNKNOV^N
SLUICE SYSTEM I — BWI
E =
« Z
/
/
^ES
/
f
\
\
\
\
\
\
"
\
\
\
\
\
\
\
v
I . •
; : •
PRECIPITATOR
ABC
0
PRECIPITATOR
E
F G H I
UNAVAILABILITY
FIG. 3:
PRECIPITATOR UNAVAILABILITY
Greece, J., "Electrostatic Precipitators - An Operator's View"
Specialty Conference: Design, Operation, and Maintenance of
High Efficiency Particulate Control Equipment
St. Louis, Missouri, March 1973
1-37
-------
TABLE 3: PRECIPITATOR START-UP CHECKLIST (REF. 9)
PRECIPITATOR START-UP CHECKLIST
1. Check line voltage for proper phase and magnitude.
2. Inspect transformer-rectifier tanks for signs of oil leaks or physical
damage. Check oil tank gauge. Refill if necessary (follow manufacturer's
instructions).
3. Check hopper discharge valves and dust handling equipment.
4. Inspect exhaut fan.
5. Follow key-interlock procedures for opening precipitator access doors.
Inspect interior of precipitator. Remove any foreign materials (tools,
rags, cleaning materials, etc.) from inside the unit.
6. Disconnect high-voltage conductor at support insulator and check resistance
between discharge system and ground. Reading should be 100 megohms or
greater.
7. Inspect all rapper drivers for proper position (follow manufacturer's in-
structions) .
8. Check rotation and alignment of all gear motors and drives that have been
serviced.
9. Inspect access doors for operation and alignment. Lock them. Return door
keys to their proper location in key-interlock transfer blocks.
10. Check condition of all explosion-relief devices (if applicable).
11. Inspect precipitator control cabinets for evidence of loose connections.
12. Complete procedure outlined in the key-interlock instructions to return
all keys to operating position.
13. Preheat support insulators at least 2 hours before energizing the precipi-
tator. Start insulator vent system (if applicable).
14. Activate dust-discharge and dust-handling systems.
15. Start collecting surface and discharge-electrode rapping systems. Opera-
tion should be continuous during start-up.
16. Activate gas distribution plate rapping system (if applicable).
17. Turn on high-voltage current as soon as gas flow has been started to the pre-
cipitator (by activating exhaust fan, dampers, or slide gates) and the tem-
perature of the precipitator*s internal parts exceeds dewpoint of the gas.
(High voltage, however, should not be activated if there is a possibility
of combustible gases being present in the precipitator).
18. Set precipitator operating control on "automatic".
19. Turn off insulator heaters or set on "automatic".
20. Turn off continuous rapping and set on "automatic".
1-38
-------
PRECIPITATOR SHORT-TIME SHUT-DOWN CHECKLIST
1. Turn on insulator heating system and de-energize high-voltage system.
2. Keep rapping systems for collecting-surface plates, discharge wires,
and gas-distribution plates activated unless precipitator is to be
entered. (If entry is planned, follow key-interlock system procedures
to open precipitator access doors.)
3. Keep dust discharge system operating continuously.
A. Operate exhaust fan at reduced flow rate.
TABLE A: PRECIPITATOR SHORT-TIME SHUT-DOWN CHECKLIST (REF. 9)
PRECIPITATOR SHUT-DOWN CHECKLIST
1. Activate insulator heating system. Leave on for at least 6 to 8
hours after precipitator is de-energized (do not turn on heating
system if maintenance or inspection work is to be done in insulator
compartments).
2. Keep rapping systems for collecting surface plates, discharge wires,
and gas-distribution plates on for several hours to help clean pre-
cipitator as thoroughly as possible.
3. Turn off exhaust .fan*
A. Follow key-interlock system procedures to open precipitator access doors.
5. Ground all high-voltage components securely. Warning signs should be used
on all switches (follow manufacturer's instructions).
6. Clean precipitator manually.
7. Discharge all dust from hopper if downtime is to be extensive.
8. Seal flues to other precipitators, stack, or any crossover flues to
prevent gases from other sources from entering precipitator (such
gases can condense and cause extensive corrosion).
TABLE 5: PRECIPITATOR SHUT-DOWN CHECKLIST (REF. 9)
1-39
-------
t
400 -
F)
AIR LOAD
(70°F)
I i i
10 15 20 25 30 35 40 45 50 55
VOLTAGE
FIG. 4: VOLTAGE-CURRENT CHARACTERISTICS (Inlet Section)
Reference:
Cunningham, R. L., "Operational Monitoring and Maintenance of Industrial
Electrostatic Precipitators for Optimum Performance", IAS Annual Meeting,
1976 (Reference 11)
1-40
-------
^40 4
so H
I 20 H
500 ^
300-1
200
DAT£
FIG. 5: CHANGES IN VOLTAGE AND CURRENT READINGS OVER TIME
Reference:
Cunningham, R. L., "Operational Monitoring and Maintenance of
Industrial Electrostatic Precipitators for Optimum Performance",
IAS Annual Meeting, 1976 (Reference 11)
1-41
-------
X
K
(J
C
Q.
O
0
I
5
C
7
8
9
10
a
1} ill i !;Tf rj
I/ll MuilyIJJ
TINf
a - Zero Emission Scan
b - Emissions Spikes
FIG. 6: OPTICAL DENSITY PRINT-OUT
Reference:
Frenkel, D. I., "Tuning Electrostatic Precipitators",
Chemical Engineering, June 19, 1978 (Reference 13)
1-42
-------
500
cn
2
o
z
U)
u
8
h-
V)
3
Q
200
100
0.05 0.10 0-15 0.20 O.K
OPTICAL VENSITY
FIG. 7: RELATIONSHIP BETWEEN OPTICAL DENSITY AND DUST CONCENTRATION
FOR A LIGNITE-FIRED BOILER
Reference:
Frenkel, D. I., "Tuning Electrostatic Precipitators", Chemical
Engineering, June 19, 1978 (Reference 13)
1-43
-------
TIME
A - Varying Sparkover Level
B - Envelope of Precipitator Current Pulses
C - Spark Quench
D - Fast Ramp
E - Setback
F - Slow Ramp
G - Rapid Turn-Off
H - Arc Quench
I - Soft Start
FIG. 8: ELECTRICAL POWER LEVELS OF AN ELECTROSTATIC PRECIPITATOR
IN AN ARCING MODE OF OPERATION
1-44
-------
eering
V-2.
Reprinted with permission. Copyright c by
Chemical Engineering. McGraw-Hill Publishing
Co. May 26,1975 issue.
Selecting and
Specifying
Electrostatic
Precipitators
Industrial electrostatic precipitators are
complex devices. There are many added-cost
features that will pay off in better operation
and lower maintenance, but are likely to be
omitted in the low-bid specification. Also,
proper erection and inspection procedures
are vital if you expect to receive trouble-
free service and high efficiencies.
GILBERT G. SCHNEIDER. THEODORE I. HORZELLA, JACK COOPER and PHILIP J. STRIEGL, Enviro Energy Corp.
There are many details that.you must be aware of if
you want to select and specify precipitators intelligently.
For example, inclusion of many specialized design-
features will enable the precipitator to be erected easily
and to be operated and maintained with the fewest prob-
lems. But since most pfecipitators are bought on a bid
basis, these features are likely to be omitted (to provide
for the lowest cost), unless you have specified that they
must be included.
Careful attention to detail during the erection of the
precipitator will pay dividends during startup, and in
later operation. Here, again, you must know which prob-
lems must be avoided.
One of the things that complicates the purchase of
electrostatic precipitators is that there is much "art" in-
volved in the selection of the equipment by the vendor.
This selection relies more on experience with previously
sold precipitators than on solid engineering data and cal-
culations. Depending on the supplier's experience bank,
it is perfectly reasonable for an.engineer lo receive ven-
dors' bids that, for the same gas flow, vary in size by fac-
tors of two or more.
In the design of this type of equipment, it should be
noted that size increases directly with gas volume for a
constant efficiency but increases exponentially as effi-
ciency requirements rise. That is, costs increase exponen-
tially with increase of efficiency.
Before we go into details of the precipitators, let us
discuss the paniculate-containing gases that will be going
through them.
Particle Size of the Pollutant
Particulate air-pollution-control problems involve par-
ticles under 100 microns (jan) in size. Although the par-
ticles are not spherical, the particle size is expressed as
MAY 26,1975/CHEMICAL ENGINEERING
2-1
-------
C j
Transformcr-rfjctilier s(it
Vibrator to shake loose
particles that may collect
on discharge electrodes
Magnetic-impulse rapper
to agitate collecting
electrodes and knock
collected particles loose
Bushing through which
high-voltage connection •_
is made to discharge
electrode support frame
Perforated plate to help
smooth out inlet flow —
and distribute it evenly
across entire cross-section
High-tension support
frames
Collecting electrode plates
Weights at bottom of
discharge electrode wires
to keep them plumb
COMMERCIAL electrostatic
precipitator, showing
major features—Fig. 1
the diameter of an equivalent sphere that would follow
the settling rate of Stokes' law.
Particles are often described, depending on their size or
nature, as:
• Dust—Particles from 0.1 to 100 microns (urn).
• Mist—Liquid droplets suspended in a gas.
• Fume—Solid particles or liquid droplets that are
formed by condensation from a vapor.
In most industrial applications, particle size interests
us only insofar as it affects the capability of the air-pollu-
tion-control equipment. With the electrostatic precipi-
tator, as with other industrial air-pollution-control de-
vices, the larger particles are easier to collect (except
when the particle is large but extremely fluffy).
There is a basic difference between the electrostatic
precipitation principle and the mechanical methods
(used in centrifugal separation, wet scrubbing and gas fil-
tration). In the precipitator, the electrical forces are ap-
plied only to the suspended particles. In the mechanical
methods, the complete gas stream is subject to externally
applied forces, resulting in a much higher consumption
of energy for the collection process. The size of the pre-
cipitator will be affected by the particle size, but the
energy consumption for the collection process will re-
main almost constant.
Properties of the Particles
Certain physical and chemical properties of the par-
ticles are important because they affect the properties of
^he agglomerates that result when the particles reach a
collecting surface.
For example: Mists will form liquid droplets that flow
by gravity into the hopper. Some metallurgical fumes,
such as zinc and lead oxides, form low-bulk-density lay-
ers that will break (during rapping of the collection
CHEMICAL ENGINEERING/MAY 26,1975
2-2
-------
ELECTROSTATIC PRECIPITATORS . . .
VARIATION in size (SCA) supplied for any specified efficiency (coal-fired boilers)—Fig. 2
plates) into fluffy agglomerates, which float in the gas
stream. Other fine dusts, such as cement, form a rela-
tively dense agglomerate that quickly falls into the hop-
per during rapping. Finally there are dusts, such as those
produced in no-contact (low odor) kraft recover)- boilers,
which are very tacky and difficult to remove by rapping;
special attention must be given to a proper rapper system
to minimize operational problems.
The electrical conductivity of the particle is most im-
portant. (However, in the field of electrostatic precipi-
tation, the reciprocal property, "resistivity," is used for
the sake of numerical convenience.) Nonconductive or
high-resistive particles, as they deposit on the surface of
the collecting electrode, form an electrical insulating
layer that prevents the movement of ions to ground.
Thus the flow of current from the discharge electrode to
the collecting electrode is reduced, and the voltage dif-
ferential is increased until sparking occurs. A resistivity
of the particles above 2 X 1011 ohm-cm is generally con-
sidered the limit for proper electrostatic precipitation.*
Excessive conductivity (or low resistivity) of particles,
such as carbon black and other carbonaceous materials,
causes the particles to immediately lose their charge as
they contact the collecting electrode. Thus the particles
are not retained on the collecting electrode surface, but
rather reenter the gas stream.
Conditioning To Modify Particle Resistivity
As White points out in "Industrial Electrostatic Precip-
itation," "adaptation of conditioning methods to prac-
tical situations requires a broad knowledge of the basic
principles and contingent factors to obtain useful re-
sults."
Conditioning to modify particle resistivity may consist
of:
• Addition of chemicals, or water vapor, to the gas
stream.
• Modification of the material producing the dust.
• Change of the gas temperature.
More often than not, a combination of these methods
will be chosen.
Addition of chemicals (or water vapor), even in very
small amounts, has shown remarkable effect on the re-
sistivity of the particles. Ammonia, sulfuric acid, sodium
chloride, sulfur trioxide, and other substances, added in
trace amounts, have improved the operation of some
precipitators from an unacceptable performance to one
that meets air-pollution-control requirements.
In most cases, the conditioning chemical is adsorbed
on the surface of the particles. However, several cases of
conditioning of weak basic particles by using both strong
acids and water vapor suggests that the acid is adsorbed
on the particle surface, and that water vapor is then ab-
sorbed. A similar situation occurs when conditioning
weak-acid particles with a strong base and water vapor.
Gas Flow
Gas flow is critical in the design and operation of an
electrostatic precipitator. The basic principle on which
the precipitator works—the migration of minute particles
to the collecting electrode—involves a finite length of
time. If the gas velocity in any of the passages around the
collecting electrodes exceeds the design gas velocity,
some particles will not have adequate time to reach the
collecting electrode.
After the particles have been deposited on the col-
lecting electrodes, they are made to fall into the hoppers
by rapping the electrodes. During this operation, good
gas-flow patterns are critical to avoid reentrainment of
particles in the gas stream. This reentrainment is possibly
the most devastating effect caused by poor gas-flow.
At times, excessive gas velocity—due to unbalanced gas
flow—may cause reentrainment from the hoppers. Un-
even gas flow may result from part of the gas moving
•The resistivity ot various dust particles, under operating variables such as
gas temperature, gas moisture, and chemical composition ol the gas. is
MAY 26,1975/CHEMICAL ENGINEERING
2-3
-------
Panicle size
ENERGY consumption of dust collectors—Fig: 3
through the hopper space rather than between the col-
lecting plates. This gas "sneakby" through the hoppers
will result in very poor performance. With today's re-
quirements of efficiencies above 99%, a 1% sneakby
would make it impossible to meet the required effi-
ciency. Baffles in the hoppers, when properly designed
and installed, will usually correct such eas^flow. imbal-
ance. ;
Modification of the materials that produce the dust has
been successfully applied to coal-fired boilers. Coal that
contained natural conditioning agents (such as sulfates)
was mixed with low-sulfur coal of high resistivity. The
resulting mixture produced a dust that could be effec-
tively collected in an electrostatic precipitator.
Changing the temperature of the particulate-contain-
ing gas is a widely accepted practice. In general, tem-
perature in the 200 to 400° F of range will produce high-
resistivity dust when natural conditioning agents are ab-
sent. (This is typical of low-sulfur-coal-fired boilers.)
One answer is to locate the precipitator ahead of the
heat exchanger that heats incoming air (and, con-
sequently cools the hot gas). The gas-flow upstream of
the air heater (at about 700° F) represents about 1.5
times the volume oi gas How aownsxream of the air
heater (at about 30CTF). IT would appeal tnat a larger
precipitator would be required to handle the gases at
higher temperature (higher volume). However, in actual
practice (with low-sulfur coals), because of the high-re-
sistivity problems in the 200 to 400°F range, it is the
smaller gas volume that requires the larger precipitator.
It should be kept in mind, though, that the "hot" pre-
cipitator (upstream of the air heater) may not necessarily
be the universal answer for low-sulfur coal. Some lost
sulfur coals contain other impurities that may have a
conditioning etlect. thus making tne "cold" precipitator
(downstream of the air heater) the more economical se-
lectioru.
This should warn prospective users of precipitators
that dusts from untested sources should be tried out in a
pilot precipitator or, if feasible, in a full-scale existing in-
stallation. Such test work will support a final decision on
a "hot" or "cold" precipitator.
The best precipitator will function no better than the
gas-flow distribution allows. The flue design, both up-
stream and downstream of the precipitator, has a direct
effect on gas distribution. Transitions and turning vnnes
must be designed in accord with principles developed in
wind-tunnel tests. Diffusion plates or screens at the ore-
cipitator inlet will convert a flow of creat turbulence into
a multitude of small turbulences immediately down-
stream of each opening of the nlate. These small turbu-
lences will JDromptlv fade out, causing a ngnturbulent
flow through the precipitator. Diffusion devices are usu-
ally perforated plates that are provided with rappers to
dislodge any dust buildup.
Selection of the Precipitator
Precipitators are selected using the Deutsch equation
or some of its modifications, and applying numerous ex-
perience factors.
Where:
T) = collection efficiency, %; w = migration velocity,
cm/s; A = area of the collecting electrodes, ft2; V = gas
flow, thousand ftVmin; e = base of natural logarithms.
The units in the above equation are used by U.S. man-
ufacturers of precipitators to express the migration veloc-
ity in whole numbers. The above equation is used to cal-
culate the collecting area (A) when the gas flow (V) and
the collection efficiency required (•/]) are known. The
equipment designer will select a migration velocity (w)
from his experience file. This migration velocity will
vary, depending on operating conditions, source of the
dust, and temperature and chemical composition of the
gases. The ranges of such variations are too large for us
to present meaningful data for the chemical engineer.
The migration velocity (w) is the average rate at which
particles are charged, and conveyed to the collecting
electrode (where they lose their electrical charge and are
removed into the hoppers). Those panicles that are not
collected, or which are reentrained into the gas stream,
are factored-in in this migration velocity. At the present
state of the art, these factors are determined by experi-
ence.
"Experience" factors, such as particle size, particle ag-
glomeration, gas velocity, gas temperature, gas composi-
tion, and chemical and physical properties of the par-
ticles, are all grouped into this so-called migration
velocity.
Also included as experience factors are design peculi-
arities typical to the specific application, such as configu-
ration of the ductwork, use of distribution plates, the
type of process producing the particulates, and many
other subjective-judgment elements the designer wishes
to incorporate.
It is easy to understand that today's precipitator design
is strongly tinted by the experience, and subjective deci-
sions, of the individual selecting the equipment.
Attempts to quantify the experience factors and to in-
corporate them in computer programs are underway by
numerous engineers and scientists. The precipitator in-
CHEMICAL ENGINEERING/MAY 26,1975
2-4
-------
ELECTROSTATIC PRECIPfTATORS
USER'S cost of erected precipitators
of various design efficiencies.
Variation in particle resistivity
makes for a much wider cost range
for cold precipitators—Fig. 4
A 99 " 99.2 99.4 99.6 99.8 vVcA 99 99.2 99.4 99.6
HT^JJ" \5"v-V?\ "> • Design efficiency^^'.^V---;'',
99.8
dustry, however, has been very slow in employing the
computer in this important task; hence, the selection of
equipment is made today in basically the same manner
as in the early 1940s. Most of the computer studies have
come from outside the electrostatic-precipitator industry.
The greatest difficulty in applying the resources of elec-
tronic data processing to the selection of precipitators is
the lack of complete field-test results covering all of the
variables that affect the basic concept of migration veloc-
ity. The designers of electrostatic precipitators use the
following parameters as guidelines in equipment selec-
tion:
Dry Precipitators—Face velocity (velocity of the
gas across the precipitator) is kept in the range from 1 to
15 ft/s, depending on the dust source.
Aspect ratio (collecting-plate total depth divided by
collecting-plate height) is kept above about 1.0.
The electrical system is arranged in bus sections, each
bus section representing any portion of the precipitator
that can be independently energized.
The number of fields (number of bus sections arranged
in the direction of gas flow) are calculated thusly: As a
rule of thumb, manufacturers will use 1 field for up to
90% collection efficiency, 2 fields for up to 97%, 3 fields
for up to. 99% and 4 or more fields for efficiencies above
99%.
The number of cells (number of bus sections arranged
in parallel) are established so that if any field is shorted
out, the overall precipitator efficiency will not fall below
the specifications.
Often, more than one cell will be energized from a
common high-voltage electrical set (or source). However,
more than one field in any one cell should not be ener-
gized from the same high-voltage electrical set, since a
short would affect more than one field in the same cell,
causing a substantial reduction in collection efficiency. In
general, one high-voltage electrical set is used for up to
25,000 ft2 of collecting surface. About 55mA are supplied
per 1,000 ft2 of collecting surface.
Wet Precipitator—The wet precipitator is ideally suited
for removing acid mists or other materials that can be col-
lected as a liquid solution or suspension. The wet precipi-
tator is also useful for either high- or very-low-resistivity
dusts, provided that the process is not affected by such
wet collection of the dust. There are cases where the dust
can be removed from the collecting surface only by wet
washing of the surface, and here the wet precipitator is
the only answer—unless alternate methods of collection
(such as fabric filters) are used.
The wet precipitator may be designed similarly to a
dry one, but furnished with water sprays for contin-
uously, or periodically, washing the collecting surfaces
and the discharge electrodes. More often, though, the
wet precipitator is designed as a series of pipes in paral-
lel, each containing a single discharge-electrode in its
center. Gases flow vertically through each pipe, and the
collected mist flows down the pipe surface. Intermittent
washing or flushing is usually required. As in the dry pre-
cipitator, the basic design parameters vary with each ap-
plication.
BASIC STRUCTURAL CONSIDERATIONS
Structural engineering and design considerations are fre-
quently overlooked by the engineer who specifies and
buys electrostatic precipitators. Such details are perhaps
as important as the design of performance parameters if
the machine is to operate reliably for its expected life. It
is often presumed that the manufacturer's experience
MAY 26,1975/CHEMICAL ENGINEERING
2-5
-------
and engineering capability should be accepted at face
value. Consequently, evaluations of proposals are based
essentially upon migration velocities, face velocities,
quantity of collecting surface, electrical supply, section-
alization, number of rappers, and the like.
Of course these criteria are extremely important in
comparing competitive designs. And, in most cases, the
manufacturer's structural engineering capability is suf-
ficient. Still, it is prudent to evaluate the structural de-
sign—at least, beyond the details of casing thickness, de-
sign pressure and temperature, and construction mate-
rials (which are the items usually included in proposals).
Hence, to provide a more complete picture before fi-
nal evaluation, you should thoroughly examine (a) the
supplier's standard fabrication and erection dimensional-
tolerance system; (b) his procedures for qualifying his
subcontractors; (c) his quality-control and inspection
procedures in the shops and in the field; (d) the caliber of
the individuals designated as construction supervision or
advisers and service engineers, and (e) even the workload
in his organization to assess the level of competence that
will be applied to your precipitator.
This section of our article is not intended as a con-
struction or design manual for electrostatic precipitators.
Each manufacturer has his standards, design philosophy
and "track record." Certainly, we do not wish to evaluate
here the relative merits of the designs offered. Rather,
what follows is an attempt to provide the user's buyers,
engineers and operators with a guide that will enable
them to examine and discuss the equipment with the
-manufacturers more knowledgeably. Assurance that the
electrostatic precipitator will be designed and erected so
that it will operate both reliably, and as intended, should
be based upon considerably more than merely pages of
legalistic terms, conditions and warranties. (It is now
generally accepted that low price and "paper" guarantees
do not generate the best investment in pollution-control
equipment.)
In the competitive atmosphere of a "bid business," a
manufacturer will generally propose only the equipment
and features that he believes necessary and sufficient to
meet the contract requirements. Although he will strive
to point out his "features" at sales presentations, he will
rarely offer designs that, in his judgment, will make his
price noncompetitive. He will try to avoid increases in
proposal and engineering costs that will be caused by de-
viations from his standards, which, after all, were devel-
oped largely as a result of commercial aspects and his
knowledge of operating experience.
Therefore, a plant engineer must either carefully ex-
press his needs in the specification, recognizing the im-
pact on his cost or, alternatively, evaluate proposals with
extreme care.
Catastrophic structural failures have been rare (al-
though they have occurred). What one must be wary of,
usually, is the subtle structural problem. One such-ntoK
lem might be lack of provision for expansion, possibly
stemming from a temperature assumption that allows no
marein (causing, excessive deflection of the substructure
or the interior precipitator beams and columns). There
are^many other such problems, even insufficient attention
to fabrication and erection tolerances (which will result
CHEMICAL ENGINEERING/MAY 26,1975
in misalignment and other operating difficulties after the
precipitator is onstream).
Under the pressure of plant operation, with the natural
reluctance to shut down units because of high cost, the
true causes of malfunction are extremely difficult to as-
certain, and repairs are expensive in both time and
money. Even more disturbing is that the symptoms are
often the result of several deficiencies that overlap and
tend to mask each other. Fortunes have been spent by
manulacturers and plants in making corrections and
modifications in one shutdown after another, frequently
without any benefit.
Problem prevention usually requires no more than a
professional attention to detail, proceeding from a know-
• ledgeable understanding of functional requirements, in
logically following the design, fabrication and erection
phases of the precipitator. The cost of this review will be
far less than that of even a few days of forced shutdown.
The Support Structure
Use of standard A1SC (American Institute of Steel
Construction) allowable stresses and deflections for
Basics of Electrostatic Precipitators .
Particles in the dirty gas entering a precipitator are
charged by the discharge electrode. The particles then
migrate to the collecting electrodes (collecting plates),
where they adhere and lose their charge. The particles
are generally dislodged from the collecting plates by vi-
brators or rappers that are attached to the plates, and
fall into hoppers below the plates. In some cases, how-
ever, it is necessary to wash the particles off; liquid par-
ticles may drip off by themselves. "
2-6
-------
ELECTROSTATIC PRECIPITATORS . . .
structural-steel design may not provide a rigid enough
"platform" to support the multiple columns of a precipi-
tator. Some deflection is inevitable, but it is the relative
deflections between support points that must be exam-
ined for the effects on internal alignment.
In an attempt to minimize these relative deflections.
manufacturers generally place high restrictions on the al-
lowable deflections, frequently leading to the need for
massive structural girders. Additional structural columns
to grade should be studied as an alternative (to reduce
the support of the precipitator by members in flexure).
Because the casing expands in the plan view with in-
creasing temperature, it is necessary to provide for slid-
ing at the support points, with resulting frictional forces.
This friction causes torsion on the structural member
(since the sliding takes place across its top flange), and
consequent bending in the precipitator columns. Also,
the vertical loads will be eccentric to the steel member in
either the hot or cold position (or both), depending upon
the detail dimensioning. Most designers will offset the
precipitator columns relative to the steel for half the cal-
culated expansion.
The hoppers (with insulation and lagging), moving
with the casing, must clear the flanges of the support
girders in both the cold and hot positions. To provide the
space required, in view of the large girders (usually need-
ing rather wide flanges), the precipilator columns are ex-
tended—in effect raising the casing in relation to the steel.
The bending moments caused by friction forces at the
base are thus increased and the columns must be checked
carefully. Errors occur here frequently because of poor
coordination between designers of the precipitator and
those who design the structural steel.
To reduce the frictional resistance, Lubrite (or other
friction-reducing plates) are interposed between the
structural steel and the precipitator-column base. When
these are used in the design, one should carefully check
allowable bearing-pressures and dimensional details in
the hot and cold positions. Installation must be super-
vised to assure that the sliding surfaces are horizontal.
Fabrication and erection tolerances usually dictate that
"shim packs" be installed between the precipitator base
and the structure, thereby allowing for vertical adjust-
ment of the casing. Logically, the base of the precipitator
should be detailed "high" in the event that the structural
tolerance is on the "plus" side at erection—otherwise.
shims would be of no value.
Perhaps some adjustable "jackbolt" provision could be
made so that shims could be fitted after the loads have
been applied rather than as they are commonly used to-
day—that is, as a tool to overcome tolerances "to a level"
after erection of the steel.
To control expansion movements, most precipitators
are anchored to the structure at one column base, with all
others allowed to grow radially from the one point. If all
the column bases are bolted to the steel, it is necessary to
make provision for movement by using slotted holes.
Considering fabrication and erection tolerances and the
possibility of overtightened bolts, many designs incorpo-
rate shear-bar guides and stops at a limited number of
columns along the two lines that radiate from the anchor
column parallel to the structural steel.
These bars both control the expansion and transmit
lateral shear at the base (from wind or seismic loads.
which may be in excess of frictional loads). These shear
bars must be located accurately and must be deep
enough to allow for the maximum shim pack. Therefore,
they should be placed after the casing is installed. Admit-
tedly, access to the shear bars at this time is somewhat
difficult and, occasionally, an erector may place the shear
bars prematurely. This is another reason for careful field
quality-inspection.
Casings
Functionally, the precipitator casing both houses and
supports the collecting surfaces and discharge systems,
forming a gaslight enclosure between the inlet and outlet
plenums or flue connections. Therefore, the casing is sub-
jected to, and must be designed for:
• Static or dead loads of all components, including
any equipment located on the roof, superstructure
weights, hoppers, and dust loads.
• Roof live-loads and snow-loads.
• Loads and movements imposed by connecting flues.
• Wind and seismic loads.
• Internal gas pressure (or vacuum).
• Dynamic loading imposed by vibrators and rappers.
Additionally, the casing will be subjected to the ele-
vated temperature of the flue gas: hence, the design must
provide for the casing's overall expansion. Superimposed
upon this expansion relative to the surrounding structure,
there are also the differential expansions between com-
ponents, and the consideration of thermal gradients in
specific members as the stresses and deformations are in-
fluenced by end-restraints (connections).
The structural components of the casing are:
Upper Beams—These beams support the collecting sur-
faces and discharge systems, transformer and other elec-
trical equipment on the roof, as well as the roof casing it-
self, and part of the superstructure and other loads.
Because these beams are somewhat shielded from the gas
stream, they are subjected to temperature differentials
between bottom and top, particularly during the tran-
sient states of startup and shutdown.
Columns That Support the Upper Beams—These col-
umns are subjected to lateral loads: i.e., pressure from
wind that is transmitted through the casings, seismic
forces, and flue loads. Since the columns form part of the
heated enclosure, the bases must be permitted to "slide"
relative to the cold steel substructure. Therefore, any
friction loading results in additional column stresses.
In multicell or multichamber precipitators, interior
columns are required to reasonably limit the spans of the
beams. In the direction of gas flow, an interior partition
ties the columns together, acting as a diaphragm. Fre-
quently, this partition is a double wall having insulation
• between two panels to provide a heat break in the event
one chamber is shut down. Certain designs employ brac-
ing in lieu of panels. Placed at ninety degrees to the gas
flow, the columns are either designed to be self-support-
ing for the full height, or use K-bracing of one form or
another.
Some casings are developed as horizontal panels, with.
2-7
MAY26,1975/CHEMICAL ENGINEERING
-------
the "columns" formed by massive vertical "stifleners."
. Lower Beams—A series of lower-beams (and baffles)
tie the columns together, supporting hoppers in most
cases.
Casing—The casing consists of stiffened panels de-
signed to transmit pressure and wind loads to the col-
umns. Also, external loads exerted by plenums and flues
(or stub stacks) may be transmitted through the stiffeners
of the casing sheets.
Roof—The casing roof consists of secondary members
and a cover plate that supports the discharge system
(through electrical insulators), transformer-rectifier sets,
control panels, etc.
Stresses in the Structure
We must readily appreciate the possibilities of thermal
stresses in a structure of the foregoing type, resulting
from differential heating and cooling of the elements.
Aside from the potential for structural failure, partic-
ularly at connections, excessive deformation of the casing
structure may cause misalignment of the discharge and
collecting systems. If such deformation is purely elastic,
internal inspection of the precipitator in the cold condi-
tion will reveal nothing. Therefore, the best protection is
a rigorous and thoroughly understood design analysis,
paying particular attention to deflections and thermal
stresses.
It would be well to become aware of the manufac-
turer's design criteria prior to award of the contract, and
to include a review of all stress analysis in parallel with
the approval of drawings during the engineering phase of
the contract.
The increased allowable stresses for A-36 steel over A-
7 has tended to "lighten" structures. However, the modu-
lus of elasticity (governing deflections) has not increased
and. therefore, deflections are greater in many modern
designs. As the temperature increases, the modulus de-
creases, so that deflections in the hot condition should be
carefully examined.
Fortunately, most precipitators operate at gas tem-
peratures below 800°F, so the phenomenon of creep
rarely needs be considered. However, because of (1)
marked reductions in allowable stresses, (2) the possi-
bility of graphitization, and (3) decreased oxidation cor-
rosion resistance in carbon steel (A-36) at temperatures
in excess of about 750°F, manufacturers look to other
steels for construction.
Because of availability and cost, there is a tendency in
the direction of certain proprietary materials such as
Corten or Mayari R as opposed to "pedigreed" alloys;
i.e., those recognized by the ASME codes as acceptable
for elevated temperatures.
We would recommend to the plant engineer that he
carefully discuss these materials with the metallurgists
employed by the suppliers. In our opinion, there is little
or no structural improvement made through the use of
these materials (in place of A-36) up to gas temperatures
of 800°F. Actually, designs to higher stress levels, a rea-
son for selecting Corten A. again may result in greater
deflections of members.
The primary benefit of using these proprietary steels
mayjie in some increased resistance to certain corrosive
media at low temperatures and during shutdown. Note
that although the gas temperature may wisely be used as
a design temperature,'the metal temperatures of main
structural members should be about 50 dec lower.
Inspection and maintenance of both collecting and
(most particularly) discharge sytems. is (in many designs)
via key-interlocked doors through the roof casing (usu-
ally one door per dust-plate section). The operator has to
crawl under casing stiffeners and over and around sus-
pension hardware. The floor of the crawl space is usually
the discharge-electrode support system, with the implica-
tions of scraped shins and sore knees. Specifying min-
imum clearances will eliminate the competitive tendency
among manufacturers to reduce casing, rapper shaft, and
suspension hardware costs by providing uncomfortably
low headroom. Walkways and access doors between
fields are a worthwhile investment for inspection, clean-
ing and general maintenance of the precipitator inter-
nals.
Miscellaneous
Pains must be taken during erection of casing, col-
lecting and discharge systems to assure alignment. How-
ever, because of the effects on deflections of the struc-
ture, final alignment should be checked after all the loads
have been imposed on the precipitator. These loads must
include attachment of both the inlet and outlet flues,
which often impose significant forces and moments on
the casing. Frequently, the precipitator erection is "com-
pleted" by one contractor, with another installing the
flues at a later date. Normal fabrication tolerances may
require considerable "jacking" of the materials in order
to make welded connections. This process may seriously
impair alignment of internals.
It is preferable that alignment be rechecked after a
short period of initial operation, allowing the system a
sort of "shakedown run" through a few temperature
cycles, because of the possibility that some weld connec-
tions may yield plastically.
Designs that separately support inlet and outlet ple-
nums to relieve the casing of the loads must include sup-
ports that compensate for expansion of the precipitator
relative to those supports. Pressure loads on the plenum
faces may transmit torsional loads onto precipitator col-
umns, depending upon the connection design.
Roofs, Penthouses and Superstructures
Insulator-Compartment Designs—Some designs pro-
vide separate insulator compartments, with the roof cas-
ing covered by insulation, and with a walking-surface
deck plate. This deck plate will be cold relative to the
casing. It must be adequately supported, either through
rigid insulation or metal framing, and sloped for drain-
age (so as to prevent water and ice accumulation).
Clearances must be provided for the movements of in-
sulator compartments, rapper-shaft sleeves and any other
equipment or equipment supports that will move as the
casing and precipitator structure expands.
Designs for any of the deck plate's metallic supports
CHEMICAL ENGINEERING/MAY 26,1975
2-8
-------
ELECTROSTATIC PRECIPITATORS . . .
that are attached to the casing must be examined for ex-
pansion provision. The criticality increases with higher
gas temperatures and larger precipitator size. The ex-
posed deck plate must be maintained watertight to pre-
vent damage to insulation.
Penthouse—Penthouse designs extend the casing over
the precipitator to form single large insulator compart-
ments, eliminating the need for bus ducts.
The operating temperature of the penthouse is con-
trolled by radiation and convection from the precipitator
roof, conduction through the casing, and the flow of pres-
surizing air required to purge insulators. Heat losses de-
pend upon the method of insulation. Frequently, the in-
sulation is attached to the underside of the penthouse
roof and to the inside of the walls so that the roof is cold
steel, presenting a walking surface.
The structure is then required to absorb'the expansion
differential between the precipitator and the penthouse,
with reactions transmitted into the casing, with a poten-
tial for causing distortion and misalignment.
Rapper shafts extending through the height of the
penthouse, with rappers mounted above the roof, must
be protected against any forces that may cause misalign-
ment or binding.
If pressure testing of the precipitator shell along with
the flue system is desired, it is important that the pent-
house design-pressure be specified. Frequently, pent-
houses are designed of thinner materials than the precip-
itator casing. At best, pressure tests of precipitators are
difficult and costly because of the many penetrations that
must be temporarily sealed.
Weather Enclosures—Superstructures, or weather en-
closures, may be provided to protect personnel and
equipment; this will be money well spent, since it will as-
sure maintenance inspections on a routine basis. These
structures are cold, so connections to the precipitator
must transmit vertical and lateral loads while also allow-
ing for differential expansion.
Under weather enclosures, the precipitator roof is an
insulated walking surface but the surface need not be
weathertight (i.e., a sealed deck plate is not necessary).
However, in some monolithic castable designs that fail to
provide for differential movements in the casing, the
penetrations and the castable do not last very long. Fis-
sures, and the ultimate development of rubble, allow
radiation and connection from the casing, which will
make the area extremely uncomfortable for maintenance
personnel.
Weather enclosures must be properly ventilated for
both the comfort of personnel and the protection of elec-
trical equipment and controls that are located on the pre-
cipitator roof. Undersized roof vents or gravity ridges
may be less costly than liberally sized louvered blowers
located along the walls; but in the long run, the oper-
ating difference will be appreciated.
Hoppers
Whether suspended from the casing or supported di-
rectly on the substructure that is interposed between the
casing and the steel, hoppers are required for both collec-
tion and temporary storage of the dust. The simplest and
most- common hopper is pyramidal, converging to a
round or square discharge. Frequentlv. the.honpm are
baffled at the division between two djisJ-.jilaLp-sp'ai™* t^
prevent gas bypassing the treater.
In ceuain application's, where the dust is to be re-
moved by screw conveyors, the pyramid converges to an
elongated opening along the length of the conveyor. In
others, where the dust agglomerates in a sticky fashion
(e.g., salt cake) and has a tendency to build up on any
sloping surface^the hoppers are eliminated and the cas-
ing is extended down to form a flat-bottomed box under
the precipitator. The dust is removed by drag conveyors
that cover the entire bottom plate.
For many applications, however, hoppers slopes will
shed the dust (provided the angles are steep enough).
The valley angle (the angle between the corner of the
hopper and a horizontal plane) should be checked as the
governing minimum .slope. Although valley angles as low
as 52 deg have been used successfully, for some pro-
cesses, 60 deg is an often-specified minimum.
hOppers must be kept clean and dry. Therefore, al-
though many designs do not require vibrators (they are
both costly and require maintenance), it may be prudent
to examine past experience in operating plants of the
same or similar process and install mounting provision
for vibrators anyway (to avoid later costly removal of in-
sulation and lagging if operation shows the need for vi-
brators).
Moisture-laden dust that hits cold steel hoppers, has a
tendency to stick. Therefore, insulation of hoppers is vi-
tal. However, this is sometimes not sufficient, and .addi-
tional heating of the hoppers may be advisable, albeit ex-
pensive.
Installing hopper heating after operation is a costly
construction affair involving building of scaffolding, and
removal and replacement of the insulation and lagging.
Therefore, in marginal cases, provision might be made in
the design of the hopper and insulation to allow for fu-
ture installation of strip heaters. Other common methods
of hopper heating are attachment of mineral-insulated
heating cables or by installing steam tracing. These de-
vices are best applied at initial erection.
The discharge of the hopper should be as large as
practical and the inner surfaces free of all projections
and rough welds. Internal ladder rungs, attached by
welding to hopper walls, in addition to presenting pro-
jections for dust buildup, provide a hazard to personnel
in the event of weld failure caused by corrosion or exter-
nal vibrators.
When a baffle extends too far down into a hopper,
there is a danger of its acting'as a "choke," causing bridg-
ing between the baffle and one or both sides of the hop-
per. Stopping the baffle a liberal distance (say 2 ft) clear
of the sloping hopper wall should not, usually, allow gas
bypassing. A gas sweep under a baffle of this type, con-
sidering the pressure drop of the turn, is possibly (and
even probably) a symptom of poor gas distribution to the
precipitator (that is, a downward jet at the entrance).
Access to hoppers should be via external, key-inter-
locked doors. Bolt-on doors through baffles should be
avoided because of the dangerous possibility of dust ac-
cumulation on the far side of the door. Liberal "poke-
2-9
MAY 26,1975/CHEMICAL ENGINEERING
-------
hole" ports should be provided to allow for clearing k
blockage at the discharge.
Overfilling of hoppers is a cause of precipitator prob-
lems. Dust buildup is capable of lifting discharge sys-
tems, shorting out sections, and frequently causing elec-
trode breakage.
Level alarms are extremely valuable, provided they
are kept in working order. Too often, because they are
located near the top of the hoppers (even so high as to
place them above the bottom of the structural steel sup'
porting the precipitator), they are inaccessible for peifl-
odic inspection and maintenance. Also, the temperature
of the atmosphere in this confined area may be sufficient
to cause the alarm mechanism to fail; this is a point for
critical review of details prior to installation of the instru-
ment.
Hopper capacity should be carefully checked to pro-
vide a reasonable lime for minor maintenance of the
dust-removal system.
With certain types of dust, either those that are pyro-
phoric or contain high levels of carbon, the danger of fire
increases if hoppers are allowed to overfill and smolder,
up to the levels where a spark from the precipitator may
ignite the mixture. The smoldering material, itself form-
ing clinkers, may structurally damage the hopper without
actually breaking out in flame. Poorly placed (or faulty)
level alarms cannot be the primary cause of the damage,
but the preventive alarm function is one that is certainly
S-l^Wire support anglers
^::V'h-\^^£ife: *
-.. <,•.;--.•* •.!•£*-.,-•
\: .-.."•.-'••.•.-i. iSSv--ip-.i3srJSav.--j,
^ -V-<-B^** ^ir'^T'uXjTKSKCj^T**J^^*3eT**J(3*\i&M<'J*™'V>lr^' J •* *
'•'^•M^S^^^i^^^mj^-i
'' £'. -'-'• ' Discharge electrode K^SfS^f'^r**^ •'
(loop rings top and bottom)
• Plate hanger bolt
v.- ' - ..• lk*3i **C"JNJ-" -iff*
•>,_t- .H.^Un^L^«^tiM;&ail3x;^,:%M
^-•.^v.^/^ffptf"*^^^1 >*N:
^^-i^^:,- <-^
'.. .-'i —i^« -*'fr' j»- ^^^^
I' .. • -r: - •.• •*•.•-*'-,•••-.»
I'. . •-:. •kj.-it-vc-f.
E' ' ~~ ':&^i^f
fc *...- ".itV. '!-• -i^ff'^T
•.^.-... -,
•• j>p.. v.'SSjtej'.i .Keif -•'••
- —•* ,^,-~,.^!>^rft,:^:,f<^ .. •
"'
..... .'•'-?•••;•«-£ "^~ —" Opzel plate "•«.•-,•;i
."•- •>/!** -*;l v**-4 • *,v,(jitffl*rv*te«-i»** j-t-tt .r-.
:\*^g^^£l^%'fJ5ir '
. .1 Plate spacer bar-5^
DETAIL of wire-weight system precipitator electrode construction—Fig. 5
CHEMICAL ENGINEERING/MAY 26.1975
2-10
-------
ELECTROSTATIC PRECIPITATORS . . .
well worthwhile designing and maintaining properly.
Anywhere from 60% to 70% of the dust will be re-
moved by way of the inlet hoppers. However, in the
event of inlet-field failure, the dust load will be trans-
ferred to the next hopper downstream, and so forth. This
point is particularly important in sizing both discharge
systems and conveyors. As for conveyor selection, the
Conveyor Equipment Mfrs. Assn. places a manual at the
disposal of the plant engineer. Just a few points are
worth mentioning here. There is no better assurance of
performance than a liberally sized, low-speed conveyor,
that normally operates at 15-45% of the trough loading,
depending upon the dust. The loading should be based
upon the lowest anticipated dust density.
Alignment of the conveyors is important, and to a
great extent it will depend upon the alignment of hopper
connections. Because of the difficulty in erecting multiple
hoppers to close alignment tolerances, field-adjustable
flange connections are recommended. Also, provision for
expansion between hopper connections and conveyor
troughs must not be overlooked.
Discharge Systems
There are two basic designs of discharge-electrode sys-
tems offered today by manufacturers.
Wire-Weight System—The wire-weight system consists
of individual electrode wires suspended from an upper
support-frame. The wires are best shrouded in some
fashion to prevent arcing to exposed, sharp, ground
edges, or where the electrical clearance is reduced by
passing the tops and bottoms of the collecting dust plates.
The wires are held taut by suitable weights suspended
from their bottoms. The weights in turn are spaced by a
guide frame. The frame must be stabilized against swing-
ing, an action that may be generated mechanically by the
gas stream, through "electrical wind," by an improperly
functioning automatic voltage-control, or some combina-
tion of these.
Commonly, the stabilization is accomplished by
trusses extending from the upper support-frame to the
guide frame. Rapper energy, transmitted through the
trusses, aids in keeping the lower guide-frames clean.
Guide frames of a design that permits dust buildup
high enough to raise the weights may cause slackening of
the wires, arcing, and eventual wire failure.
Guide irarnes"tnat are stabilized by ceramic, or other,
insulators from the casings or hoppers can be a mainte-
nance problem. Dust buildup on the insulators during
operation, although resistive in some cases, presents a
source of leakage to ground. Moisture gathered during
shutdown (or low-load operation) might lead to com-
plete failure of anjnsulator.
Hlectrode wire failure will be virtually nil, assuming:
• Reasonable care, during erection, in alignment of
the casings and surfaces.
• A well-designed support, guide and stabilizer sys-
tem.
• Reliable, properly adjusted automatic voltage-con-
trols.
• Good operating maintenance of the dust-handling
system.
Unfortunately, there are many precipitators in which
the foregoing preconditions have not been met; repeated
wire failure (really the symptom, not the problem) has
been the culprit in the eyes of the operators.
Rigid Wire Frame—The rigid wire frame was furnished
by U.S. suppliers prior to 1950 and then virtually aban-
doned (in favor of the wire-weight design) because of its
many reliability and operating problems. Recently, U.S..
licensees of foreign manufacturers have reintroduced
frame electrodes to this country, and there are quite a
few modern installations both in operation and on order
(as there are wire-weight designs in European plants).
The rigid frame requires a high degree of quality con-
trol, both in fabrication and erection, and is intrinsically
more costly. Replacement or repair is an expensive, time-
consuming undertaking. It is about the same as attempt-
ing to replace a dust-collecting plate.
At lower temperatures, up to 400°F, warpage of the
frames is uncommon, but for operating temperatures
above 400°F, or with cyclical operation, potential defor-
mation of the frames becomes seriously undesirable.
The rigid frame employs wider gas lanes, or ducts, to
provide electrical clearance between the frame and the
dust-collecting plate. This.leadsJaUigej cosines to house
the re.ouired surface-areas.
It is important that the engineer be fully aware of the
differences and the requirements of each design philoso-
phy in detail, so that he avoids incorrect evaluations of
one versus another.
The erection sequence usually consists of casings and
hoppers first, followed by collecting surfaces and, then,
discharge systems. If the casings are not erected to true
dimensions, plumbed vertically, and square cornered in
the plan view, attempts will often be made to compen-
sate during the installation of collecting surfaces, i.e., us-
ing guides that should be free of frictional loads as
"jacks," and so forth. Then, the discharge system, which
should hang freely, is stabilized in an offset position to
maintain, as best as possible, the wire-to-plate centers.
This kind of construction will most probably prove an
operating headache from the first day on; again, this is a
vital reason for in-depth, step-by-step quality control and
inspection, regardless of the pressure of construction
schedules.
Discharge systems are supported from the casing
through standoff electrical insulators. These must be kept
in a clean, dry condition during operation, to prevent
dust- or moisture-coatings from accumulating, because
such coatings present a leakage path to ground. Wet ac-
cumulations can be very common during shutdown as
the moisture in the gas condenses.
Warmed, filtered pressurizing-air supplies must be
adequately designed to prevent such a problem. The sys-
tem must provide distribution to a multiplicity of insula-
tors, none of which may be allowed to "starve" because
of disproportionate flow. This design problem is similar
to that of balancing an air-conditioning system. Some
method for checking distribution should be provided to
the operators. Maintenance routines of changing filters
and checking heater elements also should be established
as soon as the system is operational.
Most commonly used electrical insulators lose dielec-
2-11
MAY 26,1975/CHEMICAL ENGINEERING
-------
trie strength as temperature increases. Although the max-
imum temperature varies with the insulating material,
400°F is a probable limit. Therefore, it is necessary that
electrical insulators be isolated thermally from hot gases.
The purge-air system normally suffices, but insulators
mounted on hot casing-steel may be affected by conduc-
tion, at least for several inches along the length. Fortu-
nately, most electrical insulators retain structural strength
under higher temperatures and also act somewhat as
thermal insulators, so that, if the electrical path is long
enough, the effect of the conducted heat is limited to a
short distance up the insulator.
Collecting Surfaces
The keys to any successful collecting surface configura-
tion are: •
Dust-Plate Trueness—It is important to ensure true-
ness of the individual dust plate, i.e., it must be free of
kinks or excessive "oil canning." This trueness will de-
pend upon care in fabrication, the packaging for ship-
ment, and how the surfaces are stored r nd handled in the
field. Most manufacturers will supply complete proce-
dural details if requested to do so.
Dust-plate bundles should be stored on edge, on
closely-spaced, level dunnage that has been positioned to
take the loads directly from the shipping frames on the
bundles. If stored for long periods, bundles should be
protected from weather.
Most damage occurs during unpackaging and raising.
Supervisory experience is the best guide. Although many
foremen are thoroughly competent, having been in-
volved in the erection of other precipitators, it is wise not
to leave this operation solely to the devices of the work-
men. Removing a damaged dust plate after the unit is'
"buttoned up" is difficult and costly. Close inspection as
dust plates are installed will pay dividends.
The effect of wind on deformation of surfaces during
raising is minimized by dust-plate packaging designs that
permit installation of entire bundles into the casing shell
prior to their being opened.
Depending upon orientation of the gas inlet, high
winds passing through during construction—in the direc-
tion of gas flow—may damage plates, and this may go un-
noticed if inspection is not constant.
Ruggedness of the Support System (and Its Dimen-
sional Tolerances)—This system supports the dust
plates and, in many designs, must transmit rapper en-
ergy to them. Be aware that a manufacturer's "stan-
dard" design may not be sufficiently rugged for all types
of rappers. The design must allow for alignment adjust-
ment at erection and should also allow for readjustment,
if necessary, after shakedown operation. Particular atten-
tion should be paid to the effects of vibration and impact
loading (notch sensitivity) at all welded points.
Alignment—Sufficient, adequate spacers must be
provided to maintain alignment, while also allowing for
possible temperature variation between dust plates (and,
of course, temperature differentials between the casings
and dust plates).
Often, the surfaces are guided at the bottom by using
interior plates whose primary function is to provide a gas
TYPICAL transformer-rectifier (T-R) set
for powering precipitator—Fig. 6
baffle. Such baffle plates may be somewhat thin, and may
expand more than exterior casings; hence they fre-
quently distort (or buckle). If this happens, the collect-
ing surfaces may be pulled out of alignment through the
guides.
Rapper Anvils—Rapper anvils that are attached to
either dust-plate supports or rapper header beams re-
quire particular attention because of the duty to which
they are subjected. Since alignment is extremely critical,
designs that permit bending of flanges, or other local de-
formations, must be questioned.
Alignment Tolerances—Beware of "tolerances"
given on drawings to govern alignment unless there are
full explanations of:
• How materials may be installed to those tolerances.
CHEMICAL ENGINEERING/MAY 26,1975
2-12
-------
ELECTROSTATIC PRECIPITATORS
Rapper-
T-R sets
_ Pipe and
guard run
T-R SET, rappers and other items of precipitator equipment mounted on the roof—Fig. 7
• How inspection and checking is intended. Inspection
by "eyeballing" is always somewhat subjective.
• How adjustment is provided to meet the require-
ments. .
Baffles—The baffles between the extreme outer dust
plates and the casings are necessary to prevent bypassing
of untreated gases, because there are no discharge elec-
trodes in this space.
Many designs call for such baffles to be installed after
the surfaces are in and aligned. Unless there are walk-
ways between fields (and the baffles are thus made acces-
sible), such installation is almost impossible because of
workspace limitations. Erectors often add these baffles to
the casing plates while they are still on the ground, or be-
fore the dust plates are installed. The result is frequently
a less-than-desirable closing of the space. Even inspec-
tion is difficult. Therefore this feature is best discussed
with the manufacturer during the design stage.
Rappers
Rappers are used to remove dust from the collecting
and discharge surfaces, and their effectiveness and relia-
bility are vital. The types generally furnished are:
• Electromagnetic impulse, either single or multiple.
• Electric vibrators.
• Pneumatic impulse.
• Various mechanical hammers. These are usually as-
sociated with foreign designs but are sometimes fur-
nished by others for special applications.
Each manufacturer has developed rapper applications
for compatibility with his suspension system and rapper
schedule (number of surfaces per rapper), based upon his
experience and tests. Generally, pneumatic rappers will
impart more energy than either electromagnetic rappers
or electric vibrators, and will remove tenacious dusts
more readily. However, it is important to be certain that
all the hardware in the system is designed to withstand
such high-energy forces. Changing from electrical vibra-
tors to pneumatic rappers (in an attempt to improve op-
eration) without also "beefing up" the hardware has led
to structural failure.
Mechanical hammers are frequently very effective. But
moving parts in the dirty.gas stream become a mainte-
nance problem. Any repairs will require shutdown of a
chamber or system.
In the final installation, it is important that there is no
binding of rapper shafts against casings so that the
energy is directed where it belongs—to the surfaces to be
cleaned. The design of rapper shafts through penthouses
should be examined thoroughly with regard to the ex-
pansion differential between the penthouse and the cas-
ing.
Rapper controls should be readily adjustable for inten-
sity, sequence and cycle time. Optimizing these for best
rapper performance with the aid of a dust density meter
in the outlet or stack) is a good practice.
It is prudent to check—at least once a week, but prefer-
ably once a day—that all rappers and the controls are in
working order. A manual control that allows operation
of one rapper at a time, coupled with an indicator panel
in the main control room that monitors the automatic
control, is an excellent feature. (Indicators that simply
monitor power to a rapper may not tell the full story of
the operation.)
Equipment on the Roof
The first-cost economics of locating automatic voltage-
controls and rapper controls on the precipitator roof, or
"operating floor," are attractive. However, the reliability
of control elements that are subjected to the hot and gen-
erally dusty atmosphere makes it well worth evaluating
the additional wiring and conduit cost that results from
locating the controls in the main air-conditioned control
room.
2-13
MAY 26,1975/CHEMICAL ENGINEERING
-------
In engineering a precipitator, manufacturers may not
provide a detailed layout of all of the components on the
roof, thereby leaving conduit or piping runs to a contrac-
tor's judgment.
Frequently, the problems of close spacing of equip-
ment, and of access to doors and cabinets, do not become
evident until after erection. Further, the contractor tends
to install the lowest-cost system possible in keeping with
his specifications. This may result in an overall layout
that is not conducive to good maintenance. The best pro-
tection is a concise specification and a detailed drawing
study.
Although infrequent, transformer-rectifier (T-R) sets
do sometimes fail. The methods of removal and replace-
ment should be thoroughly understood and agreed upon.
Again, although the first cost may be higher, transformer
rectifiers located on a separate platform—with good ac-
cess for removal—could be a sound investment. A spare
T-R set, in storage on the platform, would make for min-
imum downtime after a failure.
A kilovoltmeter in the control panel of the T-R set is
more indicative of what is happening in the precipitator
than is a voltmeter on the transformer's primary side.
When a kV meter is not provided, T-R sets may be
equipped originally with provision for attaching such a
meter for checking and test purposes.
Gas Distribution, Gas Proportioning, and Flues
Uniform gas distribution, with the gases entering per-
pendicularly to the face of the precipitator, is important
to proper operation. High velocity jets may either cause
erosion of dust from collecting surfaces or permit vol-
umes of gas to move through the machine relatively un-
treated.
Flow-model studies (and the distribution devices in-
stalled—most often one or more perforated plates) are
not always effective in developing the desired distribu-
tion. However, flow-model studies will generally not be
conducted unless specified by the customer although they
are at least qualitative indicators of what is going on. Un-
fortunately, most flow devices installed are fixed by de-
sign, so changes or adjustments require costly shutdowns.
The velocity of the gas entering the precipitator is so
low, and the areas are so broad, that a pitot-tube traverse
to check velocity is impractical. Instead, the most com-
monly used tool is the hot-wire anemometer. It would be
worthwhile to allow a period in the construction schedule
of two weeks to one month to allow for conducting tests
and making adjustments before operation. Most pro-
grams have a time schedule that precludes this step as a
luxury. If these adjustments are neglected, an equivalent
period may have to be spent in making the changes after
startup.
There are devices such as the "Konitest" that may be
used effectively, but they require that specially designed
hardware be installed initially in order to make them
practical. Attention must be paid to flue designs to avoid
both the close coupling that makes distribution more dif-
ficult, and surfaces that may allow for gross dust accumu-
lation. Of what value is a flow-model study if the dust
builds up for several feet on the bottom of a flue?
Any distribution device must be kept clean through
adequate rapping.
Multiple-chamber precipitators require some means
for gas proportioning. These are most commonly louver
dampers at the outlet. Guillotine shutofT dampers at the
inlet should not be used for proportioning since they
tend to destroy proper gas distribution to a chamber.
During the initial layout phase of a project, attention
must be given to the adequate location of test ports for
both inlet and outlet sampling. After locating the ports in
a section of the flue where a reasonably uniform velocity
profile may be expected, it is advisable to provide proper
platforms and weather protection for the test crews.
ECONOMICS
Installed Cost
The installed costs of electrostatic precipitators vary
considerably, depending upon construction location,
whether they are new or retrofitted, and on the season of
the year. The fob. cost, however, is more-or-less predict-
able—it is basically a function of the area of collecting
surface provided.
After the collecting-surface area has been determined
(as described on p. 97), the packaging configuration is es-
tablished. The package is then subjected to various con-
straints, such as the length-to-height ratio (L/H), the
contact time, the number of fields, the gas velocity, etc. If
the relationships between the quantity of gas and the
package meet the criteria of past experience, that par-
ticular geometry will be priced and offered to the cus-
tomer. The specific geometry is also affected by a sup-
plier's standard modules for length, height and width.
Variations In Installed Cost
The limitations of the selection criteria as influenced
by suppliers' experience, plus variations in packaging
geometry, result in installed costs that vary by several
hundred percent for similar applications.
Electrostatic precipitation is a mature technology
wherein the fundamental principles have not changed.
The changes presently occurring are in the refinement of
application, such as attempts to maximize corona power,
improve gas distribution, optimize rapping techniques
and prevent reentrainment.
The fundamental change that has taken place in the
recent past is that the •technology is being required to op-
erate at its boundary conditions, which results in expo-
nential changes in costs. For example, older performance
curves that related efficiency with cost were relatively flat
and predictable in the efficiency range for which most
precipitators were bought (90 to 98%). But these curves
became asymptotic as the efficiency passed 99%. In these
days of EPA and state regulations, one seldom sees pre-
cipitator specifications calling for less than 99% effi-
ciency, and this is the area where prior knowledge is
practically nonexistent both for performance and for cost
data. This requirement imposes a need to completely un-
derstand the effects of independent variables on electro-
static precipitator efficiency.
CHEMICAL ENGINEERING/MAY 26.1975
2-14
-------
ELEC7PL0SVATIC PHECIPITATORS . . .
Cost Effectiveness of Efficiency Models
The current state of the art in predicting efficiencies
from the independent variables may show a 1% differ-
ence between the predicted and observed values. This is
a good statistical fit, but that 1% difference at 96-97% ef-
ficiency level translates into an almost 10% difference in
the precipitator's fob. cost. The same 1% difference in fit
about the 99% efficiency point can result in a 25% differ-
ential in cost.
Thus, to be cost effective, there must be significant im-
provement in the currently used efficiency models. The
current sizings generally contain a plethora of contin-
gency and safety factors to ensure compliance with the
customer's requirements. This is because the penalties
for being wrong are so high that they can bankrupt a
small company and seriously damage the profitability of
even the very large companies. These factors have re-
sulted in a marked tendency toward overkill in precipi-
tator design, and most users have had to fund for signifi-
cant added costs for pollution control over what they had
expected.
Cost Effectiveness of Precipitator Geometry
The specific geometry of the enclosure for the collect-
ing surface that has been specified by the efficiency
model, affects the installed cost.
One precipitator supplier can produce a relatively
small precipitator in 35 different "standard" combina-
tions, to meet various application criteria. The selection
of the least-cost configuration in conjunction with an im-
proved efficiency model is necessary to provide a truly
cost-effective precipitator to the users.
Operating Costs
The variables that have an effect on installed cost also
can affect operating costs. The major electrical require-
ments are a function of the design power-density (watts
per square foot of collecting surface). Power require-
ments range from 0.00019 kW/actual cfm to 0.00040
kW/acfm, according to a recent study of TVA installa-
tions. Additional operating costs can be incurred, de-
pending upon the specific installation requirements. Typ-
ical requirements are as follows:
• Rapper system—1 kVA/rapper panel.
• Control and signal power—0.25 kVA/T-R control
panel.
• Insulator-compartment vent system—4 kVA/
compartment.
These are generally ignored in cost comparisons, as
they are not significant when compared to the corona
power requirements. Another general guideline is that
the annual operating costs are 10% of the installed cost.
Maintenance Costs
Costs of maintaining a precipitator are influenced by
relative size, efficiency requirements and design parame-
ters. A review of TVA's costs shows a range of $0.01 to
$0.03/actual cfm of gas treated. Typical items of mainte-
nance are rappers, rapper anvils, electrode wires, ash-
handling-system parts, curtains and electrical controls.*
Meet the Authors
Gilbert G. Schneider is Executive Vice-President. Enviro Energy Corp.. Suite
220. 16161 Ventura Blvd.. Encino. CA 91436. An expert in air-pollution control,
he has been sales manager for the Western Precipitation Div. of Joy Manufac-
turing Co. He has a B.A. in mathematics from Wotford College. S.C.. and a
B.Ch.E. (rom Rensselaer Polytechnic Institute. Troy. N.Y. He is a guest lec-
turer at University of Southern California and is a member of AlChE.
"Theodore I. Horzella is Vice-President of Enviro Energy Corp. He received his
B.5. in chemistry from Santa Maria University. Chile, and his M.S.Ch.E. from
Iowa State University. Ames. Iowa. He has also done graduate work in busi-
ness administration at the University of California. He is a member of AlChE,
Sigma Xi. and the American Management Assns.
Jack Cooper is President of Enviro Energy Corp. He was previously Manager
of Design Engineering and Manager of Equipment Construction at Foster
Wheeler Corp. and was Manager of Engineering at Western Precipitation Div.
of Joy Manutacturing Co. He received his B.S. in mechanical engineering from
Polytechnic Institute of New York, and is a Registered Professional Engineer
in the state of New Jersey.
Philip J. Striegl is Vice-President of Enviro Energy Corp. He was formerly as-
socialed with Allis-Chalmers Corp. and Joy Manufacturing Co. plants. He has
a B.S. in metallurgical engineering from Illinois' Institute of Technology, and
has done graduate work in business administration at Marquette University
and the University of Wisconsin.
Reprints of this 15-page report on precipitators will be available shortly. Check No 227 on the reprint order form in the
back of this or any subsequent issue. Price: $2.
2-15
MAY 26,1975/CHEMICAL ENGINEERING
-------
V-3.
ELECTROSTATIC PRECIPITATORS IN INDUSTRY
by
Robert L. Bump
Research-Cottrell, Inc.
Copyright fc) by Chemical Engineering. McGraw-Hill
Publishing Co. Reprinted with permission, January 11,
1977 issue.
3-1
-------
Electrostatic precipitators
in industry
Here is an overview of the subject, touching on theory, design, sizing,
controls, component reliability, efficiency, the upgrading of old
equipment and the retrofitting of new, and the conditioning of gases.
Robert L. Bump, Research-Cottrell, Inc.
r] For over half a century, electrostatic precipitation
has been the method of choice to control particulate
emissions at industrial installations ranging from ce-
ment plants and pulp and paper mills to oil refineries
and coke ovens. In most cases, the participates to be
collected are by-products of combustion. In others, they
are dust, fibers or other small solids from a production
process.
In the past decade, precipitator design—spurred by
increasingly stringent emission regulations—has ad-
vanced at an especially rapid rate. Efficiencies and
availability records not considered possible a few years
ago are now routinely achieved. Although the geomet-
rically accelerating cost curves associated with higher
and higher purity standards are well known, the cost of
precipitators has not risen exponentially as might have
been expected. Why not? This article explores some of
the reasons.
General precipitator design
A modern precipitator system, whether it was
created to treat flue gas from a heat source or to deal
with particulates spilling from process streams, is likely
to be far superior to any unit that could have been
built 10 or 20 years ago. Among the factors underlying
this superiority are these: far-more-sophisticated math-
ematical techniques for predicting precipitator per-
formance; superior construction materials; comput-
erized data banks of technical information based on 50
years or more of experience in building industrial pre-
cipitators; availability of high-quality auxiliaries such
as flues, dampers and handling systems; design im-
provements growing both out of experience with earlier
precipitators and out of accelerating research pro-
grams.
Precipitator theory
Electrostatic precipitation is a physical process by
which a particulate suspended in a gas stream is
churged electrically and, under the influence of the
Typical precipitation
process
Dust on..
precipitator wall
3-3
-------
, Transformer-rectifier
Top end frames--.
High-voltage
conductor
High-tension ^,--
support-'"'
insulators
Perforated
distribution —
plates
Bottom end frames-""
Upper high-tension /
hanger assembly
(hanger and hanger frame)
. Ground switch box
on transformer
, Discharge-electrode
vibrator
Insulator compartments
^,- Handrails
'_ Collecting-electrode
rappers
'•Hot roof
Side frames
—Discharge electrode
Access door between
~~ collecting plate sections
-Precipitator
base plate
\ Slide plate package
Upper high-tension wire' /
support frame / • u • .
' / Horizontal »
Sway bracing / bracing strut •
Hopper ! /
!/
Steadying bars Hopper baffle Collecting electrodes
Stilts \
x
X Support structure
v^ . cap plate
Lower high-tension
\ steadying frame
\*.^---' J - • . ,..
electrical field, separated from the gas stream. The sys-
tem that does this (Fig. 1) consists of a positively
charged (grounded) collecting surface placed in juxta-
position to a negatively charged emitting electrode. A
high-voltage dc charge is imposed on the emitting elec-
trode, setting up an electrical field between the emitter
and the grounded surface. The dust particles pass be-
tween the electrodes, where they are negatively charged
and diverted to the oppositely charged collecting sur-
face.
Periodically, the collected particles must be removed
from (he collecting surface. This is done by vibrating or
rapping the surface to dislodge the dust. The dislodged
dust drops below the electrical-treatment zone and is
collected for ultimate disposal.
A commercial precipitator (Fig. 2) comprises sym-
metrical sections of collecting surfaces, discharge elec-
trodes, suitable rapping devices, dust hoppers, and an
enveloping casing and the necessary electrical energiz-
ing sets.
Mathematics and design
One major source of improvement in precipitator de-
sign is in the sophistication of the mathematical tech-
f Ml Ml* 4| I.Ni.lNtfrH!N<« |AM
3-4
-------
Type of process
Size or production
rate of process
Gas volume
Temperature
Gas analysis
Type of fuel"
° For particulato control on powor bolloro
Fuel analysis0
Dust analysis
Particle size
Resistivity
Efficiency required
Space limitations
Cliques employed to predict the exact size of a precipi-
iator to handle a specific task. When emission
oiandards were less stringent, larger margins of error
were tolerable. Now that standards have risen to im-
pose efficiencies of 99% or more, there is less room for
error.
Major advances in precipitator design techniques
have resulted in precipitators more precisely "tai-
lored"—and therefore more efficient in operation—to a
specific installation. The Deutsch equation, once the
definitive tool for precipitator sizing, is no longer ade-
quate to meet current demands for efficiencies well in
excess of 98%. A modified Deutsch equation, now in
use, factors in many of the newer practical considera-
tions inadequately expressed in earlier theoretical ap-
proaches to sizing.
Too conservative a design produces unacceptably
high equipment costs. Too "lean" a design means
unacceptable operating and maintenance costs—not to
mention stiff fines for out-of-compliance operation.
Sizing
The primary factors in precipitator sizing have been
fsxe velocity (the speed at which the gas travels through
the precipitator), migration velocity (the speed at which
the dust particle travels toward the plate under the in-
fluence of the electrical field) and aspect ratio (the ratio
of precipitator height to its length).
In recent years, however, a more sophisticated ap-
proach to sizing has evolved. Extensive investigation of
the relationship of process and operating variables to
predicted performance, combined with a wealth of ac-
tual field experience correlating operational versus pre-
dicted performance, has been pulled together to create
a central computerized bank of essential information.
From this data bank, which considers type of process,
detailed paniculate analysis, temperature, particle size
and dust resistivity, a precipitator sizing program has
been developed that generates a variety of acceptable
options. If, for example, several process variations are
possible (e.g., variety of fuels in a power boiler), the
program considers all of them—in contrast to the old,
often inadequate method .of selecting a migration ve-
locity for a single operating condition.
The modern result is a properly sized unit with less
guesswork and more certainty of predictable and
proper performance than ever befor'e. Equally impor-
tant, a way of mathematically modeling the effects of
other alternatives is also ea.sily available.
1,600
1,400
1,200
1,000
* 800
93
o
2 600
o
400
200
60 70 80 90 95 98 99 99.S 99.8 99.9 99.99
Precipitator efficiency, %
To obtain such a mathematical model, however, the
purchaser must furnish detailed process information.
Table I lists the minimum data required to size a pre-
cipitator. Without these data, assumptions would have
to be made that might not clearly identify the perti-
nent points affecting the proper equipment selection.
Also to be considered are operating and maintenance
procedures, including specific schedules for process
downtime and process-equipment inspection and re-
conditioning. These factors can influence original de-
sign parameters relating to cost and reliability.
Electrical sectaonalization
In order to ensure optimization of power input and
to afford the highest percentage of onstream reliability,
the modern precipitator is divided into a substantially
greater number of independent electrical sections than
was previously done.
The efficiency of a precipitator is a direct function of
the power input (Fig. 3). Any condition that adversely
affects power input should be avoided in the basic de-
sign of the precipitator. Proper alignment and stability
of the high-voltage system is essential.
Theoretically, the most efficient precipitator would
be one ia,which each individual discharge electrode has
its own power supply in order to maximize power in-
put. This is highly impractical. However, it is both
practical and advisable to have the precipitator di-
vided into a number of separately energized electrical
sections that can be individually isolated. This practice
not only allows to some extent for variations and strati-
fication in temperature, dust loadings and so forth, but
I Ml Ml' 41 I Mi)rxt>MNO JAM'ABk If fill
3.-5
-------
Power __
supply "
Electrical
section
y^i$|sl!iLL,'r
Gas flow
I
ffi r " r\ w i #7 '
fo r*~>~rn ^
'''' r~7~r^i n '
CM ,, 1 /i^ ' .. ?1Q
»5 . I /l ' #11
#3 I A"H I ftij
iL"L. , , • *^r - -^
Right
r— O
r— O
;
i O {
«, ... -. ^-^ i
L. *J
Wrortg
- ... : | i • '•
k&-SlJ '• ' ^ '--AJrb'i; JJtljJ
it renders a smaller section of the prccipitator vul-
nerable to external malfunctions such as dust-removal
problems. When such problems occur, one section can
generally be shut down for dust removal while other
sections remain onstream.
For example, today a typical 99%-efficient precipi-
tator (Fig. 4) on a 350,000-lb/h boiler would have six
separate electrical sections in series and two in parallel,
each with its own power supply. Ten years ago, such an
approach would not even have been considered. There
would have been three large sections.
In addition, it is common practice to provide some
redundancy so that outage of one or more sections does
not adversely impact on efficiency.
Automated controls
A relatively new field of development is the appli-
cation of sophisticated electronic measuring and con-
trol devices to precipitator control circuitry. Power is
now held at an optimum level automatically and de-
pendably, despite wide variations in gas and dust con-
ditions.
Component reliability
Keeping the precipitator on-line requires increased
reliability in the components used. To this end, re-
search and testing is directed constantly toward up-
grading existing components and developing new
hardware. This includes such areas as plate and dis-
charge-electrode designs, suspension systems, rapping
processes and mechanics.
A major impetus for research in all these areas is
matching component life to maintenance schedules or
turnaround so that unscheduled shutdowns due to
component breakdown can be avoided.
In addition, "fall-out" benefits are being realized.
During one study on rapping systems, for example,
cracking was noticed between rods and frames. Investi-
gation revealed that a solid rapper rod acted as a heat
sink during welding to the rapper frame and upset
welding heat balance. Subsequent tests on welding
techniques suggested a better way—hollow rapper rods
at the welded end. Now the welded surfaces of both rod
and frame respond equally to welding heat cycles. New
welds exhibit 85% more resistance to cracking than
those made with solid rods.
Factors affecting efficiency
In addition to the accumulation of data on precipi-
tator sizing and design, there has been a substantial ac-
cumulation of data on operation procedures and firing
practices that can cause the precipitator to lose effi-
ciency.
Gas volume— A precipitator is a volumetric device. For
example, any increase in boiler load that results in ex-
cessive flow through the precipitator will cause a loss of
efficiency. A precipitator designed for 3 ft/s face veloc-
ity and an efficiency of 99% will drop to 96.5% if the
velocity increases to 4 ft/s, a 33% load increase.
Temperature—A change in operating temperature may
also have an effect on precipitator efficiency. Particle
resistivity varies greatly in the temperature range of
200 to 400°F. Ignoring the effects of temperature on gas
volume, the impact of temperature on efficiency would
be, assuming 99% guarantee at 325°F (on a fly-ash ap-
plication):
Temperature, F
200
325
400
Efficiency,'
99.9+
99
99.5
Obviously, there is benefit to be derived in operating
below or above the 300 to 350°F level.
Fuel—Any significant change in the type of fuel being
fired will have an effect on the preci pita tor's perform-
ance. For example, a change from a 2%-sulfur bitu-
minous coal to a 0.5%-sulfur, subbituminous Western
coal can result in a design efficiency of 99.5% dropping
to 907« or less. Other chemical constituents, such as so-
dium oxide, in the ash can have an effect on perform-
ance by reducing bulk resistivity.
The unit should be designed for the worst expected
fuel.
Inlet loading—Since a precipitator is designed to re-
move a certain percentage (by weight) of the entering
i III Mil Al Mll.INKkl.INIt )AM A»y II, lv»
-------
material, ail things being equal, an increase of 50% at
the inlet will result in the same increase at the outlet.
Therefore, if an operating change involves an increase
in percentage of dust, a corresponding increase at the
outlet—resulting in greater opacity—can be expected.
Carbon—Variations in firing practice or coal pulveri-
zation that affect the quantity of combustible in the fly
ash also impact on precipitator performance. Carbo-
naceous materials readily take on an electrical charge
in a precipitator, but lose their charge quickly and are
readily reentrained. Not only is the carbon particle
very conductive, it is large and light compared with the
other constituents making up fly ash.
These are the major variables to be considered if a
deterioration in performance is to be avoided.
Age as a factor in performance
The question is often asked whether or not precipi-
tator performance deteriorates with age. The answer,
based on available operating experience, is "No;" how-
ever, there are two basic factors involved.
First, operating conditions that affect gas volume,
temperature, gas and dust composition and so forth
cannot be changed. Second, the precipitator must be
maintained properly to the extent that the internals re-
main in good alignment and are adequately cleaned by
the rapping system.
Meeting new clean-air standards
Assuming that precipitator performance has not de-
teriorated, there is still a major problem facing chem-
ical plant operators. That is the ever-increasing strin-
gency of clean-air codes, both new and revised, whether
on the local, state or federal level. Collection efficiency
requirements in the range of 99% and more are becom-
ing the norm. Few precipitators installed before 1970
are rated this high.
To meet the new requirements, the plant operator
has two choices. He can either upgrade existing equip-
ment, or replace it with new collection equipment.
Upgrading existing equipment
Nothing short of additional equipment will yield
99% performance where 90% is installed. However,
there are some areas where improvement can be ob-
tained.
Gas distribution—Before efficiency requirements went
to the 99% level, good gas distribution was not as criti-
cal as it is now. Therefore, it may be possible to im-
prove the efficiency of older units by improving the
flow pattern.
Ash conditioning—As mentioned earlier, a fuel change
can result in a low or high resistivity situation, which
can be controlled by injecting trace chemicals into the
gas stream to make the dust precipitable. Such condi-
tioning can have a markedly beneficial effect on collec-
tion efficiency.
Energization—Older precipitators are generally
equipped with fewer electrical energizing sets than cur-
rent practice dictates and may not be as responsive to
varying operating conditions. To this extent, efficiency
may be improved by increased electrical sectionaliza-
tion.
Gas flow
I, ,1
EielcLL
-
F.ifikL2.-
'"""
—
-
-
Fieid.3.
_
-Baffles
Right
Gas flow
Wrong
New equipment
The decision to install new collection equipment in
series with existing equipment, or to replace the exist-
ing equipment entirely, is affected by several considera-
tions. These include space, efficiency of the existing col-
lection devices, their condition, pressure drop as related
to possible need for a new fan and other operation con-
siderations, and, of course, cost.
For example, assume an old, low-efficiency mechani-
cal collector in poor condition is taking space where a
new electrostatic precipitator can go. If the mechanical
collector is left in place, the induced-draft fan cannot
handle the pressure drop across the new equipment. In
this instance, the mechanical collector should be re-
moved and replaced with the new equipment.
In another example, a 5-yr-old precipitator is in
good condition and operating at 95% efficiency. A new
i III MH .41 iM.INimi'w. JAM M.. a iv.
-------
clean-air code calk for 99% efficiency. Since there is
•pace downstream of the existing prccipitator, the most
feasible—and economical—solution is to add a new,
small precipitator in series with the existing one.
Retrofitting alternatives
In some instances an existing precipitator can be
made longer and higher so as to meet new require-
ments. However, this approach requires both a lengthy
outage and the space necessary to do the modification.
Another sometimes viable approach is to "double-
deck" a new precipitator over an existing unit. An al-
ternative to this approach, assuming space is available,
is to duct from the old precipitator to the new one and
then double back to the existing stack.
When the precipitator is located on a building top,
additional problems occur. In most cases, the building
cannot support the weight of additional or new (that is,
larger) equipment. This can be resolved by putting up
a new support structure that penetrates the building
roof and runs down to grade. An alternative solution is
to, where possible, locate the new unit at grade and
duct down to it.
Loss of efficiency occurs when gas bypasses the elec-
trostatic zone in a precipitator. This can occur between
the end plates and the shell, over the top of the electri-
cal fields or in the hoppers (Fig. 5) if proper design care
isn't taken.
As good flow control in a precipitator is achieved,
there is a marked increase in collection efficiency. Pre-
cipitators have gone, for example, from 96% to 99.5%
efficiency simply by corrections in flow control alone.
Proper use of perforated plates, turning vanes and baf-
fles is essential.
Complete and accurate information on fuel analysis
and ash chemical composition is essential. The pre-
ferred data are discrete analyses, rather than merely an
indication of expected ranges of constituents. With
these data, the sizing program can determine the worst-
case combination of constituents.
Work has also been done on cold-side sizing. Analy-
sis of performance data on cold-side units and pilot
precipitator work has indicated that there is significant
deviation from the performance predicted by the com-
monly used Deutsch-Anderson equation at efficiency
levels in excess of 98%. This led to the development of a
modified equation.
In a typical example, use of the original equation
would have resulted in a precipitator about -15%
smaller. In addition, the program requires an identi-
fication of the boiler type, coal, mass mean particle
size,* and sulfur and sodium oxide contents of the ash.
Based on these inputs, the necessary collection area is
indicated.
Obviously, as improvements in sizing technique de-
velop, improvements in other design areas are neces-
sary if precipitator efficiency is to be increased and
maintained.
Gas flow distribution—Higher efficiencies demand more
and more emphasis on the need for uniform gas flow.
Detailed laboratory model studies are often employed
to develop the most economical configuration for a new
*lfc4iurirt "f • fitfiiu Ir iif •vrmur rn«»
1012
0.5-1.0% Scoal
/
/
/
x
10"
1.5-2.0% Scoal
10'°
250 300 350
Temperature, "F
400
450
precipitator or to identify flow problems in existing in-
stallations. In many instances, a redistribution of gas
flow through the precipitator can increase collection ef-
ficiency by several percentage points.
Component reliability—The need for continuous on-
line availability has required constant attention to the
development of precipitator component reliability
without extensive increase in equipment costs.
As a matter of fact, developments in precipitator de-
sign have advanced to the point where buyers dealing
with established, reputable suppliers can be assured
that they are purchasing the system and component re-
liability they need.
Precipitator placement—Long-range studies of precipi-
t a tors in flue-gas streams indicate that gas temperature
can affect precipitator performance. In certain appli-
cations, precipitators have shown that high operating
temperatures are not only feasible, but economically
desirable. Hot-side application can eliminate the un-
certainty about dust resistivity that can result in dew-
point corrosion, poor precipitator performance and low
efficiency.
Ash conditioning—Either moisture or chemical treat-
ment of flue gas can alter the electrical properties of
dust particles and enhance their "precipitability." This
permits greater flexibility in the choice of fuel, and
more-efficient particulate collection over a wide range
of operating conditions.
3-8
|AM At I IJ l-fli
-------
The first preciphator iostaHations were cold-side op-
erations with flue-gas temperatures seldom exceeding
SOOT. In these installations, the electrical resistivity of
the ash is established by a surface-conduction mecha-
nism sensitive to the presence of minute quantities of
sulfur trioxide, sodium oxide and other hydrophilic
species, as well as to the partial pressure of water vapor
in the system. The quantities and interreactions of
these substances are not readily predictable. Therefore,
low-temperature resistivity is variable and unpredic-
table.
However, in recent years many installations have
been made ahead of the air preheater where tempera-
tures are in the 650 to 850T range. In these operations,
resistivity is established by bulk chemical analysis of
the ash, since conduction proceeds via a volume-con-
duction mechanism. It has been found that the predict-
ability of fly-ash resistivity is not only much more re-
liable at elevated temperatures, it is virtually certain to
be within a predictable range at 650 to 850 T.
Fig. 6 shows the effect of temperature on resistivity
for a given low-sulfur fuel.
Because of the accompanying decrease in gas density
at the higher temperatures, requirements in the corona
starting potential in hot-side precipitators is greatly re-
duced. In addition, the elevated temperatures have
been found to reduce sparkover level. The net result is a
more efficient use of the electrical power input to the
precipitator.
The final area of significant difference between hot-
and cold-side operation is in the physical properties of
the fly ash as related first to removal from the elec-
trodes by rapping and then to removal from the precip-
itator dust hopper. Experience has shown that fly ash
from a cold-side, high-resistivity operation is more ad-
hesive than hot-side ash. To keep the electrodes rela-
tively clear of this insulating layer usually requires
more intense and more frequent rapping. This, in turn,
limits gas velocity to about 4 ft/s to prevent exce.ssive
reentrainment of the fly ash.
Conversely, at elevated temperatures it is common-
place today to use gas velocities in the 5 to 5.5 ft/s
range.
Removal of collected ash from hoppers has some-
times been a difficult and troublesome chore on cold-
side installations. Hopper plugging not only reduces
precipitator efficiency but, also, can cause serious dam-
age to the internals. This may include distortion of
lower high-tension framework, bowing discharge elec-
trodes and accelerating failures. Moreover, ash buildup
in the hoppers increases possibility of dust reen-
trainment and loss of efficiency.
Since the fly ash in high-temperature operations is
almost fluid, and the gas temperature is far above the
dew point, there have been essentially no problems re-
ported with ash handling.
Ash conditioning
To combat fuel-supply uncertainties and shortages,
many plants are burning lower-sulfur fuels. These low-
sulfur fuels may generate more ash and gas per Btu,
and the ash has an electrical resistivity several orders of
magnitude greater than thai from higher-sulfur coals.
1
I
(3) Perforated
distributor plates
Right
(1) Perforated
distributor plate-v
Wrong
13
This, too, will affect precipitator operating efficiency.
For precipitators to perform at the required degree of'
efficiency, these ash particles can be treated chemically
to enhance their charge retention.
To accomplish this, small quantities of SO;t are in-
jected into the flue gas. This reduces the electrical re-
sistivity of the fly ash, making the dust more amenable
to collection in the precipitator.
By improving the electrical operation, the gas-condi-
tioning system can significantly improve precipitator
performance and minimize the need to enlarge the pre-
cipitator for low-sulfur fuels.
Such a system can also be used as an "efficiency"
backup system on high-sulfur-fuel operations at a re-
duced capital expenditure.
There are several types of commercial SO;< gas-con-
ditioning systems: direct injection or evaporation of liq-
uid SO,; catalytic conversion of SO3; vaporization of
M IV.IMJfc'V, |\M Ml \
-------
^^y^^^^^^fe^r- _ -^
Capital jnvflitmont.
thousand dollars
Process
Molten sulfur
Liquid SO2
Liquid SO3
250 MW 500 MW
373
320
281
Sulfuric-acid evaporation 650
'Based on condition
ng of a low-sulfur Westem-coa
565
512
441
1,020
flue gas to an
tYlb SO3
250 MW
4.70
8.56
8.92
7.38
optimum resistivity
Operating costs*
500 MW
3.76
7.73
8.00
6.16
level estin
rf/kWh
250 MW
0.008
0.0146
0.0152
0.0123
nated at 55 ppm SO-j.
500 MW
0.006
0.0132
0.0136
0.0103
sulfuric acid; and sulfur burning followed by the cata-
lytic conversion of SO2 to SOs.
SOy injection—The simplest system to design and
build is a liquid-SO3 conditioning system. Liquid SO3
is readily available in commercial quantities and is
clear, colorless, stable, not particularly corrosive, and
has a fairly low vapor-pressure. However, it is highly
toxic and requires reasonable safeguards. SO3 is highly
hygroscopic and, when dispersed in air, immediately
forms an extremely dangerous sulfuric acid mist.
In operation, liquid SO.-j is first metered into a vapor-
izer, then air-diluted. This maintains a constant con-
trollable volume of gas flowing through the injection
manifold (which provides adequate dispersion in the
flue). The mixture is conveyed in heated lines to the in-
jection point to prevent H2SO4 condensation and cor-
rosion.
Acid vaporization—In this system, sulfuric acid is
heated above its boiling point, vaporized, diluted with
air and then injected into the flue ahead of the precipi-
tator. Water vapor is always present in the acid vapor-
ization system, so a heating system is necessary to keep
gas temperature above the dew point. In addition, the
manifold inside the flue must be insulated to prevent
corrosion and premature condensation.
Gas distribution
As recently as 10 to 15 years ago, fully one-third of
the particles treated in the precipitator were treated
twice because of reentrainment caused by improper gas
flow within the precipitator. Because of this, precipi-
tators were usually sized larger than would otherwise
be necessary.
However, increasingly stringent collection-efficiency
requirements have tightened the criteria for what con-
stitutes "good" gas distribution. The very high collec-
tion-efficiency levels currently required are realized
with a gas-distribution pattern that permits maximum
utilization of the gas-treatment zone.
Nothing will downgrade overall precipitator per-
formance as thoroughly as maldistribution of flue
gases. Careful attention to design of the flue leading to
and from the precipitator (as shown in Fig. 7) can
create a more uniform overall gas flow within the pre-
cipitator.
Dust reentrainment from the hoppers due to im-
proper gas flow is a frequent cause of performance defi-
ciency in high-efficiency precipitators. Variable-poros-
ity distribution plates, fitted at the precipitator outlet,
and proper baffling inhibit those pressure gradients
that would normally promote hopper sweepage. The
constricting device changes resistance across the gas
stream and corrects what is considered to be a little-
known and often-misunderstood flow phenomenon.
Catalytic conversion—In the catalytic conversion of SO2
to SO t, liquid SO2 is vaporized in a steam-heated va-
porizer. This vapor is then mixed with enough air to
produce a mixture containing approximately 8% SO2
by volume. This mixture is heated in an electric heater
to about 840°F, and fed to a single-stage converter.
About 70 to 75% of the SO2 can be converted to SO3
and injected into the flue gas.
Sulfur burning—In this system, molten sulfur is
pumped' from a storage tank to the sulfur burner. Liq-
uid sulfur is atomized with .high-velocity air and com-
pletely burned to SO2 in the combustion chamber. The
effluent SC>2—air mixture flows out of the sulfur burner
at about 1,600°F and, after cooling to 650CF in an air
cooler, is converted catalytically in a one-stage bed of
vanadium oxide to SO3. Conversion efficiency is about
72%. The dilute SO3 gas, at 1,120'F, is then trans-
ported to the precipitator distribution manifold for gas-
conditioning purposes.
Table II shows an economic comparison of the pro-
cesses.
In addition to economics, several other factors favor
the sulfur-burning system over the others. The inert
sulfur is easily handled and stored. Furthermore, the
only location of a corrosive or hazardous problem is
upstream of the catalytic converter.
The author
Knhrrl Hump ix product manager for
industrial prmpitators at Research
Cmtrrll Inc.. Box 750, Bound Brook, NJ
OH8(5.
In this position, he analyzes customers*
nerds and provides technical backup.
Hr luis had 30 yean of experience with
prei'ipitators and related equipment.
Before joining Research Coltrell, he
held several engineering positions in the
field of air-quality control. He holds
degrees in engineering and economics.
i III Ml* .\l I •
I V. JAM AMV
-------
V-4.
ELECTROSTATIC PRECIPITATOR MAINTENANCE SURVEY
APCA TC-1 Particulate Committee
Copyright Qc) 1976 by the Air Pollution Control
Association. Reprinted with permission from the
Journal of the Air Pollution Control Association,
Vol. 26, No. 11.
4-1
-------
Electrostatic Precipitator Maintenance Survey
APCA TC-1 Particulale Committee
Principal Author: Robert L. Bump
In the latter part of 1974, the TC-1 Committee ol the APCA under-
took a survey of four (4) major user Industries of electrostatic pre-
clpltators. The purpose of the survey was to establish the users' de-
gree of satisfaction with this equipment from an operational and a
maintenance viewpoint. Specific areas of maintenance requirements
were Investigated as well as the nature of difficulty experienced. The
174 responses received covered user experience wHh 243 preclpl-
taiors of various manufacturers. This paper reports on the results of
thto survey and gives the statistics derived. The conclusions reached
should be beneficial to the user and the manufacturer In the areas
of product Improvement and maintenance.
TMs Is the first of this type of survey handled by the TC-1 Com-
mtttee, and It will be followed by Fabric Filter and Scrubber Infor-
mation gathering questionnaires. Problem areas so defined will be
enlarged upon In future seminars planned under APCA auspices.
The Federal Clean Air Act began an era of increased public
awareness of the importance of protecting our environment.
Individual states and municipalities have implemented and
policed the laws which have been promulgated regarding
permissible emission from various processes. In addition to
the ability of a specific control device to limit emissions to the
prescribed level, it is also obvious that it is important that the
control be consistently effective. There would be little benefit
from guaranteed performance during the efficiency tests and
long periods of sub-standard operation due to equipment
malfunction thereafter.
1 n order to assess the experience of the major user industries
in recent years regarding actual operation and maintenance.
the APCA TC-1 Committee embarked upon a survey in 1974.
This survey was confined to electrostatic precipitators and
four major industries were canvassed—electric utilities, ce-
ment, paper, and metallurgical. The objective was to obtain
constructive input relative to actual experience in general and
in certain specific areas. This information would provide the
basis for the generation of statistics and recommendations
from consulting engineers and manufacturers. In addition, it
provides users with an insight into whether their experience
is unique or shared by others in their own and other indus-
tries.
A quick reference to a typical electrostatic precipitator will
show the major components, i.e., collecting surfaces on which
the material is precipitated; discharge electrodes which create
the high voltage, uni-directional field; rapping system which
dislodges the collected material; suspension insulators which
support and insulate the high voltage system; and dust re-
moval which consists of hoppers, drag scrapers or a hydraulic
mode of evacuating the dust. A malfunction in any one of these
areas usually results in outage of a portion of the precipita-
tor.
The actual survey format used is shown in Figure 1. Effort
was made to keep the form as simple, yet meaningful, as
possible and, at the same time, to cover the areas which are
commonly acknowledged to be troublesome. Approximately
400 forms were mailed and 174 responses were received
(43.5%) which pertained to 243 precipitators of various man-
ufacturers. Sixty-three electric utilities reported on 88 pre-
cipitators; 53 cement plants reported on 70 precipitators; 36
paper mills reported on 49 units; and 22 metallurgical pro-
cesses reported on 36 precipitators. The equipment reported
on has been in service from a minimum of 3 months to a
maximum of 50 years. Average service life was in the 7 to 10
year range. It will be noted that percentages do not add up to
100% in most cases. This reflects the failure of respondents
to address every question. The results are based on total re-
sponse to a specific'question as a percentage of total respon-
dents from the industry.
November 1976 Volume 26. No. 11
4-2
1061
-------
Figure 1.
Company:
Location:
Application:
Years of Service:
'Equipment Manufacturer:
Over-All Experience with the Precipitator:
Precipitator maintenance survey form.
Temperature:
Operation:
Maintenance:
Excellent
Excellent
Good
Good
Fair
Fair
Poor
Poor
Areas of Maintenance Requirement
A. Discharge Electrode Failures
Frequency:
Type Failure:
Magnitude of Problem:
Type of Electrode:
Frequent
Fatigue
Major
Weighted
Material
Shape: Round
Infrequent
Corrosion
Minor
Rigid
Size _
Square
Very Seldom
Arcing
No Problem
Barbed
B. Rappers or Vibrator Failures
Frequency:
Type Failure:
Magnitude of Problem:
Type of Rapper:
Adequacy of Cleaning:
Frequent
Major
Pneumatic Vibrator
Electric Vibrator
Electric Impulse
Mechanical Rappers
Good Fair
Infrequent
Minor
Poor
Make.
Make
Make .
Make.
Very Seldom
No Problem
Collecting Plate Failures
Frequency:
Type of Failure:
Magnitude of Problem:
Type of Plate
Frequent Infrequent
Connection Points Weld
Major Minor
Roll Formed Welded
Very Seldom
Corrosion
No Problem
Material
Gauge
D. Dust Removal System Failures
Frequency: Frequent Infrequent
Type of Failure: Screw Conveyors Dust Valves
Magnitude of Problem: Major Minor
Type of Dust Removal: Screw Conveyors Pneumatic
Very Seldom
Pluggage
No Problem-
Hydraulic
E. Insulator Failures
Frequency: Frequent Infrequent Very Seldom
Type of Failure (Cause):
Magnitude of Problem: Major Minor No Problem
Type of Insulator: Porcelain Silica Alumina
Whtt do you consider to be your major precipitator maintenance problem both from the reliability and the expense point
of view. A, B, C. D or E.
Other Precipitalor Maintenance Problems:
*If you have equipment of more than one manufacturer, list the names here and designate them as A, B, C, etc. in filling
out the data.
1062
4-3
Journal of the Air Pollution Control Associatir
-------
The first point of inquiry had to do with overall experience
with the equipment from an operational and maintenance
viewpoint. The results are shown in Tables I and II. The re-
sults on an individual industry basis were reasonably consis-
tent one to the other. It did emerge, however, that the paper
and cement industries consider maintenance to be a greater
problem than the others.
Table I. Operation of prccipitalors.
Table IV. K;ippcr/vibraliir failure frequency.
Industry
Utilities
Cement
Paper
Metallurgical
Average %
Excellent
14.8%
15.7
18.4
16.7
16.0
Good
45.5%
52.9
55.1
63.9
52.3
Fair
29.5%
25.7
16.3
5.6
22.3
Poor
10.2%
5.7
8.2
5.6
7.8
Table II. Prccipitator maintenance.
Industry
Utilities
Cement
Paper
Metallurgical
Average %
Excellent
13.6%
5.8
8.2
2.8
8.6
Good
52.3%
50.0
46.9
69.4
53.1
Fair
1 3.6%
37.1
36.7
19.4
25.9
Poor
20.5%
5.7
6.1
0
10.3
The survey then dealt with specific areas of potential dif-
ficulty. The first had to do with failure of discharge electrodes,
with results presented in Table III. Of the three major types
of failure normally experienced (fatigue, corrosion, or elec-
trical arcing), 61.7% indicated that electrical erosion (arcing)
was the principal cause of failure. Corrosion and fatigue fail-
ures ranked second and third.
Table 111. Discharge electrode failure frequency.
Industry
Utilities
Cement
Paper
Metallurgical
Average %
Frequent
29.5%
25.7
16.3
5.6
22.2
Major
22.6%
Infrequent Very seldom
38.6%
47.1
44.9
58.3
45.3
Magnitude of problem
Minor No
53.1%
28.4%
25.7
30.6
33.3
28.8
problem
21.0%
Failures in the precipitator rapping system were the next
'point of interest. These systems are normally electric or
pneumatic vibrators or electromagnetic or mechanical impact
type rappers. Referring to Table IV. As would be expected,
the data indicated that the vibratory type of cleaning mech-
anism, whether pneumatic or electric, is a higher maintenance
item than the impulse type.
Collecting surfaces were the next point of interest. These
are normally fabricated or roll formed of 18-20 gauge material,
24-36 ft high, suspended at the top and guided at the bottom.
Industry
Utilities
Cement
Paper
Metallurgical
Average %
Frequent
0.1%
31.4
26.5
30.6
22.2
Major
10.3%
Good
58.4%
Infrequent
Very seldom
38.6% 47.7%
35.8 31.4
34.7 26.5
30.6 ll.l
35.8 33.3
Magnitude of problem
Minor
No problem
53.1% 28.4%
Adequacy of cleaning
Fair
32.1%
Poor
5.8%
Regarding problems of collecting surface origin, the poll in-
dicated (Table V) the major cause of collecting surface failure
cited was fatigue at the points of plate suspension. Corrosion
was cited as the second major cause.
Removal of dust, once precipitated, has historically been
one of the major causes of precipitator malfunction, as well
as a contributory factor to other maintenance requirements
such as discharge electrode failure. The survey indicated, re-
ferring to Table VI, by far the majority of the problems ex-
perienced were with dust hopper pluggage. Screw conveyors
and dust valves were ranked second and third.
Suspension insulators, manufactured of glazed porcelain,
fused silica, or alumina oxide are used to support and isolate
the high voltage elements of a precipitator. These insulators
are somewhat vulnerable to failure due to electrical arc over
Table V. Collecting plate failure frequency.
Industry
Utilities
Cement
Paper
Metallurgical
Average %
Table VI. Dust
Industry
Utilities
Cement
Dry paper
Wet paper
Metallurgical
Average %
Frequent
4:5%
7.1
16.3
19.4
9.9
Major
17.3%
removal system
Frequent
36.4%
27.1
25.0
2.4
16.7
24.7
Major
24.7%
Infrequent
Very seldom
7.9% 68.2%
14.3 64.3
32.7 42.9
11.1 33.3
15.2 56.8
Magnitude of problem
Minor
32.1%
failure frequency-
Infrequent
No problem
45.3%
Very seldom
42.0% 20.%
40.0 27.1
50.0 1 2.5
9.8 63.4
38.9 25.0
35.8 30.0
Magnitude of problem
Minor
49.0%
No problem
17.3%
November 1976 Volume 26. No. 11
4-4
1063
-------
Table VII. Insulator failure frequency.
Industry l;ret|ucn( Infrequent Very sclle VIII. Major maintenance problems.
Collect- Dust
Discharge Rapper/ ing removal
electrode's vibrator plates systems Insulators
Utilities
Cement
Paper
Metallurgical
Average %
35.2%
34.3
20.4
25.0
30.5
5:7%
25.7
34.7
33.3
21.4
13.6%
11.4
16.3
22.2
14.8
31.8%
34.3
4.1
19.4
25.1
1 . 1%
1.4
6.1
25.0
5.8
2. Discharge electrodes are the principal source of malfunction
and the area where design expertise should be directed. This
point seems to l>e recognized by the manufacturers and there
is evidence that design improvements and developments are
in progress.
3. Careful attention to the design, operation, and maintenance
of the dust removal system is extremely important. It is sig-
nificant to note that the industry which reported the highest
incidence of discharge electrode failure also reported the
highest degree of hopper pluggage. Dust build-up into the high
voltage system, in addition to inhibiting efficient performance,
can cause serious damage and accelerated discharge electrode
failure.
The TC-1 Committee of the APCA suggests that close co-
operation between user and supplier, coupled with an ex-
change of information between the various user industries,
will ultimately result in the mutual development of an elec-
trostatic precipitator which fills the needs of all concerned
Finally, it should be pointed out that this survey is only the
beginning of a comprehensive study of experience with high
efficiency collectors of various types. Moreover, the scope of
data presented herein is somewhat preliminary in that con-
siderably more detail can be derived statistically from the
information received. The TC-1 Committee will continue to
work with this data base and will report additional findings,
conclusions, and recommendations at a future time.
Conclusions
There are several conclusions which may be drawn from the
results of this survey:
1. Although there is obviously room for improvement on the
part of precipitator manufacturers, the majority of the users
are satisfied with the precipitator as a functioning piece of
equipment. Only 7.8% gave a "Poor" rating.
Mr. Hump is Product Manager, Uesearch-Cottrell, P. O.
Box 750, Bound Brook, N.I 08805. This is a survey report of
the AI'CA TC-1 Particulale Committee of which Jacob Katz
is chairman. Mr. Katz' address is 4525 Main Street, Munhall,
I'A-15120.
1064
4-5
Journal of the Air Pollution Control Assoclatioi
-------
V-5 .
MAINTENANCE PROGRAM AND PROCEDURES
TO OPTIMIZE ELECTROSTATIC PRECIPITATORS
by
J. Katz
Precipitator Technology, Inc.
Copyright (c) 1975 by the Institute of Electrical and
Electronics Engineers, Inc. Reprinted with permission
from IEEE Transactions on Industry Application, Vol. IA-11,
No. 6.
5-1
-------
674
IKKK T« \SSACTIONS ON' INDI'KTHY APPI.IC XTIONH, VOL. IA-11, NO. Ct, NOVEMBEH/DECCMBKK 107ft
Maintenance Program and Procedures to Optimize
Electrostatic Precipitators
JACOB KATZ
Abstract—The electrostatic precipitator has proven to be a highly
ruccessful piece of equipment for the removal of suspended particu-
Ute matter in the gas streams discharged from cement kilns.
Whether the high collection efficiency of the precipitator will be
maiztaiced on a continuous basis can often be related to proper
maintenance procedures. This paper is primarily offered as a non-
technical coverage of some problems that may face the maintenance
department of a cement plant after the installation of the precipitator.
Whether these problems actually arise are subject to many variables,
including the thought and care given to reliability features in the
original design. Discussion is presented to facilitate a better under-
standing of how precipitator design and kiln operating parameters
can affect maintenance procedures. I
PRECIPITATOR APPLICATION l-'OK
CEMENT
General
THE MAJOR application fur
in the cement industry is
from the gases leaving the feed
producing kiln. Both wet :ind dry process rotary kilns have
. utilized precipitators successfully.
Early cement kiln precipitate
vILNS
electrostatic precipitators
to collect the particulate
or back end of the cement
rs, prior to the recent need
to meet tight environmental standards, were primarily
designed for 90 percent to 9*i percent collection efficiencies.
This level of collection simulated the philosophy of pre-
cipitator application for other industries as well. Un-
fortunately, this level of design efficiency usually left
little performance margin, so that the failure of single
components often led to unsatisfactory stack appearances.
The maintenance department was, therefore, under con-
tinuous pressure to keep all components in service.
Several points of interest are listed.
1) Low efficiency precipitators can increase the main-
tenance burden.
2) Physical size of the precipitator (closely related to
efficiency levels) is of less concern to maintenance personnel
than the integrity of its component parts.
3) As precipitators increase in size to attain greater per-
formance levels, the need for integrity of component parts
increases at a much greater rate.
Paper TOD-74-110. approi'ed by the Cornell Industry Com-
mittee of the IEEE Industry Applications Siciety for presentation
at the r.'74 IEEE Cement Industry Technical Conference. Mexico
City. Mexico, M.ty 13-16. Maiium-ript released for publication
Deremtwr 30, 11174.
The author is with Preripitator TerhnoloKy, Inc., Munholl, Pa.
15120.
5-2
These points, while appearing to be in conflict, actually
stress the vulnerability of the large new precipitators to
maintenance difii» ulties. unless extreme care is taken in the
design of the component parts.
Clinker Production
The production of Portland Cement utilizes an assort-
ment of raw materials which are dried, calcined, and
formed into a clinker inside a rotating kiln. Wliile the
actual raw mix may vary greatly, the main composition
consists of calcium carbonate, silica, and alumina, or nhale
material.
The major difference between the wet and dry processes
lies in the method of grinding the raw materials, and the
state of the feed that enters the kiln. The wet process
grinds the raw materials in a wet state, and the slurry
feed consists of 33 percent to 44 percent .by weight of
water. In the dry process, the raw materials are dried
before grinding, and the resultant product is fed into the
back end of the kiln in a nearly dry state.
Process Variables Important to Precipitation
The chemical reactions and theories about what occurs
inside the kiln is beyond the scope of this paper. We arc
concerned, however, with the composition of the waste-
gases and particulale matter that enter the precipitator.
The following process variables are considered important
to precipitator performance and will be discussed in greater
detail with regard to how they affect maintenance prob-
lems:
1) concentration of particulate in kiln gases;
2) size distribution of waste dust;
3) moisture content of kiln gases;
. 4) gas temperature of kiln gases;
*>) alkalie and chloride content of particulate matter in
kiln gases.
THEORIES AND DESIGN OF PUECIPITATORS
General
There are countless differences in the desipi and hard-
ware among the various vendors of electrostatic precipita-
tors and this writer n?sumes that the reader is .familiar
with the physical components of u precipitator. Without
the use of mathematical equations the precipitator will
generally increase in collection performance under the
following conditions.
-------
F.I.KCTIlf>ST.XTir PUKClfrTATOHrf
r.?:.
}) When the area of tin- collect ing surface is increased in
relation to the gas (low rate or c|iiantity of waste gas pas-
sing through the collector per Riven unit of time, the
coll u performance will improve.
1, ,nversely, for any given size precipitator, reduction
of gas flow rate is beneficial.
3) An increase in the physical size of the individual
particles passing through the precipitator is beneficial.
4) A decrease in the viscosity of the waste gases will
generally increase the collection of particulate matter if all
other factors are equal. Viscosity will decrease with a re-
duction of gas temperature.
.')) A minor increase of the electric Held in the precipitator
can often materially improve collection performance. The
voltage fields provide the driving force for movement of
particles toward the collecting surfaces.
Effect of Precipitator Size for Ceinenl Kiln Applications
Large prccipitators are synonymous with high collection
efficiency. The cross sectional area is generally large enough
to obtain internal gas velocities as low as 3 ft/s for the
Jl9 percent plus levels of performance. Several factors are
noted for maintenance concern.
1) With very large precipitator designs, the inlet fields
will collect a large percentage of du>t. and difficulties in
hopper evacuation may result .'This is especially important
with installations having high inlet loadings—over 5
grains actual cubic feet.
2) Dust buildup c:tn readily occur in duct work and
emr Chambers directly ahead of the precipitator as the
pas . icity decreases rapidly to levels that exist within
the collector.
3) With the large precipitator, maintenance personnel
must identify and define repetitive problems and solve
them quickly. As stated earlier, large numbers of compo-
nent parts can provide a severe maintenance headache if
troubles cause a snowballing effect.
Poirtr Input to the Precipitator
The precipitator collects material from the gas stream
by using electrical forces. The object is to increase the po-
tential difference betwtvn electrode systems in order to
maintain a high charging condition. A limiting factor is
the electrical breakdown of the dielectric or gas spnce
(called a sparkover) which then becomes a localized path
of high current flow at one point in the electrode system.
For reference, the energization of a precipitator can be
described briefly.
1) Transformation of a low voltage supply to suitable
high voltage-alternating current. (Seel'ig. 1.)
. 2) Rectification of this high voltage ac into a pulsating
dc.
3) Connection of this output of the rectifier is made to an
internal precipitator electrode system that is insulated
from ground. This part of the precipitator is at a negative
higi ential and is now the source of energization.
•i .10 high potential on the negative electrode (which
could be a win* with a minimum diameter of 0.10 in)
PREeiPtTATOR
(a)
PRECIPITATOR
(-) PROPER
RETURN^ tt
(b)
FIR. 1. Two possible rectifier circuits for electrostatic ]>recipitalors.
(a) Full-wave circuit schematic, (b) Double half-wave circuit
schematic.
causes a localized corona breakdown of the gas surrounding
the electrode.
5) This corona sheath supplies the flow of electrons
(either separately or more generally attached to gas
molecules) through the gas space toward the passive
electrode or collector plate which is at ground or positive
potential. It is this electrical charging of particles by
negative ion attachment that provides the transport
mechanism for the particles to move through the gas
stream to the collecting surfaces under the influence of the
electric field. The distance between the 2 electrode systems
can be 4 to G in.
G) The grounded electrode (collector plates), which also
acts as the primary depositor of material collected, returns
the total precipitator current to the rectifier so as to
complete the electrical circuit. (See l-'ig. 1.)
The voltage transformers are especially designed for
precipitator sendee with the ability to withstand winding
stress when a severe sparkover occurs inside the collector.
Ratings of transformers normally range from lii to 120
kVA, with 45 000 to of> 000 secondary winding voltages
and secondary output load currents to 2200 niA dc.
Modern rectifiers arc of the silicon typo and are usually
mounted inside the transformer tank. Conversion kits ure
available to modify earlier installations containing recti-
fiers of the vacuum tube or mechanical type. One recent
improvement in the control of power input has been the
silicon-controlled rectifier (SCR) automatic voltage con-
trol circuit which minimizes the effect of elcetricnl dis-
turbances in the precipitator.
Important items of the control circuitry are the meter*
which monitor the variations in the electrical power input.
-------
r,7r,
IKKB TRANSACTIONS OX IXDCSTItV APPLICATIONS, XOVEilBER/DECEMBEK I07.1,
NOTE EFFECT OF
SMALL INCREASE OF
VOLTAGE ON EFFICWCY
INCREASE IN KILOVOLTS
TWICAL RANGE 36 TO 60 KV
•ie 2. Prerii.itator peak voltac*. A lypirn! electnwtntir pre-
cipitai'T peak voltage versus collwiion rll.ciency eurvi-rorona
r ver.-us efficiency curve woulil f«ll«»w wmilar pattern.
The most commonly used meters an*
1) voltage across the primary winding of the transformer
in ac volts;
•J) current in the primary winding <>f the transformer
in ac amperes;
3) ground return current from tin- prccipitator in dc
amperes, which is usually designated as milliamprres
in the smaller power supplies.
Two other meters sometimes found on control panels are
4) precipitator voltage through an appropriate voltage
divider and noted as average de kilovolts;
3) sparkmetcr which integrates tin- electrical break-
downs in the precipitator as average sparks per
minute.
The power supply must be matched correctly for the
precipitator section or service exported, or several of the
following difficulties can arise.
li'The impedance of the power supply, including a
ballast resistance or reactor in the primary winding of the
transformer, may not be sufficient to dampen the severity
of the electrical breakdowns in the preoipitator. This con-
dition is especially likely to occur if t he *x>wer supply ruing
is much larger than the operating power level.
'1\ If the physical size of the proeipitator is too great for
the size of the power supply, then the inherently lower
precipitator voltages may cause decreased performance.
S) The gas and paniculate conditions can drastically
alter the voltage-current relationship ami produce iowtT
voltage tields than expected because of a limitation iv. the
rating of t ho power supply.
High levels of voltage and useful corona powo.r in th<>
prccipitator, nil other conditions equal, will produce
high collection efficiencies. Fig. 2 shows a typical perform-
ance curve of the effect on efficiency by changes in the peak
voltage of a prccipitator. This simple curve can onlv
represent one situation because each prccipitator will have
its own characteristic curve based on many factors. The
important point to remember is that small changes in
voltage can produce substantial changes in power, hence,
changes in the efficiency of the collector, especially at the
lower levels of power input.
ELECTRICAL CHARACTERISTICS OF
CEMENT KILN PRECIPITATORS
Design Characteristics
Every cement kiln precipitator will have specific elec-
trical characteristics before the onset of gas and dust
conditions. The design characteristics are primarily n>-
lated to the following factors:
1) the distance between the high voltage electrode and
collector surface;
2) high voltage electrode design (especially corona
producing surface);
3) spacing of high voltage electrodes;
4) collector surface design;
5) power supply characteristics;
G) collector surface and electrode length related to
power supply.
An important set of data which constitutes the elec-
trical characteristics of a clean precipitator is obtained
prior to the initial operation under gas and participate
conditions. Thi.s data, called air-load readings, represent
precipitator sections energized without air movement und
at ambient temperatures. It is best to energize the smallest
possible electrical section. Sparkover should not be pn-scnt
at the maximum rated input of power.
Air load readings for each section provide a maintenance
tool for use after each outage, especially after internal
work or cleaning has occurred. Comparison of readings on
identical electrical sections can isolate areas of trouble
over a long term period.
Gas Conditions and Electrical Pourr Input
The electrical characteristics of a cement kiln precipita-
tor under normal process conditions can show a variety of
meter readings. These readings, like the air load readings, .
should be recorded periodically to help ascertain malfunc-
tions and process changes. Cement kiln 'prccipUator* will
generally exhibit minimal sparkovers ps|M*ci»lIy for wet
process applications. Usually the dry process pn-cipitator
will have sparkovers in the inlet fields or sections, nnd in
succeeding sections as well, if moisture contents below
12 to 15 percent by volume exist in the ga* stream.
-------
: ELECTRO«T\T!C PRECII'ITtTORS
To aid maintenance personnel in evaluating changes in
pjrocipitator electrical readings, the following general
jjuir*~"\ PS should help.
1; .iC higher the gas temperature entering the prccipita-
tor, the lower the voltage will appear compared to the
eJectrical current reading.
2) When the moisture content of the kiln gases increases,
for'any given condition, the voltage will also tend to
increase in value.
3) If reduced voltage exists because of sparkovcr, a rise
in moisture may allow for an increase in the prccipitator
voltage level.
4) An increase in the concentration of particulate will
tend to elevate voltages and reduce current flow.
5) A decrease in the particle size of particulate entering
the precipitator will tend to raise voltage while suppressing
oarrent flow.
6) A higher gas velocity through the precipitator will
tend to raise voltages and depress currents.
7) Air inlcakage may cause sparkover in localized areas
resulting in reduced voltages.
S) Under gas conditions, a number of precipitator fields
in scries energized by individual-power supplies will show
varying electrical reading?. The voltage-current ratio will
decrease in the direction of gas flow.
MAINTENANCE EFFECTS AND
ELECTRICAL READINGS
' operation factors that have been discussed earlier
did not take into consideration the maintenance variables
that can complicate the analysis of the voltage-current
observations. This section will combine some of these
factors and show the cause and effect relationship for the
purpose of maintenance aid. —
Alignment of Electrodes
The net effect on the electrical readings by the char-
acteristics of the gas and particulate will often be contin-
gent upon the proper spacing or alignment between the
electrodes. Meter readings may indicate one effect, while
n close spacing or even a specific electrode- design will
cause a spark-sensitive precipitator. The comparison of
similar electrical sections in a multiunit precipitator is
useful to identify internal defects. There is no substitute
for careful internal measurements and inspections.
The degree of misalignment allowable in » prccipitatur
is dependent on a number of factors. It is. however, recom-
mended that deviations of the high voltage electrode? be
kept less than J in from the center of a H-in gas passage.
One misaligned discharge electrode (pos>ibly 1 in off
center) .can control the power input vi an electrical section
containing hundreds' of parallel electrodes. This particular
electrode could cause electrical breakdowns of tin- section
atr ' -ced voltages. With proper automatic voltagi- control
air .pi-dance, this condition might continue for many
months. Without proper spturkovfr control, the high spark
rate will produce metal erosion and consequently result
in electrode damage and failure. °
Discharge Electrode Buildup
While the negative electrode is the source of energisa-
tion, it can also develop a coating of particulate. It is not
unusual to observe a 1- to 3-in diameter buildup on certain
wires, for example, and at various sections of the electrode.
Large wire buildups may or may not be critical on the
electrical characteristics dependent on the nature of the
buildup. For example, a nonuniform buildup in the shape
of beads over the total length of wire will not basically
alter the meter readings or precipitator performance. If
however, a uniform buildup of fine particulate coats the
electrode so that it simulates a wire of larger diameter, then
higher voltages may be required to initiate corona current.
This condition can lead to a voltage sensitive precipitator
in which sparkover occurs, because the voltage gradient is
too great for the physical spacing.
Collector Electrode Buildup
Partieulate buildup on the collector electrode or plate
will substantially affect electrical readings. A thin resistive
layer, can. cause a spark-sensitive precipitator. If not
resistive, material buildups of 1- to 2-in thicknesses can
occur without an electrical breakdown of the remaining
space.
Since the grounded collector plate acts as a receptor for
the precipitator current, a particulate layer made up of
particles that resist a passage of electrons will cause n
resistive- sheath to form. This layer will tend to reduce
current flow (consider this an increase of dielectric be-
tween the electrodes) and can raise the indicated voltage on
thr meter. The term used to describe the ability of the
dust layer to pass current is called resistivity. Measure-
ments are sometimes made to help determine the resistiv-
ity value of particulate usually expressed in ohm-centi-
meters. The use of this number as a sole criterion of resistiv-
ity can provide erroneous conclusions because many other
factors also affect the sparkover characteristic of a precipi-
tator section. It is well to know, however, that the oc-
currence of sparkover. which causes lower power inputs,
can be overcome by methods of process modification or gas
conditioning.
It is practically impossible to operate with collator
electrodes in a completely clean condition. Normally § in
to i in buildups will be found on the collector surface.
Unfortunately, the higher resistive material will tend to
have better cohesive forces holding the particles together
on the collector. The ease where thick buildups occur cam
also produce changes in the voltage-current readings.
Even though the voltage gradient may be similar for wider
spaci-s, the increase in gas velocity through the narrowed
passage will decrease the overall collector efficiency. The
indicated voltage for the electrical section will be lowered
by the narrowed passages.
5-5
-------
f.TS
TRANSACTIONS ON INDt'KTRT APPLICATIONS, NnVF.MBEH/DF.CEMbtJl 107')
SPARKOVEH CHARACTERISTICS
The level of sparkovcr can affect the performance of the
precipitator cither by quantity or severity. Usually
minimal sparkovers occur on the latter (discharge end)
sections of the precipitator. The sparkover should be
extinguished quickly and especially not extend into a
power arc of several cycles. Whether or not this condition
can occur is based on the electrical design of the section
and power supply.
If a spark meter is not available, the movement of the
primary circuit meter needles can be used to indicate the
sparkover severity. Some basic concepts are as follows.
• 11 When a sparkover occur* on a modern power supply,
controlled at the thres.hhold level, both the voltage and
current meter needles can dip sharply.
21 The degree of needle movement will usually depend
on the ratio of the operating current to the rated current
of t lie power supply.
3 i With some automatic voltage controls and older power
supplies, a sparkover will be seen by a sharp drop on the
voltmeter with an equally sharp rise on the ammeter
needle.
4 With a well controlled power supply at approxi-
mately 100 -parks per minute, the decay of the voltmeter
may lie l.~> V per fluctuation.
.Yi At the level of 100 sparks per minute, there should be
a number of cycles where the meters come to rest before
the fluctuations begin again.
i»' Valid readings taken durmg the condition of "•) should
record the high point of the voltmeter and low point of the
ammeter.
71 It is possible that the internal sparkover could occur
on a practically continuous basis so that no meter fluctua-
tion is MTU. This condition will show an abnormally low
volt age and high current. ....
Xiiimnai-ij <•/ Sparkorer Conditions
Bccau-*- the presence of sparkover is a fundamental
occurrence in precipitators, the following key points will
be summarized.
11 Minimal sparkover is usually observed in wet process
precipitators. Inlet fields are most susceptable.
Ji Sparkover can be prevalent in dry process precipita-
tor?. This can occur in several fields in series and possibly
in all Held*.
:>: Sparkover will normally occur randomly within a
pn-cipitator section unless there are defects at specific loca-
tions. This can involve misalignment of electrodes or cold
air inleakage. The effect of the inleakage of cold air on
electrical disturbances usually depends on the magnitude
oi air inflnw which can be more severe when the induced
draft fan follows the precipitator. The inlet field is most
sensitive to sparkover when sufficient air inleakage occurs
between the b:ick end ol the kiln and the precipitator to
dilute the moisture benefit of the ga> stream. Another key
factor is a ili>rontinuity of the collector surface due to
ci'ii>truction damage or design. A severe mal-gas distribu-
tion entering the precipitator could also sensitize a local
area. Heavy dust buildups on electrodes add to sensitivity
for sparkover.
4) Heavy sparkover, especially if localized, will result in
electrode damage if allowed to continue over a long period
of time. Power supplies with manual control arc mostly
vulnerable to this condition. However, automatic control
of voltage, if not working properly, can also produce
elect rode damage.
OTHER AREAS OF MAINTENANCE
Much time was devoted to the electrical monitoring and
trouble detection by the observance of sparkovcr char-
acteristics. Several trouble areas of precipitators are
normally first found by drastic changes in the electrical
meter readings.
fluildup of Dust in Hoppers
The buildup of flue dust in a hopper will eventually con-
tact the high voltage electrode frame and produce a
grounded condition for that particular field. With auto-
matic voltage control, the voltage will decay to zero or to
some low value while the current is maintained at current
limit. With manual operation of the power supply, current
will generally rise to cause trip-out of the power supply.
The hopper buildup problem is considered most im-
portant because of the possible damage to the electrode
system, in addition to the loss of precipitation and trtan-
hours for correction. Key points are the following.
1) Make sure the dust evacuation or removal equipment
is sufficient to keep ahead of the collected material. This
problem can, and usually does occur in the inlet lu.ppcr.s.
•J) Keep hopper side walls, especially the apex of hoppi-rs,
above the dew point of the gases. Proper insulation and
heat tracing are important. Eliminate openings in insula-
tion which may cause air movement by convection.
3) Minimize air leakage into the hopper by way of the
conveyor system.
4) Design factors such as slope of hopper walls, type of
corners, internal baffles, and capacity can affect mainten-
nance difficulties.
")) Exert caution in the use of hopper sidewall vibrators
if substantial dust buildup occurs in the hopper. The
design of the hopper apex should include sufficient means
to unplug the hopper. Once the apex of the hopper is dust
free, judicial use of a vibrator or hammer blow* oil
.strike pads can dislodge any upper buildups. The uw of
compressed air can also be helpful.
0) Methods to detect substantial hopper buildups and
even de-energize the electrical field before the dust can
reach the high voltage frame is highly recommended.
While it is preferable to place more emphasis on the pn«-
vcntion of buildup, reliable detection methods arc im-
portant.
Start-up and Shutdown Problems
Periods of start-up and shutdown of the process :m«
critical for the precipitator. It is difficult to discuss all the
5-6
-------
: EI-ECTWO.VT \-ric
aspects of this important .area, but the following comments
should help.
I) If hopprr or support insulator heater elements are
liable, make sure those heat sources an- in operation at
«cast three hours before start-up.
2) Ascertain the combustible level in the gas exiting the
kiln before electrical energization.
3) It is generally preferable to preheat the precipitator
proper to as high a temperature as possible before energiza-
tion of the power supplies. Gas temperatures of 180-200°F
at the exit of the precipitator are recommended. If precip-
itator operation is required before this temperature range
is reached, it is suggested that the outlet electrical fields
bv energized first at low power settings.
•4) Mace all rapper equipment in service prior to the
start-up of kiln.
3) Make sure all hopper evacuation equipment is in
operation before startup of the kiln.
6) Upon shutdown of process, it is suggested that the
electrical sections of precipitator be de-energized before
the gas temperature falls below 200-2">00F at the exit of
prsrable.
Maintenance of Support Insulators
All insulator surfaces exposed to gas conditions provide
potential electrical paths to ground in the precipitator
proper. Insulators are used in the top support structure
for the high voltage discharge electrode frame. Another
insulator application isolates the high voltage frame
rapping mechanism from ground potential. Other insula-
tors are sometimes used at the bottom of high voltage
frame to provide stabilization. The insulator is usually
a tub design or post for the top support structure. En-
closures are usually made up of individual compartments
for the cement industry, but enclosure designs can vary.
Some points of concern are the following.
1) Energization of the precipitator without the proper
protection of insulator surface conditions can lead to
electrical leakage and insulator failures. Preheating of
some precipitators are required to minimize this condition.
2) Whenever precipitator fields are energized from a cold
start, each power supply should be testi-d at low power
input in the manual mode of control to del ermine whether
insulator leakage exists. This can bv detected by low
voltage-high current readings, with a tendency to vary
sharply. The T-K sets should be quickly de-rncrgizpd.
Frequent starts and stops by trained personnel can possibly
•> out the leakage paths without permanent damage to
iiie insulator surface. Otherwise, nil additional waiting
p?ratd brforv energization is indicated.
:\) Insulation of the insulator compartments and heated
airblowninto compartments are effective aids to minimise
insulator failures. Heater strips or elements can also be
used in the compartments to keep insulator surfaces nbrws
the dew point.
4) The inner surface of a tub-design insulator is exposed
to gas conditons. Cleaning of this surface during phut-
downs is recommended.
5) Replacement of an insulator is a critical procedure.
Extreme care must be taken to prevent misalignment
of the high voltage frame, and this is especially important
with a two-point suspension system. An J-in deviation at
the top support pivot can produce a spacing problem at the
bottom of the precipitator.
Rapper Maintenance
Maintenance procedures for the care of rappers can
be critical for cement kiln precipitators. The object of
rapping is to dislodge the collected material from the
electrode surfaces with minimal reentrainment of dust into
the gas stream or physical damage to the components.
The failure of the rapper can often lead to electrical dis-
turbances because of excessive material buildup on elec-
trode surfaces. The variety of rapper designs will generally
include pneumatic, impact, electrical vibration, and
electromagnetic impact. Other modes of rapping can also
be supplied.
Reliability of the rapper system starts with proper
design by the manufacturer, but plant maintenance can
often correct repetitive trouble spots. Some critical areas
include the following.
1) Compressed air used for pneumatic rappers should be
moisture free. For some systems, insulation of the air
line exposed to cold ambient ternperatures is suggested.
.2) A clean ambient environment for the rapper control
cabinets is especially important for apparatus such as cam
timers and armature relays.
3) Periodic check and adjustment of anvil gaps of the
electric \ibrators may be required.
4) Check for binding of rapper shafts that extend
through the precipitator shell.
o) Breakage of bolts ma}' be due to oversized bolt holes.
0) Failures of springs or coils in the electromagnetic type
of rapper sometimes occur with extended periods of high
intensity rapping.
7) Even' component of a rapper system exposed to gas
and dust conditions should be carefully checked during
each outage for wear of rotating parts, alignment of shafts,
and binding at packing glands.
A maintenance program will be enhanced by a knowl-
edge of the ramifications of rapper performance. Some key
points include the following.
1) The initial pattern of rapper frequency and intensity
may require modification if there is a substantial change
in the process. Internal inspections and observations of
electrical disturbances can determine the extent of this
modification.
5-7
-------
080
IEEE TRANSACTION* ON I MM -TRY APPLICATIONS, NOVEMBER DECEMBER ItlTS
2) The dust buildups on electrode systems will not clean
uniformly with normal rapper designs. There is nipping
attenuation and harmonics that occur in any Riven elec-
trode structure. The concern is that the random buildup
areas do not adversely effect the electrical power input to
the prccipitator. It is better to add several rappers than to
operate existing rappers above a feasible maintenance
limit.
3) Gas. dust, and fume characteristics are critical to
the effective removal of collected materials on the elec-
trodes. High alkalie levels at certain gas temperatures can
aggravate buildup problems. In this case, it may be de-
sirable to modify the cause of the buildup rather than en-
large the rapper system.
4) The type of rapping blow can be critical. If an electric
vibrator is ineffective in a certain installation, the impact
blow may be found beneficial.
Failures of Electrodes
Whether failures of discharge electrodes or damage to
collector surfaces occur in cement • kiln precipitators
depends on both design features and the process character-
istics. The following comments may be beneficial.
I) Isolated failures of discharge electrodes can readily
occur during the first months of operation. Fabrication or
construction defects may surface during this early period.
Continuous electrode failures are abnormal, and an at-
tempt should be made to determine the cause of the fail-
ures.
2) Patterns and locations of failures should be noted.
The type of failure break will indicate whether it is cati.«ed
by metal fatigue or electrical erosion.
3) Repeated failures'of discharge electrodes opposite
obstructions on collector surfaces are indicative of elec-
trical erosion. The presence of severe sparkover will usually
coincide with this condition.
4). Corrosion of electrode surfaces will generally occur at
the bottom or outer walls of the precipitator. Air inleakagu
may cau.-e .some localized corrosion damage.
it) If mechanical breaks are noted in the electrodes
closest to the rapper mechanism, then the reduction of
rapper impact is mandatory.
G) If electrode damage is noted at the bottom corners
opposite a precipitator section, then the oscillation of the
high voltage discharge frame may be suspect.
7) The most critical location of wire electrode failures
may occur opposite the top and bottom terminations of
the collector surfaces under certain design conditions.
S) The buildup of hopper dust that extends into the
bottom structure of a precipitator section can produce
localized electrode damage.
9) Severe temperature gradients can cause some bowing
and distort ion of collector surfaces.
10) Tin- use of electrical readings and characteristics
over a period of t hue may be ini|M>rtant to ascertain reasons
for electrode f:iilnre>. Sparkover may not be tolerated on
Home, installations depending on the degree of electric
<*ontrol or ttiiflt:tn!'*:i! desK'ii.
SUMMARY
This paper is primarily offered as a nontechnical cover-
ape of some problems that may face the maintenance de-
partment of a cement plant after installation of an elec-
trostatic precipitator on kiln gases. Whether these prob-
lems actually arise are subject to many variables including
the thought and care given to reliability features in the
original design.,
The following references can help plant personnel gain
greater insight into the theory and application of pre-
cipitators. Remember, however, that the success or failure
of. any given installation will often rest on process effects,
frequent comprehensive internal inspections, and correc-
tion of repetitive maintenance problems. This care can
sometimes result in acceptable performance from a mar-
ginal collector.
REFERENCES
il| Terminology for Electrostatic Precipitators, Industrial Gas Clean-
ing Inst., Inc., 1110 .Summer St., P.O. Box 1333, Stamford, Conn.
OGU04.
[2] Criteria for Performance Guarantee Determinations, pub. no. 3,
Inilustri.il Gas Cleanine Inst., Inc., 1116 Summer St., P.O. Box
1333, Stamford. Conn. (Hj'JfM.
[3] Information Kei/uircil for the Preparation of Bidding Sped feat iant
far Electrostatic Precipilator*, EP 5, Industrial Gas Cleaning Inst.,
Inc.. 1110 Summer St., P.O. Box 1333, Stamford, Conn. «»'J04.
[4] H. J. White, Industrial Electrostatic Precipitation. Reading,
Mass.: Addi-son-Wesley, I'.Hio.
[.">] S. Oglesby, Jr., and G. R. Nichols, .4 Manual of Prreipitaior
Technology, part 1. fundamentals, document PB 1073X0: part
II, applications areas, document PB HtG.'iSl; National Technical
Information Service. Springfield, Va., 1070.
[6| M. Robinson and X. Fri.sch, .4 ,\[anual of Electrostatic Prt-
dpitator Ttclwingy, part III; National Technical Information
Service, Springfield, \ a., document PB I'JMT'J, l'.)70.
(7) M. Robinson, "Air pollution "control," part 1, in Electrostatic
Precipitation. New York: Intel-science. l'J71.
Jacob Katz received the U.S. decree in elec-
trical engineering and the M.S. degree in
hygiene from the t'niversily of Pittsburgh.
Pittsburgh, Pa., in 194'J and 1069, respec-
tively.
From 19.V) to 19*>6 he wa,s Test and Service
Engineer with Hesearch-Cottrell, Inc., where
he worked primarily on electrostatic prccioiia-
tor applications. From 19.">6 to 1963 he was
___^___ _^___ with the Duquesne Works of thi> United
States Steel Corporation. From 1064 to 1068
he was nn Adjunct Instructor of air pollution studies at the Graduate
School of Public Health, I'niversity of Pittsburgh, Pittsburgh, Pa-
Since 1064 he has worked as a consultant, based in Munlmll. Pa.,
in the air pollution field. He is presently with Prccipitator Tech-
nology, Inc., Munhall, Pa. His ex]>erience includes work on the
evaluation and upgrading of electrostatic precipituiorx for the ce-
ment, utility, and iron and steel industries with .specialized work in
bource sampling of paniculate emissions from a variety of industrial
processes.
Mr. Katz is a member of the Pennsylvania and National Society
of Professional KnRineers, the Association of Irim anil Slct-l Kngi-
iieew, the American Society of Mec-hauirul Eiiinnccr*, Power '£"«=*
Code 21/27, subcommittee PTC 21/27, anil coniiniUcc 1TC '•>•
He is idio a member of the Air Pollution Control Assuriation,
chairman of I heir committee TC-1 and subcommittee TIM.
5-8
-------
V-6 .
OPERATIONAL MONITORING AND MAINTENANCE OF INDUSTRIAL
ELECTROSTATIC PRECIPITATORS FOR OPTIMUM PERFORMANCE
by
R. L. Cunningham, P.E.
Eastman Kodak Co.
Copyright (cT) 1976 by the Institute of Electrical and
Electronics Engineers, Inc. Reprinted with permission
from IEEE Conference Record, IAS Annual Meeting, October 11-
14, 1976, Chicago, Illinois (No. 76 - CH 1122-lIA).
6-1
-------
OPERATIONAL MONITORING AND MAINTENANCE OF INDUSTRIAL
ELECTROSTATIC PRECIPITATORS FOR OPTIMUM PERFORMANCE
By:
Richard L. Cunningham, P.E.
Utilities Division
Eastman Kodak Company
Rochester. New York H650
BOILER
ABSTRACT
This paper is directed primarily to operating and
maintenance personnel. Designers or precipitator
manufacturers also may be interested in this informa-
tion, h'e wish to outline our experiences and our
methods in order to assist others in maintaining
optimum precipitator performance.
A first portion presents Kodak's expectation of
operations in monitoring units and details the tools
and methods we have employed. The use of voltage and
current meters, opacity monitors, graphs, visual.
checks, and training is detailed with a view to pro-
blems and their solution.
The second portion of this paper details main-
tenance and its scheduling. Also, it examines several
corrective actions which have been taken.
Introduction
Kodak Park is Eastman Kodak Company's largest
industrial complex. Located in Rochester, New York,
this photographic and chemical production facility is
like a small, self-contained city. The complex pro-
vides most utility services -- steam, electric,
refrigeration, compressed gases, waste disposal, waste
water treatment, and telephone. More than 30,000
employees work here, ir. more than 200 buildings
spread over 1,900 acres.
' As in small cities, concerns are always present
that our need for highly reliable and economical
utilities -- critical in today's complex manufacturing
processes -- be properly balanced by our respect for
the environment, which has been of extreme importance
at Kodak since the company's earliest days. This
requires, among other things, that continual, reliable
and optimum performance is achieved on our electro-
static precipitators (E.S.P.'s). Literature is avail-
able with general discussions recommending that units
be maintained and watched closely, but there is not
ouch detail on how this should be done. Scant help
is offered on the use of monitoring procedures and
equipment. There is a need for guides to continual
operation and maintenance and this need has been
voiced outside Kodak.
We wish to share our experiences with you in this
presentation, hoping you may profit from our experi-
ences and that we may learn from yours.
Our facilities are shown in these tables:
Type
.Two Multiple Retort
Underfeed Stokers
.Two Multiple Retort
Underfeed Stokers
.One Oil Fired with
Forney-Ver1oop
Burners
.One Oil and Coal
Fired Cyclone
.One Oil and Coal
Fired Cyclone
.One Oil and Coal
Fired Cyclone
.One Oil and Coal
Fired Cyclone
.One Refuse/Oil
Tangentially Fired
Steam Load
(1000»/hr)
340
(Combined Load)
400
(Conbined Load)
450
400
400
550
400
150
Design
Effi-
ciency
95%
95%
90%
95%
95%
98.3%
96.5%
99%
E.S.P^
Design
Gas
Volume
(1000 ACFM)
181
181
162
152
152
204
161
101.5
Design
Gas
Temp .
450°F
4SO°F
354°F
330°F
330"F
325*F
3SO*F
625'F
Manufacturer
Research-
Cot trell
Research-
Cottrell
Research-
Cottrell
Research-
Cottrell
Research-
Cottrell
Research-
Cottrell
Universal Oil
Products
Wheelabratof-
.Frye
Our work includes operation, maintenance, and de-
sign modifications of these facilities. The toilers
are located in three different buildings approximately
1 1/2 miles apart. The E.S.P.'s are all outdoors.
6-2
-------
rated 45 kV (avc.) and from 250 to 750 mA.
Program Set-up
andin Was Put ln
and filled in once each trick by
as they made checks of each unit'o
room. We now record the primary volts and
amps secondary volts and amps, spark rate
""
.
tenance schedules.
e°°d
3).
4).
A need for good communication of problems
coordination of work, and documentation of
operation and maintenance.
Aersod t0 i^PreSS ?Perators and maintenance
a^nsng,ClCan flue gases via"proper E^sT"'
ana boiler operation.
A need for training personnel in the theory
maintenance, and operation of E.S.P.'s"
ful rL tCTPeratu«. and t
or tuel. The electricians and an engineer
» the operating department review the
and opacity charts. Besides observing
h±vBS a"d Charts> 0P"«ors inspect asn
handling systems to insure proper operation
and to prevent high hopper-ash problems!
2). An operating procedure was written to help
guide operators in monitoring and operating
°pacity versus
characteristics
o
t«
standard deviations above it. is drawn. Any
opacity above this second line now is re
berch r,3 defe«iVe ^-P-ent report t"
be checked and responded to. For example
if the boiler is operating at 400.000 i
hr. and the opacity is 10%, a report is
required (see Figure 1). This report
6).
»K Provide 8""es to determine
whether an E.S.P. is malfunctioning or is
operating at optimum efficiency.
A need to overcome such difficulties as
B Ordor t
" " d
20
H>
I"
Reodings in this region
require o defective ••
equipment report. •;:'.••'
500
Operations
6-3
PARALLEL LINE TWO STANDARD UEVIATlONS AWAY
___^ jtGNLINE
WO 400 450~
Boiler steam load [lO5 Ibs / hr]
Figure I
'HM. ^ "f :current CV-») characteristic
vo t^e } " draW by first Gaining
voltage versus current after any main-
tenance shutdown (unit is cleaned air i,
at ambient temperature) for each field-
±ct h! CfalCUlatio" to compensate for'
expected flue gas temperature durine
operation is madr (For theory
.
oc
units with very low resistivity
S'iSJV;! operatine curvc
to right of the temperature-compensated
not
-------
theoretical curve.) Any combination of
voltage and current reading to the left of
this curve is reported on a defective
equipment report.
400
n '300
Q
200-
100 •
VI CHARACTERISTIC INLET SECTION
Reodings in this
region require o
defective
equipment report
50
40
30
8 20
Temperoture compensated
curve (500°F)
Air load curve ol 70° F
500
400
300
10 15 20 25
30 35
Kilovolts
Figure 2
40 45
50
55
200
GRAPHS of KILOVOLTS and MILLIAMPS from
LOG RECORDED ONCE per TRICK
Reodings by dote
Figure 3
.The transformer-rectifier ratings are
posted so that the operators can report
current or voltage readings that are zero
or over the ratings,
.Notification of problems is usually done
with a written note on a standard form
reporting defective equipment. This
report is seen by operating department.
supervision and a utilities electrician.
A reply is written down when the situation
is corrected and kept on file for the
given unit to document E.S.P. problems
and corrective action taken.
.Kilovolts (kV) and milliatnperes (mA) are~
currently being charted for each high-
voltage bus. We have been experimenting
to see if using these graphs like a
Quality Control Chart is feasible. When
the process level (kV or mA) changes
significantly or a trend occurs, this
question should be asked, "Why?". The
answer may be an E.S.P. malfunctior or
just a significant change in boiler load
or fuel characteristics (say'percent ash).
An example of such a graph is shewn in
Figure No. 3.
3).
.In addition, operators check for abnormal
noises, insure the ash handling system is
functioning correctly, note high sparking
conditions, che-ck lights that indicate
rapper operation, and check that air flow
(at intakes) to support bushing compart-
ments is adequate.
.All the above items are operating pro-
cedures reviewed with each shift by the
operating department supervision. They
included procedures to insure that warn
( > 225eF) dry air has been supplied to
support bushing compartments for more than
30 minutes before energizing the E.S.P.
during start-up.
Additional monitoring of the rappers,
vibrators, and.air supply to high voltage
support bushings is done once per week via
a physical inspection on the roof. A log
check-off is made by the electrical
operations group. This check consists of
insuring that each rapper operates (in
correct sequence), insuring that air pres-
sure to bushing compartments is greater
than in E.S.P., and insuring that positive
nitrogen pressures are present in the high-
voltage bus duct (pipe and guard) to insure
dry environment on two of our units.
6-4
-------
Maintenance
1). Maintenance electricians also examine the
logs and graphs used by the operating
department to check for possible problems.
2). The electricians are responsible for re-
sponding to most of the defective
equipment reports indicating possible
problems. Routine maintenance items such
as broken wires or control circuit pro-
blems are easily handled without seeking
engineering assistance.
3). Weekly checks (besides roof inspections)
consist of general checks of control room
and warm air system to support bushing
compartments.
4). An annual maintenance check-list has been
developed for each unit. The effort
required for these checks is at least 400
labor hours per unit. An abbreviated
listing is:
.Before securing the unit, check rapper
operation, warm air system, and air load
test (right after boiler shutdown).
.Secure and ground 'the E.S.P.
.Before cleaning the unit, check for
grounds, failures of voltage divider
stack or diodes, ash build-ups on wires
or plates, pluggage of diffusion plate,
and indications of arcing on (or cracks
in) insulators. We use a motorized 2SOO
volt megger for the above electrical
checks. The high voltage frame and the
high voltage bus are moggernd together
and separately.
.After washing, make a more thorough in-
ternal inspection. Check for alignment
(see Figure 4), evidence of corrosion,
condition of rappers and rapper rods,
structural integrity (such as condition
of hanging bolts), existence of bowed
or torn plates, evidence of arcing, and
foreign material.
Collector Plate Support Channel Leading Edge
Top View
I Not to Scale
Tolerance ± *
O - High voltage electrode
Figure 4
5).
.Check and clean all insulators, replace if
cracked.
.Check bus duct, warm air supply system,
and all compartments for dust, moisture,
tracking, or holes.
.Check transformer-rectifier .for oil leaks,
oil level, oil dielectric strength, oil
appearance; check the surge-arrestor air
gap, metering and control networks for
tracking to ground, cleanliness of
ground switch connections, need to replace
door seals, etc.
.Check vibrators and rappers—inspect
rapper boot seals and support rods, check
vibrator for water, corrosion, or grounds.
.Check warm air supply fans and motors.
.Check control system -- wiring, terminals,
devices, meter calibration; recondition
distributor switch and timer; check diode
alid SCR foreward and reverse bias resis-
tance with megger; check cabinet vent fan
and motor.
.Recheck interior for foreign material,
close access doors and account for all
personnel.
.Megger frame and bus (together and
separately) with 2500 V. power megger.
.Lock all doors and retrieve all interlock
keys.
.Energize unit and take static air load
test. Note voltage at which corona
starts. Use these data to make V-I
characteristic curve for each field.
Double check waveforms with scope (current,
voltage, and SCR firing).
.Adjust controls for optimum: voltage
limit and current limit at maximum
equipment ratings, spark limit at less
than 100 per minute, sensivity pot to
respond to sparking and not to 60 Hz line
signals.
.Repair roof, if necessary, after all work
is done. This is necessary to insure
insulation integrity if the unit does not
have pent-house or double roof.
Opacity monitors arc checked once each month
(on site) by instrument maintenance per-
sonnel. This insures blowers are operating,
alignment of unit is okay, lenses arc clean,
etc. We have had filters made to allow
calibration bench checks of units. A
standard report form is being set up to re-
cord all maintenance and calibration work
done on opacity monitors.
Design Changes and Problem Areas
.To assist in controlling sparking we have
modified our units with Research-Cottrell's
newest SCR proportional rate controller.
6-5
-------
.We are investigating replacing some old
saturable core controls with new SCR
controls.
.Some of the bolts supporting the
collecting plate frame have worn badly
where they pass through the channel
iron to which the plates are held.
We have replaced these 5/8 inch bolts
with 1 inch bolts.
.To prevent cracked support bushing in-
sulators (crocks), we are modifying our
warm air supply system for the support
bushing compartments. In some cases
this has required two fans and motors in
parallel for one unit to insure a re-
liable air supply.
.We make sure that there is proper
temperature in the insulator compart-
ments to prevent condensation ( > 225°F
and
-------
be replaced if faulty. Develop a
periodic maintenance check on these
controls as with control circuits for
other equipment.
.On one unit our pipe and guard pro-
. tcction for high voltage bus rusted
away. The primary cause was being in
the path of the exhaust head water
vapor. We have supplied nitrogen to
the bus duct, insulated the duct, and
insulated the support bushing enclosure
to maintain high voltage bus system
integrity.
.We have had to lift rectifier apparatus
out of the transformer oil two or three
inches with a crane to replace low vol-
tage bushings which have cracked.
Special care to protect insulating fluid
from contamination was required. . Be
careful in tightening the bolts on these
insulators I
Training sessions for operators, maintenance
personnel and engineers are required to:
1). Develop an awareness of the need for
operational monitoring, maintenance, and
documentation to insure optimum cleaning
of flue gases.
2). Develop a better understanding of the
basics of precipitator operation.
3). Develop an appreciation of the extent
of environmental regulations. Many
people do not realize the detailed
requirements to be met in testing, in
rpporr retention, in siaking reports, etc.
that are presently or will be in effect.
In summary, there is a very definite need
for:
.Management support
.Logs to monitor E.S.P.'s
.Opacity meters to monitor E.S.P.'s
.Operating procedures
.Interpretation of logs
.Continual training for maintenance
personnel and engineers
.One engineer (or group) responsible for
proper E.S.P. functioning.
BIBLIOGRAPHY
[1] Engelbrecht, Heinz L. -- "Electrostatic
Precipitator Operation and Industr^il
Applications", seminar manual given at
Kodak Park, Rochester, N.Y., 1974.
Mr. Engelbrecht employed by Wheelabrator-
Frye, Inc.
[2] IBID. -- "Electrostatic Precipitator Operation
and Maintenance Seminar".
[3] Hall, H.J., "Design and Application of High
Voltage Power Supplies in Electrostatic
Precipitation", Journal of A.P.C.A.,
Vol. 25. No. 2, Feb., 197S, pp!32-138.
[4] Nichols, Grady B. and Couch J.P., "An Electro-
static Precipitator Performance Model"
Final S.R.I, report for E.P.A. under
contract No. CPA70-166, 7/6/72.
[5] Ogleshy, S. and Nichols, Grady B., "A Manual of
Electrostatic Precipitator Technology,
Parts I and II", SRI report under National
Air Pollution Control Adra. Contract
C.P.A. 22-69-73, Aug., 1970.
[6] Ramsdell, Roger G., "Design Criteria for
Precipitator for Modern Central Station
Power Plants", report at Consolidated
Edison Co., of N.Y., Inc., 1968.
[7] Reynolds, J.P. (et alia), "Calculating Collection
Efficiencies for Electrostatic Precipita-
tors", APCA Journal, Vol. 25, No. 6,
pp610-616, June, 1975.
[8] Schnedicr, Gilbert, G. (et alica), "Selecting
and Specifying Electrostatic Precipitators",
Chemical Engineering, May 26, 1976,
pp94-108.
[9] White, H.J. -- "Industrial Electrostatic
Precipitation", Addition-Wesley, 1963.
[10] IBID. -- "Resistivity Problems in Electrostatic
Precipitation", Journal of A.P.C.A.,
April, 1974, Vol. 24, No. 4, pp314-338.
6-7
-------
EPA-600/8-77-020b
December 1977
V-7.
PARTICULATE CONTROL HIGHLIGHTS:
AN ELECTROSTATIC PRECIPITATOR
PERFORMANCE MODEL
by
J. McDonald and L. Felix
Southern Research Institute
2000 Ninth Avenue, South
Birmingham, Alabama 35205
Contract No. 68-02-2114
Program Element No. EHE624
EPA Project Officer: Dennis C. Drehmel
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park. N.C. 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 20460
-------
ABSTRACT
Electrostatic precipitators are widely used for controlling emissions of fly ash and other dusts
from industrial sources. Research on the process of electrostatic precipitation has resulted in a
computerized mathematical model that can be used for estimating collection efficiency for pre-
cipitators of different designs operating under various conditions. Mathematical expressions
based on theory are used for calculating electric fields and dust particle charging rates. Empirical
corrections are made for non-ideal effects such as a non-uniform gas velocity distribution. The
model is expected to aid in improving precipitator design and in selecting optimum operating
conditions.
THE COVER:
The EPA has sponsored research to develop
a computer model to predict electrostatic
precipitator performance. The model is
available to industry and the public upon
request. A reference to the computer
model is given at the end of this report.
-------
CONTENTS
Abstract ii
Modeling a Precipitator 3
Validating the Precipitator Model 6
Applications 9
FIGURES
Figure 1. Schematic diagram of an electrostatic precipitator collecting dust 2
Figure 2. Particle charge vs. electric field strength for laboratory aerosols
of four different diameters 4
Figure 3. Particle charge vs. diameter for three values of electric field 5
Figure 4. Particle charge vs. Not for three values of electric field 5
Figure 5. Average current density at the collection plate vs. corona voltage 5
Figure 6. Electric potential vs. position between the corona wire and
collection plate 6
Figure 7. Electric field of the collection plate vs. position 6
Figure 8. Simplified flow chart of the computer program to calculate
precipitator performance 7
Figure 9. Experimental and predicted migration velocities for a laboratory
precipitator 8
Figure 10. Experimental and predicted collection efficiency vs. particle
diameter for a laboratory scale precipitator 8
Figure 11. Experimental and predicted migration velocity vs. particle diameter for
a full scale precipitator 9
Figure 12. Experimental and predicted migration velocities vs. particle diameter
for a full scale precipitator 9
in
-------
AN ELECTROSTATIC PRECIPITATOR PERFORMANCE MODEL
The availability of high speed digital comput-
ers makes it possible for the engineer to examine
complex industrial processes by constructing
mathematical models of them which can be used,
for example, to show the effect that a variation
in a process parameter such as temperature or
pressure will have on the rate or direction of the
process. An example of the use of this tech-
nique is the modeling of the process of electro-
static precipitation which is used for removing
dust and ash from industrial exhaust gases.
Paniculate air pollution is produced by many
.industrial processes, such as metallurgical smelt-
ers, iron arid steel furnaces, incinerators, electric
power generating plants, and cement kilns. Elec-
trostatic precipitators, sometimes called prccipi-
tators, arc used in all of these industries to con-
trol air pollution.
Well designed electrostatic precipitators typical-
ly remove better than 98% of the dust in the
exhaust gas they treat. The collected dust can
be re-introduced into the manufacturing process,
sold to other industries for raw material, Or
disposed of, for example, in a landfill.
One of the largest sources of industrial air
pollution that must be controlled is the fly ash
produced in coal fired electrical power plants.
Electrostatic precipitators are widely used in the
power industry and in 1976 they were used to
remove an estimated 40 million tons of fly ash
from coal fired boiler stack gases in the United
States.
The widespread use of precipitators provided
the impetus for research by the Environmental
Protection Agency into the operating mechanisms
of these control devices to obtain information
that can be used in the design of more efficient
equipment. As part of this effort, a mathemati-
cal model of the electrostatic precipitation pro-
cess has been developed.
Figure 1 shows a schematic drawing of an elec-
trostatic precipitator. The precipitator shown is
typical of those which are used to collect fly ash.
The dust laden flue gas enters the precipitator
from the left and flows between negatively
charged wire electrodes and nearby grounded
plate electrodes. The wire electrode is charged
to a high potential (20-40 kV) by an unfiltcrcd
dc power supply outside the precipitator housing.
The applied voltage is high enough to produce a
visible corona discharge in the gas immediately
surrounding the wire electrodes. Electrons set
free in the discharge collide with gas molecules
producing gas ions that in turn collide with dust
particles and give them negative charges. In the
strong electric field between the wire and plate
electrodes the electrically charged dust particles
migrate to the plates where they arc deposited,
giving up their charge. Eventually a thick layer
of dust builds up on the plates. With vertically
mounted wire and plate electrodes the accumu-
lated dust layer can be conveniently removed
from the plate by periodically rapping it by
means of an automatic hammer. The dislodged
dust layer falls into hoppers in the bottom of
the precipitator housing, from which it is re-
moved for disposal. The plates continue to col-
lect dust until they arc rapped again.
Most industrial precipitators are quite large
because large volumes of particulate laden flue
gases must be treated. A large electric utility
power boiler burning coal may require several
precipitators, each of which will typically con-
tain over 500 collection plates 10 meters high
and 3 meters wide. Each precipitator will treat
a million cubic meters of flue gas per hour, re-
cover several tons of fly ash during that time,
and cost perhaps $5 million. On such a scale,
the need for accurate design predictions of the
and geometry of precipitator components is
apparent. Also, as precipitators are applied to
various industrial processes, the scaling rules dis-
covered by precipitator manufacturers for one
application may not work in another.
-------
HIGH-VOLTAGE
SUPPLY
PLATE
RAPPING
SYSTEM
DUST
LADEN
AIR
CORONA WIRES
GROUNDED
COLLECTION
ELECTRODES
CLEANED
AIR
DUST COLLECTION
HOPPERS
Figure 1. Schematic diagram of an electrostatic precipitator collecting dust.
-------
MODELING A PRECIPITATOR
Most of the models that one sees arc physical
entities - a miniature representation of some air-
craft or ship, for example. The quality of the
model is in direct proportion with the accuracy
which the original design is minutely reproduced.
Another kind of model is the abstract construct.
Thus, a theory, for example, is a model because
it seeks to represent how something in nature
works or acts. Instead of wood or metal, a
theory is a model made up of facts, each fact
pieced together with another fact until some
representation of nature has been made. The
quality of this model is judged by how well it
predicts what nature will do in the situations
that it was designed to model.
Therefore any object or phenomenon can be
modeled. What is important is that the model
can be either a concrete or an abstract structure.
A mathematical model of some process is then
no more than a representation of the process by
mathematical formulae tied together with some
overriding procedure or logic. This report deals
with a mathematical model of electrostatic pre-
cipitation; the model is simply some fundamental
theories of physical processes tied together by
the logic of a computer program.
The idea of modeling the electrostatic precipi-
tation process has great appeal if only because of
economic considerations. On a more fundamen-
tal level, the modeling of any complex process
is useful because it promotes an understanding
which is otherwise only available from a costly
"cut and try" approach.
Modeling the electrostatic precipitation pro-
cess is complicated because a variety of physical
phenomena must be accounted for in order to
predict prccipitator performance. The process
is also sensitive to a number of parameters which
must be accurately measured or estimated. The
efficiency of particle collection for a given par-
ticle size is a function of ash or dust properties
(chemical composition, resistivity, density, parti-
cle size distribution), precipitator operating par-
ameters (applied voltage, temperature, gas com-
position, gas flow rate) and prccipitator geometry
(collecting plate area, internal dimensions).
Historically, the first aspect of precipitator
performance to be studied was the effect of var-
ious precipitator operating parameters on collec-
tion efficiency. The first successful electrostatic
prccipitators for controlling industrial dust emis-
sions were developed by F. G. Cottrcll in 1910.
Shortly afterwards, one of Cottrcll's associates,
Evald Anderson, recognized that the efficiency
of dust collection was exponentially related to
such parameters as gas velocity and collecting
plate area. In 1922 the German investigator W.
Deutsch put this relationship into a more com-
prehensive form that incorporated concepts from
electrical theory. The equation developed by
Deutsch predicts prccipitator collection efficiency
at a particular particle size for turbulent flow con-
ditions and depends upon three parameters: the
area of the grounded collection electrode, the vol-
ume flow rate of the gas passing through the pre-
cipitator, and the migration velocity of the dust
particle to the collection electrode. The last of
these, the migration velocity, is the net velocity
of the dust particle to the collection electrode
resulting from the opposition of two forces, the
force of electrostatic attraction and the viscous
drag of the gas, which retards movement of the
particle. The migration velocity depends on the
charge on the particle, the electric field near the
collection electrode, the gas viscosity, the parti-
cle diameter, and an empirical correction factor
called the Cunningham or slip correction factor.
The Deutsch equation is idealized in that it
assumes thorough mixing of the gas due to tur-
bulent flow, a uniform concentration of uniform-
ly sized (monodisperse) dust particles, and a con-
stant migration velocity for these particles. Any
comprehensive modeling effort must make allow-
ance for these restrictions. In the computer
modeling scheme which has been developed, the
precipitator was divided into short sections and
the Deutsch equation applied to each section,
over several particle size ranges.
Two other fundamental aspects of precipitator
operation which must be described before any
model is built are particle charging and electric
field estimation, both of which are needed to
find the migration velocity.
Finding the charge acquired by a dust particle
in the presence of free gas ions and an electric
field is a complex calculation. Briefly, there are
two ways in which a dust particle can acquire
charge in a precipitator. If the particle is larger
than one or two microns in diameter then the
applied electric field is responsible for most of
-------
the charge on the particle. This type of charging,
called field charging, depends on an induced elec-
tric field to be set up on the dust particle. Then
ions moving in the electric field set up on the
particle arc attracted to it, impact, and give it
charge. The particles continue to acquire charge
until the resident charge on the particle is large
enough to repel the incoming ions. The particle
has then reached a saturation charge and can gain
further charge only by random collisions with
energetic ions. This second process, the diffusion
of ionic charge to dust particles, is the predomi-
nant charging mechanism for particles smaller
than about one micron in diameter. For particles
near one micron in size both charging mechanisms
operate and the particle gains charge by field
charging and diffusion charging.
Theories which describe particle charging typi-
cally do well in estimating particle charge for
either diffusion charging or field charging condi-
tions, but in the particle size range where both
types of charging occur, a simple sum of the
charging due to each mechanism is incorrect. A
more sophisticated theory is needed. Fortunately,
recent work sponsored by the Environmental
Protection Agency has produced a more compre-
hensive theory of particle charging. This theory
agrees with experiment to within 25%. For par-
ticle sizes and charging times in the range of
interest for precipitator operation, the agreement
with experiment is within 15%.
Figures 2 through 4 show comparisons of
theory and experiment for a variety of experi-
mental charging conditions. Figure 2 shows
particle charge as a function of charging field
strength for four particle sizes. Here the pro-
duct of the charging ion concentration, No, and
the time that the particle is charged, t, is equal
to 1.0 x 10^3 scc/m3. This Not product is in
the correct range for precipitator operation but
is lower than a more usual value of 4 x 10^3
scc/m3. Figure 3 shows particle charge as a
function of particle diameter for three charging
field strengths. The value of 3.6 x 10$ volts/
meter is probably most representative of precip-
itator operation. As in Figure 2 the Not product
is 1.0 x 1013 scc/m3. Figure 4 shows particle
charge as a function of the Not product for
several charging field strengths; these data are
for a particle diameter of 0.28 jim.
One last fundamental aspect of precipitator
operation must be described before a model of
electrostatic precipitation is possible. This is the
500
100
O
cc
o
Ul
o
K
<
d - 0.312
d - 0.109
•THEORY
02468
CHARGING FIELD STRENGTH, kV/cm
Figure 2. Particle charge vs. electric field strength
for laboratory aerosols of four different
diameters. Not = 1 x W1^ sec/nr*.
calculation of the electric field inside the precipi-
tator as a function of position. A correct value
of the electric field is needed to calculate both
migration velocity and particle charged
The equations which describe the behavior of
the electric field in a precipitator are well known.
The difficulty is their solution. Their solution
is obtained by numerically solving the appropri-
ate partial differential equations subject to the
wire-plate geometrical configuration of the elec-
trostatic precipitator. A computer program was
written to perform the calculations and yield a
voltage-current relationship for a given wire-plate
geometry. The distribution of voltage, electric
field, and charge density arc also calculated by
the computer program for each corona wire
voltage and the associated current to the collec-
tion electrode. The agreement between theory
and experiment is within 15%.
Figures 5 through 7 show how the predictions
of this computer program agree with measure-
ments made of the current density, electric field,
-------
and potential values at various places in a wire-
plate electrode system. Figure 5 shows the aver-
age current density at the collecting electrode
(plate) as a function of the voltage applied to the
wire. In this experiment a 1.3 mm wire was used.
Here the agreement between theory and experi-
ment is excellent. Excellent agreement is also
seen in Figure 6, which presents a comparison of
predicted and measured potential as a function
of the distance between the corona wires and
the grounded collection plate. Results for two
wire diameters, 1.016 mm and 0.3048 mm, are
shown. Figure 7 shows the electric field at the
collection plate as a function of displacement.
Corona wires arc located directly across from
the points -10, 0, and 10 cm at the plate. Posi-
tions -5 and 5 correspond to positions at
the plate, midway between corona wires. Again,
the agreement with theory is good, and within
8%.
Now a computer model of the electrostatic
precipitation process can be constructed. The
ID3
102
o
tr
o
P 101
a:
• E - 6.0 x 104 V/m
* E - 3.6 x 10= V/m
• E - 1.08 x 10° V/m
N0t - 1.0 x 1013 sec/m3
THEORY
I I I I
0.2 0.4 0.6 0.8 1.0
PARTICLE DIAMETER, ^m
1.2
1.4
Figure 3. Particle charge vs. diameter for three
values of electric field.
60
50
40
C3
-------
computer model is simply a codified procedure
which uses a mathematical description of each
of the fundamental aspects of prccipitator oper-
ation discussed above to predict the behuvior of
an actual precipitator. As discussed above, the
method used is to break the prccipitator into
many small sections. As the simplified flow
diagram, Figure 8 shows, the particle-size dis-
tribution entering the prccipitator is broken
down into a number of narrow size bands with .
a median particle size calculated for each band.
Calculations arc made separately for each size
band as the dust moves through the segmented
precipitator. In each segment of the precipitator,
the electric field, particle charge, migration veloc-
ity, and collection efficiency are calculated for
40
30
z
UJ
O 20
10
• EXPERIMENTAL
— THEORETICAL
1.02 mm
WIRE DIAMETER
0 2
L.
PLATE
4 6 S 10
DISPLACEMENT, cm WIRE
Figure 6. Electric potential vs. position between
the corona wire and collection plate.
3.0
2.0
ui
o 1.0
, MEASURED
(WITH DISCHARGE)
THEORETICAL
MEASURED
(WITHOUT DISCHARGE)
\
THEORETICAL
0.0
DISPLACEMENT, cm
Figure 7. Electric field of the collection plate
vs. position. Corona wires are directly
across from positions —10,0, 10,
the median particle size and the percent collected
is subtracted from the concentration entering
that segment. This procedure is repeated for the
next and each succeeding segment until the
entire precipitator has been traversed. In this
way each size band passes through the simulated
prccipitator and an overall collection efficiency
is found for the various median sizes. The pre-
cipitator has then been modeled. That is, its
collection efficiency has been predicted over the
range of particle sizes which experiment has
shown that it must collect.
VALIDATING THE PRECIPITATOR MODEL
In order to validate a modeling procedure, the
predictions of the model must be compared with
the behavior of actual systems. This precipitator
-------
Read Input Data
\
Divide precipitator into N segments.
Start with first segment.
Calculate correction factors to allow (or
gas sneakage.
fcgk^v.-.v.v.v.v.-.-'-'-'-.'.V.'.V
J:::::::::4y"sj ••'•••••••••••••• ••••••• ••
Calculate migration velocity. If non-
ideal effects are to be included use
correction (actors generated above to
modify the migration velocities.
Is this the last segment of the
precipitator?
H bh
Move to next segment of the
precipitator.
|B»
C^flpsp
\ji*:
.-.-.-.-.,-
P°|!
•.•••.••Kfffl
N
f
y^--m\
s*
VLX ><£J
\
••• '»«s>. _ _ _f
_ _
Divide particle concentration distribution
into M segments. Start with smallest
particle size.
Are there non ideal effects to be
Calculate collection efficiency for this
segment of the precipitator at this particle
size.
Increment particle size to next largest
size.
V
Print out results; overall efficiency and
other pertinent data.
—
M
st&wvx
VMt*Sf:-ttt:-f<:
>J '
• ty/
\ 1
I Yes
V
(^ No
"^
Calculate electric field values, voltages
current densities, etc., for a chosen
segment.
ft, .
Calculate particle charge for a chosen
Subtract off the amount of dust
collected from the total concentration
entering this segment.
I 1
\/
Is this the largest particle size used?
End of program.
Figure 8. Simplified flow chart of the computer program to
calculate precipitator performance.
-------
mode! has been compared with measured migra-
tion velocities and collection efficiencies for labo-
ratory scale and full scale electrostatic prccipita-
tors. Figure 9 shows the comparison of ideally
calculated migration velocities and collection
efficiencies with experimentally measured values
obtained from a laboratory scale precipitator.
The values obtained in Figure 9 were taken for
three different current densities. The good
agreement with laboratory data indicates that
the model is fundamentally sound. Other
measurements made with the laboratory scale
precipitator indicate that perhaps 8% of the
paniculate laden air does not pass through the
charging regions. If this sneakagc is taken into
account, even better agreement with theory is
achieved, as is shown in Figure 10.
When the precipitator model is compared with
field data and an attempt is made to simulate
the behavior of full scale prccipitators, non-ideal
effects must be included or else the agreement
is generally poor. Therefore, the precipitator
model is not complete until these effects are
allowed for. In a real precipitator, the gas ve-
locity across a duct may be very nonuniform,
the flue gas stream can bypass the electrified
regions (sncakage) and particles that arc once
collected can be rccntraincd when the collecting
100.0
u
O
2
O
O
S
10.0
1.0
THEORETICAL
EXPERIMENTAL
0.1
1.0
PARTICLE DIAMETER,
10.0
Figure 9. Experimental and predicted migration
velocities for a laboratory precipitator.
99.99
99.98
99.95
* 99.9
(j
z
i1 99.8
u
LU
O
p 99.5
O
<-> 99.0
98.0
95.0
90.0
THEORETICAL
EXPERIMENTAL
CORRECTED
FOR 8%
SNEAKAGE
I I i I
0.1
1.0
PARTICLE DIAMETER,
10.0
Figure 10. Experimental and predicted collec-
tion efficiency vs. particle diameter
for a laboratory scale precipitator.
plates are cleaned (rapping rcentrainmcnt). All
of these non-ideal effects are to some extent
design related. However, even with careful
design they usually are reduced but not elimi-
nated.
The net result of the non-ideal effects is to
lower the ideal collection efficiency of the pre-
cipitator. Since the mathematical model of the
precipitator is based on an exponential equation
for individual particle sizes, it is convenient to
represent non-ideal effects in the form of correc-
tion factors which apply to the exponential argu-
ment. The correction factors are used to modify
the ideally calculated migration velocities. The
resulting "apparent" migration velocities are
empirical quantities and arc no longer related to
the actual migration velocities in the real precipi-
tator being modeled. The determination of the
correction factors is an involved task which re-
quires the correlation of large amounts of field
8
-------
information, taken at existing electrostatic pre-
cipitators. These results have also shown that
the current density, applied voltage, and particle
size distribution arc the most important variables
in the calculation of overall mass collection effi-
ciency for a given collection electrode arca-pre-
cipitator gas flow ratio. The theoretical calcula-
tion of ideal overall collection efficiency of a
typical boiler effluent in an electrostatic pre-
cipitator generally predicts a higher value than
is observed. Corrections to the idealized or theo-
retical collection efficiency to estimate the
effects of non-uniform gas flow, rcentrainment
of dust due to rapping, and gas sneakage all
reduce the overall values of calculated efficiency
to the range of values obtained from field
measurements. The calculations suggest that the
theoretical model may be used as a basis for
quantifying performance under field conditions
when sufficient data on the major non-idealities
are available. Considerable effort has been
expended to learn about modeling non-ideal
effects and their inclusion in the precipitator
model. To date the results are promising; how-
ever, much study and evaluation remains to be
done.
Figures 11 and 12 show experimentally
measured and model predicted values of migra-
tion velocity and collection efficiency as a func-
tion of particle diameter for a full scale precipi-
tator. This precipitator collected fly ash from a
coal fired power boiler and operated at an aver-
age temperature of 150°C. These figures il-
lustrate the kind of agreement which is currently
realized. Two curves are shown on each graph.
O
28.0
24.0
20.0
16.0
12.0
< 8.0
(T
O
S 4.0
THEORETICAL-
THEORETICAL
CORRECTED
. I
0.1 0.2 0.4 1.0 2.0
PARTICLE DIAMETER, p
4.0
10.0
Figure 11. Experimental and predicted migration
velocity vs. particle diameter for a
full scale precipitator.
O
O
u
99.99
99.90
99.8
99
98
95
90
80
60
0.
THEORETICAL-^/ '
EXPERIMENTAL
\ ..^<
THEORETICAL
CORRECTED
i i i
i i i
1.0
PARTICLE DIAMETER, pm
10.0
Figure 12. Experimental and predicted migration
velocities vs. particle diameter for a
full scale precipitator.
The upper curve is an "ideal" calculation. The
lower curve takes into account a correction for
a non-ideal gas velocity distribution. Other non-
ideal effects were not taken into account; how-
ever, a continuing effort to model these effects
is underway.
The theory has been compared with a broad
range of laboratory and field data. The results
of these comparisons indicate that the mathe-
matical model provides a basis for indicating
performance trends caused by changes in pre-
cipitator geometry, electrical conditions, and
particle-size distribution.
APPLICATIONS
Precipitator size depends on the quantity of
gas flow, the gas composition, the collection effi-
ciency, the electrical properties of the dust, and
the size distribution of the dust. Present practice
is to base the size on that of an existing precipi-
tator collecting dust from a similar source, on
pilot plant tests, or from empirical relationships.
One of the unknown factors in design is the
allowable current density. Selection of the design
current density involves a prediction of the resis-
tivity of the dust to be collected. If the resistiv-
ity is low then high current densities are possi-
ble. High resistivity dusts are difficult to col-
lect and precipitators must be operated at
reduced current densities. These dusts are often
encountered in flue gas streams from power
boilers burning low sulfur content coals. The
-------
art of precipitator design is based to a great
extend on being able to recognize the relevant
factors influencing resistivity and allowable
current density.
In the electric power industry many types of
empirical relationships have been developed to
permit the selection of design parameters from
coal composition. But none of these relation-
ships are founded in a consistent theory of pre-
cipitator operation. Even these relationships
arc not appropriate for some of the high effi-
ciency precipitators currently being installed.
What is needed, and what the Environmental
Protection Agency is attempting to provide with
the mathematical model of electrostatic precipi-
tation is a theoretical base for prediction of
electrostatic precipitator design parameters.
Cost considerations alone suggest that a useful
mathematical model of electrostatic precipita-
tion would benefit both the manufacturer and
the user of these devices. The actual dollar
savings are dependent on precipitator size,
operating temperature, gas volumetric flow rate,
collection plate area and difficulty of erection.
But all of these factors, with the exclusion of
the physical construction, can be estimated with
the help of the precipitator model. Further-
more, savings would be introduced at the design
stage.
Another useful application of the modeling
effort is in troubleshooting problems in existing
precipitators. The remedy to a problem can be
tried out on the computer before money and
time are commitcd. Once the fix is determined,
costs can be realistically estimated because all
of the needed modifications have been deter-
mined in advance.
With this mathematical model of electrostatic
precipitation, the Environmental Protection
Agency hopes that precipitator design can move
in the direction of a science rather than an art.
It is recognized that the model is not perfect,
especially in a comprehensive estimation of non-
ideal effects. However, a continuing effort of
research and development is underway to im-
prove the model and insure its applicability to
a wide range of gas cleaning situations.*
* A more detailed description of the computer
model is contained in "A Mathematical Model of
of Electrostatic Precipitators", by J. P. Gooch,
J. R. McDonald, and S. Oglesby. Jr. 1975.
NTIS-PB 246188. This report can be ordered
from the National Technical Information
Service, 5285 Port Royal Road, Springfield,
VA 22161.
10
-------
TECHNICAL REPORT DATA '
(Please read luitrtictiuiti on the /ri\~ne lic/wi' c
NO.
EPA-600/8-77-020b
2.
4. TITLE AND SUBTITLE
Participate Control Highlights: An Electrostatic
Precipitator Performance Model
3. RECIPIENT'S ACCbbSIONNO.
5. REPORT DATE
December 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHORlS)
J. McDonald and L. Felix
8. PERFORMING ORGANIZATION REPORT NO.
SORI-EAS-77-675
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Southern Research Institute
2000 Ninth Avenue, South
Birmingham, Alabama 35205
10. PROGRAM ELEMENT NO.
EHE624
11. CONTRACT/GRANT NO.
68-02-2114
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AJMD PERIOD COVERED
Task Final: 11/76-11/77
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES IERL_RTP project officer is Dennis C. Drehmel, Mail Drop 61,
919/541-2925.
16. ABSTRACT
The report describes a computerized mathematical model that can be used
to estimate the collection efficiency of electrostatic precipitators (ESPs) of different
designs, operating under various conditions. (ESPs are widely used to control emis-
sions of fly ash and other dusts from industrial sources.) Mathematical expressions
based on theory are used to calculate electric fields and dust particle charging rates.
Empirical corrections are made for non-ideal effects such as a non-uniform gas
velocity distribution. The model is expected to aid in improving ESP design and in
selecting optimum ESP operating conditions.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
C. COSATI 1'icld/ClOUp
Air Pollution
Electrostatic Precip-
itators
Mathematical Models
Collection
Efficiency
Estimating
Fly Ash
Dust
Air Pollution Control
Stationary Sources
Collection Efficiency
Particulates
13 B
12A
14B
21B
11G
i. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Reporl)
Unclassified
21. NO. OF PAGES
14
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
CPA Form 2220-1 (9-73)
11
-------
|