United States
Environmental Protection
Agency
Office of Research and
Development
Washington, DC 20460
EPA/625/1-85/017
September 1985
Technology Transfer
&EPA Manual
Operation and Maintenance
Manual for Electrostatic
Precipitators
-------
PBM21.6785
EPA/625/1-85/017
September 1985
OPERATION AND MAINTENANCE MANUAL
FOR ELECTROSTATIC PRECIPITATORS
by
PEI Associates, Inc.
11499 Chester Road
Cincinnati, Ohio 45246
Contract No. 68-02-3919
Project No. 3587
EPA Project Officer
Louis S. Hovis
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
AIR AND ENERGY ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
NATIONAL TECHNICAL
INFO«MATIONSERVICE
-------
-------
TECHNICAL REPORT DATA
(Pti'sse resit Imintciions mi ilic rcrrrsr before completing)
1 REPORT NO-
EPA/625/1-85/017
4. TITLE ANOSUBTITLE
Operation and Maintenance Manual for Electrostatic
Precipitators
6. PERFORMING ORGANIZATION CODE
5. H,
PI86-216785TIS_
September ISflS
7. AUTHQRlS!
M.F. Szabo, R. D. Hawks, G, L. Sanders, and
F. D. Hall
8, PERFORMING ORGANIZATION REPORT NO,
9, PERFORMING ORGANIZATION NAME AND ADDRESS
PEI Associates, Inc.
11499 Chester Road
Cincinnati, Ohio 45246-0101
10, PROGRAM ELEMiNT NO.
II. CONTRACT/GRANT NO.
68-02-3919
13. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 1985
14, SPONSORING AGENCV CODE
EPA/600/13
15. SUPPLEMENTARY NOTES ^EERL project officer is Louis S. Hovis. Mail Drop 61, 919/541-
3374.
is. ABSTRACT
manual focuses on the operation and maintenance (O/M) of typical elec-
trostatic precipitators (ESPs). It summarizes available information on theory and
design in sufficient detail to provide a basic background for the O/M portions of the
manual. Although O/M- related air pollution problems cannot be completely elimina-
ted, they can be minimized by the conscientious application of a well planned O/M
program. The causes of such problems often vary widely, and their effects on deter-
iorating performance may be direct, indirect, or synergistic. Process, particle,
mechanical, environmental, and gas-flow-dynamics factors dictate that O/M pro-
grams and troubleshooting be approached from a total system or process/plantwide
viewpoint. The variable nature of these factors also requires that O/M programs be
individualized and specifically tailored to the needs of the process and installation
served. Effective O/M also affects equipment reliability, on-line availability, con-
tinuing regulatory compliance, and regulatory agency /source relations. Lack of
timely and proper O/M leads to gradual equipment deterioration which, in turn, in-
creases the probability of equipment failure and decreases both its reliability and
on-line availability. The last two items can decrease plant productivity if process
operations are forced to be curtailed or shut down for control equipment outages,
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDEO TERMS
(-', COSATi ] leld/Group
Pollution
Electrostatic Precipitators
Operations
Maintenance
Reliability
Particles
Dust
Pollution Control
Stationary Sources
Particulate
13B
131
14G
15E
14 D
11G
13. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS fTilit Rfpttrll
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (This pagej
Unclassified
22.
EPA Form 2210-1 (9-JJ)
-------
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
11
-------
-------
ABSTRACT
The manual focuses on the operation and maintenance (O/M) of typical
electrostatic precipitators (ESPs). It summarizes available information
on theory and design in sufficient detail to provide a basic background
for the O/M portions of the manual. Although O/M-related air pollution
problems cannot be completely eliminated, they can be minimized by the
conscientious application of a well planned O/M program. The causes
of such problems often vary widely, and their effects on deteriorating
performance may be direct, indirect, or synergistic. Process, particle,
mechanical, environmental, and gas-flow-dynamics factors dictate that
O/M programs and troubleshooting be approached from a total system
or process/plantwide viewpoint. The variable nature of these factors
also requires that O/M programs be individualized and specifically
tailored to the needs of the process aid installation served. Effective
O/M also affects equipment reliability, on-line availability, continuing
regulatory compliance, and regulatory agency/source relations. Lack
of timely and proper G/M leads to gradual equipment deterioration
which, in turn, increases the probability of equipment failure and
decreases both its reliability and on-line availability. The last two
items can decrease plant productivity if process operations are forced
to be curtailed or shut down for control equipment outages.
-------
-------
CONTENTS
Figures vii
Tables xii
1. Introduction 1-1
1.1 Scope and content 1-2
1.2 Intended use of manual 1-4
2. Overview of ESP Theory, Design, and O&M Considerations 2-1
2.1 Basic theory and principles of electrostatic
precipitation 2-1
2.2 ESP systems and components 2-21
2.3 ESP O&M considerations " 2-63
References for Section 2 2-67
3. ESP Performance Monitoring 3-1
3.1 Key operating parameters and their measurement 3-1
3.2 Instrumentation systems and components 3-12
3.3 Performance tests and parameter monitoring 3-21
3.4 Recordkeeping practices and procedures 3-29
4. Performance Evaluation, Problem Diagnosis, and Problem
Solutions 4-1
4.1 Performance evaluation 4-2
4.2 Problem diagnosis . 4-21
4.3 Corrective actions 4-41
References for Section 4 4-56
5. O&M Practices 5-1
5.1 Operating practices 5-1
5.2 Preventive maintenance 5-8
Preceding page Wank
-------
CONTENTS (continued)
Page
6, Inspection Methods and Procedures 6-1
6.1 Preconstruction and construction inspections 6-2
6,2 External inspection 6-11
6.3 Internal inspection 6-38
References for Section 6 6-72
7. Safety 7-1
7.1 Electrical hazards 7-1
7.2 Hopper entry . 7-6
7.3 Confined-area entry 7-9
7.4 Worker protection 7-15
8. Model O&M Plan 8-1
8.1 Management and staff 8-2
8.2 Maintenance manuals 8-5
8.3 Operating manuals 8-7
8.4 Spare parts 8-9
8.5 Work order systems 8-10
8.6 Computerized tracking 8-16
8.7 V-I curves 8-23
8.8 Procedures for handling malfunction 8-23
References for Section 8 8-26
Appendices A - ESP Applications in Cement Industry A-l
B - ESP Applications in Kraft Pulp Industry B-l
C - ESP Applications in Iron and Steel Industry C-l
D - ESP Applications in Municipal Incinerators D-l
E - Data Sheets and Example Checklists E-l
Glossary of Terminology G-l
VI
-------
FIGURES
Number Page
2-1 Basic Process Involved in Electrostatic Precipitation 2-2
2-2 Typical Curve Showing Efficiency as a Function of
Particle Size for an ESP Collecting Fly Ash 2-5
2-3 Resistivity of Several Dusts at Various Temperatures 2-7
2-4 Effect of Temperature on Collection Efficiency of
an ESP in a Cement Preheat Kiln Application 2-8
2-5 Effect of Gas Volume (reduced SCA) on Outlet Loading 2-10
2-6 Moisture Conditioning of Cement Kiln Dust 2-18
2-7 Flow Diagram of Sulfur-Burning Gas Conditioning System 2-20
2-8 Typical Wire-Weight Electrostatic Precipitator with
Top Housing 2-22
2-9 Typical Rigid-Frame ESP 2-23
2-10 Typical Rigid-Electrode-Type ESP 2-24
2-11 Concentric-Plate Wet ESP 2-26
2-12 Circular-Plate Wet ESP 2-27
2-13 Flat-Plate-Type Wet ESP 2-28
2-14 Parallel and Series Sectionalization of an ESP 2-32
2-15 Typical Roof and Casing Construction for a Rigid-
Frame ESP 2-35
2-16 Various Designs of Collection Electrodes 2-38
2-17 Comparison of Wire-Weight and Rigid-Frame ESP Designs 2-41
2-18 Typical Installation of Magnetic-Impulse, Gravity-
Impact Rapper 2-43
VI 1
-------
FIGURES (continued)
Number Page
2-19 Typical Pneumatic Rapper Assembly 2-44
2-20 Typical Electric Vibrator Type Rapper 2-45
2-21 Tumbling-Hammer Assembly for Use with Rigid-Frame
Discharge Electrode and Collecting-Surface
Rapping System 2-47
2-22 Vacuum System for Solids Removal 2-49
2-23 Pluggage of Perforated Plates at the Inlet to an ESP 2-51
2-24 . Examples of Two Inlet Plenum Designs that Generally
Cause Gas Distribution Problems 2-52
2-25 Two Methods of Spreading the Gas Pattern at Expansion
Inlet Plenums 2-52
2-26 Electrostatic Precipitator Power Supply Circuit 2-55
2-27 Typical ESP Control Cabinet and T-R Set Instrumentation 2-57
2-28 Typical Rapper Control Panel 2-59
3-1 Typical Cascade Impactor System 3-7
3-2 Sampling Train with Cascade Impactor 3-8
3-3 Typical T-R Set Control Panel 3-13
3-4 Example of Rapping Spikes on a Transmissometer Strip
Chart ' 3-18
4-1 Typical Plot Plan Layout for Recording ESP Operating
Data 4-6
4-2 Comparison of T-R Set Trip Patterns for Two Different
Days 4-7
4-3 Graphical Display of Plate Area Out of Service Over
a 3-Day Period 4-8
4-4 Graphical Plot of Secondary Current vs. Field for a
3-Chamber ESP 4-9
4-5a Example of Graphical Displays of Secondary Current
and Voltage vs. Day of Operation 4-10
-------
FIGURES (continued)
Number Page
4-5b Example of Graphical Displays of Secondary Current
and Voltage vs. Day of Operation 4-11
4-6 Typical Air-Load Test V-I Curve for an ESP on a
Recovery Boiler with Normal Dust Layer 4-13
4-7 Variation of Voltage Current Characteristics with
Collecting Plate Contamination 4-15
4-8 Effect of Dust Layer Thickness on V-I Curve 4-16
4-9 Comparison of Typical Air Load and Gas Load V-I
Curves 4-18
4-10 Comparison of V-I Curves for High Resistivity at
Air Load and Gas Load Conditions 4-18
4-11 V-I Curves Demonstrating Particulate Space Charge
Effect in a Cold Side Precipitator Collecting Fly Ash 4-19
4-12 Typical V-I Curves for a Cold Side ESP Operating at
Moderate Ash Resistivity 4-20
4-13 V-I Characteristics of Inlet Section of ESP Collecting
High Resistivity Ash 4-24
4-14 Air Load V-I Curve for ESP Field with Insulator
Tracking 4-27
4-15 V-I Curve for a Field with Excessive Wire Buildup 4-29
4-16 Air Load V-I Curve Pattern Generated by Alignment
Problems 4-39
5-1 Items that the Operator or ESP Coordinator Should
Record Daily 5-11
5-2 Items that the Operator or ESP Coordinator Should
Record Weekly 5-13
5-3 Items that the ESP Coordinator Should Record
Quarterly 5-15
5-4 Items that the Operator or'ESP Coordinator Should
Check Annually 5-21
IX
-------
FIGURES (continued)
Number Page
6-1 Rapper Air Line Trap and Filter 6-14
6-2 Rapper Boots 6-16
6-3 Penthouse Heater and Fan 6-18
6-4 Example of Rapper Boot Corrosion 6-20
6-5 Electrical Conduit Corrosion 6-22
6-6 Example of Secondary Current Pattern for Two
Chambers with Chamber A Having Maintenance
Problems that Limit Power Input 6-31
6-7 Corona Power Versus Collection Efficiency for a Coal-
Fired Utility Boiler 6-37
6-8 Velocity Distribution Patterns Resulting from Improper
Gas Distribution . 6-40
6-9 Chronic Distribution Plate Pluggage Problem 6-42
6-10 Example of Bottle Weight Deposits 6-44
6-11 Accumulation of Dust on Rapper Header Beams 6-46
6-12 Baffle Hopper Center!ine 6-48
6-13 Upper Baffle Used as Plate Suspension 6-49
6-14 Scraper Blade Passing Under a Baffle in a Deflected
Position 6-50
6-15 Drag Chain Assembly 6-52
6-16 Water Patterns Caused by Cold Conduit in the Penthouse
Roof 6-55
6-17 Resin-Type Insulator 6-57
6-18 Access to a Typical Insulator Enclosure and Support
Insulator for the Discharge Wire Frame 6-58
6-19 Insulator Compartment Showing Dust Deposits 6-59
6-20 Example of Plate Cracks 6-60
-------
FIGURES (continued)
Number Page
6-21 Examples of Alignment Rake and Anti-Sway Insulator 6-63
6-22 Falling Hammer Rapper 6-64
6-23 Photograph of Broken Plate Stabilizing Bracket 6-66
6-24 Lateral Movement of Upper or Lower Wire Frame
(End Elevation View) 6-6?
6-25 Longitudinal Movement of Upper or Lower Wire Frame
(Side Elevation View) 6-67
6-26 Lateral Movement of Upper or Lower Wire Frame (Plan View) 6-68
6-27 Rotation of Upper or Lower Wire Frame (Plan View) 6-68
6-28 Jack Screws 6-70
6-29 Notched Alignment Spacer 6-71
7-1 Control Cabinet Key Interlocks 7-3
7-2 T-R Set Ground Switch Key Interlocks 7-4
7-3 Ground Clips 7-7
7-4 Nomograph Developed by McKarns and Brief Incorporating
the Revised Fort Knox Coefficients 7-21
8-1 Organizational Chart for Centrally Coordinated ESP
O&M Program 8-4
8-2 Outline for ESP Maintenance Manual 8-6
8-3 ESP Operating Manual Outline 8-8
8-4 Example of Five-Level Priority System 8-13
8-5a Example of Work Order Form 8-17
8-5b Example of Work Order Form 8-18
8-6 Example of Department Profile 8-20
8-7 Example of Maintenance Summary 8-21
8-8 Example of Repair/Service History 8-22
-------
-------
TABLES
Kumber
2-1 ESP Characteristics Associated with Different Levels
of Resistivity 2-6
2-2 Input Data for EPA/SORI ESP Computer Model 2-14
2-3 Percent of Total Capacity Treated or Flange-to-
Flange Orders for Various ESP Applications 2-16
2-4 Operating Conditions of Electrostatic Precipitators 2-16
2-5 Reaction Mechanisms of Major Conditioning Agents 2-19
6-1 Erection Sequence 6-3
6-2 Key Process Parameters for Utility and Industrial
Boilers 6-28
6-3 Key Process Parameters for Cement Kilns 6-28
6-4 Key Process Parameters for Kraft Recovery Boilers 6-28
7-1 Effects of Various Levels of Oxygen on Persons 7-12
7-2 Allowable Concentrations for Entry into Confined Spaces 7-13
7-3 Applications Presenting Potential Eye Hazards 7-16
7-4 Maximum Permissible Sound Levels for Intermittent
Noise 7-17
7-5 ACGIH Threshold Limit Values for Nonimpulsive Noise 7-17
7-6 Explanation of Values in Belding and Hatch HSI 7-19
7-7 Heat Production for Various Levels of Exertion 7-22
7-8 Metabolic Body Heat Production as a Function of Activity 7-2?
-------
-------
ACKNOWLEDGE!
This manual was prepared for the U.S. Environmental Protection Agency's Air
and Energy Engineering Research Laboratory under contract No. 68-02-3919.
Michael F. Szabo was the project manager and provided technical coordi-
nation/editing In the preparation of this manual. The primary authors of
this manual were Gary L, Saunders, Ronald L. Hawks, and Michael F, Szabo.
Other contributing authors were David R. Dunbar and William F. Kemner.
Jack A. Wunderle provided senior review and assisted in developing the de-
tailed report outline. Marty H. Phillips provided editorial services and
the page layout design. Jerry Day coordinated typing and graphics for
the report and provided final review.
A review panel consisting of 34 members provided input throughout this
project. They are listed below in alphabetical order:
Name
Charles A. Altin
Ralph Altman
William S. Becker
Eli Bell
William S. Bellanger
Robert Brown
Steven Burgert
John M. Clouse
Jim Cummings
Duane Durst
Heinz Engelbrecht
David Ensor/David Coy
Kirk Foster
F.W. Giaccone
Wally Hadder
James Hambright
Norman Kutujian
Affijiatign
Ebasco Services, Inc.
Electric Power Research Institute
State and Territorial Air Pollution Program
Administrator
Texas Air Control Board
U.S. EPA Region 3
Environmental Elements Co.
East Penn Manufacturing Co.
Colorado Department of Health
U.S. EPA Office of Policy Analysis
U.S. EPA Region 7
Wheelabrator-Frye
Research Triangle Institute
U.S. EPA Stationary Source Enforcement .
Division
U.S. EPA Region 2
Virginia Electric Power Company
Pennsylvania Bureau of Air Quality Control
U.S. EPA Center for Environmental Research
Information
xm
-------
Name Affiliation
John Lytle Tennessee Valley Authority
Richard McRam'e Southern Company Services
Grady Nichols Southern Research Institute
Sid Orem Industrial Gas Cleaning Institute
John Paul Montgomery County, Ohio Regional Air
Pollution Control Agency
Charles Pratt U.S. EPA - Training
Richard Renninger National Crushed Stone Association
John Richards Richards Engineering
A.C. Schneeberger Portland Cement Association
Eugene J, Sciassia Erie County, New York Department of Envi-
ronment and Planning
Don Shephard Virginia Air Pollution Control Board
Lon Torrez U.S. EPA Region 5
William Voshell U.S. EPA Region 4
Glenn Wood Weyerhauser Corp.
Howard Wright U.S. EPA Stationary Source Enforcement
Division
Earl Young American Iron and Steel Institute
The comments of the review panel on the topics to be covered, the de-
tailed outline of the manual, and the draft manual were very helpful and have
contributed to the success of this project.
Finally, the cooperation and assistance of the project officer,
Louis 5. Hovis, in completing this manual are greatly appreciated.
xiv
-------
SECTION 1
INTRODUCTION
The success of an air pollution abatement program ultimately depends
upon effective operation and maintenance (O&M) of the installed air pollution
control equipment. Regardless of how well an air pollution control system is
designed, poor O&M will lead to the deterioration of its various components
and a resulting decrease in its particulatt removal efficiency.
Effective Q&H also affects equipment reliability, on-line availability,
continuing regulatory compliance, and regulatory agency/source relations.
Lack cf timely and proper O&M leads to a gradual deterioration in the equip-
ment, which in turn increases the probability of equipment failure and de-
creases both its reliability and ort-line availability. These latter two
items can decrease plant productivity if process operations ere forced to be
curtailed or shut down to minimize emissions during air pollution control
equipment outages. Frequent violations of emission limits car. result in more
inspections, potential fines for noncompliance, and in some cases, mandatory
shutdown until emission problems are solved.
This manual focuses on the operation and maintenance of typical electro-
static precipitators (ESP's). The overview presented in Section 2 summarizes
the available information on theory and design in sufficient detail to provide
a basic background for the Q&K portions of the manual. Numerous documents
are available if the reader desires a more rigorous treatment of ESP theory
and design.
Although O&M-related air pollution problems cannot be completely elimi-
nated, they can be minimized by the conscientious application of a well-
planned O&M program. The causes of such problems often vary widely, and
their effects on deteriorating performance may be direct, indirect, or syner-
gistic.
SECTION 1-INTRODUCTIQN
-------
Process, particle, mechanical, environmental, and gas-flow-dynamics
factors dictate that O&M programs and trouble-shooting actions be approached
from a total system or process/piantwide viewpoint. The variable nature of
these factors also requires that O&M programs be individualized and specifi-
cally tailored to the needs of the process and installation served. .
1.1 SCOPE AND CONTENT
Section 2 outlines the basic theory and principles of electrostatic
precipitation in sufficient detail to provide the background for and under-
standing of the manual sections that follow. It describes common types of
electrostatic precipitators (ESP's) and their components, notes typical ESP
applications, discusses factors that affect performance, and lists the limits
or constraints to ESP application. It also presents information on recent
developments, research, and trends in the use and application of ESP equip-
ment that are or may be useful in improving O&M. Discussions cover typical
causes of poor ESP performance and how to prevent or minimize them; design,
construction, and installation considerations that affect O&M; and the basic
elements of an O&M program designed to attain and maintain optimal ESP per-
formance.
Section 3 discusses performance monitoring as a major element in an O&M
program. Initial discussion centers on the key parameters that define inlet/
outlet gas stream conditions and ESP system operation. This discussion
covers what is to be measured (or monitored), where the measurements should
be taken, and the purpose of the measurements. Instrumentation systems are
described and evaluated with respect to their use as performance monitors.
Also discussed are the operating principle of monitors; their purpose; the
use, advantages, and limitations of the output data. Performance tests,
baseline assessments, and the role of parameter monitoring are briefly
discussed, and the importance of good recordkeeping practices and procedures
is stressed in regard to recordkeeping frequency, quality assurance, and
records maintenance and retention.
Section 4 describes the use of performance monitoring and other data in
the evaluation of control system performance, in the discovery of real or
SECTION 1-INTROOUCTION 1-2
-------
impending problems, and in the diagnosis and correction of causes of poor
performance. Initial discussion centers on how the various data collected
and recorded can be organized and used to track the performance of the over-
all control system and its components on both a short-term and long-term
basis. Tracking procedures and trends analysis methods for assessment of
current or impending performance deterioration are described. This section
then focuses on the determination of probable causes of ESP operating prob-
lems, malfunction, and deteriorating performance. Recognition of the symp-
toms of deteriorating performance and problem diagnosis are covered. The
actions generally required to restore the ESP to satisfactory operation are
discussed. Detailed instructions are not given because specific corrective
measures are highly system-dependent. Final discussion covers followup
techniques for determining the success of corrective actions and verifying
restored performance.
Section 5 presents guidelines for general O&M practices and procedures
that can be used to improve and sustain control equipment performance and
reliability. General guidance, rather than specific instructions, is given
because of the unique nature of the various ESP control systems and the
process streams they serve. This section prescribes the basic elements of
good operating practice and preventive maintenance programs that can be used
as the basis and framework for tailored, installation-specific programs. The
section addresses proper startup/shutdown procedures and normal operating
practices that prevent damage to equipment, minimize excessive emissions, and
optimize service and performance. Schedules are suggested for inspection/ob-
servation of equipment items and for performing preventive maintenance on ESP
systems and system components. These schedules indicate when and what to
look for and why.
Section 6 presents methods and procedures for the detailed inspection of
ESP systems and their components. Step-by-step procedures and techniques are
provided for conoucting external and internal inspections at both large and
small ESP installations. Inspection during the pre-operational construction
phase and the performance demonstration (baselining) period are addressed.
The portable instrumentation and safety equipment needs during inspection are
listed, and example inspection checklists are provided. Section 7 presents
SECTION 1-INTRODUCTION
1-3
-------
safety considerations and special precautionary measures for major source
installations.
Section 8 summarizes the more important elements of an adequate O&M
program by addressing the key items to be included in a model O&M plan.
These areas include management and staff responsibilities, maintenance and
operations manuals, spare parts, work order systems, computerized tracking,
V-I curves, and procedures for handling malfunctions.
The appendices include a glossary of terms and industry specific sec-
tions on cement (Appendix A), kraft pulp mill recovery boilers (Appendix B),
iron and steel industry (Appendix C), and municipal incineration (Appendix
D). A set of example checklists for recording various O&M activities as well
as example bid specification forms are presented in Appendix E.
1.2 INTENDED USE OF MANUAL
The increasing interest of both government and industry in proper O&M
has created a need for informative O&M manuals to assist both source and
control agency personnel. Obviously, no O&M manual for general application,
guidance, and use can provide a solution to all the the many and varied
O&M-related problems and combinations thereof. The objective of this manual
is to present the elements of a sound and systematic maintenance, operation,
surveillance, and diagnostic program that will promote the continuous, satis-
factory performance of electrostatic precipitators at a high level of avail-
ability. The technical materials, procedures, techniques, and practices pre-
sented can be readily incorporated or adapted to fulfill the basic require-
ments of a site-specific O&M program.
The advice and suggestions of an advisory panel of governmental and
industrial representatives were sought regarding the intended audience of the
manual, its topical content, and the depth and level of detail to be devoted
to each topic. Several panel members also provided technical data, general
information, and operating experience that were useful in the preparation of
the manual and critically reviewed draft sections as they were completed.
The manual is aimed at plant engineers, plant O&M personnel, and agency
inspectors. Its intent is to serve as an educational tool, not an enforcement
tool. Although the authors have focused on practical and proven O&M, they
SECTION 1-tNTRODUCTlQN 1-4
-------
have also integrated relevant data from the literature, equipment manufac-
turers' files, operating and service manuals, field service reports, plant/
equipment operating records, and case history experience. Emphasis is on
operating practices; preventive maintenance procedures; performance monitor-
ing; recordkeeping; and finding, diagnosing, and solving problems.
The plant engineer with responsibility for compliance with environmental
requirements will find the following of particular interest:
0 ESP design considerations to avoid O&M problems and facilitate
maintenance
0 Construction phase inspection to discover and/or prevent fabrica-
tion and installation errors
0 Performance monitoring/evaluation and trends analyses to discover,
diagnose, and correct real or impending problems
0 Good O&M practices and key elements in an O&M plan
0 Inspection methods and procedures
For plant operating and maintenance personnel, the topics of primary
interest include:
0 Operating practices and preventive maintenance
0 Performance monitoring/record keeping/evaluation
0 Malfunction and problem diagnosis/correction
0 Inspection methods and procedures
For agency inspectors, the subjects of primary interest include those
sections that address:
0 Inspection methods and procedures
0 Major elements of good O&M practice
0 Common compliance-related problems encountered at major sources
where ESP control systems are used
0 Parameter monitoring/recordkeeping/evaluation of system performance
0 ESP designs/installations that minimize O&M problems and provide
safe, unhindered access for maintenance and inspection
SECTION 1-INTRODUCTION
-------
The manual may also be used for a variety of secondary purposes. Col-
leges, universities, and technical schools that include air pollution courses
in their curricula have a need for O&M information. The manual can also
provide guidance on the general principles of O&M for new inspectors, plant
engineers-in-training, and operators. Equipment manufacturers may find the
manual useful as a guidance document regarding standard content and format in
the preparation of equipment O&M manuals. Plant engineering personnel and
consulting engineers will find the guidelines and principles set forth in the
manual useful in their preparation of specifications and operating procedures.
SECTION 1-INTRODUCTION
-------
SECTION 2
OVERVIEW OF ESP THEORY, DESIGN,
AND O&M CONSIDERATIONS
This section provides an overview of ESP theory, design, and O&M consid-
erations and sets the stage for more detailed treatment of QiM in later
sections of this manual.
2.1 BASIC THEORY AND PRINCIPLES OF ELECTROSTATIC PRECIPITATION
2.1,1 Operating Principles
The basic principles of the electrostatic precipitation process are 1)
development of a high-voltage direct current that is used to electrically
charge (transfer to) particles in the gas stream (almost all commercial ESP's
have negative polarity), 2) development of an electric field in the space
between the discharge electrode and the positively charged collection elec-
trode that propels the negatively charged ions and participate matter toward
the collection electrode, and 3) removal of the collected participate by use
of a rapping mechanism (or water flushing in the case of a wet collector).
These basic principles of the electrostatic precipitation process are illus-
trated in Figure 2-1.
The electrostatic precipitation process occurs within an enclosed cham-
ber; a high-voltage transformer (to step up the line voltage) and a rectifier
(to convert AC voltage to DC) provide the power input. The precipitation
chamber has a shell made of metal, tile, or Fiberglass Reinforced Plastic
(FRP). Suspended within this shell are the grounded collecting electrodes
(usually plates), which are connected to the grounded steel framework of the
supporting structure and to an earth-driven ground. Suspended between the
collection plates are the discharge electrodes (also known as corona elec-
trodes, which are insulated from ground and negatively charged with voltages
ranging from 20 kV to 100 kV. The large difference in voltage between the
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN. AND O&M CONSIDERATIONS 2-1
-------
EARTHED COLLECTOR
ELECTRODE AT
POSITIVE POLARITY
ELECTRICAL CHARGED
FIELD PARTICLE
DISCHARGE ELECTRODE
AT NEGATIVE POLARITY
UNCHARGED
PARTICLES
PARTICLES ATTRACTED
TO COLLECTOR ELECTRODE
AND FORMING DUST LAYER
HIGH VOLTAGE
CURRENT SUPPLY
ro
i
Figure 2-1. Basic processes involved in electrostatic precipitation.
Source: Lodge Cottrell
-------
negatively charged discharge electrode and positively charged collection
electrode creates the electric field that drives the negatively charged ions
and particles toward the collection electrode. The particles may travel some
distance through the ESP before they are collected or they may be collected
more than one time. Some particles lose their charge rapidly after being
collected and are lost through reentrainment in the gas stream.
The last segment of the process covers the removal of the dust from the
collection electrodes. In dry ESP's, this is accomplished by periodic strik-
ing of the collection and discharge electrode with a rapping device which can
be activated by a solenoid, air pressure, or gravity after release of a
magnetic field, or mechanically through a series of rotating cams, hammers,
or vibrators. The participate is collected 1n hoppers and then conveyed to
storage or disposal.
In wet ESP's, the collected participate is removed by an intermittent or
continuous stream of water or other conducting fluid that flows down over the
collection electrodes and into a receiving sump.
2.1.2 GasStream Factors Affecting Electrostatic Precipitation
Several important gas stream and particulate properties dictate how well
an ESP will collect a given particulate matter. They include particle size
distribution, flow rate, and resistivity which is influenced by the chemical
composition, density of the particulate, and process temperature. These
factors can also affect the corrosiveness of the dust and the ability to
remove the dust from the plates and wires. Following are brief discussions
of some of these properties.
Particle Size Distribution--
The coarser the size of a particle, the easier it is for an ESP to col-
lect it. Particles in the 0.2- to 0.4-ym diameter range are the most diffi-
cult to collect because in this size range, the fundamental field charging
mechanism gives way to diffusion charging by thermal ions (random collisions)
as a charging mechanism for very small particles.
A large percentage of small particles (<1 ym) in the gas stream can sup-
press the generation of the charging corona in the inlet field of an ESP, and
thus reduce the number of particles collected. Source personnel should have
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND OiM CONSIDERATIONS
2-3
-------
a good idea of the expected particle size of the participate before
purchasing an ESP, and the particle size distribution should be determined
for the full range of the operating conditions. Performance as a function of
particle size can be predicted by use of a computer model, as discussed later
in this section. Figure 2-2 presents a typical plot of particle size versus
efficiency.
Resistivity--
This parameter is a measure of how easy or difficult it is for a given
particle to conduct electricity. The higher the measured resistivity (the
value being expressed in ohm-cm), the harder it is for the particle to trans-
fer the charge. Resistivity is influenced by the chemical composition of the
gas stream and particulate, the moisture content of the gas stream, and the
temperature.
Resistivity must be kept within reasonable limits for the ESP to perform
8 ID
as designed. The preferred range is 10 to 10 ohm-cm. Table 2-1 presents
the effects of various levels of resistivity on ESP operating characteristics.
This discussion on resistivity applies to dry ESP's only; resistivity is
not important to the operation of a wet ESP.
Temperature--
The effect of temperature on resistivity and (ultimately on ESP collec-
tion efficiency) can be significant in some processes. Figure 2-3 illustrates
the variation in resistivity with temperature for several different industrial
dusts. Figure 2-4 shows the effect of temperature on ESP efficiency in a
cement preheat kiln application in which the gas stream is normally condi-
tioned and the temperature is reduced by a water spray tower. This figure
illustrates the effect of temperature that is allowed to rise. Although not
all temperature effects are this dramatic, the source should be aware of how
resistivity varies with temperature in their particular process application.
Gas Volume/Velocity--
An ESP will operate best when the gas volume keeps the velocity within a
typical range of 3.5 to 5.5 ft/s. Designers usually calculate a hypothetical
average value for gas velocity from the gas flow and the cross section of the
precipitator, ignoring the localized variances within the precipitator. The
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN. AND OSM CONSIDERATIONS . .
-------
99.9
PRECIPITATOR OPERATING CONDITIONS
TEMPERATURE 310° F
2
SCA 285 ft'/1000 acfm
CURRENT DENSITY 20 na/ft2
EFFICIENCY 99.6%
93
0.1 0.2 0.3 0.5 0.7 1 2
PARTICLE DIAMETER, ym
3457
Figure 2-2. Typical curve showing efficiency as a function of particle
size for an ESP collecting fly ashJ
(Permission granted by APCA)
2-5
-------
TABLE 2-1. ESP CHARACTERISTICS ASSOCIATED WITH
DIFFERENT LEVELS OF RESISTIVITY2
Resistivity level,
ohm-cm
ESP characteristics
Less than 10
8
108 to 1010
10
11
Sreiter than 10
12
(1) Normal operating voltage and current levels unless
dust layer is thick enough to reduce plate clearances
and cause higher current levels
(2) Reduced electrical force component retaining collect-
ed dust, vulnerable to high reentrainment losses
(3) Negligible voltage drop across dust layer
(4) Reduced collection performance due to (2)
(1) Normal operating voltage and current levels
(2) Negligible voltage drop across dust layer
(3) Sufficient electrical force component retaining
collected dust
(4) High collection performance due to (1), (2), and (3)
(I) Reduced operating voltage and current levels with
high spark rates
(2) Significant voltage loss across dust layer
(3) Moderate electrical force component retaining
collected dust
(4) Reduced collection performance due to (1) and (2)
(1) Reduced operating voltage levels; high operating
current levels if power supply controller is not
operating properly
(2) Very significant voltage loss across dust layer
(3) High electrical force component retaining collected
dust
(4) Seriously reduced collection performance due to (1),
(2), and probable back corona
Typicalvalues
Operating voltage :
Operating current density:
Dust layer thickness :
30 to 70 kV, dependent on design factors
5 to 50 nA/cm
i to 1 in.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN. AND O&M CONSIDERATIONS
2-6
-------
10'
10
12
12 10
10
10-
10C
10
I I I I
7_
0 50 TOO 150 200 250 300 350
TEMPERATURE,°F
Figure 2-3. Resistivity of several dusts at various temperatures.3
2-7
-------
100
M 98
•5
5 96
1 94
LL.
U.
LU
s 90
I—
<-> 88
u
o 86
84
82
200 300 400 500 600 700
OPERATING TEMPERATURE,°F
800
Figure 2-4. Effect of temperature on collection
efficiency of an ESP in a cement preheat kiln application.1*
(Permission granted by All is-Chalmers, Corp.)
2-8
-------
primary importance of the hypothetical gas velocity is to minimize potential
losses through rapping and reentrainment. Above some critical velocity,
these losses tend to increase rapidly because of the aerodynamic forces on
the particles. This critical velocity is a function of gas flow, plate
configuration, precipitator size, and other factors, such as resistivity.
Figure 2-5 illustrates the effect of higher-than-optimum gas volume, using an
outlet loading of 0,01 gr/acf is used as the base point. As shown, a gas
volume of 10 percent over design increases the outlet loading by 50 percent,
to 0.015 gr/acf.
Many ESP's are designed with some redundancy in treating the expected
amount of flue gas. Nevertheless, the source should be aware of the design
limits of the gas volume and take this information into account when consid-
ering process changes that will increase gas flow. Excessive air inleakage
can also cause higher-than-expected gas volumes, but this problem can be
remedied by proper design and maintenance of seals and expansion joints.
A final consideration is that of low gas volumes. If velocity is al-
lowed to drop below 2 to 3 ft/s, performance problems can occur as a result
of maldistribution of gas flow and dropout of dust in the ducts leading to
the ESP. Building sufficient flexibility into the design of the ESP (e.g., a
dampering system that allows a portion of the ESP to be closed off during
periods of low gas flow) can minimize the problem.
Fuel-Related Parameters--
A decrease in the sulfur content of coal will generally result in an
increase in resistivity and a reduction in the collection efficiency of the
ESP. A switch from 2 percent Eastern bituminous coal to 0.5 percent Western
subbituminous coal can cause an ESP designed for 99.5 percent collection
efficiency to operate at 90 percent or less. Adequate amounts of certain
chemical constituents of the particulate (e.g., sodium and iron oxide) can
reduce resistivity and improve performance. Thus, is imperative that the
source obtain an analysis of the ash or process dust and be prepared to
design the ESP based on the worst fuel or process dust expected.
Because of its low resistivity, carbon is another constituent that can
reduce ESP performance. The carbon particle is conductive, but it loses its
charge quickly and becomes reentrained from the collection plates. This is
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS -
-------
_ -Aw
P=e
0,03
10.025
•a
b 0.02
QJ
5 0,05
o 0.01
§0.005
o
90 95 TOO 105 110
DESIGN GAS VOLUME,%
115 120 125
Figure 2-5. Effect of gas volume (reduced SCA) on outlet loading.4
(Permission granted by Allis-Chalmers, Corp.)
2-10
-------
aggravated by the fact that carbon is lighter than other constituents in the
flue gas. This is a problem on coal-fired stoker boilers and coke oven
underfire applications, for example, where the combustible content of the ash
may range from 25 to 50 percent. The ESP's for these units are larger and
have lower face velocity than those for applications where resistivity levels
are normal .
2.1.3 ESP Design Equations and Hodels
Over the past 85+ years of applied ESP technology, a number of techniques
have been used to estimate the amount of collection area required to produce
the desired collection efficiency. All of these techniques, however, are
based on the original Deutsch-Anderson equation, which is as follows:
n = 1 - e" W¥ (Eq. 1)
where n = ESP collection efficiency
A = Total collection electrode surface area
V = Gas flow rate
W = Migration velocity of the particles
e = Base of natural logarithms
The main problem with the Deutsch-Anderson equation (and the reason so
many attempts have been made to modify it) is that it does not take into
account the fact that 1) industrial process particulate matter is not mono-
disperse, and 2) the particle size distribution of dust suspended in the gas
stream (and thus the migration velocity, i.e., how quickly the charged gas
particles move to the grounded collection electrode) changes as the gas
stream moves through the ESP. Also the equation does not account for other
nonideal occurrences (such as gas turbulence and particle reentrainment) and
assumes uniform electrical conditions throughout the ESP. When w is deter-
mined empirically, of course, these nonideal factors are accounted for.
The most well-known and frequently used variation on the Deutsch-Anderson
equation is the Matts-Ohnfeldt version, which is derived as:
n - 1 - e- (Eq. 2)
SECTION 2-OVERVIEW OP ESP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-11
-------
where n = ESP collection efficiency
A = Total collection electrode surface area
v = Gas flow rate
w, = Modified migration velocity of particles
k = Dimensionless parameter
e = Base of natural logarithms
The value of the exponent k depends on the process being evaluated (most
commonly 0.5 for typical fly ash applications). When k = 1, the Matts-Ohnfeldt
again becomes the Deutsch-Anderson equation. The w. value in the Matts-Ohnfeldt
equation can be assumed to be independent of charging voltage and current
levels and of particle size distribution within an ESP as the gas stream moves
through it. If other gas stream changes occur, however, such as chemical
composition, resistivity, or particle size distribution, w, will be affected
just as the conventional w is affected. Section 3 presents a discussion of a
modified form of these relatively simple equations that a source operator will
find very useful in estimating the current performance of an ESP in comparison
with a baseline estimate by using the same equation in conjunction with a
stack test.
Although the above equations form the basis of most sizing techniques,
the total sizing procedure is much more involved. Each manufacturer has its
own method of sizing, often involving the use of computer models, and always
involving the use of some judgment because no model is capable of accounting
for all of the variables that affect ESP performance.
Buyer participation also varies. Whereas some buyers rely heavily on the
ESP manufacturer for determining proper sizing, in recent years other buyers
have begun to take a more active role either directly or through use of their
A/E. In fact, the A/E may make the final decision on what the minimum size of
the ESP will be, as well as the types of components to be used.
EPA/Southern Research Model--
The best known and most widely used performance model for ESP's is one
developed and refined by Southern Research Institute (SoRI) for the U.S.
Environmental Protection Agency over the past 10 years.
SECTION 8-OVERVIEW OF ESP THEORY. DESIGN. AND O&M CONSIDERATIONS
2-12
-------
The EPA/SoRI ESP model " is a valuable tool for examining and evaluat-
ing 1) gas/particulate characteristics, 2) design specifications, and 3) low
or reduced process operating conditions that affect precipitator performance.
Based on the inputs presented in Table 2-2, the model also can study typical
problems and deficiencies of precipitator performance in light of actual
performance results. The model is designed to accomplish the following;
0 Predict collection efficiency as a function of particle size, elec-
trical operating conditions, and gas/particulate properties.
0 Calculate clean-plate, clean-air, and voltage-current character-
istics.
0 Determine particle charging levels by unipolar ions.
0 Use empirical correction factors to adjust migration velocity
results.
0 Account for nonideal effects of gas distribution, gas bypass, and
reentrainment from nonrapping sources.
0 Account for rapping reentrainment.
0 Predict trends caused by changes in specific collection area, volt-
age, current, particulate loading, and particle size.
Given accurate input data, the model usually can estimate emissions
within ±20 percent of measured values. Such predictions are possible because
a relationship can be established between secondary voltage and current
levels (corona power) and emission levels through iterative computation by
the model. Once the empirical factors are adjusted and agreement is reached,
reasonable estimates of emission levels under other ESP conditions can be
made. Although this model is obviously not simple, its complexity has been
reduced sufficiently for it to be used with a programmable -calculator. The
calculator version, although not always as accurate as the full-size model,
is still a useful tool, especially for installations that do not have complex
O&M problems,
Other Sizing Techniques—
The use of pilot-scale ESP's can help in sizing a full-scale unit and
also for modeling flow patterns. The main problem with use of pilot-scale
SECTION 2-OVEHVIEW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-13
-------
TABLE 2-2. INPUT DATA FOR EPA/SORI ESP_COHPUTER HQDEL7"10
ESP specifications Gas/particulate specifications
Estimated efficiency Gas flow rate
Precipitator length Gas pressure
Superficial gas velocity Gas temperature
Fraction of sneakage/reentrainment Gas viscosity
Normalized standard deviation of gas Particulate concentration
velocity distribution
Particulate resistivity
Number of stages for sneakage/reentrainment
Particulate density
Number of electrical sections in direction
of gas flow Particle size distribution
For each electrical section Dielectric constant
Length • Ion speed
Area
Applied voltage
Current
Corona wire radius
Corona wire length
Wire-to-wire spacing (1/2)
Wire-to-plate spacing
Number of wires per linear section
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN. AND O&M CONSIDERATIONS 2-14
-------
units is the scale-up factor because pilot-scale units usually perform better
than full-scale units. This presents some uncertainty in the choice of
proper scale-up factor.
Combustors are also useful for characterization of potential coal to be
used in boilers. The installations available in the United States are small,
and the data they provide are qualitative. Much additional information is
needed for use with full-scale units.
2.1.4 ESP Applications
Dry ESP's are used in all basic industries and also in some specialized
applications. The electric utility industry is the biggest user, but other
large users include the cement industry (rotary kilns), the pulp and paper
industry (kraft recovery boilers, coal, and hogged fuel boilers), municipal
incinerators, ferrous metallurgical applications (BOF, sinter, scarfing),
nonferrous metallurgical applications (copper, lead, zinc, and aluminum
smelting), the petroleum industry (fluid catalytic crackers, detarring), the
chemical industry (sulfuric acid plants), and industrial boilers of all
types. The particulate matter from these sources can generally acquire an
electrical charge quite well, and an ESP can be designed to treat large gas
volumes at high temperatures (up to 2000°F) and several atmospheres of pres-
sure.
Table 2-3 summarizes the percent of total capacity treated (1912 to
1969) and percent of total orders (1971 to 1980) for major areas of ESP
application. These data show that ESP applications in the iron and steel and
rock products industries have decreased in importance, whereas those in the
pulp and paper and miscellaneous categories have increased in importance.
Table 2-4 presents operating conditions of ESP's in major application areas.
These data show the wide range of temperature, pressure, and particulate
concentration under which an ESP can operate.
Problem Applications--
Wet ESP's are generally used for applications where the potential for
explosion is high (closed-hood BOF in the steel industry), where particulates
are very sticky, and for high-resistivity applications. Where moisture or
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-15
-------
TABLE 2-3. PERCENT OF TOTAL CAPACITY TREATED
(1912-1969) OR FLANGE-TO-FLANGE ORDERS (1971-1980) .,
FOR VARIOUS ESP APPLICATIONS (UNITED STATES AND CANADA)1^
Application
Utility industry
Pulp and paper industry
Iron and steel industry
Rock products industry
Chemicals industry
Nonferrous metals
Petroleum Industry
Miscellaneous
1912-1969
78.6
5.1
6.4
6.4
1.0
1.5
0.7
0.3
1965-1969
88.9
4.5
3,4
2.1
0.4
_
_
0.7
1971-1975
77.3
6.7
2.5
2.6
-
5.0
_
5.8
1976-1980
80.3
7.2
2.9
3.2
-
1.0
- (
5.4
1980
0.4)
Source: Research-Cottrell, Inc.
Somerville, NJ.
TABLE 2-4. OPERATING CONDITIONS OF ELECTROSTATIC PRECIPITATORS
12
Application area
Electric utility
Pul p and paper
Iron and steel
Rock products
Chemical process
Nonferrous metals
Petroleum
Refuse combustion
Miscellaneous
Temperature,
°F
225
225
70
350
80
70
70
450
90
- 900
- 375
- 600
- 700
- 800
- 1100
- 850
- 550
- 1700
Pressure,
psia
14.7
14.7
14.7
14.7
14.7
14.7
7.5 - 164
14.7
65 - 825
Concentration,
gr/scf
1.5
1.0
0.01
3.7
0.2
0.01
0.8
0.5
lO'5
Efficiency,
/a
- 7.5 90
- 9.0 90
- 3.0 85
- 156
- 50
- 45
- 40
- 4.0
- 3.0
92
85
90
80
95
95
- 99.6
- 99.5
- 99.8
- 99.9
- 99.9
- 99.9
- 99.7
- 99.2
- 99.5
Source: Research-Cottrell, Inc.
Somerville, NJ.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND OiM CONSIDERATIONS
2-16
-------
chemical substances are needed to increase the conductivity of the particu-
late (mostly low-sulfur coal applications), dry ESP's can be equipped with a
conditioning system.
Moisture not only reduces the resistivity of most dusts and fumes at
temperatures below 250° to 300°F, but also greatly enhances the effect of
chemical conditioning agents. Moisture conditioning is performed by steam
injection, water sprays, or wetting the raw materials before they enter the
ESP. The lower the gas stream temperature, the better the conditioning
effect is. Figure 2-6 presents an example of this effect for cement kiln
dust. Proper spray nozzle design, adequate chamber space, and proper temper-
ature control are imperative; otherwise, too much water can be provided and
the particulate matter will cake on the interior of the ESP.
Chemical conditioning agents that are in use or under study include
sulfur trioxide, sulfuric acid, ammonia, ammonium sulfate, triethyl amine,
compounds of sodium, and compounds of transition metals. Although high-
resistivity problems are most commonly treated using conditioning agents,
low-resistivity problems are also treatable (e.g., ammonia has been uti-
lized). The ppm of these compounds required to provide the desired result is
highly dependent on the application. Table 2-5 lists the conditioning agents
and their mechanisms of operation.
In the United States, sulfur trioxide (SO,) and sulfuric acid are the
most successful and widely used conditioning agents on coal-fired utility
boilers. The primary mechanism is condensation or adsorption on ash. The
handling of both of these highly corrosive and toxic liquids is different
because they must be vaporized before they are injected into the flue gas.
Figure 2-7 shows a typical SO, conditioning system.
Although flue gas conditioning often improves ESP performance by reduc-
ing dust resistivity or through other mechanisms, conditioning agents should
not be considered cure-alls for ESP problems. For example, they cannot
correct problems associated with a poorly designed ESP, poor gas distribu-
tion, misaligned plates and wires, or inadequate rapping. Thus, any existing
installation should be carefully evaluated to determine that poor ESP per-
formance is due entirely to resistivity problems. Conditions for injection
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND OtM CONSIDERATIONS
2-17
-------
10
12,
10
11
E
O
>
i—i
I—
C/l
10
10
10"
200 300 400 500 600 700
TEMPERATURE, °F
Figure 2-6. Moisture conditioning of cement kiln dust.
(Permission granted by APCA.)
13
2-18
-------
TABLE 2-5. REACTION MECHANISMS OF MAJOR CONDITIONING AGENTS
14
Conditioning agent
Mechanism(s) of action
Sulfur trioxide and sulfuric
acid
Ammonia
Ammonium sulfate'
Triethyl amine
Sodium compounds
Compounds of transition
metals
Potassium sulfate and
sodium chloride
Condensation and adsorption on fly ash surfaces;
may also increase cohesiveness of fly ash.
Reduces resistivity
Mechanism is not clear; various ones proposed:
Modifies resistivity
Increases ash cohesiveness
Enhances space charge effect
Little is known about the actual mechanism;
claims are made for the following:
Modifies resistivity (depends upon injection
temperature)
Increases ash cohesiveness
Enhances space charge effect
Experimental data lacking to substantiate which
of these is predominant
Particle agglomeration claimed; no supporting
data
Natural conditioner if added with coal. Resis-
tivity modifier if injected into gas stream
Postulated that they catalyze oxidation of S02
to S03; no definitive tests with fly ash to
verify this postulation
In cement and lime kiln ESP's:
Resistivity modifiers in the gas stream
NaCl--natural conditioner when mixed with
coal
If injection occurs at a temperature greater than about 600°F, dissociation
into ammonia and sulfur trioxide results. Depending upon the ash, S02 may
preferentially interact with flyash as S03 conditioning. The remainder re-
combines with ammonia to add to the space charge as well as increase the
cohesivity of the ash.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-19
-------
LIQUID SULFUR
CONTROLLED TO
BOO' - 825* F
CONDITIONED
FLUE GAS TO
PRECIPITATOR
figure 2-7. Flow diagram of sulfur-burning flue gas conditioning system.
(Courtesy of Wahlco, Inc.)
2-20
-------
of a chemical conditioning agent should also be carefully studied. Inade-
quate mixing of the conditioner can cause performance to be below that
expected.
The use of wet ESP's also can overcome resistivity problems. A wet ESP
can be used alone or in conjunction with wet scrubbers, which remove both
particulate and gaseous pollutants (such as fluorides). Wet ESP's are used
to control a variety of industrial processes, including sulfuric acid mist;
coke oven off-gas, blast furnaces, detarring, basic oxygen furnaces, scarf-
ers» and cupolas in the iron and steel industry, and aluminum potlines. The
conditioning of the incoming gas stream and continual washing of the internal
components with water eliminate resistivity and reentrainment problems. Some
gaseous pollutant removal can also occur, but such removal is limited by the
solubility of the gaseous component of the wash liquor. Organics that con-
dense are also collected in a wet ESP. Significant collection of submicron
particles is also possible with a wet ESP.
2.2 ESP SYSTEMS AND COMPONENTS
This section discusses the major types of ESP's and describes their
major components in terms of typical design features and construction proce-
dures that are related to the operation and maintenance of the equipment over
its useful life. New design and research and development are also discussed
briefly.
2.2.1 Dry ESP's
The major distinction between different types of dry ESP's is the type
of corona discharge system used. The three most common discharge electrode
configurations used are 1} wires suspended or tensioned by weights (weighted
wire), 2) wires suspended in a rigid frame, and 3) rigid electrode. The
rigid electrode does not utilize wires but creates a corona on spikes welded
or otherwise attached to a rigid mast support. Figures 2-8 through 2-10 show
a typical wire-weight (American type) ESP, a rigid-frame (European type) ESP,
and rigid electrode type ESP, respectively. Other differences in the design
of wire-weight and rigid-type ESP's are discussed under the appropriate
component (e.g., rapping equipment, etc.).
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN. AND O&M CONSIDERATIONS
2-21
-------
BUS OUCT
INSULATOR
COMPARTMENT
VENTILATION SYSTEM
HIGH VOLTAGE
SYSTEM RAPPER
INSULATOR
COMPARTMENT
RAILING
HIGH VOLTAGE
SYSTEM UPPER
SUPPORT FRAME
24 in.
MANHOLE
TRANSFORMER/RECTIFIER
REACTOR
PRIMARY LOAD
RAPPER
'CONTROL PANEL
ELECTRICAL
EQUIPMENT
PLAT FOR*
COLLECTING
SURFACES
HIGH VOLTAGE
ELECTRODES
WITH HEIGHT
COLLECTING
SURFACE
RAPPERS
HOPPER
Figure 2-8, Typical wire-weight electrostatic precipitator
with top housing.
(Courtesy of Western Precipitation)
2-22
-------
WLAHIUI HOOF
BOX omoens
BOX GIRDER ACCESS MATCH
DISCHAdGE
ELECTRODE
RAPPER MOTOR
INSULATION
TRANSFORMER
HECTIFIEH
HOT ROOF
CASINO
CONSTRUCTION
COLLECTING
ELECTRODE
RAPPER MOTOR
GAS
DISTRIBUTION
PLATES
COLLECTING
ELECTRODES
HOPPERS
COLLECTING
ELECTRODE
HAPPERS
DISCHARGE
ELECTRODES
Figure 2-9. Typical rigid-frame (European type) ESP.
(Courtesy of Wheelabrator Frye)
-------
Figure 2-10. Typical rigid-electrode-type ESP
(courtesy of Environmental Elements Inc.).
2-24
-------
The wire-weight design was the typical American ESP from the late 1950's
to the mid-1970's. Since then, users have shown an overall preference
(heavily influenced by the utility industry) for rigid-type ESP's, because of
their more conservative and higher cost design, the ability to provide longer
discharge electrodes (generally greater than 36 ft) without an increase in
breakage rate, and their ability to provide higher rapping force without
damage to internal components (especially important in removing the highly
resistive fly ash generated by low-sulfur coal). U.S. manufacturers are now
offering ESP's of rigid-frame design and some of the so-called "hybrid"
design in which a rigid-type electrode is combined with an American-type
rapping system (e.g., magnetic impulse, gravity impact, pneumatic impact)
instead of the European-type mechanical hammers. In 1980, more than 75
12
percent of all ESP orders in the United States were for rigid-frame ESP's.
One other application of the wire-weight ESP (and to a very limited
extent, the rigid-frame-type ESP), the hot-side ESP, is used primarily in the
electric utility industry. This unit is placed upstream of the air pre-
heater, where temperatures range from 600° to 700°F. Normally, the higher
gas temperature dramatically reduces the resistivity of low-sulfur coal ash,
which makes the installation of a hot-side ESP more economical than the larg-
er size cold-side ESP that would be needed on the same type of installation.
In the early to mid-1970's, approximately 100 hot-side units were installed
on utility boilers firing low-sulfur coal. Results showed that units firing
coal with low sodium content experienced high resistivity and reduced per-
formance. It was found that sodium conditioning reduced resistivity and
improved performance on units firing low-sodium coal; however, no hot-side
ESP's have been sold in the United States since 1977.
Wet ESP's—
The major differences in the types of wet ESP's available today are as
follows: the shape of the collector, whether treatment of the gas stream is
vertical or horizontal, whether incoming gas is preconditioned with water
sprays, and whether the entire ESP is operated wet. Figures 2-11 through
2-13 show three different types of wet ESP's, two of the circular-plate/pipe
SECTION 2-OVERV1EW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-25
-------
CLEAN GAS
DISCHARGE
HOOD
PRECIPlTATtm
PBECONDITIONER
WATER
DISTRIBUTOR
GAS INLET
PRECONDITIQNER
DRAIN
COLLECTION
CYLINDER
EMITTING
ELECTRODE
VENTURI/DRAIN
CUTTER
STR&IGHTENING
VANES
ACCESS MANWAV
PRECIPITATOR
ORA;N
Figure 2-11. Concentric-plate wet ESP.
(Courtesy of Fluid Ionics, Inc.)
2-Z6
-------
Figure 2-12. Circular-plate wet ESP (detarring operations),
(Courtesy of Environmental Elements, Inc.)
2-27
-------
GAS
OUTLET
HOOD
PBECIPJTATOR
EMITTING
ELECTRODE
VENTURI/DRAIN
GUTTER
PRECONDITIONER
ENTRY INLET
ACCESS
MANWAY
HIGH VOLTAGE
INSULATOR
PHECIPITATOB
DRAIN
PRECONDITIONED
DRAIN
ACCESS
MANWAY
GAS
INLET
Figure 2-13. Flat-plate-type wet ESP.
(Courtesy of Fluid Ionics, Inc.)
2-23
-------
variety and one of the square or rectangular flat-plate type. Casing can be
constructed of steel or FRP, and discharge electrodes can be carbon steel or
special alloys, depending on the corrosiveness of the gas stream.
In circular-plate wet ESP's, the circular plates are irrigated continu-
ously; this provides the electrical ground for attracting the particles and
also removes them from the plate. It can generally handle flow rates of
30|000 to 100,000 cfrn. Preconditioning sprays remove a significant amount of
participate by impaction. Pressure drop through these units usually ranges
from 1 to 3 inches of water.
Rectangular flat-plate units operate in basically the same manner as the
circular-plate wet ESP's. Water sprays precondition the incoming gas and
provide some initial particulate removal. Because the water sprays are
located over the top of the electrostatic fields and, collection plates are
also continuously irrigated. The collected particulate flows downward into a
trough that is sloped to a drain for treatment. The last section of this
type of wet ESP is sometimes operated dry to remove entrained water droplets
from the gas stream.
The preconditioner liquor and the ESP liquor are generally treated
separately so that the cleanest liquor can be returned to the ESP after
treatment.
2.2.2 BuyerResponsibilities in ESP Specificationand Installation
The buyer should specify the supplier's scope of participation, i.e., to
what extent, it includes design and furnishing of the material, equipment, and
tools necessary to install the ESP (e.g., turnkey or equipment only). The .
scope may include items that are not specifically mentioned in the specifi-
cations but that may be necessary to complete the work. Receiving, unload-
ing, handling, and storage of materials, equipment, and accessories and their
complete erection should be covered, as well as a schedule of these events.
Startup procedures and return of the site to its original condition should
also be addressed.
The buyer should describe the location, provide drawings of the existing
system or planned site, and indicate the proposed location of the new ESP's.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND O»M CONSIDERATIONS
2-29
-------
The scope should also include the condition of existing gas ducts, breeching,
and stacks. In addition, the buyer should provide a list of preferred mate-
rials or designs.
The plant engineer should be knowledgeable regarding functional ESP
requirements and use this knowledge to follow the design, fabrication, and
erection phases of the program. This attention to detail throughout the
various phases will pay for itself by reducing the chance of forced shutdown
to correct problems that could have been solved before the ESP was con-
structed and put into operation. The buyer should thoroughly examine the
following: 1} the supplier's system standards for fabrication and erection
dimensional tolerances', 2) procedures for qualifying the subcontractors; 3)
quality control and inspection procedures in the field; 4) the caliber of
individuals designated as construction supervisors, advisors, or service
engineers; and 5) the organization's work load (to assess the level of
competence that will be applied the design and construction of the ESP).
The buyer should also determine what services the seller offers in the
way of the training of operators, providing comprehensive O&M manuals (in-
cluding recommended spare parts), assisting in the actual startup of the new
ESP, and providing followup maintenance and/or troubleshooting expertise
after the warranty period is over. Although the need for these services will
depend on the expertise of the buyer, it is important to determine what the
seller is capable of offering.
2.2.3 Component Design, Construction, andInstallation Considerations With
Respect to Operation and Maintenance'"
Because ESP components vary greatly among manufacturers, it is impos-
sible to discuss each design. Instead, general information on each component
is presented to assist the plant engineer in determining what to look for in
the design and installation of various major components.
Structural Sizing--
The use of various techniques for sizing ESP's was discussed in Section
2.1. Once the required collection area has been calculated, one of the first
SECTtON 2-OVERVIEW OF ESP THEORY. DESIGN. AND O&M CONSIDERATIONS
2-30
-------
structural parameters to be determined is the width of the ESP. This value
depends on the total number of ducts, which is calculated as follows;
TV) (so) (PH)
Total number of ducts -
where acfm = volumetric throughput of gas, actual cubic feet per minute
TV = treatment velocity of gas, ft/s
PS = plate spacing, ft
PH = plate height, ft
Treatment velocity, which is a function of resistivity and particle size of
the dust, ranges from 3.0 to 5.5 ft/s in most applications.
The ESP manufacturer determines plate spacing (based on experience with
different types of dust) by velocity distribution across the precipitator and
by the plate type. Plate spacing usually ranges from 6 to 15 inches, and
9-inch spacing is most common in the United States (weighted wire) and 12 to
14 inches for rigid-type ESP; however, ESP designers are now showing a great
deal of interest in larger spacings.
In the selection of plate height, consideration must be given to simul-
taneously maintaining the required treatment velocity and an adequate aspect
ratio and to the limitations posed by structural stability and overall de-
sign. Aspect ratio is defined as the ratio of the effective length to the
height of gas passage. Although space limitations often determine ESP dimen-
sions, the aspect ratio should be high enough to allow collection of reen-
trained dust carried forward from inlet and middle sections. In practice,
aspect ratios range from 0.5 to 1.5. For efficiencies of 99 percent or
higher, the aspect ratio should be at least 1.0 to 1.5 to minimize carryover
of collected dust, and some installations may approach 2.0.
The total number of ducts dictates the width of the box. Mechanical
sectionalization across the gas flow (parallel) separates the ESP into
chambers (each of which can be isolated from the other). Mechanical sectional'
ization in the direction of gas flow (series) separates the ESP into fields.
Figure 2-14 shows both parallel and series mechanical sectlonalization.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND OiM CONSIDERATIONS
2-31
-------
ce
UJ
eo
UJ
cn
cr
LU
CO
4TH FIELD
3RD FIELD
2ND FIELD
1ST FIELD
(DIRECTION OF GAS FLOW'
INDICATED BY ARROW)
Figure 2-14, Parallel and series sectionalization (mechanical
of an ESP.'6
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-32
-------
Energization and reliability can be improved by limiting the total
number of ducts per unit. The number of chambers in an ESP depends on the
total number of ducts, as determined from Equation 3. The required number of
ESP's depends on the degree of reliability needed, space limitations at the
site, and the relative ease of distributing the effluent gas to the ESP's.
The length of the ESP can be calculated by use of the following equa-
tion:
Treatmpnt ipnath - -, Total collecting plate area ,_ .,
Treatment length - (No. of ESPts)(chambers/ESP)(ducts/chamber)(PH)(2) (E^ 4)
The design treatment length Is determined by the selection of an integer
value of standard section lengths from those offered by the ESP manufacturer.
For example, if four sections are required, two of one length and two of
another, the structural considerations (such as hopper spans) and some perform-
ance criteria determine the positioning of the sections in the direction of
the gas flow. The size of the transformer rectifier (T-R) sets is selected
to provide lower current density at the inlet, where corona suppression is
likely to decrease collection efficiency, and higher current density at the
outlet, where although the percentage of fine particles is greater, the
overall gas stream is cleaner, which allows high current density.
Mechanical sections result from the chamber and series sectionalization
of the ESP. Hopper selection, in turn, is based on the size of these mechan-
ical sections.
Casings —
The ESP's casing is gas-tight and weatherproof. The inlet and outlet
connections, the shell, hoppers, inspection doors, and insulator housing are
the major casing parts. The shell and insulation housing form a grounded
steel chamber that completely encloses all the high-voltage equipment to
ensure the safety of personnel.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-33
-------
The casing for most applications is fabricated of a steel suitable for
the application (especially for the particular process and heat range). The
shell is reinforced to handle maximum positive or negative environmental
stresses, such as those imposed by wind, snow, and earthquake. Figure 2-15
shows the typical casing and roof construction.
If inspection and maintenance of both collection and discharge electrode
systems are made through the roof casing (usually one door per dust-plate
section), as it is in many wire-weight designs, the operator must crawl under
the casing stiffeners and over and around the suspension hardware. Because
the floor of the crawl space is usually the discharge electrode support
system, inspection may be difficult. Thus, if the buyer specifies minimum
clearances, he/she will eliminate the tendency among manufacturers (for
competitive reasons) 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, cleaning, and
general maintenance of the ESP's internals, but these additional shell pene-
trations also increase the potential for inleakage and corrosion.
If the design provides separate insulator compartments, a roof casing
covered by insulation, and a walkway surface deck plate, the plate must be
watertight and sloped for drainage, and the entire structure must be ade-
quately supported either through rigid insulation or metal framing. Clear-
ances must also be provided for movement of the insulator compartments,
rapper shafts, or any other equipment or supports that will move as the ESP
casing and structure expand.
The provision of a weather enclosure or superstructure should be encour-
aged to facilitate routine inspections. Care must be taken to provide for
differential movements in the casing and proper ventilation.
Dust Hoppers--
Hoppers collect the precipitated dust and deliver it to a common point
for discharge. The most common hoppers are pyramidal and converge to a round
or square discharge. If the dust is to be removed by screw conveyor, the
hopper usually converges to an elongated opening that runs the length of the
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN. AND OSM CONSIDERATIONS
2-34
-------
ROOF CONSTRUCTION
CASING CONSTRUCTION
Figure 2-15. Typical roof and casing construction for a rigid-frame ESP.
Source: Wheelabrator-Frye-Lurgi.
ro
i
-------
conveyor. Hoppers are not recommended for applications where the dust is
very sticky and may build up on sloping surfaces. Instead, the casing should
be extended to form a flat-bottomed box under the ESP. The dust is removed
by drag conveyors.
Hopper plugging is a major problem. Although manufacturers have produced
designs incorporating vibrators, heaters, poke holes, baffles (emphasis on
proper location), large discharge flanges, and steep hopper wall angles (55
to 65 degrees) to reduce these problems, they still persist.
A number of improvements could be made in the design of hoppers. The
first consideration should be to provide continuous evacuation of the hopper
so it will not be used as a storage device. Hopper aspect ratio (height to
width) is another important consideration; the correct ratio will minimize
reentrainment caused by gas sneakage to the hoppers. Low aspect ratio
hoppers can be corrected by vertical baffling.
In the sizing of hoppers, consideration should be given to the fact that
80 to 90 percent of the collected dust is removed in the first field. A
conservatively designed dust-removal system will keep pluggage to a minimum.
The trend is toward large-sized hoppers so that operators can respond to
hopper plugging before electrical grounding or physical damage is done to the
electrodes. Although this trend is a valid one, some thought must be given
to the time required to remove the accumulated ash. It is probably better
not to specify a certain storage time. Stainless steel fillets or lower-end
cladding also should be considered to reduce dust bridging in these larger-
size hoppers. /
Some manufacturers offer a high-ash, fail-safe system that automatically
phases back or deenergizes high-voltage equipment when high ash levels are
detected. Some kind of reliable ash-level detection, either the nuclear or
capacitance type, is recommended for all hopper designs. If the preliminary
design indicates a potential for problems with ash discharge, the discharge
flange should be no less than 12 in. in diameter. Heaters in the discharge
throat and up to one-third the height of the hopper have proved to be espe-
cially beneficial. A low-temperature probe and alarm also might be consid-
ered.
SECTION 2-OVERVtEW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-36
-------
During construction, checks should be made to determine if the transi-
tion from a rectangular hopper to a round outlet is accomplished without
ledges or projections; this will help to reduce plugging. Baffles should not
extend too far into the hopper (which can increase plugging), and vibrators
should be mounted at the baffle line to eliminate the formation of rat holes.
The rapping controls should be interlocked with the ash-removal system so
that rapping cannot occur unless the hopper is being evacuated.
External key interlocks should be installed to provide safe access to
hoppers. Bolt-on doors through baffles should not be installed because they
can cause dangerous dust accumulation on the interior side of the door.
Enough "poke hole" ports should be provided to allow for cleaning a blockage
at the discharge. Enclosing the hopper areas will help to reduce heat loss
in the hopper and discharge system. Alignment of conveyors is very impor-
tant, and depends on the alignment of hopper connections. Field-adjustable
flange connections are recommended.
Collecting Electrodes—
Collecting electrodes (plates) are the grounded components on which the
clust collects. Many shapes of flat collecting electrodes are used in ESP's,
as shown in Figure 2-16. Some ESP's ere designed with cylindrical collection
surfaces.
Collecting plates are commercially available in lengths of 3 to 12 ft
(wire weight), or 6 to greater than 15 ft (rigid frame) and heights of 9 to
36 ft (wire weight) or as high as 50 ft for rigid frame designs. These
panels generally are grouped with the ESP to form independently rapped col-
lecting modules. A variety of plates are commercially available, but their
functional characteristics do not vary substantially. When assembled, col-
lecting plates should be straight and parallel with the discharge electrodes.
Correct alignment requires that care be exercised during fabrication, ship-
ping, storage in the field, and erection.
The plate support system must be rugged because in many designs it must
also transmit rapper energy to the plates. Each design should be examined
with regard to its operating limits with various types of rappers. The
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-37
-------
OPZEL PLATES
<%- <
^•VERTICLE
BAFFLES-
POCKETS
n_n_rn_
• * *
_ 1 i f *i i
rn_^ GAo^TXOW <<<<<<
11 — i ^ ^ ^ ^ ^
OFFSET PLATES V PLATES
,. 18IN. .. 2
_ _
_ _
J J
SHIELDED
33 £ I ~
PLATES STRIP PLATES
n BFT- x
f 1 4
n J '
IN
FLAT PLATES
Figure 2-16. Various designs of collecting electrodes.""
(Permission granted by APCA.)
2-38
-------
effects of vibration and impact loading at all welded points should be con-
sidered. Consideration should also be given to the adjustment of any nec-
essary plate alignment after shakedown. Enough spacers should be provided to
maintain alignment and allow for temperature variations.
Rapper anvils attached to either plate supports or rapper header beams
should be durable enough to withstand the stress of rapping and to maintain
alignment (no bending of flanges or other local deformations).
If an ESP is installed properly, collecting plates pose no maintenance
problems during normal operation. Replacement or repair of a collecting
plate because of warpage or breakage is a time-consuming and expensive task.
The following items should be considered in the erection of collecting
plates:15
° Collection plate - The trueness of the dust plate depends on the
care exercised in fabrication, packaging, storage, and handling in
the field. Most damage occurs while plates are being unpacked and
raised.
c Collecting plate support structure - The design of this structure
must be flexible enough that the structure can be adjusted and
readjusted for proper alignment during construction and shakedown.
Proper alignment is critical for the rapping anvils.
0 Baff1es - Baffles are installed between the casing and the outermost
collection plate to prevent gas sneakage out of the main zone of
precipitation. Enough room should be left for installation and
inspection of these baffles to ensure proper closing of the space.
Discharge Electrodes--
Discharge electrodes are metal, and the type is determined by the compo-
sition of the gas stream. The electrodes may be cylindrical or square wire,
barbed wire, or stamped or formed strips of metal of various configurations
(as shown in Figure 2-16). The shape of these electrodes determines the
current voltage characteristics; the smaller the wire or the more pointed its
surface, the greater the value of current for a given voltage.
Discharge electrodes are mounted 1n various ways. They may be suspended
from an insulating superstructure with weights at the bottom holding them
tightly in place, or they may be rigidly mounted on masts or frames. The ad-
vantage of the rigid-type discharge electrode system over a wire-weight
system is that it lessens the chance of a broken wire falling against a plate
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND OiM CONSIDERATIONS
2-39
-------
and shorting out that section of the ESP. The weighted-wire type must be
stabilized to avoid its swinging in the gas stream. Examples of the wire-
weight and rigid-wire systems are shown in Figure 2-17.
When properly designed, both wire-weight and rigid-frame discharge elec-
trode configurations have excellent collection capability. The low initial
cost of a weighted-wire design is typically offset by high maintenance costs
resulting partially from wire breakage. The reverse is true of rigid-frame
designs; in this case, the high initial costs are usually offset by low main-
tenance costs. Warping of the discharge electrode frame caused by wide ther-
mal swings is generally not a problem if a unit is properly designed. Both
designs can deliver similar electrical power levels to the ESP for particu-
late matter collection. Generally, however, rigid-frame ESP's operate at
higher voltages and lower current densities than wire-weight ESP's in a given
application because of the wider spacing between the discharge electrodes and
the collecting electrodes. These voltage-current characteristics may be
better suited to collection of high-resistivity dusts.
Because of problems with discharge electrodes design is especially im-
portant in areas related to electrical erosion, mechanical fatigue corrosion,
and inadequate rapping. When high-current sparks or continuous sparking must
be tolerated, the use of large, formed discharge electrodes will provide much
better protection against erosion of the discharge electrode than will small-
er sizes (of either wire-weight or rigid-frame electrodes). Shrouds should
be included at both the top and bottom of wire-weight electrodes, and all
interelectrode high-voltage and grounded surfaces should have smooth surfaces
to minimize spark-over. Transformer-rectifier sets should be well matched to
the ESP load, and automatic spark controllers should keep voltage close to
the sparking threshold. Contact between the electrode and the stabilizing
frame should be solid to prevent sparking. For rigid-frame discharge elec-
trodes, substantial reinforcement is required at the point where the elec-
trode is attached to the support frame, to ensure that a significant amount
of metal must be lost before failure occurs. The use of alloyed metals is
recommended for all discharge electrodes to minimize corrosion and fatigue.
Mechanical connections in the discharge electrode structure should be
designed so that flexing and reduction in cross-sectional area at junction
SECTION 2-OVERV1EW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-40
-------
SUPPORT INSULATOR
HOUSING
HIGH VOLTAGE
BUS DUCT
VIBRATION
ISOLATORS
PROTECTOR5*^
TUBE
TENSIONING WEIGHT
DISCHARGE ELECTRODE
BUS CONDUCTOR
HIGH VOLTAGE
SWITCH
'TRANSFORM R-
'RECTIF1ER
DISCHARGE ELECTRODE
WEIGHT GUIDE FRAME
HANGER
ROD
RAPPER ANVIL
SPACER BRACKET
FUSED SILICA SUSPENSION INSULATOR
DISCHARGE ELECTRODES
ro
i
Typical discharge electrode system -
penthouse arrangement
(Courtesy of Environmental Elements Co.)
Typical rigid-frame electrode system
(Courtesy of Wheelabrator Frye)
Figure 2-17. Comparison of wire-weight and rigid-frame ESP designs.
-------
points are minimized. Connections should be vibration- and stress-resistant,
and electrodes should be allowed to rotate slightly at their mounting points.
Keeping the total unbraced length of electrode as short as possible will
minimize mechanical fatigue.
Rappers/Vibrators--
Rappers are categorized according to their use on wire-weight or rigid-
frame ESP's. On wire-weight ESP's, rapping impulses are provided by either
single-impulse or vibratory rappers, which are activated either electrically
or pneumatically. Figures 2-18 through 2-20 show examples of typical rappers
for wire-weight ESP's,
Electromagnetic or pneumatic impulse-type rappers usually work better on
collecting electrodes and in difficult applications because a vibrator gener-
ally cannot generate sufficient operating energies without being damaged.
The magnetic-impulse, gravity-impact rapper is a solenoid electromagnet con-
sisting of a steel plunger surrounded by a concentric coil; both are enclosed
in a watertight steel case. The control unit contains all the components
(except the rapper) needed to distribute and control the power to the rappers
for optimum precipitation,
During normal operation, a d.c. pulse through the rapper coil supplies
the energy to move the steel plunger. The magnetic field of the coil raises
the plunger, which is then allowed to fall beck and strike a rapper bar con-
nected to the collecting electrodes within the ESP. The shock transmitted to
the electrodes dislodges the accumulated dust.
The electromagnetic rappers also have a coil (energized by alternating
current). Each time the coil is energized, vibration is transmitted to the
high-tension v/ire-supporting frame and/or collecting plates through a rod.
The number of vibrators applied depends on the number of high-tension frames
and/or collecting plates in the system. The control unit contains all the
components necessary for operation of the vibrators, including a means of
adjusting the vibration intensity and the length of the vibration period.
Alternating current is supplied to the discharge-wire vibrators through a
multiple-cam timer that provides the sequencing and time cycle for energiza-
tion of the vibrators.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN. AND O&M CONSIDERATIONS
2-42
-------
FLEXIBLE CONDUIT.
TOP OF RAPPER ROD
PRECIPITATQR OR
TOP HOUSING ROOF
BANK OF
COLLECTING
ELECTRODES
COVER
COIL ASSEMBLY t,
PLUNGER GUIDE
LOWER CASING
ADJUST ING BOLT
7-3/8" OPTIHUM OUTPUT
(VAR1BIE 6-7/8' TO 9-7/8")
NIPPLE
*
RAPPER ROD
ANVIL BEAM
Figure 2-18. Typical installation of Magnetic Impulse, Gravity Impact (M.I.G.I.}
Rapper (courtesy of Environmental Elements Co.)-
2-43
-------
10 It
16
19
1. RAPPER ASSEMBLY
2. HOSE CLAMP
3. A!R HOSE
4. AIR HOSE NIPPLE
5. SOLENOID VALVE
6. ELECTRICAL CONNECTOR
7. U-BOLT
8. SUPPORT
9. STRAINER
10.UNION
11.GLOBE VALVE
12.PIPE
13, CONDUIT
14. PRESSURE REGULATOR
IS. AIR INLET FITTING
16. SAME AS 2
17. SAME AS 3
18. PREC1PITATOR FRAME
19. GROUND CABLE
20. SCREW AND Nbl
21. RAPPER ROD
Figure 2-19. Typical pneumatic rapper assembly.
(Courtesy of Environmental Elements Co.)
2-44
-------
Figure 2-20. Typical electric vibrator type rapper.
2-45
-------
Nonmalleable high-strength alloyed steels should be used for the hard-
ware components because their resonance properties and strengths match those
required to receive and deliver impacts with few mechanical failures.
The electrical controls should be adjustable so that the rappers can be
issembled into different groups and each group can be adjusted independently
for optimum rapping frequency and intensity. The controls should be manually
adjustable so they can provide adequate release of dust from collecting
plates and simultaneously prevent undesirable stack puffing.
Failures of rapper rod connections to carbon steel electrode systems can
be minimized by designing welds that are large and strong enough to withstand
impacts and by careful welding. Proper selection of rod material and protec-
tive shrouding in sealed areas will minimize corrosion problems.
Problems related to ground faults also occur in the ESP's conduit system
because of lack of seals at connections, poor-quality wire terminations, and
19
use of low-quality wire.
In some applications, the magnetic-impulse, gravity-impact rapper is
also used to clean the ESP's discharge wires. In this case, the rapper
energy is imparted to the electrode-supporting frame in the normal manner,
but an insulator .isolates the rapper from the high voltage of the electrode-
supporting frame.
The number of rappers, size of rappers, and rapping frequencies vary
according to the manufacturer and the nature of the dust. Generally, one
o
rapper unit is required for 1200 to 1600 ft of collecting area. Discharge
electrode rappers serve from 1000. to 7000 ft of wire per rapper. Intensity
of rapping generally ranges from about 5 to 50 ft/lb, and rapping intervals
are adjustable over a range of approximately 30 to 600 seconds.
Rigid-frame ESP's generally have mechanical-hammer rappers. Each frame
is rapped by one hammer assembly mounted on a shaft. (See Figure 2-21.) A
low-speed gear motor is linked to the hammer shaft by a drive insulator,
fork, and linkage assembly. Rapping intensity is governed by the hammer
weight, and rapping frequency is governed by the speed of the shaft rotation.
Acceleration forces in discrete places on the collecting surface plates
should be measured and mathematical relationships between hammer weights,
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN. AND OiM CONSIDERATIONS
2-46
-------
GEAR MOTOR DRIVE
DISCHARGE ELECTRODE
RAPPER ASSEMBLY
COLLECTING SURFACE
RAPPER ASSEMBLY
ro
i
Figure 2-21. Tumbling-hammer assembly for use with rigid-frame discharge electrode
and collecting-surface rapping system.
(Courtesy of Wheelabrator Frye, Inc.)
-------
lift angles, and plate dimensions should be established and confirmed in
laboratory and field testing. Uniform acceleration on the discharge wire
frame is also important for efficient dust removal without the wire being
destroyed by its own vibrations.
Solids Removal Equipment--
In large systems such as those in utility applications, solids can be
removed from ESP's by a pressure or vacuum system (see Figure 2-22), A screw
conveyor can be used for this purpose in many smaller industrial applications.
Dust can also be wet-sluiced directly from the hoppers-. Once conveyed from
the hoppers, the dust can be disposed of dry, or it can be wet-sluiced to a
holding pond.
Removal from the hopper—An air seel is required at each hopper dis-
charge. Air locks provide a positive seal, but tipping or air-operated
slide-gate check valves are also used for this purpose. The use of heaters,
vibrators, and/or diffusers is often considered because of the occasional
bridging that occurs in the hoppers. In trough-type hoppers, a paddle-type
conveyor provides the best means of transporting the dust to the air lock.
Dust valves are often oversized to help facilitate removal of>dust from the
hopper.
Pneumati c systerns --The length of a vacuum system is limited by the con-
figuration of the discharge system and the altitude above sea level. When
the limits for vacuum systems are exceeded, pressure systems are applied.
When the number of hoppers exceeds about 20 and the length of the system is
too greet for a vacuum system, combination vacuum/pressure systems may be
used.
A vacuum is produced either hydraulically or by use of mechanical vacuum
pumps. Positive displacement blowers are used with pressure systems. Vacuum
systems are equipped with electric valves and slide gates, whereas pressure
systems have air locks and slide gates.
Materials of construction are extremely important in the selection of a
solids-removal system. The chemical composition of both the dust and convey-
ing air and the temperatures at various points in the conveying system should
be determined.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-48
-------
VALVES
DRY FLY ASH
UNLOADING
CHUTE
Figure 2-22. Vacuum system for solids removal.
(Courtesy of Allen Sherman Hoff Inc.)
2-49
-------
When material characteristics (material density, particle sizes, and
concentrations, and the physical characteristics of the conveying air or gas)
are known, the required conveying velocity can be determined. Setting the
design rate at 20 percent above the theoretical maximum conveying capacity
will usu&lly prevent plugging.
Facilities for storing pneumatic dust generally are equipped with cy-
clones, and often with a fabric filter. The stored dust is conditioned with
water and/or a wetting agent and then either transported (by truck or rail)
to a disposal site or mixed with water and pumped to a disposal pond.
Gas Distribution Equipment—
Proper gas flow distribution is critical for optimum precipitator per-
formance. Areas of high velocity can cause erosion and reentrainment of dust
from collecting surfaces or can allow gas to move through the ESP virtually
untreated. Improper distribution of gas flow in ducts leading to the ESP
causes dust to accumulate on surfaces and results in high pressure losses.
Devices such as turning vanes, diffusers, baffles, and perforated plates
are used to maintain and improve the distribution of the gas flow. A diffus-
er consists of a woven screen or a thin plate with-a regular pattern of small
openings. A diffuser breaks large-scale turbulence into many small-scale
turbulent zones, which, in turn, decay rapidly and within a short distance
coalesce into a relatively low-intensity turbulent flow field. The use of
two or three diffusers in series provides better flow than only one diffusion
plate could achieve. Cleaning of the gas distribution devices may entail
rapping.
The design of inlet and outlet nozzles of ESP plenums and their distri-
bution devices must be uniform. Poor design of inlet plenums can result in
pluggage such as that shown in Figure 2-23. Figure 2-24 shows an example of
poorly designed inlet plenums. Figure 2-25 shows two methods of improving
gas distribution at the inlet plenum.
In multiple-chamber ESP's, louver-type dampers should be used for gas
proportioning instead of guillotine shutoff dampers, because guillotine-type
dampers tend to destroy proper gas distribution to a chamber.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND 0*M CONSIDERATIONS
2-50
-------
***** *****
• *••• ••*•
VA*.V W
&•_•-•.• i **. •
Figure 2-23. Pluggage of perforated plates at the inlet to an ESP.
2-51
-------
GAS FLOW
ELEVATION
ELEVATION
\
Figure 2-24. Examples of two inlet plenum designs that generally cause
gas distribution problems.*Q
(Permission granted by Precipitator Technology, Inc.)
PERFORATED
PLATES
GAS FLOW
Figure 2-25. Two methods of spreading the gas pattern at
expansion inlet plenums.'0
(Permission granted by Precipitator Technology, Inc.)
2-52
-------
Poor gas distribution can cause gas sneakage through hoppers. Expansion
plenums or top-entry plenums cause gas vectors to be directed toward the
hopper; if multiple perforated plates do not fit well in the lower portion of
the plenum or if the lower portion has been cut away because of dust buildup,
gas is channeled into the hoppers. Bypassing the active collection portion
of the ESP and/or reentrainment is the end result.
Gas Flow Models--
Gas flow models are used to determine the location and configuration of
gas flow control devices. Although flow model studies are not always able to
develop the desired distribution, they can at least provide a qualitative
indicator of the distribution.
Temperature and dust loading distributions are also important to effi-
cient ESP operations. Although the temperature of the flue gas is generally
assumed to be uniform, this is not always true. The effects of gas tempera-
ture on ESP electrical characteristics should be a design consideration, as
well as modeling of dust distribution.
If dust loading distributions are not modeled, the dust is assumed to be
distributed evenly in the gas; as long as the gas distribution is of a prede-
fined quality, no dust deposition problems should occur. Nevertheless, prob-
lems such as poor duct design, poor flow patterns at the inlet nozzle of the
ESP plenum, and flow and wall obstructions can cause unexpected dust deposi-
tion.20
Power Supplies —
The power supply to an ESP consists of four basic components: a step-up
transformer, a high-voltage rectifier, a control element, and a control sys-
tem sensor. The system is designed to provide voltage at the highest level
possible without causing arc-over (sustained sparking) between the discharge
electrode and the collection surface.
The T-R set converts low-voltage alternating current to high-voltage un-
directional current suitable for energizing the ESP. The T-R sets and radio-
frequency (RF) choke coils are submerged in a tank filled with a dielectric
fluid. The RF chokes are designed to prevent high-frequency transient volt-
age spikes caused by the ESP from damaging the silicon diode rectifiers. The
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND OiM CONSIDERATIONS
2-53
-------
automatic control system Is designed to maintain optimum voltage and current
in response to changes in the characteristics and concentrations of the dust.
Figure 2-26 shows the components of a typical automatic voltage control
system.
The T-R sets should be matched to ESP load. The ESP will perform best
when all T-R sets operate at 70 to 100 percent of the rated load without ex-
cessive sparking or transient disturbances, which reduce the maximum continu-
ous-load voltage and corona power inputs. Over a wide range of gas tempera-
tures and pressures in different applications, practical operating voltages
range from 15 to 80 kV at average corona current densities of 10 to 70 mA/lQQQ
21
2
ft of collecting area.
The following are the most common T-R set output ratings:1
70 kVp, 45 kV avg. 250 to 1500 mA D.C., 16 to 100 kV, 9- to 10-in. ducts
78 kVp, 50 kV avg. 250 to 1500 mA D.C., IE to 111 kVa, 11- to 12-in.
ducts
80 kVp, 55 kV avg. 250 to 1500 mA D.C., 12-in. ducts (from some vendors)
At currents over 1500 mA, internal impedances of the T-R sets are low,
which makes stable automatic control more difficult to achieve. Design
should call for the highest possible impedance that is commensurate with the
application and performance requirements. With smaller T-R sets, this often
means more sectionalization. The high internal impedance of the smaller T-R
sets facilitates spark quenching as well as providing more suitable wave
forms. Smaller electrical sections localize the effects of electrode misalign-
ment and permit higher voltages in the remaining sections.
High-temperature gases (700° to 800°F) require T-R sets with lower volt-
age because the density of the gas is lower. High-pressure gases in corona
quench situations [high space charge in the interelectrode space; e.g., acid-
mist or very wide (15- to 25-in.) ducts] require extra high voltage (lower
current ratings than in conventional use); conversely, low gas density and/or
low dust concentrations require higher currents at lower voltages. In gener-
al, current ratings should increase from inlet to outlet fields (3 to 5 times
for many fly ash ESP's).
Generally, name-brand T-R sets rarely fail. Problems are generally
related to quality control: defective components; moisture in oil due to
SECTION 2-OVERVIEW OP ESP THEORY, DESIGN. AND O&M CONSIDERATIONS
2-54
-------
TRANSFORMER/RECTIFIER
CURRENT
LIMITING
REACTOR
" AUTOMATIC VOLTAGE CONTROL
RS 232
r INTERFACE
r-LINE
T SYNC.
tCESSOR
MODULE
t*
u
ESP
en
en
Figure 2-26. Electrostatic precipitator power supply circuit.
(Courtesy of Lodge Cottrell, Inc.)
-------
improper coll baking or vacuum fill techniques; metal bits, rust, or scale in
tanks; incompatibility of solid insulation materials with certain cooling
liquids; overfilling with oil and insufficient expansion space; poor mechani-
cal bracing or mounting of transformer coils and other components; internal
sparking due to inadequate spacing and electrical field concentration points;
21
and mishandling in shipment or installation.
Hand hole-cover plates should be provided for access to rectifiers,
radio frequency (RF) diodes, and voltage dividers. Also desirable (and
expensive) are large crane trolley systems for quick replacement of T-R sets,
Another alternative is to include some additional redundancy of plate area in
the design to compensate for T-R set outages.
Silicon-controlled rectifiers should be carefully mounted in suitable
heat sink assemblies and tightened with a torque wrench to manufacturer's
specifications. All sensitive leads from the T-R sets to the automatic
21
voltage-control cabinet should be shielded in coaxial cable.
Instrumentation--
Instrumentation necessary for proper monitoring of ESP operation can be
categorized by location; i.e., T-R sets, rappers/vibrators, hoppers/dust
removal systems, and external items.
T-R sets—Power input is the most important measure of the ESP perform-
ance. Thus, any new ESP should be equipped with the following:
Primary current meters
Primary voltage meters
Secondary current meters
Secondary voltage meters
Spark rate meter (optional)
These meters are considered essential for performance evaluation and trouble-
shooting. Figure 2-27 shows a typical control cabinet and T-R set instrumen-
tation.
Data loggers (mainly for digital automatic control systems) are avail-
able to help speed up troubleshooting and reduce operating labor. Oscillo-
scopes are also useful in evaluating power supply performance and identifying
the type of sparking (multiple-burst versus single-arc).
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND O4M CONSIDERATIONS
2-56
-------
!
Figure 2-27. Typical ESP control cabinet and T-R set instrumentation.
(Courtesy of Environmental Elements, Inc.)
2-57
-------
It is also possible to use feedback signals from transmissometers, full
hopper detectors, gas conditioning systems, rappers, and suitable process
fault indicators in conjunction with the automatic control unit to achieve
22
optimum performance under all conditions. An example of this is automatic
phase-back of T-R sets when hoppers are overfilled, which prevents discharge
wires from burning.
Rapper_s/v1brators—Microprocessor-type technology is available for a
high degree of rapper control flexibility and ease of maintenance. For
example, new controls can test each circuit before energizing it and thus
prevent control damage from ground faults. If a ground fault does occur, the
control will automatically bypass the grounded circuit and indicate the
1 Q
problem on a Light Emitting Diode (LED) display. This permits early loca-
tion of the problem and expedites its solution.
Instrumentation should be used in conjunction with a transmissometer for
troubleshooting ESP problems. Separate rapping instrumentation should be
provided for each field. In the case of wire-weight electrodes, readings of
frequency, intensity, and cycle time can be used with T-R set controls for
proper setting of rapper frequency and intensity (see Figure 2-28).
In the case of rigid-frame, mechanical rappers, cycle time and rap fre-
quency of both internal and external rappers are easy to measure. Individual
operation of internal rappers is not easily instrumented, nor is intensity
control possible without a shutdown of the ESP.
Hoppers--
Instrurrentation should be provided for detecting full hoppers, for the
operation of the dust valve, and for the dust-removal system. Level detec-
tors can utilize gamma radiation, capacitance, pressure differential, or
20
temperature. Alarms should be located such that hoppers never become com-
pletely filled, but frequent alarms should be avoided. A low-temperature
probe and alarm can be used in conjunction with the level detector. Control
panel lights indicate the operation of hopper heaters and vibrators.
Zero-motion switches are used on rotary air lock valves and on screw
conveyors to detect malfunctions. Pressure switches and alarms are normally
used to detect operating problems in pneumatic dust handling systerrs.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-58
-------
Figure 2-28. Typical rapper control panel
(frequency, intensity, and cycle time).
2-59
-------
Accessibility and Safety--
Safe, convenient access (walkways, hatches, etc.) must be.provided for
entry and servicing of ESP's and ancillary equipment during shutdown. . Some-
times the design of the ESP should be such that individual chambers can be
prepared for safe entry while the balance of the chambers are on line. Items
to which access is needed are discharge wire mountings, hoppers, penthouses,
rappers, instrumentation, etc. For rigid-frame ESP's, adequate clearance
should be provided between collecting surfaces and interior walkways for the
replacement of rigid electrodes through side doors. Access to collection
plates and inlet baffles is necessary to allow cleaning during shutdown.
Such accessibility requires the proper location of hatches, walkways, lad-
ders, and handrails. Hopper access doors should be wide enough for ladders
to be placed in the hoppers, and maintenance personnel should be able to
reach the bottom of the discharge electrode frame by ladder from the hopper
access platform. All potential electrical shock hazards must be addressed
by the use of grounding devices and electrical system, lockout procedures.
Measures also must be taken to purge enclosures of hot toxic gases before
entrance of maintenance personnel.
Erection Sequence—
Usually the casings and hoppers are erected first, and then the collect-
ing surfaces and discharge systems. A good quality control and inspection
program must be followed during the erection of an ESP, despite the pressures
of construction schedules. If the casing is not erected to true dimensions
and attempts are made to compensate for this error during the installation of
the collection surfaces, poor alignment of wires and plates can result.
Allowing 2 weeks to a month in the construction schedule for conducting
tests of gas velocity and making adjustments in the distribution system
(something that is often precluded) will save the time required to make these
adjustments after startup.
2.2.4 Areas of_R_esearch and Development
Some of the new designs and concepts that have been researched and
tested over the past 10 years include 1) wide spacing between discharge and
collection electrodes; 2) two-stage charging, i.e., the use of a pre-charging
SECTION 2-OVERVIEW OF ISP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-60
-------
electrode or ionizer ahead of the inlet field of the precipitators; and 3)
pulse-charging. Researched in the past, these concepts have recently re-
ceived renewed attention. Each of these technologies is described briefly,
Wide Spacing--
The design of a wide-space ESP is such that the plate-to-plate spacing
is in excess of the 9-inch spacing that has been standard in wire-weight
ESP's for many years. Spacings of from 10 to 24 inches have been used in
Europe and Japan, and more than 160 wide-space ESP's were in operation in
Japan at the end of 1978.
Higher voltages are necessary for establishing the electric field in
these ESP's, but migration velocities also increase with higher voltages.
Thus, some cost savings can be realized because the ESP collection plate area
required is smaller, with the wider spacing.
Under normal conditions, a wide-space ESP will operate at about the same
current density and have less tendency to spark. Also, minor misalignments
will not be as noticeable and additional space is available for inspections.
With high inlet loadings of fine dust, however, the wide-space ESP is
more sensitive to space charge effects and excessive sparking may occur,
especially in the inlet field.
The design of the wide-space ESP will require higher voltage power
supplies, and the optimum bus section size may be different from the standard
size because of the higher voltage. The relationship between SCA and collec-
tion efficiency will also differ from standard relationships.
Two-Stage Charging--
Two-stage charging is currently being investigated on a pilot-scale
level as a possible means of solving the problem of collecting high-resis-
tivity dust. In a two-stage design, the charging and collecting functions
are separated. The particles are first charged upstream of the main ESP unit
by a precharger or ionizer, and then collected in the main ESP unit by use of
a high electric field. The apparent advantage of such an arrangement is the
use of a small precharging inlet section and the design of the collector
section, which produces a high electric field without generating a corona,
SECTION 2-OVERVIEW OF ISP THEORY, DESIGN, AND O4M CONSIDERATIONS
2-61
-------
and in some cases a lower than normal current density, which thus eliminates
back-corona.
Several two-stage designs are currently available and being used in
small, low-dust-loading situations such as air conditioning/ventilation
systems, condensed oil mists, and other industrial applications in textile,
rubber, vinyl, asphalt, carpet, printing, grain drying, and food industries,
Current research is directed at extending the two-stage design to handle high
dust-loading industrial sources and for improved collection of high-resistiv-
??
ity dusts.
The five main types of precharges now being tested are:
1) The Air Pollution Systems high-intensity ionizer
2) The trielectrode precharger
3) The boxer charger
4} The ion beam charger
5) The cooled-pipe charger
Work is continuing to improve the precharging mechanism and to deal with
the problem of keeping a high electric field in the collector section without
corona generation (or perhaps low current density for recharging reentrained
dust), and the two-stage concept appears to have considerable-potential for
use in controlling high-resistivity dusts.
Pulse Charging--
Pulse energization has been experimented with in the past few years as a
means of upgrading ESP performance on high-resistivity fly ash without adding
plate area. In pulse energization, a high-voltage pulse is superimposed on
the base voltage to enhance ESP performance on high-resistivity dust. Retro-
fitting of an existing power supply is relatively simple, and the pulse unit
does not require much maintenance.
The lack of performance data on the effectiveness of pulse energization
has prevented interested companies from determining how much improvement they
can expect. Recent pilot- and full-scale operating data have been analyzed,
and a method published for estimating how much improvement can be expected
23
for a given situation.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND DIM CONSIDERATIONS
2-62
-------
2.3 ESP O&M CONSIDERATIONS
2.3.1 Typical ESP Failure Modes and Causes of Poor Performance
The several causes of poor performance in an ESP can be divided into the
following distinct categories:
1) Fundamental Problems - This category includes gas stream character-
istics such as high resistivity, unusually fine particle size, which
can be accounted for in the design of the ESP; and overall design
inadequacies (poor gas flow distribution, inadequate plate area,
inadequate or unstable energization equipment not matched to process
characteristics, improper rappers for the process particulate being
collected). Because these problems will cause O&M difficulties
throughout the life of the ESP, they are essentially independent of
a good O&M program. As discussed in Section 2.2, care must be taken
during the design and specification of each component of the ESP if
the plant O&M personnel are expected to keep the ESP operating
within prescribed air pollution control limits. In other words, if
the ESP is poorly designed, a proper O&M program may only serve to
keep the ESP operating marginally within compliance or at some
minimum level above the compliance limit. In these instances, the
design-related problem must be corrected before the O&M program can
be truly effective.
2) Mechanical Problems - These problems include electrode alignment
(i.e., warped plates, close clearances, twisted frames), wire break-
age, cracked plates, air inleakage, cracked insulators, dust depos-.
its, and plugged hoppers. Problems such as these will generally be
discernible through a review of V-I curves or in the course of
routine external or internal inspections. Improper design or con-
struction may contribute to these problems, and efforts should be
made to find the cause of the problem instead of blindly replacing
the component.
3) Operational Problems - This category of problems includes process
upsets that degrade ESP performance, inadequate power input or
failure of T-R sets, electrical sections out of service, improper
operation or failure of rappers, and dust removal valve failures.
These problems also can be influenced by poor design, and the re-
sulting degradation in performance can be immediate or occur over a
period of time (e.q., when rappers fail and dust deposits build up
on wires or plates).
Also important are the interdependence of the various ESP components and
the cascading effect of one problem creating other problems. These consider-
ations are discussed in more detail in later sections.
SECTION 2-OVERVIEW OF ESP THEORY. DESIGN. AND OiM CONSIDERATIONS
2-63
-------
A well-developed and adequate O&M plan will allow the plant to discover
problems before they have a chance to create other, more serious problems.
It will also help to determine why certain problems are happening or recur-
ring when the cause of the problem is not immediately obvious. The latter is
possible through adequate recordkeeping and the use of these data to develop
trends or otherwise isolate reasons for malfunctions or gradual deterioration
in performance.
Previous studies have shown performance histories are better at plants
where recordkeeping practices are adequate and plant personnel utilize these
24
data than at plants where recordkeeping has received little emphasis. The
necessary records for an O&M plan are discussed in the next subsection.
2,3.2 Establishing an Adequate Operation and Maintenance Program
Why should a plant make a concerted effort to maintain its ESP properly?
The most convincing reason, outside of the necessity to meet applicable par-
ticulate emission regulations, is one of economics. An ESP is an expensive
piece of equipment, and even well-designed equipment will deteriorate rapidly
if improperly maintained and will have to be replaced long before it should
be necessary. Not only can proper O&M save the plant money, it can also
contribute to good relations with the local control agency by showing good
faith in its efforts to comply with air regulations.
An ESP is unlikely to receive proper O&M without management support and
the willingness to provide its employees with proper training. Management
must instill an attitude of alert, intelligent attention to the operation of
the ESP instead of waiting for a malfunction to occur before acting. This
requires a consistent monitoring program entailing the maintenance of de-
tailed documentation of all ESP operations.
Although each plant has its own method of conducting an O&M program,
past experience has shown that plants that assign one individual the respon-
sibility of tying all the pieces of the program together operate better than
those where different departments look after only a certain portion of the
program and have little knowledge of how that portion impacts the overall
program. In other words, a plant needs to coordinate the operation,
SECTION 2-OVERVIIW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS
2-64
-------
maintenance, and troubleshooting components of its program if it expects to
be on top of the situation.
Some companies that have several plants have found it to be advantageous
to set up a central coordinating office to monitor the O&M status at each
plant. The resulting improved communications can provide an opportunity to
develop standardized reporting forms, assistance in personnel training,
interpretation of operating data, and routine inspections. With knowledgeable
people in the central coordinating office, the plants have somewhere to go
for assistance in solving problems for their specific kind of ESP.
Another resource that plants can draw upon is the manufacturer's field
service engineer. This person is involved in pre-operational inspections to
ensure proper assembly of ESP components; to set up the various controls
within prescribed limits; to check proper operation, actual energization of
the T-R sets, and the dust discharging system; to fine-tune the unit after
initial startup; and finally, to instruct plant personnel on how to perform
these functions.
Although experienced field service engineers can be very helpful as a
resource for assistance in troubleshooting, manufacturers are generally
plagued with a high turnover rate. Thus, the plant should be wary of in-
experienced people, who may incorrectly diagnose operating problems or be
unaware of proper correction procedures. This only adds to the confusion by
misleading O&M personnel.
The training and motivation of employees assigned to monitor and main-
tain the ESP are critical factors. These duties should not be assigned to
inexperienced people who do not understand how the ESP works or the purpose
behind their assigned tasks. The employee must know what management expects
and should receive encouragement for a job well done.
Regular training courses should be held by in-house personnel or by the
use of outside expertise so that operators and maintenance personnel are
instructed on everything they need to know in regard to the ESP. This should
include written instructions and "hands-on" sessions on safety, how to make
inspections while the ESP is both in and out of service, how to take electri-
cal readings, perform routine maintenance, investigate grounds or other
SECTION 2-OVERVIEW OF iSP THEORY, DESIGN, AND OiM CONSIDERATIONS
2-65
-------
problems, and how'to record and use data. Training provides the knowledge
necessary for proper operation and maintenance of the ESP and makes the
employees' job easier because they will understand why they are taking elec-
trical readings or searching for broken wires.
In summary, the three separate components of an adequate plan for long
ESP life are operation, maintenance, and troubleshooting. Each plant should
have its own O&M procedures manuals, blueprints, and a complete set of ESP
specifications; an adequate supply and record of spare parts; written proce-
dures for addressing malfunctions', and formalized audit procedures.
Records should be kept on ESP operating conditions (process logs, fuel
records, gas temperature, ESP power levels, etc.). equipment conditions (in-
ternal inspections; daily inspections of rappers, hoppers, T-R set trips,
etc.)* maintenance (work orders, current work in progress, deferred work),
and troubleshooting/diagnostic analysis (component failure frequency and
locations, impact of process changes on ESP performance, and other trend-
related analyses). Each of these areas is discussed in detail in later
sections of this manual.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN. AND O&M CONSIDERATIONS
2-66
-------
REFERENCES FOR SECTION 2
1. White, H. J. Electrostatic Precipitation of Fly Ash - Precipitator
Design.- JAPCA 27(3). 1977. p. 214.
2, Szabo, M. F., Y. M. Shah, and S, P. Schliesser. EPA Inspection Manual
For Evaluation of Electrostatic Precipitator Performance. PEDCo Environ-
mental , Inc. EPA-340/1-79-007. March, 1981.
3. Liegang, D. The Effect of Gas Temperature Upon the Performance and
Design of Electrostatic Precipitators. Staub-Reinhalt, Luft, Vol. 28,
No. 10. October 1968.
4. Riley, Oohn D. Common Process Variables Which Can Affect Electrostatic
Precipitator Performance. In APCA Proceedings of a Specialty Conference
on Operation and Maintenance of Electrostatic Precipitators. Dearborn,
Michigan. April 10-12, 1978.
5. Deutsch, W. Ann der Physik, 68:335 (1922).
6. Matts, S,, and P. 0. Ohnfeldt. Efficient Gas Cleaning with S. F. Elec-
trostatic Precipitators. Flakt. A. B. Swenska FlaktFabriken. June
1973.
7. Gooch, J. P., J. R. McDonald, and S. Oglesby, Jr. A Mathematical Model
of Electrostatic Precipitation. EPA-650/2-75-037. April 1975.
8. McDonald, J. R. A Mathematical Model of Electrostatic Precipitation
(Revision 1): Volume I. Modeling and Programming. EPA-600/7-78-llla.
June 1978.
9. McDonald, J. R. A Mathematical Model of Electrostatic Precipitation
(Revision 1): Volume II. User Manual. EPA-600/7-78-lllb. June 1978.
10. Mosley, R. B., M. H. Anderson, and J. R. McDonald. A Mathematical Model
of Electrostatic Precipitation (Revision 2). EPA-600/7-80-034. Febru-
ary 1980.
11. Cowen, S. J,, D. S. Ensor, and L. E. Sparks. TI-59 Programmable Calcu-
lator Programs for In-Stack Opacity, Venturi Scrubbers, and Electro-
static Precipitators. EPA-600/8-80-024. May 1980.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN. AND O&M CONSIDERATIONS 2-67
-------
. 12. Walker, A, 8., and G. W. Gawreluk. Performance Capability and Utiliza-
tion of Electrostatic Precipitators Past and Future, in Proceedings -
International Conference on Electrostatic Precipitation - Sponsored by
Electric Power Research Institute, Industrial Gas Cleaning Institute,
and Air Pollution Control Association. Monterey, California. October
1981.
13. White, H. J. Resistivity Problems in Electrostatic Precipitation.
JAPCA. Vol. 24, No. 4. April 1974.
14. U.S. Environmental Protection Agency. Control Techniques for Particu-
late Emissions from Stationary Sources, Volume 1. Office of Air Quali-
ty Planning and Standards. Research Triangle Park, North Carolina.
July 1980.
15. Schneider, G. R., et al. Selecting and Specifying Electrostatic Precip-
itators. Chemical Engineering. 82(11):94-108. May 26, 1975.
16. Szabo, N. F., and R. W. Gerstle. Operation and Maintenance of Particu-
late Control Devices on Coal-Fired Utility Boilers. PEDCo Environmen-
tal, Inc. EPA-600/2-77-129. July 1977.
17. PEDCo Environmental, Inc. Design Considerations for Particulate Control
Equipment. Prepared for U.S. Environmental Protection Agency, Region V.
Contract No. 68-02-2535. December 1979.
18. Stewart, L. L., and K. H. Pearson. The Application of Multiple Level
Rapping and Improved Collecting Plate Design to Provide Significant
Improvement in Electrostatic Precipitator Performance. Peabody Sturdivant
Company. Presented at the 76th Annual Meeting of the Air Pollution
Control Association, June 20-25, 1983, Atlanta, Georgia.
19. Lynch, J. G., and 0. S. Kelly. A Review of Rapper System Problems
Associated with Industrial Electrostatic Precipitators. In; A Special-
ty Conference on Operation and Maintenance of Electrostatic Precipita-
tors. Michigan Chapter - East Central Section, Air Pollution Control
Association. April 1978.
20. Katz, J. The Art of Electrostatic Precipitation. Precipitator Tech-
nology, Inc. Munhall, Pennsylvania. 1979.
21. Hall, H. J. Design, Application, Operation, and Maintenance Techniques
for Problems in the Electrical Energization of Electrostatic Precipita-
tors. In; A Specialty Conference on Operation and Maintenance of
Electrostatic Precipitators. Michigan Chapter - East Central Section,
Air Pollution Control Association. April 1978.
22. Hall, H. J. Design and Application of High Voltage Power Supplies in
Electrostatic Precipitation. In: Proceedings of Symposium on Electro-
static Precipitators for the Control of Fine Particles, Pensacola,
Florida, September 30 - October 2, 1974. EPA-650/2-75-016. January
1975.
SECTION 2-OVERVIEW OF ESP THEORY, DESIGN, AND O&M CONSIDERATIONS 2-68
-------
23. Rinard, G., M. U. Durham, and D. E. Rugg. Two Stage Electrostatic
Precipitation. Denver Research Institute. In: Proceedings: Interna-
tional Conference on Electrostatic Precipitation. Sponsored by Electric
Power Research Institute, Industrial Gas Cleaning Institute, and Air
Pollution Control Association. Monterey, California. October 1981,
24. Puille, W.» et al. Estimate ESP Efficiency Gains from Pulse Energiza-
tion. Power. May 1984.
25. Schliesser, S. P., and J. R. Richards. Development of Guideline Docu-
ment for State Operating and Maintenance Reccrdkeeping Programs. Pre-
pared for Engineering Science Co. under EPA Contract No. 68-01-4146,
Task No. 72. March 1981.
SECTION 2-OVERVIEW OF ESP THEORY. DESIGN, AND O&M CONSIDERATIONS
2-69
-------
-------
SECTION 3
ESP PERFORMANCE MONITORING
Performance monitoring is a key factor In establishing good operation
and maintenance procedures for an ESP. It includes measurement of key oper-
ating parameters by both continuous and intermittent methods, comparison of
these parameters with baseline and/or design values, and the establishment of
recordkeeping practices. These monitoring data are useful in performance
evaluation and problem diagnosis. In this section, the key operating data
and procedures used in performance monitoring are discussed. Interpretation
of the data is covered in Section 4.
3.1 KEY OPERATING PARAMETERS AND THEIR MEASUREMENT
Several operating parameters are indicative of a likely change in per-
formance. Some of these parameters are easily measured and monitored on a
continuous basis, whereas others must be measured only periodically because
of the expense and/or difficulty in measurement. Host of these parameters,
however, directly affect ESP performance. The following typical parameters
are discussed here; gas volume and gas velocity through the ESP; tempera-
ture, moisture, and chemical composition of the gas; particle size distribu-
tion and concentrationi resistivity of the particulate; and power input.
Many of these factors are interrelated.
3.1.1 Gas Volume and Velocity
According to predictive equations and models, a decrease in gas volume
results in an increase in collection efficiency and vice versa. Although the
improvement or deterioration in performance is not nearly as great as the
Deutsch-Anderson equation predicts, the equation is qualitatively correct. A
decrease in gas volume results in an increase of the SCA (ft of plate area/
1000 acfm}, a decrease in gas velocity through the ESP, an increase in the
SECTION 3-ESP PERFORMANCE MONITORING
3-1
-------
treatment time (during which the participate is subjected to the electric
field charging and collecting mechanisms), and hence, improved performance.
A decrease in velocity may also reduce rapping reentrainment and enhance the
collection of the fine particles in the 0.1 to 1.0 ym range, which are excep-
tionally difficult for most ESP's to collect.
Gas flow distribution is a very important aspect of gas flow through the
ESP. Ideally, the gas flow distribution should be uniform throughout the ESP
(top to bottom, side to side). Actually, however, gas flow through the ESP
is not evenly distributed, and ESP manufacturers settle for what they consid-
er an acceptable variation. (Standards recommended by the Industrial Gas
Cleaning Institute have been set for gas flow distribution. Based on a
velocity sampling routine, 85' percent of the points should be within 15
percent of the average velocity and 99 percent should be within 1.4 times the
average velocity.) Generally, uneven gas flow through the ESP results in
lost performance because the reduction in collection efficiency in areas of
high gas flow is not compensated for by the improved performance in areas of
lower flow. Gas distribution can also affect gas sneakage through the ESP.
The use of gas distribution devices such as perforated plates and turning
vanes and good ductwork design help to provide good gas distribution.
Total gas volume is usually measured by using a pi tot tube traverse.
The method is usually a combination of EPA Reference Methods 1 and 2; the
duct is divided into equal areas, and each area is sampled to arrive at an
average velocity through the duct. When the average velocity and the duct
cross-sectional area are known, the average gas volume can be determined.
Because most facilities do not routinely measure gas volume, other indirect
indicators may be used to estimate the volume. These include fan operating
parameters, production rate, and a combination of other gas condition parame-
ters. Because gas volume is not routinely monitored, neither is the actual
SCA on a day-to-day basis.
Measurement of gas flow distribution through the ESP is even less com-
mon. Because the flow measurements are obtained in the ESP rather than the
ductwork (where total gas volumetric flow rates are usually measured), more
sensitive instrumentation is needed for the low gas velocities. The instru-
ment typically specified is a calibrated hot-wire anemometer. The anemometer
SECTION 3-ESP PERFORMANCE MONITORING
3-2
-------
test is usually performed at the inlet and outlet of the ESP, but occasional-
ly they are performed at some mid-point between the inlet and outlet (usually
between two fields). Care must be taken to assure that internal ESP struc-
tural members do not interfere with the sampling points.
Gas flow distribution tests are conducted when the process is inopera-
tive and the ESP and ductwork are relatively cool. This often limits the
amount of gas volume that can be drawn through the ESP to less than 50 per-
cent of the normal operating flow; however, the relative velocities at each
point are assumed to remain the same throughout the normal operating range of
the ESP. A large number of points are sampled by this technique. The actual
number depends upon the ESP design, but 200 to 500 individual readings per
ESP are not unusual. With a good sampling protocol, any severe variations
should become readily apparent.
3.1.2 Gas Temperature
Monitoring the temperature of the gas stream can provide useful informa-
tion about the performance of an ESP and can provide useful clues for diag-
nosing both ESP performance and process operating conditions. The major
concern in temperature measurement is to avoid sampling at a stratified point
where the measured temperature is not representative of the bulk gas flow.
Thermocouples with digital, analog, or strip chart display are typical.
The effect of temperature is most important as it relates to the resis-
tivity of the particulate and as an indicator of excessive inleakage into the
gas stream. In moderately sized ESP's, changes in dust resistivity can pro-
duce large changes in performance (as evidenced by power Input to the ESP and
opacity readings). In some cases, when the resistivity versus temperature
curve is steep, a change of only 10° to 15°F may substantially change ESP
performance because it causes a shift in resistivity. This is particularly
true where high resistivity is a problem (recall Figure 2-3). Lowering the
temperature slightly to increase condensation or adsorption of surface con-
ductivity-enhancing materials is usually one available option, if neither
corrosion nor sticky particles pose a problem.
Temperature can also affect gas properties to such an extent that they
will change the relative levels of voltage and current and the density and
viscosity of the gas stream, which affect particle migration parameters.
SECTION 3-ESP PERFORMANCE MONITORING
3-3
-------
These effects, however, may go unnoticed on many precipitators, as resis-
tivity effects may overshadow them.
Lastly, comparison of inlet and outlet temperatures may be useful in the
diagnosis of excessive inleakage into the ESP, Even the best constructed and
insulated ESP will experience some temperature drop, which can range from 1°
to 2°F on smaller ESP's or up to 25°F on very large ESP's. In any case, some
acceptable difference or maximum differential should be set, and when exceed-
ed, this should be an indicator of improper operation or a maintenance prob-
lem that must be corrected.
3.1.3 Chemical Composition and Moisture
The chemical composition of both the particulate entering the ESP and
the flue gas can affect ESP performance, although in somewhat different ways,
In many process applications, either the gas composition or key indicators of
gas composition are usually available on a continuous or real-time basis.
Chemical composition of the particulate matter, however, is often not avail-
able except on an intermittent, grab-sample basis.
The operation of an ESP depends on electronegative gases (such as oxy-
gen, water vapor, carbon dioxide, and sulfur dioxide/trioxide) to generate an
effective corona and to transport the electrons from the discharge electrode
to the collection plate. The presence of one or more of these gases is
necessary to enhance the ESP performance, and the relative level in the gas
stream is not always important to ESP operation. Levels of CCL or 00, howev-
L. t-
er, are often monitored on combustion sources as a measure of excess air and
combustion efficiency and not as an indicator of the potential ESP operation,
In most processes, these electronegative gases are available and are not a
direct concern to operators.
The presence of water vapor and/or acid gases may prove useful as resis-
tivity modifiers or conditioners, and they may be necessary for proper ESP
performance. On the other hand, they may cause a sticky particulate that is
difficult to remove (see Appendix B, Kraft Pulp Recovery Boiler for discus-
sion of SOp generation as an example).
The chemical composition of the particulate matter also influences ESP
performance. Specifically, it greatly influences the range of resistivity
with which the ESP will have to operate. The presence of certain compounds
SECTION 3-ESP PERFORMANCE MONITORING
3-4
-------
such as alkalies, calcium, or other components can be used to predict resis-
tivity problems. In addition, chemical composition can change with particle
size, which may change ESP performance at the inlet, mid, and outlet sections
and further complicate prediction of ESP performance on a day-to-day basis.
From a practical standpoint, the chemical composition of the dust and
gas stream is a dynamic quantity, and any monitoring scheme may only point
out an optimum range and the variability. Monitoring the level of certain
compounds may prove useful in some instances; for example, in the combustion
of coal, sulfur content, combustibles content, and chemical composition of
the ash may provide supporting evidence when problems occur. In many in-
stances, however, chemical composition is either not monitored or it is
monitored for other purposes.
3.1.4 Particle Concentrations and Size
Electrostatic precipitators can be designed for a wide range of mass
loadings to provide satisfactory performance when combined with other opera-
ting and design parameters. They have been designed to collect loadings from
several tenths of a grain per actual cubic feet of gas to values exceeding
1-QO gr/acf. Within limitations, changes in the mass loading do not seriously
affect an ESP's performance, although some changes in outlet concentration
can occur. Other factors (e.g., design values of SCA, superficial velocity,
and electrical sectional ization and such physical properties as resistivity
and particle size distribution) are usually more important to ESP performance
when mass loading changes occur.
Mass loading at the inlet and outlet of the ESP is usually measured by
standard EPA reference methods. The difference between the amount of materi-
al in the outlet gas stream and the inlet gas stream provides the basis for
removal efficiency calculations. The use of the reference sampling methods,
however, can be difficult on very-high-efficiency ESP's or on ESP's serving
processes that generate very high mass loadings at the inlet. When outlet
mass loadings are very low, long sampling times may be required to collect
enough material to be weighed accurately. Also, simultaneous sampling of
inlet loadings during the entire test period may not always be possible if
the loadings are so high that the sampling train becomes overloaded. In some
instances, a series of probes inserted for 1 to 15 minutes to take "grab"
SECTION 3-ESP PERFORMANCE MONITORING
3-5
-------
samples of the inlet concentration may be all that is technically possible.
Although this may not provide as accurate a value for inlet mass loading as
would an "integrated" sample taken concurrently with the outlet emissions
test, it will give a reasonable value to work with.
As previously mentioned, a change in mass loading may have little effect
on ESP performance compared with the importance of other parameters. Never-
theless, a discussion of loading effects on ESP performance would not be
complete without a discussion of the effects of particle size distribution.
The particle charging mechanisms were discussed in Section 2. In the
particle size ranges where field charging dominates (above 1 urn) and diffu-
sion charging dominates (below 0,1 ym), the ESP usually performs reasonably
well. It is in this region between 0.1 and 1.0 pm, however, that most ESP's
have difficulty collecting particulate because neither charging mechanism
dominates. The minimum ESP collection efficiency is usually on particles
between 0.4 and 0.8 ym in diameter. Thus, if a change in loading is also
accompanied by a change in particle size distribution, the magnitude of these
combined changes must be evaluated to predict ESP performance. In many in-
stances a shift in particle distribution toward the 0.1 to 1.0 ym range could
be detrimental to ESP performance even though the mass loading decreases, be-
cause these particles are lighter than 5, 10, or 50 ym particles. In other
words, the total weight entering the ESP can decrease while the number of
particles actually increase, and this increase in the number of particles can
be detrimental to ESP performance if excessive numbers are in the 0.1 to 1.0
ym range.
Particle size distribution is usually determined through the use of
cascade impactors, Various types of cascade impactors are available with
different particle cut sizes and for different loadings. A typical cascade
impactor system is presented in Figure 3-1. The cascade impactor is usually
placed on a standard sampling probe and inserted into the gas stream for
isokinetic sampling of the particulate. A sampling train with a cascade
impactor is illustrated in Figure 3-2. After sampling is completed, each
stage of the impactor is weighed in the lab and compared against its initial
weight to determine distribution. Because the impactor consists of several
stages (usually 5 to 9) and each stage corresponds to a progressively smaller
SECTION 3-ESP PERFORMANCE MONITORING
3-6
-------
JET SIZE
JET VELOCITY 9
.0100" Di».
164 FT/SEC
,01OO" DM
77.O FT/SEC
,0135" Di».
42.3 FT«EC
.0210" Di.
17.SO FT/SiC
,0280" Di*
9.81 FT/SiC
.O360"
5.91 FT/SEC
.0465" Di*
3.57 FT/SEC
.0636" Di*
1.91 FT/SEC
Figure 3-1, Typical cascade impactor system (courtesy of Andersen Samplers, Inc.).
3-7
-------
METER BOX
Figure 3-2. Sampling train with cascade impactor.
3-8
-------
particle size range, the weight gain of each successive stage provides a
weight distribution of particle sizes.
Cascade impactors have two limitations: the flow rate cannot be varied
during the test run, and multiple-point samples are not usually possible on a
single sample train. The careful selection of sampling location is required
to avoid errors caused by stratification and to provide the representative
sample necessary for obtaining valid results. The particle capture charac-
teristics of a cascade impactor are calibrated against a given flow rate.
Thus, the stated particle size range for any given stage in the impactor is
referenced against a fixed flow rate. Changes in the reference flow rate to
provide isokinetic sampling in the stack will change the particle size range
that each impactor stage will capture. If the chosen flow rate is different
from the reference value, calibration curves are available for each impactor
to correct for changes in the particle size sensitivity of each impactor
stage. Thus, the flow rate through the impactor cannot be changed once it
has been established. This necessitates single-point sampling, which is
essentially a grab sample. The situation is even worse at the inlet, where
sample times may be limited to only 1 to 2 minutes because of mass loading.
More than a single-point sample may be obtained by the use of multiple cas-
cade impactors to sample a number of different points. This is both equip-
ment- and labor-intensive; however, it may provide an indication of the
representative nature of a single-point sample. In some instances, sampling
at an "average" isokinetic rate is used for traverses.
3.1.5 Resistivity of Participates
The particulate resistivity is important to the control of the electri-
cal characteristics of the ESP. Whereas resistivity has little to do with
how much charge a particle will accept (that is related to particle size), it
is a controlling factor in how much voltage and current are applied in each
field of the ESP. The voltage and current levels determine the migration
rate of charged particles and the charging rate of the particulate matter.
Resistivity primarily affects the plate. When resistivity is outside of a
very narrow range, ESP performance deteriorates. The optimum resistivity
o in
range is typically 10 to 10 ohm-cm.
SECTION 3-ESP PERrORMANCE MONITORING
3-9
-------
The resistivity of a given dust is usually controlled by its chemical
composition, the composition of the gas stream (particularly the presence of
conditioning agents such as water vapor, SO, HC1, etc.). and the gas stream
temperature. Resistivity is generally not a function of particle size al-
though some slight effects may be apparent between large and small particles
due to compaction on the plate. The resistivity of large and small particles
can be substantially different, however, if their compositions are signifi-
cantly different as a result of process operating characteristics. The
resistivity of a dust is not a static quantity; it varies with process condi-
tions and feed characteristics. Designers of an ESP can only hope that the
resistivity will stay within a relatively narrow range over the life of the
unit.
Dust resistivity is usually measured at the ESP inlet by one of two
methods: in situ and laboratory (bulk) measurement. In situ methods usually
involve collecting the particulate under actual gas stream conditions, mea-
suring changes in the voltage/current characteristics for comparison with
clean conditions, and using these changes to determine dust resistivity. The
limitations of this method are that resistivity changes due to temperature
cannot be measured and actual dust layer thickness is difficult to measure.
Bulk measurement takes place in the lab after an isokinetic dust sample is
collected and prepared. The difficulty with the bulk measurement method is
that actual gas and particulate conditions cannot be duplicated; however, a
resistivity-versus-temperature profile can be obtained with this method. The
resistivity value obtained by two methods can differ by one order of magni-
tude or more. The value of resistivity obtained by a point-to-plane in situ
method is probably more representative of the actual dust resistivity, but
both methods provide some indication of resistivity.
3.1.6 Power Input
The power input to the ESP can be a useful parameter in monitoring ESP
performance. The value of power input for each field and for the total ESP
indicates how much work is being done to collect the particulate. In most
situations, the use of power input as a monitoring parameter can help in the
evaluation of ESP performance, but some caution must be exercised.
SECTION 3-ESP PERFORMANCE. MONITORING
3-10
-------
The T-R's of most modern ESP's are equipped with primary voltage and
current meters on the low-voltage (a.c.) side of the transformer and second-
ary voltage and current meters on the high-voltage rectified (d.c.) side of
the transformer. The terms primary and secondary refer to the side of the
transformer that is being monitored', the input side is the primary side of
the transformer. Older models may have only primary meters and, perhaps,
secondary current meters. When both voltage and current meters are available
on the T-R control cabinet, the power input can be estimated. Each T-R meter
reading must be recorded.
When only the primary meters are available, the values,.for a.c. voltage
and current are recorded and multiplied; however, when secondary meters are
available, d.c. kilovolts and mi Hi amps also should be recorded and multi-
plied. When both primary meters and secondary meters are available, the
products of voltage and current should be compared. These values represent
the number of watts being drawn by the ESP; in all cases, the secondary power
output (in watts) is less than the primary power input to the T-R. The
primary and secondary meter values should not be multiplied; however, this is
done occasionally to aid in the evaluation of the ESP performance (e.g.,
primary voltage to secondary current),
The power inputs calculated for each T-R set and for the ESP do not
represent the true power entering the T-R or the effective power entering the
ESP; however, they are sufficiently accurate for the purpose of monitoring
and evaluating ESP performance. These values indicate just how well each of
the sections is working when compared with the actual voltage and current
characteristics. The ratio of secondary power (obtained from the product of
the secondary meter readings) to the primary power input will usually range
from 0.5 to 0.9; the overall average for most ESP's is between 0.70 and 0.75.
In general, as the operating current approaches the rated current of the T-R
it appears to be more efficient in its utilization of power. This is due to
a number of factors, including SCR conduction time, resistance of the dust
layer, and capacitance of the ESP. The actual voltage and current readings
that are used to calculate power will be controlled by the gas composition,
dust composition, gas temperature, and physical arrangement within the ESP.
Thus, as one moves from inlet fields towards outlet fields, the apparent
SECTION 3-ESP PERFORMANCE MONITORING
3-11
-------
secondary power/primary power ratio increases in most ESP's because the ESP's
tend to operate to their rated current output. When ESP's only have primary
voltage and current meters, the power input may be estimated by obtaining the
multiplication product.
3.2 INSTRUMENTATION SYSTEMS AND COMPONENTS
Numerous instruments may be used to monitor ESP performance and per-
formance changes. These include primary and secondary voltage and current
meters, spark meters, rapper monitors, transmissometers, hopper level indica-
tors, and the usual temperature sensors. Other monitors may be used to
determine gas conditions, such as oxygen levels; combustibles, CO, and CCL
content; and SCL and NO emissions. Combined with process data, many of
c* X
these instruments are important for determination of the day-to-day perform-
ance of the ESP. Some of the newer digital controls for ESP's will even
allow these instruments to be linked together through the appropriate inter-
face for automatic optimization of ESP performance. Each of the instruments
is discussed briefly with regard to its usefulness in ESP performance evalua-
tion.
3.2.1 Voltageand Current Meters
The readings from voltage and current meters indicate how the ESP is
performing. The individual readings themselves are usually not important,
but patterns created by these readings are. Because several different param-
eters influence the electrical readings of the ESP, it is usually the trends
in the electrical readings that are used for diagnosis of operating problems.
Figure 3-3 illustrates a typical T-R set control panel used on modern
ESP's. The four meters shown in the circuit are primary voltage, primary
current, secondary voltage, and secondary current. Spark rate meters are
also used but are not shown in this diagram. The terms "primary" and "secon-
dary" meters are defined relative to the transformer primary (low-voltage)
and secondary (high-voltage) windings.
Typical primary voltage and current meters are direct, in-line meters;
i.e., the are tied directly to the increasing line voltage and current being
applied to the transformer. Note: Because of its placement in the circuit,
SECTION 3-ESP PERFORMANCE MONITORING
3-12
-------
13
1. KILOVOLTS
2. INTERLOCK
3. MILL1AMPS
4. ON
5. OFF
6. STOP
7. START
8.LOCAL-REMOTE
9. MANUAL ADJUST
10. AUTO-MANUAL
11. OPERATE
1?. GROUND
13. AMPS (PRIMARY)
14. VOLTS (PRIMARY)
Figure 3-3. Typical T-R set control panel
(Courtesy of American Air Filter.)
3-13
-------
••the voltage applied to the T-R primary will be less than the input line
voltage (nominally 440 to 480 volts a.c.» single-phase), and this should be
reflected in the primary voltage meter. In some applications, however, the
meter does not measure current flow or voltage directly, but it provides an
indirect or proportional measurement through a transformer, and sometimes
through an amplified circuit. In most circumstances, the primary meters are
1 fairly accurate, whether they measure directly or indirectly.
The secondary meters are always an indirect method of measurement be-
cause of the operating voltages encountered on the secondary side of the
transformer. The usual location of the secondary current meter is the ground
return leg of the rectifier circuit, which may use a set of calibrated resis-
tors or an amplifier circuit to determine the current flow. The secondary
voltage is, usually measured through a voltage divider resistance network off
of the T-R output. Again, with calibrated resistors used to substantially
reduce the current flow to the meter, the applied voltage can be measured
indirectly or through an amplifier circuit. It should be noted that not all
T-R's that display secondary voltage measure, even indirectly, the voltage on
the secondary side of the transformer. Some circuits measure the primary
voltage and then assume the secondary voltage to be proportional to the
primary voltage. Review of the T-R schematics will usually indicate which
system is used.
Another difference between the primary and secondary values is the level
of voltage and current that is represented. Typically, the values reflected
on the primary meters are the root mean square (RMS) values for current and
voltage; however, the values displayed by the secondary meters are usually
average rather than the RMS values. Thus, although changes in voltage and
current that occur on the primary meter are usually reflected on the secon-
dary meters, the relative magnitudes may differ because of the way the values
are measured.
3.2.2 Spark Meter
The spark meter is usually a relatively'Simple circuit consisting of a
calibrated meter placed on a resistor/capacitor circuit. The capacitor is
charged by voltage pulses fed by the spark transient detector circuit. The
larger the number of pulses fed to the capacitor, the higher is the voltage
SECTION 3-ESP PERFORMANCE MONITORING
3-14
-------
stored by the capacitor and discharged across the resistive circuit. Al-
though this type of circuit also may be used in digital-type displays, more
sophisticated circuitry is normally used to take advantage of the digital
control technology. Because of its simplicity, this circuit works well if it
is maintained. In many instances, however, this system has failed and the
spark meter has become either inoperative or inaccurate. Where analog meters
are provided, the spark rates usually can be determined by counting the
number of sudden needle deflections over a period of time. This technique is
useful for spark rates up to about 150 to 180 per minute. Above this rate
sparks become difficult to count.
Some newer controls are equipped with spark indicators in the form of
LED's (light-emitting diodes) rather than a spark rate meter. In many in-
stances, these may also be used to indicate spark rate. Some systems may
further divide sparks as light (or "split") sparks, moderate sparks, and
heavy sparks {or arcs) for the convenience of setting up the controls. This
feature may be useful in establishing maximum voltage/current levels for
efficient operation. The spark measurement can be made on either the primary
or .secondary side of the transformer.
3.2.3 Rapper Horn'tors
Monitoring equipment for rapper control systems has been relatively
limited, ranging from no instrumentation or indication of rapper operation to
equipment that indicates intensity and operation cycle time. What is availa-
ble depends greatly on the rapper type and equipment manufacturer.
The magnetic-impulse gravity-impact (MIGI) rapper can be equipped to
indicate both the operation of the rappers and their relative intensity. The
indicator may be a bulb or LED that is activated when the control circuit
fires the signal to activate a rapper. More sophisticated rapper controls
indicate which field or which rapper is being activated. In addition, the
relative rapping intensity can be monitored by a current pulse sent to indi-
vidual rappers, and a panel meter may be provided to indicate the percent of
full current (maximum rapping intensity). Again, advances have allowed
rapper instrumentation not only to control the length of time between rapping
for each rapper, but to control individual rapper intensity. Most existing
SECTION 3-ESP PERFORMANCE MONITORING
3-15
-------
.rapper controls, however, will only provide a uniform rapping intensity to
the entire ESP or to individual fields.
The rapping intensity of mechanical, falling-hammer rappers cannot be
varied easily; therefore, the length of the rapping cycle alone is controlled
for each field. Timers are usually provided to indicate the dwell time
between rapping cycles and the length of time that the rapper drives are
activated. These two times can be varied to optimize rapping reentrainment.
Air-driven pneumatic rappers and electric vibrators usually have minimal
instrumentation; however, they may be equipped with air pressure gauges and
voltage meters, respectively, to provide an indication of rapping intensity.
Mo monitors are used for direct measurement of rapping intensity within an
operating ESP.
3.2.4 Transmi'ssometers
Transmissometers can be useful both in determining performance levels
and in optimizing ESP performance. A facility may have one or more monitors
that indicate opacity from various ESP outlet ducts and from the stack itself.
Opacity also may be measured on a real-time basis or over selectable averaging
periods.
The opacity monitor simply compares the amount of light generated and
transmitted by the instrument against the quantity received by the receiver.
The difference, which is caused by absorption, reflection, refraction, and
light scattering by the particles in the gas stream, is the opacity of the
gas stream. Opacity is a function of particle size, concentration, and path
length. Most opacity monitors are calibrated to display opacity at the stack
outlet path length. Most of the opacity monitors being installed today are
double-pass monitors; i.e., the light beam is passed through the gas stream
and reflected back across to a transceiver. This arrangement is advantageous
for several reasons: 1) automatic checking of the zero and span of the moni-
tor is possible when the process is operational; 2} because the path length
is longer, the monitor is more sensitive to slight variations in opacity; and
3) the electronics package is all located on one side of the stack as a
transceiver. Although single-pass transmissiometers are available at a lower
cost (and sensitivity), the double-pass monitor can meet the requirements for
SECTION S-ESP PERFORMANCE MONITORINQ
3-16
-------
zero and span in Performance Specification 1, Appendix B,-4Q CFR 60. Monitor
siting requirements are also discussed in this specification.
For many sources, mass-opacity correlations can be developed to provide
a relative indication of ESP performance. Although site-specific, these
correlations can provide plant and agency personnel with an indication of
relative performance levels for a given opacity and deterioration of perform-
ance that requires attention by plant personnel. In addition, the opacity
monitor can be used to optimize spark rate, voltage/current levels, and
rapping cycles, even though the conditions within the ESP are not static.
In some instances, it may take 6 months to optimize ESP rapping patterns
and intensity to obtain the best electrical conditions and minimum reentrain-
ment of partlculate. One difficulty is the time required for the ESP to
establish a new "equilibrium" dust layer on the plates. This is complicated
by the ever-changing conditions within the ESP. In high-efficiency ESP's,
however, reentrainment may account for 50 to 70 percent of the total outlet
emissions, and optimization of the rapping pattern may prove beneficial.
Transmissometer strip charts have been observed on well-operated and moder-
ately sized ESP's that exhibit practically no rapping reentrainment spikes.
Rapping reentrainment must be observed with the monitor operating in a real-
time or nonintegrating mode (also called a "zero" integration time) such as
the example shown in Figure 3-4. Rapping spikes tend to get smoothed out in
integrated averages ^such as the 6-minute average commonly in use. The inte-
grated average does provide a good indication of average opacity and emissions,
however.
When parallel ESP's or chambers are used, an opacity monitor is often
placed in each outlet duct, as well as on the stack, to measure the opacity
of the combined emissions. Although the stack monitor is .commonly used to
indicate stack opacity (averaging opacities from different ducts can be dif-
ficult), the Individual duct monitors can be useful in indicating the per-
formance of each ESP or chamber and in troubleshooting. Although this option-
is often not required and it represents an additional expense, it can be very
useful, particularly on relatively large ESP's.
New systems are available in which the opacity monitor data can be used
as input for the T-R controller, for example, the new digital microprocessor
SiCTlON 3-ESP PERFORMANCE MONITORING
3-17
-------
Figure 3-4. Example of rapping spikes on a transmissometer
strip chart.
3-18
-------
designs that are being sold or developed for new installations. Use of the
opacity monitor data can decrease power input throughout the ESP to maintain
some opacity level preselected by the source. If the opacity increases, the
controller increases power input accordingly until the opacity limit, spark
limit, current limit, or voltage limit is reached. This system (often sold
as an energy saver) can save a substantial quantity of energy on large,
high-efficiency ESP's, and at reduced gas loads. In many cases, reduction of
ESP power does not significantly alter ESP performance because reentrainment
and gas sneakage constitute the largest source of emissions, and additional
power often does not reduce these emissions significantly. In some observed
cases, reducing power by one-half did not change the performance. For units
typically operated at 1000 to 1500 watts/1000 acfm, power levels of 500 to
750 watts/1000 acfm still provide acceptable collection efficiencies.
3.2.5 Hopper Level Indicators
Hopper level indicators could more accurately be called high hopper lev-
el alarms because they do not actually measure dust levels inside the hopper.
Instead they send an alarm that the dust level within a hopper is higher than
the level detector and that corrective action is necessary.1- The level detec-
tor should be placed high enough that "normal" dust levels will not continu-
ously set off the alarm, but low enough to allow adequate response time to
clear the hopper before the dust reaches the discharge frame and causes the
T-R to trip or to misalign the ESP. Not all ESP's use or need hopper level
indicators.
Two types of level indicators are commonly in use (although others are
available). The older of the two is the capacitance probe, which is inserted
into the hopper. As dust builds up around the probe, a change in the capaci-
tance occurs and triggers an alarm. Although these systems are generally
reliable, they can be subject to dust buildup and false alarms in some situa-
tions. A newer system, currently in vogue, is the nuclear or radioactive
detector. These systems utilize a shielded Cesium radioisotope to generate a
radioactive beam that is received by a detector on the opposite side of the
hopper. Two of the advantages of this system are they do not include a probe
that is subject to dust buildup and more than one hopper can be monitored by
one radioactive source. The major drawback is thatthe plant personnel would
SECTION 3-ESP PERFORMANCE MONITORING
3-19
-------
, be dealing with a low-level radioactive source and adequate safety precau-
tions must be taken. These detectors are provided with safety interlocks to
prevent exposure of plant personnel when maintenance is required.
Hopper level detectors should normally be located between one-half and
two-thirds of the way up the side of the hopper. As long as hoppers are not
used for storage, this should provide an adequate safety margin. It should
be remembered that it takes much longer to fill the upper 2 feet of a pyramid
hopper than the lower 2 feet.
Other indirect methods are available for determining whether the hopper
is emptying properly. On vacuum discharge/conveying systems, experienced
operators can usually tell where the hopper is plugged or if a "rat-hole" is
formed by checking the time and vacuum drawn on each hopper as dust is re-
moved. On systems that use a screw conveying system, the current drawn by
the conveyor motor can serve as an indicator of dust removal. Another simple
method for determining hopper pluggage is through a thermometer located
approximately two-thirds of the way up on the hopper. If dust covers the
probe because of hopper buildup, the temperature will begin to drop, which
signals the need for plant personnel to take corrective action.
3-2.6 His eel 1 a neou s Egu i pment
CO and Combustibles Analyzers-
Some source categories, most notably portland cement plants, use CO or
combustible analyzers as a safety device to detect high levels and to prevent
high levels that form a combustible mixture from entering an energized ESP
where a spark could ignite or detonate the mixture and cause substantial
damage to the ESP. Although this is usually related to process malfunctions,
the damage to the ESP can seriously hinder performance and limit production
rates in the future.
Each of several instruments and measurement methods on the market today
has its advantages and limitations. All have the common function of indicat-
ing an undesirable condition, sending an alarm, and possibly automatically
deenergizing the ESP until the danger has passed. Many sources monitor CO
where excursions are expected, because CO in combination with oxygen at the
right concentration can form a combustible mixture. Other compounds [e.g.,
methane (CHa) and hydrogen (H^)] may be more important in the prevention of
SECTION 3-ESP PERFORMANCE MONITORING
3-20
-------
explosions, however. These compounds increase the range of explosive mix-
tures by changing the upper and lower limits of the CQ-air mixture and en-
hance the possibility of explosion within the ESP. Most CO monitors do not
monitor combustibles and are relatively insensitive to them. Thus, an explo-
sive mixture can exist within an ESP without the limiting CO value that sets
off an alarm ever being reached.
Two other problems may occur with regard to monitoring CO and combusti-
bles. First, monitors that rely on catalytic combustion to determine combus-
tibles must have sufficient available oxygen to complete the reaction to form
C£L and water. This level is generally greater than 3 percent CL in the flue
gas. The second and perhaps more serious problem is the monitor response
time. Monitor response times on the order of 45 seconds to 2 minutes are not
unusual. In cases where the CO or combustible spike is detected because of
some process upset, the danger often has passed by the time the monitor
responds. Monitors and special probes are available that have much shorter
response times (2 to 5 seconds), but these tend to be quite expensive and
have a relatively short life. The cost may be more than offset, however, by
the cost of having to repair or rebuild an ESP and other components as a
result of an explosion.
3.3 PERFORMANCE TESTS AND PARAMETER MONITORING
The operating characteristics of an ESP are such that several concepts
are useful in a performance evaluation. These include parameter monitoring
and baseline assessments. These concepts, which will be defined further,
form the basis for good recordkeeping and a preventive maintenance program
aimed at achieving continuous compliance of the controlled source.
From a regulatory standpoint, compliance is determined by a performance
test involving the use of a Reference Method, such as Method 5 or 17. In
between these periodic performance tests, compliance also may be determined
by the use of opacity observations according to the requirements of Reference
Method 9. Because these emission tests represent only a small segment of
time in the daily operation of the process and ESP, the performance during
the emissions test may not be representative of characteristic daily opera-
tions. Nevertheless, these emissions tests do afford the opportunity to
SECTION 3-ESP PERFORMANCE MONITORING
3-21
-------
document process and ESP operating conditions that influence performance. By
providing a known level of performance, these values serve as a benchmark or
baseline condition for future comparisons with data collected during routine
parameter monitoring and recordkeeping or as part of diagnostic troubleshoot-
ing, The establishment of these baseline conditions makes 1t possible for a
number of parameters to be compared to determine their effect on performance.
This comparison is particularly useful for ESP's because many parameters can
affect performance to varying degrees. It is the magnitude of these changes
that is important. Baseline conditions may include both air-load and gas-
load tests in addition to the data obtained during emission testing.
Parameter monitoring, an extension of baselining the ESP and process
equipment, forms the basis of diagnostic recordkeeping and preventive mainte-
nance. Several key parameters are usually monitored to track ESP performance.
Generally, parameter monitoring includes both process and ESP data because
both are important to ESP performance. An analysis of these key parameters
and a comparison with baseline values can define many performance problems,
indicate the need for maintenance, and define operating trends within the
ESP. For example, at a utility burning several types of coal, performance
may deteriorate or be enhanced by one type of coal or another. Rather than
accepting these performance changes as part of the daily variation, these
fuel-related variations may be avoided by exercising special operating pre-
cautions or perhaps eliminating a certain coal type. Determining the cause
of variations can often be more beneficial than merely accepting the problem
at face value. Parameter monitoring is most useful in moderate-sized to mod-
erately large ESP's. Undersized or oversized ESP's tend to be less sensitive
to changes in the key operating parameters in the efficiency ranges of great-
est interest.
3.3.1 Performanee Test_s
The performance test often is the deciding factor for the acceptance of
a new ESP, and many agencies require periodic testing (anywhere from quarter-
ly to once every 3 to 5 years). The initial performance test certifies that
the ESP is designed to be capable of meeting the specifications. These
initial performance tests may also include tests with sections of the ESP out
of service to meet special requirements of the permit, specifications, or
SECTION 3-ESP PERFORMANCE MONITORING
3-22
-------
regulatory requirements, The initial performance tests may also include
inlet tests to establish mass loading, collection efficiency, and, in some
cases, inlet particle size characteristics.
Testing requirements vary from site to site, and they should be estab-
lished in a testing protocol", however, one of two test methods is generally
specified for determining particulate emission rates. These are EPA Refer-
ence Methods 5 and 17 (40 CFR 60, Appendix A). In both methods, a sample is
removed isokinetically from various sample locations to prevent the sample
results from being biased. The main difference between the two methods is
the location of the filter in the sample train. The Method 5 sample train
uses an external filter held in a temperature-controlled hot box. The sample
passes through the heated sample probe and filter into the impinger train for
removal of condensible materials (water, condensible organics). The particu-
late emission rate usually will be determined from the probe and filter catch
alone (front-half catch); however, some regulatory limitations specify the
use of both front- and back-half catches (which include the impinger catch
minus water). In most cases, the specified temperature is 250° ± 25°F for
the filter of a Method 5 sample train, but special conditions allow a temper-
ature up to 32GP ± 2S°F.
Method 17, on the other hand, uses an in-stack filter to capture partic-
ulate. After the filter temperature has been allowed to equilibrate to stack
conditions, the sample is drawn through the nozzle and into the filter. The
sample is then passed through a set of impingers to remove condensibles from
the gas stream.
The two methods often do not provide equivalent results, even when the
flue gas temperature is the same as the hot box temperature. First, Method 5
defines particulate as the material that is captured on a filter at the hot
box temperature (nominally 250°F), although the temperature of the gas stream
passing through the filter may be substantially different than the hot box
temperature. This is important because many "participates" are temperature-
dependent, i.e., they exist below a certain temperature, but they may remain
in a gaseous form above a given temperature. Theoretically, particulate mat-
ter is referenced to a particular temperature. Second, some losses are nor-
mally associated with recovering the particulate from the in-stack filter of
SECTION 3-ESP PERFORMANCE MONITORING
3-23
-------
the Method 17 sample train, and a correction must be applied (e.g., 0.04
gr/scf for kraft recovery boilers). Lastly, because Method 17 is referenced
to the stack temperature, the definition of what constitutes "particulate"
may be different. Method 17 is usually reserved for participates or process-
es that do not involve a particular temperature dependency. When this is the
case, the results of both methods are usually in relatively close agreement.
There has been widespread discussion as to which method should be used
when a choice is allowed. Each method can be manipulated to provide the most
favorable emission rate. Method 17 may more accurately reflect the condi-
tions the ESP may encounter; however, Method 5 attempts to standardize the
operating temperature so that differences in temperature and temperature
dependency are minimized.
In addition to overall particulate emission rates, some regulations
limit the emission of fine particles. Such limits require particle size
analysis, either by microscopic methods or by the use of cascade impactors.
Cascade impactors are placed in the stack in a manner similar to the place-
ment of an in-stack filter. An impactor consists of a series of perforated
plates and target or impact stages in which an impact medium (grease or
fiberglass substrate) is used. As the gas and particulates pass through the
impactor, they are accelerated to higher velocities. The particulate matter
has difficulty staying with the flow streamlines and its inertia carries it
to impact the target stage. Each stage is sized to capture a predetermined
particle size range at a given flow rate. Calibration curves and corrections
to the particle size ranges are provided by the equipment manufacturers.
There are two problems related to the use of cascade impactors. First,
the gas stream must be sampled isokinetically to avoid skewing the particle
size distribution. Under-isokinetic sampling (at a probe velocity less than
that of the stack or duct) usually results in a distribution skewed toward
large particles, whereas over-isokinetic sampling favors smaller particulate.
Also, the particle size distribution may have little bearing on the Method 5
results because of the temperature-dependency of the particulate. Second,
the isokinetic sampling requirement means that the sample must be drawn at a
given flow rate and the flow rate cannot be varied to maintain calibration of
SECTION 3-ESP PERFORMANCE MONITORING
3-24
-------
the impactor. Varying the flow rate would vary the particle size distribu-
tion each stage would capture. This usually results in single-point sampling
being used, with all the limitations associated with the representativeness
of the sample. In some cases, multipoint sampling may be carried out, but
careful planning and, for multiple impactors, good flow distribution are
necessary.
Opacity is usually monitored during the performance test with an opacity
monitor and/or by Method 9 observations. Some efforts have been made to
correlate mass and opacity to provide a performance indicator by conducting
multiple tests while altering ESP performance (by turning power down or off
on certain T-R's) to generate upscale data points. Opacity readings taken
during each test run (either monitor data or Method 9 observations) are then
averaged.
Several sources have done some work to establish mass/opacity correla-
tions, most notably in utility applications, which prompt the following
general observations. First, mass/opacity correlations appear to depend on a
number of industry- and site-specific factors, including size distribution,
stack path length, and process-related factors. Second, at a number of
sources, consistent relationships have been found over a period of time
(regardless of process load) provided neither the process nor the ESP is
experiencing severe malfunctions. The process operation seems to be the
controlling factor in most cases. For example, much of the data for utility
applications suggests that the mass/opacity relationship is relatively con-
stant for a given source; however, a burner problem that produces high carbon
content in the flash can shift the mass/opacity. A similar shift occurs on
kraft recovery boilers and some cement kilns. In general, when a condensible
particulate is present, the mass/opacity correlation may not be reliable.
Third, opacity monitor data tend to produce "tighter" or better correlations
than Method 9 when 95 percent confidence intervals are calculated for the
correlation. Observer biasing can result from changing background conditions
or from "between-observer" differences; whereas a properly maintained monitor
usually is not subject to biasing problems. Lastly, confidence intervals
tend to become very large at the extreme ends of the curves when mass and/or
opacity is either very low or high. At the low mass/opacity end of the
SECTION 3-ESP PERFORMANCE MONITORING
3-25
-------
curve, the relative errors, particularly in test methods, can become substan-
tial. Although the mass loading is very high at the upper end, little change
may occur in opacity. In fact, so much of the ESP may be deenergized that it
may not behave as it would if it were running normally, it may behave as
though it were designed differently and significantly undersized. In the
opacity ranges of interest to most agencies and sources, however, the confi-
dence intervals can be quite tight. This opacity correlation, although not
usually used for compliance determinations, can be useful in evaluating
operation and maintenance, which was the original intent behind the require-
ment for continuous emission monitors.
3.3.2 Baseline Assessment
Although baseline assessment actually should begin before a new ESP is
operated, the establishment of baseline conditions for an ESP during a per-
formance test provides a basis for comparison in future evaluations of the
ESP. The baseline serves as a reference point, and the types and magnitude
of shifts from baseline conditions are important in evaluating ESP perform-
ance.
For a new ESP, baselining includes an air-load test prior to operation
while the plates and wires are in a clean condition. After work has been
completed and the final physical checks have been made, the ESP is closed and
prepared for energization. An air-load test should be run on each field to
generate a V-I curve. The values for the voltage/current relationship should
be similar for fields of the same design. Items such as T-R capacity, square
feet/T-R (sectionalization), and wire design (barbed vs. smooth) will influ-
ence the shape of the curves, as will any physical defect in construction.
This test serves as a check on the electrical capacity of the unit as well as
a check on the construction of the ESP while it is still in a clean condition
and problems can be corrected. When the unit has been started up and operat-
ed, the air-load curves will be different because of the residual dust on the
plates.
Air-load tests prior to a performance test (in case of a shutdown prior
to the test) will also confirm the electrical performance of the ESP. Gas-
load tests also can be performed before and during a performance test. The
gas-load test is performedunder actual gas flow and particulateconditions
SECTION 3-E5P PERFORMANCE MONITORING
3-26
-------
to generate a V-I curve for each field. Several items should be noted. The
shape and length of the curves will vary somewhat because the process and ESP
form a dynamic system. The absolute values of voltage and current are not as
important as the trends observed within the ESP; these trends help in the
evaluation of the ESP's operating condition during the performance tests.
Gas-load tests can be performed before, during, and after a performance test,
with one cautionary note. Because conducting a gas-load test often upsets
ESP performance for 5 to 20 minutes, some time should be allowed for the ESP
to stabilize before testing (particularly between runs of the performance
test).
Most of the effort expended during the collection of baseline perform-
ance readings is in the area of process and ESP operating data. Process data
may include production weight, raw material and product feed characteristics,
operating temperatures and pressures, combustion air settings, and cycle
times (for cyclic processes). The ESP data will include electrical readings
(usually several readings per run), temperature, gas flow rate, opacity,
rapping cycle, and excess air levels.
Although accurate predictions cannot be made of the exact effect a
change in most of these parameters will have on performance, a qualitative
evaluation can often be made when values deviate from baseline conditions,
and these deviation values are useful in parameter monitoring.
3.3.3 Parameter Monitoring
Parameter monitoring usually plays a key role in an overall operation
and maintenance plan, particularly one that stresses preventive maintenance.
Such monitoring also forms the basis for a recordkeeping program that places
emphasis on diagnostics. Typically, daily operating data are reduced to the
data on a few key parameters that are monitored. Acceptable ranges may be
established for various parameters (by use of baseline test data) that re-
quire further data analysis or perhaps some other action if the values fall
outside a given range. Care must be taken not to rely on just one parameter
as an indicator, as other factors, both design- and operation-related, usual-
ly must be considered. Typical parameters that can be monitored include
opacity, corona power input, gas flow rate through the ESP, gas temperature
SECTION 3-ESP PERFORMANCE MONITORING
3-27
-------
and oxygen content, process operating rates, and conditioning systems (if
used).
Many sources use opacity levels as the first indicator of performance
changes. In general, opacity is a good indicator and tool for this purpose.
If used in conjunction with mass/opacity correlations, it can help in the
scheduling of maintenance and in the reduction or optimization of ESP power
input. It is not wise to rely on opacity data alone, however, as such reli-
ance can cause one to overlook problems that can affect long-term performance
(e.g., hopper pluggage may not significantly increase opacity at reduced
load, but it may misalign the affected fields and reduce their performance at
full load or in other difficult operating situations).
Another useful parameter is the corona power input to the ESP, which can
be thought of as a measure of the work done to remove the particulate. Coro-
na power input can be obtained by multiplying the voltage and current values
of either the primary or secondary side of the transformer. As noted earli-
er, the apparent power input on the primary and secondary side will differ
because of circuitry and the metering of these values. Values from the
secondary meters are preferred. As a general rule, performance improves as
the total power input increases. This is normally the case when resistivity
is normal to moderately high, assuming most other factors are "normal" and
components are in a state of good repair. One should not rely solely on
power input, however. The pattern or trends in power input throughout the
ESP are important in a performance evaluation. Also in some cases, although
the apparent power input is high, the performance is poor. For example, when
dust resistivity is low or very high or when spark rates are very high, cor-
rective measures will usually lower power input, but will also substantially
improve performance.
Some ESP's are relatively insensitive to power input changes. This con-
dition is usually limited to high-efficiency ESP's that are generously sized
and sectionalized. The normal power input of some of these ESP's may be
reduced by one-half to two-thirds without causing any substantial change in
performance. The emissions from the ESP's are caused primarily by rapping
reentrainment and gas sneakage, both of which are relatively unaffected by
SECTION 3-ESP PERFORMANCE MONITORING
3-28
-------
the level of power input. In this case, power reduction for energy conserva-
tion is probably a useful option.
Varying corona power input affects power density (watts/square foot of
plate area). This may be tracked two ways: 1) by obtaining an overall value
for the ESP, or 2) by checking the power density in each field from inlet to
outlet of the ESP. Power density should increase from inlet to outlet as the
particulate matter is removed from the gas stream (the maximum value is usu-
2
ally less than 4 watts/ft ). Power density accounts for the differences or
normalizes the values for power input'in each field that are caused by dif-
ferent field size. Most normally operating ESP's will show an overall power
2 2
density of 1 to 2 watts/ft ; values of 0.10 to 0.50 watt/ft are more common
for high-resistivity dusts.
The gas volume passing through the ESP is important to the actual SCA,
the superficial velocity, and the treatment time. The temperature of the gas
stream, the excess-air values (for combustion sources), and the production
rate all influence the gas volume entering the ESP. If gas volume is known
or estimated, the specific corona power (watts/1000 acfm) can be calculated.
This value tends to account for changes in performance due to different loads
and power input because removal efficiency generally increases as the specif-
ic corona power increases. The same cautionary remarks that apply to overall
power input also apply to specific corona power. The values obtained for
specific corona power input may be misleading if other factors are not con-
sidered.
3.4 RECORDKEEPING PRACTICES AND PROCEDURES
Recordkeeping practices for ESP's range from none to maintaining exten-
sive logs of operating data and maintenance activities and storing them on
computer disks. The data obtained by parameter monitoring form a basis for
recordkeeping, as this type of data usually indicates ESP performance. Rec-
ordkeeping allows plant personnel to track ESP performance, evaluate trends,
identify potential problem areas, and arrive at appropriate solutions. The
magnitude of the recordkeeping activity will depend on a combination of fac-
tors, such as personnel availability, size of the ESP, and the level of main-
tenance required. For moderately sized, well-designed, and well-operated
SECTION 3-ESP PERFORMANCE MONITORING
3-29
-------
ESP's, maintaining both daily operating records and maintenance records
should not be too cumbersome; however, only records of key operating parame-
ters should be maintained to avoid accumulating a mountain of unnecessary
information.
Recordkeeping practices can be separated into two major areas, operating
records and maintenance records, each of which can be further divided into
subcategories. When setting up a recordkeeping program, one should give at-
tention to both areas because they are required to provide a complete operat-
ing history of the ESP. This operating history is useful in an evaluation of
future performance, maintenance trends, and operating characteristics that
may increase the life of the unit and minimize emissions. Even though rec-
ordkeeping programs are site-specific, they should be set up to provide
diagnostic and troubleshooting information, rather than merely for the sake
of recordkeeping. This approach makes the effort both worthwhile and cost-
effective.
Other supplementary records that should be maintained as part of the
permanent file for operation and maintenance include data from air-load tests
conducted on the unit, all baseline assessments that include both process and
ESP operating data, and data from emission tests. A spare parts inventory
listing should also be maintained, with periodic updates so that parts may be
obtained and installed in a timely manner.
3.4.1 Operating Records
As mentioned previously, the specifics on what parameters will be moni-
tored and recorded and at what frequency will be largely site-specific.
Nonetheless, the factors that are generally important in parameter monitoring
will also be the ones recorded as part of a recordkeeping program. Such data
would typically include the process operating rate, T-R set readings, opacity
monitor readings, and perhaps some reduced data on power input, power densi-
ty, or specific corona power. These data probably should be gathered at
least once per shift. The greater the frequency of data gathering, the more
sensitive the operators will be to process or ESP operational problems, but
the amount of data to manipulate also increases. The optimal frequency may
SECTION 3-ESP PERFORMANCE MONITORING
3-30
-------
be every 4 hours (twice per shift). If sudden and dramatic- changes in per-
formance occur, if the source is highly variable, or if the ESP operation is
extremely sensitive, shorter monitoring intervals are required. On gener-
ously sized ESP's, intervals of once or twice a day may be adequate.
In addition to the numerical values of the operating parameters, a check
list should be included to confirm operation of rappers, hopper systems (or
other dust-removal systems), the absence of inleakage, and the other general
physical considerations that can adversely influence ESP performance.
3,4.2 Mainten a nee R e co rd s
Maintenance records provide an operating history of an ESP. They can
indicate what has failed, where, and how often; what kind of problems are
typical; and what has been done about them. These records can be used 1n
conjunction with a spare parts inventory to maintain and update a current
list of available parts and the costs of these parts.
The work order system provides one of the better ways to keep mainte-
nance records. When properly designed and used, this system can provide
Information on the suspected problem, the problem actually found, the correc-
tive action taken, time and parts required, and any additional pertinent
information. The system may involve the use of triplicate carbon forms or it
may be computerized. As long as a centralized system is provided for each
maintenance activity, the work order approach usually works out well.
Another approach is to use a log book in which a summary of maintenance
activities is recorded. Although not as flexible as a work order system
(e.g., copies of individual work orders can be sent to various appropriate
departments), it does provide a centralized record and is probably better
suited for the small ESP facility.
In addition to these centralized records, a record should be maintained
of all periodic checks or inspections. These should include the periodic
weekly, monthly, semiannual, and annual checks of the ESP that make up part
of a preventive maintenance program. Specific maintenance items identified
by these periodic inspections should be included in the recordkeeping
process. The items to be checked are discussed in more details in Section 6.
SECTION 3-ESP PERFORMANCE MONITORING
3-31
-------
3.4.3 Retrieval of Records
A computerized storage and retrieval system is ideal for recordkeeping,
A computer can manipulate and retrieve data in a variety of forms (depending
upon the software) and also may be useful in identifying trends. A computer-
ized system is not for everyone, however. The larger the data set to be
handled, the more likely it is that a computer can help to analyze and sort
data. For a small source with an ESP that presents few problems and that has
a manageable set of operating parameters to be monitored, a computer system
could be very wasteful (unless computing capability is already available).
Also, it is sometimes easier to pull the pages from a file manually, do a
little arithmetic, and come up with the answer than to find the appropriate
disks and files, load the software, and execute the program to display "the
answer."
Retention time is also a site-specific variable. If records are main-
tained only to meet a regulatory requirement and are not used or evaluated,
they can probably be disposed of at the end of the statuatory limitation
(typically 2 years). It can be argued that these records should not be
destroyed because if the ESP (or process) should fail prematurely, the data
preserved in the records could be used as an example of what not to do. On
some ESP's in service today, records going back 10 to 12 years have been kept
to track the performance, cost, and system response to various situations and
the most effective ways to accomplish things. These records serve as a
learning tool to optimize performance and minimize emissions, which is the
underlying purpose of recordkeeping. Some of these records may very well be
kept throughout the life of the equipment. After several years, however,
summaries of operation and maintenance activities are more desirable than the
actual records themselves. These can be created concurrently with the daily
operating and maintenance records for future use. If needed, actual data can
then be retrieved for further evaluation.
SECTION 3-ESP PERFORMANCE MONITORING
3-32
-------
SECTION 4
PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS,
AND PROBLEM SOLUTIONS
Although ESP performance 1s complex and sensitive to a number of varia-
bles associated with the process, it can usually be related to its internal
electrical characteristics. These electrical characteristics, which are
usually monitored by the T-R set control cabinet voltage and current read-
ings, serve as the basis for performance evaluation and troubleshooting of
the ESP. Thus, it is important for personnel responsible for evaluating or
maintaining ESP performance to maintain good records and to understand the
significance of these recorded values. Proper evaluation of the data may
have a significant effect on both the short- and long-term performance of the
ESP as well as maintenance requirements.
Difficulties often arise in the interpretation of the data because plant
personnel do not realize what is "normal" in their ESP or because they do not
understand the importance of these values to ESP performance. Establishing
what is "normal, good performance" through good recordkeeping practices helps
to provide a data base against which responsible personnel may compare daily
operating values. This "baseline" condition serves as a benchmark against
which to compare changes in operating conditions of either the process or ESP
and to help decide what effect, if any, these changes have had on ESP per-
formance. Although an emission test is the ultimate indicator of compliance
with emission limitations, some method should be available for agency person-
nel or plant personnel to evaluate the nature and magnitude of any problem(s)
in an operating ESP without resorting to a stack test every time there is a
suspected problem.
A further complication in the evaluation of ESP performance is that a
change in operating characteristics may be a symptom of not just one problem,
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
-------
but several unrelated problems. Enough data must be gathered for the symp-
toms to reduce the potential problems to one or two possibilities. In addi-
tion, synerglstic or "chain-reaction" effects often result in one problem or
failure leading to another, which in turn leads to a third failure and so on.
This sometimes makes it difficult to decide what the initial failure was and
what corrective action is needed.
This section discusses the uses of available data in evaluating ESP
performance and diagnosing the more common ESP problems and the corrective
actions available for both long- and short-term improvement of performance.
Much of the long-term improvement relies on good recordkeeping practices
(both operation and maintenance records) as discussed in Section 3.4.
4.1 PERFORMANCE EVALUATION
Most of the performance changes that occur in ESP's are reflected in the
electrical characteristics that are monitored and controlled by the T-R set
control cabinets. These changes may be caused by a failure of some internal
ESP component or by a change in process operation. Because some changes are
very subtle (e.g., a change in gas temperature or excess air level, a change
in primary/secondary air ratios in a recovery boiler, or a shift in the feed
material characteristics), monitoring and recording the pertinent operating
parameters are important aspects of a performance evaluation.
At some sites all important parameters may not be monitored, and some
that are monitored and recorded may be of little use in a performance evalua-
tion. Although a site-specific evaluation will depend on the process in-
volved and the instrumentation available, most ESP's should be equipped with
at least primary voltage, primary current, and secondary (or precipitator)
current monitoring equipment. Newer designs will incorporate a secondary
voltage monitor and may forego altogether the monitoring of primary voltage
and current. These meters will be used to assess ESP performance and to
diagnose operating problems.
Obtaining and maintaining these data for diagnostic purposes can become
quite cumbersome, particularly with some of the newer ESP designs that are
generously sized and highly sectionalized. Plant personnel may find it
relatively simple to keep track of three, four, or five T-R's and some key
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS 4-2
-------
process operating parameters on a day-to-day basis; but maintaining records
of readings on anywhere from 10 to 200 T-R's becomes a very difficult, if not
completely unmanageable, task. Data acquisition and retrieval systems help
in these situations, but detailed analysis may also be necessary. Thus, in
any performance monitoring or audit program, one must first decide what will
be monitored and recorded and in what form. Again, this will depend on site-
specific equipment and design factors.
Two considerations are necessary in any performance audit or evaluation
of an operating ESP. The first concerns the design factors that are built
into the ESP. These include such parameters as the specific collection area
(SCA), number of fields, number of T-R's, electrical sectionalization, T-R
set capacity, design superficial velocity and treatment time, aspect ratio,
and partlculate characteristics. This background information permits the
auditor or evaluator to determine what the ESP was designed to do and whether
operating parameters have changed significantly from design. The second
consideration concerns the use of baseline data to establish normal or good
operating conditions. These baseline data could consist of values recorded
during an emissions test or could be a compilation of operating records to
establish normal operating conditions.
No single parameter should be used to evaluate performance; a combina-
tion of factors is more likely to be reliable. Although some parameters are
more important and have greater effect than others, it is usually the combi-
nation of these parameters that determines performance of the ESP.
Depending on the situation, at least one person should be responsible
for overseeing the operation of the ESP's, for reducing the data to a usable
form, performing the evaluations, identifying potential problems, and helping
to schedule maintenance. Other personnel may gather data, but one person
should understand the significance of the gathered data.
How often and how much data should be gathered is a site-specific deci-
sion that will depend on equipment size, design factors, and personnel avail-
ability. The purpose of these data is to provide sufficient information for
an effective evaluation of performance. Extraneous data of limited value
should not be collected unless a specific problem is expected or encountered.
When a program of recordkeeping is just being established, however, it is
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS. AND PROBLEM SOLUTIONS
4-3
-------
better to err on the side of having more data than needed at the beginning.
Unnecessary duplication of recordkeeping should be avoided. This is partic-
ularly true of process information involving final quality data that may
already be retained by plant personnel. These process records should be
available to the personnel responsible for monitoring ESP performance if more
detailed data are needed beyond that recorded for the ESP performance evalua-
tion,
4.1.1 Data Col 1ection and Compi1ati on
Recording T-R Set Data—
The primary indicators of performance are the electrical operating con-
ditions monitored at the T-R control cabinet. These conditions are reflected
in the primary voltage, primary current, secondary voltage, and secondary
current. Even if all these are not monitored (on older ESP's secondary
voltage often is not monitored, and on some newer ones primary voltage and
current are not monitored), the values provided should be recorded.
The level of effort required for this task depends on the size of the
ESP and on the number of parameters monitored. For relatively small ESP's
equipped with two to five T-R sets, very little time is required to record
the data. The time required for the larger and more sectionalized ESP's,
however, can be substantial.
The T-R data may be recorded in tabular form with the appropriate data
for each T-R set. Again, for ESP's with a small number of T-R sets, this
form makes it relatively easy to assemble the data and to track inlet, cen-
ter, and outlet field performance. Plant personnel will be looking for
certain patterns that are indicators of ESP performance levels. For larger
ESP's the tabular form speeds up data gathering, but it does not immediately
provide a visual pattern of ESP performance. For example, in most well-
designed, operated, and maintained ESP's, the current tends to increase from
the inlet field to the outlet field. In a large ESP with 10 or more fields,
1t may be difficult to visualize this effect if the physical placement of the
T-R control cabinets or the arrangement of the T-R numbering system do not
permit recording the data from inlet to outlet.
When the tabular form is less than satisfactory, a more graphical
approachcanbe taken. Several graphicalapproaches are available for
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSJS, AND PROBLEM SOLUTIONS
4-4
-------
obtaining these data in a more useful form. The simplest of these is to draw
the ESP plot plan with the relative position of the plate area of each T-R
blocked out and to place the electrical data for each T-R in the appropriate
box {see Figure 4-1). This is useful for evaluating the performance of large
ESP's and those having fields of different dimensions. When the data gather-
ing is completed, a look at the values for each field will quickly indicate
if the desired pattern is there. This graphical representation will also
show how many fields are out of service and how severe the problem may be
(see Figure 4-2). Total plate area of out-service for the entire ESP can
then be observed over a number of days, as shown in Figure 4-3.
Another graphical method is to plot the electrical data on a graph for
each field from inlet to outlet (one for each chamber or grouping of T-R's,
if necessary). This also allows a visual evaluation of the data for charac-
teristic patterns (see Figure 4-4). Usually, all electrical parameters do
not have to be plotted, as secondary current and voltage are good first
indicators. For ESP's with different plate areas per T-R, it may be useful
to normalize the data by dividing the values by the square feet of plate for
each T-R. The resulting current densities should reflect the desired pattern
of increasing current from inlet to outlet. (Note: In some cases, the same
ESP will have T-R sets with energized fields varying from 1.5 to 12 feet in
depth. These ESP's may have been designed this way, or they may have changed
over time. In either case, the non-normalized values may reveal strange
electrical characteristics,}
Although both of these graphical techniques are good for collecting and
visualizing electrical characteristics, they have the shortcoming of only
depicting ESP performance during the period when the values are taken-, they
cannot reflect trends in performance over time.
Another graphical method that can be used to evaluate long-term changes
in ESP performance involves plotting the values of interest on a time chart
(time on x-axis, voltage/current on y-axis). Two examples of this technique
are shown in Figure 4-5. This chart allows maintenance personnel to note any
changes that are occurring and the rate of change. Although this system of
data compilation provides an excellent visual analysis of operating trends,
it does not provide a good means of comparing voltages and currents directly
SECTION 4-PEBFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-5
-------
3|H
t
3 G
]
3 F
3?E
i
3-D
3|C
«*
31 B
*
I
2
2
H
G
P
2|F
2:0
2
2
2
C
B
1
1
1
1
1
1
1
1
H
G
F
E
D
C
B
A
W
f
W i E W i E
Figure 4-1. Typical plot plan layout for recording ESP operating data.
4-6
-------
No. T/R SETS: 24
No. CHAMBERS: 3
No. FIELDS: 8
ELECTRICAL FIELDS: IB
DATE: 3/30/ai
No. SECTIONS OUT: 5
% PLATE OUT: 10
3|H
3?
3
G
F
•T
3
3
3
3
D
C
a
A
T
2:G
|
2-F
• !•
2eE
2:0
:
B
I
2sC
z
I
2:B
2 A
1SH
:
:
i
1SF
:
r
1-D
S
f
1 SC
i
£
'!•
£
1JA
W|E W|E W|E
DAILY T/R SET TRIP PATTERN
tio. T/R SETS: 24
No. CHAMBERS: 3
No. FIELDS: a
ELECTRICAL FIELDS: 48
DATE: 6/18/81
No. SECTIONS OUT: 17
% PLATE OUT: 35
3-E
3:0
3JC
s
3j*
2:E
2D
2IB
1
_j.
1cQ
1£F
1'
1:E
-•-
°
3
T
4,
W | E WjE W | E
DAILY T/R SET TRIP PATTERN
Figure 4-2. Comparison of T-R set trip patterns for two different days.
-------
•0
IU
o so
£
O
<
"
30
CO
i »«
BOILER R1 10/80
IS 20
DAYS
2S
Figure 4-3. Graphical display of plate area of service over a
30-day period.
4-8
-------
oc
cc
1200
1000
800
600
UJ
to
200
O NORTH
A CENTER
D SOUTH
FIELD NUMBER
Figure 4-4. Graphical plot of secondary current vs. field for a
3-chamber ESP.
4-9
-------
30
25
20
500
HQQ
300
200
100
I • • • • I ' ' ' ' I ' ' ' 'I
VOLTAGE = •
CURRENT «= •
Fl*ld 2A
10 15 20
DAYS
25
BROKEN
£
3031
Figure 4-5a. Example of graphical displays of secondary current
and voltage vs. day of operation.
4-10
-------
no
35
» -JQ
25
20
MA
800
35
>JO
25
20!
mA
800
600
400
200
FIELD 1A
FIELD 2A
FIELD 3A
FIELD 4 A
10
15
20
25
KV
35 '
30
25
1000
800
600 *
200
KV
35
30
25
1000
800
600 <
400
200
Figure 4-5b. Example of graphical displays of secondary current
and voltage vs. day of operation.
4-11
-------
to see if the desired patterns exist. This presents a real problem on larger
ESP's with many fields, but a relatively minor one on ESP's that have only
three or four fields. Another advantage of this graphical method is that it
permits the plotting of other parameters such as gas temperature, 0? content,
process load, opacity, and important feed characteristics. This capacity may
provide correlating data to help diagnose the source of any problems that
occur. For example, consider an ESP on a utility boiler that normally burned
coal with a 0.8 and 1.0 percent sulfur content. When a coal with a lower
sulfur content was introduced because normal coal supplies were interrupted,
the effects of the change on ESP performance were visible within 24 hours
when plotted on a time chart. Although the use of this coal was terminated
after two weeks, it took several months of operation before the ESP returned
to its normal performance level. Although the classic symptoms of high
resistivity due to the lower sulfur content were identifiable from the meter
readings, the severity of the effects of this process change on the ESP was
much more visible in the trend graphs than in the tabular data gathered
during each shift.
Other data that should be collected from the T-R cabinets include spark
rate, evaluation of abnormal or severe sparking conditions, controller status
(auto or manual), and identification of bus sections out of service. This
additional information helps in the evaluation of ESP performance.
Air Load/Gas Load V-I Curves--
In addition to the routine panel meter readings, other electrical tests
of interest to personnel responsible for evaluating and maintaining ESP's
include the air load and gas load V-I (voltage-current) tests, which may be
conducted on virtually all ESP's. Air load tests are generally conducted or,
cool, inoperative ESP's through which no gas is flowing. This test should be
conducted when the ESP is new, after the first shutdown, and everytime off-
line maintenance is performed on the ESP. These airload V-! curves serve as
the basis for comparison in the evaluation of ESP maintenance and perform-
ance. A typical air load curve is shown in figure 4-6.
Generating a V-I curve, a simple procedure, can be done with either
primary or secondary meters. A deenergized T-R set on manual control is
energized (but with zero voltage and current), and the power to the T-R set
SECTION 4-PERFORMANCE EVALUATION. PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-12
-------
1200
< 1000
OC
OC
Z>
O
>-
O
o
UJ
800
600
400
200
I
FIELD 2
ASPARKQVER
20 30 «40 50 60
SECONDARY VOLTAGE, kV
Figure 4-6. Typical air-load test V-I curve for an ESP on a recovery boiler
with normal dust layer.
4-13
-------
is increased manually. At corona initiation the meters should move suddenly
and the voltage and near zero current level should be recorded. (It is
sometimes difficult to identify this point precisely, so the lowest practical
value should be recorded.) After corona initiation is achieved, the power
should be increased at predetermined increments, say, every 50 or 100 mini-
amps of secondary current or every 10 volts of AC primary voltage (the incre-
ment is discretionary), and the values recorded. This procedure should be
continued until either sparking occurs, the current limit is achieved, or the
voltage limit is achieved. This procedure is then applied to each T-R. One
difficulty that sometimes arises is activation of the undervoltage trip
circuit in the control cabinet. Either increasing the time for response or
decreasing the activation voltage will prevent the T-R from tripping out
during the test. This problem is worse with some T-R cabinet designs than
with others.
When the air-load tests have been completed for each field, the voltage/
current curves are plotted. When ESP's are equipped with identical fields
throughout, the curves for each field should be nearly identical. The curves
should be similar for ESP's with varying field dimensions or T-R sizes. In
most cases, the curves also should be similar to those generated when the
unit was new, but shifted slightly to the right because of dust on the wires
and plates. These curves should become part of the permanent record on the
ESP. This effect is demonstrated in Figures 4-7 and 4-8.
The use of the air-load curves enables plant personnel to identify which
field(s) may be experiencing difficulty. Comparison with an air-load test
run just before a unit is serviced will confirm whether the maintenance work
corrected the problem(s). A check should be made to be sure all tools, soda
cans, rags, raincoats, and magazines are removed from the ESP prior to its
startup.
The advantage of air-load tests is that because they are performed under
near identical conditions each time, curves can be compared. One of the dis-
advantages is that the internal conditions are not always the same as during
normal operation. For example, misalignment may appear or disappear when the
ESP is cooled (expansion/contraction), and dust buildup may be removed by
rapping during ESP shutdown.
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS. AND PROBLEM SOLUTIONS
4-14
-------
0,10
3 0-05
o
ec
O
u
0
10
RELATIVELY
CLEAN PLATES
AFTER PERIOD OF
INTENSIVE RAPPING
VERY DIRTY
PLATES
20 30
VOLTAGE, kV
Figure 4-7, Variation of voltage current characteristics with collecting
plate contamination.
4-15
-------
I
z
LU
o
LU
cc
sc
u
70
60
50
30
20
10
SECOND FIELD / /
CLEAN ELECTRODE/ /
, SECOND FIELD
WITH 1 on LAYER
10 20 30 HQ 50 60
APPLIED VOLTAGE, kilovolts
Figure 4-8. Effect of dust layer thickness on V-I curve.
4-16
-------
The gas load V-I curve, on the other hand, is generated during the
normal operation of the process while the ESP is energized. The procedure
for generating the V-I curve is the same except that gas-load V-I curves are
always generated from the outlet fields first and move toward the inlet.
This prevents the upstream flow that is being checked from disturbing the V-I
curve of the downstream field readings. Although such disturbances would be
short-lived (usually 2 minutes, but sometimes up to 20 minutes), working from
outlet to inlet also speeds up the process.
The curves generated under gas load will be similar to air-load curves.
They will generally be shifted to the left under gas load conditions, howev-
er, and the shape of the curve will be different for each field depending on
the presence of particulate in the gas stream (see Figure 4-9).
The pattern in the V-I curves under gas load conditions is similar to
what is shown in Figure 4-9. As shown, the gas-load curve is to the left of
the air-load curve. Both curves shift to the left from inlet to outlet
(characteristic of most ESP's operating under moderate resistivity). The end
point of each curve is the sparking voltage/current level, or maximum attain-
able by the T-R. These points represent the characteristic rise in current
from inlet to outlet that is normally seen on the ESP panel meters. Problems
characterized by the air load curves will normally also be reflected in the
gas-load curve, but some problems may show up in one set of curves and not in
the other (e.g., high resistivity as shown in Figure 4-10, some misalignment
problems).
Another item of importance is that gas-load curves vary from day to day,
even minute by minute. Curve positions may change as dust builds up and is
then removed from the plates; as gas flow, particulate chemistry loading, and
temperature change; and as resistivity changes (for examples, see Figures
4-11 and 4-12). Nonetheless, they still should maintain a characteristic
pattern. Gas-load curves are normally used to isolate the cause of a suspected
problem rather than on a day-to-day basis; however, they can be used daily if
necessary. Several facilities equipped with analog T-R controllers manually
set the voltage/current limits every shift because the controllers find it
difficult to recognize the back corona conditions of their high resistivity
dust. By establishing where back corona begins, plant personnel are able to
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS. AND PROBLEM SOLUTIONS
4-17
-------
Figure 4-9. Comparison of typical air load and gas load V-I curves,
AIR-LOAD
Figure 4-10.
Comparison of V-I curves for high resistivity at air load
and gas load conditions.
4-18
-------
.o
a
E
ro
O
c
CO
c
i
Z
LU
O
I—
z
LU
DC
QC
HO.U
44.0
40,0
36,0
32,0
28,0
24.0
20,0
16,0
12,0
8,0
4,0
n.n
i i i i i i i i i i
- "INLET ELECTRICAL FIELD J
* MIDDLE ELECTRICAL FIELD
" » OUTLET ELECTRICAL FIELD . i «"
_
* •
• * • -
_ * _
• * •
V ™
A
mm ™ NHM
A
* • ~
_ • • _
*
9* ^
. 1. 11 »• .*i * i 1 1 J 1
8 12 16 20 24 28 32 36 40
VOLTAGE, kV
Figure 4-11. V-I curves demonstrating particulate space charge effect in a cold
side precipitator collecting fly ashj
4-19
-------
40
35
30
«:
c 25
w
1 20
IXI
CC
§ 15
o
ID
TR NO. 1
TR NO. 3
Tft NO. 5
TR NO, 7
* "
A _
Mi » V
%l
1
16
20
2^ 28
VOLTAGE, kV
32
36
Figure 4-12.
Typical V-I curves for a cold side ESP operating at
moderate ash resistivity. 1
4-20
-------
obtain the maximum voltage and current possible without wasting power or
degrading ESP performance.
Other possible data that could aid in the evaluation of trends and long-
term performance include a plot of wire failures within the ESP and their
frequency, frequency of hopper pluggage, and a plot of the percent of the ESP
deenergized on a daily basis. This last item can be used in combination with
opacity and electrical data to define when maintenance work is needed and
whether it is addressing the problems encountered, and to aid in the schedul-
ing of routine and preventive maintenance.
It is evident that obtaining good operating data and maintaining good
records help in the maintenance of ESP performance by providing a historical
data base that can be used to evaluate daily operating performance. Record-
keeping alone, however, will not guarantee satisfactory long-term perform-
ance. Analysis of the data and an understanding of the fundamental design
features and limitations and the operating characteristics, of the ESP are
necessary to correct minor problems before they become major.
4.2 PROBLEM DIAGNOSIS
Many ESP operating problems are reflected in the electrical operating
characteristics. In a typical, well-designed, operated, and maintained ESP
(without resistivity problems), a pattern of increasing current and decreased
sparking from the inlet to the outlet fields would be expected. The operat-
ing voltage may be somewhat low because of sparking at the inlet and increase
in the second and third fields. The voltage may begin to drop as the ges
approaches the outlet because the gas is relatively clean. Although this is
not a problem in every ESP, the tendency in most ESP's is for the current to
increase from inlet to outlet. It is important to be familiar with the oper-
ating characteristics of the ESP and to know what is typical. The record-
keeping discussed in Section 3 helps one to become more knowledgeable.
One of the difficulties in assessing ESP performance is that many dif-
ferent problems produce the same electrical characteristics on the panel
meters. For this reason, plant personnel obtain additional data to reduce
the number of possible causes to one or two. In addition, synergism often
causes the original problem or failure to lead to additional problems that
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-21
-------
can cascade into even more problems. When this occurs, it is difficult to
identify the original cause of a problem. Nevertheless, it is usually impor-
tant to identify and correct all causal factors rather than to treat the
symptom. Again, the key to diagnostic troubleshooting is to know the precip-
itator's characteristics; to understand what the meter readings mean; and to
use all the process, opacity, and electrical data to assist in the evalua-
tion. An internal inspection may even be necessary to confirm or eliminate
possible sources of problems (Section 6.0).
Most major performance problems can be categorized into the following
areas: resistivity, hopper pluggage, air Inleakage, dust buildup, wire
breakage, rapper failure, inadequate power supplies and/or plate area, chang-
es in particle size, and misalignment of ESP components. Some of these
problems are related to design limitations, operational changes, maintenance
procedures, or a combination thereof. The identification of these problems
and their effect on ESP performance are discussed here; possible corrective
measures are discussed later (in Section 4,3).
4.2.1 Problems Related to Resistivity
The concept of resistivity and its effect on ESP design were introduced
earlier. Briefly, the resistivity of the dust on the collection plate af-
fects the acceptable current density through the dust layer, the ability to
remove the dust from the plates, and indirectly, the corona charging process.
Much attention has been given to high resistivity conditions in utility fly
ash applications. Because the optimum resistivity range for ESP operation is
relatively narrow, however, both high and low resistivity cause problems.
When a unit is designed with modest plate area, sectionalization, and power-
input capabilities, poor ESP performance can result from excursions outside
the optimum resistivity range. At sources where resistivity changes are
intermittent, modification of operating procedures may improve performance
temporarily. At sources where the dust remains outside the design resistivi-
ty characteristics, however, expensive retrofitting or modification may be
required.
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-22
-------
High Resistivity—
The most common resistivity problem is that caused by high dust resis-
tivity. Because of their inability to release or transfer electrical charge,
the particles acquire charge from the corona charging process and migrate to
the collection plate. Once at the collection plate, the particles neither
give up very much of their acquired charge nor easily pass the corona current
to the grounded collection plates. As the dust layer buildup continues, the
resistance to current flow increases, and the controller responds by "opening
up" the SCR more to increase the voltage level. This is demonstrated in the
V-I curve presented in Figure 4-13. Although this would occur with almost
all particulates, the detrimental effect on ESP performance is more pro-
nounced when particle resistivity is high.
The voltage drop across the dust layer may be substantial. The dust
layer voltage drop (which depends on the resistivity and thickness of the
dust layer) can be approximated by Ohm's law. As the current increases, the
voltage drop also increases. Under high resistivity conditions, however,
very high voltage drops may occur at low current density*, when the voltage
drop exceeds 15 to 20 kV/cm, the dust layer will break down electrically. As
the resistivity climbs, the current level at which this breakdown occurs
decreases. It is this breakdown of the dust layer that diminishes the ESP
performance.
n -i A
The optimum resistivity range is generally between 10 and 10 ohm-cm.
Performance of the ESP generally does not diminish until approximately 2 x
10 ohm-cm. High-temperature gas streams at the 600° to 700°F level could
possibly exhibit resistivity problems at 10 ohm-cm because of low gas
density. Conversely, high altitudes will also tend to reduce the resistivity
level at which problems may occur. At this level the breakdown of the dust
layer may be limited, but it can be aggravated by unequal buildup on the
plates, and the response of the controller to increase the operating voltage
may exceed the voltage required for spark propagation. Thus, when the dust
layer does break down, the resistance to current flow is suddenly reduced and
a spark is formed. An identifying characteristic of high resistivity is the
tendency toward high spark rates at low current levels throughout the ESP,
SECTION 4-PERFORMANC1 EVALUATION, PROBLEM DIAGNOSIS. AND PROBLfM SOLUTIONS
4-23
-------
in
a
E
ro
DC
DC
D
O
DC
Q
Z
o
o
uu
V3
0,3
0.2
0,1
0,0
WITHOUT DUST LAYER
WITH DUST LAYER
0
1
10 20 30
APPLIED VOLTAGE, kV
50
Figure 4-13. V-I characteristics of inlet section of ESP
collecting high resistivity ash.
4-24
-------
which often makes it difficult for the T-R controller to respond and function
adequately,
13
As the resistivity climbs well into the 10 ohm-cm range, the sparking
may be sharply reduced or become nonexistent because the dust layer voltage
exceeds the breakdown threshold at such low current density that insufficient
voltage is applied to the wire to propagate a spark across the interelectrode
space. Also, the breakdown of the dust layer may be widespread across the
plate. This condition, known as "back-corona," is characterized by a dis-
charge of positive ions from the plates that may reduce the charging of
particles and even reduce the voltage below the level of negative corona
initiation. Although back-corona is more pronounced and well developed at
higher resistivities, it also occurs during sparking. In the latter case it
is somewhat limited to small "craters" that, form on the plate opposite the
wires.
Severe sparking can cause excessive charging off-time, spark "blasting"
of particulate of the plate, broken wires due to electrical erosion, and
reduced average current levels. It is the reduced current levels that gener-
ally lead to deteriorated performance. Because the current level is indica-
tive of the charging process, the low current and voltage levels that occur
inside an ESP operating with high resistivity dust generally reflect slower
charging rates and smaller voltage forces applied to the dust to force migra-
tion to the plate. The effect is an undersized ESP; if high resistivity is
expected to continue, the design could be modified to accommodate this prob-
lem and thereby improve performance.
High resistivity also tends to promote rapping problems, as the electri-
cal properties of the dust tend to make it very tenacious. High voltage drop
through the dust layer and the retention of electrical charge by the parti-
cles make the dust difficult to remove because of its strong attraction to
the plate. In addition to the reduced migration and collection rate associ-
ated with high resistivity dust, greater rapping forces usually required to
dislodge the dust may also aggravate or cause a rapping reentrainment prob-
lem. Important items to remember are ]) difficulty in removing the high-
resistivity dust is related to the electrical characteristics, not to the
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-25
-------
sticky or cohesive nature of the dust; and 2) the ESP must be able to with-
stand the necessary increased rapping forces without sustaining damage to
insulators or plate support systems. Figure 4-14 shows an example V-I curve
for an ESP field with insulator tracking (i.e., current leakage) problems.
Low Resistivity--
Low dust resistivity can be just as detrimental to the performance of an
ESP as high resistivity. Low resistivity refers to the inability of parti-
cles to retain a charge once they have been collected on the plate. As in
the normal- and high-resistivity cases, the ability of the participate matter
to obtain a charge is not affected by Its resistivity; particle charging
occurs by the previously discussed charging mechanisms, which are dependent
on particle size. Once the particles are at the collection plate, however,
they release much of their acquired charge and are capable of passing the
corona current quite easily. Thus, attractive and repulsive electrical
forces that are normally at work at higher resistivities are lacking, and the
binding forces ("holding power") to the plate are considerably lessened.
Particle reentrainment is a substantial problem at low resistivity, and ESP
performance appears to be very sensitive to contributors of reentrainment,
such as poor rapping or poor gas distribution.
The voltage drop across the dust layer on the plate is usually small.
The lower resistance to current flow than in the optimum- and high-resistivi-
ty ranges means lower operating voltages are required to obtain substantial
current flow. Thus, operating voltages and currents are typically close to
clean plate conditions, even when there is some dust accumulation on the
plate. A typical low-resistivity condition, then, is characterized by low
operating voltages and high current flow, which would be reflected in the T-R
panel meter readings. These electrical conditions may look very similar to
those of high resistivity with well-developed back-corona. In any case, the
result is usually the same—reduced ESP performance.
Despite the large flow of current under the low-resistivity conditions,
the corresponding low voltages yield lower migration velocities to the plate.
Thus, particles of a given size take longer to reach the plate than would be
expected. When combined with substantial reentrainment, the result is poor
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
-------
5 10 15 20 25 30 35 40 U5 50
SECONDARY VOLTAGE, kV
Figure 4-14, Air load V-I curve for ESP field with insulator tracking,
4-27
-------
ESP performance. In this case, the large flow of power to the ESP represents
a waste of power.
The low-resistivity problem typically results from the chemical charac-
teristics of the participate and not from temperature. The participate may
be enriched with compounds that are inherently low in resistivity, either
because of poor operation of the process or the inherent nature of the process,
Examples of such enrichment include excessive carbon levels in fly ash (due
to poor combustion), the presence of naturally occurring alkalies in wood
ash, iron oxide in steel-making operations, or the presence of other low-
resistivity materials in the dust. Over-conditioning may also occur in some
process operations, such as the burning of high-sulfur coals or the presence
of high SO, levels in the gas stream, which lower the inherent resistivity of
9
the dust. In some instances, large ESP's with SCA's greater than 750 ft /1000
acfm have performed poorly because of the failure to comprehend fully the
difficulty involved in collecting a low-resistivity dust. Although some
corrective actions are available, they are sometimes more difficult to imple-
ment than those for high resistivity. Fortunately, the low-resistivity
problem is not as common as the high-resistivity problem.
4.2.2 Excessive Dust Accumulations on Electrodes
Where no ash resistivity problem exists, the cause for excessive dust
accumulation in an ESP is often external. When buildup of material on the
discharge electrodes and collecting electrodes or plates is difficult.to
distinguish in an operating ESP, differences in the V-I curves can often
point up the nature of the problem (see Figure 4-15).
Buildup of material on the discharge electrodes (whether straight-wired,
barbed-wired, or rigid) often means an increase in voltage to maintain a giv-
en operating current. The effect of dust buildup on the discharge electrodes
is usually equivalent to changing the effective wire size diameter, and since
the corona starting voltage is strongly a function of wire diameter, the
corona starting voltage tends to increase and the whole V-I curve tends to
shift to the right. Sparking tends to occur"at about the same voltage unless
resistivity is high. This effect on corona starting voltage is usually more
pronounced when straight wires are uniformly coated with a heavy dust, and
less pronounced on barbed wires and rigid electrodes or when the dust layer
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-28
-------
<
E
LLJ
GC
CC
D
O
cc
O
u
ui
W
1200
1000
800
600
200
NORMAL OR
BASELINE CURVE]
FIELD WITH
DISCHARGE
ELECTRODE
BUILDUP —
20 30 10
SECONDARY VOLTAGE,
Figure 4-15. V-I curve for a field with excessive wire buildup.
4-29
-------
is not uniform. Barbed wires and rigid electrodes tend to keep the "points"
relatively clean and to maintain a small effective wire diameter and, there-
fore, a low corona starting voltage. Nevertheless, a higher voltage would
still be required to overspread the wire with the corona discharge where the
wire buildup had occurred. Thus, buildup on the discharge electrodes would
still be characterized by a higher voltage to maintain a given current level.
Under normal operating conditions, most of the dust would be collected
at the plate and relatively little would collect on the wires. The dust that
collects on the wires is usually the dust that enters the corona discharge
area with the proper trajectory to attach to the wire. The material collect-
ed on the plate is usually allowed to build up for some specified length of
time to take advantage of certain cohesive forces between particles and then
dislodged by activation of a rapper. This dust buildup usually changes the
electrical characteristics of the field and causes a shift in voltage and
current over the period of the buildup. This variation in the amount of dust
on the plates is one reason why readings of panel meters may vary from obser-
vation to observation. In general, however, dust buildup on a clean plate
(as on electrode wires) increases the voltage to maintain a given current
level. This effect is normally most apparent on a middle or outlet field of
an ESP, where the time between rapping periods is sufficiently long to allow
a substantial dust layer to build up. Electrical readings taken just before
and just after a rapping cycle should indicate decreased operating voltage
(as reflected by the primary voltmeters) and a decreased or constant current
level.. The dust layer presents a resistance to the current flow, and the
operating voltage must be increased to overcome this resistance. The dust
layer has relatively little effect on corona starting voltage.
If the dust layer buildup were relatively even, it might be expected to
continue up to the T-R set capacity. In practice, however, dust buildup
usually reaches a thickness of between 3/4 and 1 in. under normal resistivity
conditions before performance is markedly reduced. As the thickness increas-
es and operating voltages increase, however, the clearance between the dis-
charge electrode and the surface of the dust layer diminishes, which encour-
ages a sparking condition within the precipitator. The precipitator controls
then respond by decreasing operating voltage and current, which lowers the
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS. AND PROBLEM SOLUTIONS
4-30
-------
charging and migration rates. This in turn causes the volume occupied by the
ash within the precipitator to Increase, and the gas velocity between the
plates must also increase to maintain a given flow rate. If the increased
velocity exceeds 6 to 8 ft/s, reentraimnent is likely to occur and reduce
performance further. In some cases the dust layer will become self-limiting
because of the gas velocity through the ESP.
The usual cause for buildup on the collection plates or discharge wires
is failure of the rapping system or an inadequate rapping system. The rapping
system must provide sufficient force to dislodge the dust without damaging
the ESP or causing excessive reentrainment. The failure of one or two iso-
lated rappers does not usually degrade ESP performance significantly. The
failure of an entire rapper control system or all the rappers in one field,
however, can cause a noticeable decrease in ESP performance, particularly
with high-resistivity dust. Therefore, rapper operation should be checked at
least once per day, or perhaps even once per shift. A convenient time to
make this check is during routine T-R set readings.
Rapper operation may be difficult to check on some ESP's because the
time periods between rapper activation can range from 1 to 8 hours on the
outlet field. One method of checking rapper operation involves installing a
maintenance-check cycle that allows a check of all rappers in 2 to 5 minutes
by following a simple rapping pattern. The cycle would be activated by plant
personnel, who would interrupt the normal rapping cycle and note any rappers
that fail to operate. After the cycle, the rappers would resume their normal
operation. Maintenance of rapper operation is important to optimum ESP
performance. (Note: In rare cases, rapping is not necessary. These usually
involve low-resistivity dust that requires very little energy to remove or
actually dislodges from its own weight. Also, no rappers are associated with
the collection surface in wet ESP's, but they may be used for the discharge
electrodes.)
Excessive dust buildup also may result from sticky dusts or dewpoint
conditions. In some cases, the dusts may be removed by increasing the tem-
perature, but in many cases the ESP must be entered and washed out. If
sticky particulates are expected (such as tars and asphalts), a wet-wall ESP
is usually appropriate because problems can occur when large quantities of
SECTION 4-PEflFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-31
-------
sticky particles enter a dry ESP. Among the cases where this may be a prob-
lem are ESP's applied to wood-fired boilers, municipal incinerators, and some
coal-fired boilers. The problem usually occurs when improper combustion
yields a partially combusted, sticky, hydrocarbon material. This can also
present a fire hazard and potential low-resistivity problems. Some utilities
have also experienced the problem when an ESP was energized while the oil
guns were still in use during startup or when stable operation had not yet
been achieved.
Sticky particulate can also become a problem when the temperature falls
below dewpoint conditions. Although acid dewpoint is usually of greater con-
cern in most applications, moisture dewpoint is important. When dewpoint
conditions are reached, liquid droplets tend to form that can bind the par-
ticulate to the plate and wires (and also accelerate corrosion). Carryover
of water droplets or excessive moisture can also cause this problem (e.g.,
improper atomization of water in spray cooling of the gas or failure of a
waterwall or economizer tube in a boiler). In some instances the dust layer
that has built up can be removed by increasing the intensity and frequency of
the rapping while raising the temperature to "dry out" the dust layer. In
most cases, however, it is necessary to shut the unit down and wash out or
chisel out the buildup to clean the plates. Localized problems can occur
where inleakage causes localized decreases in gas temperature.
In the pulp and paper industry, sticky particulate has been noted during
periods of high excess air levels in recovery boilers burning black liquor.
Concentrations of SO, tend to increase. These result in a raised acid dew-
point, since the SO,, is absorbed on the particulate at the relatively low
temperatures in the economizer and ESP. This sticky salt cake can be diffi-
cult to remove from both the economizers and ESP's. The situation can be
further aggravated by the combustion of residual fuel oil containing vanadi-
um. The vanadium oxides with SCL in the gas stream, and excess 0, tends to
convert greater quantities of S0? to SO- than would naturally occur. The
sodium/vanadium salt complexes formed can also make it more difficult to
remove the salt cake. Some of the vanadium complexes may also be insoluble
in water and be difficult to wash off the plates during an ESP washdown.
SECTION 4-PERFORMANCE EVALUATION. PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-32
-------
4.2.3 HireBreakage
Some ESP's operate for 10 to 15 years without experiencing a single wire
breakage, whereas others experience severe problems causing one or more
sections to be out of service nearly every day of operation. Much time and
effort have been expended to determine the causes of wire breakage. One of
the advantages of a rigid-frame or rigid-electrode ESP is that this type uses
shorter wires or no wires at all. Although most of the new ESP's are of the
rigid-frame and rigid-electrode type (and some weighted-wire systems have
also been retrofitted to rigid electrode)» the most common ESP in service
today is still the weighted-wire; therefore, the nature, severity, and loca-
tions of wire failures cannot be overlooked.
Wires usually fail in one of three areas: at the top of the wire, at
the bottom of the wire, and wherever misalignment or slack wires reduce the
clearance between the wire and plate. Wire failure may be due to electrical
erosion, mechanical erosion, corrosion, or some combination of these. Nhen
wire failures occur, they usually short-out the field where they are located,
and in some cases, may short-out an adjacent field. Thus, the failure of one
wire can cause the loss of collection efficiencies in an entire field or bus
section. In some smaller ESP applications, this can represent one-third to
one-half of the charging/collecting area and thus substantially limit ESP
performance. One of the advantages of higher sectionalization is that wire
failure affects smaller areas so ESP performance does not suffer as much.
Some ESP's are designed to meet emission standards with some percentage of
the ESP deenergized, whereas others may not have any margin to cover down-
time. Inlet fields are usually more important to ESP operation than outlet
fields.
Sparking usually occurs at points where there is close clearance within
a field (due to a warped plate, misaligned guidance frames, or bowed wires).
The maximum operating voltage is usually limited by these close tolerance
areas because the spark-over voltage is lowered by the reduction in the
distance between the wire and the plate. Under normal circumstances random
sparking does little damage to the ESP. During sparking, most of the power
supplied to energize the field is directed to the location of the spark, and
the voltage field around the remaining wires collapses. The considerable
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS. AND PROBLEM SOLUTIONS
4-33
-------
quantity of energy available during the spark is usually sufficient to vapor-
ize a small quantity of metal. When sparking continues to occur at the same
location, the wire usually "necks down" because of electrical erosion until
it is unable to withstand the tension and it breaks. Misalignment of the
discharge electrodes relative to the plates increases the potential for
broken wires, decreases the operating voltage and current because of spark-
ing, and decreases the performance potential of that field in the ESP.
Although the breakage of wires at the top and bottom where the wire
passes through the field can be aggravated by misalignment, the distortion of
the electrical field at the edges of the plate tends to be the cause of
breakage. This distortion of the field, which occurs where the wire passes
the end of the plate, tends to promote sparking and gradual electrical ero-
sion of the wires. The methods available to minimize this particular failure
are discussed in Section 4.3.
Both design considerations and the failure to maintain alignment gener-
ally contribute to mechanical erosion (or wear) of the wire. In some de-
signs, the lower guide frame guides the wires or their weight hooks (not the
weights themselves) into alignment with the plates. When alignment is good,
the guide frame or grid allows the wires or weight hooks to float freely
within their respective openings. When the position of the wire guide frame
shifts, however, the wire or weight hook rubs the wire frame within the
particulate-laden gas stream. Failures of this type usually result from a
combination of mechanical and electrical erosion. Corrosion may also con-
tribute to this failure. Microsparking action between the guide frame and
the wire or weight hook apparently causes the electrical erosion. The same
type of failure also can occur in some rigid frame designs where the wires
ride in the frame.
Another failure that sometimes occurs involves crossed wires. This
happens when those who replace wires do not check to see that the replacement
wire does not cross another wire. Eventually, the resulting wearing action
breaks one or both wires. If one of the wires does survive, it is usually
worn down enough to promote greater sparking at the point of contact until it
finally does break. When wires are replaced, care should be taken to see
that wires are not crossed. Any wires that are found to be exceptionally
SECTION 4-PERFORMANCE EVALUATION. PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-34
-------
long and slack should be replaced; they should not be crossed with another
wire to achieve the desired length.
Corrosion of the wires can also lead to wire failures. Corrosion, an
electrochemical reaction, can occur for several reasons, the most common
being acid dewpoint. When the rate of corrosion is slow and generally spread
throughout the ESP, it may not lead to a single wire failure for 5 to 10
years. When the rate of corrosion is high because of long periods below the
acid dewpoint, failures are frequent. In these cases the corrosion problem
is more likely to be a localized one; e.g., in places where cooling of the
gas stream occurs, such as inleakage points and the walls of the ESP. Corro-
sion-related wire failures can also be aggravated by startup/shutdown proce-
dures that allow the gas streams to pass through the dewpoint many times.
Many facilities have experienced wire breakage problems during the initial
process shakedown period when the process operation may not be continuous.
Once steady operation has been achieved, wire breakage problems tend to
diminish at most plants. Some applications that routinely start up and shut
down (small "peaking" utilities, for example) have had relatively few prob-
lems with wire breakage. Good operating practices and startup/shutdown
procedures help to minimize this problem.
Another cause of wire failure is wire crimping. These crimps usually
occur at the top and bottom of the wires where they attach to the upper wire
frame or bottle weight', however, a crimp may occur at any point along the
wire. A crimp can mechanically weaken and thin a wire, it can cause a dis-
tortion of the electric field along the wire and promote sparking, and it can
subject the wire to a stress corrosion failure (materials under stress tend
to corrode more rapidly than those not under stress). Because a crimp cre-
ates a residual stress point, all three mechanisms may be at work in this
situation.
Wire failure should not be a severe maintenance problem or operating
limitation in a well-designed ESP. Excessive wire failures are usually a
symptom of a more fundamental problem. Plant personnel should maintain
records of wire failure locations. Although ESP performance will generally
not suffer with up to approximately 10 percent of the wires removed, these
records should be maintained to help avoid a condition in which entire gas
SECTION 4-PERFORMANGE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
-------
lanes may be deenergized. Improved sectionalization helps to minimize the
effect of a broken wire on ESP performance, but performance usually begins to
suffer when large percentages of the ESP are deenergized.
4.2.4 Hopper Pluggage
Perhaps no other problem (except fire or explosion) has the potential
for degrading ESP performance as much as hopper pluggage. Hopper pluggage
can permanently damage an ESP and severely affect both short-term and long-
term performance. Hopper pluggage is difficult to diagnose because its
effect is not immediately apparent on the T-R panel meters. Depending on its
location, a hopper can usually be filled in 4 to 24 hours. In many cases,
the effect of pluggage does not show up on the electrical readings until the
hopper is nearly full.
The electrical reaction to most plugged hoppers is the same as that for
internal misalignment, a loose wire in the ESP, or excessive dust buildup on
the plates. Typical symptoms include heavy or "bursty" sparking in the
field(s) over the plugged hopper and reduced voltage and current in response
to the reduced clearance and higher spark rate. In weighted-wire designs,
the dust may raise the weight and cause slack wires and increased arcing
within the ESP. In many cases, this will trip the T-R off-line because of
overcurrent or undervoltage protection circuits. In some situations, the
sparking continues even as the dust builds between the plate and the wire;
whereas in others, the voltage- continues to decrease as the current increases
and little or no sparking occurs. This drain of power away from corona gen-
eration renders the field performance virtually useless. The flow of current
also can cause the formation of a dust clinker resulting from the heating of
the dust between the wire and plate.
The buildup of dust under and into the collection area can cause the
plate or discharge electrode guide frames to shift. The buildup can also
place these frames under enough pressure to distort them or to cause perma-
nent warping of the collection plate(s). If this happens, performance of the
affected field remains diminished by misalignment, even after the hopper is
cleared.
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-36
-------
The causes of hopper pluggage include such things as obstructions due to
fallen wires and/or bottle weights, inadequately sized solids-removal equip-
ment, use of hoppers for dust storage, inadequate insulation and hopper
heating, and inleakage through access doors. Most dusts flow best when they
are hot, and cooling the dusts also can promote a hopper pluggage problem.
Hopper pluggage can begin and perpetuate a cycle of failure in the ESP.
For example, in one ESP, a severely plugged hopper misaligned both the plates
and the wire guide grid. When the hopper was cleared, the performance of
this field had decreased and the wires and weight hooks were rubbing the
lower guide and causing erosion of the metal. When the metal eventually wore
through, hopper pluggage increased as weights (and sometimes wires) fell into
the hopper, plugging the throat, and allowed the hopper to fill again and
cause more misalignment. The rate of failure continued to increase until it
was almost an everyday occurrence. This problem, which has occurred more
than once in different applications, points out how one relatively simple
problem can lead to more complicated and costly problems.
In most pyramid-shaped hoppers, the rate of buildup lessens as the
hopper is filled (because of the geometry of the inverted pyramid). Hopper
level indicators or alarms should provide some margin of safety so that plant
personnel can respond before the hopper is filled. The rate of deposition in
the hopper also will diminish when the top of the dust layer interferes with
the electrical characteristics of the field, which reduces the collection
efficiency. Lastly, reentrainment of the dust from the hopper can also limit
how far up into the field the dust can go. Although buildups as deep as 4 ft
have been observed, they usually are limited to 12 to 18 in. up from the
bottom of the plates.
4.2.5 Misalignment
As mentioned several times in the previous sections, misalignment is
both a contributor to and a result of component failures. In general, most
ESP's are not affected by a misalignment of less than about 3/16 1n. Indeed,
some tolerance must be provided for expansion and contraction of the compo-
nents. Beyond this limit, however, misalignment can become a limiting factor
in ESP performance and is usually visually evident during an internal inspec-
tion of the ESP. Whether caused by warped plates, misaligned or skewed
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS. AND PROBLEM SOLUTIONS
4-37
-------
discharge guide frames, insulator failure, or failure to maintain ESP "box-
squareness," misalignment reduces the operating voltage and current required
for sparking. The V-I curve would indicate a somewhat lower voltage to
achieve a low current level with the sparking voltage and current greatly
reduced (see Figure 4-16). Since the maximum operating voltage/current
levels are dependent on the path of least resistance in a field, any point of
close tolerance will control these levels.
4.2.6 Unusually Fine Particle Size
Unusually fine particles present a problem if 1) the ESP was not de-
signed to handle them, or 2) a process change or modification shifts the
particle size distribution into the range where ESP performance is poorest.
A shift in particle size distribution tends to alter electrical characteris-
tics and increase the number of particles emitted in the light-scattering
size ranges (opacity).
As was discussed in Section 2, there are two basic charging mechanisms:
field charging and diffusion charging. Although field charging tends to dom-
inate in the ESP and acts on particles greater than 1 micrometer in diameter,
it cannot charge and capture smaller particles. Diffusion charging, on the
other hand, works well for particles smaller than 0.1 micrometer in diameter.
On particles between 0.1 and 1.0 micrometer in diameter, and particularly in
the range of 0.2 to 0.5 micrometer, performance of the ESP diminishes consid-
erably. Because neither charging mechanism is very effective, particles in
this range are more difficult to charge; and once charged, they are easily
bumped around by the gas stream, which makes them difficult to collect. The
collection efficiency of an ESP can drop from as high as 99.9+ percent on
particles sized above 1.0 micrometer and below 0.1 micrometer, to only 85 to
90 percent on particles in the 0.2- to 0.5-micrometer diameter range, depend-
ing upon the type of source being controlled. If a significant quantity of
particles fall into this range, the ESP design must be altered to accommodate
the fine particles.
Two significant electrical effects of fine particles are space charge
and corona quenching, which occur when heavy loadings of fine particles enter
the ESP. At moderate resistivities, the space-charge effects normally occur
in the inlet or perhaps the second field of ESP's. Because it takes time to
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-38
-------
500
oc
ir
u
300
200
o
O
I 100
REFERENCE OR
BASELINE CURVE,
FIELD WITH
ALIGNMENT PROBLEMS
10 20 30 40
SECONDARY VOLTAGE, kV
50
60
Figure 4-16, Air load V-I curve pattern generated by alignment problems,
4-39
-------
charge the particles and then to force them to migrate to the plate, a cloud
of negatively charged particles forms in the gas stream. This cloud inter-
feres with corona generation process and impedes the flow of ions from the
wire to the gas stream. The T-R controller responds by increasing the oper-
ating voltage to maintain current flow and corona generation. The increase
in voltage usually causes increased spark rates, which may in turn reduce the
voltage and current to maintain a reasonable spark rate. As the particles
move through the ESP and are collected by the plates, the gas stream becomes
cleaner. As a. result, the voltage level will usually decrease but current
levels will increase markedly. As quantities of fine particles are in-
creased, the space charging effect may progress further into the ESP.
Corona quenching can also result when the quantity of particles is so
great that relatively few electrons even reach the plate in the inlet. This
condition is characterized by very high voltages and extremely low current.
An example of this type of situation would be a raw mill where an ESP is used
to control particulates from a preheater or precalcining kiln and all of the
material leaving the mill's preheater/precalciner enter the ESP. Grain
loadings up to 165 to 200 gr/acf could be encountered, and the ESP must be
able to handle this quantity of material.
4.2.7 Inleakage
Inleakage is often overlooked as an operating problem. In some instanc-
es, it can be beneficial to ESP performance, but in most cases its effect is
detrimental. Some of the causes of inleakage, which may occur at the process
itself or at the ESP, are leaking access doors, leaking ductwork, and even
open sample ports.
Inleakage usually cools the gas stream, and it can also introduce addi-
tional moisture. The result is often localized corrosion of the ESP shell,
plates, and wires. The temperature differential also could cause electrical
disturbances (sparking) in the field. Finally, the introduction of ambient
air can affect the gas distribution near the point of entry. The primary
entrance paths are through the access doors. • Inleakage through hopper doors
may reentrain and excessively cool the dust in the hopper, which can cause
both reentrainment in the gas stream and hopper pluggage. Inleakage through
the access doors is normally accompanied by an audible in-rush of air.
SECTION 4-PERFORMANCE EVALUATION, PROSLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-40
-------
Inleakage Is also accompanied-by an increase 1n gas"vo:1uine. In some
processes, a certain amount of Inleakage is expected. For example, applica-
tion of Lungstrom regenerative air* heaters on power boilers or recovery boil-
ers is normally accompanied by an increase in flue gas oxygen. For utility
boilers the increase may be from 4.5 percent oxygen at the inlet to 6.5 per-
cent at the outlet. For other boilers the percentage increase may be smaller
when measured by the 0, content, but 20 to 40 percent increases in gas vol-
umes are typical and the ESP must be sized accordingly. Excessive gas volume
due to air inleakage, however, can cause an increase in emissions due to
higher velocities through the ESP and greater reentrainment of particulate.
For example, at a kraft recovery boiler, an ESP that was designed for a
superficial velocity of just under 6 ft/s was operating at over 12 ft/s to
handle an increased firing rate, Increased excess air, and inleakage down-
stream of the boiler. Because the velocities were so high through the ESP,
the captured material was blown off the plate and the source was unable to
meet emission standards,
4.2,8 Summary
Familiarity with ESP operating characteristics, failure modes, and de-
sign factors aid in the diagnosis of ESP performance. As pointed out in the
discussion, many problems produce similar symptoms in T-R set meter readings.
Gathering process data and generating V-I curves add data that are useful for
diagnostic troubleshooting, but usually a process of elimination is necessary
to narrow the possible causes of problems to one or two areas. To the extent
possible, maintenance personnel should try to determine the cause of the
problem, not merely treat a symptom. This approach often identifies the
corrective actions necessary to avoid future problems, whereas the "band-aid"
approach is more likely to cause more problems in the future. Recordkeeping
is also very important in the evaluation of both short- and long-term ESP
performance.
4.3 CORRECTIVE ACTIONS
When the data collected indicate that a problem exists, plant personnel
must decide what action should be taken. Sometimes the initial cause of the
SECTION 4-PERFORMANCE EVALUATION. PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-41
-------
problem is hard to define, even though results and symptoms clearly indicate
its existence. In other cases, the problem is easily identifiable, but more
than one choice for corrective action is available. The options available to
plant personnel for the various problem areas discussed in Section 4.2 are
presented here.
4.3.1 Correction of Resistiy ity-Related Prob1 ems
When examining the alternatives for correcting resistivity problems,
plant personnel should ask three questions; I) How often does this problem
occur? 2) Can the process or materials be (economically, environmentally)
changed to minimize or eliminate the problem? 3) Is the operation of the ESP
optimal, or are there design limitations that cannot be overcome? The an-
swers to these questions may eliminate certain options and enhance the feasi-
bility of others.
For long-term resistivity problems, an option with high initial costs
may be the most cost-effective when lost production and increased maintenance
are considered. When resistivity problems are intermittent, however, chances
in process operating parameters may offer a better solution.
One of the simplest changes in process operation is to raise or lower
the gas temperature to increase either the surface conduction or bulk conduc-
tion mechanisms. This is particularly useful when the resistivity-versus-
temperature curves are steep and have a relatively sharp peak. In some situ-
ations, a change of only 20° to 30°F may be all that is required to modify
the resistivity and improve ESP performance. The disadvantages of changing
the gas temperature include an increased potential for corrosion from acid
dewpoint conditions if the temperature is lowered, and an increase in energy
loss due to a higher gas temperature. The peak resistivity often occurs at
or neer the optimum temperature for the process. In addition, some multi-
chambered ESP's may show symptoms of high resistivity in some chambers be-
cause of the difference in operating temperatures between chambers.
The addition of moisture to the gas stream may be an acceptable method
for conditioning the gas stream and improving performance. The moisture
changes the dewpoint levels, enhances the conduction of the participate
matter, and by increasing the dielectric strength of the gas, the electric
field is less likely to break down and spark. The evaporation of water
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-42
-------
droplets used to condition the gas stream will also cool the gas stream and
further condition it. The water droplets must be properly atomized to pro-
vide good evaporation without excessive water use or carryover into the ESP.
This often requires the use of air-atomized sonic nozzles. In some ESP
applications, the use of moisture conditioning is essential to the capture of
particulate. Two examples are salt cake recovery in kraft mills and cement
dust control in cement plants. Whether the presence of moisture is inherent
to the process (as in the case of kraft recovery process or wet process
cement kilns) or due to external addition (as in dry processes, some preheat
situations, or precalciner cement kilns), the lack of moisture would make it
difficult to capture the particulate matter because of its high resistivity.
Another alternative solution for ash resistivity problems involves a
change in process feed or operation to modify the chemical characteristics of
the dust to be captured. For example, the coal supply for coal-fired boilers
could be changed or blended to change the resistivity characteristics of the
fly ash (this may or may not include an increase in coal sulfur). A change
or blending of coal may improve collection efficiency, but care must be taken
to avoid boiler problems such as slagging or fouling of boiler tubes due to
the incompatibility of ash and furnace conditions. Another alternative
available to a boiler operator is to lower the combustion efficiency and
allow more unburned carbon into the ESP. Although this may reduce boiler
efficiency, the presence of small amounts of carbon can act as a conditioning
agent and reduce the resistivity of the fly ash. The trade-off in some situ-
ations is worthwhile because the slight loss in efficiency is offset by the
ability to maintain load rather than to reduce the load to maintain the
opacity limitation. Care must be taken not to overuse this solution, howev-
er. In some processes, such as cement kilns or municipal incinerators, it
may not be possible to change or control the feed characteristics to maintain
acceptable resistivity characteristics.
Low-resistivity problems are often more difficult to correct. Low
superficial gas velocities and optimization of rapper operation and rapping
pattern can help to minimize reentrainment problems commonly encountered with
low-resistivity dusts; however, generating the force necessary to keep the
collected dust on the plates is difficult. If gas temperature is a problem,
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-43
-------
one alternative is to increase the temperature towards the peak resistivity.
The problem is usually related to the process characteristics rather than gas
temperature, however. Previously cited examples include high-sulfur fuel,
excessive carbon levels, and the presence of natural but excessive levels of
conditioning agents (e.g., alkalies) in the dust. Ash from wood combustion
is a good example of this. Where possible, process changes will usually
help—such as changing the fuel sulfur level or improving combustion effi-
ciency to minimize the carbon content of the ash.
Conditioning agents have also been used to correct some low-resistivity
problems. One of the better known applications is ammonia injection in gas
streams containing large quantities of SCu/SQ.. Although tests have indicat-
£ O
ed that (as in the high-resistivity cases) no apparent change occurs in
resistivity, the agglomeration characteristic of particles on the plates
changes and results in improved particle-to-particle cohesion. This helps to
reduce the reentrainment problems that characterize low-resistivity dusts.
In addition, the ammonia may react with flue gas constituents to form fine
ammonium salt particles; e.g., ammonia combines with the SO- to form ammonium
sulfate. These fine particles increase the space charge and cause higher
operating voltages and the application of higher forces to the particles at
the plates.
The resistivity-related problems associated with ESP's have been well
studied in recent years, and the prediction of and design considerations for
resistivity have become more precise. Some combination of corrective actions
is usually available to the plant that experiences a high- or low-resistivity
problem. Each corrective measure has both an economic and practical consid-
eration, and site-specific factors will govern which options are chosen.
The remaining alternatives for correcting resistivity problems are more
complex and more expensive. One of these is to retrofit additional ESP plate
area to improve collection efficiency. This may entail the use of new de-
signs, such as wide plate spacings. Retrofitting also may be difficult
because of equipment placement and space limitations. Another alternative is
to change the T-R secticmalization and perhaps to use low-power (secondary
current limit) T-R's and/or pulse energization. The application of pulse
energization, although commercially available, has been relatively limited.
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-44
-------
The use of lower-power T-R's and increased sectionalization also can improve
performance by matching electrical capabilities to the demands of high resis-
tivity. The most common alternative, however, is the use of chemical condi-
tioning to modify the resistivity characteristics of the participate or to
reduce their effects on ESP performance.
In addition to proprietary chemical additives, typical chemical addi-
tives include SO,, sulfuric acid, ammonia, and soluble alkali salts, Most
rely on the chemical additive to improve the surface conduction mechanism of
charge transfer and to lower the effective resistivity of the particulate.
(One exception to this is ammonia injection, which generally has little ef-
fect on dust resistivity, but may alter the electrical characteristics of the
gas stream and the agglomerating characteristics of the ash on the plates,
The effects are often unpredictable. This has been used more on low-resis-
tivity problems caused by high-sulfur fuels.) The quantities of chemical
additives required and their effectiveness vary; however, in a typical coal
boiler application where the SO- option is often selected, the concentration
required ranges from 8 to 40 ppm. This combination has been very successful,
and some new ESP systems are being installed with a conditioning system and
smaller SCA as an economic alternative to large-SCA ESP's (without condition-
ing) or fabric filters.
4,3.2 Correction ofSodium Depletion Problems
Options are available for solving the sodium depletion problem noted in
Section 2 for hot-side ESP's at some utility boilers. Although periodic
washdown restores the ESP performance to acceptable levels, it is not an
altogether acceptable solution to the problem for most utilities. One analo-
gous solution that has been proposed is the modification of the rapping
system to rap the plates clem and remove the sodium-depleted ash layer at
the plate surface. The modification may have to be substantial, as not all
ESP's car withstand the high rapping forces necessary to cover the entire
plate, nor can all rapping designs supply it. It may be an alternative on
new ESP's, however, although high-intensity rapping alone is not likely to be
an effective solution to sodium depletion.
Several methods of ash conditioning have been attempted with varying
success. Because the electrical conduction mechanism jn the ash layer of a
SECTION 4-PERFOBMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-45
-------
hot-side ESP is bulk conduction -at normal operating temperatures, it relies
on alkali materials such as sodium to serve as electron carriers. As the
sodium content of the ash becomes depleted, resistivity increases. The
conditioning methods replace this sodium, usually with sodium carbonate, and
thereby reduce the resistivity. Two methods of introduction have been at-
tempted: introduction into the furnace zone with the coal and injection into
the flue gas preceding the ESP.
Although the injection of dry sodium carbonate into the gas stream has
provided some improvement in performance, large quantities of sodium car-
bonate may be required to effect reportedly small improvements. The addition
of 1 to 2 Ib/ton of coal to condition the ash produced in the furnace zone
has been more successful. The sodium apparently becomes homogeneously mixed
in the ash and thus is more effective in reducing the sodium depletion. The
basic premise of this conditioning method is that the higher concentration of
sodium in the dust layer increases the diffusion of sodium towards the plate
and overcomes the electrically related diffusion away from the plate. Al-
though primarily a high-temperature phenomenon, sodium depletion apparently
can occur at lower temperatures as well, but it is not nearly as severe.
Boiler slagging is one, .of the problems associated with boiler operation
when the sodium is mixed with the fuel. The increase in alkalies, particu-
larly sodium, tends to lower the ash fusion temperature. At full load, the
ash becomes liquid in the furnace and causes all sorts of operating difficul-
ties. The heat release rates become too high for the new ash characteris-
tics, and the rate must be dropped to cool the furnace. At some stations,
this may mean derating the load by 20 to 35 percent, and unfavorable econom-
ics may result. The balance point between load reduction and sodium deple-
tion is sometimes difficult to establish in these situations, particularly
when coal characteristics vary widely.
4.3.3 Corrective Actions for Dust Accumulation Problems
The most common cause of excessive dust accumulation on the electrodes
is failure of the rapper-control system. Unless there is reason to suspect
otherwise (e.g., known high resistivity potential or other indications of
hopper pluggage), this should be cne of the first areas checked if power
SECTION 4-PERFOfiMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-46
-------
input Into the ESP decreases markedly. This problem is1 relatively easy to
rectify.
Rapper failures may involve individual rappers, entire electrical fields,
or the entire ESP. Although isolated rapper failures are usually not severe,
they can affect the performance of the field they are serving. An easy
approach to this problem is to have a few rappers assembled in the spare
parts inventory for quick change-out when a malfunctioning rapper is found.
When the malfunctioning rapper is disassembled and the defective component(s)
replaced, the rebuilt rapper is then placed into the spare parts inventory.
The reasons for individual rapper failure depends on the rapper type.
Magnetic-impulse, gravity-impact (MIGI) rappers may fail because of a short
in the coil that lifts the rapper. Electric vibrators may fail because the
proper air gap or sealing from the elements has not been maintained. Air-
activated or pneumatic rappers often fail because water and/or oil enter the
compressed air lines or because the solenoid fails to open the air supply
line. Internal falling-hammer failures are more difficult to diagnose if the
problems are inside the ESP. Problems such as misalignment of hammers with
the anvils or a broken drive shaft usually cannot be diagnosed until the ESP
is shut down for an internal inspection. Two of the more common problems
with internal rappers are failure of the drive motor and failure of the gear
reduction system.
Failure of the rappers serving one field of the ESP can usually be
traced to the rapper control cabinet(s). Depending on the design, rapper
controls are usually in separate cabinets for both the discharge electrodes
and the collection plates. The collection plate controls nay be further
separated into individual cabinets for each field, or one cabinet may contain
the controls for all the rappers on the ESP. If a failure at the control
cabinet is the suspected cause for excessive dust buildup in the ESP, the
first check should be to see if the power is on and that the fuse or circuit
breaker has not been opened. A check should be made to ascertain proper
operation of the switch and drive on some older systems that use rotary
switches to activate the rappers. After these basic checks, corrective
actions become more complicated and dependent on the rapper manufacturer.
The manufacturer usually outlines specific procedures for testing the rapper
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS. AND PROBLEM SOLUTIONS
4-47
-------
control circuit boards for failed components. These are beyond the scope of
this manual; however, some interesting problems have been noted, particularly
with MIGI rappers. The most interesting problem is the continual failure of
the rapper control card components. Failure or lack of a diode to protect
the transistors from charges carried back to the rapper control cabinet may
contribute to this failure. Most rappers are constructed with internal
diodes for such protection, but not all of them. When no protection is
provided for the cards, there is a potential for many card failures. Another
problem involves the electrical arrangement of the rappers in such a manner
that failure of one disables all of them. In some cases, this arrangement
makes it nearly impossible to find the rapper that malfunctioned. Rappers
wired in this manner should be rewired for ease of maintenance, i.e., in
parallel instead of series.
When dust buildup is suspected and the rappers are in good operating
order, the available options are to increase rapping frequency or to increase
rapping intensity. Both options have certain advantages and disadvantages.
Not all rapper control systems can control both frequency and intensity;
however, if possible, an increase in rapping frequency is a good first choice
(usually the only choice.with internal rappers). Many dusts respond well to
this, whether the buildup is caused by an increase in dust generation rate or
an increase in resistivity. Rapping more frequently reduces the maximum dust
layer thickness if sufficient rapping energy is provided.
If an increase in rapping frequency does not improve electrical charac-
teristics after several hours, an increase in intensity may be required.
Large increases in rapping energy should be avoided, however. Increases
probably should not exceed 50 percent. Some ESP's are designed to withstand
only limited rapping forces before damage to the ESP can occur. High-resis-
tivity or sticky dusts usually require increased rapping intensity. If
increasing rapping frequency and intensity fail to remove the dust (particu-
larly high-resistivity dusts), a procedure called "power-off" rapping may
help. Removing the power from the field to be rapped greatly diminishes the
electric forces holding the dust to the plate and may allow the dust to be
removed. Usually only one field at a time is rapped with the power off—for
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS. AND PROBLEM SOLUTIONS
4-48
-------
a period of 15 minutes to an-hour.. The power is usually turned off manually,
although an automatic power-off option is available from some manufacturers
on new T-R/rapper control systems. The disadvantage of this procedure is
that it may increase the emissions from the ESP.
When all other measures fail to remove the material from the plates or
discharge electrodes, the remaining option is to shut the ESP down and wash
it out. This option should be kept at a minimum, as it may increase corro-
sion. Also, this can be an unpleasant task on cold January or February days.
This procedure generally will remove the dust from the internal components;
however, in some instances the dust layer must be scraped off, a time-consum-
ing and difficult task.
All of the procedures discussed assume that the ESP has sufficient
rapping capability, including appropriate rapping force, and that the design
does not assign the cleaning of too much plate area to a single rapper. One
aspect of the rigid-electrode or rigid-frame designs (of utility ESP's with
large plates) that tends to overcome dust accumulation problems is the use of
at least one internal rapper per plate. Two to four plates per rapper are
the standard on smaller rigid-frame ESP's. In weighted-wire designs, the
rapper may have to cover from two to six plates', however, the rapping inten-
sity of these rappers is controllable. Additional rappers can be retrofitted
to an ESP to reduce the plate area per rapper and to increase rapping effec-
tiveness.
4.3.4 Corrective Actions for MireBreakage
The general approach to correcting a wire breakage problem is to find
the broken wire in the field and clip it out during the next convenient out-
age. Most of the time a broken wire is not replaced, and the bottle weight
is removed to prevent its falling into the hopper. If wire failure is ran-
dom, as many as 10 percent of the wires can be removed without significantly
deteriorating ESP performance. Records should be kept of wire failure loca-
tions and dates to ascertain that they are indeed random.
If a pattern of failures begins to show, it should be interpreted as e
symptom of some other problem. For example, a pattern showing that wires are
SECTION ^-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-49
-------
failing in one area of the ESP or at the same location on the wire (top,
bottom, middle, etc.) should alert plant personnel that the problems go
beyond just a broken wire.
Several wire failure mechanisms were discussed earlier in this section.
The most common is failure of the wire either at a plate/wire misalignment
point or where the wire passes the edge of the plates in the collecting field
(end effect). Localized corrosion due to inleakage may be reduced by reduc-
ing inleakage (sealing access doors, maintaining duct integrity). Crimping
of the wires, which can cause excessive sparking or corrosion, may be a
manufacturing and/or installation defect. In these instances, these wires
may have to be replaced.
The problems with misalignment and end effects can be solved in various
ways. When a misalignment problem is localized within an area of the ESP
(e.g., because of three or four warped plates), plant personnel may opt to
remove the wires in the gas lanes. This will usually improve operating
voltages and current and prevent excessive wire breakage. This option should
not be exercised unless the plates cannot be straightened. The danger of
this procedure is that it may allow dust-laden gas to pass through the ESP
essentially untreated if other adjacent upstream and downstream fields happen
to be deenergized. To avoid this possibility, some plants weld plate steel
into position to block the flow of gas down these lanes and force it to flow
down other lanes. When the entire field is misaligned due to plate warpage
or guide misalignment, clipping wires is not a sound solution.
Two methods are used to minimize excessive sparking due to end effects
et the plates. The first is the use of wire shrouds that extend 6 to 18
inches from both ends of the wires. These shrouds are approximately 3/8 to
1/2 inch in diameter and generate corona only at very high operating voltag-
es. In addition, it takes a long time for the spark to cause electrical ero-
sion of the large effective wire diameter. The second method (designed into
new ESP's) is the use of rounded plates at the top and bottom rather than
sharp edges with an effective diameter equivalent to the plate thickness. In
the new designs, the ends are rounded and approximately 2 to 3 inches in di-
ameter. By reducing the distortion of the electrical field at the end of the
SECTION 4-PERFORMANCE EVALUATION, PRO1LEM DIAGNOSIS AND PROBLEM SOLUTIONS
4-50
-------
plates, this new design reduces sparking. This latter change is difficult to
retrofit into an existing ESP, but wires with shrouds can be used to replace
most unshrouded wires and should be considered when wholesale wire replace-
ment is scheduled.
When the 10 percent random wire failure rate has been reached or if
more than 5 to 10 wires in any gas lane have been removed (depending upon ESP
design), replacement of broken wires should be considered. This does not
mean all the wires must be replaced. If the wire breakage rate appears to be
on the increase, however, it may be advisable to replace all the wires.
During replacement, care must be taken to avoid crossing wires and causing
premature failure due to wear.
4.3.5 Corrective Actions for Hopper Pluggage
When hopper pluggage is detected, immediate action should be taken to
clear the pluggage and empty the hopper. Maintenance personnel should give
this problem highest priority, as failure to respond can significantly reduce
long-term ESP performance.
Hopper pluggage can be caused by foreign objects such as wires or weights,
cooling of the dust in the hopper or hopper throat, an undersized dust-convey-
ing system, or the introduction of moisture into the hopper. Hoppers should
be equipped with level detectors. They also should be insulated, and in many
cases should be equipped with heaters. Rod-out capabilities are necessary,
and a method of removing the hopper throat for emptying would be advantageous.
If a hopper is plugged but not filled, one appropriate action is to
place the T-R controller for the field(s) above the plugged hopper in the
manual mode to reduce the collection rate until the hopper is ready to be
cleared. If the hopper is completely filled and the T-R has not tripped
automatically, it should be turned off until the hopper has been cleared.
The T-R must be deenergized while clearing of the hopper is in progress, to
prevent electrocution of maintenance personnel. When ESP's are overdesigned
in terms of sectionalization and plate area, their hoppers and dust removal
system are often underdesigned. Because their Inlet fields collect more
material than was planned, the overload causes the hoppers to plug while the
outlet fields remain virtually empty. When this occurs, the dust removal
SECTION 4-PERFOHMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-51
-------
from the gas stream can-be spread more evenly in the ESP by reducing power
input to the fields that are plugged. This permanent reduction in power may
change the pattern within the ESP and adversely affect its performance. The
ESP must be adequately sized for this option to be exercised. (Note: Mal-
distribution of the gas stream can produce similar effects.)
The use of vibrators on the hopper does not necessarily enhance the
flowing properties of all dusts. In fact, it can worsen the situation by
compacting the dust in the hopper. Striking the hopper, however, can dis-
lodge "bridges" of dust attached to the hopper walls. The hopper wall or
throat should not be struck, as the impact may cause damage that provides a
future site for hopper bridging and pluggage; rather, reinforced strike
plates should be installed for the occasional strike that may be needed.
The use of hopper "stones" to fluidize the dust can provide mixed re-
sults. Placed near the bottom of the hopper, they may contribute to the
pluggage problem by closing off open areas in the hopper. These stones,
which are similar to those used to aerate aquariums, but much larger, must be
supplied with dry, heated air. A high-quality dryer is needed to remove the
moisture and oil, and the air must be heated so it will not cool the dust in
the hoppers.
Excessive cooling of the dust can cause problems. Condensation of acid
or moisture can solidify some dusts, make others extremely sticky, and simply
cause some not to flow well. Hopper heaters and insulation usually help keep
the hoppers warm. Pluggage of hoppers on the windward or north side of the
ESP in the winter can cause excessive amounts of heat to be carried away
from the hoppers. In many cases this can be corrected by constructing a
windbreak or enclosing the hoppers to reduce cooling effects.
Finally, hoppers are usually sloped so as to provide good flowing char-
acteristics. Gas sneakage baffles that project too far into the hopper
sometimes contribute to the bridging problem. These baffles are desirable
for preventing gas sneakage, and if bridging occurs, they can be shortened to
minimize hopper pluggage.
If dust is suspected to have reached the plates and wires during an in-
cident of hopper pluggage, a gas-load V-I curve should be generated to deter-
mine that no buildup, clinkers, or serious misalignment has occurred in the
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-52
-------
field(s). The T-R should then be returned to normal operation. When mainte-
nance personnel clear the hoppers or open a hopper access door, care must be
taken to avoid electrocution or being overcome by the dust or gas stream.
4.3.6 Corrective Actions for Misalignment
To perform the necessary work for correcting misalignment in the ESP
requires an outage. Depending upon the extent of the misalignment, solutions
range from simple application of heat and pressure to complete removal and
replacement of the plates or discharge guide frames. As the repairs become
more complex, the time required for correction also increases.
Failure of support, compression, or stand-off insulators for the dis-
charge electrode system can cause widespread misalignment within the ESP
between the wires and plates. Hopper pluggage can also shift the lower guide
frame and contribute to the failure of standoff insulators. As indicated
previously, reliable hopper level indicators are a means of avoiding mis-
alignment due to hopper pluggage. The replacement of the insulators and
subsequent realignment is relatively easy and straightforward if none of the
internal ESP components have been bent.
Misalignment due to bent plates, bent wire, or bent rigid frames is more
difficult to correct. If the point of misalignment is near the edge of the
field, adequate access may be available to attempt corrective actions; how-
ever, misalignment that is halfway into the field, where access is poor, may
be difficult to correct.
Plate straightening can be attempted by one of several methods. The
simplest is to bend the plate back into shape by use of a small hydraulic
press. Sometimes heating with a torch is alternated with water quenching to
relieve the stress on the plate while returning it to position. Another
option (for small sections of plate) is to remove the warped section with a
cutting torch and replace it. (Note: Some plants merely cut away the warped
section without replacing it, although this is poor practice.) This approach
is generally limited to a section of plate edge small enough to fit through
the access doors. Central portions of the plate area are generally not
accessible. Care must be taken to remove all burrs and to smooth the plate
before operation is begun again.
SECTION 4-PERFORMANCE EVALUATION, PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-53
-------
Some plants have tried replacement of plate sections or panels in ESP
designs where the plate is composed of individual sections approximately 18
inches wide. In this time-consuming procedure, the plate sections to be
removed are undipped and detached from the plate hangers and guide. If
enough room is available in the ESP for the panel and lifting equipment, the
new panel or panels are brought in through the access door and the plate is
reassembled. In major rebuilding involving replacement of large portions of
plate area, it is often easier to remove the roof of the ESP and to replace
the plates with a crane.
Bent wire frames or lower guide frames often cause the wires to slacken
and bow towards the plates. Distorted lower guide frames are often difficult
to straighten and may have to be replaced; however, if the distortion is not
too severe and only a few wires are slack, it may be worthwhile just to
remove those wires. On rigid-frame designs, space limitations often dictate
the extent, of straightening that can be done to the discharge frame. The
wires may be tightened by crimping them in the direction of gas flow. This
tightens the wires and prevents bowing towards the plate. It can also in-
crease sparking and damage to the wire. Because the wires in rigid frame
ESP's are usually of a larger diameter, however, it is usually an acceptable
trade-off.
In summary, if a general misalignment is caused by a shift in guide
frame components, it usually can be corrected by realigning the frame. Plate
warpage or wire frame warpage may be more difficult to correct, however. All
ESP's have a certain freedom of motion to allow expansion and contraction.
Checks should be made to see that this freedom is maintained and that there
is no binding before the other approaches to correcting misalignment are
considered.
4.3.7 Corrective Actions for Inleakage
The corrective action for inleakage is straightforward. Unless the
design calls for the admission of ambient air for a specific purpose, any
spot of inleakage should be sealed. Such an approach reduces the total gas
SECTION 4-PERFORMANCE EVALUATION. PROBLEM DIAGNOSIS, AND PROBLEM SOLUTIONS
4-54
-------
volume to the ESP's, can help to prevent acid dewpoint problems, contributes
to more stable operation, and can enhance the gas distribution into the ESP,
All access doors should remain sealed during routine operation. This
includes hopper access doors and penthouse doors. On newer designs, the use
of double doors for sealing and insulating has become popular to minimize in-
leakage and door corrosion. Gaskets should be checked periodically and
replaced if damaged. Any corrosion around doors or expansion joints should
be corrected immediately, as the corrosion rate tends to accelerate around
these points.
A routine check of oxygen content and temperature in combustion flue gas
is a useful indicator of inleakage. Such checks should be conducted along
the ductwork from the process exit to the ESP outlet. (Note: A system under
positive pressure generally does not have inleakage problems, but outleakage
may contribute to fugitive emissions and exterior corrosion.) Any sudden
increase in (L in a nonstratified gas stream indicates inleakage. This is
usually accompanied by a corresponding decrease in temperature. The point of
inleakage will usually be between the sampling locations where the change in
Op occurred. The causes can be a broken or worn seal, a hole in the duct-
work, or sampling ports that someone left open. Appropriate action should be
taken.
4.3.8 Summary
Because problems are often site- and design-specific, extensive coverage
of corrective action is not possible. Nevertheless, the major problems have
been addressed. As previously stated, it is useful to understand the nature
and magnitude of the effects of various ESP problems. Before selecting any
corrective action, one must first determine the cause of the problem so that
the proper corrective action is chosen rather than one that merely treats the
symptom. Long-term ESP performance usually benefits from this approach.
SECTION 4-PERFOWMANCI EVALUATION, PROBLEM DIAGNOSIS. AND PROBLEM SOLUTIONS
4-55
-------
REFERENCES FOR SECTION 4
1. McDonald, 0. R, and A, H. Dean. A Manual For the Use of Electrostatic
Precipitators to Collect Fly Ash Particles. EPA-600/8-80-Q25. May
1980.
SECTION 4-PERFORMANCE EVALUATION. PROBLEM DIAGNOSIS. AND PROBLEM SOLUTIONS
4-56
-------
SECTION 5
O&M PRACTICES
5.1 OPERATING PRACTICES
Operating practices can significantly affect daily and long-term ESP
performance. Because they are site specific, only the most general of prac-
tices can be presented here, and their characteristics must be considered in
the establishment of operating practices and procedures. These practices and
procedures should be-straightforward and cover most of the situations ex-
pected to be encountered, and personnel should be trained so that these
practices become routine.
5.1.1 Startup Practices
Startup practices greatly affect the subsequent operation of an ESP and
may be as important to performance as daily operating checks and maintenance
practices. Major concerns during the startup of ESP's on most processes are
corrosion and buildup of material on plates, wires, and insulators, which can
severely limit ESP performance. For some processes, the potential hazard of
fire and explosion resulting from unstable operations is a primary concern.
Startup practices should be geared to address these potential problems.
Personnel safety should be the foremost consideration in any startup
procedure. Before anyone enters an ESP for inspection or maintenance, ground
devices should be installed to be certain that all built-up charge is dis-
charged and that T-R sets are disabled and main breakers are locked out so
they pose no threat to personnel. When preparing for startup of the ESP,
these safety devices must be removed so that the T-R can operate. One person
or small group of persons should be responsible for checking the ESP to
StCTION S-O&M PRACTICES
5-1
-------
ascertain that it has been cleared of all personnel, tools, parts, and other
miscellaneous materials and that the scheduled maintenance has been complet-
ed. For the sake of discussion, let us assume that an inspection for mainte-
nance, alignment, etc. is performed separately. When the check for clearance
of personnel, tools, and equipment has been completed, the ESP should be
closed up and locked by use of a safety key interlock system. When several
persons are involved in the final checkout for startup, the responsibilities
of each person or group should be clearly outlined. Checklists are useful
reminders for the individual items to be checked and serve later as written
reference that specific items have been checked.
When the ESP has been closed up and the keys for the interlock system
have been returned to their appropriate locations, an air-load test should be
run for each T-R set and, if time permits, for each bus section to ascertain
that all maintenance has been completed, all foreign matter removed, and that
the ESP is ready for operation. The air-load test is conducted under ambient
conditions with little or no air flow during energization of each section.
The test result should be an air-load curve for each T-R, which should become
part of the permanent record for the unit. Points of interest are the volt-
age at corona initiation, -the shape of the voltage/current curve, and the
voltage and current values when sparking occurs. The preferred readings are
secondary voltage and secondary current, but primary voltage and primary cur-
rent may be used. When secondary voltage meters are not available, primary
voltage versus secondary current is sometimes used. The voltage and current
should be increased gradually until sparking occurs. If sparking does not
occur, the power should be increased until the current limit is reached.
(Note: The voltage limit is rarely reached.)
For fields of similar geometry (same T-R capacity, same wires, wire-to-
2
plate spacing, ft /T-R, etc.), the V-I curves that are generated should be
similar. Slight differences may occur because of internal alignment or
buildups, but generally each curve should be nearly identical under air-load
conditions. Any major deviation between the curves or between a curve and a
reference curve (when the unit was new) may indicate an internal problem.
Spotting a problem from these curves may afford the opportunity to correct
SECTION 5-04M PRACTICES
-------
that problem before startup. This procedure also provides .a final check that
the scheduled maintenance was performed during the ESP shutdown.
The symptoms of various problems are shown in Section 4. The air-load
test curves will differ from gas-load conditions, when operating temperatures
and participate cause the curves to shift. Any similarity between air-load
and gas-load curves is coincidental.
Insulator heaters and hopper heaters should be turned on 2 to 12 hours
prior to startup. Purge-air systems also should be activated at this time.
Because environmental regulations generally do not allow bypass of the con-
trol system during startup, some portion of the system must be in service.
Condensation should be minimized in these areas to prevent insulator leakage
or tracking and to prevent pluggage of hoppers at startup.
Even if the ESP is not energized, the rapping system and hopper evacua-
tion system should be in operation during startup to remove any dust that has
settled. Because gas loads and particulate loadings are much below normal
operating loads, the ESP can be an effective settling chamber and remove the
large-diameter particulates entering it. An increase in rapping intensity is
normally suggested to dislodge any wet (from condensation) or sticky ^articu-
late that has collected on the wires and plates to prevent buildup problems
from affecting performance as full load conditions are approached. Rapping
intensity can then be returned to normal after operating temperature is
reached.
When and how much of the ESP should be energized is a much-debated topic
in many applications. Many believe that energization should not be attempted
until the moisture dewpoint has been exceeded for several hours, and some
like to wait until the acid dewpoint temperature has been surpassed, to avoid
excessive sparking and to minimize buildup on the ESP internals. Newer, mod-
ern, power supplies, however, should be able to energize the ESP with little
or no sparking, whereas older controllers probably should be set in manual
control below the spark threshold. The concern with sticky participate and
sparking is two-fold. At best, if the sticky particles fuse or clinker on
the plates and wires, the ESP will have to be shut down and washed out. At
worst, unstable combustion could lead to unburned carbon, hydrocarbons, and
SECTION 5-0&M PRACTICES 5-3
-------
CO in the ESP, which may-be set afire or result in an explosion that endan-
gers personnel and could also permanently destroy the ESP's performance
potential. Substantial misalignment can occur within an ESP as a result of a
fire or explosion. Because of the tendency for an ESP to spark during start-
up and provide a possible ignition source, some plants have approached the
explosion potential by lengthening their startup times. For example, in some
cement kiln applications, startup times, of up to 12 hours are not uncommon to
avoid CO explosions that may be caused by unstable operation during startup.
It is ironic that the monitors used to detect CO or combustibles and deener-
gize the ESP to prevent explosions or fire are so slow in detecting a "spike"
that by the time the danger is monitored the combustible mixture has passed
through the ESP. If ignited by a spark, the resulting fire or explosion is
noted by the continuous monitor from 15 seconds to 2| minutes after the fact
and merely serves as an epitaph to the damaged ESP.
Energizing the entire ESP during startup is usually unnecessary; only
enough of the ESP should be energized to maintain emissions below opacity
limitations. This means that the ESP is brought on line stepwise; the number
of T-R's energized is increased as particulate and gas load increase. Reconi-
mendations vary as to which fields should be energized first. Usually either
inlet or outlet fields are recommended for initial energization, and there
are good arguments for either approach. The argument for inlet field ener-
gization is based or, the fact that scouring action on the wires and plates
tends to occur when loads are increased; because these fields experience the
heaviest loading at full capacity, any deterioration of performance due to
buildup of particulate will be minimized. The argument for outlet field
energization is that the cleaning action of the heavier particulate may never
occur and that these charged and collected undesirable particles can pene-
trate into the second (and possibly third) field and cause unsatisfactory
performance of the ESP. If fouling occurs in the ESP, it occurs in the out-
let field, whose function is usually the final cleanup of rapping emissions
and fine particle capture, and this field is less crucial to mass removal
than are the inlet and second fields. This is a version of controlled damage
SECTION 5-O&M PRACTICES
-------
potential. It is further argued that damage due to fire'or explosion is con-
fined to one field. This may be true for fire potential, but not necessarily
for an explosive condition.
As the process load and gas temperature increase, more of the ESP is
energized and higher power levels may be achievable. Some manufacturers have
controllers which automatically increase power levels based on opacity read-
ings. As full load is approached, the entire ESP should be energized, placed
in automatic control, and optimized for power input and spark rate. After
several hours of stable operation, rapper intensity should be set back to
normal levels.
5.1.2 Routine Operation
During routine daily operations, some variations will occur in operating
voltages and currents, depending on dust characteristics and process opera-
tions. Unless variations are extreme, daily requirements include parameter
monitoring and recordkeeping, preventive maintenance, evaluation for malfunc-
tions, and response to malfunctions. If no problems arose during startup,
performance should be governed by daily operations. (Parameter monitoring,
recordkeeping, and evaluation for malfunctions were discussed in Sections 3
and 4, and preventive maintenance will be discussed in Section 5.2.)
Response to malfunctions is important in avoiding deteriorating perform-
ance and possible damage to the ESP. These malfunctions have a tendency tc
cascade and build upon themselves. For example, a broken wire can cause
deterioration of ESP performance in a roundabout way. The broken wire may
result in a T-R trip end a weight or wire in the hopper, which may in turn
plug the hopper. This allows dust to build up, which warps plates and mis-
aligns wire guide frames, and ultimately results in diminished ESP perform-
ance even'after the T-R operation is restored. Routine checks of voltage and
current levels, rapper operation, dust removal, opacity, and process opera-
tion are necessary to catch problems early.
Some operations experience gradual deterioration in performance, usually
due to resistivity problems, but the generation of sticky particulate can
also limit performance. One approach to this problem is to try to optimize
SECTION 5-O&M PRACTICES c r
D-D
-------
ESP performance and to ma'ke changes in the process feed or operation. This
may include conditioning of or restrictions on process materials to prevent
the generation of the participate that causes this degradation with time.
Another approach is to allow the performance to degrade and schedule periodic
washdowns of the ESP. Some combination of both procedures may be applicable,
wherein some conditioning or restriction of fuels or process materials slows
the performance degradation and allows scheduled shutdowns to be more widely
spaced. This latter approach generally requires more careful monitoring to
avoid rapid and uncontrollable degradation.
5.1.3 Shutdown Practices
Except for emergency shutdown procedures, the process should be essen-
tially the reverse of startup procedures, but much simpler. As process load
decreases, the ESP generally can be deenergized one field at a time. As in
startup, the selection of the first field to be deenergized is a site-specif-
ic decision; however, deenergization of the inlet fields is usually favored.
As each field is deenergized the electrical field that held the dust to the
plates is released. Having the next field in line energized reduces the
quantity of emissions. In addition, the reduction in particulate load will
reduce the quantity of material on the plates. Again, higher than normal
emissions may be allowed during shutdown, but the extent of the emissions
should be,< reduced insofar as possible.
When the process is shut down, all remaining T-R's should be deener-
gized. Again, sequential deenergization toward the outlet is preferred, but
it should be done relatively quickly to prevent unnecessary sparking, conden-
sation, cr insulator buildup. The rappers and hopper evacuation system
should be allowed to run anywhere from several hours up to 24 hours to remove
as much dust as possible from the ESP. (Under special circumstances, rapper
operation may be terminated at shutdown to retain dust on the plates so pat-
terns of buildup can be observed during a subsequent internal inspection and
washdown; however, because turning off the T-R's will affect these patterns,
little may be gained from turning the rappers off.) Note: Shutdown proce-
dures can have a bearing both on the maintenance required during the outage
and on the success of the next startup.
SECTION S-04M PRACTICES
5-6
-------
In general, shutdowns are better controlled and present less problems
than startups do, primarily because the equipment is warm and has been under
relatively stable operation. In the case of an emergency shutdown, however,
one practice is almost universal—the entire ESP is tripped off line. Emer-
gency conditions may include anything from a fuel feed problem in a cement
kiln or boiler that presents a potential fire or explosion hazard to a rup-
tured tube in a boiler or a turbine trip that takes the boiler off line.
Under some circumstances, the ESP power supplies are interlocked to these
systems rather than relying on a monitor to measure possible combustible or
explosive conditions in the gas stream. The sudden release of emissions due
to deenergization can be substantial and, in some cases, in excess of normal
operating emissions for the rest of the year; however, this may be necessary
to avoid expensive repairs and excessive downtime. Some consideration should
be given to both the short- and long-term effects. Overreaction to a problem
is not a solution. For example, rupturing of waterwall or economizer tubes
usually calls for immediate deenergizing of the ESP even though the increased
water in the gas stream may improve performance dramatically. If the dust is
not removed from the ESP immediately, the usual highly undesirable result is
mud in the ESP. On the other hand, failure of the tubes invthe superheater.
or reheater area of a boiler may not necessitate immediate and complete
deenergization of the ESP, even though the boiler is tripped out of service
rather quickly. Again, the reader is reminded of the cement application,
where the current generation of the monitors is probably inadequate to pro-
tect against a fire or explosion,
If a fire or explosion does occur, the ESP should be deenergized.
Although a certain amount of damage has probably already been done, deener-
gizing may help to contain the damage. Fires should be allowed to burn out
by themselves, but in the case of a hopper fire, the hopper should be emptied
through the ash conveying system. The hopper door should never be opened,
and water or steam should never be used to put out a fire. The inrush of air
may increase the burning rate or set up an explosive mixture, and under
certain circumstances water can be reduced to hydrogen, which increases the
explosive potential.
SECTION 5-O4M PRACTICES
-------
5.2 PREVENTIVE MAINTENANCE
The goal of preventive maintenance is to maintain the long-term perform-
ance of the ESP and to reduce or minimize the failure of various components
that affect ESP performance. An important aspect of preventive maintenance
is routine inspection of the ESP, both internally and externally. These
inspections include daily or shift inspections, weekly inspections, monthly
or quarterly inspections, and outage inspections (the only time internal
inspections can be performed). Depending on the unit's operating history and
the manufacturer's recommendations, internal inspections can be performed
quarterly, semiannually, or annually. As the time interval increases, the
amount of action required usually increases. Daily and weekly inspections
may require checks of operating parameters and general operating conditions;
whereas monthly or quarterly inspections require specific actions regardless
of performance of the ESP.
5-2.1 Daily Inspection and Maintenance
Most often the daily inspection or shift inspection will be conducted as
part of a parameter monitoring and recordkeeping plan. The purpose of this
routine and frequent inspection is to identify the existence of any operating
problems before they develop Into more serious and possibly more damaging
failures. It is extremely important to have all ancillary equipment equipped
with alarms,^and plant personnel should respond to these alarms immediately.
As has been discussed in Section 4, some problems can and have led to long-
term degradation of ESP performance because they failed to be diagnosed or no
action was taken to correct them,
The instrumentation available for most ESP's provides the first indi-
cator of performance problems. Process operating data and ESP corona power
levels should be recorded and compared against baseline values or normal
values established for the source. The corona power values may be obtained
from primary voltage, primary current, secondary voltage, and secondary cur-
rent levels for each field in the ESP. In most applications, opacity values
also may provide an indication of ESP performance. Although most regulations
require the data to be reported on the basis of integrated 6-rninute averages,
SECTION 5-O&M PRACTICES
-------
the ability to observe the magnitude and frequency of individual rapping
spikes is beneficial in optimizing ESP performance. Any changes in these
values, either from previous normal readings or baseline conditions, are
important because they may indicate the need for further investigation and/or
maintenance.
Once the readings have been obtained and checked for any apparent chang-
es, some simple external checks are in order. If the operating values of a
field or fields show considerable change, the remainder of the inspection
should concentrate on attempting to identify the cause of the observed chang-
es. The electrical readings should identify at least one and possibly more
of the potential causes for the change. In addition to the electrical read-
ings, the following items also should be reviewed.
The operation of the dust discharge system should be checked. All con-
veyors, airlocks, valves, and other associated equipment should be operating
for continuous removal of the collected dust. If hopper heaters are neces-
sary to maintain the hopper temperature, the current levels at each hopper
will indicate whether these heaters are operating. Hopper throats should be
warm to the touch; a cold hopper throat may indicate that the hopper is
plugged. If a vacuum system is used, vacuum charts may provide a useful
indication of proper hopper emptying. Of course, the operation of indicator
lights on hopper level alarm systems also should be checked. While in the
hopper area, the one performing the inspection should check all access doors
for audible inleakage or dust discharge.
The operation of the rappers should be checked. This may be difficult
on some systems because of the delay times between rapping cycles and a
"maintenance-check" mode. Although it is not necessary to check the opera-
tion of every rapper, almost every rapper should activate. .. The purpose of
this check is to determine if the entire field or the entire ESP rapping sys-
tem is out of service. If possible, individual rappers that are not operat-
ing should be identified so that appropriate maintenance may be performed.
If monitored, the frequency and intensity of rapping also should be noted.
Other checks of the ESP externals include sparking or arcing in the T-R
high-voltage bus duct, localized sparking (usually reflected by T-R read-
ings), and audible inleakage around all other access hatches on the ESP.
SECTION 5-OSM PRACTICES 5-9
-------
Although these problems will generally affect ESP performance, if they are
corrected in a timely manner, the effect will not be long-term. Figure 5-1
summarizes the data that the operator or ESP coordinator should record daily.
5.2.2 Meekly Inspectionand Maintenance
The best way to start a weekly inspection is with a brief review of the
daily or shift inspection data. This review should attempt to identify
any apparent trends in the key operating parameters and to determine whether
a change is needed in some operating practice or maintenance procedure. In
addition, this review should confirm that all requested or required mainte-
nance has been completed satisfactorily or has been scheduled in a timely
manner. Lastly, a week is generally sufficient time for a change in opera-
tion (e.g., rapping intensity and timing, some process changes, gas-condi-
tioning systems operation) to surface in ESP performance, even though longer
periods may be necessary to establish the trend.
After reviewing the previous week's operating data and comparing it
against normal or baseline values, the inspector should make a physical check
of the ESP, including daily or shift inspection activities. The items cov-
ered in the following paragraphs also should be checked.
When the T-R readings are obtained, the cabinet air filters should be
checked and cleaned or replaced. Although some cabinets may not be equipped
with cooling fans, the air intakes and vents should be filtered to keep the
cabinet internals clean. Dust buildup on circuit boards and heat sinks can
cause excessive heat buildup and electrical problems within the control
circuitry. It is generally recommended that T-R control cabinets be placed
in a temperature-controlled, clean environment to minimize failure of con-
troller circuitry. Integrated circuits cannot withstand the high tempera-
tures that may accompany a hot, dirty environment.
Rapper operation should be checked more thoroughly during the weekly
inspection. Each rapper or rapper system should activate, and those that do
not should be scheduled for repair or replacement. Proper operation of
internal falling-hammer systems is more difficult to ascertain, but rapper
drive motors should operate when activated by the rapper controller. If
SECTION 5-OAM PRACTICES
5-10
-------
DAILY INSPECTION CHECKLIST
0 Corona power levels (i.e., primary current, primary voltage, secondary
current, secondary voltage) by field and chamber, (twice a shift)
Q Process operating conditions [i.e., firing rates, steam (lb/h)t flue gas
temperature, flue gas oxygen, etc.]. The normal operator's log may
serve this purpose, (hourly)
0 Rapper conditions (i.e., rappers out, rapper sequence, rapper intensity,
rapping frequency by field and chamber, (once a day)
0 Dust discharge system (conveyors, air locks, valves for proper opera-
tion, hopper levels, wet-bottom liquor levels.
8 Opacity (i.e., absolute value of current 6-minute average and range or
magnitude of rapper spiking) for each chamber duct if feasible. (2-hour
intervals)
0 Abnormal operating conditions (i.e., bus duct arcing, T-R set control
problems, T-R set trips excessive sparking), (twice'a shift)
0 Audible air inleakage (i.e., location and severity), (once a day)
Figure 5-1. Items that the operator or ESP coordinator should record daily.
SECTION 5-Q&M PRACTICES
-------
rapper settings are provided (i.e., rapping intensity, duration, and frequen-
cy), these should be recorded. Changes in rapper operation should be made as
necessary or desired for optimum operating characteristics. The new settings
should also be recorded, and the performance should be assessed during the
next week of operation.
While on the roof of the ESP, the inspector should note the operating
temperature and oil level in the high-voltage transformer. Generally, indi-
cators on the transformer will show the desired operating range. In addi-
tion, the operation of any insulator purge air and heating systems should be
checked. Air filters on the purge air system should be checked and cleaned
or replaced. Failure of the purge air system may cause fouling or condensa-
tion on insulator surfaces, which can result in electrical tracking or break-
age of the insulator. In negative pressure systems, if insulators can be
contaminated by variations in the ID fan operation cr during startups, the
insulator pressurization and heating system should be checked more often.
All access hatches should be checked for inleakage, both by listening
and feeling around the hatch. The hatch should be fully closed and locked.
If inleakage occurs, the door gasket (or possibly the door) should be
replaced. Inleakage can cause excessive sparking in localized areas, cor-
rosion, reentrainment of particulates, and wire breakage due to reduced
temperatures-. Figure 5-2 summarizes the items that the operator or ESP
coordinator should record weekly.
5.2.3 Monthly to Quarterly Inspections and Maintenance
Whether a monthly or quarterly inspection is required depends on the
manufacturer's recommendations and on selected site-specific criteria; how-
ever, all of the recommended procedures should be followed at least quarter-
ly.
The control cabinets for T-R sets and rappers should be cleaned (vacu-
umed) to remove any accumulated dust and dirt inside the cabinet. The exte-
rior of the cabinets for rapper controls should be checked for proper seal of
the door gaskets. If allowed to deteriorate, these gaskets will permit water
and dust to enter the cabinet and may result in failure of the rapper control
cabinet. In addition, all switch contacts within the rapper control cabinet
should be cleaned,
SECTION 5-O1M PRACTICES _
-------
WEEKLY INSPECTION CHECKLIST
Trends analysis (plot gas load V-I curves for each field and chamber and
other key parameters to check for changes in values as compared with
baseline).
Check and clean or replace T-R set cabinet air filters and insulator
purge air and heating system filters.
Audible air inleakage (i.e., location and severity).
Abnormal conditions (i.e., bus duct- arcing, penthouse and shell heat
systems, insulator heaters, T-R set oil levels, and temperature).
Flue gas conditions exiting the ESP (i.e., temperature and oxygen con-
tent).
More extensive rapper checks (also optimize rapper operation if needed).
Figure 5-2. Items that the operator or ESP coordinator should record weekly.
SECTION 5-OSM PRACTICES
5-13
-------
All rappers should be checked for proper operation. Checks should in-
clude proper striking of anvils for MIG.I rappers, proper lift, and energy
transfer. Pneumatic systems and vibrators should be adjusted as necessary to
transfer the proper amount of rapping energy to the collection or discharge
systems. Although ESP's equipped with falling-hammer rappers are more diffi-
cult to check, gear-reduction drive systems should be checked for proper
sealing. The boots may fail over a period of time and allow water and cool
air to enter the ESP. This can cause local corrosion or ESP operating prob-
lems due to inleakage.
Hopper heaters should be checked for proper operation if the ESP is so
equipped. As previously mentioned, the purpose of these heaters is to help
the dust stay warm and permit it to flow smoothly. Hopper-level alarm sys-
tems should be checked for proper operation.
The ESP's instrumentation should be checked and calibrated. This in-
cludes all voltage and current meters. The primary voltage and current
meters should be calibrated as AC RMS values, whereas the secondary voltage
and current meters should be calibrated against a DC power supply. The
transmissometer should also be cleaned and realigned during this inspection.
Figure 5-3 summarizes the items that the operator or ESP coordinator should
record quarterly.
5.2.4 •Semiannual Inspection and Maintenance
Depending on the operation, the semiannual inspection may correspond
with a maintenance shutdown or outage, and internal inspections are conduct-
ed, during such outages. Considerations for an internal inspection are
discussed under the Annual Inspection subsection.
In addition to the recommended procedures for quarterly and weekly
inspection and maintenance, semiannual procedures should include the lubrica-
tion of all door hinges and closure mechanisms, the cleaning and lubrication
(with graphite) of key interlocks, and a check of all ground connections
[grounding straps and sampling and testing transformer oil for maintenance of
dielectric strength (insulating capability)].
SECTION 5-OtM PRACTICES
5-14
-------
QUARTERLY INSPECTION CHECKLIST
Internal inspection of shell for corrosion (i.e., doors, hatches, in-
sulator housings, wet-bottom liquor level, dry bottom, roof area).
Effectiveness of rapping (i.e., buildup of dust on discharge electrodes
and plates).
Gas distribution (i.e., buildup of dust on distribution plates and
turning vanes).
Dust accumulation (i.e., buildup of dust on shell and support members
that could result in grounds or promote advanced corrosion).
Major misalignment of plates (i.e., visual check of plate alignment).
Rapper, vibrator, and T-R control cabinets (motors, lubrication, etc.).
Rapper distribution switch contacts (i.e., wear arcing, etc.) that are
now used infrequently.
r '
Vibrator cam contacts (i.e., wear, arcing, etc.) that are no longer
used.
Rapper assembly.(i.e., loose bolts, ground wires, water in air lines,
solenoids, etc.).
Vibrator and rapper seals (i.e., air inleakage, wear, deterioration).
T-R set controllers (i.e., low-voltage trip point, over-current trip
point, spark rate, etc.).
Vibrator air pressure settings.
Figure 5-3. Items that the ESP coordinator should record quarterly.
SECTION 5-O&M PRACTICES
5-15
-------
5.2.5 Annual Inspection -(Outage) and_Maint_enance
The process and its ESP generally should be shut down at least once a
year for a more complete Inspection, including a check of internal condi-
tions. The design of some ESP's is such that they may be isolated or by-
passed without a process shutdown. In other situations, however, outages may
occur on a more frequent basis. In all cases, however, adherence to estab-
lished safety and confined-area entry procedures cannot be overemphasized.
Safety interlocks should never be bypassed to enter the ESP.
Before anyone enters an ESP, an air-load check of each field is recom-
mended. This serves as a record for comparison when the ESP maintenance is
complete and all scheduled maintenance has been performed. When these air-
load tests have been completed, the inspection is ready to begin (Note:
gas-load tests may be used just prior to shutdown). During the inspection
more attention may be focused on certain selected areas where readings are
abnormal or unusual.
With the key interlock system, the actual opening of the ESP can be a
relatively time-consuming operation; however, this system provides a method
of locking out and grounding the power supply to prevent accidental energiza-
tion while personnel are inside. In older ESP's, hopper systems were gener-
ally excluded from the key interlock system, but most new systems include the
hopper access doors in the system.
When the key interlock system procedures have been completed and the
doors opened, the grounding straps should be attached to a wire inside the
access door. This establishes a positive ground to bleed away any voltage
retained by the plates (the ESP behaves like a large capacitor) and to pre-
vent energization of a field if, for some unforeseen reason, the interlock
system should fail. These grounding straps are recommended for each field in
the ESP.
Before anyone enters the unit, the confined work area should be sampled
and evaluated for oxygen deficiencies and the presence of toxic substances
and combustible materials. It is generally recommended that the ESP be
cooled and purged prior to entry; however, with proper safety equipment and
precautions, inspections can be performed without this extra step. Routine
SECTION 5-O&M PRACTICES
5-16
-------
monitoring of internal conditions should be part of the safety plan, but
continuous personal monitors with alarms are preferred. The point of initial
entry into the ESP will depend on access and maintenance considerations. For
the purpose of this section, entry is assumed to be through the side of the
ESP at the bottom of the plates.
The initial inspection of a collecting and discharge electrode system
should be used to observe several items. The first is whether there is a
buildup of material on the surfaces. Generally, 1/8 to 1/4 inch of material
will remain on the plate even in its clean, "rapped-down" condition. Too
much buildup can lead to both energization and gas distribution problems
within the ESP and is indicative of poor or ineffective rapping. Clean metal
conditions may indicate low resistivity, high gas velocity, or too much
rapping. Buildup on wires should be minimal.
The nature, extent, and locations of any buildups within the ESP should
be noted. While checking for buildups, the alignment of the wires and plates
also should be checked. Any bowing or skewing of the alignment of more than
+_ 1/2 inch is usually visible, and its location should be noted for correc-
tive action. This check is conducted for each lane of each field within the
ESP. A misaligned ESP usually causes reduced voltage and power input and
increased sparking.
Checks for broken wires are usually conducted in fields where short-
circuits are found. The broken wires should be clipped and removed, as
should the bottle weight. The location of the broken wire and where the wire
failed should be recorded as part of the permanent record. Generally, wires
are not replaced individually as they fail because the performance of the ESP
does not suffer greatly if several wires are missing as long as the wire
failures are random. On the other hand, failure in the same location gener-
ally indicates a problem. Wires also should be the proper length. It is not
unusual for individuals to cross wires to shorten them so they will fit
within the upper and lower guide frame; however, the rubbing of the crossed
wires during operation will cause wire failure.
The upper and lower discharge guide frame assembly should be aligned so
that equal spacing is maintained, not only from the top to the bottom of the
SECTION 5-0&M PRACTICES
5-17
-------
plate, but also from the leading edge to the trailing edge of the plates.
The frames should be level in both parallel and perpendicular planes to the
gas flow. Frames that are not level or twisted frames may cause excessive
tension on some wires and insufficient tension on others (slack wires). When
checking the upper discharge frame, the upper support beams should be checked
for excessive dust buildup. These buildups will sometimes cause intermittent
high sparking in the upper section of the ESP, which reduces collection effi-
ciency. Additional baffles may be necessary to prevent such dust buildups.
All insulators should be checked and cleaned to remove dust accumula-
tion. They should also be checked for any evidence of insulator tracking.
The inspection should include both the inside of the large support bushing
insulators at the top of the ESP and the discharge rapper insulators. Any
insulators that are broken, chipped, cracked, or glaze-damaged should be
removed and replaced.
Hoppers should be emptied during shutdown. Some buildup may be found in
the corners or in the upper portion where the hopper joins the ESP housing,
and these buildups should be removed. In flat-bottom ESP's in recovery
boiler applications, the space between agitators should be checked and any
' buildups should be removed. • All hopper level detectors should be checked and
repaired as necessary, and defective hopper heaters should be replaced.
Since nuclear hopper level detectors can expose maintenance personnel to
radiation, the built-in shields should be put in place before any personnel
enter the hopper. All dust discharge valves should be checked and cleaned
and repaired as necessary. These valves should be maintained to prevent any
inleekage of air into the hopper, which can aggravate hopper pluggage and
cause dust reentrainment.
The interior of the ESP should be inspected for corrosion of the shell,
plates, and wires. Localized areas of corrosion may indicate points of in-
leakage, which cause temperatures to fall below the acid dewpoint. Sonic
testing can be utilized to determine the thickness of the shell. Besides
reducing the strength of materials, corrosion can cause scaling of the metal
components, which interferes with the clearances within the ESP and reduces
performance. Corrosion and excessive dust in the penthouse or insulator
SECTION 5-O&M PRACTICES
5-18
-------
housing may indicate that Insufficient purge air is being-supplied to the
insulator housing, which will cause condensation of flue gas in this area.
Rapper rod connections or anvils for the discharge system and the plates
should be checked. Loose, broken, or bent connections should be repaired to
allow rapping force to be transferred efficiently into the ESP. Failure of
the rapper rods may not be evident from dust buildup or. the plates and wires
if sufficient force is still being applied.
The T-R's should be checked and cleaned during shutdown. All contacts
should be removed, cleaned, and adjusted. All electrical connections should
be checked for proper tightness. Loose connections and dirty contacts can
cause electrical erosion and wasted energy, and may lead to failure of the
T-R. The high voltage line, bushings, and insulators should be checked and
cleaned. Surge arresters should be checked and replaced if necessary. The
high-voltage bus duct should be checked for dust buildup and corrosion, which
could lead to grounding of the transformer secondary. All leaks in the bus
duct should be repaired during the reassembly to keep moisture and inleakage -
to a minimum. All insulators contained within the bus duct should be checked
and cleaned or replaced as necessary. All insulators which are chipped,
cracked, or damaged with glaze due to electrical tracking should be replaced.
Tightness of the connections for the high-voltage bus duct should be checked.
Transformer switchgear should be cleaned and adjusted for proper contact.
Some general areas requiring attention during the annual outage include
a check of door gaskets for proper seal. It may be worthwhile to replace
these gaskets annually or biannually to minimize inleakage. Another area is
the inlet and outlet distribution system, which should be checked for plug-
gage and dust buildup. This includes the inlet and outlet ductwork. Build-
ups that result in poor gas and dust distribution within the ESP should be
removed. If the buildups are substantial and recurrent, some modification
may be needed to minimize them. Expansion joint seals should be checked for
integrity and replaced if necessary. Failure of expansion joints can lead to,
excessive inleakage and increased corrosion. Lastly, water washdown of the
ESP is generally not recommended unless a dust is present that will adversely
affect ESP performance after startup and there is no other way of removing
SECTION 5-O&M PRACTICES g -
-------
the material from the plates. Water-washing an ESP can accelerate corrosion
and cause rust and scaling, which interfere with the electrical performance
of the ESP; therefore, the ESP must be dry before operations resume. Figure
5-4 summarizes the items that the operator or ESP coordinator should check
during the annual outage inspection.
At the completion of the outage, all personnel, tools, and other materi-
als used inside the ESP should be accounted for. A final safety check should
be completed for each section of the ESP to determine that all personnel have
exited before the ESP is closed up. An air-load test of each field should
then be performed. This final air-load test will indicate whether the sched-
uled maintenance was in fact completed. The air-load test will also detect
any mistakes or forgotten items and will serve as a record or certification
of readiness for operation. This air-load test should also become part of
the permanent records kept by the plant maintenance personnel.
SECTION 5-OSM PRACTICES
5-20
-------
ANNUAL INSPECTION CHECKLIST
Transformer Enclosure
HV line, insulators, bushings, and terminals
Electrical connections
Broken surge arrestors
High-Voltage Bus Duct
Corrosion of duct
Wall and post insulators
Electrical connections
Penthouse, Rappers, Vibrators
Upper rapper rod alignment
Rapper rod insulators
Ash accumulation
Insulator clamps
Lower rapper rod alignment
Support insulator heaters
Dust in penthouse area
Corrosion in penthouse area
Water inleakage
HV connections
HV support insulators
Rapper rod insulator alignment
Collecting Surface Anvil Beam
Hanger rods
Ash buildup
Weld between anvil beam and lower rapper rod
Upper Discharge Electrode Frame Assembly
Welds between hanger pipe and hanger frame
Discharge frame support bolts
Support beam welds
Upper frame level ness and alignment to gas stream
Lower Discharge Electrode Frame Assembly
Weight guide rings
Level ness of frame
Distortion of the frame
Figure 5-4, Items that the operator or ESP coordinator
should check annually, (continued)
SECTION 5-OSM PRACTICES
-------
Stabilization Insulators
Dust buildup and electrical tracking
Broken insulators
Collecting Electrodes
Dust deposits; location and amount
Plate alignment
Plate plumbness
Plate warpage
Discharge Electrode Assembly
Location of dust buildup and amount
Broken wires
Wire alignment
Weight alignment and movement
Hoppers
Dust buildup
Level detectors
Heaters
Vibrators
Chain wear, tightness, and alignment
Dust buildup in corners and walls
Dust Discharge System
Condition of valves, air locks, conveyors
General
Corrosion
Interlocks
Ground system
Turning vanes, distribution plates, and ductwork
Figure 5-4 (continued)
SECTION 5-O&M PRACTICES
5-22
-------
SECTION 6
INSPECTION METHODS AND PROCEDURES
This section presents detailed techniques and procedures for conducting
the following inspections of ESP systems and associated components; precon-
struction and construction inspections; external inspections; internal in-
spections; special inspections and observations that should be made with
respect to air-load and gas-load tests, baselining, and performance tests.
Safety considerations during inspections are presented, and the use of porta-
ble instruments and safety equipment is discussed.
The purpose of any ESP inspection is to determine the current operating
status and to detect deviations that may reduce performance or cause failure
at some future date. For this reason, inspection programs must be designed
to derive maximum benefit from the information gathered during the inspec-
tion.
A properly designed inspection program can be used for three purposes:
recordkeeping, preventive maintenance, and diagnostic analysis. Depending on
its purpose, the inspection may be conducted by operators, maintenance staff,
regulatory agency Inspectors, outside consultants, or vendor representa-
tives.
The external ESP inspection (on-line) usually is limited in scope to
critical components. Checks of the operating status of the unit are limited
to gas volume, general operating characteristics (moisture, temperature,
oxygen, SO-, etc.), rapper function, dust removal system function, insulator
temperature/heater function, electrical readings, and process conditions.
From this information, internal conditions and emission rates may be inferred
and/or calculated.
The internal inspection provides information on corrosion, alignment,
insulation condition, points of air inleakage, effectiveness of rapper func-
tion, and distribution plate pluggage.
SECTION 6-INSPECT1ON METHODS *ND PROCEDURES
6-1
-------
Taken together, internal and external inspections provide continuous in-
formation to aid in the operation and maintenance of the unit.
6.1 PRECONSTRUCTION AND CONSTRUCTION INSPECTIONS
Reconstruction inspections are necessary to ensure that the correct ESP
components and structural members are received and properly stored for subse-
quent assembly. Many of the later O&M problems with an ESP can be traced to
improper design, fabrication, and storage of critical components. This
section provides guidance for establishing a preconstruction/constructicn
acceptance program that will help to reduce subsequent maintenance problems
that are related to poor project planning, and management and unit construc-
tion. Table 6-1 presents an example of an ESP erection sequence. Each manu-
facturer has its his own sequence; this one is provided for illustrative
purposes only.
6.1.1 nans an d Spec if i cation s
Many maintenance problems are created during the critical period of
initial design. The special limitations and conditions placed on the unit by
the specific process application must be considered. Design of the shell
must meet site-specific constraints such as adjacent structures, crane movement,
overhead clearances, and stack conditions. In this early design stage, O&M
personnel must be involved to define projected O&M requirements. A thorough
review of each component, subassembly, and structure is critical. Many de-
signs that appear to be adequate on paper and to meet all safety requirements
cannot be properly maintained after they are built.
A list should be prepared of all known or expected maintenance (repair,
replacement, etc.) tasks that may be required on the unit. The design engi-
neers, construction supervisors, operators, and maintenance staff should then
review the proposed design to determine if these tasks can be completed
properly. Many problems and maintenance expenses can be avoided if minor
changes are made at this stage of the design.
When the basic design (shell size, orientation, location, ductwork,
etc.) has been specified, a complete laboratory flow model study should be
made to determine the acceptability of the general design. Detailed engineering
SECTION 6-INSPECTiON METHODS AND PROCEDURES c ~
t-d
-------
TABLE 6-1. ERECTION SEQUENCE
Construction phase
Tolerances
Comment
Foundations
Structural steel
Subassembly of
hoppers
Set hoppers
Assemble side panels
Install side panels
and base plates
Base girders
installation
Rod bracing and
guide wires
Roof girders
Weld-bolted
connections
Seal welding
Inlet and Outlet
nozzles and
perforated
plates
(continued)
Elevation and bolt
position
+ 1/16 in.
1/8 in. weld
1/8 in,
1/8 in.
individual
diagonally
Level, plumb
dimensions.
to + 1/16 in,
and diagonal
Shim supports
elevation.
Temporary support
Flat work table for subsec-
tion welding horizontally
on ground
Alignment-check diagonals,
center!irie position
Girders parallel to each
other, right angles to
shell» top/bottom flanges
level, and same elevation
Tack weld shell and bolts/
nuts. Exterior bolts are to
be seal welded
All shell seams
Assemble perforated plate,
install on outlet nozzle,
and seal-weld nozzle to
shell. Assemble perforated.
plate, install on inlet noz-
zle, and seal-weld nozzle to
shell
SECTION 6-INSPECT1ON METHODS AND PROCEDURES
6-3
-------
TABLE 6-1 (continued)
Construction phase
Tolerances
Comment
Lift hoppers into
position
Assembly and
installation of
collecting plates
Plate warpage
+ 1/16 in. overall
length
Vertical flashing
Plate rapper bars,
anvils, and stop
(internal hammer
design only)
Electrode support
frame stabil-
izers
Electrode support
frame
Diagonals
+ 1/16 in,
Roof panels
Support insulators
Bolt to structural steel,
seal weld to side panels,
and base girders
Establish ESP center!ine.
Tack-weld captive spacer bar
on roof girder lower flange.
Lift plate cradle into field
and secure. Move each plate
into position using field
trolley, Plate section sub-
assembly anvils, stop plates,
connecting links, etc.
Align collection plates
Seal-weld between outside
plate and shell in each
field (leading and trailing
edges)
Align rapper hammers and
anvils for center strike.
Straighten on ground before
installation
Assemble on ground. Install
hammer assembly, anvils.
Temporarily install elec-
trode frame. Hang elec-
trodes. Straighten
electrodes to j;l/16 in.
Align and seal weld
Install insulator support
gasket. Install insulator
support. Install insulator.
Install Insulator cap.
Mount support springs
(continued)
SECTION 6-INSPECTION METHODS AND PROCEDURES
6-4
-------
TABLE 6-1 (continued!
Construction phase
Tolerances
Comment
Alignment of elec-
trode support
frame
Alignment of stabi-
lizers and
electrodes
Insulator housing
Plate rapper
Power supply and
controls
Insulation
Safety interlocks
Conveyors, dump
valves, air
locks
Miscellaneous
equipment
Lift electrode frame from
temporary support. Align
electrode support frame to
center line of collection
surfaces. Level support
frame. Tack-weld insulator
support to roof
Mount electrode frame stabi-
lizers. Align upper and
lower frames and align elec-
trodes +1/16 in. Hang elec-
trode weights. Recheck
electrode alignment
Assemble and install insula-
tor housing to roof. Seal-
weld. Install rapper
assembly. Install insula-
tor heaters
Install rappers and/or rap-
per drives. Position and
rotate rapper hammer shaft
for proper strike
Install T/R. Install bus
duct, insulators, bus bars
SECTION 6-1NSPECTION METHODS AND PROCEDURES
6-5
-------
design and components can then be specified. The flow distribution is criti-
cal to achieving the design emission levels and must conform at least to IGCI
specifications. Model studies only define the potential (optimum) flow pat-
tern in the unit and not necessarily the operating flow patterns under dust
conditions. Thus, several studies may be performed (with artificial deposits
on turning vanes, distribution plates, or other surfaces) to determine the
limits of a unit. The use of visual indicators (such as smoke) can indicate
areas of turbulence or material deposition. Elimination of these areas can
reduce maintenance problems in the full-scale unit and improve the overall
on-line performance.
6.1.2 Specification of Materials
The vendor or manufacturer usually specifies the materials of construc-
tion as part of his bid submittal. In many cases, however, the purchaser,
based on in-plant experience, may specify in detail the components, materials,
or construction practices. Most specifications must be accepted as complying
with national ratings or codes. Deviation from accepted normal practice must
be clearly specified. Typical sources of ratings and codes are as follows:
Federal Specifications
Naval Facilities Engineering Command (NAVFAC)
U.S. Environmental Protection Agency (EPA)
American Iron and Steel Institute (AISI)
American Society for Testing and Materials (ASTH)
Industrial Gas Cleaning Institute (IGCI)
National Electrical Manufacturers Association (NEMA)
Occupational Safety and Health Administration (OSHA)
Sheet Metal and Air Conditioning Contractors National
Association (SMACNA)
American Institute of Steel Construction (AISC)
International Electro Technical Commission (IEC)
Nuclear Regulatory Commission (NRC)
6.1.3 Selection of a Contractor
The selection of a general contractor to erect the ESP is very important.
Because of the size of the job, bid requests may yield a large number of sub-
mittals from general contractors that have little or no experience in con-
structing ESP's. General contractors who are well qualified to complete
major projects such as bridges, buildings, etc., may not be able to assemble
an ESP to the required tolerances. Qualifications of potential bidders
SfCTION 6-INSPECTION METHODS AND PROCEDURES cc
-------
should be reviewed and previous jobs should be inspected for subsequent qual-
ity control and maintenance problems related to the original construction.
It is generally more efficient to use the manufacturer or vendor as the
general contractor because a single firm then has direct responsibility for
the manufacture, delivery, storage, and erection of the unit; however, the
use of others may not necessarily result in an inferior product.
6.1.4 Scheduling
Scheduling is necessary to maintain continuous and steady progress
toward completion of the construction. The scheduling of major milestones is
completed during the conception and design phases of the project. One area
that is not always adequately addressed is the scheduling and receiving of
material. The potential problems created by critical components arriving too
early in the construction process can be greater than those resulting from
delayed shipments, Hany ESP components are sensitive to storage conditions,
and long storage can severely damage them and cause failure during operation.
The time must be projected for receipt of parts and storage options must be
considered. An assessment of penalties may be considered for late delivery
and also for early delivery that entails special handling and storage.
Close coordination between the plant and the vendor or manufacturer of
the components is necessary to maintain projected construction schedules.
6.1.5 Inspection of Manufacturer'sFacilities
The purchaser may reserve the option of inspecting the manufacturer's
fabrication facilities prior to bid selection or after the contract award to
ensure proper quality control during component fabrication. Most manufactur-
ers will welcome a review of their quality control and production procedures.
This may enable defects and/or fabrication errors to be discovered prior to
shipment. Methods of crating, shipping, and special handling also may be
addressed that will prevent damage in transit.
6.1.6 Quality Assurance/Qualjty Control
A quality assurance/quality control (QA/QC) program must be established
for evaluation of each phase of the construction to certify that all compo-
nent subassemblies and systems are manufactured and installed so as to meet
orexceed designspecifications. The program can be divided intotwo major
SECTION 6-INSPECTION METHODS AND PROCEDURES £ -,
o-/
-------
elements: 1) inspection and acceptance of materials, and 2) quality control
of construction and erection.
Inspection and acceptance of materials is a continuous process involving
verification of the quality of the part or component at various stages of
construction. During the design phase, acceptance is based on the estab-
lished specifications; at this point, quality assurance and reliability of
the product are determined, and changes are made as necessary.
The second phase involves the review and acceptance of the manufactur-
er's quality control program to prevent the production and shipment of defec-
tive components. The purchaser may require special tests or quality control
checks, In the third phase, the purchaser or the general contractor must
have a quality acceptance program for material received at the job site.
This inspection is very important because it verifies receipt of the proper
component and certifies that no in-transit damage has occurred. Acceptance
of materials at this point releases the manufacturer from all claims except
hidden damage or in-service failure (design flaw). Acceptance tests (non-
destructive or destructive) may be performed on a sample of components,
depending on past history of problems with this component.
Because components may be damaged during storage, a quality control
check should be made of material as it is removed from storage, prior to its
installation. No damaged material should be installed in the ESP.
The final phase of the QA/QC program involves quality control of con-
struction practices and installation. Quality control checks should be com-
pleted at each major step in construction before further progress is allowed.
Such checks prevent hidden defects or substandard construction that may
require major renovation or demolition at a later date. The program may
include alignment or clearance checks, welds, surface preparation, and mate-
rials of construction.
6.1.7 Onsite Material Storage
Quality control procedures for onsite storage of components requires
special consideration. Plates, wire frames, electrodes, and insulators are
critical components in the performance and reliability of the ESP. Improper
onsite storage of these components can result in chronic maintenance problems
after installation.
SECTION 6-INSPECTION METHODS AND PROCEDURES
6-8
-------
The manufacturer fabricates plates and plate panel components (stiffen-
ers, alignment rakes, etc.) to precision tolerances, and these components
must be assembled in the final unit to a clearance of +1/4 in, (discharge
electrode to plate). If these items are stored flat, without adequate sup-
port, stress may cause alignment problems with the plates when they are
placed in the heated gas stream (bows, bends, twist, etc.). Plates should be
stored on edge and adequately supported. Wire frames also can be distorted
as a result of improper storage, which prevents their alignment in the gas
lane. Pipe frames (masts) are composed of hollow pipes, and in winter
months, water may freeze in the inner surfaces and cause swelling and/or
rupture of the pipes.
All components must be protected from weather to prevent corrosion if
extended storage (greater than 1 yr) is required. Proper scheduling of the
receipt of components will generally limit exposure to less than 60 days.
Structural components (beams, plate, etc.) usually will be painted with a
primer and will not require special treatment. Plates and wires are general-
ly received unpainted, but surface corrosion is not a problem during short
exposure.
Electrical and mechanical components (such as rappers,-insulators, T/R
sets, and control cabinets) should be stored in climate-controlled areas and
protected from dust and physical damage.
6.1.8 Preoperational Inspection and Testing
Acceptance of the ESP after installation involves a series of predeter-
mined steps including preoperational planning, acceptance testing, system
transfer, and project completion.
Preoperational planning involves the transfer of technical information
from the vendor and general contractor to plant operators 'and maintenance
staff. This can include the development of operation and maintenance manu-
als, a work order system, recordkeeping requirements, and troubleshooting
procedures. It can also include both internal and vendor training of the
staff at a time when the components of the system are new, clean, and cold.
A schedule and procedure should be set up to verify the "as installed"
versus the "as specified" system with regard to system components, materials
of construction, etc. Checklists of step-by-step evaluation procedures
SECTION 6-INSPiCTION MiTMODS AND PROCEDURES
6-9
-------
should be prepared and the responsibilities of each person involved in the
evaluation should be spelled out.
Several preoperational tests also can be conducted before the ESP system
is transferred to operating personnel. The following are examples:
0 Function_a1_ tests --Functional tests may be performed on an individual
component to determine initial acceptance before a full gas-load evalua-
tion. Functional tests include continuity checks of electrical con-
trols, alarms, and motor circuits; verification of the rotation of
motors, fans, and screw conveyors; end checks on the operation of major
systems such as rappers and ash handling systems.
0 Tests for air inleakage—The ESP shell should be checked for potential
aTr inleakage at welds, access doors, and ash valves. Welds may be
checked visually with vacuum boxes. All points of air inleakage should
be repaired before further acceptance tests are made,
° Tests for electrical continuity and grounds—Physical inspection of the
electrical distribution system should be made, and continuity of elec-
trical connections should be verified. The electrical system also
should be inspected to determine the possible presence of unintentional
grounds. After the physical check, the electrical bus should be discon-
nected and each field should be meggered. A 1000 volt megger should be
used and the resistance should be between 100 megs and infinity.
0 Alignment tests—An internal inspection of the wire and plate alignment
should be conducted to determine proper clearances. Any point of close
clearance, warped plates, or poor fabrication should be corrected before
air-load tests are conducted.
0 Gas distribution — If any possibility of nonoptimum gas flow distribution
exists, hot-wire anemometer tests should be conducted. Fans should be
dampered, and total gas volume and temperature for the test (Reynolds
Number) should be defined. Velocity profiles should be produced at pre-
determined vertical locations and horizontal positions. The nonoptimum
distribution may be downstream of the inlet distribution devices (be-
tween lanes) or at the outlet upstream of the last distribution devices.
The average velocity (in feet per minute) and deviation should be calcu-
lated and compared against. I6CI acceptance values as a minimum. Many
users need and have much tighter requirements for gas flow distribution.
Visual evaluation also may be made by using smoke systems and internal
lighting. Any deviation from predicted performance should be corrected
prior to startup.
0 Ayr-load tests--The ESP should be closed, and key interlock procedures
should be followed. Each T/R set should be energized, and after the
control system function has been verified, the power should be increased
to the current limit or flash-over (spark) point. The secondary volt-
age/current relationship should be plotted and the corona initiation and
SECTION 8-INSPECTION METHODS AND PROCEDURES
6-10
-------
spark points noted. Electrical clearances can be judged, based on com-
parison between fields. If any field deviates from predicted (design)
values by _+10 percent, the alignment of the field should be reconfirmed,
6.1.9 System Transfer
An orderly transfer of the ESP's operation and maintenance should be
made from the vendor and/or general contractor to the plant personnel. This
transfer should occur over a short period of time, as operators and mainte-
nance staff become trained and certified.
6.1.10 Project Completion
An ESP project is completed in phases: construction, mechanical and
electrical check-out, startup, system operation, and acceptance tests.
Acceptance tests generally are defined as verification of ESP efficiency or
outlet mass loading under full load conditions. This test should be made in
accordance with established EPA test protocol so that the manufacturer's
guarantee acceptance test can be submitted to the appropriate regulatory
agencies as initial proof of compliance. To determine reliability of the
unit, a second acceptance test should be conducted after a period of sus-
tained service (i.e., 6 months to 1 year). During this period, any hidden or
previously undetermined defects may surface that limit the long-term removal
efficiency of the unit. It is also good practice to establish a means for
correcting from a given test condition to the contract test condition since
volume flow rates frequently differ from specifications. Plots of guaranteed
or outlet loading vs. volume flow rate are an acceptable method for this.
6.2 EXTERNAL INSPECTION
Because an external inspection of an ESP is generally conducted when the
unit is in operation, it cannot provide direct information on the internal
condition of the ESP. Internal conditions of the unit are reflected, howev-
er, in the electrical readings in each field. The external inspection in-
volves checking major system components and determining their impact on
electrical readings and indirectly on ESP collection efficiency. Components
that are inspected include rappers and rapper controls, penthouse purge-air
SECTION 6-INSPECTION METHODS AND PROCEDURES
6-11
-------
and heater systems, shell condition, ash-handling system, and flue gas char-
acteristics. The following sections describe the procedures that may be used
to evaluate the status of an ESP by inspecting these components. Because
each vendor's equipment design is different, several of these items may not
be directly applicable to all units.
6.2.1 Rappers and Rapper Control
The purpose of rappers is to remove dust that has been collected on the
ESP plates, gas turning vanes, distribution plates, and discharge electrodes.
If dust is not removed from the plates, the power input to the individual
plate (electrical field) sections may be reduced. As dust depth increases,
the voltage drop across the dust layer increases, which requires an increase
in the T/R set secondary voltage to maintain a constant secondary current.
Dust buildup on wires increases the effective wire diameter and corona initi-
ation voltage. Heavy accumulation can completely quench the corona and/or
cause premature sparkover as clearances are reduced. Buildup of dust on the
turning vanes and perforated plates changes the gas distribution and velocity
vector entering the ESP. Gas-flow models and efficiency calculations per-
formed as part of the unit design assume that these surfaces are reasonably
clean and the gas distribution is correct.
The level of dust collected on each surface is a function of dust load-
ing, dust characteristics, gas characteristics, and unit design. Require-
ments for intensity and frequency of rapping may be unique for each unit and
may change with process conditions and the age of the unit. It is important
that each rapper function at design rapping intensity.
The four general types of rappers are pneumatic-impact, magnetic-impulse
gravity-impact (MIGJ), electric or pneumatic vibrators, and falling hammers.
Each provides a different level and frequency of vibration to the surface.
The application of each is based or, the vendor's preference and the ESP
application. Pneumatic rappers are generally used on plates and wires in the
kraft pulp and cement industries, and MIGI rappers are generally used on
plates and wires in industrial and utility boiler applications. Vibrators
are also used on wires in the cement industry, and falling hammers are used
for internal and external rapping in almost all applications.
SECTION 6-INSPECtlON METHODS AND PROCEDURES
6-12
-------
Rappers that are external to the ESP shell are easier to evaluate and
maintain than are internal rappers. The first step in the inspection of
external rappers is to verify that all rappers are functioning. This check
requires the application of either electrical impulse or air pressure to each
rapper to determine if it is operating. In the case of vibrators or pneumat-
ic rappers, the inspector should note the sound of the rapper during its op-
eration. Weak activity indicates needed repair. Manual triggers, solenoids,
or electric relays may be installed to aid in the evaluation.
For proper operation of pneumatic rappers, the air must be dry and
clean. When condensed water mixes with lubricating oils that are placed in
the air lines for rapper lubrication, an emulsion forms that can combine with
rust and cause seizure of the rapper piston. Air for the rappers should be
equivalent to instrument air (dew point - 40°F). Air dryers, traps, and
filters should be used as necessary to ensure proper operation of the rapper
(see Figure 6-1). If air lines have been subjected to high levels of moisture
and air quality is later improved, a secondary problem may occur; rust may
flake off the line walls and foul the solenoid valves. It may be necessary
to place filters prior to the solenoids to remove particles.
The fact that.the rapper is functioning does not necessarily mean the
rapping energy is being transferred to the collecting surface. Each rapper
should be checked while it is operating to determine if it may have become
disconnected from the rapper shaft and/or insulator.
Alignment of rapper shafts and rappers should be noted. Any movement in
plate and wire frame alignment can cause misalignment of the rapper shaft and
roof opening, with the possible result of breakece of insulators, rapper
shafts, and/or anvils.
If the ESP does not have a penthouse and the roof is constructed of
concrete, binding of the rapper shaft can cause structural failure. Dust
buildup in the rapper top opening (nipple) can harden and cause the rapper to
seize. Rapping under these conditions imparts rapping energy to the roof
structure and can cause cracking. Each rapper should be moved laterally
during the inspection to determine if the shaft is free in the opening.
Rubber boots around the rapper shafts should be inspected to determine
if they are airtight. The rubber should be flexible and free of cracks (see
SECTION 6-INSPECT1ON METHODS AND PROCEDURES
6-13
-------
Figure 6-1. Rapper air line trap and filter.
6-14
-------
Figure 6-2). In many cases, the boot may be installed with the rapper shaft
at one elevation, and over a period of time the shaft may shorten (as a
result of wear, roof distortion, etc.) and cause the boot to fold. The folds
hold water, which increases the rate of failure and inleakage. Exposure to
high temperatures also increases the failure rate. It should be noted that
Hypalorr^and VitorKmaterials are now available to withstand high tempera-
tyres and chemical attack.
Manual activation of rappers does not ensure their proper automatic
operation because the automatic controls may not be functioning. Periodically,
the system should be checked completely to evaluate the control system.
Because rapper operation may be less frequent in the outlet fields and the
total number of rappers on the ESP may be excessive, the inspector cannot
wait for each rapper to activate in the automatic mode. In this case,
inspectors can use indirect methods, such as placing coins, washers, or other
items on the rapper surface, to check rapper operation. If the rapper is
activated, these items will fall off; when the inspector returns after a
rapper cycle, it will be readily apparent which rappers did not activate.
Depending on the complexity of the control circuits, rapper,function in the
automatic mode may also be confirmed by decreasing rapper frequency on the
control board and pacing the rappers as they activate.
Evaluating fall ing-hammer rappers during ESP operation is difficult
because the only visible moving component is the rapper shaft and drive.
First, the inspector should note that the shaft is turning and estimate the
approximate rate of rotation. If the number of hammers attached to the shaft
is known, an equal number of rapping strikes should occur during each rota-
tion. Because these strikes cannot be observed externally, indirect methods
must be used. One effective method is to place a stethoscope against the ESP
shell in the area of the rapper shaft. Each strike can be heard, and if the
numeric location of the rapping sequence is known, the approximate location
of poor or missing anvils can be identified. Common failures of these rap-
pers involve offcenter impact on the anvils, slippage of hammers on the
shaft, and separation of the anvil shaft from the plate. An internal inspec-
tion is necessary for a complete evaluation of these rappers. Seals around
SECTION 6-INSPECTION METHODS *ND PROCEDURES
6-15
-------
Figure 6-2, Rapper boots.
6-16
-------
the rapper shaft should be checked to determine if air inleakage is occur-
ring,
6.2,2 Penthouse Purge F_ans_and Air Heaters
Depending on the vendor, ESP purge fans may be located in a penthouse or
in an insulator enclosure (see Figure 6-3). These fans are designed to
provide a constant flow of clean, dry air across the insulator surface where
the wire frame support shafts (rods) penetrate the shell to prevent dust
contained in the flue gases from being deposited on the insulators and caus-
ing electrical tracking. Generally, only a small amount of purge air is
required, i.e., just enough to pressurize the penthouse or enclosure above
the pressure in the gas treatment zone.
The inspector should verify that the purge fans are functioning. The
amount of air being provided is not usually determined, but the effectiveness
can be evaluated during an internal inspection of the penthouse or insulator
enclosure. Dust accumulation may indicate poor purge rates. Filters should
be used on fen intakes where foreign matter is present to prevent dust frorc
being blown into the penthouse, and these filters should be changed periodi-
cally to prevent reduction in purge volume. Ammeters may serve as an indi-
rect indicator of fan volume and when it is time to change the filters.
Purge-air heaters may or may not be required on individual units, de-
pending or- operating temperatures and design. The purpose of the heaters is
to prevent local quenching of hot flue gas as cold outside purge air is
introduced. This quenching can result in condensation of water or other
condensible material in the flue gases in the insulator or under the ESP
roof. The extent of the condensation is site-specific and depends on the
moisture and temperature of the flue gas, the moisture and temperature of the
outside air, the temperature in the penthouse (as a result of heat loss
through the roof), and the rate of penthouse purge. If condensation tends to
form in the insulator (as indicated by hardened dust deposits, corrosion in
the inner roof, or condensation in the penthouse), heaters should be used.
Heating purge air is expensive because 5.5 kW is required for each
insulator. Contact heaters (which only require 0.4 kW/insulator) mounted on
the insulator are now in common use on many dry dust ESP's, The most criti-
cal period for penthouse heating is during startup, when metal surfaces are
SECTION 6-INSPECT1ON METHODS AND PROCEDURES
6-17
-------
Figure 6-3, Penthouse heater and fan.
6-18
-------
cold and any additional heat loss may result in condensation. The inspector
should verify heater function by using permanent ammeters or portable hand-
held amp probes for each purge-air unit,
6.2.3 Insulator Heaters
Many vendors provide direct insulator heaters to maintain wire frame
support insulators above air dew points. These heaters cannot be inspected
during ESP operation-, their operation can only be verified by measuring the
amount of current to the heater. Typically, these are resistance-type heat-
ers enclosed in a ceramic body. The inspector should determine the status of
the heaters by using permanently installed ammeters or portable amp probes.
The units may or may not be thermostatically controlled. The inspector
should define current levels by using a normal baseline; any deviation indi-
cates low heat rates, short circuits, or open circuits in the system.
As with purge-air heaters, the proper functioning of the insulator
heaters may not be critical for an individual unit, but it is critical during
startup to protect the insulators from condensation and electrical tracking.
6.2.4 ESP Shell and Doors
For proper ESP operation, the shell must be as airtight as possible.
Any inleakage of ambient air into the enclosure produces local charges in the
gas stream characteristics (temperature, moisture, oxygen, etc.). Depending
on the location of the inleakage, the electrical field may flash-over (spark)
at a lower voltage and current or cause dust buildup on the plates and wires.
If the inleakage is in the hopper area, collected dust may become reentrained
in the gas stream and lower the unit's efficiency. Severe gas inleakage (in
ducts, doors, roof, etc.) can generally suppress the flue gas temperature,
cause corrosion, and increase the gas volume to be treated (i.e., reduce
treatment time, increase the superficial velocity, and decrease the SCA).
The inspector should audibly check doors, hatches, flanges, rapper shaft
boots, etc., to determine if there is an inrush of air. Rapper boot corrosion
is shown in Figure 6-4, On cooler surfaces, air inleakage may also be deter-
mined by touch. Smoke generators also may be used as indicators of poor
sealing. Air inleakage is a constant problem that must be judged by a number
of indirect indicators, including inspection of the following: door gaskets,
SICTlON 6-INSPECTION METHODS AND PROCEDURES
6-19
-------
Figure 6-4. Example of rapper boot corrosion.
6-20
-------
shaft seal and boots, temperature drop from inlet to outlet of the collector,
increase in flue gas oxygen from inlet to outlet of the collector, and patterns
in the electrical power readings. Each point of inleakage should be noted,
and possible corrosion should be checked during an internal inspection.
In many systems, shell corrosion occurs from the inside of the unit
-(moist and corrosive gas stream) and does not become visible from the outside
until complete failure has occurred. Dust deposits may hide this type of
corrosion during internal inspection, or it may occur in an inaccessible
area. The inspector should be careful to note any areas that show distor-
tion, abnormal flexing, or any changes in appearance that could indicate thin
metal. Areas of most concern are the roof, the areas around door or hatches,
isolation plates (guillotines), and expansion joints.
6.2,5 Electrical Conduit Ducts
Wire or metal pipe conductors transfer the electrical power (DC) from
the T-R set to the discharge wire frame. This power has a high negative
potential (20 to 55 kV) and cannot be insulated by conventional methods. The
conductor is isolated from personnel by an enclosure (duct) and an air gap
between the conductor and duct to prevent electrical arcing. The conductor
is supported by post insulators located in the duct and on the T-R set.
Changes in the air in the duct due to moisture or condensation on the
insulators can cause arcing of the high voltage to ground. These arcs are
audible when the arc strikes the duct wall, and electrical meters fluctuate
wildly in the field. If they are severe, continuous, and stronger than the
control circuits are designed to handle (higher current than sparking), these
arcs can cause overheating of the transformer and linear reactor.
The inspector should take note of any indication of arcing in the duct
(and on meters) and detenrnne if the arc may have been caused by air inleak-
age into the duct, water penetration, or some external influence. An example
of conduit corrosion is shown in Figure 6-5. Heavy, uncontrolled arcing may
necessitate isolating the field to prevent permanent damage to the T-R and
control circuits.
If external causes cannot be determined, an internal inspection may be
necessary during an outage. Arcing can be caused by dust buildup on post
SECTION 0-lNSPECTlON METHODS AND PROCEDURES
6-21
-------
Figure 6-5. Electrical conduit corrosion.
5-22
-------
insulators, condensation, or reduced clearances resulting from movement of
the conductor.
Depending on vendor preference and equipment location, the conduit may
(or may not) be vented externally or with the purge-air system. If the
system is purged, vent openings should be checked to ensure proper air flow,
6.2.6 Control Cabinets
The area in which the T-R set control cabinets are located should be
clean and supplied with cool air. Transformers and resistors in the cabinets
generate heat, which must be removed through finned surfaces. Generally, a
water chiller or an air conditioning system is provided in the area to remove
heat and maintain a moderate cabinet temperature (less than 90*F). Failure
to remove heat can result in failure of solid-state printed-circuit control
boards.
Dust in cabinets also can cause electrical short circuits and increase
heat retention. If independent cabinet ventilation is used, each cabinet
should be equipped with air filters; if independent cabinet ventilation is
not used, the room air should be filtered.
The inspector should visually check the condition of the filters to
determine if dust has accumulated on electrical components. Under no circum-
stances should the inspector penetrate the cabinet while it is energized
because it presents a potential shock hazard.
The inspector should note any evidence of corrosion in electrical con-
tacts, relays, and circuit boards. Deterioration may be expected in areas
where hydrogen sulfide or sulfur dioxide is present. Severe corrosion can
cause loss of the electrical conductor or components in the printed circuit
boards. If the problem is chronic, room filters (activated carbon) may be
required.
6.2.7 Key Interlock System
The inspector should visually inspect the key interlock system to deter-
mine that all-weather enclosures are in place and that interlocks have not
been removed. The operation of the individual locks should be verified
during entry procedures associated with an internal inspection of the unit.
SECf ION 6-INSPECTION METHODS AND PROCEDURES
6-23
-------
Most interlock hardware is manufactured from brass alloys, and a certain
amount of tarnish and surface corrosion is expected in an industrial environ-
ment. Internal corrosion of locks and tumblers can produce operating prob-
lems and replacement may be necessary. A dry lubricant (graphite) should be
used on the interlock in lieu of oils or greases because the latter will hold
the dirt,
6.2.8 Ash-Handling System
The ash-handling system is the final discharge point for collected dust.
Depending on application, system components can consist of hoppers, drag bot-
toms, and wet bottoms. Regardless of the system used, however, the collected
dust must be discharged at an average rate equal to that at which it is
collected. Short-term deviations are acceptable as long as the capacity of
the hopper to store the dust 1s not exceeded.
Generally, an inspection of the ash-handling system entails verifying
that no hoppers are plugged and that the level of collected dust has not
exceeded the high-level alarm point.
In drag-bottom ash-handling systems, an inspection is conducted to
verify that the drag scraper is in motion and that material is being properly
discharged to the screw or mixer. Because the drive motor can continue to
operate after the chain has broken or the drive shaft has sheared, an exter-
nal evaluation is not adequate. Current drawn by the drive motors can be
used to indicate heavy buildup or free motion, but this may not provide full
protection.
Positive-motion indicators have been developed that ere connected to
electrical counters outside the shell and sound an alarm if contact 1Vnot
completed within a specified period of time (i.e., motion each time a scraper
blade strikes the arm). Instruments of this type prevent major drag scraper
damage and the excessive downtime that can occur when chains break and drags
become misaligned.
Heavier dust discharge rates are required in the inlet field because of
the heavier collection there. Chamber-to-chamber variation also may result
from gas-flow imbalances and dust stratification. Adjustments of each hopper
evacuation rate and frequency must be based on site-specific factors. Hop-
pers with chronic pluggage problems may be influenced by rapid cooling from
SECTION fl-INSPECTION METHODS AND PROCEDURES
6-24
-------
prevailing winds or other thermal sources, and windscreens may be required to
reduce these thermal effects. The inspector also should verify the operation
of hopper heaters (if used), vibrators, and air stones.
Inspectors may use the vacuum chart on the suction withdrawal side of
the system to evaluate hopper evacuation times. A premature vacuum break may
be an indication of rat-holing or hopper bridging. The time required to emp-
ty specific hoppers (inlet/outlet fields) should be consistent under normal
operation (full load/half load, etc.). any deviation may indicate incomplete
removal or pluggage problems,
6.2,9 Gas Stream Conditi ons
Variations in the flue gas stream can cause ESP performance to deviate
seriously from design operation. Variables that must be considered are gas
temperature, moisture content, and volume. Some of these items cannot be
determined during a routine inspection, and a Method 5 stack test may be nec-
essary; however, others may be measured directly or inferred from surrogate
factors.
Temperature--
The inspector should record the flue gas temperature at the inlet and
outlet of the ESP and compare these values against normal operating ranges.
Deviations can cause reduced power (resistivity effects), premature flash-
over (sparking), corrosion, or changes in the flue gas volume treated. Both
positive and negative deviations are important.
The temperature across an ESP is rarely uniform; a variation of 20° to
100CF is not uncommon. Multiple thermocouple locations are desirable to
determine changes from normal operation.
If gas stream temperature is controlled by evaporative coolers that
humidify the gas stream (typically used in the cement industry), the tempera-
ture can be related to flue gas moisture. In these applications, the resis-
tivity of the dust layer is controlled by the water vapor. Changes in tem-
perature may be a significant indirect indicator of the flue gas moisture.
Temperature measurements should be correlated with inspection and monitoring
of the evaporative cooler.
SECTION f-INSPECTiON METHODS AND PROCEDURES
6-25
-------
Flue Gas Oxygen--
Flue gas oxygen by itself is not critical in ESP performance, but chang-
es in flue gas oxygen indicate a change in the process operation that the ESP
is serving or possible inleakage to the flue gas stream.
Kost combustion process flue gas streams contain a residual amount of
excess oxygen that can be limited to increase combustion efficiency. An
increase in excess air results in an increase in flue gas volume at the same
boiler load. If the boiler was designed for 10.5 percent (2.01 Q2) excess
air and the level of excess air were to increase to 50 percent (7.0» O^J, the
flue gas volume would increase by 36 percent. The increased gas flow would
reduce the effective SCA of the unit and consequently the collection effi-
ciency. Duct corrosion and flange leaks also can increase the flue gas
volume over several months or years. In addition, many operators may allow
boiler excess air to increase to undesirable levels.
Oxygen content in the flue gas should be measured periodically, before
end after the ESP, with portable instruments (if permanent instruments are
not installed). Typically, boiler operating instruments are installed at the
boiler outlet or the economizer outlet, and they do not necessarily reflect
ESP conditions.
Flue Gas Volume--
The flue gas volume being handled by the ESP directly affects the per-
formance of the unit, Typically, absolute values of gas volume are deter-
mined by stack test, and they may or may not be applicable on a day-to-day
basis because of variations in excess air, inleakage, process load, or tem-
perature.
Flue gas volume may be estimated by adjusting known volumes (process
generating rates) for changes in temperature or excess air (oxygen content).
These methods (generally defined as F-factor estimates) can produce gas
volume estimates within +10 percent of the true gas flow.
Another method of estimating gas volume is by noting the current drawn
by gas moving devices or fans. If current is used, the absolute value must
be normalized to the same fan rpm and flue gas temperature. This method also
may be used where parallel chambers are vented through two identical fan
SECTION 6-INSPECTION METHODS AND PROCEDURES
6-26
-------
units. For optimum performance, each chamber should treat the same gas-
volume.
6.2.10 Process Influences
When making an external inspection of the ESP, the inspector must consi-
der the constraints and conditions placed on the unit by the process to which
it is applied. Flue gas volume, temperature, moisture dust loading, particle
size distribution, particle chemistry, resistivity, and gas composition are
all determined by the process variables. During inspections, the tendency is
to ignore the influence of the process variables on ESP performance. The
inspector must document all process operating conditions that have an impact
on the flue gas or particulate conditions. On the surface, these process
conditions may not appear to influence ESP operation, but they can have an
indirect impact.
Tables 6-2, 6-3, and 6-4 list several important process parameters for
utility and industrial boilers, cement kilns, and kraft recovery boiler
applications, respectively. The items are not all-inclusive; a review of
each source is required to define site-specific parameters.
6.2.11 Opacity
The inspector should observe the condition of the ESP stack and make
opacity readings in accordance with EPA Hethod 9 procedures. These readings
should be made at 15-second intervals and averaged over a 6-ininute period.
The total opacity evaluation time should be at least 30 minutes and preferably
extend over one ESP rapper cycle. The 6-minute averages should be plotted to
identify any cyclic patterns that might indicate rapper failure.
If the ESP is equipped with an opacity monitor, the inspector should
record the current 6-minute average opacity and note any patterns that are
evident in the previous 2 to 4 hours of strip chart recordings. With the
monitor placed in the real-time output mode (integration time set to zero),
the frequency and magnitude of rapper reentrainment spikes should be evaluat-
ed. Rapper reentrainment spikes are generally caused by outlet field rap-
pers. If the spike pattern indicates that reentrainment is resulting from
inlet or midfields, overrapping may be occurring when gas velocity is too
high in the treatment zones.
SECTION 6-INSPECTION METHODS AND PROCEDURES
6-27
-------
TABLE 6-2. KEY PROCESS PARAMETERS FOR UTILITY AND INDUSTRIAL BOILERS
Flue gas temperature (economizer outlet)
Flue gas oxygen (economizer outlet), %
Fuel feed rate, tons/h
Fuel moisture (wood, bark, etc.), %
Fuel sulfur (coal), %
Fuel ash (coal, wood, etc.), «
Fly ash carbon (wood/coal), %
Pyrite reject rate (coal), tons/h
Ash composition (coal)
TABLE 6-3. KEY PROCESS PARAMETERS FOR CEMENT KILNS
Process (dry, wet, preheater, precalciner)
Raw material feed rate, tons/h
Composition (CaCCu, FeO, CaO, AKO, NaJ3, K?0, SO,
Moisture, % i
Clinker production rate, tons/h
Fuel rate, tons/h
Fuel type
Burning zone temperature, °F
Back-end temperature, °F
Flue gas oxygen, %
Insuflation rate, tons/h
Insuflation composition (Ne^O, K^O, etc.)
Kiln speed (revolutions per hourf
Type of cement produced (Types I through V)
', Al", etc.)
TABLE 6-4. KEY PROCESS PARAMETERS
FOR KRAFT RECOVERY BOILERS
Boiler type
Evaporator type
•Black liquor rate (BL), gal/min
Black liquor solids (BLS), %
Black liquor (chlorides), %
Soot blowing rate, 103 Ib/h
BLS into evaporator, I
BLS out of evaporator, A
Flue gas oxygen, %
Flue gas temperature, °F
Char bed height, ft
Char bed temperature, °F
Primary air temperature, °F
Primary air flow, 103 Ib/h
Brine rate, gal/min
Brine solids, %
Oil firing rate, gal/min
Salt cake pH (ESP)
SECTION 6-INSPECTlON METHODS AND PROCEDURES
6-28
-------
If monitors are installed on multiple chambers, average and instantane-
ous opacity values for each chamber should be compared. Deviations that
cannot be explained by gas-flow deviations, particle segregation, and strati-
fication may indicate rapping problems, clearance deviations, or distribution
plate pluggage. Deviations should also be correlated with the ESP power
input and pattern in each chamber.
6.2,12 Electrical Power Readings
The electrical power readings in each field provide an indication of ESP,.
conditions. Whereas the primary current and voltage reading provide a signif-
icant amount of information on the performance of the unit (because they are
on the line side of the T-R), they cannot directly measure the current and
voltage characteristics of the electrical field between the plates and wires.
The secondary current and voltage (as measured in most modern ESP control
circuits)» however, provide detailed information on the electrical field
force and current flow in each section of the ESP.
For example, no voltage on the secondary side may be due to an open
primary circuit. The circuit breaker may be open or tripped or a reactor
secondary may be open. High voltage on the primary side and no voltage on
the secondary side may be due to a faulty, open, or disconnected T-R set; an
open bus; or a faulty rectifier. Low voltage on the secondary side coupled
with low voltage on the primary side could be the result of leaks in the
high-voltage insulation, buildup of dust in the discharge system, excessive
dust on the electrodes, or swinging electrodes.
No secondary current and no secondary voltage indicate an open primary
circuit. Irregular secondary current coupled with low secondary voltage
indicates a high-resistance short in the circuit. This condition may be
caused by excessive dust or arcing.
The inspector should record both primary and secondary meter readings
during each inspection. In any reading, deviation from normal values should
be noted, and the data should be correlated with other data obtained during
the inspection (i.e., rapper function, flow distribution, process conditions,
internal inspection data, etc.). The secondary current levels in each field/
bus section should be plotted for each field from inlet to outlet. Generally,
SECTION C-INSPECTION METHODS AND PROCEDURES
6-29
-------
current levels should increase as dust is removed from the gas stream. The
rate of increase and final values are site-specific and vary with gas stream
conditions. A baseline or reference value should be established during
acceptance tests or compliance tests.
If the internal conditions and gas conditions (flow rate, temperature,
dust loading, etc.) are similar in parallel chambers, the electrical readings
should be comparable. Deviations should be used to target investigation
during the internal inspection.
Power levels vary with process conditions that modify ash and gas stream
characteristics. Deviations that cannot be related to physical ranges in the
unit can be correlated with process variables, such as fuel changes, excess
air, boiler load, or raw material chemistry.
In recording and evaluating ESP meter readings, the inspector should
recognize that the first indicator of ESP performance levels 'is the trend
observed in the voltage and current levels over time. Specifically, the ESP
secondary current levels should show a gradual rise from the inlet to the
outlet fields (see Figure 6-6). No numerical scale is given for secondary
current milliamps in Figure 6-6 because the secondary current in milliamps is
a function of the number of feet of wire per field, and that information was -
not available for this example. Depending on field size, the secondary
current level seldom exceeds 250 mA in the inlet fields. Typical designs
call for a current level of 0.02 mA per linear foot of wire length in the
inlet fields. Sparking in the inlet fields is usually indicated by deflection
of the meters. When the meters indicate progression toward the outlet field,
the secondary current levels should increase and sparking should decrease,
with almost no sparking occurring at the outlet. The T-R current levels of
most ESP outlet fields should be at least 85 percent of the T-R set current
rating (e.g., if the secondary current rating is 1000 mA on the outlet T-R, a
reading of at least 850 mA can normally be expected in the design) and in the
range of 0.06 to 0.10 mA per linear foot of discharge wire.
Another trend that is evident in some ESP's is a gradual decrease in the
secondary voltage from inlet to outlet. In larger ESP's (with five fields or
more), voltages often decrease in the inlet field (due to sparking), increase
in the middle fields, and then decrease slightly in the outlet fields.
SECTION 6-1NSPECTIQN METHODS AND PROCEDURES
6-30
-------
'CHAMBER B
(CHAMBER A
FIELD
Figure 6-6. Example of secondary current pattern for two chambers with
Chamber A having maintenance problems that limit power input.
6-31
-------
The gas entering the-inlet field of an ESP contains the greatest concen-
tration of particles. Because the greatest quantity of particle charging
occurs in this field, many ions are captured during charging. The rate of
•charge transfer from the discharge to the collection electrodes is thus
reduced because the mass of the particles migrating toward the plates is
considerably larger than that of the ions (ion mobility). In addition, more
-force is required to move the electrons from the corona discharge to the
collecting plates because the. charged particles in the interelectrode space
-act as a large negative cloud that repels the ions. The greatest amount of
sparking occurs in the first field, as this large cloud of charged particles
increases the electric field adjacent to the plate, leading to gas breakdown
(spark formation).
As the gas moves through the ESP and the particles migrate to the plate,
there are fewer particles to capture and inhibit the flow of electrons;
therefore, less force (voltage) is required to obtain a high current flow,
"Usually the increase in current is much greater than the decrease in voltage,
and the net effect is an increase 1n power input from ESP inlet to outlet.
The relationship or ratio between voltage and current levels for each
•T-R is not constant throughout the ESP (from inlet to outlet). Changes in
the voltage-current relationship are due to capacitance and resistance char-
acteristics of the particulates and to inefficiencies in the circuitry of the
•T-R set at various operating levels. For the same reasons, the relationship
between the primary and secondary voltage and current levels also will not be
constant from inlet to outlet. Thus, general relationships between primary
and secondary meter readings are difficult to establish and will change with
dust characteristics. Over very narrow operating ranges within a T-R, howev-
er, the relationship between primary and secondary voltage and current levels
may be considered linear. This linear relationship is useful in the evalua-
tion of ESP performance if a secondary meter is out of service and the corre-
sponding primary meter on the other side of the transformer is operating.
Whereas the trends in the voltage and current levels are important in
evaluating ESP performance, corona power input is one of the most useful
indicators of ESP performance. Secondary meters are preferred for determin-
ing corona power because they are more indicative of actual power input tc
SECTION 6-INSPECTION METHODS AND PROCEDURES
6-32
-------
the ESP. Primary meters may be used for this calculation,---however, if they
are the only meters available, provided an appropriate efficiency factor is
applied. Corona power input is simply the product of the secondary voltage
multiplied by the secondary current to yield watts of power to the ESP field
frorr the T-R, This value should be calculated for each field of the ESP.
When both primary and secondary meters are available, the primary voltage and
current levels should be multiplied to yield the primary power input to the
T-R in watts. The value of the primary meter product must be higher than
that obtained from the secondary meter product. If the secondary power
product (corona power) is the higher of the two, the values on the meters are
Incorrect. Isolating the malfunctioning meter, however, may be very diffi-
cult.
Electrical losses occur in the T-R during increases in voltage and rec-
tification to an unfiltered DC wave form. Losses in the T-R control cir-
cuitry also reduce the efficiency of the transfer from primary input power to
secondary corona power. The apparent efficiency factor for the T-R can be
calculated by use of the ratio of secondary power to primary power:
apparent T-R efficiency - (I,. 1)
As mentioned in Section 3.1.1, this efficiency value is apparent because
it encompasses two different methods of measuring current and voltage (i.e.,
RMS on the primary side vs. average on the secondary side}. The apparent T-R
efficiency value usually ranges from 0.55 to 0.85, although the theoretical
maximum conversion efficiency of a silicon T-R is only 0.77. Apparent values
as high as 0.90 are occasionally noted. Again, with the exception of a
malfunctioning meter, the actual values obtained from the meters are not as
important as the patterns that are created by these values. In general, the
value of the T-R efficiency increases as the T-R approaches its rated output
current level. Thus, T-R efficiencies tend to be lower in the inlet fields
(0.55 to 0.60) and higher in the outlet fields {0.80 to 0.85) because of lim-
itations imposed on the electrical operating characteristics by particulate
loading and space-charging. A value of 0.70 to 0.75 is usually appropriate
for use as an average for all fields in the ESP.
SECTION 6-INSPECTION METHODS AND PROCEDURES
6-33
-------
Another item that should be checked in mu Hi chambered ESP's with T-R's
for each chamber is the balance of power across the ESP, The secondary
current level and power input for each field should be approximately equal
across the ESP. Although some relatively small differences {2 to 51) may
occur because of rapper sequencing, slight gas-flow imbalances, or differenc-
es in internal alignment, these differences should not be large in most ESP
designs, as chambers with equal gas flow and equal power levels give the best
ESP performance. Temperature variations across the face of the ESP, however,
may upset the balance of power.
When the power levels have been calculated and the patterns have been
checked (as previously discussed), the corona power from each T-R should be
added together to determine total corona power to the ESP (in watts). When
multi chambered ESP's having T-R's serving each individual chamber are in-
stalled, the corona power levels should be totalled for each chamber (to
indicate balanced power) and an overall total corona power level should be
calculated for all chambers.
If available, baseline test values should be compared with the actual
secondary current and corona power levels. If the gas volume through the ESP
is identical to the test value and the meter readings are also very similar,
ESP performance is usually much the same as that observed during the stack
test. Comparing design power levels with performance levels is usually
impossible, as power levels are rarely included in design information.
A key indicator of ESP performance is the specific corona power, which
is a useful value in determining whether ESP performance has changed signif-
icantly over time.
Specific corona power is calculated by the following equation:
Specific corona po«er . <*<•
This value may be calculated for the entire ESP or for individual chambers.
In general, the higher the value of the specific corona power, the higher the
ESP removal efficiency is. Thus, an evaluation of the specific corona power
indicates whether the ESP performance would be expected to increase or de-
crease.
SECTION e-INSPECTION METHODS AND PROCEDURES
6-34
-------
2
Power density (watts/ft plate area) is another useful indicator of ESP
performance. The value for power density may be calculated by the following
equation:
Power density - ™rof Power inPut (Eq. 3)
ft plate area
This calculation may be performed for each field by substituting into
the equation the T-R corona power and the plate area associated with that
T-R. The increasing power density that should be evident from inlet to
2 2
outlet may range from 0.25 W/ft at the inlet to as high as 5 W/ft in the
outlet field. This calculation is useful for ESP's in which the fields ire
not of equal size. Although the increase of secondary current and corona
power does not progress smoothly, this calculation will help "normalize" the
values. A corresponding calculation may be performed on the secondary cur-
?
rent (current density, mA/ft } to normalize the data.
By substituting total corona power and total plate area into the equa-
tion, one can usually determine the overall power density. Typical values
2
are 1 to 2 W/ft » and these usually indicate §ood performance. As with
specific corona power, however, performance may be poor even. though the power
density is high. Again, the inspector must rely on his/her background and
experience to make judgments regarding performance.
The inspector can use a baseline test to calculate ESP efficiency. To-
tal corona power, gas volume, and efficiency or penetration (1 - efficiency)
are calculated in the baseline test, and these values can be used, along with
a modified version of the Deutsch-Anderson equation, to calculate a constant
for use in subsequent calculations:
Pt - e"°-06 K
-------
This equation relates corona power input and gas flow to ESP performance,
whereas the original form of the Deutsch-Anderson equation assumed maximum
and nonvarying field density and related particle migration rate to the SCA
of the ESP. The particle migration rate, however, is a function of corona
power input; thus, corona power changes the particle migration rate and the
ESP efficiency. This equation is useful for predicting changes in efficiency
if wide variations in the specific corona power have not occurred or the
power distribution within the ESP is not seriously altered. Figure 6-7
illustrates the efficiency of a coal-fired utility boiler ESP as a function
of corona power.
The value of "K" will typically be in the range of 0.1 to 0.25 for most
kraft recovery boilers and 0.35 to 0.55 for utility boilers. When the value
of the constant has been calculated, it may be applied in subsequent calcu-
lations in which inspection data are used. Because the equation usually will
overpredict performance within narrow operating ranges, the predicted effi-
ciency is based on the assumption that all the power is used to collect par-
ticulate. If the specific corona power decreases substantially, however, the
equation may greatly underpredict performance because the equation becomes
very sensitive to variations in the specific corona power. In fact, the val-
ue of the constant begins to change with substantial increases or decreases
of specific corona power, and the predicted performance can be incorrect by
as much as half an order of magnitude when applied over a very wide range of
specific corona power. The value of the constant and the behavior of the ESP
will also change if T-R's become inoperative (particularly those in the inlet
fields). Nevertheless, the equation will provide a gross indication of a
performance shift.
The underlying limitation of the equation is the assumption that power
input and changes in power input are the only factors that affect ESP per-
formance once the SCA and gas volume are fixed and that the value of the con-
stant and the effective migration velocity are related by a linear relation-
ship. Actually, the relationship in the value of the exponent is probably a
power function similar to that presented in the Matts-Ohnfeldt equation,
which makes the equation less sensitive to variations in power input. Care
SECTION 6-INSPECTION METHODS AND PROCEDURES
G-36
-------
99.0
9G.O-
100 200 300 350 400
CORONA POWER, w/1000 icfm
Figure 6-7. Corona power versus collection efficiency for a coal-fired
utility boiler.
6-37
-------
must be exercised when an equation is used at conditions different from
baseline values.
In summary, the external evaluation of ESP data includes plotting the
secondary current to discern an increase from inlet to outlet; calculating
corona power from primary or secondary meters; evaluating the balance between
the readings in parallel fields; calculating total corona power, power
density, and specific corona power; and comparing these values with baseline
readings. If conditions are nearly identical, performance is likely to be
similar. Even if conditions are not identical, the inspector can draw some
conclusions regarding performance based on major deviations from the estab-
lished baseline conditions. For small changes of specific corona power, an
indication of the magnitude of any performance shift may be estimated by use
of a modified form of the Deutsch-Anderson equation. This equation should
not be used when gross changes in ESP operation have occurred unless the
values are modified appropriately.
6.3 INTERNAL INSPECTION
Proper O&M must include the internal inspection of an ESP, including
ductwork, the electrical system, and the ash handling system. Periodic
internal inspections are usually part of routine preventive maintenance
activities or part of troubleshooting (diagnostic) evaluations. The data
obtained must be combined with on-line evaluations to assess the ESP's condi-
tion adequately and to define needed maintenance or changes in the operation
of the unit.
Before an internal inspection is begun, the reason for and scope of the
inspection should be defined. The detail and scope of any internal inspec- .
tion may be limited by time availability, resources, personnel, and accessi-
bility. Annual internal inspections are generally necessary to determine
hidden problems, to conduct scheduled repairs, or to make design changes.
These more lengthy and involved inspections require exceptional planning and
coordination.
Any diagnostic testing that is to be done during an outage, should take
place before the dust layers on electrodes and plates (air load tests, dust
_depth^ dust analysis^ etc.) are disturbed.
SECTION 8-INSPECTiON METHODS AND PROCEDURES
6-38
-------
The normal inspection procedure Involves several access areas in the ESP
and the entry of many personnel during outages (prerap periods, post-rap
periods, post-wash-out, preclose-up, etc.). Each inspection is performed
under different conditions to acquire specific information.
The following sections describe the areas that should be inspected, the
method of Inspection, possible malfunctions, and the effect these deficien-
cies may have on ESP operation.
6.3.1 Inlet Plenum
The inspector should check the gas-turning vanes in the ductwork entering
the ESP. Vanes should be relatively free of dust accumulation and not eroded
or corroded. Gas-turning devices are designed through scale model studies,
and changes in gas-flow sectors or volume (chamber bias) can result in serious
maldistribution of gas velocity over the ESP inlet. Deviation of the average
P
velocity from the IGCI standard reduces the potential collection efficiency
of the collector especially for marginally-sized units. Figure 6-8 shows gas
velocity distribution patterns resulting from improper gas distribution at
the inlet plenum of the unit. As shown, the entry of high gas velocities in
several areas of the collector results in lower gas treatment time, turbu-
lence, high superficial velocity, and increased rapping reentrainment.
Areas of dust accumulation in the inlet nozzles should be noted, and
changes in gas distribution, vibrators, or rapper frequency should be made to
eliminate these deposits. Increases in the depth of dust deposits can change
the inlet gas velocity profiles. Extension of the perforated plate to the
floor of the dust can impound the dust and prohibit it from flowing into the
inlet field hopper. Slotted openings at the bottom of the distribution plate
can allow dust movement without changing the velocity across the plate. If
the dust deposits continue to be a problem and normal remedies do not work,
the rate of accumulation can be estimated and the dust can be removed during
scheduled outages before it interferes with ESP performance.
The inlet plenum should be inspected for corrosion In areas where air
inleakage has occurred. These areas may be associated with hard dust depos-
its, poor exterior insulation, or surfaces with maximum heat loss (prevailing
wind direction). Each penetration point should be repaired and sealed.
SECTION 6-INSPECTION METHODS AND PROCEDURES
6-39
-------
Voo '
10
-1
t IPO 100
. iQQ sp
200
300
en
i
Figure 6-8. Velocity distribution patterns resulting from improper
gas distribution. (Numbers are actual values in various portions
of the ESP inlet. The average velocity through the unit is 166 ft/min.)
-------
Wash-out may be required to reveal metal surfaces, and probing may be neces-
sary to determine the extent of deterioration. The service, application,
history, and age of the unit will determine the level of detail needed.
The distribution plate should be inspected for wear and deposits. Loca-
tion of deposits should be noted and correlated with rapper operation to
determine what corrective action should be taken. Among the numerous reasons
for the pluggage of distribution plates are gas velocity distribution, dust
stratification, dust composition, low gas temperature, moisture, or ineffec-
tive rapping. When a portion of the plate becomes covered with dust, the re-
sistance to gas flow increases and causes gas flow to shift to cleaner areas.
When clean, the distribution plate offers a uniform resistance to flow by
evening out the velocity vectors entering the ESP from the duct. Pluggage of
the plate starves areas (lanes) of the collector of both particulate and gas
flow. The resulting shift in gas flow reduces the physical size of the
collector with respect to treated gas volume (i.e., decreases SCA, increases
superficial velocity, reduces treatment time, and reduces efficiency). It is
critical that the distribution plate be as clean as possible and that effec-
tive rapping be maintained to remove accumulated dust.
Most distribution plates are rapped from the upper edge, which is not as
effective as horizontal (face) rapping. Changes in rapping direction, fre-
quency, and intensity are often required to remove chronic dust accumula-
tions. A change in rapping direction may necessitate structural changes in
the distribution plate supports or an attachment to accommodate the addition-
al stress on the welds.
Severe plate pluggage may result in a gas pressure sufficient to distort
and collapse the plate. Figure 6-9 shows a chronic distribution plate plug-
gage problem that has resulted in reduced efficiency and distortion of the
plate into the electrical field.
Dust accumulation in the inlet nozzles can also place abnormal stress on
the distribution plate and cause the plate to collapse into the inlet electri-
cal field.
6.3.2 Deposit! on Electrodes
The electrode is the electrical component that provides the high elec-
tric field in which particles are placed. The electrode generates the corona
SECTION B-INSPECTION METHODS AND PROCEDURES c »,
O-*U
-------
Figure 6-9. Chronic distribution plate pluggage problem.
6-42
-------
that provides electrons, and subsequently, Ions for particle charging. The
diameter of the electrode or spike is critical in determining the voltage at
which corons is initiated. The inspector should visually check the depth,
location, and hardness of deposits on wires before and after rapping. An
evaluation before rapping indicates areas of close clearance or ineffective
rapping. Instead of being uniform along the electrode length, deposits may
be heavier near the bottom of the wire or they may occur at the outlet of the
electric field (plate bottom). These deposits are comonly referred to as
donuts or dumbells. Areas of heavy accumulation can be correlated with
electrical power readings (V-I curves) prior to shutdown. Buildup on weights
and the lower wire frame should be noted as possible areas of grounds, and
the free movement of bottle weights in the weight guides should be checked.
All abnormally heavy deposits should be removed manually before startup.
Depending on dust composition and hardness, simple rapping with wood poles
may dislodge the deposits. In some cases, dropping a heavy weight with a
wire rider (slot) down each wire and retrieving it by rope will remove hard
crusty material. Dust deposits on the wires should be evaluated from the
lower wire frame area and upper wire frame aree.
1 Bottle weights can be freed in the wire frame guides by'striking them
with a heavy object (hammer, pipe, etc.). Care must be taken not to dislodge
the wire from the bottle weights. Wires are held in place by the weight of
the bottle. If the bottle becomes frozen, the wire may partially move out of
the seat and later become free during normal rapping. Figure 6-10 shows a
heavy accumulation of dust on bottle weights that can reduce clearance and
cause eventual grounding of the electrical field.
The inspector should note the location of any wires removed because they
were grounded (clipped wires) or those removed because of close clearances.
Random removal of up to 10 percent of wires does not reduce ESP efficiency,
however, removal of several wires in one lane creates a path through which
gas may pass without particle charging. Any areas where a large number of
wirts have been removed should be investigated.
If an electrical short was identified during the external inspection,
the indicated bus section should be checked from the top frame to determine
SECTION 6-INSPECTION METHODS AND PROCEDURES
-------
Figure 6-10. Example of bottle weight deposits.
6-44
-------
if a wire has burned off and is grounded. The grounded wire should be re-
moved.
6.3.3 Plate Deposits
The thickness of dust deposits on collection surfaces should be evaluat-
ed to determine if rapping is effective. This evaluation should be made at
shutdown, prior to the rapping and/or washout of the ESP. Areas that showed
low power levels during the external inspection should be targeted for spe-
cial analysis.
Because dust removal from plates is not uniform, dust tends to fall in
large layers when the plates are rapped. Adherence is a function of dust
composition and thickness, plate design, rapping method, rapping intensity
and frequency, and electrical field strength.
Dust removal is best at the top of the plate (top rap) and worst at the
bottom of the plate. Variations in dust thickness from the top to the better;,
are normal and usually should not interfere with collection efficiency.
The resistivity of the dust layer is more important to collection effi-
ciency than its thickness is. Heavy dust deposits of conductive dusts (e.g.,
•those produced by high-sulfur coal) may not reduce power input, but moderate
dust layers (less than 1/16 in.) of high-resistivity dusts (e.g., those
produced by low-sulfur coal) will severely reduce power input and collection
efficiency. Layers of resistive dust are hard to remove by heavy rapping
forces, whereas conductive dusts will generally fall off the plates in large
sheets. Variations in the thickness of dust deposits from area to area
within the collecting fields are an important indicator of potential rapper
problems.
6.3.4 Deposits on Wire Frames
Dust accumulations on upper wire frames, plate rapper beams, anvils, and
the roof structure should be checked because these accumulations can indicate
gas velocity vectors above the collection fields and possible gas sneakage,
Figure 6-11 shows an accumulation of dust on the rapper header beams of the
upper wire frame. Because of the inlet plenum design, deposits such as these
are more likely to occur in the inlet mechanical field, where gas vectors are
upward. Heavy deposits can cause electrical grounding of the high-voltage
SECTION 6-INSPECTION METHODS AND PROCEDURES 5.45
-------
Figure 6-11. Accumulation of dust on rapper header beams,
5-46
-------
frame. They also can add mass to the wire frames and plate system, which
tends to dampen rapping energy and reduce rapping effectiveness.
6.3.5 Antisneakage Baffles
Gas baffles between the fields of the ESP prevent flue gases from pass-
ing above or below the gas treatment zone. The location and design of these
baffles vary with the manufacturer and the age of the unit. In general,
however, the baffle is placed between fields at the hopper valleys, at the
center!ine of hoppers, or between fields in the roof.
In units equipped with hoppers or with drag chains whose movement is
perpendicular to the gas flow, the baffle is a fixed sheet of metal. Figure
6-12 shows the location of a hopper center!ine baffle. These baffles should
be solid partitions with openings only for structural bracing, passage of
level detector beams, or maintenance access. The baffle should extend down
into the hopper throat to prevent sneakage, but not to the point of interfer-
ing with hopper discharge.
In many designs, the baffle (particularly the upper baffle) also serves
as a structural support member. Figure 6-13 shows an upper baffle used as a
plate suspension.
In drag bottom systems in which the drag moves parallel to the gas flow,
baffling is critical to the prevention of gas sneakage below the treatment
zone. The baffle must be free to move out of the path of the drag scraper
blades (top and bottom) and must have penetration points for the drag chains.
Over years of operation, the baffle plate may become distorted, misaligned,
and shifted into e deflected position. Baffle plates also may be torn by the
scraper blades and have to be removed from the unit. Figure 6-14 shows a
scraper blade passing under a baffle in the deflected position.
The inspector should verify that all baffles are in their proper posi-
tion and free to move and that baffle sections do not interfere with each
other during motion. Movement of baffles during chain drag operation should
be checked from outside of the ESP. A visual evaluation can be made through
the access door or from structural walkways between fields above the drag
area. The inspector must not enter the hopper when drags are in operation.
As a means of avoiding complete loss of resistance to gas flow under the
fields, the spacing of the drag chain blades is set to prevent successive
SECTION 6-INSPECTION METHODS AND PROCEDURES
6-47
-------
Figure 6-12. Baffle hopper center!ine.
6-48
-------
Figure 6-13. Upper baffle used as plate suspension.
6-49
-------
Figure 6-14. Scrapper blade passing under a baffle in a
deflected position.
6-50
-------
baffles from being deflected by the scraper blades. As drag chain blades and
chains are replaced during maintenance and repair, the position of blades may
change, which can result in simultaneous deflection of the baffles. Adjacent
parallel drags should be offset to minimize the baffle opening at any in-
stant. The inspector should check the sequencing of baffle operation from
the outside access door.
Failure to maintain positive baffling can allow flue gas to pass under
the treatment field and result in the reentrainment of particulate matter
into the outlet plenum. Because dust is stripped from hoppers during rap-
ping, sneakage generally increases the severity of rapper spikes. These
spikes are most severe when the unit is operated above design gas volumes
(higher velocity), when the top inlet is used and gas distribution is poor,
or when the vector of the gas entering the unit is downward into the inlet
hopper,
6.3.6 DragChains
Drag chains remove the dust as it is rapped from the collection plate
surfaces. These drags may be perpendicular or parallel to gas flow, and they
operate on a large chain sproket (head and tail) that allows*them to pass
over the ESP bottom and then return in the opposite direction (Figure 6-15),
Attached to each pair of chains are drag scrapers (L-shaped plates) that push
the dust to the end of the chain and discharge it into a transversely mounted
screw conveyor.
To be effective, movement of the drag scrapers must be parallel to the
hopper bottom and not be deflected upward (which causes the scraper to pass
over the dust layer). Deflection can be prevented by adjusting the chain
drive sprocket pillow blocks to maintain chain tension.
Each chain must be parallel to the other, and the tension must be iden-
tical. Misalignment can cause chains to break or bind and/or the dust to
shift to one side during operation.
The drive shaft is usually equipped with shear pins to prevent major
damage if the drag chain should become entangled in the baffles, with other
chains, or with the bottle weights.
Because of the possibility of binding, drag chain scrapers are separatee
from each other and from the wall surfaces. This clearance can result in an
SECTION 6-INSPECTION METHODS AND PROCEDURES
6-51
-------
Figure 6-15. Drag chain assembly.
6-52
-------
accumulation of deposits that may eventually interfere with electrical clear-
ances and cause grounds. The inspector should note the location of any such
deposits, and these should be removed periodically as a preventive mainte-
nance measure.
Drag chains are considered to be the highest maintenance item in an ESP,
Any deviation from optimum operation can result in extensive downtime for re-
pair and maintenance,
6.3.7 Penthouse and Insulator Enclosures
The penthouse is located between the weather enclosure (roof) on which
the rapper and T-R sets are located and the inner roof .containing the flue
gases. This area contains the support (barrel) insulators that are used to
hang the electrical wire frames; it does not contain flue gases, and it is
pressurized to prevent dust penetration through rapper access openings and
insulators.
The inspector should determine whether dust is entering the enclosure,
and if so, the probable point of entry. Oust deposits can cause electrical
tracking and surface cracking on insulators. Penetration can be expected if .
the penthouse purge fans were ineffective or were not operating during the
external inspection. The inspector should also check for any areas of corro-
sion or any indications of water flow from condensation or rain penetration.
The insulators should be checked for signs of condensation, cracks, dust
deposits, or chips on the upper edge. If insulator heaters were not operat-
ing during the external inspection, continuity checks should be made of the
heating elements.
Condensation is the result of hot, moist gases contacting colder surfac-
es. The problems described in this section are related to the paper industry
where the high-power usage for heated air is accepted by the industry as a
necessary cost of operation. The underside of the penthouse roof and exter-
nal penthouse sides are insulated. The bus ducts must also be insulated.
The penthouse is pressurized and insulators purged with heated air. This
requires 5.5 kW and 100 cfm per insulator. Failure of the heaters results in
cooling of the hot roof in contact with the process gases. Corrosion and
dust in the penthouse is indication of a failure of the pressurizing fan.
SICTION 6-1N5PECTION METHODS AND PROCEDURES
6-53
-------
A more cost-effective means used on other dry-dust applications such as
incinerators, industrial power boilers, cement, lime, etc. will insulate the
top surface of the hot roof and external penthouse sides. The cold penthouse
roof and bus ducts are not insulated. The insulator surfaces are maintained
above outside ambient by radiation from the process and a 400-watt contact
heater. The penthouse is pressurized with ambient air at the rate of 100 cfm
per insulator. With all internal surfaces above ambient, condensation will
not form. Corrosion and dust buildup in the penthouse are an indication of
reverse flow of process gases into the penthouse due to pressurizing fan
failure.
If the underside of the weather roof is insulated, blocks of insulation
that become loose and fall can create cold spots; these areas should be
repaired. The conduit duct from the T-R set to the penthouse may be vented
through the penthouse fans and insulated (if the roof is insulated) to keep
the temperature of the duct above the condensation point. In some cases,
however, the duct is neither insulated nor purged, which results in a cold
spot in the roof. Water may condense, run down the sides of the conduit, and
cause insulator arcing and failure. In these areas, rapid metal corrosion
can occur from the inside of the conduit. Figure 6-16 shows water patterns
caused by cold conduit in the penthouse roof.
The introduction of purge air into the penthouse can cause erosion
and/or corrosion of the metal in the inner roof below the inlet of the purge
air. This area should be inspected for metal thinning, and repairs should be
made as necessary. Impingement baffles may be required for adequate distribu-
tion of the purge air.
Penetration openings (pipe nipples) of the plate rapper shafts should be
inspected for dust accumulation and' shaft binding. Deposits should be
cleared, and the opening should be sealed with high-temperature rapper boots.
Any major misalignment of rapper shafts through the weather roof to the
inner roof should be noted. Major lateral movement can indicate changes in
plate alignment and result in possible rapper shaft or anvil failure.
Discharge frame rapper shaft insulators should be checked for cracks and
evidence of electrical tracking. Ceramic insulators should be inspected to
SECTION 6-1NSPECTION METHODS AND PROCEDURES
6-54
-------
Figure 6-16.
Water patterns caused by cold conduit in the
penthouse roof.
6-55
-------
ensure that end caps are bonded and not loose, and resin-type insulators with
tapered connections should be checked for tightness (Figure 6-17), as any
loosening of the connection can result in incomplete rapping energy transfer
(bouncing).
An insulator compartment may be used when the design does not include a
full penthouse. These compartments are referred to by various names, includ-
ing "coffin box" or "dog houses." Access into these enclosures is limited,
and the inspector may need to use small mirrors to inspect areas of the insu-
lators. The compartment should be free of dust and the insulators should be
clean as shown in Figure 6-18. Because of the small volume of air required,
purging can be effective and complete. Figure 6-19, however, illustrates
dust deposits in an insulator compartment.
6.3.8 Rapge_r_Shafts and Anvils
Rapper shafts and anvils that are attached to the collection plate head-
er beam should be inspected for tightness and possible weld failure. For a
complete inspection in this area, the surfaces may have to be washed to
remove dust deposits; however, washing should be avoided whenever possible to
reduce corrosion.
"' Bolted and welded components of the plate rapping header assembly should
be inspected periodically to ensure proper transmission of rapping energy.
6.3.9 Plate Warpage and Cracks
Heavy rapping forces can cause structural failure in some systems be-
cause of weld breaks in the plate header support and roof structure, chips in
the plate attachment, and channels or splits in the plates (Figure 6-20).
After a wash-down, the inspector should inspect the plate support and attach-
ment hardware for misalignment, distortion, or breakage.
Strengthening of the hardware may be necessary if intense rapping forces
are required to remove resistive ash or sticky salt cake. Depending on
manufacturer, the failure can result in alignment changes or plate dropping,
bowing, lateral movement, or tears.
If the unit has been subjected to high temperature,-fire, or explosion,
plates can become warped, which reduces electrical clearances between plates
and wires. This limits the voltage at which flash-over occurs and reduces
SECTION 8-INSPECTlON METHODS AND PROCEDURES
6-56
-------
Figure 6-17, Resin-type insulator.
6-57
-------
Figure 6-18. Access to a typical insulator enclosure and support
insulator for the discharge wire frame.
6-58
-------
Figure 6-19. Insulator compartment showing dust deposits,
6-59
-------
Figure 6-20. Example of plate cracks.
6-60
-------
total power Input to the entire electrical field. Local areas of plate warp-
age should be identified and the plates should be straightened or replaced.
Correction methods include stiffeners, heat treatment, patching, or replace-
ment.
If the ESP has sufficient redundancy, the wires in the area of the warp-
age may be removed to reduce the clearance problem. This allows electrical
power in the field to be increased to normal levels.
Allowable deviation from the plate design is ±1/16 inch. Normally,
plate warpage of less than 1/2 inch cannot be detected visually. Visual
inspection usually is conducted from the upper wire frame, looking down on
the plate and wire. Distances between wire and adjacent plates are compared
at several points along the wire length to judge areas of distortion.
Bus section V-I curves (air load) are useful for limiting the areas re-
quiring visual inspection. Because the V-I curve can show misalignment,
frame clearance problems, and other factors, a process of elimination may be
required.
6.3.10 Antlsway Insulators
Severa.l methods are used to support and align the lower: wire frame. If
the wire frame is supported by pipes attached to the upper frame, squareness
of the frame and support pipes must be maintained to enable alignment of the
plate and wire system. The wire frame is aligned during installation of the
system, end the components are bolted and welded to a ±l/8-inch tolerance.
Under normal operation, the wire frame subassenbly does not change shape, but
the frame may change orientation with respect to the plates and box as a
unit. This is covered under plate alignment. Also, excessive force from an
explosion or fire can distort the frame.
Many systems allow partial support and alignment of the lower wire frame
separate from the upper wire frame. These systems are more susceptible to
lower wire frame movement independent of the upper wire frame. Stand-off or
antisway devices are used between the wire frame and ESP shell to maintain
alignment. Because the wire frame is ma-intained at high voltage, the antl-
*sway devices must be nonconductive. The insulators may be constructed of
resin, ceramic, or alumina, and they may be attached to the side wall (hori-
zontal) or to the hopper wall (vertical).
SECTION S-1NSPECTION METHODS AND PROCEDURES
6-61
-------
In wire-frame systems, alignment rakes may be used at the leading and
trailing edges of the frames, and stand-off insulators may be attached to the
alignment rakes. Examples of alignment values and stand-off insulators are
presented in Figure 6-21.
If abnormal movement of the lower frames or rakes results from hopper
overflow, drag chain breaks, or overheating of the unit, the insulation may
be stressed to the point of failure. The inspector should check the integri-
ty and relative position of all insulators.
The attachment of the insulator mounting hardware cannot be rigid; it
must be free to move with thermal expansion of the shell and wire frame
system. Also, insulators must be relatively free of dust, moisture, and
coatings that make them conductive. Electrical tracking can cause insulator
failure or grounding of the electrical frame. The inspector should check all
insulators for dust deposits and clean each insulator with a nonorganic
cleaner.
6.3.11 Falling-Hammer Rappers
Falling-hammer rappers are used to dislodge dust from the plate and wire
frames. Each hammer impacts the surface once during the shaft's rotation.
Multiple hammers are attached to the shaft with U-bolts and clamps. As the
hammer reaches the apex of the shaft rotation, it swings and free-falls in an
arc, striking an anvil attached to the wire frame or plates (Figure 6-22).
If the hammers become loose on the shaft, they will not rotate and rap-
ping ceases. The inspector should note any free hanging or missing hammers.
Hammers and anvils can also become misaligned, which results in a hammer
strike that is offcenter. This can shear both hammers and anvils and cause
abnormal wear. Each hammer/anvil pair should be inspected for wear. Anvils
can also become detached from the plate and wire frame rakes, which results
in poor rapping.
6.3.12 Plate and ^ire Alignment
For practical purposes, the collection plates in each collection field
should be vertical and parallel to the ESP walls. Each plate pair forms a
gas lane in which particle charging and collection occurs.
SECTION 8-1NSPECTION METHODS AND PROCEDURES
6-62
-------
WIRE BREAKAGE AREA (TYP.
in.
TYPICAL CRIMP
\
30 ft
1/2, In, LONG x 3/16 In, WIDE
1/2 In.
1/2 in.
©
WI*BEr*BlTAvo^E 3/8 in, - SHROUDED WIRE
AREA
B)
ANTI-SWAY
INSULATOR
GRID
Figure 6-21. Examples of a) alignment rake and b) anti-sway (standoff) insulator.
6-63
-------
Figure 6-22. Falling hammer rapper.
6-64
-------
Each electrode nust be centered in the gas lane and be parallel to the
plate surface. Because the plate and electrode systems are independently
positioned and supported in the field and cannot make contact, each system
must be aligned accurately. Any reduction 1n clearance between a grounded
surface (plate, stiffener, brace, alignment bar, etc.) will result in a lower
spark-over voltage and current. Figure 6-23 shows a broken plate stabilizing
bracket, which could result in reduced clearance between the wire and plate
frames. This reduction in current limits power input and collection efficiency.
The normal accepted tolerance between wire and plate is ±1/4 inch.
If the plate system is assumed to be parallel to the shell and to serve
as a fixed reference, the upper and lower wire frame can then be shifted to
the center of the wires between the plates. Normally, the upper wire frame
is supported and hung from four or more support insulators, each of which can
be adjusted.
In general, each wire frame can move in two rectilinear directions (per--
pendicular and parallel to the gas stream) and can rotate through a limited
arc around a vertical line (movement is lateral around the set of supports).
Rotation of the frame about a horizontal, line can occur if the upper frame is
not level. A 1/16-inch change in elevation can be magnified to an approxi-
mate 1/2-inch movement of the lower frame in the lateral direction over the
length of a 30-foot plate.
Either the lower or upper frame in the direction of gas flow may cause
close clearances to occur at the edge of the plate near the stiffeners or
alignment rakes (Figure 6-24).
Independent movement of either the upper or lower frame perpendicular to
gas flow may cause close tolerances to occur at each plate edge, either
entering or exiting the plate area (Figure 6-25).
Movement of both frames perpendicular to gas flow can create a clearance
problem along the total length of the wire (Figure 6-26).
A shift of two of the support insulators perpendicular to the gas stream
can cause the wire frame to rotate and result in close clearances at the
corner electrode of each wire frame (Figure 6-27).
SECTION 6-INSPECTION METHODS AND PROCEDURES
6-65
-------
Figure 6-23. Photograph of broken plate stabilizing bracket.
6-66
-------
1
I WIRE FRAME
PLATES
Figure 6-24. Lateral movement of upper or lower wire frame (end elevation view),
WIRE FRAME
PLATES
a FT
Figure 6-25. Longitudinal movement of upper or lower wire frame (side elevation view),
6-67
-------
t*
0
3 PLATES
FRAME
<1
Figure 6-26. Lateral movement of upper or lower wire frame (plan view!
3 PLATES
WIRE FRAME
Figure 6-27. Rotation of upper or lower wire frame (plan view),
6-68
-------
Distortion of the insulator support seat (inner roof) can allow one or
more of the corners of the wire frame to move downward and cause close clear-
ance problems in the area diagonally opposite the insulator at the lower wire
frame.
Changes in alignment are made by moving the support insulators (singly
or in pairs) laterally or perpendicularly to the gas stream, depending on the
nature of the alignment problem. Frame elevation is changed by moving sup-
port adjustment screws or other devices (shims, washers, etc.)- Figure 6-28
shows jack screws used to move the insulator horizontally and the screw and
nut used to move the frame vertically.
The inspector should check wire-to-plate clearances visually at each
corner of the lower and upper wire grid (eight points). The direction of the
close clearance at each point will determine how the wire frames should be
moved. Any field with alignment problems mey be correlated with corona power
readings determined during external inspections. Accurate realignment of the
wire -frame should be accomplished by using a notched alignment scale or
spacer (Figure 6-29).
SECTION 6-INSPiCTION METHODS AND PROCEDURES
6-69
-------
Figure 6-28. Jack screws.
6-70
-------
-1!>
-f
*
$_
a
a.
VI
a
z
c
a.
c:
01
"O
0!
(J
+J
o
C\J
I
3
6-71
-------
REFERENCES FOR SECTION 6
1, White, H. J. Electrostatic Precipitation of Fly Ash. JAPCA Vol. 27 No.
3. P 213. March, 1977.
2. Industrial Gas Cleaning Institute Gas Flow Model Studies. Publication
EP-7.
3. PEDCo Environmental, Inc. Analysis of Participate Matter Emission Limit
Variance For the Coal-Fired Power Plants in North Carolina. EPA Contract
No. 68-02-3512, Task No. 34. March, 1982.
SECTION 6-INSPECTION METHODS AND PROCEDURES 6-72
-------
SECTION 7
SAFETY
The safety of plant personnel during all aspects of ESP operation ano
maintenance and Agency personnel during inspections is of ultimate importance.
Areas of concern include electrical shock hazard, confined area entry (oxygen
deficiency and toxic gases), hazardous materials (dust, metals, etc.), chemical
burns, eye injury, and normal industrial safety concerns such as moving
equipment, falls, etc. In ESP's, many of these concerns are simultaneous and
can result in potentially serious injuries to personnel. With proper planning,
safety equipment, and established procedures, operation and maintenance and
inspections can be performed safely without risk of injury.
Hany of the potential hazards and proper procedures for addressing them
are discussed in the following subsections. Further information on confined-
area entry and manufacturers of safety appliances can be found in specific
vendor maintenance manuals on installed units, Occupational Safety and Health
Administration (OSHA) publications, and National Institute of Occupational
Safety and Health (NIOSH} publications.
7.1 ELECTRICAL HAZARDS
Electrical shock is the greatest concern in the operation of an ESP.
During the particle-charging process, high direct-current voltages are gener-
ated by the T-R set and transferred to the discharge electrodes. All portions
of the electrical system outside of the ESP shell must be insulated or isolated
from potential contact. All access points to the electrical distribution
system must be closed and bolted or key-interlocked to limit inadvertent
access to the system. In addition, a clear, legible sign must be attached to
each component indicating the nature of the hazard (i.e., extremely dangerous
high voltage). Areas requiring warning signs include (but are not limited
to) T-R set control cabinets, high-voltage conduit, the T-R set, shell access
SECTION 7-SAFETY _
-------
doors (inlet/outlet plenems, top access doors, insulator compartments, and
penthouse)t hopper access doors, and rapper control cabinets.
Key interlocks are used to assure that the ESP has been deenergized
before personnel enter the ESP or electrical distribution system. The key-
interlock system consists of the use of keys in a number of sequenced steps
that must be followed to deenergize and permit ESP components to be opened.
Each interlock system is especially designed and installed for each unit, and
'the procedure and equipment can be expected to vary from unit to unit. For
specific instructions, the inspector should refer to the manufacturer's
literature.
Most key-interlock systems begin with the main circuit breakers on the
T-R set control panels. Opening the circuit breakers releases a mechanical
interlock that allows keys contained in the circuit breaker mechanism to be
rotated and removed. (See Figure 7-1.) Rotation of the keys locks the
breaker in the open position and prevents its operation. One key is released
for each T-R set control cabinet.
The keys from the control cabinet are inserted into the interlock on the
T-R set grounding switches; one key is used for each T-R set. (See Figure
7-2.) Insertion of-'the key and rotation allows the T-R set grounding switch
to be moved from the operation (closed position) to a ground position. The
switch cannot normally be interlocked in an open position; it must be fully
••closed to a ground position, which completes the electrical circuit from the
deenergized transformer to ground. This position ensures safety by 1) discon-
necting the T-R set from the electrical distribution system, and 2) grounding
the T-R set. Any attempt to energize the T-R set would result in an immediate
.short circuit to ground.
When the T-R set ground switch has been positioned, a second set of keys
located on the T-R set ground switch may be rotated, which locks the switch
in position. Removal of these keys ensures that the switch cannot be moved
into the open or operating position.
Keys from the T-R set ground switch may be placed into the master key
interlocks in the transfer block. After keys from all ground switches are in
place, keys located in the transfer block can be rotated. Rotation of these
keys locks the T-R set switch keys in place and prevents their removal. The
transferblock keys may be removed and used to openaccess doors on the ESP.
SECTION 7-SAFETV
7-2
-------
i
'^*^1®^1^^
Figure 7-1. Control cabinet key interlocks.
7-3
-------
Figure 7-2. T-R set ground switch key interlocks.
7-4
-------
The number of keys needed depends on the number of T-R sets and access
doors on the individual units. Maintenance of the key-interlock system is
critical in any safety program. Locks are equipped with weather-tight covers
(caps, lock boxes, etc.)* and each cover must be replaced after removal and
maintained in that position. Locks should be lubricated (graphite only, no
oils), and preventive maintenance should be performed as necessary to ensure
clean, trouble-free operation.
Under no circumstances should the key-interlock system be bypassed or
short-circuited to gain access to the ESP without proper and complete deener-
gization of the unit. If special studies or-analyses require access to an
energized unit, a complete safety analysis, including proper safety procedures,
must be established for each occurrence.
Following the key-interlock procedure provides two points of electrical
safety: 1) the T-R set main breaker (mechanical lockout), and 2) the T-R set
ground switch (electrical isolation and ground). Many companies further
isolate the unit for positive assurance of grounding of the electrical
distribution system. Access doors are removed from the electrical conduit
between the T-R set and wire frame, and the lead wires are disconnected from
the wire frame.
A potential for nonisolation of the T-R set ground switch exists in many
older designs that use an immersed switch in the transformer fluids. In
these designs, the switch arm mechanically connected to the knife switch with
set screws, clamps, etc. can become loosened with use and will not close
fully. Because the switch is not visible, its position cannot be verified.
As a further safety measure, many companies use a permanent ground switch or
wire that must be thrown or placed in position on the high-voltage bushing in
the penthouse or insulation enclosure before further access is possible.
Placement of this device on an energized T-R results in an immediate short
and a T-R trip.
More modern designs require an air switch on the T-R set, and the contacts
are visible through a window. This ensures that the T-R is grounded and
allows visual verification of the switch position.
Grounding straps bolted to the ESP shell must be provided at each access
point to prevent potential electrical hazards. The strap is attached to the
electrical discharge system nearest the entry point and clamped in place
SECTION 7-SAFETY _ ,.
-------
prior to entry. (See Figure 7-3.) The corona wire-plate system acts like a
capacitor and discharges slowly after a T-R set trip. To prevent potential
shock, the ground strap should be connected to the plate system and discharge
electrode system.
Ground straps should be checked routinely to ensure their continuity,
and the insulated attachment devices (fiberglass or wooden handles) should be
inspected for damage. If exposed to weather, wooden handles can deteriorate;
and if exposed to dust and metal particles, they can become conductive because
the dust and metal particles become imbedded in the surface.
Although rare, it is possible to lock up and energize an ESP with person-
nel inside. The design of the key-interlock system only requires the following
of a step-by-step procedure to energize a unit. To prevent the possible
closure of the unit with personnel inside, a tag or personnel lock system is
required. Central Operation should be notified whenever anyone enters the
unit. Tags should be placed on the main breaker, T-R set switch, and entry
door advising that men are working inside. Entry should never be made alone
without proper notification and proper tagging of the access door. If a
procedure is established for use of a two-man buddy system (one man on the .
outside), that person cannot leave this position without being replaced.
7.2 HOPPER ENTRY
Hoppers present special safety hazards. Although access to an ESP
hopper does not put one in direct contact with the electrical system, a
broken electrode (wire) represents a potential electrical shock hazard.
It is generally recommended that hopper doors be interlocked and that
the dcors be opened only after the unit has been deenergized. For economic
reasons, however, many companies use padlocks instead of a key-interlock
system. In principle, this practice is as safe as the key-interlock system
if proper safety procedures are followed. Workers, however, tend to remove
the lock ind open hopper doors prematurely to cool a unit quickly or to clear
a hopper pluggage. The danger created by opening the doors is not so much
one of electrical hazard as the discharge of hot ash impounded in the hopper.
In the opening of hopper doors, the inspectors must take care to ensure
that no accumulation of collected dust is impounded behind the inner door.
SECTION 7-8AFfTY
7-6
-------
Figure 7-3. Ground clips.
7-7
-------
Before hopper doors are opened, an internal inspection also must he made from
the top of the collecting surfaces to be certain no buildup is present in the
corners of the unit or in the valleys of pyramidal hoppers. Dust that has
accumulated in valleys or corners may break loose during entry into the
hopper and cause minor injury. In some cases, more serious injury or suffoca-
tion car, result from dust falling on the inspector and possibly burying him.
If lower side access doors are available and catwalks (beams, etc.) are
provided between fields, an inspection should be made from the lower level.
Entry into hoppers for purposes other than maintenance should be avoided.
Maintenance first should be attempted from outside of the hopper. If the
hopper must be entered, steps should be taken to dislodge and discharge dust
before such entry. This can be accomplished by mechanical vibration (vibra-
tors, hammers, etc.), poking, prodding, or air lancing. Complete removal can
be accomplished by washing with a high-pressure water hose.
Removal of accumulated dust should be made from the lower catwalk at the
bottom of the ESP; it must not be attempted from inside the hopper. Care
should be taken to ensure that any dust accumulation in the inlet and outlet
plenums (nozzles) is removed. This material can become dislodged, move en
masse into the inlet or outlet field hoppers, and completely fill the hopper.
Finally, before the hopper doors are opened, the inlet plenum should be
checked, and any dust should be moved into the hopper for discharge.
Hopper doors should not be opened during ESP operation because hot ash
could overflow onto the operator. This ash is very fluid, and it can quickly
engulf and severely burn a person. Even after prolonged storage, ash tempera-
tures can be 300° to 400°F in cold-side ESP's and 700° to 800° F in hot-side
units.
All hopper doors should be equipped with safety chains or double latches
to prevent complete opening upon release. This can slow the loss of ash in
the event of accidental opening of a full hopper,
Most hopper inner doors have design features that, if properly used,
will ensure that no door is opened when dust is impounded behind it. First,
a pipe coupling with plug that can be removed should be installed in the
door; this would allow visual verification. Second, a pressure-type latch
should be used that allows a portion of the door seal to be released to
SECTION 7-SAFETY
-------
create a gap between the door and sealing jam. This partial release would
allow accumulated dust to flow out and indicate a partially full hopper
without the possibility of the door opening fully.
A normal practice calls for full discharge of the hoppers before entry
and after each period of dust removal. A full hopper may be determined by
striking the door with a hammer. If the hopper is empty, this will produce a
resounding ring, indicating there is nothing against the inside surface; if
the hopper is full, the blow will produce a dull thud.
A further warning regarding hopper entry involves the use of handgrips,
footholds, etc. in the hopper. Because of the possible dust buildup on
protruding objects, manufacturers have purposely avoided the use of handholds
and footholds in hopper interiors. The steep valley angles and dust layer
create a potential for a fall and injury for persons entering the door.
Because of the angles and small door openings, back injuries are the most
common injury other than abrasions. Outside access equipment (scaffolds,
ladder, handholds, etc.) should be installed to minimize the awkwardness of
hopper-door entry. Installing handholds inside the hopper is also helpful.
If nuclear hopper level detectors are used, the radiation source (beam)
should be shielded from the outside prior to entry. This shielding should be
part of the interlock system for the doors.
Hopper evacuation systems (screws, drag chains, agitators, etc.) should
not be operated when personnel are inside the hopper area or in an area from
which they could fall into the hoppers. Dust accumulation that is discharged
into the hopper can be considered a line bottom with moving equipment. The
dust becomes fluid and results in a nonsolid footing. Scaffolds on which
personnel may be standing can shift and float, and persons inside could
become engulfed in the fluidized ash.
7.3 CONFINED-AREA ENTRY
A confined space is an enclosure in which dangerous air contamination
cannot be prevented or removed by natural ventilation through opening of the
space. Access to the enclosed area also may be restricted so that it is
difficult for.personnel to escape or be rescued. The most common examples of
confined spaces are storage tanks, tank cars, or vats. Depressed areas
SECTION 7-SAFETY
-------
{e.g., trenches, sumps, wells) also may have poor ventilation and considered
confined spaces. An ESP falls under the general definition of confined
space, and as such, requires special procedures and precautions with regard
to entry.
Potential dangers presented by confined space fall into three categories:
oxygen deficiency, explosion, and exposure to toxic chemicals and agents.
Anyone entering the ESP for inspection or maintenance must assess the risks
and potential dangers and follow specific safety requirements for each category.
7.3.1 Oxygen Deficiency
Oxygen deficiency is the most common hazard. Any gas introduced into a
confined space displaces the atmosphere and reduces the oxygen content below
the normal value of 20.B percent. Out-gassing of combustible gases (methane,
H«S, organic vapors, etc.) from collected particulate can result in local
pockets with reduced oxygen levels. Further, application of the ESP to
combustion sources (e.g., utility boilers, industrial boilers, cement kilns,
recovery boilers, incinerators,) produces an atmosphere that is extremely low
in oxygen content (2 to 10 percent). Purging of the unit during cooling does
not always completely replace the flue gases with ambient air, and local
pockets may remain.
Reduction in oxygen pressure below normal conditions has increasingly
severe effects and eventually leads to death. Oxygen levels of less than
16.5 percent result in rapid disability and death. Table 7-1 shows the
effects of reduced oxygen concentrations for various lengths of time. Because
of the subtle effects of oxygen deficiency, the average person does not
recognize the symptoms and may ignore the danger. By the time the person
does recognize the problem, he may no longer be able to remove himself from
the dangerous environment.
7.3.2 Expjosion
Explosive atmospheres can be created in confined spaces by the evaporation
of volatile components or improper purging of the ESP when the process is
shut down. Three elements are necessary to initiate an explosion: oxygen, a
flammable gas or vapor, and a.n ignition source. A flammable atmosphere is
defined as one in which a gas concentration is between two extremes: 1} the
SECTION 7-SAFETY , ,g
-------
lower explosive limit (LEL) and the upper explosive limit (UEL). A mixture
of gas and oxygen in a concentration between these two values can explode if
a source of ignition is present. With regard to ESP inspection and mainten-
ance, explosive gases normally consist of methane, hydrogen, carbon monoxide
(CO), and mixed organic vapors. The gases most commonly present at ESP
shutdown are CO and methane.
Possible sources of ignition include cigarettes, matches, welding,
cutting torches, and grinding equipment. Sparks can also be generated by
static electricity and electrical discharge through grounding straps. The
best means of preventing explosion is to dilute the flammable gas below the
LEL by ventilation. It is not safe to assume that a source of ignition can
be eliminated ano to allow work to continue in a potentially explosive atmos-
phere.
Work in a confined area may release flammable gases that, once released,
can increase in concentration. Constant ventilation should be provided to
maintain the concentration below the LEL.
Because many vapors are htavier than air, pockets of flammable gases
also may develop. An effective monitoring program checks concentrations at
multiple locations and times during the exposure period.
7.3.3 Exposure to Toxic Chemicals and Agents
Depending on the application of the ESP, collected dust may contain
toxic chemicals or harmful physical agents. These compounds may exist in the
system or be created as a result of operations in the confined area. Inhala-
tion, ingestion, or skin contact can have adverse health effects. Most
agents have threshold limit doses below which harmful effects do not occur;
exposure above these threshold doses can cause acute or chronic symptoms,
depending on the compound. A quantitative assessment of each compound and
the threshold dose levels must be made before anyone enters the ESP. Typical
toxic chemicals or species in the ESP environment include CO, hLS, Total
Reduced Sulfur (TRS) gases, arsenic, cadmium, beryllium, lead, alkali, and
acids. If repair work is being conducted, organic solvents, zinc, or cadmium
also may be present. Table 7-1 lists the allowable concentrations of several
compounds in confined spaces for entry to be permissible.
SECTION 7-SAFETY
7-11
-------
TABLE 7-1. EFFECTS OF VARIOUS LEVELS OF OXYGEN ON PERSONS3
Concentration,
percent
Duration
Effect1
20,9
19.5
16.5
12-16
10-14
6-10
below 6
Indefinite
Not stated
Not stated
Seconds to minutes
Seconds to minutes
Seconds
Seconds
Usual oxygen content of air
Minimum oxygen content for oxygen-
deficient atmospheres (OSHA
Standards)
Lowest limit of acceptable stan-
dards reported in literature for
entry without air-supplied res-
pirators
Increased pulse and respiration,
some coordination loss
Disturbed respiration, fatigue,
emotional upset
Nausea, vomiting, inability to move
freely, loss of consciousness
Convulsions, gasping respiration
followed by cessation of breathing
and cardiac standstill.
Data from correspondence of Robert A. Scala, Ph.D., REHD, Exxon
Corporation, March 26, 1974.
Effects - Only trained individuals know the warning signals of a low oxygen
supply. The average person fails to recognize the danger until
he is too weak to rescue himself. Signs include an increased
rate of respiration and circulation that accelerates the onset of
more profound effects, such as loss of consciousness, irregular
heart action, and muscular twitching. Unconsciousness and death
can be sudden.
As noted in Table 7-2, entry may be permitted within certain limitations
provided the person is equipped with appropriate approved respiratory protec-
tion. An assessment of the hazard, concentration, permissible exposure, and
protective equipment must be made before anyone is allowed to enter the
equipment.
SECTION 7-SAFETY
7-12
-------
TABLE 7-2. ALLOWABLE CONCENTRATIONS FOR ENTRY INTO CONFINED SPACES
Agenta
Hydrocarbons
Oxygen
H2S
Carbon monoxide
so2
Without breathing-
air equipment
1% LEL max.
19.5-23.5%
10 ppm max.
30 ppm max.
5 ppm max.
With air-supply
breathing equipment
20% LEL max.
16.5% min.
300 ppm
ZOO ppm
500 ppm
No entry.
permitted
Above 20% LEL
Below 16.5*
>20 ppm
>200 ppm
>10 ppm
If other contaminants are present, the industrial hygienist or REHD should
. be consulted for the appropriate allowable limits.
Work may be performed in oxygen-free atmospheres if backup systems are
available, such as air-line respirators, self-contained breathing apparatus,
, and an emergency oxygen escape pack.
Adapted from: The Industrial Environment - Its Evaluation and Control.
NIOSH, 1973.
Each facility must establish a confined-space entry policy that includes
recognition of the hazards, atmospheric testing and analysis, ventilation
requirements, selection and use of protective equipment, training and education
of personnel, and administrative procedures.
An important component of the policy is recognition of the potential
hazard, which requires complete knowledge of the industrial process and wesh
area. A cursory examination cannot prevent serious deficiencies; a detailed
analysis is recommended.
The second policy component involves ambient air monitoring. An initial
certification of gaseous concentrations must be made before entry is permitted.
This certification must be made by a qualified safety officer with properly
calibrated and maintained equipment. In general, a permit to enter (with a
time limit) may be issued and displayed at the point of entry. Assuming that
oxygen and gas levels do not change with time can be dangerous; an effective
program should include periodic revaluations of concentrations after initial
entry.
Hazard Recognition-
Each worker should be trained in the use of protective equipment, poten-
tial hazards, early warning signs of exposure (symptoms), and rescue procedures
SECTION 7-SAFETY
7-13
-------
(first aid, CPR, itc.)« It is most important that each person be aware that
multiple fatalities can occur if proper rescue procedures are not followed.
If a worker 1s affected within the confined area and cannot remove himself,
rescue personnel must not enter the area without complete self-contained
breathing equipment. If the first worker is affected by an unknown agent, it
is highly probable that rescue personnel will be similarly affected unless
they have the proper protective equipment. Because the causal agent is not
known, maximum protection must be used during the rescue attempt.
Atmospheric Testing and Analysis--
Gas monitoring usually is conducted to determine percent oxygen, percent
lower explosive limit (hexane/methane, heptane, etc.), and hydrocarbon concen-
tration (in parts per million), and carbon monoxide levels (in parts per
million). If hydrogen sulfide or other toxic gases are suspected, additional
analyses may be conducted with detection tubes or continuous gas samplers.
The use of continuous gas samplers with an audible alarm is recommended. The
initial measurements should be performed according to the following suggested
procedures:
1. The ESP to-be entered should be emptied, purged, cleaned, and
ventilated to the maximum extent possible. All entry ports should
be opened to facilitate mixing. All electrical and mechanical
equipment must be locked out and posted. The confined space must
be completely isolated by closing dampers, using guillotine dampers,
or installing blanks.
2. A gas tester should check the vessel's oxygen content, explosivity,
and toxic chemical concentration by first sampling all entry ports
and then sampling inside the space with probes (while he remains
outside). Caution should be used when testing for combustible
gases, as many meters need an oxygen level close to ambient levels
to operate correctly. This is one reason that the space should be
purged and vented before testing. Voids, sub-enclosures, and other
areas where pockets of gas could collect should also be tested.
3. When initial gas test results show that the space has sufficient
oxygen, the gas tester can enter the space and complete the initial
testing by examining areas inaccessible from outside the shell.
He/she should wear an air-supplied positive-pressure respirator
during these measurements. Special care should be taken to test
all breathing zone areas.
4. If the results of the initial tests show that a flammable atmosphere
still exists, additional purging and ventilation are required to
SECTION T-SAFETY _
-------
lower the concentration to 10 percent of the LEL before entry may
be permitted.
5. If testing shows an oxygen-deficient atmosphere or toxic concentra-
tions, all personnel entering the space must use an appropriate
air-supplied respirator.
After the initial gas testing has been performed, dust, mists, fumes,
and any other chemical agents present should be evaluated by either an indus-
trial hygienist or a trained technician. The results will indicate if addi-
tional control measures are necessary. Physical agents such as noise, heat,
and radiation must also be evaluated, and if any are present, the appropriate
control measures (e.g., providing ear protection or rotating employees)
should be instigated.
The specified respiratory protection should be based on the hazard
assessment, I.e., the type of contaminant, its concentration, and the exposure
time. The type of respiratory equipment required for each species is specified
by NIOSH.
Respirators include basic particle-removing devices (dust, aerosol,
mist, etc.), air-purifying respirators (gas, vapors, etc.), and air-supplying
respirators (air-line, self-contained).
7.4 WORKER PROTECTION
Dust collected by ESP's is very fine and usually contains a high percen-
tage of particles with diameters of less than 5 urn. The dust also may.be
sharp-edged or crystalline in nature. All surfaces in the ESP are coated
with dust, and this material can be easily dislodged and suspended during
internal inspections. Thermal drafts and/or cooling fans cause constant
motion of dust particles in the gas stream.
7.4.1 Eye Protection
Eye protection is necessary to prevent dust from entering the eyes.
Goggle-type protection is generally not effective because of the inability of
the frames to form a tight seal against the worker's face. Effective eye
protection consists of full-face protection, a snorkling mask, or eye goggles.
Eyes also may be subject to chemical damage as a result of the dust
composition or species condensed onto the dust particles. The most common
SECTION T-SAFETY 7 -_
-------
active agents are sulfuric acid on fly ash particles and alkali agents in ESP
applications at cement facilities. Each plant should collect samples of ESP
dust and specify eyewash solutions suitable for removing or neutralizing the
active components. Table 7-3 summarizes the kinds of applications where
potential eye hazards may exist.
TABLE 7-3. APPLICATIONS PRESENTING POTENTIAL EYE HAZARDS
Application
Fly ash
Cement
Potential active species
Sulfuric acid
Alkali
PH
Acid
Alkaline
Kraft recovery
Municipal incineration
Copper converter
(NaOK, Na2S04
Sodium sulfate
Sulfuric acid
Hydrochloric acid
Sulfuric acid
etc.)
Alkaline
Acid
Acid
7.4.2 Hearing Protection
" The ESP shell surrounds a large open area, and the metal walls that teno
to reflect and amplify sound energy. When inspectors are inside the unit,
they should use proper hearing protection to limit sound levels to maximum
permitted exposure. Many types of hearing protection devices (cotton, pre-
molded inserts, foam, ear muffs, etc.) are available; selection depends on
individual preference and expected sound levels.
The major sources of sound energy are sonic horns and rappers. Because
of their dangerous level of energy and sound pressure, sonic horns should be
tier1 into the key-interlock system to prevent their activation while persons
are inside the unit.
Activation of either internal failing hammer rappers or external rappers
may become necessary during an inspection and evaluation of an ESP unit. The
impact of these rappers results in high short-term sound levels and dislodges
dust from plates.
Limits of worker exposure to noise are based on both duration? of expo-
sures and sound levels (dBA). Permissible levels for intermittent noise and
SECTION 7-SAFETY
7-16
-------
nonimpulsive levels are presented in Tables 7-4 and 7-5.
TABLE 7-4, MAXIMUM PERMISSIBLE SOUND LEVELS FOR INTERMITTENT NOISE3
{A weighted sound level, dBA)
Total time/8 hours
8 hours
6 hours
4 hours
2 hours
1 hour
4 hour
i hour
8 minutes
4 minutes
2 minutes
Number of occurrences per day
1
89
90
91
93
96
100
104
108
113
123
3
89
92
94
98
102
105
109
114
125
7
89
95
98
102
106
109
115
125
15
89
97
101
105
109
114
124
35
89
97
103
108
114
125
75
89
94 ,
101
113
125
>. 160
89
93
99
117
125
Source; The Industrial Environment - Its Evaluation and Control.
NIOSH, 1973, Page 327.
TABLE 7-5. ACGIH THRESHOLD LIMIT VALUES FOR NONIMPULSIVE NOISE
Duration, hours/day
8.00
6.00
4.00
3.00
2.00
1.50
1.00
0.75
0.50
0.25
Permissible sound level, dBA
90
92
95
97
100
102
105
107
110
115
Source: The Industrial Environment - Its Evaluation and Control.
NIOSH, 1973, Page 327.
SECTION 7-SAFETY
7-17
-------
7,4.3 Skin Irritation
Depending on its composition, the dust collected in the ESP can be
acidic, alkaline, hygroscopic, or abrasive. Skin contact with this dust can
result in burns or irritation. Workers can limit skin contact area and thus
prevent potential irritation by wearing long-sleeved shirts and gloves during
internal inspections. Depending on temperature conditions and activity
levels, coveralls or other full covering may be used.
7.4.4 Thermal Stress
Thermal stress associated with inspections and maintenance of an ESP and
its components must be considered in defining the time required for repairs.
Because of the dusty, humid conditions and limited access, thermal effects
may be severe. Also if the time available for purging and cooling is limited,
entry may have to be made under elevated temperatures.
The thermal stress placed on the worker is a function of several variables,
such as air velocity, evaporation rate, humidity, temperature, radiation, and
metabolic rate (work). In effect, the stress is indicated by the need to
evaporate perspiration.
A Heat Stress Index developed by fielding and Hatch in 1955 incorporates
environmental heat [radiation (R), convection (C), and metabolic {H)3 into an
expression of stress in terms of requirement for evaporation of perspiration.
Algebraically the function may be stated as follows:
• H 1 R JL C = E req.
The resulting physiological strain is determined by the ratio of stress
(E req.) to the maximum capacity of the environment (E max.). The resulting
value is defined as the Heat Stress Index (HSI), which is calculated as;
The values for E req. and E max, are calculated at the maximum exposure
time, based on the HSI defined. Generally, HSI maximum acceptable values are
established for an 8-hour work day.
Table 7-6 indicates expected physiological and hygienic implications of
an 8-hour exposure at various heat-stress levels.
SECTION 7-SAFETY
-------
TABLE 7-6. EXPLANATION OF VALUES IN BELOING AND HATCH HSIa
Index of
Heat Stress
(HSI)
Physiological and hygienic implications of 8-hour exposures to
various heat stresses
-20
-10
0
+10
20
30
40
50
60
70
80
90
100
Mild cold strain. This condition frequently exists in areas
where men recover from exposure to heat.
No thermal strain.
Mild to moderate heat strain. When a job involves higher
intellectual functions, dexterity, or alertness, subtle to
substantial decrements in performance may be expected. In
performance of heavy physical work, little decrement expected
unless ability of individuals to perform such work under no
thermal stress is marginal.
Severe heat strain, involving a threat to health unless
persons ire physically fit. A break-in period required for
those not previously acclimatized. Some decrement in
performance of physical work is to be expected. Medical
selection of personnel desirable because these conditions are
unsuitable for anyone with cardiovascular or respiratory
impairment or with chronic dermatitis. These working conditions
are also unsuitable for activities requiring sustained mental
effort.
Very severe heat strain. Only a small percentage of the popu-
lation can be expected to qualify for this work. Personnel
should be selected by medical examination and by trial on the
job (after acclimatization). Special measures are needed to
assure adequate water ana salt intake. Amelioration of working
conditions by any feasible means is highly desirable, and may
be expected to decrease the health hazard and simultaneously
increase efficiency. Slight "indisposition" that in most jobs
would be insufficient to affect performance may render workers
unfit for this exposure.
The maximum strain tolerated daily by fit, acclimatized, young
men.
Adapted from fielding and Hatch, "Index for Evaluating Heat Stress in Terms
of Resulting Physiologic Strains," Heating, Piping and Air Conditioning,
19S5.
SECTION 7-SAFETY
7-19
-------
A nomograph may be used to evaluate acceptable exposure times under
various conditions. Figure 7-4 shows the methodology for calculating exposure
time. Constants and variables used in the nomograph are as follows:
R - 17.5 (Tw - 95)
C * 0.756 V0'6 (Ta - 95)
E max. « 2.8 V0'6 (42-PWa)
where R = radiant heat exchange, Btu/h
C = convective heat exchange, Btu/h
E max. = maximum evaporative heat loss, Btu/h
Tw = mean radiant temperature, °F
Ta = air temperature, °F
V = air velocity, ft/min
Pk'a = vapor pressure, mm hg
Twb r wet bulb temperature, °F
M = metabolic rate, Btu/h
Tg = globe temperature, °F
The example presented here illustrates the use of the nomograph under
the following conditions: Tg = 130°F, Ta = 100°F, Twb = 80°F, V = 50 ft/min,
M = 2000 Btu/h, and dew point = 73°F.
Step 1. Determine convection. Connect velocity (column I) with air
temperature (Ta) column II and read on column III.
Step 2. Determine £ max. Connect velocity (column I) and dew point
column IV and read E max. on column V.
Step 3. Determine constant K. Connect velocity (column I) with Tg-Td
(column VI) and read K on column VII.
Step 4. Determine Tw. Connect K (column VII) and Tg (column VIII) and
read Tw on column IX.
Step 5. Extend line in Step 4 to column X and read R.
Step 6. Connect R column X with M (column XI) read R + to on column
XII.
Step 7. Connect C (column III) with R + M (column XII) and read E req.
on column XIII.
SECTION 7-SAFETV n^
7-20
-------
MO'
«so
MO-
BO-
HO
ISO
m
JOO-
in
IK
100-
W
13
Irekin linti indi-
cate iolvlten to ttKompla in t«xi-
ft'9k*tl lines end Olio COlvmnl art
MNRbtrid to indicot* iMpi in Ihi
tOlution,
c««r, nut r ' »•
12M
1000;
tec
«CI
X^ '
too-
^*%M
!*'
_^'zt
IOC-
*°,
60
JO
40
30
*
, (-MO)'
(-500).
' (-400).
. (-J001-
' t-JOOi-
I- ^50f
"(-100).
*****'
' ("»)'
^-••w.
. <•»)•
(•>«•
t-'S)'
(.10)
(-»)•
•1900 -
• 1000 JOO •
no iso-
•600
•SOO „;
•400 *•'*'"
•joax^* so-
^r 43.
•JOO JO-
150 •»'
Too 'IJ"
•15 10.0"
52 6.0-
"<0" 4,0-
00 3.0-
•30 2.6-
IJ 1 50-
.,o '•00;
0.50
J
X*
^*
M ••
50
60
&5
•JO <70|-
75 „
82 <»t-
^^^
>lt
W I*"'
11?)-
TO w-
Jl— "
9? <*"'
(«h
9]
(«)•
140
.» :
IJO
11}
uo_-*
^» ^^
105
104
103
103
101
100
-»
'?!
•»7
100
ao
to
JO
40
3C
r
1}
8.0
*fl
5.0
4.0
3.0
3.1
} 0
!,i
1.0
m K
(B), TU#£Wtym »n; TP*EIATye[ t£^j;SE£ IIP<3S*E (N). |tj. M4lL«-^
"uh £t,5> MSl*TlQk (t? / ^rrtsjl T'*--47":
423
900-
J75
IK-
US '
*J
32 JC
— -
^~
\
0
|
^v
1
IJ
9
»A,
V
N
^
rJ5«
24i 9003.
•240 4790
s «
-220 4COC •
a?so
•2)C
1500-
•700
32 SO
' ^-96 9000*
•no ^££s: *75Q
*** MOO-
n/6
• 1*0 -M
-— ~^~~T5eo-
17JO
•IJO |ac
'ni tow-
•100
** MO-
<
•M^
^
•1BO
1?J .
iw ;
143
•160 .
115
,iso*";
us
^
130
115 •
120
•115
no
105 ,
100 .
n ""
S730 -,
•ww - ;
"MOO *"'/'"
J-
'400° »-l^i
-4JOO ^ff* 4.
•4000 j,
•W» fffft*
-------
Step 8. Connect E max. (column V) with E re'q. column XII! and read
allowable exposure time on XIV.
Metabolic rate varies with exertion and work expended, and an estimate
of K must be made for each effort expended in the ESP inspection or repair.
Examples of M for several levels of activity are provided in Tables 7-7 and
7-8.
When the work involves lifting, pushing, or carrying loads; cranking;
etc., the heat equivalent of the external work (W) is subtracted from the
total energy output to obtain heat produced in the body (M).
TABLE 7-7. HEAT PRODUCTION FOk VARIOUS LEVELS OF EXERTION3
Activity
cal/m2h
Sleeping
Sitting quietly
Working at a desk, driving a car, standing, minimum
movement
Sentry duty, standing at machine.and doing light work
Walking 2.5 mph on level, moderate work
Walking 3.5 mph on level, moderately hard work
Walking 3.5 mph on level with 45-lb load, hard work
40
50
80
100
150
200
300
Adapted from: The Industrial Environment - Its Evaluation and Control.
NIOSH, 1973.
TABLE 7-8. METABOLIC BODY HEAT PRODUCTION AS A FUNCTION OF ACTIVITY3
Activity
kcal/h
At rest (seated)
Light machine work
Walking, 3.5 mph on level
Forging
Shovel ing
Slag removal
90
200
300
390
450-600
700
Adapted from: The Industrial Environment - Its Evaluation and Control.
NIOSH, 1973.
SECTION 7-SAFETY
7-22
-------
SECTION 8
MODEL O&M PLAN
Generally, one or more individuals at a plant site have the responsi-
bility of ensuring that an ESP is operated and maintained so that it meets
design particulate matter removal efficiencies and the plant complies with
regulatory emission limits.
Unfortunately, most O&M personnel do not receive in-depth training on
the theory of ESP operation, diagnostic analysis, and the problems and mal-
functions that may occur over the life of the unit. Plant personnel tend to
learn about the operation of the specific unit and to gain operating experi-
ence as a result of day-to-day operating problems. This so-called "on-the-
job" training can result in early deterioration or catastrophic failures that
could have been avoided.
The purpose of this section is to present the basic elements of an O&M
program that will prevent premature ESP failure. This program is not all-
inclusive, and it does not address all potential failure mechanisms. Never-
theless, it ensures the user adequate knowledge to establish a plan of ac-
tion, maintain a reasonable spare parts inventory, and keep the necessary
records for analysis and correction of deficiencies in ESP operation.
The overall goal of an O&M plan is to prevent unit failures. If fail-
ures do occur, however, the plan must include adequate procedures to limit
the extent and duration of excess emissions, to limit damage to the equip-
ment, and to effect changes in the operation of the unit to prevent recur-
rence of the failure. The ideal O&M program includes requirements for rec-
ordkeeping, diagnostic analysis, trend analysis, process analysis, and an
external and internal inspection program.
The components of an O&M plan are management, personnel, preventive
maintenance, inspection program, specific maintenance procedures, and inter-
nal plant audits. The most important of these are management and personnel.
SECTION 8-MODEL OSM PLAN
8-1
-------
Without a properly trained and motivated staff and the full support of plant
management, no O&H program can be effective,
8.1 MANAGEMENT AND STAFF
Personnel operating and servicing the ESP must be familiar with the com-
ponents of the unit, the theory of operation, limitations of the device, and
proper procedures for repair and preventive maintenance.
For optimum performance, one person (i.e., a coordinator) should be
responsible for ESP O&M. All requests for repair and/or investigation of
abnormal operation-go through this individual for coordination of efforts.
When repairs are completed, final reports also should be transmitted to the
originating staff through the ESP coordinator. Thus, the coordinator will be
aware of all maintenance that has been performed, chronic or acute operating
problems, and any work that is in progress.
The coordinator, in consultation with the operation (process) personnel
and management, also can arrange for and schedule all required maintenance.
He/she can assign priority to repairs and order the necessary repair compo-
nents, which sometimes can be received and checked out prior to installation.
Such coordination does not eliminate the need for specialists (electricians,
pipe fitters, welders, etc.) but it does avoid duplication of effort and
helps to ensure an efficient operation.
Many ESP failures and operating problems are caused by mechanical
deficiencies. These are indicated by changes in electrical power readings.
By evaluating process conditions, electrical readings, inspection reports,
and the physical condition of the unit, the coordinator can evaluate the
overall condition of the unit and reconmend process modifications and/or
repairs.
The number of support staff required for proper operation and mainte-
nance of a unit is a function of unit size, design, and operating history.
Staff requirements must be assessed periodically to ensure that the right
personnel are available for normal levels of'maintenance. Additional staff
will generally be needed for such activities as a major rebuilding of the
unit and/or structural changes. This additional staff may include plant
personnel, outside hourly laborers, or contracted personnel from service
SECTION 8-MOOEL O&M PLAN o.o
-------
companies or ESP vendors. In all cases, outside personnel should be super-
vised by experienced plant personnel. The services of laboratory personnel
end computer analysts may also be needed. The coordinator should be respon-
sible for final acceptance and approval of all repairs. Figure 8-1 presents
the general concept and staff organizational chart for a centrally coordinat-
ed O&M program.
As with any highly technical process, the O&M staff responsible for the
ESP must have adequate knowledge to operate and repair the equipment.
Many components of an ESP are not unique, and special knowledge is not
required regarding the components themselves; however, the arrangement and
installation of these components are unique in most applications, and special
knowledge and care is necessary to achieve their optimum performance.
Many plants have a high rate of personnel turnover, and new employees
are assigned to work on an ESP who may have had no previous contact with air
pollution control equipment. To provide the necessary technical expertise,
the source must establish a formal training program for each employee as-
signed to ESP maintenance and operation.
An optimum training program should include the operators, supervisors,
and maintenance staff. Many electric utilities, kraft pulp mills, end cement
plants indicated that more than 70 percent of the ESP problems are non-ESP
related. Changes in operation that affect resistivity, particle size, tem-
perature, and the carbon content of ash entering the unit have a detrimental
effect on removal efficiency. The process operator has control over many of
these variables. An understanding of the cause-and-effect relationship be-
tween process conditions and ESP can help to avoid many performance problems.
Safety is an important aspect of any training program. Each person assoc-
iated with the unit should have complete instructions regarding electrical
hazards, confined area entry, first aid, and lock-out/tag-out procedures.
Thus, a typical ESP training program should include safety, theory of
operation, a physical description of the unit, a review of subsystems, normal
operation (indicators), and abnormal operations (common failure mechanisms),
troubleshooting procedures, a preventive maintenance program, and recordkeep-
The O&K program should emphasize optimum and continuous performance of
the unit. The staff should never get the impression that less-than-optimum
SECTION 8-MODEL O4M PLAN
-------
MANAGEMENT
ESP
COORDINATOR
MAINTENANCE
SUPERVISOR
ELECTRICIANS
Figure 8-1, Organizational chart for centrally coordinated
ESP O&M program.
SECTION 8-MODEL 0&M PLAN
8-4
-------
ESP performance is acceptable. Redundancy is established in the unit solely
to provide a margin of safety for achieving compliance during emergency
situations. Once a pattern 1s established that allows a less-than-optimum
condition to exist (i.e., reliance on built-in redundancy), less-than-optimum
performance becomes the norm, and the margin of safety begins to erode.
To reenforce the training program, followup written material should be
prepared. Each plant should prepare and continually update an ESP operating
manual and an ESP maintenance manual for each unit, A generic manual usually
is not adequate because each vendor's design philosophy varies. The use of
actual photographs, slides, and drawings aid in the overall understanding of
the unit and reduce lost time during repair work.
Training material and courses available from manufacturers and vendors
should be reviewed and presented as appropriate. Further, staff members
responsible for each unit should attend workshops, seminars, and training
courses presented by the Electric Power Research Institute (EPRI), the Asso-
ciation of Pulp and Paper Industries (TAPP1), the Portland Cement Association
(PCA), EPA, and other organizations to increase the scope of the knowledge
and to keep current with evolving technology.
E.Z MAINTENANCE MANUALS
Specific maintenance manuals should be developed for each ESP at a
source. The basic elements of design and overall operation should be spe-
cific to each ESP and should incorporate the manufacturer's literature'and
in-house experience with the particular type of unit. The manual should
relate to the physical aspects of the unit. Descriptions should be brief and
to the point; long narratives without direct application should be avoided.
Figure 8-2 presents a suggested outline for a typical manual. The manu-
al should begin with such basic concepts as ESP description and operation.
It can then continue with a section on component parts which should include
detailed drawings and an explanation of the function of each component and
its normal condition.
The next section covers the internal inspection and maintenance proce-
dure, which is extremely critical in maintaining performance. Periodic
SECTION B-MODEL OIM PLAN
8-5
-------
PREC1PITATOR DESCRIPTION (GENERAL)
1, Collection lone
2. Power Supply
3. Ash Removal System
DESCRIPTION OF OPERATION
1, lonizitlon Concepts
2. Ash Removal (Plates and Wires)
SAFETY
1. Interlock System
2. Tigging Procedure
3. Grounding Rods
COMPONENT DESCRIPTION
I. Collection Plate System
2. Emitter Wire System
3. Power Supply (T-R Set)
4, Linear Reactor
5. Control Cabinet (AVC)
6. Plate Rapper System
7. Wire Rapper System,,. ,
INTERNAL INSPECTION AND MAINTENANCE
1. Electrode Alignment
a. Plate to Wire Clearance
b. Bowed Plates
c. Loose Plate Clips
2, Inlet and Outlet Ducts
a. Ash Build-up
b. Gas Distribution Devices
3. Hoppers
a. Ash Build-up in Hoppers
b. Hopper Heater Operation
4. All Areas
a. Corrosion
EXTERNAL INSPECTION AND MAINTENANCE
1, T-R Sets
B. Oil Leakage
b. Oil Sample
c. Loose Connections
1. Control Cabinet
t. Cleanliness
b. Loose Connections
c. Air Filter
3, Linear Reactor
a. Loose Connections
4. Irsulitors (Support, Antisway,
Bus Duct, T-R Bushings)
a. Cleanliness
b. Cracks
c. Tracking
5. Plate Rappers and Wire Vibrators
a. Check Operation
6. Air Leakage
a. Expansion Jcints
b. Door Gaskets
c. Rapper Rod Penetrations
d. Hoppers
7. Interlocks
a. Check Operation
b. Lubricate
REENERGIZING ESP AFTER MAINTENANCE
1. Air Load
APPENDIX
1. Inspection and Maintenance
Checklist
2. ESP Layout Details
Figure 8-2. Outline for ESP Maintenance ManualJ
(Copyright April 1983, EPRI Report CS-2908, "Proceedings: Conference
on Electrostatic Precipitator Technology for Coal Fired Power Plants."
Reprinted with Permission.)
SECTION 8-MODEL OIM PLAN
8-6
-------
checks are necessary to maintain alignment, to remove accumulated ash de-
posits, and to prevent air inleakage. The section on external inspection and
maintenance includes all supporting equipment, such as T-R sets, control
cabinets, linear reactors, insulators, rappers, etc. Each of these sections
should provide a procedure for evaluating the component. The manual should
identify key operating parameters, define normal operation, and identify
indicators of possible deviations from normal condition. Key operating
parameters include temperature, air pressure, voltage drops, current levels,
or other parameters that can be used to establish the basic operating condi-
tion of the unit.
After evaluation of conditions, a procedure must be presented to replace,
repair, or isolate each component. Unless a proper procedure is followed,
the corrective action could result in further damage to the unit, excessive
emissions, or repeated failure.
8.3 OPERATING MANUALS
Whereas maintenance manuals are designed to facilitate physical repairs
to the ESP, operating manuals are needed to establish an operating norm or
baseline for each unit. Maintenance of the physical structure cannot ensure
adequate performance of the unit because gas stream conditions such as tem-
perature, SO, content, moisture, and gas volume affect the charging and
collection mechanisms.
The operating manual should parallel the maintenance manual in terms of
introductory material so that the operators and maintenance personnel have
the same basic understanding of the components and their function and of the
overall operating theory. Additional information should be provided on the
effects of major operating variables such as gas volume, gas temperature, and
resistivity. The manual also should discuss the effects of air inleakage on
power levels (sparking) and the points where inleakage may occur (hoppers,
doors, expansion points, etc.)- Figure 8-3 presents an outline for an oper-
ating manual.
With regard to fuel combustion sources, the manual should discuss the
effects of such process variables as burner conditions, burner alignment, and
pulverizer fineness, which change the ash particle composition and size
SECTION 8-MODEL O&M PLAN c.7
-------
A. DESCRIPTION OF PRECIPITATE)*
INDICATIONS
• 1, Power Supply
2. Collection Zone
3, Material Removal Apparatus
B. DESCRIPTION OF OPERATION
1. lonliation Concepts
2. Material Removal (Plates and Wires)
C. OPERATIONAL FACTORS
1. Gas Volume
e. Excess Air
b. Air Inleakage
|1) Hoppers
(2) Access Doors
(3) Expansion Joints
(4) Test Ports
(5) Air Heater Leakage
(6) Boiler Inspection Ports
2, Gas Temperature
a. High Temperature
b. Low Temperature
(1) Acid Dew Point
3. Carbon Carryover
a. Burner Conditions
b. Burner Alignment
c. Pulverizer Fineness
d. Oxygen
4, Rapper Hal-functions
a. Mire Rappers
b. Plate Rappers
D. ASH REMOVAL SYSTEM MALFUNCTION
1. Plugged Hopper
2. Low Vacuum
a. Excess Air Inleakage
b. Valves Stuck Open
E. METER READING
1. Normal or High Primary and Secondary Cur-
rent - No Primary and Secondary Voltage
a. Wire Shorted to Ground
(1) Slack Wire
(2) Broken Wire
b. Full Hopper
{1} Vacuum system failed
(2) Bus section open
2. Normal or High Primary and Secondary
Volts - Ho Primary and Secondary Current
a. Open Circuit
(1) T-R Failure
(2) Bus Section Open
3. Fluctuating Voltage and Current
a. Bus Section Swinging
(1) Antisway Insulator Broken
F. STARTUP
1. Safety Check
a. Personnel Clear
b. Ground Straps Removed
2. Ash Removal System On
3. ICVS System On
4. Power On To T-R's
5. Rappers On
6. Hopper Heaters On
G. SHUTDOWN
1. Power Down To T-R's
2. Precipitator Keyed Out
3. Rappers Turned Off 1 Hour After T-R's
4. If Long Outage, ICVS System, Hopper
Heaters, and Ash Removal System
Turned Off
Figure 8-3. ESP Operating Manual outline,1
(Copyright April 1983, EPRI Report CS-2908, "Proceedings: Conference
on Electrostatic Precipitator Technology for Coal Fired Power Plants."
Reprinted with Permission.)
SECTION 8-MODEL O&M PLAN
8-8
-------
distribution. An expected normal range of values and Indicator points should
be established as reference points for the operator.
Another important section of the manual is the one which deals with the
cause-and-effect relationships between meter readings and performance. This
diagnostic section can be generic, in that it provides basic information such
as indications of grounds, swinging wires, etc.; however, it should also
include data that are unique to a specific unit.
Startup and shutdown procedures should be established, and step-by-step
instructions should be provided to ensure sequenced outage of equipment to
aid in maintenance activities and to eliminate startup problems.
8.4 SPARE PARTS
An inventory of spare parts should be maintained to replace failed parts
as needed. Because all components or subassemblies cannot be stocked, a
rational system must be developed that establishes a reasonable inventory of
spare parts. Decisions regarding which components to include in the spare
parts inventory should be based on the following:
1. Probability of failure
2, Cost of components
3. Replacement time (installation)
4. Whether the part can be stored as a component or subassembly (i.e.,
rapper coil vs. rapper, rapper cam vs. cam shaft contacts)
5. In-house technical repair capabilities
6. Available space
The probability of failure can be developed from outside studies (e.g.,
EPRI), vendor reconmendations, and a history of the unit. It is reasonable
to assume that components subjected to heat, dust, weather, or wear are the
most likely to fail. Components of this type are no different from those in
process service, and reasonable judgment must be used in deciding what to
stock. Mechanical and electrical maintenance staff members should be con-
sulted for recommendations concerning some items that should be stocked and
the number required. Adjustments can be made as operating experience is
gained. Items that fall into this category include rappers, drive belts,
SECTION 8-MODEL O&M PLAN
8-9
-------
switch contacts, rapper motor drives, rapper switch cams, rapper boots,
complete insulator heaters, and level indicators.
If a unit is subjected to chronic hopper pluggage problems, antisway
insulators may receive abnormal stress and thus be subject to potential
failure. Also, if the unit is a cycling unit, the probability of insulator
tracking and failure is increased and may warrant the stocking of support
insulators.
Another factor in defining a spare parts inventory is the cost of indi-
vidual components. Although stocking rappers or rapper components is not
costly, stocking a spare transformer can be quite costly. Maintaining an
extensive inventory of high-cost items that have low probability of failure
is not justified.
The time required to receive the part from the vendor and the time re-
quired to replace the part on the unit also influence whether an item should
be stocked. For example, an insulator could be obtained from the vendor in
less than 24 hours. If the shutdown, cooldown, and purge period for the unit
is 18 hours and 2 hours are required to remove the broken insulator, little
time can be saved by stocking the insulator. On the other hand, if the lead
time for a critical part is a matter of weeks or months, or if a component
must be specially built, stocking such items is advantageous.
Many plants have a highly trained electronics and mechanical shop whose
staff can repair or rebuild components to meet original design specifica-
tions. The availability of this service can greatly reduce the need to
maintain component parts or subassemblies. In these cases, one replacement
can be stocked for installation during the period when repairs are being
made. For example, many printed circuit boards can be repaired internally,
which reduces the need to stock a complete line of electronic spare parts.
8.5 WORK ORDER SYSTEMS
A work order system is a valuable tool that allows the ESP coordinator
to track unit performance over a period of time. Work order and computer
tracking systems are generally designed to ensure that the work has been
completed and that charges for labor and parts are correctly assigned for
accounting and planning purposes. With minor changes in the work order form
SECTION i-MODEL Q&M PLAN
8-10
-------
and in the computer programs, the work order also can permit continuous up-
dating of failure-frequency records and can indicate whether the maintenance
performed has been effective in preventing repeated failures. In general,
the work order serves three basic functions;
1. It authorizes and defines the work to be performed.
• 2. It verifies that maintenance has been performed.
3. It permits the direct impact of cost and parts data to be entered
into a central computerized data handling system.
To perform these functions effectively, the work order form must be specific,
and the data fields must be large enough to handle detailed requests and to
provide specific responses. In many computerized systems, the data entry
cannot accommodate a narrative request and specific details are lost.
Most systems can accommodate simple repair jobs because they do not
involve multiple repairs, staff requirements, or parts delays. Major
repairs, however, become lost in the system as major events because they are
subdivided into smaller jobs that the system can handle. Because of this
constraint, a large repair project with many components (e.g., a transformer
failure, control panel repair, or insulator failure) that may have a common
cause appears to be a number of unrelated events in the tracking system.
For diagnostic purposes, a subroutine in the work order system is nec-
essary that links repairs, parts, and location of failure in an event-time
profile. Further, the exact location of component failures must be clearly
defined. In effect, it is more important to know the pattern of failure than
the cost of the failure.
The goal of the work order system can be summarized in the following
items:
0 To provide systematic screening and authorization of requested
work.
0 To provide the necessary information for planning and coordination
of future work.
0 To provide cost information for future planning.
0 To instruct management and craftsmen in the performance of repair
work.
SECTION 8-MODEL OSM PLAN
8-11
-------
0 To estimate manpower, time, and materials for completing the
repair.
0 To define the equipment that may need to be replaced, repaired, or
redesigned (work order request for analysis of performance of
components, special study, or consultation, etc.).
Repairs to the unit may be superficial or cosmetic in nature or they may
be of an urgent nature and require emergency response to prevent damage or
failure. In a major facility, numerous work order requests may be submitted
as a result of daily inspections or operator analysis. Completing the jobs
in a reasonable time requires scheduling the staff and ordering and receiving
parts in an organized manner.
For effective implementation of the work order system, the request must
be assigned a level of priority as to completion time. These priority as-
signments must take into consideration plant and personnel safety, the poten-
tial effect on emissions, potential damage to the equipment, maintenance per-
sonnel availability, part availability, and boiler or process availability.
Obviously, all jobs cannot be assigned the highest priority. Careful as-
signment of priority is the most critical part of the work order system and
the assignment must be made as quickly as possible after requests are re-
ceived. An example of a five-level priority system is provided in Figure
8-4.
If a work order request is too detailed, it will require extensive time
to complete. Also, a very complex form leads to superficial entries and
erroneous data. The form should concentrate on the key elements required to
document the need for repair, the response to the need {e.g., repairs com-
pleted), parts used, and manpower expended. Although a multipage form is
not recommended, such a form may be used for certain purposes. For example,
the first page can be a narrative describing the nature of the problem or
repair required and the response to the need. It is very important that the
maintenance staff indicate the cause of the failure and possible changes that
would prevent recurrence. It is not adequate to simply make a repair to
malfunctioning rapper controls and respond that "the repairs have been made."
Unless a detailed analysis is made of the reason for the failure, the event
may be repeated several times. Treating the symptom (making the repair;
SECTION 8-MODEL O»M PLAN 8-12.
-------
WORKQRDER PRIORITY SYSTEM
Priority Action
1 Emergency repair.
2 Urgent repair to be completed during the
day.
3, 4 Work which may be delayed and completed
in the future.
5 Work which may be delayed until a
scheduled outage.
Figure 8-4. Example of five-level priority system
8-13
-------
replacing parts, fuses; etc.) is not sufficient; the cause of the failure
must be treated.
In summary, the following is a list of how the key areas of a work order
o
request are addressed:
1. Date - The date is the day the problem was identified or the job
was assigned if it originated in the planning, environmental, or
engineering sections.
2, Approved by - This indicates who authorized the work to be com-
pleted, that the request has been entered into the system, and that
it has been assigned a priority and schedule for response. The
maintenance supervisor or ESP coordinator may approve the request,
depending on staff and the size of the facility. When emergency
repairs are required, the work order may be completed after the
fact, and approval is not required.
3- Priority - Priority is assigned according to job urgency on a scale
of 1 to 5.
4, Wprk order number - The work order request number is the tracking
control number necessary to retrieve the information from the
computer data system.
5. Continuing _gr related work order numbers - If the job request is a
continuation of previous requests or represents a continuing prob-
lem area, the related number should be entered.
6. Equ jpment _numb_e r - All major equipment in an ESP should be assigned
an identifying number that associates the repair with the equip-
ment. The numbering system can include process area, major process
component, ESP number, ESP section (i.e., chamber, field, Inlet,
outlet, etc.), equipment number, and component. This numeric iden-
tification can be established by using a field of grouped numbers.
For example, the following could be used:
ID number
XX - XXX - XX - XXX - XXX - XXX
—component
•equipment number
ESP section
—ESP number
process subcomponent
I—process area
SECTION 8-MODEL 0»M PLAN
8-14
-------
If the facility only has one ESP and one process, the first five
numbers (two groups) may not be required, and the entry is thus
simplified. The purpose of the ID system 1s to enable analysis of
the number of events and cost of repair in preselected areas of the
ESP, The fineness or detail of the equipment ID definition will
specify the detail available in later analyses.
7. Description ofwork - The request for repair is usually a narrative
describing the nature of the failure, the part to be replaced, or
the work to be completed. The description must be detailed but
brief because the number of characters that can be entered into the
computerized data system is limited. Additional pages of lengthy
instruction regarding procedures may be attached to the request
(not for computer entry).
8. E_|timated labor - Assignment of personnel and scheduling of outages
of certain equipment require the inclusion of an estimate of man-
hours, the number of in-house staff needed, and whether outside
labor is needed. The more complex jobs may be broken down into
steps, with different personnel and crafts assigned for specific
responsibilities. Manpower and procedures in the request should be
consistent with procedures and policies established in the O&M
manual.
9. Material requirernents - In many jobs, maintenance crews will remove
components before a detailed analysis of the needed materials can
be completed; this can extend an outage while components or parts
are ordered and received from vendors or retrieved^from the spare
parts inventory. Generally the cause of the failure should be
identified at the time the work order request is filled, and spe-
cific materials needs should be identified before any removal ef-
fort begins. If the job supervisor knows in advance what materials
are to be replaced, expended, or removed, efficiency is increased
and outage time reduced. Also, if parts are not available, orders
may be placed and received prior to the outage. Material require-
ments are not limited to parts; they also include tools, safety
equipment, etc.
10. Action taken - This section of the request is the most important
part of the computerized tracking system. A narrative description
of the repair conducted should be provided in response to the work
order request. The data must be accurate and clearly respond to
the work order request.
11. Materials_ replaced - An itemized list of components replaced should
be provided for tracking purposes. If the component has a pre-
selected ID number (spare parts inventory number), this number
should be included.
SECTION 8-MODEL OiM PLAN
-------
Actual manhours expended in the repair can be indicated by work order
number on separate time cards and/or job control cards by craft and personnel
number.
Copies of work orders for the ESP should be retained for future ref-
erence. The ESP coordinator should review these work orders routinely and
make design changes or equipment changes as required to reduce failure or
downtime. An equipment log also should be maintained, and the work should be
.summarized and dated to provide a history of maintenance to the unit.
Figure 8-5 shows a simplified work order request form. Changes in
design for individual applications and equipment must be made to meet site-
specific requirements.
8.6 COMPUTERIZED TRACKING
8.6,1 Work Orders
If the work completed and parts used in the ESP have been entered in the
computerized work order system with sufficient detail, maintenance and man-
agement personnel can evaluate the effectiveness of ESP maintenance.
Preventive maintenance .(.PM) manhours versus repair, manhour.s .also ran,be
'compared to evaluate" the effectiveness of the current PM program. The level
of detail may allow tracking of the impact of PM on particular subgroups
(e.g., rappers, hoppers) as changes are made in PM procedures. The effec-
. tiveness of the PM program may be further evaluated by the required number of
emergency repairs versus scheduled repairs over a period of time (i.e.,
priority 2 versus priority 5, etc.).
Although an evaluation of the frequency of component failures is an
important benefit of computerized systems, improving ESP performance is the
ultimate objective. The actual number of failures generally are not realized
by the maintenance staff when several people are involved in performing
maintenance. The temporal and spatial distribution of failures can provide
insight into the cause of many failures. For example, 1f T-R set trips occur
more frequently during wet, cold weather, the cause could be related to mois-
ture in the conduit or a roof leak. If, however, T-R trips are associated
SECTION B-MODEt O&M PLAN
8-16
-------
*OUIU »T»TKX
W0«< MOMIT
lOuimiNT I AVMLAKE . BATt
ma r«a
I I I t-
iMIHT N*Mt 0* Kit TITli
COOi
' ' '
roccusf CLASS
ocscmmoN of worn rr CKAFT SKILLS
UtN 1 MOuKI
TOT*L
tSTlMnTfO
14AX/HOLBS
MiOUlRIWENT
o*rc »tc D
OtUMA TO
DCICHIPtlOM
1TCCK NO OTY
UUO
SMCUU. loutc moumio
ACCEPTED 1*
Figure 8-5a. Example of work order form.
(Copyright April 1983, EPRI Report CS-2908, "Proceedings: Conference
on Electrostatic Precipitator Technology for Coal Fired Power Plants."
Reprinted with Permission.) g_17
-------
MIT
SYSTEH
JUESTSTEH
WWaT
SUBCWWEKT
MlNTBtWCE RMUEST FORM
0 0 0 0 0 0
Anton TO;
PRIMITV:
BMI:
Bun STATUS;
Foitmm;
0'
BITE:
Jfil SlATUl:
DiTI.
US6S:
NAttxDuRS
COST
Figure 8-5b. Example of work order form.3
(Copyright April 1983, EPRI Report CS-29Q8, "Proceedings: Conference
on Electrostatic Precipitator Technology for Coal Fired Power g_]g
Plants." Reprinted with Permission,}
-------
with hot, humid conditions, the electronic components may be overly sensitive
to temperature. Cyclic events can be correlated with process operation or
fuel changes that modify the ash characteristics.
It should be emphasized that the purpose of the computerized tracking
system is not to satisfy the needs of the accountants or programmers, or to
state that the plant has such a system. Rather, the purpose of a computer-
ized tracking system is to provide the necessary information to analyze ESP
maintenance practices and to reduce component failures, excess emissions, and
outage time. The maintenance staff and ESP coordinator must clearly define
the kind of data required, the level of detail, and the type of analysis'
required prior to the preparation of the data-handling and report-writing
software. Examples of output may be man-hours by department, man-hours by
equipment ID, number of repairs, number of events, number of parts, and
frequency of events. Figures 8-6, 8-7, and 8-8 are examples of outputs from
a computerized- tracking system.
8.6.2 ESP Operating Parameters
In addition to tracking work orders, the computer can be used to develop-
correlations between process and ESP parameters and observed emission pro-
files. Depending on the type of cycles expected in. process operation, the
data may be continuously input into the system or it may be entered from
operating logs or daily inspection reports once or twice a week.
The key data for tracking performance are ESP power levels (i.e., sec-
ondary meter reading by field), opacity (i.e., 6-minute averages), SOg lev-
els, boiler load (or associated parameter proportional to gas flow volume),
flue gas temperature, and fuel quality data (i.e., fuel source, sulfur,
fineness, etc.).
The dampening effect of dust layers oh plates, coal"'bunker hold-up time,
and bunker rat hole problems may prohibit exact temporal correlations. In
many cases, coal changes may have a final effect on power levels several
hours after its introduction into the furnace. Combustion problems (i.e.,
residual carbon) will generally surface immediately because carbon is carried
through the ESP by reentrainment.
SECTION 8-MODEL OiM PLAN 8-19
-------
NUMBER OF REPAIRS
BY DEPARTMENT
LABOR MANHOURS
BY DEPARTMENT
450'
400-
1100
1000-1
!
350'
(/>
o:
5 300-
LU
(31
° 250'
UJ
CO
Z 200-
150
100-
50
n
1
*S8
*A
^
*•&
2*S"
1
«s
1
MS*
1
g
58
IS
3*5*
%
JS
1
•55
ifi
»
1
1
5ft
•ft
v5
aSJ
3ft
S£
I
•^•*
•ft
K»
900-
800-
g 700-
=j
o
z
1 600'
500-
400'
300-
™ 200-
1 ' 100-
*<**
«^»*
«J!jj
1££.. n
(•••i
*••
MHUl
MM
••1
=
JJJJ
•••
i
s
1
^B
i^
S
s
=
•MM
S
•jji
|
^^
jjU'
•MM
--
^
•J™,
^•t i
—
^^
•Mi
•MB
s
1
«••
§
»
s
s
*^
ym
S
|
^^
*••
MMI
••
f=i
•ml
=
• *
-g
L ^B
) OH
i ^B
1 «^
_.
5
^^1
i MHi
rz
.—
MECH ELEC INST
MECH ELEC INST
Figure 8-6. Example of Department Profile.
(Copyright April 1983, EPRI Report CS-2908, "Proceedings: Conference
on Electrostatic Precipitator Technology for Coal Fired Power Plants."
Reprinted with Permission.)
8-20
-------
100-
90
80-
70
j
* 60-
ce:
40-
30'
20-
10
0
vwA
WAS
NO. REPAIRS MAT'L COST MANHOURS
LEGEND
FGD
Figure 8-7. Example of Maintenance Summary.
(Copyright April 1983, EPRI Report CS-2908, "Proceedings: Conference
on Electrostatic Precipitator Technology for Coal Fired Power Plants."
Reprinted with Permission.)
8-21
-------
EQUIPMENT SITE UNIT PUTT IEBUEST Jfil NAME
DCSCDIPIIOM
1*4431
114431
14445!
U443I
U445I
114431
114431
IM432
U4452
114433
114453
114433
114433
114434
U4434
114434
11443)
114435
141455
U4433
114455
114413
144435
114431
'4(454
114454
114431
114431
114431
114457
114437
U4437
114437
14/11/gl 112
IJ/27/81 112
tf/i*/8i t\i
•f/I7/ll 2
If/21/11 12
11/11/82 12
12/14/12 12
11/11/81 112
11/31/81 112
11/27/11 113
11/11/82 112
11/11/82 2
11/17/82 2
11/24/11 2
ff/fl/ll 12
12/11/12 2
ll/D/li 2
ll/ll/S; 2
1I/I9/9I 112
12/22/11 2
12/24/91 ||2
I«/lt/8l 2
11/29/83 112
12/18/91 2
II/II/M 112
11/27/11 12
M/27/81 112
12/11/81 12
i»/n/ar 2
11/24/81 1
ii/ts/ai 112
I7/I2/SI 112
12/12/11 112
I 31171
3 233*1
] S37IA
] 12711
2 34*51
3 55131
2 3IJ84
1 32447
1 32436
2 23?24
2 3444*
1 34814
2 1 7111
4 53711
1 13241
2 imt
2 14f3l
4 S2J43
2 Itlli
] IU47
4 S37II
3 74I4S
2 13781
3 31747
2 21744
1 2177)
2 32543
4 3374J
2 litf?
2 43BU
2 siti4
3
2»f-tl
2*r-ll
I2-AF-I1 HOPPEI TOP 6MI
2AFII
2W-I1
2AF II
2Af-I2
2Af-«2
2AF-U
2ir-Il
12 AF-13 HOPPER VCHT VHVC
2«f 13
2AH4
2HF-U
12-AF-H MGPPEi
2*f-l5
JlfMS
2AF-I5 HOOPER
12 «F 13 NUV« FEEDER
2*f-U
2AF-I5
:»F 13 IOTTOH STONE
IJdF-U VINT LIKE
2AF-U »\l
-------
8.7 V-I CURVES
Air-load tests and gas-load V-I curves are a fundamental part of ESP
maintenance and troubleshooting. The V-I curve represents the voltage-
current relationship on the secondary side of the T-R set. Details of the
shape of the curve, corona initiation, and the termination point are con-
trolled by the physical condition of the ESP and gas stream conditions.
Preparation of a baseline or reference V-I curve for each ESP field under
ideal or optimum conditions will permit evaluation of changes in internal
conditions and operating conditions at a later date.
Two types of V-I curves can be produced--air-load or gas-load. The
air-load test, as implied by the name, is prepared when the ESP is not re-
ceiving particulate matter or hot gas from the process. The gas-load test is
conducted when the ESP is receiving flue gas and particulate (i.e., during
normal operation). Procedures for producing and utilizing both types of V-I
curves are discussed in Sections 3.3 and 4,1, and will not be repeated here,
8.8 PROCEDURES FOR HANDLING MALFUNCTION
Many malfunctions are of an emergency nature and require prompt action
by maintenance staff to reduce emissions or prevent damage to the unit. On
some units, predictable but unpreventable malfunctions can be identified;
such malfunctions include hopper pluggages, wire breaks, T-R set trips,
shorts, and conduit arcing. These problems as well as corrective actions are
discussed in Sections 4.2 and 4.3.
An effective OSM program should include established written procedures
to be followed when malfunctions occur. Having a predetermined plan of
action reduces lost time, increases efficiency, and reduces excessive emis-
sions. The procedures should contain the following basic elements; malfunc-
tion anticipated, effect of malfunction on emissions, effect of malfunction
on equipment if allowed to continue, required operation-related action, and
maintenance requirements or procedure. Examples of malfunction procedures
are presented here.
SECTION S-MODEL O&M PLAN g-?3
-------
8.8.1 Example 1 - Foreseeable Malfunction
An example of a foreseeable malfunction that cannot be prevented is hop-
per pluggage. Although the short-term effect on emissions is not critical,
if the pluggage is allowed to continue, the ash will overflow the hopper and
back up into the electric field. This will cause a trip-out of the T-R set,
which in turn, may cause increased emissions. Although the T-R trip does not
cause damage initially, overflow into the electric field can cause a movement
in plate or wire frame alignment. Because the impact of the hopper pluggage
increases while the time required to clear the problem increases, the proce-
dure for taking action should depend on the extent of the problem. A T-R set
trip should not be the first indication of hopper pluggage. Pluggages should
be identified through operator inspection, evacuation vacuum charts, etc.
If the problem can be cleared before the high hopper level alarm sounds,
operations will not be involved. If the pluggage is significant, operations
may be requested to reduce the load (i.e., reduce the participate matter
collection rate) or deenergize the field to reduce participate matter collec-
tion. Obviously, modifications will be made to the procedure depending on
the location of the hopper pluggage (inlet/outlet), level of dust in the
hopper when the pluggage is identified, and the time required to make re-
pairs,
8.8.2 Example 2 - Internal Malfunction
When an ESP malfunction requires internal repairs, one chamber of the
ESP may have to be deenergized. For emission reduction through the remaining
chambers, a process reduction may be specified to keep the superficial veloc-
ity near design values.
8.8.3 Example 3 - ExtensiveMalfunction in T-R Set
A T-R set malfunction that involves extensive repairs (rectifier/trans-
former, etc.) taking several weeks or months would require a field to be out
for an extended period, which would result in increased emissions. If adja-
cent T-R sets have sufficient capacity (secondary current), the isolated
field can be jumped to an active energized section. This jump is made inter-
nally between wire frames.
SECTION 8-MODEL O4M PLAN B-24
-------
8.8-4 Example 4 - Isolati onof Portion of T-RSet
Malfunction procedures should include the isolation of shorted elec-
trical sections down to the smallest possible area. This may involve the
isolation of one-half of the T-R set if it is double-half or full-wave,
Records should be maintained showing any isolation, jumping or other pre-
ventive measure that may have been taken.
SECTION B-MODEL OIM PLAN 8-25
-------
REFERENCES FOR SECTION 8
1. Vuchetich, M. A., and R. J. Savoie. Electrostatic Precipitator Training
Program and Operation and Maintenance Manual Development at Consumers
Power Company. In: Proceedings Conference on Electrostatic Precipitator
Technology For Coal-Fired Power Plants. EPRI CS-2908 - April 1983.
2. Rose, W. 0. Fossil Maintenance Documentation at Duke Power Company.
In: Proceedings Conference on Electrostatic Precipitator Technology For
Coal-Fired Power Plants. EPRI CS-2908 - April 1983,
3. Lagomarsino, J.,~and P. Goldbrunner. Computer Compilation of
Particulate Control Equipment Maintenance Records. Burns and Roe Inc.
In: Proceedings Conference on Electrostatic Precipitator Technology For
Coal-Fired Power Plants. EPRI CS-2908 - April 1983.
SECTION B-MODEL Q&M PLAN
-------
APPENDIX A
ESP APPLICATIONS IN CEMENT INDUSTRY
BACKGROUND
The manufacture of Portland cement is basically a calcining operation in
which rock ind earth are mixed and burned slowly in a rotating kiln. The
burning drives off carbon dioxide (CO,) and leaves a clinker product that is
<- i
composed primarily of calcium silicates. The basic raw materials used in
Portland cement manufacturing are calcium carbonate, silicon oxide, alumina,
ferric oxides, and small amounts of sulfate, alkali, and carbonaceous mate-
rials.2
The rock and earth from the quarry are crushed, screened, and ground to
the appropriate size for mixing and blending before they are fed (dry or as a
slurry) into the back or feed end of the kiln. Dry process feed usually
contains less than 1 percent water by weight. A typical wet process feed
contains 34 to 40 percent water by weight. The electrical characteristics of
the particulates generated by these two processes vary widely.
Figure A-l is a flow diagram of the Portland cement process. As shown,
crushing sometimes takes place in two or three stages. Crushing, screening,
and grinding operations are typically vented directly to the atmosphere and
are potential sources of particulate matter emissions. The emission rate
depends on the kind of raw material and its moisture content, the charac-
teristics of the crusher, and the kind of control equipment used (usually
fabric filters) and its operation and condition.
The rotary kiln is the major source of emissions at a portland cement
plant. The rotary kiln has three stages of operation: feed, fuel burning
and clinker cooling and handling. The raw materials are fed into an ele-
vated and inclined refractory-lined steel cylinder that rotates at approxi-
mately 50 to 90 revolutions per hour. As the kiln rotates, its slightly
APPENDIX A-ESP APPLICATIONS IN CEMENT INDUSTRY A-l
-------
ruti
,
.
UJ
GYPSUM
T I
PRODUCT
SI WAGE
UJJ
TO TRUCK.
BOX CAR
PACKAGING
L
1 i
•
\
RR (
I
fN)
Figure A-1. Schematic diagram of Portland cement process flow.
-------
inclined position causes the feed to travel slowly downward and be exposed to
increasing heat. Water is evaporated from the feed with the aid of heat
exchangers. As the temperature increases, organic compounds are volatilized.
At about midsection of the kiln, calcium and magnesium carbonates are decom-
posed and CO- is liberated. Calcium oxide and magnesium oxide are also
formed. At 2700°F, approximately 20 to 30 percent of the charge is converted
to liquid. It is while the charge is in this state that the chemical reac-
tions proceed and the material turns incandescent.
The kiln consumes large quantities of fuel and is a large source of
particulate matter emissions. Design features that reduce particulate
matter emissions include the use of larger kiln diameters at the feed end and
the addition of suspension preheaters..
Depending on its alkali content, the dust collected in the initial
stages of the kiln operation often can be returned to the kiln. This reduces
disposal problems and effects a cost savings in raw material. The dust may
be returned directly by mixing it with the kiln feed and introducing it in a
parallel feeder. Another method of returning the dust is by insufflation;
the dry dust is returned to the burning zone either through the fuel pipe or
2
by a separate pipe running parallel to the fuel pipe.
The clinker from the kiln rolls into a clinker cooler. As the clinker
cooler reduces the temperature of the clinker, it recovers the heat from the
clinker to preheat the primary or secondary combustion air in the kiln.
After the clinker is cooled, it may be taken to a storage area or transferred
to finishing mills. The finishing mills are usually rotary ball mills.
Sometimes these mills are sprayed with water to keep them sufficiently cool
and to minimize dehydration of the gypsum, which is added at a rate of 5
percent, to control setting times.
Although the design of the equipment used in the cement process is not
complex, several design features or operating characteristics can affect the
amount and quality of material found in the effluent. These include:
c Extra fine grinding in the mills seems to generate more particulate
matter in the effluent.
D Higher speeds of kiln rotation tend to generate more particulate
matter.
APPENDIX A-ESP APPLICATIONS IN CiMENT INDUSTRY
-------
0 The type and size of chain section used in the kiln (as a heat
exchanger and to provide intimate mixing of the wet slurry and
waste gases) can either increase or reduce the quantity of released
dust. Short sections of dense curtain chains might help reduce
dust generation, whereas loop systems might allow more to be
released.
0 More draft than required tends to generate more dust carryover,
0 Long dry-process kilns usually contain a minimal chain section, but
flue gas temperatures are usually controlled at the back end by
water sprays and dilution air on the older units.
° Recently, the industry has begun to use several new types of heat
exchangers (cyclone preheaters, for example) on dry-process rotary
kilns; these new exchangers preheat the raw feed by intimate mixing
with the waste gases,
0 Insufflation (the method of returning collected material back into
the kiln by blowing it into the burner flame or including it with
the pulverized coal) has been processed less frequently in recent
years. Although beneficial in some ways, this method has caused
some difficulties in the ESP by increasing the circulation of dust
concentrations and causing higher alkali content.
PROCESS VARIABLES AFFECTING ESP PERFORMANCE IN CEMENT INDUSTRY
The major application for ESP's in the cement industry is for collection
of the particulate matter leaving the feed or back end of the kiln. Both
wet- and dry-process rotary kilns have successfully used ESP's. Several key
process varieties, however, effect ESP performance. These include:
0 Concentration of particulate matter in the kiln gas
0 Size distribution of the waste dust
0 Moisture content of kiln gases
0 Gas temperature of kiln gases
0 Alkali and chloride content of the particulate matter in the kiln
gas,
Based on a number of field measurements of waste dust in the discharge
gases of wet process kilns, the particulate emissions range from 1.5 to 6.0
gr/acf or 30 to 50 Ib dust per barrel of clinker. Emissions from the dry
APPENDIX A-ESP APPLICATIONS IN CEMENT INDUSTRY A-4
-------
process will normally be 10 to 20 percent higher at similar-kiln production
rates.
Large particles leaving the back end of the kiln are usually separated
out in the dust plenum, which serves as a drop-out chamber. Thus, the mater-
ial entering the ESP is finer in size than fly ash, but it exhibits a similar
heterogeneous distribution.
Moisture levels in the wet process gas stream closely approximate the
water percentage of the slurry in a tight system. Moisture in the dry pro-
cess kiln depends primarily on the quantity of water spray conditioning used
to control back-end temperatures. Moisture levels are also increased by the
combustion of'fuel. If the moisture level rises above 10 to 20 percent,
care must be taken to maintain the ESP temperature above the dewpoint to
avoid potential corrosion problems. Where precalciners are used, however,
the moisture level should be approximately 4 to-5 percent to ensure adequate
ESP performance.
The major effects of temperature are reflected in the modification of
the electrical characteristics and the reactions of the particles as they are
deposited on the plates. The electrical characteristics of almost all parti-
culate matter collected by an ESP vary widely in the temperature range of
200° to 750°F. At the lower end of the range, electrical characteristics are
affected by condensation and surface leakage; at the higher end of the range,
they are affected by conductivity changes in the bulk material. The real
effects at any given temperature depend on the moisture level and chemical
composition of the particulate matter. The greater concern, however, is
whether the ESP is operating in the critical temperature zones for effective
control of the particulate matter. The most critical range for ESP cement
applications is 350° to 400°. Within these critical zones, electrical read-
ings may vary with temperature shifts as small as 10° to 15°F.
The alkali compounds (potassium and sodium), sulfur, and chloride are
considered the volatile components. When the alkali material in the feed
slurry is heated to approximately 1000° to 1500eF, it vaporizes and becomes
entrained in the kiln combustion gases. As the gases reach the feed end of
the kiln, their temperature is lower (because of the heat exchange with the
chains and the heat loss resulting from water evaporation in the slurry) and
APPENDIX A-ESP APPLICATIONS IN CEMENT INDUSTRY A-5
-------
the volatile compounds condense on the feed slurry nodules. The normal
movement of the slurry down the kiln returns the volatiles to the hotter area
of the kiln, where they are revaporized. This vaporizing/condensing cycle
continues until an equilibrium is reached between the feed alkali and the
final loss from the kiln through the clinker product and the particulate in
the exhausted combustion gases.
EFFECTS OF PARTICULATE CHEMISTRY ON ESP PERFORMANCE
The condensed alkali compounds appear as submicron size or fine parti-
cles, composed primarily of potassium hydroxide, potassium chloride, potassium
sulfate, sodium hydroxide, sodium sulfate, and sodium chloride. The specific
chemistry of the alkali compounds depends on kiln temperature, slurry chem-
istry, and back-end temperature. Chemical properties of typical volatile
components are provided in Table A-l and Figure A-2, The volatile compounds
also condense on larger particles that are entrained by the flue gases in the
drying slurry feed materials. These larger particles, which are primarily
CaCQ,, CaO, and SiO*. are chemically close to the composition of the feed
slurry.
TABLE A-l. CHEMICAL PROPERTIES OF VOLATILE COMPONENTS OF PARTICIPATES
FROM CEMENT KILN PROCESSING
Compound
Oxide
Carbonate
Sulfate
Chloride
Hydroxide
K
Melting
point, °C
Decomp.
894
1074
768
360
Boiling
point, °C
350
Decomp.
1689
1411
1320
Na
Melting
point»°C
Sub! im.
850
884
801
328
Boil ing
point, QC
1275
Decomp.
1440
1390
The emissions from the kiln typically form a bimodal particle size
distribution with a submicron size fraction and a supermicron size fraction.
Although an ESP is effective in collecting both particle size fractions, the
probability of collecting the larger particles is greater. The larger nonal-
kali particles are collected in the front fields of the ESP, whereas the
APPENDIX A-ESP APPLICATIONS IN CEMENT INDUSTRY
A-6
-------
soo
76Q j-
700
600
500
400
300
200
100
700 800 900 1000 1100 1200 1300I40Q
TbUu
ISOO'F
2700°F
TEMPERATURE,"C
Figure A-2. Chemical properties of alkali volatile materials
in cement kiln processing.
A-?
-------
harder-to-collect finer alkali particles tend to pass through the inlet
fields. A typical analysis of the dust found in ESP hoppers shows that the
alkali content increases from inlet to outlet fields.
Table A-2 presents typical composition data for successive fields on a
wet-process kiln ESP, Figure A-3 shows the decrease in nonalkali particu-
lates {CaCTand SiO) as a percentage of total dust composition in successive
fields of a typical ESP. The enrichment of alkali particulates (Na20, KJ),
and S03) results primarily from the high removal rate for nonalkali materials
in the inlet field. The close association of KgO and SO. is expected tecause
a major portion of the alkali particulates is composed of potassium sulfate
(Na2$04). This increasing percentage is not always found, especially if most
of the alkali condenses onto larger nonalkali particles. In this case, the
particle size/chemistry is more homogeneous and the ESP does not segregate
the dust chemically by fields.
TABLE A-2. TYPICAL COMPOSITION OF DUST COLLECTED IN SUCCESSIVE FIELDS
OF AN ESP SERVING A WET PROCESS CEMENT KILN (ROCK FEED)1*
Chemical
compound
Na20
K20
Li20
MgO
CaO
A12°3
Si02
Fe2°3
LOD
so3
Ti02
Composition, %
Inlet
0.47
5.80
0.34
0.41
41.98
7.98
13.48
1.84
19.91
6.84
0.26
Field 1
0.50
7.00
0.24
0.85
43.26
6.15
12.80
1.90
18.96
7.19
0,25
Field 2
0.74
12.05
1.00
1.09
39.41
2.14
11.72
1.84
17.89
11.15
0.22
Field 3
0.98
19.80
1.64
0.20
29.09
2.15
8.76
1.44
14.08
20.31
0.17
Field 4
1.72
35.80
2.16
0.12
8.19
1.75
3.72
0.63
7.85
37.03
0.06
APPENDIX A-ESP APPLICATIONS IN CEMENT INDUSTRY
A-8
-------
50
45
40
30-
20-
10-
INLET
1 2 3
FIELD HUHBER
Figure A-3. Selective removal of nonalkali particulates in inlet fields
of ESP and enrichment of alkali dust in outlet fields.5
A-9
-------
The chemistry of the participates also affects their resistivity.
Soluble alkali components (Na9SO. and Nad ) have proved to be effective in
c t
-------
10
100 200 300
GAS TEMPERATURE,°C
Figure A-4. Resistivity of dust from cement making processes.
(©1978 IEEE.)
A-11
-------
v
o
»
>-
I13
3 IF
4
2
10
1/1
'$
6
4i
10
10'
25 - 40% MOISTURE
@ 4,000 VOLTS
25% MOISTURE
40% MOISTURE
200 300 400 500
TEMPERATURE.°F
600
700
Figure A-5. Typical resistivity of cement dust collected at the inlet
of an ESP serving a wet-process cement kiln (rock feed) as a
function of moisture content.5
A-12
-------
10
13
8
6
10
12
B
6
4
o
I
o
IS)
IS)
yj
ct
10
*l
6
4
10
10
25 - 40% MOISTURE § 4,000 VOLTS
• 25X MOISTURE
A 40X MOISTURE
I I I I
200 300 400 500
TEMPERATURE,°F
600
700
Figure A-6. Typical resistivity of cement dust collected in the first field
of an ESP serving a wet-process cement kiln frock feed) as a function
of moisture content.5
A-13
-------
10
13
ft
6
-"s
6
4
u
o
ft
>-
f-
*—<
>
*—<
IS)
2
jll
8
6
4
,o'8'
6
4
108
6
4
10
25 - 403 MOISTURE 0 4,000 VOLTS
» 25« MOISTURE
A 40% MOISTURE
_] I I I
200 300 400 500
TEMPERATURE,8F
600
700
Figure A-7, Typical resistivity of cement dust collected in the second field
of an ESP serving a wet-process cement kiln (rock feed) as a function
of moisture content.5
A-14
-------
10
13
u
o
>-
>12B
6
4
2
O11
8
6
4
*l
6
10!
10
MOISTURE % 4,000 VOLTS
* 25« MOISTURE
A 40% MOISTURE
i I I
200 300 400 500
TEMPERATURE,«F
600
700
Figure A-8. Typical resistivity of cement dust collected in the third field
of an ESP serving a wet-process cement kiln (rock feed) as a function
of moisture content.^
A-15
-------
10
13
8
ft
10
E
u
I
o
.12
B
6
4
2
311
8
6
4
10
2
10
8
6
10
10
25 - 40% MOISTURE 6 4,000 VOLTS
• 25% MOISTURE
A 403 MOISTURE
200 300 400 500
TEMPERATURE,°F
600
700
Figure A-9. Typical resistivity of cement dust collected in the fourth field
of an ESP serving a wet-process cement kiln (rock feed) as a function
of moisture content.*
A-16
-------
500
400
I 30°
'£ 200
*•
«c
100
0
i i i i i
45 50 55 60 65 70 75 80
YEAR INSTALLED
Figure A-10. Design SCA vs. year installed for wet-process kilns.
700
600
•i 500
u
S 400
o
"c 30°
v.
-------
decreased (Figure A-ll). The purpose of larger SCA's was to increase collec-
tion efficiency. Figures A-12 and A-13 present the design collection effici-
encies versus SCA's for wet- and dry-process kilns. In general, SCA's in the
range of 300 to 400 ft /1000 acfm are needed to achieve 99+ percent control
efficiency.
OPERATING PRACTICES THAT AFFECT ESP PERFORMANCE
The grinding heat in a cement mill is almost equal to the power consump-
tion of the mill motor. Assuming that the clinker temperature is the same as
that of the finished product, all of the grinding heat has to be removed.
Approximately 20 percent of this heat may be dissipated by radiation; the
remaining 80 percent is usually removed by using air as the cooling mediums
by evaporation of water injected into the mill, or by a combination of these
methods.
If air only is used for cooling, a great amount of air is required.
Because the air will be dry, electrostatic precipitation will be less effec-
tive. Injecting water into the mill for cooling and aiming at a water content
in the vent air corresponding to a dewpoint of 60°C can reduce the required
air volume by a factor of 5. Operating conditions for the ESP are ideal at a
dewpoint of 60°C,7.
At the startup of a cold mill, no water injection is possible until the
cement temperature has risen above 110°C, During this time, ESP performance
is often less than ideal because the dewpoint of the air from the mill is too
low. When this lower efficiency is not acceptable, a special method can be
used that automatically keeps the dust loss low during the startup period,
The method Involves reducing the air flow through the mill by approximately
40 percent during the period with no water injection. The principles of this
method are as follows:
0 The ESP is energized before the fan and mill are started; this
suppresses Initial dust puffs.
0 When the fan is started, the air flow through the mill is reduced
to the minimum required to keep the mill inlet dust-free. This
reduces the air velocity in the mill and ESP and lowers the dust
concentration.
APPENDIX A-ESP APPLICATIONS IN CEMENT INDUSTRY A-18
-------
100
99
98
>: 97
z
LU
G 96
i—i
u.
u.
w 95
94
93
200
300
400
SCA.ftVlOOO a fern
500
Figure A-12. Design collection efficiency vs. SCA for wet process kilns,
200 250 300 350 400 450 500 550
SCA.ft2 /1000 afcm
Figure A-13. Design collection efficiency vs. SCA for dry process kilns.
A-19
-------
0 Startup of the ESP rapping gears is delayed to avoid dust puffs
during periods of difficult operating conditions (when dewpoint is
low). After a few minutes of operation, the dewpoint rises as a
result of evaporation of water from the gypsum and the rapping can
begin.
0 When the water injection starts, the air flow through the mill is
immediately regulated up to a level corresponding to the desired
dewpoint for normal cement mill operation, i.e., 60°C,
0 An automatic spark rate controller on the high-temperature recti-
fiers keeps the ESP voltage at optimum level during the varying
operating conditions.
0 The ESP remains energized and the rapping gears continue to operate
for a certain period after the cement mill is stopped; this cleans
the air drawn through the system by natural draft.
Figure A-14 shows results from tests run out at a Danish cement plant that
used the method just described. Because dust emissions remain low during the
startup period in spite of the reduced dewpoint, ESP migration velocity is
reduced, primarily because of the reduced air flow through the ESP and the
low inlet dust concentration.
Because the resistivity of the dust after a cyclone preheater kiln is
usually quite high, a very large ESP is required if the gases are to be
treated without water conditioning. Therefore, water conditioning is the
rule because it reduces the gas volume, reduces the resistivity of the dust,
and increases the dielectric strength of the gas. For these reasons, water
conditioning has a very pronounced, positive effect on ESP performance. This
is directly reflected in the operating voltage, as illustrated in Figures
A-15 and A-16.
Figures A-15 and A-16 show the effect of water conditioning on ESP
current-voltage characteristics for a preheater kiln with a conditioning
tower and for a preheater kiln with a raw mill. As the water conditioning
increases and the gas cools, the ESP voltage rises dramatically and perform-
ance improves. Over a wide temperature range, the current-voltage character-
istics are almost vertical, or even curve back, which indicates "back-corona"
due to high resistivity.
Figure A-17 shows a preheater kiln with a conditioning tower and a raw
mill installed in parallel. In this example, the major part of the hot kiln
APPENDIX A-ESP APPLICATIONS IN CEMINT INDUSTRY A-20
-------
nute
200
180
160
140
120
100
80
60
40
20
n
h- ' ' feOO
DUST, INLET, g/m3
_
CEMENT TEMP. ,•£,
-— •••/ _^~- '
jr DUST OllfLET, mg/Ti»3
JTO
3
AIR FLOW, m /min.
30Q.
»-^*^*
^— — " ' j loo-
__^c J
i i i_n-
-100
RO
.60
-40
-20
n
60
120
MINUTES
Figure A-14. Automatic control of operating conditions
during startup of cement mill.7
(©1978 IEEE.)
A-21
-------
Csl
E
0.4
0.2
0.1
185«C 165*C 150°C
tt 0.25
0.125
125°C
15 20 25 30 35 40
ESP VOLTAGE, kV
45 50
Figure A-15. ESP current-voltage characteristics with varying H-0
conditioning of gas from cyclone preheater kiln with a conditioning tower.
(©1978 IEEE.)
C\J
E
0.4
0.2
£ 0>1
UJ
§ 0.05
0.025
0.0125
o.
1/1
25 30 35 40 45
ESP VOLTAGE, kV
SO 55
Figure A-16. ESP current-voltage characteristics with varying
0 conditioning of gas from cyclone preheater kiln with raw mill.
(©1978 IEEE.)
A-22
-------
350°C
FEED
CONDITIONING TOWER
1050'C
"(90°C)
Figure A-17. Cyclone preheater kiln with conditioning tower and raw
(©1978 IEEE.)
A-23
-------
gas is drawn through the raw mill, where the gas obtains moisture from the
raw materials. The cooled gas from the raw mill is mixed with the remaining
part of the kiln gas, which has been cooled in the conditioning tower, and
the gas mixture enters the ESP with an ideal operating temperature and moisture.
When the raw mill is stopped, all the kiln gases pass through the conditioning
tower, where they are cooled, and acceptable ESP performance is obtained,
Thus, good ESP performance can be maintained with the mill in operation or
with the mill stopped.
Unstable ESP operation and considerably reduced efficiency can occur,
however, in the transition phases between the two modes of operation, espe-
cially in connection with startup of the raw mill. When the mill is started,
the major part of the gas is diverted from the conditioning tower to the
mill. Because the amount of water injected and evaporated in the condition-
ing tower is controlled by the gas temperature at the tower outlet, it is
reduced in proportion to the reduction in gas flow through the tower. This
reduction might be accompanied by a minor or perhaps even a major fluctuation
of the gas temperature at the tower outlet, depending on the properties of
the conditioning tower's automatic regulation equipment. The hot gases drawn
through a cold mill .must heat'the mill and the-raw materials before full-
level evaporation of water from the mill is reached. The result is a tempo-
rary humidity deficiency in the gas stream from the mill, which, when com-
bined with possible fluctuations of the temperature of the gas stream from
the conditioning tower, might result in serious deterioration of the ESP
performance during and after the changeover phase.
One method that has been used to overcome this problem is to increase
the water injection in the conditioning tower while the mill is heating up.
This can be accomplished by an automatic, temporary displacement of the set
point of the temperature regulator controlling the water injection. For
example, the displacement might be of the magnitude -25°C. This method pre-
supposes that the conditoning tower has sufficient cooling capacity to avoid
the sludge formation that would normally accompany such a temperature lower-
ing. Another method is to preheat the raw mill by drawing a small hot gas
stream through the mill before the startup. A third method involves the use
of automatic control and synchronization of damper movement, water injection
APPENDIX A-ESP APPLICATIONS IN CEMENT INDUSTRY • A-24
-------
changes, and mill startup. This method was successfully introduced at a
Greek cement plant where increased dust emissions during changeover periods
was a problem. Figure A-18 illustrates the results obtained in this case.
The dust from a grate cooler usually has a high resistivity, and the
excess air has a low moisture content; nevertheless, the fairly large particle
size of the dust makes it easily precipitable, especially if the ESP perform-
ance has been improved and stabilized by increasing the moisture content of
the excess air. A few percent moisture by volume is~sufficient, and such
small quantities of water can be injected without difficulty and evaporated
in the grate cooler above the grate at the cool end of the clinker bed.
Occasionally, however, a grate cooler may be subject to large opera-
tional variations. Rings may form in the kiln and dam up the mix, and when
the ring breaks down, excessive quantities of materials flush through the
burning zone and enter the grate cooler. When this occurs, the temperature
of the excess air may rise to 400° to 425°C.
Compliance with emission standards during periods of temporary unstable
cooler operation requires the control of ESP operating conditions. Planning
an appropriate control strategy requires detailed quantitative knowledge of
the interactions between the operating parameters, (gas temperature, moisture
content and resistivity) and ESP performance. Figure A-19 illustrates the
relationship between resistivity and air temperature and moisture content,
whereas Figure A-2Q shows the relationship between ESP migration velocity and
resistivity.
In Figure A-21, these two relationships have been combined, and the
effect of air temperature and moisture content on the migration velocity
through the ESP voltage (independent of the resistivity) has been included.
Thus, this figure shows the direct relationship between migration velocity
and air temperature and moisture content and also the effects of resistivity
as the dielectric strength of air effects. The two dashed lines in the
diagram indicate the winter and summer limits for ambient air moisture content,
and the hatched area represents possible variations in ESP operating conditions
without water conditioning, assuming excess air temperature variations from
90°C to 400°C. It should be noted that an area of low migration velocity
occurs around 180°C.
APPENDIX A-ESP APPLICATIONS IN CEMENT INDUSTRY A-25
-------
RECORDER STRIP FROM OPTICAL
DUST MONITOR
.1 "5—
•I 2-
I
t
>
^B
•
t
KILN ONLY
IN
OPERATION
""6 (25 50 75 100125 150
,1 1_,
3 n_
.Q_
8_j
"1
f
.7
f
c
%
I
*\.
V
I
i-
|_
|_
K
) t25 50 7
5 1C
012
RAV
RAW
RAM
?5H
Mil
MIL
.L S
LS'
TOP
'ART
MILL STOP
KILN AND
RAW HILL IN
OPERATION
0 m
g/N
n3
•AUTOMATIC LEVEL ADJUSTMENT IN
DUST MONITOR
Figure A-18. Dust emissions at a Greek cement plant that used
automatic control and sychronization of damper movement, water injection changes,
and mill startup to overcome problems during changeovers.?
(©1978 IEEE.)
A-26
-------
10
14
.13
o
;r n
£ 10
10
10
10
I I
X- LOG H20/kg DRY AIR
X= 0.01
X=0.03
X*0.05/1
100 200
AIR TEMPERATURE,°C
300
Figure A-19. Resistivity vs. air temperature and moisture content.'
(©1978 IEEE.)
10
11
RESISTIVITY, ohm-on
10'
Figure A-20. Precipitator migration velocity vs. resistivity.
(©1978 IEEE.) A'27
-------
Water conditioning of the excess air by injection and evaporation of
water in the cooler can change the ESP operating conditions, for example,
from point A in Figure A-21 to point B» along the line A-B. The slope of the
line A-B is determined by the proportion between the air cooling effect and
the air humidifying effect of the injected water. Any other conditioning
line (for example, A1-B1) will therefore have practically the same slope in
the diagram as A-B, The length of a conditioning line is directly propor-
tional to the amount of water injected. These assertions, of course, are
only true if the water is injected into the cooler in such a way that it
evaporates in and cools the excess air, not the clinker.
The information in Figure A-21 indicates that an appropriate control
system should be designed to maintain ESP operating conditions in areas with
high migration velocities, i.e., outside or close to the u, curve and avoid-
ing the low u area of dry air around 180°C.
With sets of water injection nozzles arranged above the cool end of the
clinker bed and controlled by the cooler exit air temperature, ESP operating
conditions can be maintained approximately along the operating control line
I-1I (Figure A-21). This method will provide satisfactory ESP efficiency and
acceptable dust emission levels during varying clinker cooler operation.
STARTUP AND SHUTDOWN PROCEDURES
As noted, periods of process startup and shutdown are critical to ESP
operation. The following items are provided as general operating rules of
thumb to follow during these periods :
0 If hopper or support insulator heater elements are available, these
heat sources must be in operation at least 3 hours before startup.
0 The combustible level in the gas exiting the kiln should be ascer-
tained before the ESP is electrically energized.
0 It is generally preferable to preheat the ESP to as high a temper-
ature as possible before energization of the power supplies. Gas
temperatures of 180° to 200°F at the exit of the ESP are recom-
mended. If ESP operation is required before this temperature range
is reached, the outlet electrical fields should be energized first
at low power settings.
APPENDIX A-ESP APPLICATIONS IN CEMENT INDUSTRY A-28
-------
400
300
E 200
TOO
I i
I-II: OPERATION CONTROL LINE
ISO-w CURVES
W1: LOW u VALUE
3'
: MEDIUM u VALUE
VALUE
tt>2
ojoi HIGH
0,02 0.04 0,06 O.OB
MOISTURE CONTENT,kg/kg DRY AIR
0.1
Figure A-21. Relationship between migration velocity w and air temperature
and moisture content including resistivity effect.'
(©1978 IEEE.)
A-29
-------
0 AH rapper equipment should be placed in service prior to startup
of kiIn.
0 All hopper evacuation equipment must be in operation before startup
of the kiln.
0 Upon shutdown of the process, the electrical sections of ESP should
be deenergized before the gas temperature falls below 200° to 250°F
at the exit of ESP. Shutdowns of the ESP should be initiated in an
orderly fashion from the inlet to outlet fields. Rapper operation
must be kept at maximum intensity. The time intervals for shutdown
of the power supplies should be gauged to minimize the discharge
from stack. Each installation may require a different procedure,
but the object is to achieve effective cleaning of the electrode
surfaces. The operation of the induced draft fan must be consid-
ered in the procedure. All conveyors and hopper systems should be
kept operable.
Effective operation of an ESP in a cement plant depends on proper design
and proper maintenance. Table A-3 presents some of the more common problems
associated with ESP operation. As indicated, most malfunctions result from
2
lack of maintenance and attention to the control system.
APPENDIX A-ESP APPLICATIONS IN CEMENT INDUSTRY A-30
-------
TABLE A-3. DETECTION AND SOLUTION OF ESP OPERATING PROBLEMS'
Control panel indicators
Primary
voltage,
a.c.
350b
285
400
350-400
240
240
400
240
Primary
current,
amps
40b
120
30
40-150
40
170
40
40b
Secondary
current,
mA
160b
500
140
100-700
200
400
160
160
ESP conditions/
panel indications
Normal operation
Gas volume and dust
load decreases.
Dust load increases.
In wet processes,
temperature increases
but resistivity is
constant. In dry
processes, tempera-
ture and resistivity
increase.
Gas temperature de-
creases.
Arcing between elec-
trodes
Added primary voltage
is required to main-
tain constant cur-
rent; spark rate
increases.
Less primary voltage
is required to main-
tain constant cur-
rent. Spark rate
increases.
ESP control
efficiency
Normal
Higher than
normal
Usually higher
than normal
Higher than
normal for wet
processes, but
lower than
normal for dry
processes
Normal unless
below dewpoint
Less than nor-
mal
Less than nor-
mal
Less than
normal
Possible problem
Higher hopper
level
Dust bridging in
in hopper
Failure of dis-
charge elec-
trode rapper
to remove
dust from
electrodes.
Failure of rapper
on collection
plate to remove
dust buildup
Problem solution
Raise process
temperature.
Increase dust
removal rate.
Use hopper vi-
brator.
Increase rapping
intensity.
Repair rapper sys-
tem.
Increase rapping
intensity.
Repair rapper
system.
-------
TABLE A-3 (continued)
Control panel indicators
Primary
voltage,
a.c.
0-350
0
Primary
current,
amps
0-40
120
Secondary
current,
mA
0-160
0
ESP conditions/
panel indications
Violent fluctuation
of indicators; high
arcing noise
No current flow
to precipitator
High spark rate
Corrosion (internal
inspection)
ESP control
efficiency
Zero to nor-
mal
Zero
Less than
normal
No immediate
effect
Possible problem
Broken electrode
with top part
swinging back
and forth
Electrical short
circuit of
transformer-
rectifier (T-R)
set, or wire
grounded out
Air leakage
through inlet
ductwork
Air leakage
through in-
spection
hatches
Inlet gas at
temperatures
less than dew-
point
Problems en-
countered dur-
ing startup and
shutdown of
kiln
Problem solution
Isolate section
until electrode
can be replaced.
Repair or replace
T-R set.
Seal points of
of leakage.
Seal hatch doors.
Maintain gas
temperature
above dewpoint.
Use insulation
and hopper
heaters.
The effects of ESP problems can only be stated on a qualitative basis.
3, Multiple-field ESP: primary voltage decreases in moving from inlet to outlet fields, and primary current
-------
REFERENCES FOR APPENDIX A
1. Katz, 0. The Art of Electrostatic Precipitation. Precipitator Tech-
nology, Inc., Munhall, Pennsylvania. January 1981.
2, U.S. Environmental Protection Agency. Portland Cement Plant Inspection
Guide. EPA 340/1-82-007, June 1982.
3. Midwest Research Institute. Participate Pollutant System Study. Vol.
Ill, Handbook of Emission Properties, Cement Manufacture. U.S. Environ-
mental Protection Agency. 1971.
4, PEOCo Environmental, Inc. Unpublished in-house data.
5, PEDCo Environmental, Inc. Report of Baseline Assessment for Inhalable
Particulate Testing of a Wet Process Cement Kiln, Prepared for U.S.
Environmental Protection Agency under No. 68-02-3158, June 1982.
6, Weber, V. P. Alkaliprobeme and alkalibeserligung bei warmespareden
Trockendrenhofen. Zement-Kalk-Gips Helf 8. August 1964.
7. Petersen, H. H. Electrostatic Precipitators for Cement Mills, Kilns,
and Coolers - Control of Operating Conditions During Transition Periods.
Presented at IEEE Cement Industry Technical Conference, Roanoke, Vir-
ginia, May 1978.
8, PEDCo Environmental, Inc. Unpublished in-house data.
9. Hawks, R., and J. Richards. Field Inspection Report—Martin Marietta
Cement Co., Martinsburg, West Virginia. Prepared by PEDCo Environmental,
Inc., for the U.S. Environmental Protection Agency under Contract No.
68-02-4147. November 1980.
10. The Mcllvaine Co. The Electrostatic Precipitator Manual. Northbrook,
Illinois. December 1976.
11. Katz, 0. Maintenance Program and Procedures to Optimize Electrostatic
Precipitators. Presented at 1974 IEEE Cement Industry Technical Confer-
ence, Mexico City, Mexico, May 1974.
APPENDIX A-ESP APPLICATIONS IN CEMENT INDUSTRY
A-33
-------
-------
APPENDIX B
ESP APPLICATIONS IN KRAFT PULP INDUSTRY
INTRODUCTION
Electrostatic precipitators servicing kraft recovery boilers are
designed to,control participate (salt cake) emissions to meet allowable
regulatory limits. The success of maintaining high collection efficiency
depends on how well an ESP was designed for the expected operation of the
associated recovery boiler and its various influences (e.g., nature of the
black liquor to be burned) and on operation of the recovery boiler in
accordance with the design parameters of the ESP(s).
Numerous factors (e.g., gas volume, participate loading, corrosion)
affect the performance of ESP's that service Kraft recovery boilers.
Effective O&M practices oftentimes can keep these factors within acceptable
limits and enable the recovery boilers to operate in compliance with par-
ticulate emissions regulations on a long-term, continuous basis. The
primary intent of this Appendix section is to describe the boiler and ESP
related factors, one by one, and to list and explain effective O&M practices
applicable to each factor. Descriptions of the boiler-related factors and
O&M practices are discussed prior to the ESP-related factors and their O&M
practices. A brief description of recovery boiler operation is presented
first.
KRAFT RECOVERY BOILER OPERATION
The kraft recovery boiler or furnace is an indirect water-walled steam
generator used to produce steam and to recover inorganic chemicals from
spent cooking liquors. (A schematic of the kraft process is shown in Figure
B-l.) The boiler consists of a large vertical combustion chamber lined with
water tubes. The heat exchanger section typically consists of a low-pressure
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
-------
WOOD CHIPS
FRESH
MAKEUP
LIQUOR
RELIEF
GASES
STEAM
DIGESTER
FLUE GAS
TO STACK
,BLOW GASES
BLOW TANK
KNOTTER
WASHING
BLEACHING
SPENT COOKING LIQUOR
JO PAPER
MILL
DO
I
PO
DIRECT-
CONTACT
EVAPORATOR
WATER
Figure B-l. Schematic of kraft process.
1
-------
boiler, superheater, and economizer. Figure B-2 shows a cross section of a
2
modern Babcock and Wilcox (B&W) boiler. The fuel used in the boiler is
spent concentrated cooking liquor (black liquor). The liquor in the burners
has a solids content of between 60 and 70 percent (depending on wood species
and yield) and is made up of organic and inorganic fractions. * The
organic fraction contains lignin derivatives, carbohydrates, soap, and
waxes. ' The inorganic portion consists primarily of sodium sulfate.
Black liquor is sprayed into the furnace at an elevated level in the
combustion chamber through a number of steam or mechanical atomizing nozzles.
The suspended liquor is burned as it falls through the combustion zone. The
following are the major steps in the combustion process:
0 The liquor is dehydrated to form a char.
0 The char is burned in a bed at the bottom of the
furnace.
0 The ash (inorganic portion) remaining after combustion of the char
is exposed to active reducing conditions to convert sodium sulfate
and other sodium-sulfur-oxygen compounds to sodium sulfide.
0 The organic materials in the upper section of the furnace are
oxidized to complete combustion.
Reduced inorganic material (smelt), which consists of a mixture of sodium
sulfide and sodium carbonate, is continuously drained from the furnace. The
ratio of sodium sulfide to sodium carbonate depends on the temperature and
the ratio of sulfur to sodium in the fired liquor.
Combustion of the char begins on the hearth of the furnace. Air for
combustion is supplied through air ports located in the furnace walls. The
primary air supply is used to initiate char combustion. The primary air
supply, which is introduced in the lower portion of the char, is kept to a
minimum to maintain the necessary reducing conditions to convert the ash to
sodium sulfide. The secondary air supply is located at a higher level in
the furnace to create the oxidizing condition necessary to control the char
bed height. A tertiary air supply may be used to complete combustion in the
upper levels of the furnace and thereby eliminate reduced sulfur compounds.
As the char bed is burned, the inorganic ash is liquified and drained to the
furnace hearth, where it is reduced.
APPENDIX 1-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
-------
Furnace
Slag Screen
Tertiary Air Ports
& Windbox
Oscillator Burner.
/Oscillator iurner
Secondary Air Ports
/ 4 Windbox
Pin Stud
Upper Limit
II
Smelt Spouts
4 Hood
Primary Air Ports
& Windbox
Green-Liquor
Reeirculation
DissolvingjJj Dissolving Tank
Tan k >3a Agitator
Figure B-2. Cross section of B&W recovery boiler.
2,7
B-4
-------
Combustion gases produced by the burning of the liquor are passed
through the heat exchanger section of the boiler before being exhausted to a
particulate control device. The gases are cooled to about 800°F in the
boiler tube bank before passing into the economizer. Temperature of the gas
leaving the economizer, which is about 750°F, is reduced in either indirect-
contact or direct-contact evaporators.
There are three types of direct contact evaporators: cyclone, venturi,
and cascade. Cyclone evaporators concentrate the black liquor by placing it
in contact with the high-temperature gas stream by using the wetted wall of
the cyclone. The cyclone removes approximately 50 percent of the uncontrolled
particulate from the gas stream. A venturi evaporator concentrates black
liquor by placing the flue gas in contact with liquor through the generation
of liquor droplets. The droplets are generated through the shearing action
of the gas stream as it passes a weir into which the liquor is being pumped.
Venturi evaporators remove approximately 85 percent of the uncontrolled
particulate emissions generated by the furnace when operated at 4 to 5 in.
HgO pressure drop. In the cascade evaporator, a thin film of liquor coats
several tubes rotated through the flue gas stream. This type of evaporator
will generally increase the black liquor solids content from 48 to 65
percent. The rate of evaporation is related to the flue gas temperature and
the cascade rotation rate. The evaporator can be operated beyond design
rates without substantial process upsets, and can remove up to 50 percent of
uncontrolled particulate emissions from the furnace.
There are three types of indirect-contact evaporators. In general',
these evaporators evaporate the liquor by use of a noncontact tube and shell
design. Because the black liquor in a noncontact evaporator design does not
come in contact with the flue gas, stripping of TRS compounds is prevented.
Evaporated water is condensed by using a tail gas condenser. Noncondensable
gases are directed to lime kilns or into the furnace primary air system for
incineration.
The high temperatures in the furnace char zone result in a partial
vaporization of sodium and sulfur from the smelt. The fume is removed from
the furnace with the combustion gases and condenses to a fine particulate
consisting of sodium sulfate (Na-SQ.) and sodium carbonate
APPENDIX 1-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
-------
Modern recovery boilers are sized for two process conditions: 1) the
heat input to the furnace, and 2) the weight of the chemicals to be recovered.
Both of these conditions affect the heat release rate as a function of
3 2
furnace volume (Btu/ft ) and furnace cross section or hearth area (Btu/ft ).
Typical design values are on the order of 9800 Btu/ft (furnace volume) and
2 8
900,000 Btu/ft {hearth area). The exact dimensions of the furnace depend
on the elemental composition, solids content, heat value, sulfidity, and
chloride content of the black liquor. Deviations of 10 percent or greater
ft
in design variables should be investigated to ensure maximum efficiency.
The manufacturers generally consider the boiler to be overloaded when the
firing rate (Btu/h) exceeds 120 percent of the rated value. Operation
outside of design values can cause tube fouling and reduced smelt recovery,
which can increase emission rates and reduce thermal efficiency.
Major Boiler-Related Factors and 0_&M Practices that Affect Uncontrolled Parti-
culate Hatter
The rate of uncontrolled participate matter from a recovery boiler and
the resulting loading to the ESP(s) depends on a number of interrelated
boiler operating variables. Several of the major ones are listed here and
9-] ?
then separately "discussed:
0 Firing rate [pounds of black liquor solids (BLS) per hour]
c Black liquor heating value (Btu/pounds BLS)
0 Black liquor concentrations (percent solids)
0 Total combustion air (excess air)
0 Primary air rate (percent of total air)
0 Secondary air rate (percent of total air)
0 Char bed temperature
In addition to affecting the uncontrolled emission rates (as generally
shown in Table B-l), these variables can also reduce ESP performance. The
following discussion addresses both of these effects.
Reductions in TRS and S0? emissions from kraft furnaces result primarily
from the optimization of process variables that cause sulfurous gases to
APPENDIX i-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
B-6
-------
TABLE B-l, SUMHARY OF THE EFFECTS OF RECOVERY BOILER PARAMETERS
ON PARTICULATE EMISSION RATES
Parameter
Firing rate
Primary air
Excess air
Smelt bed temperature
Flue gas oxygen a
Primary air temperature
Black liquor sulfidity
Change
Increase
Increase
Increase
Increase
Increase
Decrease
Increase
Effect on
participate
emission rate
Increase
Increase
Increase
Increase
Increase
Decrease
None
If increase in oxygen is a result of an increase in primary air volume.
B-7
-------
become chemically combined with sodium to form a participate emission.
Operation of the boiler under these process conditions can also increase
uncontrolled particulate emissions in the form of sodium sulfate.
Firing Rate--
The firing rate of a kraft recovery boiler is measured in pounds of
black liquor solids per unit of time (either pounds BLS/24 hours or pounds
BLS/hour), Given a specific heat value of the black liquor solids, percent
solids in the liquor, and elemental composition of the liquor, one can
define the flue gas volume produced and the boiler heat input. The firing
rate of a recovery furnace is often increased to increase pulp production
rates. This usually requires more gallons per minute of the fired liquor.
The firing rate is limited by the pumping capacity of the system, which is
based on the liquor temperature and viscosity.
Table B-2 presents the various operating conditions of a recently
inspected recovery boiler with a firing rate of about 56,000 pounds BLS/hour.
This Combustion Engineering boiler has a design process rate of 500 tons of
air-dried pulp per day or 55,400 pounds BLS/hour,
The importance of the boiler firing rate with respect to performance of
the accompanying ESP(s) is that increased firing rates increase both the
resultant flue gas volume and the uncontrolled emission rate. As more black
liquor solids are sprayed into the furnace, more air is required to support
proper combustion and more emissions per unit of time are generated.
Generally, the uncontrolled emission rate of a typical, indirect-contact,
recovery boiler is 8 gr/dscf. Actual rates, of course, are boiler-specific.
A higher-than-design flue gas volume increases the vertical gas velocity
through the furnace combustion zone and results in more particulate emissions
due to the entrainment of black liquor droplets and char particles. This
situation not only increases the particulate emission rate, but also increases
the particulate concentration and changes the nature of the particulate,
both of which can adversely affect the ESP(s).
Flue gas temperatures also increase as the flue gas velocity increases
(because of increased air volume). This occurs when the increased velocity
APPENDIX i-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
-------
TABLE B-2. OPERATING CONDITIONS OF RECENTLY INSPECTED
RECOVERY BOILER
Steam flow
Steam pressure
Steam temperature
Feedwater pressure
Feedwater temperature
Primary air flow
Primary air temperature
Black liquor temperature
Black liquor flow
Black liquor solids content
Black liquor nozzle size
Number of liquor guns
Black liquor solids firing rate
Boiler heat input
Boiler output (steam)
Boiler efficiency (calculated)
165,000 Ib/h
600 psig
700°F
880 psig
330CF
50,000 Ib/h
288° F
252°F
• 117 gal/mln
661
No. 5
6
56,062 Ib/h
347.6 x 106 Btu/h
186.6 x 106 Btu/h
53.7%
B-9
-------
decreases the efficiency of the steam tube heat transfer, The effect of
temperature on an ESP was described in the body of the manual.
Because the boiler firing rate can affect flue gas volume, flue gas
temperature, and the rate and concentration of particulate emissions (all
which can greatly affect ESP performance), this operating parameter should
be monitored carefully. Some effective O&M practices 1n this regard are as
follows:
0 The importance of recovery boiler firing rate on ESP performance
should be recognized and communicated to personnel responsible for
boiler operation and ESP operation on all shifts.
0 Based on previous stack tests (such as the ESP performance test),
the inlet and outlet grain loading air volume and temperature at
the ESP should be evaluated against the boiler firing rate.
Established baseline conditions should be used for comparison with
normally monitored and recorded boiler data.
0 The boiler operator and plant management should be kept aware of
the "acceptable" firing rate range. As additional stack test data
or other pertinent data are collected, the acceptable range should
be re-evaluated and updated as necessary,
0 Boiler personnel need to keep the environmental engineer appraised
of boiler changes that could affect flue gas conditions (e.g.,
installation of new liquor guns; change in fuel oil characteristics,
if fuel oil is used as a supplemental fuel).
0 The environmental engineer should be familiar with the "F-factor,"
which is the measured flue gas volume (dry standard cubic feet per
minute corrected to 0 percent flue gas oxygen) divided by the
boiler black liquor solids firing rate (pounds BlS/minute). The
value of the F-factor varies from mill to mill because of the
variation in species, pulp yield, makeup chemicals, and evaporator
operation. This value, however, is reasonably constant for a
specific mill.
Knowledge of the F-factor allows periodic calculation of flue gas
volumes to ensure that the boiler is not exceeding design values.
For a more accurate calculation of velocities in the combustion
chamber and the total flue gas volumes, the F-factor must be
corrected for water evaporated from the fired liquor, water of
combustion, temperature, excess air, and miscellaneous additions
such as steam from soot blowing.
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
-------
0 When the boiler firing rate, flue gas volume and temperature,
and/or participate levels are determined to be outside determined
acceptable ranges, the appropriate personnel must be consulted.
For example, a meeting of the environmental engineer, boiler
operator, and maintenance person may be in order. If the operat-
ing conditions could adversely affect the ESP, the environmental
engineer needs to know what is acceptable and how to resolve an
unacceptable situation.
Black Liquor Heating Value and Solids Content—
Based on computer models of recovery boiler operations, participate
emissions increase sharply with increases in the heat value; they increase
less, but significantly, with higher solids content in the black liquor
{Figures B-3 and B-4). These changes are primarily the result of changes in
12
the heat production rate of the char bed.
Because the heat value and solids content of black liquor are dependent
on several process variables in the pulping process (e.g., digesters,
evaporators, wood species, and harvest conditions), day-to-day firing
conditions and liquor properties may vary significantly. This variability
makes these factors difficult to control. It is important that the environ-
mental engineer know the allowable variation of the inlet grain loading the
ESP is designed to handle.
Combustion Air--
For proper and adequate combustion in a recovery furnace, the total
amount of combustion air usually must be between 110 and 125 percent of the
calculated theoretical (stoichiometric) air requirement. Incomplete combus-
tion normally occurs below this range, which is unacceptable for the operation
of a recovery boiler. Excess air (i.e., that which exceeds the theoretical
amount) is monitored and controlled in the boiler control room so that
proper combustion can be maintained.
Because the boiler operators' primary interest is in maintaining
complete combustion, it has been noted during some inspections that too much
excess air has been allowed. This can lead to particulate control problems,
and even to boiler problems that the boiler operator may not be aware of.
When the amount of excess air exceeds 125 percent (5 percent (L in flue
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
B-ll
-------
MO
wri
I" 400
I
m
IX
I
M 62 W
•lack Liquor Solid* Cancmtntim, t
5.0
4.0
1
I"
J.S
3.0
70
Figure B-3, Effect of black liquor solids concentration on uncontrolled
participate emissions.
2000
_ 1600 I-
3
S 1200 -
S2DO
SACK)
6000 MOO
of Soli* ta Kama
10
MOO
Figure B-4. Effect of black liquor heating value on uncontrolled
participate emissions.'2
B-12
-------
1? 11
gas), formulation of SO., increases in the boiler. * . The SO, that is
absorbed in the particulate at low temperatures makes it sticky, and this
sticky particulate fouls heating surfaces in the economizer and reduces heat
transfer rates. The deposits may cause in a high draft across the economizer.
The particulate (salt cake) can also cause severe operating problems when
collected on the plates of the ESP, Soot blowers cannot effectively
remove the sticky salt cake from the boilers, and normal rapping intensity
cannot remove it from the ESP plates. A method to determine what level of
SQ~ will cause sticky particulate is to prepare a 4 percent salt cake
solution in water and if the pH is <9.5» then the operator can expect to see
sticky particulate buildup.
Particulate buildup on the plates causes low collection efficiency
because it reduces the power input to the unit. The effect is more prevalent
when flue gas temperatures fall below 300°F and a combination of high boiler
excess air and ambient air in leakage occurs.
Combustion air consists of primary air and secondary air, and maintain-
ing of the proper percentages of the two is important for good boiler
operation and particulate emission control. Primary air is required to
provide combustion under reducing conditions and to maintain the temperature
in the char bed to prevent a condition called "blackout." The proper amount
of air is a compromise between maintaining sufficient combustion and reduc-
ing abnormally high vertical velocities in the furnace. High velocity
results in an accumulation of deposits on the heating surface of the boiler
(after cooling of the gas stream and condensation of fume), and this accumu-
lation causes increased particulate emissions.
A high rate of primary air (particularly at high velocities) increases
the release of sodium and sulfur from the char bed because of the greater
diffusion of the vapor from the bed. The higher air volume also increases
the combustion rate of the char, which increases the bed temperature.
Particulate emissions increase sharply when the amount of primary air
12
exceeds 45 percent of the total air volume,
Prolonged operation at low primary air volumes can increase the char
bed height, which must be reduced by an increase in bed temperature. The
most common method of reducing the bed height is to increase the primary and
APPENDIX B-ESP APPLICATIONS IN KBAFT PULP INDUSTRY
-------
secondary air volumes, which alters the smelt ratio as oxidation and tempera-
ture conditions are changed.
The secondary air should make up at least 40 percent of the total air
2
(maximum of 65 percent of the theoretical air). In boilers with a high
char bed, the secondary air has two purposes: 1) the primary purpose is to
complete combustion of CO gas released from the char bed as it moves up the
furnace walls; and 2) the secondary purpose is to provide air in the center
of the furnace to burn the char bed.
The total air volume (secondary plus primary) must be large enough to
produce complete combustion, but it also must be limited to a level that
will reduce the vertical velocity in the furnace and total flue gas volume.
Thus, good control of combustion air is obviously important to proper
boiler operation; however, in this report, it is even more important for the
control of the loading and chemical consistency of particulate matter
reaching the electrostatic precipitator(s). In summary, improper control of
combustion air can cause "blackout" (incomplete combustion of black liquor);
sharp increases in particulate emissions (e.g., that caused by primary air
exceeding 45 percent of total air); and formation of sticky particulate
(excess air greater than 125 percent, SO, absorption with 'particulate).
Improper control of combustion air can also cause increases in char bed
temperature (the adverse effects of this are discussed later) and high flue
gas velocity (the adverse effect of which were described under "firing
rate".
Effective O&M practices for maintaining proper control of combustion
air include the following:
c The boiler operator and environmental engineer must communicate
frequently and recognize each other's concern for proper control
of combustion air. Fortunately, good control of this parameter is
a shared concern,
0 The environmental engineer should try to develop relationships
between the percent excess air and primary air in the recovery
boiler and the following:
Particulate loading to the ESP (This can be done during
performance tests. If tests have already been performed,
test data and boiler operating logs can be located and
evaluated.)
APPENDIX i-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
B-14
-------
Observance of visible emissions from the ESP(s). (If this
occurs, different ESP and boiler parameters should be checked
including combustion air levels.)
Air volume to the ESP. Increases in air volume also can be
due to other factors, such as in- leakage. Excess air volume
data are available in the boiler control room, and points of
inleakage can be determined by measuring the oxygen content
with portable equipment at strategic duct locations.
Flue gas temperature to the ESP,
0 The empirically determined relationships can be very useful in the
detection of ESP operating problems. Graphical representation of
the relationships are easy to "read" and to describe to other
personnel (e.g., boiler operators, maintenance personnel, manage-
ment).
0 The environmental engineer must be familiar with the boiler
operation and understand why the boiler operator may vary the
excess air to the furnace. He/she should know how to read the
instrumentation so that he/she can periodically check the readings,
0 The environmental engineer should review ESP maintenance logs for
problems caused by sticky particulate. Maintenance personnel
should be consulted to get their opinions of the problems. The
boiler operators and operating logs (on those maintenance days)
should be consulted to see if any changes in boiler operation
occurred. There is usually a link between operating, maintenance,
and environmental problems. The environmental engineer can try to
determine these links and then get back with the operating and
maintenance personnel. Such personal associations and interest
should encourage working together to optimize operating conditions.
Char Bed Temperature—
The amount and nature of particulate emissions vary considerably,
depending on the char or smelt bed temperature. As shown in Figure B-5,
this temperature is greatly affected by the percentage of combustion air.
The smelt bed temperature is also directly influenced by the primary air
temperature. Although a figure is not available to show the effect of char
bed temperature on uncontrolled particulate emissions, Figure B-6 shows the
effect of primary air temperature of the uncontrolled particulate grain
loading. The recovery boiler firing rate, which is interconnected with the
above factors, also affects the char bed and resultant particulate emissions.
APPENDIX i-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
-------
50
o
40
35
30
100 110 120 130 140
TOTAL AIR TO UNIT.*
2,000
CD
m
o
1.900 5
T3
m
30
1,800
1,700
150
Figure B-5a. Bed temperature as a function of total air.
12
45
240
35
30
35
40 45 50 55
PRIMARY AIR.X OF TOTAL
CD
m
o
2,000
1,800
m
g
cz
•X)
1,600
60
Figure B-5b.
Char bed temperature vs. percentages of primary and total
12
combustion air.
B-16
-------
600
500
>, 400
T3
300
£
>,
T3
S 200
CM
o
100
100 200 300 400
PRIMARY AIR TEMPERATURE, °F
5 Z!
73
4 i
3
•* i-i-t
V
o
500
t-6. Effect of priamry air temperature on participate emissions.
12
B-17
-------
A qualitative discussion of the mechanisms by which particulate
emissions are increased is presented here. Dehydrated liquor (char) on the
furnace hearth is burned at a high temperature to allow the inert portion of
the liquor to melt and drain from the hearth. A mixture of sodium sulfide
and sodium carbonate must be maintained under reducing conditions to prevent
oxidation to sodium oxides.
Under normal operating conditions, elemental sodium is vaporized and
reacts to form Na,,Q. The rate of evaporation depends on the char bed
temperature and the diffusion conditions in the smelt zone. As the sodium
evaporates from the bed, 1t reacts with oxygen in the primary air zone to
14
form Na^O. The Na00 reacts with CO,, to form Na,,CO,.
2
n\J I cav t*3 W I UM Wi, WM ' V I ill lianUWn •
The char bed temperature also determines the rate of sulfur released to
the flue gas. Sulfur is commonly present in the flue gases as S, H~S, SOn,
or SO.,- The higher temperatures favor the formation of S(L and SO-. Sulfur
in the form of S and hLS reacts with excess oxygen in the oxidizing zones of
the furnace to form SO^ and SCL. The Na?CQ~ reacts with the SCL to form
Na^SOo, which is later oxidized to Na?S(L, Typically, the Na-SQ. deposits
on the heat exchanger surfaces (screen tubes, superheater, and boiler tubes)
and must be removed through continuous sootblowing. Because these deposits
reduce the heat transfer and the overall efficiency of the boiler, the
tendency is to overfire the boiler to achieve the required steam flow. This
increases both the gas temperature entering the ESP and the superficial
velocity.
The factors that influence the smelt bed (i.e., combustion air and
firing rate) have already been discussed. The O&M practices described for
those factors also apply here and therefore are not repeated. As a reminder,
the environmental engineer and the boiler operator should know basically
what takes place in the boiler. The boiler operator should at least be
aware of boiler operating factors that can affect the performance of the
downstream ESP(s). If the boiler has to be operated in a mode where the
resultant emissions may adversely impact the ESP(s) or their performance,
both the environmental and maintenance department should be notified. Only
then can alternatives to these modes of operation be discussed.
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
-------
Recovery BpilerESP' s
The theory, operation, and basic physical layout of ESP's (as described
in Section 2 of the manual) are essentially the same as those for ESP's used
on kraft mill recovery boilers, A few items that relate only to kraft
recovery boiler ESP's are discussed here,
The control efficiency of kraft recovery boiler ESP's is determined by
the initial design of the unit and operating characteristics of the recovery
boilers. The older direct-contact kraft process recovery boiler produces
salt cake that is a good electrical conductor (low resistivity), and the
high gas moisture and acid vapor improve the surface conductivity of the
dust. The newer indirect-contact (low-odor) process produces salt cake with
smaller particle size distribution and higher dust loading which requires a
larger ESP than the direct-contact process. Figure 6-7 shows the relation-
ship between the efficiency and the design specific collection area (SCA) on
modern recovery boiler ESP's. Figure B-8 shows the design superficial
velocity vs. year of installation for 20 randomly selected ESP's servicing
recovery boilers.
Two methods are used to support the discharge electrode in recovery
boiler ESP's. In the first method, referred to as the weighted-wire design,
each electrode is individually supported and tensioned between the plates.
In the second method, referred to as a rigid-frame design, the electrode is
attached to a rigid frame between the plates, A modification of the second
design is a rigid-pipe electrode system, in which the corona is generated on
the tip of spikes attached to a vertical pipe.
Collected particulate matter can be removed from the ESP by three
different methods. In the first method, referred to as a wet-bottom ESP
(Figure B-9), the salt cake is allowed to fall into an agitated pool of
black liquor in the bottom of the ESP. In the second method, referred to as
a dry- or drag-bottom ESP, the salt cake is allowed to fall onto the flat
bottom of the ESP shell, where a drag chain physically moves the material to
a discharge screw (Figure B-10). The third method of dust removal consists
of a pyramid-shaped hopper with rotary air locks and slide-gate discharges.
This design is not often used for recovery boiler ESP's because of hopper
plugging problems.
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY B-19
-------
100 H
99 -
**
*
>-
r, 98 —
97 _
DEUTSCH-
ANDERSON
MATTS-OHNFELOT
96
I / I
I I
100
200
300
400 500
DESIGN SCA, ftVlOOO tcfm
600
Figure B-7. Design SCA and efficiency of 20 recovery boiler ESP's.
B-20
-------
CD
no
ceL •)
UJ <-
O-
-------
TRANSFORMER-
RECTIFIER
HEAT JACKET
PERFORATED
DISTRIBUTION
PLATES
DISCHARGE
ELECTRODE
GROUND SWITCH BOX-
ON TRANSFORMER
DISCHARGE
ELECTRODE
VIBRATOR
COLLECTING
ELECTRODE
RAPPERS
TOP HOUSING
ACCESS DOOR
TOP HOUSING
HOT ROOF
I/ACCESS DOOR
8FTHEEN
COLLECTING
PLATE SECTIONS
COLLECTING ELECTRODES
WET BOTTOM
Figure B-9. Typical wet-bottom ESP with heat jacket.
(Courtesy of Research Dottrel"!, Inc.)
19
B-22
-------
Figure B-10. Typical weighted-wire ESP with drag bottom,16
(Courtesy of Environmental Elements Co.)
B-23
-------
O&M Practices In Spec1fic Prpb1 em Area
The subject of ESP operation and maintenance is a broad one involving
all aspects of ESP performance. It covers all the components and all
operating conditions. In general, maintenance is considered to be the
routine analysis and replacement of ESP components that have failed because
of age or abuse. Maintenance requirements can be increased by poor operating
practices or reduced by superior system design; however, detailed and
exhaustive maintenance practices do not necessarily yield superior ESP
performance,
Early identification of O&M problems reduces the extent and the occur-
rence of excess emissions and allows the plant to schedule outages or make
on-line adjustments to maintain production and operate within the prescribed
emission limits. The operator should identify those operating conditions or
variables that indicate operation outside the accepted norms for a particular
boiler/ESP system. (Normal values or conditions are established during the
initial performance stack test or are based on the accepted state of the
art.) This is generally referred to as baselining, as discussed previously
in Section 3.
The boiler-related factors and O&M practices previously described are
very important to the performance of ESP's because they affect the major
design parameters of the control equipment (i.e., flue gas volume, velocity,
and temperature and particulate loading and composition). Proper operation
of the recovery boiler will greatly reduce the overall problems with the
ESP(s). This section addresses O&M practices that can be implemented to
minimize the following potential ESP problem areas:
° Flue gas volume
0 Gas distribution and sneakage
0 Salt cake removal
0 Corrosion
The importance of maintaining adequate corona power and its usefulness
in troubleshooting malfunctions of recovery boiler ESP's are also discussed.
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
-------
Flue Gas Volume--
Because ESP performance is affected by total gas volume, a good operat-
ing practice is for the operator or environmental engineer to estimate the
volume based on the black liquor firing rate, flue gas oxygen, and tempera-
ture. Most plants monitor flue gas oxygen at the economizer outlet rather
than at the ESP outlet. An estimate of the flue gas volume must be based on
ESP outlet conditions, and the inspector should be equipped with portable
temperature measurement equipment (i.e., thermometer or thermocouple) and
portable oxygen measurement equipment (e.g., a Fyrite oxygen analyzer). The
flue gas volume may be calculated from a plant-specific F-factor (dry
standard cubic feet/pound BLS) corrected for flue gas oxygen, moisture, and
temperature. (The F-factor for a typical black liquor is approximately 51
dscf/lb BLS.) Temperature and oxygen measurements should be made at the
outlet of each chamber where possible (accessible).
The following method is used to calculate the flue gas volume at the
ESP inlet or outlet. Corrections made for the flue gas oxygen are for dry
standard gas volume, not wet gas volume.
Q =
BLS
min
20.9
\
\ + F
20.9 - %Q
2
where
!R
460s)
Sjpf- (Eq. 1)
BLS = black liquor solids firing rate to the boiler.
F. = F-factor for black liquor solids in dscf/lb BLS.
= oxygen content of flue gas at ESP inlet in percent.
F7 = standard cubic feet of water vapor generated from
combust'
solids.
combustion of hydrogen per pound of black liquor
FE = standard cubic feet of water vapor evaporated in
direct-contact evaporator.
F = standard cubic feet of water vapor added to flue gas
stream as a result of soot blowing.
T = temperature of the flue gas at the ESP inlet in °F.
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
-------
The amount of water evaporated 1n a direct-contact evaporator may be
determined by using the liquor flow rates and liquor solids content entering
and leaving the unit (see Figure B-ll). This method is a simple mass
balance based on the assumption that the total amount of solids does not
change in the evaporator, This is not strictly true because the liquor does
absorb salt cake from the flue gas stream. This effect Is considered
negligible, however, in the calculation of water lost. If a more exact
estimate is desired, estimates of absorption rates can be based on flue gas
volume (actual cubic feet/minute) and the uncontrolled boiler dust loading
(grains/actual cubic feet per minute). In general, a cascade-type evaporator
may remove 50 percent of the uncontrolled particulate. In most units this
will increase the total liquor solids mass by less than 5 percent and
usually will result in less than a 2 percent error in the real gas stream
moisture.
If a more exact or plant-specific value is desired, the F-factor can be
calculated from stack gas volume (dry standard cubic feet/minute), flue gas
oxygen content, and the firing rate of black liquor solids. When using this
method, the operator, inspector, or plant environmental personnel can make
this a day-to-day determination of the flue gas volume being treated by the
ESP without the expense of conducting stack flue gas volume determinations
with a Pitot tube.
When flue gas oxygen increases above normal ranges, the source of
inleakage should be identified immediately and appropriate repairs made to
reduce the inleakage. Failure to reduce the inleakage will not only cause
excess emissions, but the cooling effect of the ambient air may also cause
low-power input, excess sparking, and corrosion.
The amount of steam used in soot blowing is not generally measured.
Based on discussions with ESP and boiler manufacturers and limited data from
pulp ruins, however, the value is estimated to be 8 to 10 percent of the
rated steam flow of the boiler. The value is expressed in pounds of water
vapor per minute, which must be converted to standard cubic feet per minute.
The values of F and F, may be compared with the values obtained during a
stack test as a check on the validity of the derivation. The variables
affecting these values are too numerous to list here, but they include wood
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
-------
FLUE GAS (IN) /"" "\ FLUE GAS (OUT)
IQUOR (OUT)
EVAPORATOR
f
BLACK LIQUO
water evaporated (Ib/min) = A p^ ~ B p
where
A = gallons of black liquor to the evaporator
p. = density of black liquor into the evaporator In Ib/gal
B • gallons of black liquor from the evaporator
p » density of black liquor out of the evaporator in Ib/gal
% BLS.
% BLS
where
solids content of black liquor entering evaporator
% BLS * solids content of black liquor leaving evaporator
Figure B-ll. Method of calculating additional moisture in the flue gas
stream due to direct-contact evaporator.
B-27
-------
species and mix, process step variables, quantity of inorganic salt cake
recycled to the recovery boiler, percent solids in the black liquor, and
heating value of the black liquor.
When the black liquor F-factor is derived from stack tests or by
theoretical equations, it is convenient to work in terms of standard cubic
feet of gas because this allows for the addition of values without constant
correction for different gas conditions. The ESP, however, must be analyzed
at the actual gas conditions (i.e., at the measured temperature and oxygen
content). Once established, the values of F, and F+otai tend to remain
relatively constant provided no significant changes in the process occur.
The use of an established F-factor to determine gas flow through the
ESP requires relatively little calculation. Only the following are needed:
value of the F-factor (dry), firing rate of the black liquor, percent BLS,
density of the black liquor, and the temperature and oxygen content of the
flue gas.
The ESP dimensions can be obtained from engineering drawings (blueprints),
Using these dimensions and the determined gas flow rate will usually produce
superficial velocity values that are slightly lower than actual values. The
area input into the calculations does not account for the cross-sectional
area blocked by the plates and wires. The calculated value should be in the
range of 2.5 to 4.0 ft/s, and the lower values generally are recommended.
Obviously, as the superficial velocity in an ESP decreases, treatment time
will increase. Also, if the superficial velocity exceeds 8 ft/s, not only
will the treatment time drop, but reentrainment of captured particulate may
occur as a result of the high velocity stripping material off the ESP plate.
Thus, it is important to consider the gas volume through the ESP. Gas
volume is especially critical if there is a possibility of high excess air
levels resulting from air inleakage or improper boiler operation, or if high
gas volumes could occur from overfiring the recovery boiler.
Another value that should be checked as an operating practice 1s the
actual SCA. This value relates the total available plate area to the gas
2
volume (ft /I000 acfm), and when compared with design or baseline values,
indicates ESP performance capabilities. Generally, an increase in the SCA
(actual) means improved performance. Other factors are involved; therefore,
APPENDIX B-ISP APPLICATIONS IN KRAFT PULP INDUSTRY
-------
a comparison of actual SCA with design or baseline values.,,is not meaningful
by itself.
Gas Distribution and Sneakage—
Proper gas distribution into the ESP is very important to allow the
salt cake particles to have adequate residence time in the ESP. A well-
designed distribution system is necessary for good ESP operation. An
effective operating practice is to keep the distribution system clean; This
is accomplished by rappers and periodic inspection.
Once the flue gas is in the ESP and the electrostatic treatment zone,
it must be kept there so that the proper charging and collection of the
participate may occur. A series of baffles perpendicular to the gas path
are used above and below the plates to prevent gas sneakage. Gas sneakage
allows a portion of the particulate-laden flue gas to bypass the electrically
charged area. As a result, collection efficiency is reduced. The installa-
tion of these baffles perpendicular to the gas flow gives them a longer path
to the gas stream and a resistance to flow. Because gas flow follows the
path of least resistance, the gas tends to remain in the electrical treatment
zone and gas sneakage problems are reduced.
In otherwise well-designed and well-operated ESP's, gas sneakage and
gas distribution problems may account for more than 50 percent of the total
particulate emissions. An increase in ESP input power cannot result in the
capture of particulate matter in a gas stream that bypasses the treatment
zone.
The baffle plates must be designed so that they do not interfere with
the normal dust removal system. They must not extend too far below the
treatment zone, but they must be low enough to present sufficient resistance
to minimize gas sneakage. In wet-bottom ESP's, the proximity of these
baffle plates to the black liquor pool may cause them to be subject increased
possibilities of acid dew point corrosion. The placement and integrity of
these baffles should be checked during each internal inspection.
Salt Cake Removal--
The collected dust removed from the collecting surfaces must be disposed
of properly. In the kraft recovery process, the collected salt cake is
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
-------
recycled to the recovery boiler by combining it with the black liquor. This
recycling of the salt cake is usually accomplished in the bottom of the ESP.
Relatively few recovery boiler ESP's have hoppers. These ESP's are typically
flat-bottomed with a ribbon-mixer, paddle-mixer, or drag-chain conveyor to
move the salt cake to a pool, trough, or tank of either black liquor or
water for recycling to the recovery boiler. Chain breaks, misalignment of
the drags, sprocket failures, and/or motor failures are typical malfunc-
tions. Areas near these liquid recycle points and baffles between fields
and the shell walls may be prone to corrosion due to heat loss and the
raising of the acid dew point temperature. Care must be taken to detect
these problems and thus minimize excess emissions.
The conveying system or mixing system must be sized properly to remove
the quantity of dust recovered under normal maximum operating conditions.
Undersized equipment may allow buildups that can seriously affect long-term
ESP performance. In addition, the dust conveyor system must adequately
cover the ESP bottom and minimize the area that is "out of reach" of the
conveyor system.
The buildup of these deposits or the failing of the conveyor system can
have long-term effects on ESP performance. If a buildup were to reach the
treatment zone, permanent damage to the ESP components might result, i.e.,
warpage of the plates or rigid discharge frames and misalignment of wire-
weight guide frames. In addition, temporary misalignment of wires or plates
may result from the pressure of moving plates and wires, T-R's may be
tripped out because of wire plate contact, and resuspension of the dust may
occur because of the dust piling up in the treatment zone. The most serious
problem, however, is the distortion of the ESP internal components. This
distortion will reduce the ESP performance capabilities. The usual indication
of a material buildup is the tripping of a T-R set (although there are many
other causes for this) along with apparent discharge problems. Records
should indicate the time and location of the occurrence, the corrective
actions taken, and when normal operations were restored. A gas-load test
will provide preliminary indications of permanent damage, and an air-load
test should be performed at the next outage to determine if significant
damage and deterioration occurred. An internal inspection of the ESP should
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
-------
reveal the nature" and extent of-any damage incurred. In--any event, continu-
ous liquor level monitoring is required.
Corrosion--
Corrosion appears to be the most serious maintenance problem in long-
term operation of recovery boiler ESP's. Corrosion attacks the ESP shell
and internal components. Advanced corrosion is accelerated by air inleakage
through corroding areas in the ductwork, around access doors, or in areas
near the liquor-flue gas interface in wet bottom units. The rate of cor-
rosion is increased in the colder areas of the ESP. Localized cooling
occurs when heat loss through the shell is highest, i.e., where outside
stiffeners or structural columns are attached to the shell. Corrosion in
internal areas reduces rapper effectiveness so that dust cannot be effec-
tively removed from ESP collection surfaces. In addition, structural
members can be weakened or destroyed.
A survey by the Technical Association of the Pulp and Paper Industry
(TAPPI) of 19 noncontact recovery boilers installed between 1974 ancf 1979
indicated that 63 percent had some corrosion problems and 26 percent had
20
severe corrosion problems. Based on the operating conditions of the 19
boilers, the average temperature of those with serious corrosion problems
was 361°F, The average temperatrue of those reporting no corrosion problems
was 384°F.
The primary corrosive agent in kraft recovery boiler ESP's is sulfuric
acid. Flue gases from the boiler contain H«0 vapor with a high concentration
of SO,. The SO, vapor combines with the water present to form sulfuric acid
vapor (H9SQ,.). As the temperature of the gas stream is reduced, the H-SO.
21
vapor becomes saturated and forms an acid mist.
The most severe area of corrosion in wet bottom ESP's is in the area
above the liquid level and below the treatment zone. This area is baffled
and is not a part of the main gas volume passing through the ESP. Vapors
from the black liquor in the bottom are extremely corrosive. The activity
22
of the vapors Increases with high oxygen and sodium sulfide concentrations.
Also water vapor raises the local dew point temperature. The temperature of
APPfNDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
B-31
-------
the black liquor is normally below 180DF and results in a cool shell tempera-
ture surrounding the liquor. This cool shell temperature results in a
gradual decrease in shell temperature between the treatment zone and liquor
level. The lower wall temperature is usually below the acid dew point and
near the moisture dew point (which is typically around 165°F). The tempera-
ture within the ESP is not uniform and the temperature of the shell varies
due to contact with structural members, degree of insulation, exposure, and
orientation. Areas of low gas circulation in the ESP typically have the
lowest temperature and highest rate of corrosion.
The maintenance of a uniformly high temperature in the ESP is important
in reducing the rate of corrosion. The temperature may be increased by:
1. Reducing air inleakage
2. Insulating the shell
3. Heating the shell
In general, at flue gas temperatures above 350°F, a well construced,
well insulated steel shell ESP will have minimum corrosion. Units with a
flue gas temperature between 300° and 350CF have problems in areas of
highest heat loss. Units with temperatures below 300CF require supplemental
heating. The amount of heating required varies with flue gas temperature,
degree of insulation, and other environmental factors such as wind loss and
degree of exposure. Generally, the heat requirements are approximately
500,000 Btu/h.23
Maintaining Adequate Corona Power—
A key indicator of ESP performance is the corona power, which is a
useful value for determining whether ESP performance has changed significantly.
In general, the corona power in a recovery boiler ESP increases from inlet
to outlet, as dust is precipitated out of the gas stream (see Figure B-12).
2
Secondary current values are generally about 0.02 to 0.03 mA/ft in the
2
inlet field and increase to 0.08 to 0.09 mA/ft in the outlet field.
Specific corona power, which may also be calculated for the entire ESP
or for individual chambers, is calculated by the following equation:
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY g_32
-------
DC.
ce
ce
I
.10
0.08
0.06
0.04
>r
cn
o
u
0.02
2 3
FIELD NUMBER
Figure B-12. Examples of optimum secondary current distribution
in ESP serving kraft recovery boiler, assuming uniform
rapping and wire size in all fields.
B-33
-------
specie corona po.er - gt.] ua) <*>•
The volume obtained by the modified F- factor (divided by 1000) is
substituted into this equation. In general, the higher the value of the
specific corona power is, the higher the ESP removal efficiency will be.
Thus, one may determine whether ESP performance would be expected to in-
crease or decrease by evaluating the specific corona power.
Based on a review of several performance tests and discussions with ESP
vendors, the specific corona power needed to meet NSPS for kraft recovery
boilers is usually above 400 watts per 1000 acfm. Acceptable performance
may be obtained with lower specific corona power values, however, if there
are no major problems with power distribution, inleakage, or rapper operation.
Equation 2 indicates that a decrease in gas flow through the ESP may improve
performance (constant corona power is assumed). The corona power is generally
not constant, however; it increases with decreasing gas velocity (volume)
because the bulk of the particles do not penetrate far enough into the ESP
to inhibit power input. (The opposite is true when gas volume is in-
creased.) Thus, a decrease in gas volume may substantially /improve ESP
performance. This relationship is an important reason for evaluating firing
rates and excess air levels. This specific corona power may be indicative
of ESP performance, but should not be used as a sole indicator. The in-
spector must rely on his/her experience and general knowledge of the ESP
applied to the recovery boiler in question to draw any final conclusions
regarding overall ESP performance.
To detect and help prevent decreases in corona power caused by mis-
alignment, dust cake buildup insulator failure, etc., the boiler operator
should record T-R meter readings at least twice per shift. The operator
also should plot ESP power levels by field (inlet to outlet) for each
chamber. Deviations from optimum values (determined from baseline or normal
values) should be used to evaluate internal ESP conditions and to analyze
potential emission levels. If recent V-I curves are not available, the
operator should request the plant environmental engineer or electrician to
produce a V-I curve for each field. Data from the V-I curves should be used
to target the inspection of the rappers, the gas distribution system, and
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY R-34
-------
local cooling and to check for inleakage. Serious deviations from normal
values should be evaluated with respect to their impact on potential emis-
sion levels.
Typically, ESP and boiler operating conditions are not recorded during
the stack test period. Without these data, a comparative baseline (w/1000
acfm) cannot be established. The comparative baseline enables the day-to-day
operation to be accurately evaluated. Long-term degradation of corona power
levels which can occur due to the lost of rapper effectiveness, increases in
flue gas volume, misalignment, or changes in T-R set controllers, is seldom
noticed over a period of months even though the overall efficiency may be
decreasing. In most cases, immediate short-term failures such as rapper
control loss, T-R set controller failure, wire breakage, drag chain failure,
or insulator failure are indicated by changes in corona power levels that
can occur over a few minutes or several hours. Review of corona power
levels by supervisory personnel with comparison to normal values on a daily
basis can allow rapid and correct diagnosis of maintenance problems before
they result in excess emission or catastrophic failure of the unit.
Most recovery boiler ESP's are designed with at least two parallel
chambers equipped with isolation dampers. This allows internal maintenance
to be performed on one chamber while the other chamber is used to carry the
gas volume. However, in order to maintain compliance with emission limits,
the boiler load must be reduced to keep the total gas volume compatible with
the reduction in collection plate area as a result of passing all the gas
through one chamber.
In those systems with a single T-R set per field, the T-R set is
installed in a double half wave design. The T-R set controller is typically
designed to maximize power input based on the secondary siae operating
parameters. Power input to the unit is limited by the side with the lowest
spark point (i.e., closest clearance, heaviest dust cake buildup, or section
with a cold air or oxygen stratification in a single lane). In order to
determine if one chamber is limiting total power input, alternate T-R taps
should be grounded and the power to each chamber evaluated. Major deviations
in voltage or current levels at T-R set limit or spark point indicate
clearance problems, rapper failures, or salt cake buildup on the plates.
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY B-35
-------
CONCLUDING COMMENTS REGARDING Q&K FOR RECOVERY BOILER ESP's
The person responsible for ensuring compliance with applicable emission
limits and for coordinating State and Federal inspections should be familiar
with the design, operating variables, effect of changes in process conditions,
and maintenance practices with respect to the ESP(s).
Unfortunately, in most pulp mills the analysis of ESP operation is
divided among several individuals, none of whom has complete access to all
performance data or a detailed understanding of the interaction between the
process and the control equipment.
A knowledgeable key person, such as an environmental engineer, should
be given the approval of management to review operating records, work
orders, and outage reports; make informed analyses of the nature of equipment
failures; and help to direct preventive maintenance activities. This key
person would be responsible for ensuring that ESP's are properly operated
and maintained. He/she should also work hand in hand with the boiler
operator to keep him/her informed of effective O&M practices.
As an example of good recordkeeping practices, Figure B-13 presents a
specification sheet for a recovery boiler that could be kept by plant
personnel. It provides data on heat input, liquor solids rate,' steam"flow
and pressure, and gas volumes specified by the designer. Comparison of
these values with actual conditions can help the plant engineer to diagnose
ESP malfunctions. Other recommended practices related to the development of
a sound O&M plan are discussed in Section 7 (Model O&M Plan). Many of those
concepts can be considered applicable to the use of ESP's on recovery
boilers.
APPENDIX i-ESP APPLICATIONS IN KRAFT PULP INDUSTRY B_3§
-------
1
vt
-£
is
•K VI
its
i
i
IM
t—
a
X
K
UJ
\f>
Z
K
2
IM
•si
VI
Ik
gr
Z
ft*
i
IM
*-
i
i
a
u HllOfl
B X
t a
K ^
L X
ft
J
ft
tal
e
&.
5
• s
•i
e
IM
KJ
VI
DESIGN PERfORHANCE
1
Z
S
S
e
H-
IM
i
K
tW BTU-TONS/OAY
I
V*
K
2
3
a
S
1
X
g
VI
i
h»
""""""'
I
3
Z
O
«
WEIGHT. DRY SOLIDS
•
DESIGN PERFORMANCE
K
E
0
IM
e
1
1
IM
s
e
a
%
I
IM
5
e
IM
VI
0
IM
h>
X
M
•J
i
s
Z
a
3
at
S
X
er
VI
•«.
IM
VI
M
5
.a
H3ZIH
-ONOD3
£
c
I
o
5
U
3
_t
S
VH/11 U
f
M
I
5
X
1
r
a
3
M
i
^
a
1
IN
S
vi
X
X
u.
33«N»nj
i
a
S
iROP THRU ECONOMIZER
-«2d
SISA1W* 1V1N3W313|
bwon ujvSi
GROSS HEATING VALUE. BOF
BTU/U DRY SOLIDS
UPERHEATED STEAM
o
a
s
i
S
IM
C
Z
w
PSI
s
IM
tc
a
1
IM
i
sw
J *9tf1
t «Y WEIGHT. AS DELIVERED
O
s
(K
5
1
I
Z
a
3
§
0
«01WO
0 ECONOMIZER
ac
u
i
e
0;H
s
c
i
e
K
S
HI
Ml*
dwii
SPECIFIED IY
LIQUOR SPRAY OSCILLATOR NO/UNIT I
£
i
£
LOCATION: SECONDARY NO. CAPACITY/BURNER MlB]
TERTIARY HO. CAPACITY/BURNER Mtt|
STEAM CAPACITY. AUIILIARY BURNERS (APFROX) >,|
-'"-
0 FURNACE
on
IM
ro EVAPORATOR
EAVINC EVAPORATOR
UJ
O
so no?
none 11
C
a
i
SUPERHEATER
a
u
£
X
e
•^
e
10)11
•N03
dH3l
MV31S
o
I
1 PRECIPI1
i
VI
I
i
i
IM
S
X
a
i
U|
KVd
QJ
o
H-
K
r-
sw 3ntj
i
a
E
c
I
i
i
a
i
s
V*
NWJ
OJ
D3S
ec
2
;
£
IM
•J
3
O
1
S1N3H3I
i
E
a
E
B
UJ
*/i
O
K
i*>
NVJ
ai
-
STANCE
X
tllV
3-M JO 'Nl '3DH»1SIS3»
i
w.
X
1
IM
I
EVAP IN EVAPORATOR
S
D
i
X
X
a
X
•K
X
I
i
VI
s
I
i
i
VI
K
•tf
|
I
N SMELT
§
1 REDUCTION OF SALTCAKE MAKEUP
1 REDUCTION OF SULPHUR IN LIQUOR
ilEAM
NO 1 Xflfl 1 VXS i Q IVjH
IKIJ AKvnuny
O)
i.
O
T3
O
I
c
o
O)
o.
l/l
en
1/1
O)
O)
o
-O
03
O
I
00
O)
B-37
-------
REFERENCES FOR APPENDIX B
1, U.S. Environmental Protection Agency. Technical Guide for Review and
Evaluation of Compliance Schedules for Air Pollution Sources. EPA-34Q/1-
73-001-a, 1973.
2. U.S. Environmental Protection Agency. Technology Transfer. Environmen-
tal Pollution Control, Pulp and Paper Industry, Part I, Air, EPA-625/7-
76-001, October 1976.
3. Passinen, K. Chemical Composition of Spent Liquors. In: Proceedings
of the Symposium on Recovery of Pulping Chemicals, Helsinki, Finland,
1968.
4. U.S. Department of Health,'Education, and Welfare. Control of Atmospheric
Emissions in the Wood Pulping Industry. March 1970.
5. Joint Textbook Committee of the Paper Industry. Pulp and Paper Manufac-
ture. Vol. I, The Pulping of Wood. 1969.
6. Rydholm, S. A. Pulping Processes. Intersciences Publishers, New York.
1965.
7. Babcock and Wilcox. Steam/Its Generation and Use. 1978.
8. Personal Communication. J. Blue, Babcock and Wilcox Company.
9. Thoen, G. N., et al. Effect of Combustion Variables on the Releae of
Odorous Compounds From a Kraft Recovery Furnace. TAPPI, 51(8), August
1968.
10. Clement, J. U, J. H. Caulter, and S. Suda. B&W Kraft Recovery Unit
Performance Calculations. TAPPI, 46(2), February 1963.
11. Borg, A., A. Teder, and B. Warnquist. Inside a Kraft Recovery Furnace
- Studies on the Origins of Sulfur and Sodium Emission. TAPPI Environ-
mental Conference, 1973.
12. Bhada, R. K., H. B. Lange, and H. P. Markant. Air Pollution From Kraft
Recovery Units - The Effect of Operational Variables. TAPPI Environmen-
tal Conference, 1972.
13. LoCicero, P. M., and P. E. Sjolseth. Operating Experiences With the
Ace Recovery Furnace Odor Control System. TAPPI Environmental Conference.
No date given.
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
D- Jo
-------
References (continued)
14. Bauer, F. W., and R. M, Borland, Canadian Journal of Technology 32:91,
1954.
15. PEDCo Environmental, Inc. Identification of Parameters That Affect the
Particulate Emissions From Recovery Boilers. June 1982.
16. Environmental Elements Corporation, Baltimore, Maryland.
17. Personal Communication. J. Blue, Babcock and Wilcox.
18. Szabo, H. F., and Y. M. Shah. Inspection Manual for Evaluation of
Electrostatic Precipitator Performance. EPA-34Q/1-79-QQ7, March 1981.
19. Szabo, M. F., and R, W. Gerstle. Operation and Maintenance of Particu-
late Control Devices on Kraft Pulp Mill and Crushed Stone Industries.
20. Final Survey Results for Noncontact Recovery Boiler Electrostatic
Precipitators, J. S. Henderson, J. E. Sirrine Co., TAPPI, December
1980, Vol. 63, No. 12.
21. Low Temperature Corrosion by Sulfuric Acid in Power Plant Systems. V,
P. Gooch Southern Research. ESP Symposium, February 1971.
22. L. Stockman and A. Tansen Gvensk Paperstidn. 62, 907 to 914 (1959).
Abstr. Bull. Inst. Paper Chem. 30, 1164 to 1165 (1960). The Paper
Industry, June 1960, p. 215.
23. Clean Air from Paper Mill Recovery Boilers without Corrosion. J. R.
Zarfoss, Environmental Elements Corporation, Baltimore, Maryland.
National Association of Corrosion Engineers, IGCI. Atlanta, Georgia
for 1976.
APPENDIX B-ESP APPLICATIONS IN KRAFT PULP INDUSTRY
B-39
-------
-------
APPENDIX C
ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
In the steel industry, electrostatic precipitators are used on several
different processes, each with unique gas characteristics. Both wet and dry
ESP's are used. The range of operating conditions encompasses all of the
severe service conditions that were discussed in the main text of this manual.
Table C-l presents a matrix of general service conditions and gas characteris-
tics by process.
Emissions from four sources where ESP's are more commonly applied are
discussed in this appendix: BOF primary emissions, sinter plant windbox
emissions, coke oven battery stack emissions, and scarfing emissions. The
reader is assumed to have an understanding of the processes, which have been
widely described in the literature.
The normal maintenance procedures of inspection, lubrication, and record-
keeping applicable to any ESP are more critical in steel mill applications
because of the harsh, dirty environment, both internal and external, 1n which
ESP's must function. This is particularly true for BOF and sinter plant
applications. Areas of concern common to all steel industry ESP applications
include:
0 Accumulation of dust in dead areas of the ductwork and ESP
° Accumulation of dust and moisture on external structural members,
which cause corrosion.
0 Vibration and oscillation due to maladjustment of ESP internals
resulting from crane movements in the building, railroad movements,
and other disturbances.
0 Temperature and moisture fluctuations caused by air indrafts from
dampers and idle ducts.
° Stress placed on the dust handling component of the system (i.e.,
hoppers, conveyors, etc.) from high particulate loadings.
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
C-l
-------
TARLE C-l. RELATIVE SERVICE CONDITIONS AND GAS CHARACTERISTICS
Process source
Raslc oxygen furnace primary
emi ssions
Rasic oxygen furnace secondary
missions'
Sinter plant windbox
Scarfing
Coke oven battery stacks
Open hearth furnace exhaust
Blast furnace top gas
Coke oven shed gas cleaning
Electric arc furnace exhaust
Cnkn oven gas tar removal
Coal preheatnr
/
/ A*
/%
X
X
X
X
/
/
* / •-J"1'--
X
X
X
X
X
/
^ / $>
•? / <^
X
X
X
X
X
X
/
s? /
y^
X
X
X
X
/
c- /
'///
X
X
X
X
X
X
/
/•//
X
X
X
X
X
X
X
X
/
'/ft
X
X
X
X
X
X
X
X
X
«. /
<^ /
AV
X
X
X
X
„ /
^* /
-------
The following sections describe the characteristics and requirements
unique to the various steel plant applications. Supplemental detail on
process operation, ESP design, and OSM procedures is readily available in
References 1 and 2.
BASIC OXYGEN FURNACE (BOF) PRIMARY EMISSIONS
Although the BOF process is the most widely used steelmaking refining
process, the last one was built in 1978 at Kaiser Steel, The so-called
primary emissions (those resulting from the oxygen blow) are vented to a
collection hood and an ESP or scrubber. Secondary emissions are those gener-
ated by the charging and tapping operations. Roof-mounted ESP's have been
used for control of secondary emissions in Japan, but none have been installed
in the United States.
The raw materials (scrap, molten iron, and lime) are charged into the
vessel, and pure oxygen is blown into the vessel to burn the silicon and
carbon contained in the molten iron. The heat of combustion of these elements
is the major source of energy for the process1, normally, no external fuel is
used. Variations of the process include top blowing, bottom blowing, so-celled
open hooding, and closed hooding.
This discussion focuses on the top-blown, open-hood version of the
process because this is the only configuration in which ESP's are commonly
used. This configuration has fallen from favor and no new units have been
built in the United States for over 10 years. The more modern BOF's are of
the closed-hood design and use high-energy venturi scrubbers. In the open-hood
design, the hood that captures the off-gas is elevated above the furnace or
"open." This allows large volumes of air to be aspirated into the hood,
which burns the CO in the off-gas. Complete combustion is critical to the
prevention of explosions in the ESP. Monitoring for explosive gas conditions
is necessary in ESP installations, as several explosions have been experi-
enced.
A shop typically has two vessels that operate alternately and use a
common ESP. Gas volumes, which can be calculated from the oxygen rate and
A
characteristics of the charge, generally run about 2000 scfm/ton steel.
Most furnaces produce 200 to 300 tons of steel per batch, or flow rates on
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
C-3
-------
the order of 400,000 to 600,000 scfm. A 250-ton BOF requires a capacity of
775,000 acfm at 550°F and 10 in. HgO. The gas stream is made up of 234,000
cfm off-gas, 194,000 cfm indraft air, 30,000 cfm air leakage from the down
vessel, and 308,000 cfm of water vapor,
The operation of an ESP system applied to a BOF is strongly dependent on
the overall gas collection system because conditioning of the gas is required,
and failure in any part of the system can cause ESP malfunctions. Figure C-l
is a simplified diagram showing the key elements of a BOF ESP system.
The violent reaction in the refining vessel gives rise to high levels of
fine particulate, as shown in Tables C-2(a) and C-2(b). Emissions are highest
in the early part of the blow; total emission rate is about 30 Ib/ton of
steel. A complicating factor in BOF operation is the extreme cyclical
nature of the process. The oxygen blowing cycle lasts about 20 minutes.
After blowing, the testing, tapping, and recharging operations typically take
another 20 to 30 minutes. During these periods, the gas collection system is
still drafting; however, the gas volume, temperature, and particulate loading
are entirely different from those during the blow. Cooling of the ductwork
and ESP during these periods creates a potential for both condensation and
warpage. Even during the blowing period itself, the gas flow rate varies
widely, as illustrated in Figure C-2. Notwithstanding these conditions,
well-maintained and operated ESP's have proven to be very effective control
devices on BOF's. Inlet loadings are about 7 to 10 gr/scf, and an efficiency
on the order of 99.8 percent is required to meet NSPS and other applicable
regulations, which specify about 0.02 gr/dscf. Table C-3 gives performance
data on several BOF operations.
The variable gas conditions require gas conditioning. The gas leaving
the vessel is 90 percent CO and 10 percent CO. at about 3000°F. As the gas
is burned by aspirated air in the hood, temperatures in excess of 4000°F can
be developed. The first- stage cooling occurs in the water-cooled hood.
Hood design is critical and represents a high maintenance item. The charac-
teristics of hood designs are summarized in Table C-4. Although not part of
the ESP, a great deal of attention is focused on hood maintenance because of
the high cost of hoods and the severe environment in which they operate. Gas
leaving the hood enters a water spray chamber for cooling to 600°F or less
before it enters the ESP.
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
C-4
-------
NO. 2 FURNACE^ !
NO. 1 FURNACEl
T"
i
i
II
GAS COOLING
SPRAY WATER
ISOLATION DAMPERS
FOR OFF-LINE FURNACE
SPARK BOX PLAN VIEW
(ALTERNATE IN PLACE OF SPRAY WATER
WET EVAPORATION CHAMBER)
ISPI
ER)
[f^teTi
HOOD COOLING WATER
EFFLUENT
WATER TO
THICKENER
DRY
PRECIPITATORS
EXHAUST
STACK
INDUCED-DRAFT FAN
SIDE VIEW
Figure C-l. Typical configuration for a precioitator installed on a BOF.1
-------
TABLE C-2(a). TYPICAL PARTICLE SIZE DISTRIBUTION OF OPEN-
HOOD, TOP-BLOWN BOF EMISSIONS6
Particle diameter, (urn)
<1
1-65
69-90
90-110
>no
Weight percent
25
15
20
15
25
TABLE C-2{b). PARTICIPATE COMPOSITION FROM OPEN-HOOD
COLLECTION SYSTEMS6
Component
Fe total
Fe metal
Fe as FeO
Fe as FejOi,, Fe203
CaO
Si02
Weight percent
59
--
1.6
57.4
2
1
C-6
-------
SHOP A
*
u
UJ
BLOWING TIHE
Figure C-2. Variation in gas flow rate from a BOF
during the course of a heat.7
(Permission granted by APCA.)
C-7
-------
TABLE C-3. OPEN HOOD SYSTEM PERFORMANCE DATA8
Plant
U.S. Steel South Works,
Chicago, 111 - Scrubber
CF&I Steel, Pueblo, Colo. - ESP
Rebublic Steel, Buffalo, N.Y. - ESP
Wisconsin Steel, Chicago, 111. - ESP
Jones & Laughlin, Aliquippa, Pa. - ESP
Youngstown S&T, Indiana Harbor, Indiana
(now J&L Steel } - ESP
Crucible Steel Midland, Pa. - Fabric Filter
Test date
6/27/77
6/29/77
7/01/77
4/10/78
4/11/78
4/12/78
10/20/75
10/21/75
10/22/75
11/10/76
11/12/76
11/16/76
11/18/76
8/10/76
8/11/76
6/12/78
6/13/78
6/14/78
6/11/80
6/11/80
6/12/80
6/12/80
Emissions
(outlet
concentration) ,
gr/dscf
0.0039
0.0038
0.0045
0.0038
0.0062
0.0056
0.0049
0.0044
0.0073
0.0072
0.0052
0.0052
0.0220
0.0098
0.0115
0.0097
0.0094
0.0120
0.0130
0,0120
0.0092
0.0132
0.0094
0.0115
0.0040
0.0014
0.013
0.008
0.012
0.0021
0.0019
0.0032
0.0028
C-8
-------
TABLE C-4. COMPARISON OF DIFFERENT TYPES OF HOOD CONSTRUCTION9
Initial cost
Ability to take
high temperature
Ability to take
temperature change
Resistance to slag
buildup
Resistance to scaling
Maintenance cost
Refractory-
1 ined
Lowest
Poor
Poor
Poor
—
Very high
Water-
cooled
plate
panels
Low
Fair
Fair
Good
Poor
High
Formed
panels
Moderate
Good
Good
Good
Fair
Fair
Double-pass
High
Very good
Very good
Very good
Good
Low
Waterwall
Boiler
High
Very good
Very good
Fair
Good
Fair
Membrane
High
Very good
Very good
Good
Good
Fair
(Permission granted by AIME.)
o
I
-------
Water sprays are temperature-controlled; additional sprays come on as
the temperature rises during the course of the blow. Ideally, all of the
water evaporates to avoid caking in the chamber and carryover of water droplets
into the ductwork and precipitator. This requires a fine water spray and
appropriate nozzles. Nozzle erosion and nozzle pluggage due to dirty water
must both be avoided,
During the early part of the blow when the gases are colder, steam is
usually injected to maintain the moisture content in the gas. This moisture
content and the gas temperature influence dust resistivity, as shown In
Figure C-3. Although ESP's are not used on partial combustion (closed hood)
systems in the United States, Figure C-4 shows that the general dependence
of resistivity on temperature and moisture is the same for the dust from
these systems.
When the lime is charged, usually at about the 2-minute mark of the
blow, an additional burden is placed on the ESP because of the carryover of
lime particles, which have extremely high resistivity.
Electrostatic precipitators on BOF's contain six to eight chambers with
three to five fields per chamber. Collection area requirements, as shown in
9
Figure C-5, range from 200 to 600 ft /1000 acfm. The newer systems are at
the high end of this range, and requirements on some are even higher. The
systems are obviously large with concommitantly large fans and ductwork.
Multiple fans are used. The ESP should be designed so that each chamber can
be isolated for maintenance-purposes, and it should be insulated and heated
to prevent condensation.
Because of the heavy dust loading, the dust hoppers and screw conveyors
are critical elements of the system. Sensors should be used to detect drive
or screw failure before dust buildup occurs in the hopper. Hoppers should be
sized to hold at least one day's dust accumulation to provide a margin of
safety in the event of screw conveyor malfunctions; however, the goal should
be to keep the hoppers continuously evaluated. Condensation or external
moisture in any part of the system must be avoided because the high lime
content of the dust makes it cementitious.
As shown in Figures C-3 and C-4, control of temperature and moisture are
critical to ESP performance. Consequently, an evaluation of ESP 08H must
includean evaluationofthegasconditioning system. Particularly important
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
C-10
-------
1 x 10
1 x 10
100
200 300 400
GAS TEMPERATURE, °F
500
600
Figure C-3. Resistivity vs. Gas Temperature for BOF Dust.10
-------
10
12
CMC11
,10
ift
o
o
t—>
Of.
o
10"
10
100
200 300 400
TEMPERATURE,°F
500
600
o
I
ro
Figure C-4. Electrical resistivity of the partially combusted
converter dust.11
(Courtesy of Association of Iron and Steel Engineers,
extracted from Iron and Steel magazine, September, 1978, Page 44..)
-------
99.9
99.8
99.0
98.0
S 9S.O
<_>
90.0-
85.0-
80.0
II I I I I I (T
70.0L_L_J_
Note: Line for design trend
represents range of values.
DATA FROM MANUFACTURERS.
USERS. AND OPEN LITERATURE.
I I I I I I
8 S! °
8
I I lil§8if?fs?i
SPECIFIC COLLECTION AREA. ftVlOOO acfm
Figure C-5. Selected precipltator correlations for basic oxygen furnace,
based on a modified Deutsch Equation.2
o
-------
are temperature sensors, water spray controls, the steam injection control
loop, spray water quality, spray nozzle condition, and indraft air control
during nonblowing periods. Dampers can be placed in the inlet to the ESP to
minimize the indraft air and the resulting cooling of the precipitator during
down periods. All precipitators operate better at a high temperature because
of the increased ion mobility in the corona. Excessive temperature must be
avoided, however, to control resistivity and to avoid structural impairment
of the ESP components,
Because of the severe operating conditions and extremely high efficiency
required of BOF ESP systems, a spare chamber and spare fan should be provided
to permit the shutdown of chambers for routine inspection. A thorough internal
inspection of each chamber should be conducted at least twice a year (pref-
erably once a quarter). This type of schedule means at least one chamber is
always under maintenance for repair of broken and misaligned wires, broken
insulators, rappers, and screw conveyers.
The recommended minimum instrumentation for a BOF system in addition to
that described earlier for any ESP installation includes;
Low pressure alarms; For instrument air, oxygen
supply, lance cooling water,
hooding cooling water, service
water, waste gas duct, clean gas
duct, plant air
Low level alarm: For hood water cooling tower
High temperature For cooling water, dirty gas at sensors and
alarms: precipitator inlet
Failure alarm: For precipitator transformer-
rectifiers
Vibration sensors
and alarms: For all fans
High bearing temperature
sensors and alarms: For all fans
Many of the above instruments should be equipped with continuous strip
charts to record data (e.g., oxygen supply, water flow rates, temperatures at
various points, system draft at various points). In precipitator systems,
APPENDIX C-ESP APPLICATIONS IN WON AND STEIL INDUSTRY
C-14
-------
combustible content of gas (or CO concentration) is an additional important
1 ?
process variable that should be monitored, and alarms should be included. *
At a typical large plant {2.5 x 10 tons/year) where good maintenance
practices are followed, it takes 75 to 100 hours/week for proper maintenance
of an ESP system, including sprays, dust valves, hoppers, and gas cleaning
system. •
Because BOF shutdowns can be frequent, depending on production schedules,
startup procedures for the ESP are important and should include the following
fundamentals.
1) Inspection of hoppers and screw conveyors to be sure they are clear
of accumulated dust and working smoothly.
2) Air should be allowed to pass through the system to purge the
ductwork and ESP of any CO accumulations and to establish a minimum
startup temperature of 180°F before the ESP is energized. Radiant
heat from the vessel can be used for warmup, or if the system has
been shut down for a long period, steam injection can be used.
Alternatively, the blow can be started before energization. This,
of course, permits essentially uncontrolled emissions during the
warmup period.
3) Energization should be in the following order: 1) dust removal
system, 2) rapper system, and 3) T-R sets.
The normal operating temperature of BOF ESP's is 500° to 600°F. Alumina
insulators permit the ESP to function at temperatures up to 850°F during
brief periods of temperature excursion. The alternatives to such excursions
are the opening of relief dampers to allow the raw fumes to escape to the
atmosphere or the automatic shutoff of the oxygen and retraction of the
lance. The structural members of the ESP also must be designed to withstand
the higher temperatures during excursions.
Most installations have multiple induced-draft fans. The outlet plenum
and gas distribution devices should be designed to maintain uniform flow
through the multiple chambers of the ESP. Fans have a tendency to pull more
air through those chambers closest to the fan. Unless compensation is made
for the changes in the fans that are on line, this can cause excessive gas
reentrainment, uneven distribution of dust in the hoppers, and reduction in
collection efficiency.
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
C-15
-------
Wire breakage rates of 2 to 200 per year have been reported, but rates
on the order of 2 to 10 per year are more realistic.
Table C-5 presents the most common O&M problems encountered in BOF ESP
systems.
Two variations in BOF ESP applications that have been applied outside
the United States are the roof-mounted ESP for process fugitive emissions and
a novel ESP that can operate on partial combustion systems (i.e., high CO in
- the off-gas). These are not covered here in detail, but the basic arrangements
are shown in Figures C-6 and C-7. The roof-mounted ESP (Figure C-6) operates
on natural draft without fans and uses wet flushing rather than rappers. The
electro-precipitator for partial combustion systems, which has been applied
in Europe, addresses the problem of treating explosive gas with high CO
content. After it has been cleaned, the gas is usable as byproduct fuel.
The torpedo shape of the unit prevents potential stagnant areas where gas
could accumulate.
Sinter P1 arvt Wi nd boxEnriss 1_ons
The sintering process is used to agglomerate fine iron- bearing wastes
and fine ores for use in the blast furnace. Combustion air is drawn through
the materials, which are bedded on a moving grate (strand), into plenums
(called windboxes) below the grate. These windboxes discharge into a header
that carries the waste gas to a control device. A cyclone is sometimes used
as a precleaner to remove heavy particulate. The windbox gas stream contains
0.5 to 3 gr/scf of particulate and its temperature ranges from 200° to 400°F.
The temperature can fluctuate widely during startup and shutdown and cause
the gas to pass through its dewpoint. This is especially critical for gas
with higher acid concentrations. The particulate is extremely abrasive. If
roll scale or other oily material is used in the raw material mix, the exhaust
contains unburned hydrocarbon aerosols, which are extremely difficult to
remove and can cause ESP fires. At temperatures above 200°F in the outlet
side of the ESP, oil volatilizes and fine metal particles oxidize, which
causes self-ignition. These "glow fires" ca-n ignite carbon and oil mist in
14
the ESP. The exhaust contains 100 to 500 ppm of SCL from sulfur in the
mix. This creates low pH and corrosive conditions in wet systems. The
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
C-16
-------
TABLE C-5. BOF ESP OPERATION AND MAINTENANCE PROBLEMS
1,13
Problem
C«use
Potential le«eSies
Corrosion of collected suf-
fices, especially at the
bottom zone of the outlet
half of ESP
Structural damage to ESP
Insulator burnout and sub-
sequent shorting out of
sections
Broken »rires
Performance degradation
due to poor ges con-
ditioning
insulator fai tyre
Fan/motor failure
Malfunction of dust
removal equipment
Condensation due to gas
temperature going below
dew point.
Alternate heatinq and cooling.
High temperature excursions.
Moisture end dust buildup on
insulators.
Moisture and dust buildup on
Hires.
Deviation of gas tempersture
and moisture from ideil range.
High temperature excursions.
High temperature.
Bearing failure.
Vibration.
Broken screw conveyer shifts.
Plugged dust valves.
Duit bridging In hopper.
Hopper heater fjilgr*.
Hopper Vibrator failure.
Bartwp after shutdowns,
Control of cold air indraft from
down vessel or other sources.
Steam injection during early part
of blow to wtntain temperature.
Maintenance of air seals on hoppers
to prevent cold air ineraft,
ESP Insulation and/or heating.
Avoidance of tenperature ovtrloads
through temperature control system
•nd water sprays on g»s.
General operation to maintain is steady
state temperature is possible.
Design of structural components to
withstand te«|>*rature.
Avoidance of condensation, is described
above.
Providing fan and filter to supply
sir to fnsulitor housing.
Frequent (once/Meek) inspection.
Avoidance of condensation, as described
above. •
Frequent inspection and replacement
of water spray nozzles.
Use of clean or treated water w avoid
nozzle pluggage; use of *c»le inhibitors.
Inspection ana calibration of temperature
control loop.
Temperature control by use of water
spray system,
Use of alumina insulators.
Provision of fin and filter to supply
air to Insulator housing.
Frequent inspection.
Frequent cleaning of motor cooling
fins in dysty enyironrnent.
Bearings with temperature alarm.
Vibration sensor and alarm.
Conveyor on/off indicator lights
Level indicstors in hoppers, pre-
ferably high and lo* level.
Regular djst retnoval schedule.
Fr*quent Inspection (daily).
C-17
-------
NEW ROOFMOUNTED ELECTROSTATIC
PRECIPITATORS
27,000 ,m3/min (954,000 cfm )-
91.5% EFFICIENCY
170 kW , 45 kV .
NEW SHEET METAL EACH
SIDE OF BOF
CHARGING EMISSIONS.
EXISTING BOF PRIMARY
FUME DUCT
REMOVE EXISTING ROOFING
NEW SHEET METAL PARTITION
TAPPING EMISSIONS
Figure C-6. Cross section of BOF having roof-monitored ESP.12
o
I
oo
-------
COOLING STAC
INJECTION OF WATER
EVAPORATION COOLER
STACK WITH WASTE GAS TORCH
ELECTRO-PRECIPITATOR
nn
.OIL
!\ /|!
v./
1V71
Y ! Y Y
AjlAjLA
Figure C-7. BOF and Q-BOP waste gas dust collection plant."
(Courtesy of Association of Iron and Steel Engineers,
extracted from Iron and Steel magazine, September, 1978, Page 44.)
o
I
-------
sometimes significant alkali and fluoride content of the raw materials can
cause acid attack of even stainless steel components.
Sinter plant windbox emissions have been controlled by scrubbers, wet
ESP's, dry ESP's, and fabric filters; the high-pressure drop scrubbers appear
to be the most effective (>60-in. w.c.).
A typical particle size distribution in the windbox gases is as follows:
15 to 45% < 40 ym
9 to 30% < 20 ym
4 to 19% < 10 ym
1 to 101 < 5 w
The chemical composition of windbox dust is highly dependent on the raw
materials used. Table C-6 presents a range of composition. Emission rates
from the windbox average 11.1 Ib/ton of sinter, Precleaning with a cyclone
removes the heavy particles, which amount to about 20 percent of the total.
The particulate loading entering an ESP 1s therefore about 8.7 Ib/ton,
2
TABLE C-6. CHEMICAL COMPOSITION OF SINTER STRAND WIND BOX DUST
2
Compound
Fe203
Si02
CaO
MgO
A12°3
C
S
Alkali
Wt %
45-50
3-15
7-25
1-10
2-8
0.5-5
0-2.5
0-2
The flow rate depends on grate area, as shown in Figure C-8. The sinter-
ing process is operated as a continuous process, and except for the raw
material mix, operating conditions are usually uniform. The plant is typically
shut down for maintenance once a week for a period of one shift.
The particulate, which consists mainly of metallic oxides, silica, and
limestone, is very resistive. As indicated in Figure C-9 it may be the order
This limits the current densities and the applied
of 1011 to 1014 ohm-cm.
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
O2Q
-------
800
700
600
V
VI
g 500
o
400
300
200
100
12,897 + 205.6 (GRATE AREA)
I I I I
500 1000 1500 2000 ? 2500 3000 3500
GRATE AREA, ft*
Figure C-8. Flow required for sinter plant windbox control.15
C-21
-------
TEMPERATURES, °F
200 300
PLANT A
BASICITY 4,0
PLANT A
BASICITY 4.0
..' PLANT B
.-•"BASICITY i.o
PLANT C
BASICITY 4.0
PLANT D
BASICITY 4.0
PLANT E
BASICITY 1.0
100 150
TEMPERATURE»aC
200
Figure C-9. Effect of temperature and sinter basicity on resistivity
of sinter-plant participate.2
C-22
-------
voltage at the discharge electrodes to avoid back corona formation. Thus,
the particle charging process and the electric field at the plates for par-
ticulate collection are marginal. This situation appears to be analogous to
the behavior of high-resistivity fly ash found in some utility applications.
As the basicity [i.e., the amount of limestone (or dolomite) used in the
mix] increases, resistivity also increases and causes degradation in perform-
ance. More rapping and less gas velocity can overcome this effect to some
extent, but dry ESP collection is not generally effective at the high basicity
(2.5 or more) used in modern sinter plants.
Figure C-10 shows the collection area requirements for ESP's operating
on sintering machines equipped with mechanical collectors. Typically, single-
stage, horizontal-flow units have been used. Carbon steel is the predominant
material of construction, even though acid corrosion is a major concert.. The
power supply generally consists of a single-phase, high-voltage transformer;
appropriate control equipment; and s bridge rectifier circuit. The latter
may use mechanical-type rectifiers, the newer Kenotron vacuum tubes, selenium
rectifiers, or silicon rectifiers. Normal transformer ratings are between 25
and 50-kVa, 440 volts (primary), and 50 to 75 kV (secondary). The collection
area consists of dust-collection electrodes with high-voltage discharge elec-
trodes uniformly spaced and of uniform length.
Collection efficiency on the order of 98 to 99 percent is required to
meet typical state regulations. Average design parameters to achieve this
are:10
Gas velocity 4.3 ft/s
Inlet gas temperature 245CF
Electric field 8.1 kV/in.
Inlet dust loading 1.0 gr/acf
Precipicator power 71 watts/1000 cfm
Even with these parameters, opacity regulations may be impossible to meet
because of the opacity associated with fine hydrocarbons that pass through
the ESP.
The materials charged to the sinter machine can change during process
cycles (depending on the availability of wastes, ore fines, lime, etc.), and
these changes can affect the properties of the dust. Sas conditioning is not
used in the sinter plant application of ESP's. The basicity of the sinter
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
C-23
-------
99.9
99.8
99.0-
98.0-
95.0
90.0
85.0-
80.0-
70.Ol 1
Note:
Line for design trend
Represents range of values.
DATA FROM MANUFACTURERS.
USERS, AND OPEN LITERATURE.
J I
I I I
o o 9
co «»« 2
o
o
x>
o o o ooSoSo
i**» OO O\ ^J ^~ cvj r*i ^ to '
SPECIFIC COLLECTION AREA. ftVIOOO acfm
Figure C-10. Selected precipitator correlations for sinterinq process,
based on a modified Deutsch Eauation.2
-------
far override any other factor In Influencing resistivity. Changing to super-
fluxed sinter (basicity 2.5 to 3.5) has reportedly caused collection efficiency
to drop 30 percent.
In an attempt to address the resistivity and oily aerosol problem, wet
ESP's have also been used on windbox exhaust streams. Control of liquor pH
by caustic addition is essential in wet systems.
Table C-7 lists sinter plant process variables that can affect ESP
performance. Table C-8 presents specific O&M problems.
A sinter plant has many raw material conveyors and material transfer
points, which are often vented to either the windbox or the discharge-end
control device. Careful consideration must be given to the design and opera-
tion of auxiliary venting and collection from these points to avoid flow
imbalance and gas temperature and moisture fluctuations. Some plants have
multiple sinter strands vented to the ESP. In these cases, the startup of
one strand while the other is on line can cause cold air to enter the systen
and create condensation of moisture or acid gas.
When the sintering machine is started after & shutdown, it takes about
an hour for the process to reach full operating temperatures. During this
time, condensation may take place in the ductwork and in the dust collection
equipment. The ESP's are usually not powered fully until the gases have
become warm enough to evaporate the condensed moisture. A cold startup will
result in an hour or so of uncontrolled emissions. As a result of normal
downturns, startups can occur once a week. The following startup procedures
are paraphrased from reference 15.
1. Use a roll of paper to cover the empty strand. Remove the paper
when it reaches the discharge end of the machine.
2. Start sinter cooler fans.
3. Start hearth layer conveyors (if a hearth layer is used)
4. Start the dust conveyors that take the collected dust from the
windboxes and emissions controls and return it to the feed system.
5, Start windbox control unit.
a. The dust conveyors must be running before the power unit can
be started. After startup, the interlock is bypassed and the
ESP continues to run.
APPENDIX C-ESP APPLICATIONS IN (RON AND STEEL INDUSTRY
C-25
-------
TABLE C-7. PROCESS PARAMETER CHECKLIST FOR SINTER PLANT WINDBOX CONTROLLED BY ESP
Process parameter
Deviations from
normal practices
Effect on ESP
Raw material mixing
Usage of BOF dust
Usage of mil 1 scale
B.F.a flue dust
B.F. Sludge
Basicity
Sulfur throughput
Ignition
Improper mixing
Higher usage
Higher usage
Higher usage
Contaminated with oil
Higher basicity (>1.0)
Higher-sulfur material
Improper ignition
Unsintered portions will generate
emissions and increase the load
on the ESP. Excess emissions
from ESP stack.
Excessive fine materials; increased
emissions; increased KC1 and NaCl
emissions.
Excessive hydrocarbons; increase in
opacity (bluish color); danger of
ESP fires.
Increased emissions of chlorides
(KC1 and NaCl)
Excessive hydrocarbons; increased
opacity.
Increased resistivity of limestone
particles; reduced ESP efficiency
(5 to 10%).
Corrosion problems in ESP due to SO-
Unsintered portions will generate
emissions and place excessive load
on ESP.
(continued)
i
f-o
en
-------
TABLE C-7 (continued)
Process parameter
Deviations from
normal practices
Effect on ESP
'Burn-through
Gas temperature
Grate maintenance
Sinter machine seals
and ductwork main-
tenance
a) Slower than normal
b) Faster than normal
a) Higher than normal
b) Lower than normal
Poor maintenance
Poor maintenance;
excessive air leakage
Unsintered portions generate
emissions and place excessive
load on ESP.
High slag formation, bad sinter
quality, low gas flow rate to ESP.
Higher volumetric gas flow through
ESP than the designed value; poor
gas distribution and reduced
efficiency.
If below dew point, acids formed
because of S02 presence will lead
to higher corrosion of ESP parts.
Unsintered material falls through
the sinter bed and increases the
load on ESP.
Higher volumetric gas flow through
ESP than the designed values; poor
gas distribution in ESP; and reduced
ESP efficiency.
B.F. = blast furnace.
n
i
-------
TABLE C-8. SINTER PLANT ESP O&M PROBLEMS
Problem
Cause
Potential Remedies
Fires, explosions
Fan/motor failure
Dust removal equipment
manfunction
Corrosion of internals
Self ignition of oily matter
and carbon in ESP outlet or
dust bins
High temperature
Bearing failure
Vibration
Broken screw conveyer shafts
Plugged dust valves
Dust bridging in hopper
Hopper vibrator failure
Temperature excursion below
dewpoint
Maintain outlet temperature
at 200°F or less.
Minimize oily raw materials.
CO monitors and alarms in
ESP outlet.
Frequent cleaning of motor cooling
fins in dusty environment.
Bearings with temperature alarm.
Vibration sensor and alarm.
Conveyor on/off indicator lights
Level indicators in hoppers, pre-
ferably high and low level.
Regular dust removal schedule.
Frequent inspection (daily).
Minimize indrafts.
Avoid auxiliary vents on
conveyor transfer points.
Temperature sensors/alarms.
Careful startup/shutdown
procedures.
IX)
oo
-------
After starting the dust conveyer system, determine that air
(414 kPa, 60 psig) is available for rapper controls; provide
power to the rapper control panel; turn on suspension insulator
electric heaters; determine that all grounding switches are
properly set and all doors closed and locked; admit air to the
precipitator; allow 60 minutes for wannup to drive off any
condensed moisture from the insulators; put ESP power supply
into operation; after about one minute to warm up tube fila-
ments, apply plate voltage; review readings of current and
voltage to the primary of the high voltage step-up transformer
and the voltage drop across the power saturable reactor;
review power output as shown by control potentiometer.
Note: Admitting air to the ESP involves startup of the windbox
fan. The load on this fan is regulated by a hydraulic-electri-
cal mechanism that adjusts the fan inlet damper. The regulator
will maintain a manually set inlet section pressure control
point. When the main draft fan is stopped, the regulator
automatically closes the inlet damper to prevent cold air
inleakage to the ESP.
The regulator hydraulic pumping unit must be started before
the fan can be started. After the regulator is started, the
draft fan is started according to the following steps: turn
on the excitation power supply (normal or emergency) for the
motor field; actuate the switches of the high-voltage switch
gear unit; start the motor; after allowing some 20 seconds to
accelerate to full speed, automatically apply the field at the
proper instant, and pull into synchronization. If starting is
unsuccessful, the high-voltage circuit breaker will trip out,
and a waiting period is required before attempting to restart.
To stop the windbox fan, first trip the fan switch. Interlocks
reset the regulator timer, close its contacts, and thus .drive
the damper to its closed position. The a.c. power for field
excitation is then turned off, and the regulator pump unit is
stopped.
6. Start sinter machine.
Wet ESP's have been demonstrated as capable of controlling windbox
exhaust emissions. A continuous water washing of the plates prevents an
insulating layer of pollutants from forming and nullifies the effect of the
increased resistivity of the particles caused by superfluxing the sinter.
The gas stream is kept saturated by continuous sprays; as the particles and
liquid droplets collect on the plates, a film forms that continuously drips
off and carries with it the collected dust, and thus prevents any buildup.
APPENDIX C-ESP APPLICATIONS IN IBON AND STEEL INDUSTRY
C-29
-------
The approximate water requirement for thir type of system is 5 to 9 gal/1000
scf.17
The water sprays reduce the temperature of the inlet gases to 100° to
120°F; therefore, the condensible hydrocarbon portion of the total participate
emission is available for collection. Some hydrocarbons condense into a fine
mist and are collected on the plates. These small-sized hydrocarbon particles
have a low dielectric constant and are not removed as efficiently as dust
particles with dielectric constants greater than 10.
The maximum inlet loading for proper operation should be 0,25 gr/scf,
which could be achieved by installing a primary mechanical collector upstream,
Also, the pH of the recycled water should be monitored and a caustic solution
should be injected to neutralize any acidity. The combination of chlorides,
fluorides, alkalis, and SO, places great stress on the materials of con-
struction.
Coke Oven Battery Stacks
Unlike the two previous applications, the control of coke oven battery
stacks is a relatively recent development (within the last 5 years). Only
four to five batteries have installed cr experimented with ESP control of
emissions. Battery stack emissions are the products of combustion of the
fuel gas used to heat the coke oven batteries. This gas can be coke oven
gas, blast furnace gas, natural gas, or a mixture of these. Because these
gases are relatively clean, control would not be necessary were it not for
coal particles and pyrolysis products leaking from the coke oven chamber into
the flue chamber. When this leakage cannot be controlled by maintenance of
the refractory dividing wall between the oven and the flue, a control device
must be installed in the exhaust duct. Thus far, all such installations are
retrofits; no new coke oven battery has been built with a control device on
the stack.
The complicating factors in ESP application are the extremely low resis-
tivity of the particles, which are predominantly carbon; the tarry nature of
the emissions; and the sporadic nature of leakage. Although the overall
battery operation is generally uniform and continuous, leakage may be from
only certain ovens and occur at specific times during the cycle resulting in
a "puffing" nature of the emissions.
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
C-30
-------
18
Particulate emissions range from 2.6 to 124 Ib/h and average 29.5
Ib/h. The flow rate from a single battery ranges from 30,000 to 80,000 dscfm
and is proportional to the capacity of the battery and the excess air used
18 1 ft
for firing. Exhaust temperature is in the range of 400° to 600°F.
Depending on the fuel used, the gas stream can contain up to 3500 ppm of
18 18
SOp. Moisture content ranges from 5 to 15 percent. Particle sizing is
given in Figure C-1I. Electrostatic precipitators have not been particularly
effective in particulate removal, as reflected in Table C-9.
The dry ESP's used are similar to those used in coal-fired powerplants.
The design incorporates high SCA (333 to 880 ft /1000 acfm), multiple stages,
and low gas velocity. Design parameters are given in Table C-10. MRI suggests
that units should be insulated to avoid corrosion and that dust hoppers
should be heated and equipped with vibrators and hopper sides inclined at 60
20
degrees or more. The poor performance (even at high SCA's) is attributed
to the low inlet dust loading and the low resistivity of the carbon partic-
ulates. Because carbon releases the charge imparted, the particles avoid
collection; they bounce off the collecting electrode and become reentrained
in the process.
Operating experience is insufficient for the development of specific
operation and maintenance guidelines. One company has noted visible sparks
in the dust bin and ignition in the hoppers. Another found it necessary to
replace the vibrator-type rappers with electromechanical rappers and to
pressurize insulator compartments and circulate heated (25G°F) air through
20
these compartments. The poor performance of dry ESP's suggests application
will not be widespread.
A pilot-scale test was run with a wet ESP at a battery that has since
been shut down. This trial produced a particulate removal efficiency of
about 93 percent. The gas flow rate to this pilot-scale unit ranged from
about 800 to 3000 acfm. The unit contained a low-energy (1/2 in. w.c.)
fiberglass venturi gas-conditioning scrubber to cool the gas, achieve uniform
velocity, and remove larger particles. The design of the wet ESP section was
unconventional in that a water film on concentric plastic rings served as the
collecting electrode. No water spray was included; estimated water require-
ments were 230 gal/min for a full-scale unit (equivalent to 3.3 to 3.8 gal/1000
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
C-31
-------
lu.u
9.0
R 0
O . VJ
7.0
6.0
S.O
4.0
« 3.0
ut
Ul
S 2.0
o
I
o
o
at
UJ
•t
1.0
0.9
. 0.8
0.7
0.6
0.5
0.4
n 1
1 1 i I I 1 1 1 1 (
_.
A ARMCO (TESTED BY SOUTHWESTERN LAB. ) - BRINKS IMPACTOR
OP3-ST J&L (TESTED BY BETZ ENVIRONMENTAL) - ANDERSEN
• P4 J IMPACTOR KITH STAINLESS STEEL COLLECTING PLATES
• KAISER (TESTED BY APT) - U.W. MARK III IMPACTOR WITH
GREASED ALUMINUM SUBSTRATES
OSHENANAGO (TESTED BY BETZ ENVIRONMENTAL) - ANDERSEN
IMPACTOR WITH STAINLESS STEEL COLLECTING PLATES
_
NOTE: RESULTS SHOWN ARE AVERAGES
OF PARTICLE SIZE DATA CONTAINED IN
EACH TEST REPORT.
"
-
• A
A
-
•
-
-
• A
-
1 1 1 1 1 1 I 1 1 1
I 1 1
O
• ° -
0
• ^ —
o
•a
A
O
O
«D
-
O
«
-
-
-
1 i i
5 10 20 30 40 50 60 70
MASS PERCENT LESS THAN STATED SIZE
80
90 95 98 99
Rmjre C-11. Coke oven particle size distribution determined by use of cascade impactors.
18
o
I
-------
TABLE C-9. PERFORMANCE DATA FOR ELECTROSTATIC
PRECIPITATOR CONTROL OF COKE OVEN BATTERY STACK PARTICULATE EMISSIONS
18
Lone Star Steel
E. B. Germany Works
Batteries A and B
(February 1973)
Test 1
Test 2
Test 3
Armco Steel
Houston Works
Battery 1
(November 1976)
Test 1
Test 2
Test 3
Armco Steel
Houston Works
Battery 2
(November 1976)
Test 1
Test 2
Test 3
National Steel
Granite City
Battery C
(July 1979)
Test 1
Test 2
Test 3
Participate
loading, (gr/dscf)
Inlet
0.051
0.046
0.081
0.009
O.OOB
0.007
0.075
0.033
0.113
0.403
0.162
c
Outlet
0.043
0.031
0.039
0.009
0.048
0.004
0.013
0.006
. 0.011
0.370
0.164
0.105
Collection
efficiency,
weight %
12.9
32.7
49.8
2.2
33.9
45.6
80.2
79.4
89.4
Neg.
Neg.
c
Outlet
participate
emissions,
Ib/ton
0.35
0.24
0.31
0.09
0.05
0.04
0.10
0.05
0.09
6.27
2.81
c
aProbe, cyclone, and filter catch (EPA Method 5)
b,
Combined stack.
cInvalid sample.
C-33
-------
TABLE C-10. COMPARISON OF ESP DESIGN AND PERFORMANCE CHARACTERISTICS AT
LONE STAR STEEL, ARMCO STEEL HOUSTON WORKS, AND PILOT TEST
Piste area, ft*
Plate-to-plate spacing
(a) Inlet, in.
(b) Outlet, in.
Corona wire diameter, in.
Specific collection area,
ft VI .000 cfm
Ri§raticm velocity, in,/s
(a! Design
(b) Actual
Cas velocity, ft/s
Secondary voltage,* kV
Current density," m-smps/ft
Electrical sets, totil
number; eonf igyration
Mass efficiency, *
(i) Ctsijn
(b) Actual
Materials of construction
Casing
Hoppers
Discharge electrodes
Collecting surfaces
Arnce StMl
Unit No, I
«Q,720
9.0
S.O
0.09?
582C
Q.3?d
0.00-0.2*
2.06
J5 (average)
O.CSS (i*g)
3 in series
72.9
44.1
AST* A-36
ASTM -A-36
316 (stain-
less steel)
Wild steel
*r»co Steel
Unit Me. 2
14,585
S.O
9.0
0.092
562C
0.3?d
0.61-0.80e
2.06
(4| eve-age-
0.062 (i*g)
3 in series
?2.9
63.1-90.Z
ASTM A-36
ASTM A-36
316 (sts'n-
less steel }
Mild steel
Lone Star
St*el
37,595
9.0
9.0
0.092
3*2C
J.OQ*1
O.JO-0.2f
4.2
(55 «*trage]
0.06£ («»g)
3 in series
82.1
24. 4-40. S
AST* A-36
-
3D4 (stain-
less steel]
AST* A-36
Prtcfpititor
"DusWtebllt"*
940
9.0
9.0
0.092
77-218
0.40-2.16
1.5-5.0
-
0.1644
1
NA*
19.2-6S.5
Kft
NA
NA
NA
United HcGill
"Mobile EP"C
632-1868
-
C.075
O.OJS^
202-966
0.10-1. JO
1.9-5.3
19-26
i
3 in series
NA
35.3-96.1
MA
MA
NA
NA
*P11ot test ynit— dita from Lone Star Steel Company; tests performed at E.B. Sermany Koris, Batteries A
ind B (comtion stick), Ion* Star, Tews.
Not directly comparable, thicktnst of discnarfi electrode plate.
•At maximum design flew rates.
Calculated from design specifications by ysinj Deutsch's Equation,
eCalcylated by using Deutsch's Equation from acceptance test report: Armco Steel Corporation. "Sam-
pling and Analysis of Particulste Emissions from Inlet and Outlet of Coke Plant last and West Precipi-
Utors, "Houston, Te»as, Novembtr 16-16 end 22, 1S76. Based on total particwlate catch.
Calculated by uiing Otuteti'i tfluilion from aeceptanct test report: lone Star Steel Corporation,
"Source Emissions Survey of Lone Star Steel Plant, Coke 0»en Electrostatic Precipitation. Lone Star,
Texas, February 1983," Ecology Audits, Inc. Based on total participate catch.
^Design oasis.
"Ratings at w»1mum inlet grain loading.
Mot directly comparable, secondirly current readings varied from 4-t M/f1el6.
•^Replaced with 316 (stainless steel) following unit startup.
hot applicable.
C-34
-------
Wet ESP's have an inherent disadvantage of cooling the exhaust gas and
causing a loss in draft. This shortcoming necessitates the addition of an
induced-draft fan and complicates battery drafting, which is ordinarily
accomplished by natural draft.
Two suppliers (Mikropul and Fluid Ionics) reportedly offer wet ESP
designs for application in coke battery stacks. The wet ESP offers the
advantages of higher collection efficiency due to lower particle reentainment,
smaller space requirements, and fewer sparks and fires. Corrosion potential,
drafting control, the need for water treatment, and the attendant costs are
the offsetting factors.
ScarfingEmission
In the scarfing process, high-purity oxygen is blasted against the
surface of the rolled or cast steel as it exits the rolling mill or continuous
caster while it is still hot (1600° to 2000°F). This process removes surface
defects. It generates relatively low emissions of extremely fine particulate,
consisting almost exclusively of iron oxide. The process is intermittent in
that individual slabs or billets are scarfed on a piece basis. Generally, a
piece passes through the scarfing machine every 2 to 4 minutes, which generates
in a burst of emissions lasting 10 to 20 seconds. The scarfing application
is generally less demanding than the BOF or sinter plant. Emission rates are
lower; the process variability is minimal; the size of the equipment is much
smaller, and the emissions are relatively consistent in size, composition,
and temperature. The complicating factors are the fine particle size and
intermittent nature of the emissions.
The exhaust from the scarfing machine enters a smoke tunnel (duct),
where water sprays cool the gas and provide some cleaning action. Prior to
environmental controls, the gas was exhausted to the atmosphere after this
step. As in the case of sintering, the opacity of the exhaust has generally
been more difficult to control than the mass emission rate.
21
Flow rates for scarfing range from 25,000 to 125,000 scfm. Inlet
grain loading varies from 0.25 to 1.25 gr/acf. The resistivity of scarfing
fumes is shown in Figure C-12. Wet ESP's are used almost exclusively for
scarfing cor.trol because of the saturated gas conditions.
APPENDIX C-ESP APPLICATIONS IN IRON AND ST1EL INDUSTRY
C-35
-------
99.9
99.8
99.0
98.0
95.0
90.0
80.0
70.0
I I
I I
J I
o o o
oo
-------
Wet ESP's can be categorized structurally as either plate-type or pipe-
type, the latter being used principally on scarfing operations. The pipe-
type usually has a weir-overflow arrangement or a continuous spray with an
upward vertical gas flow. Plate units have either a combined presatura-
tion/continuous spray or a spray with intermittent wash and horizontal or
vertical gas flow.
Certain characteristics of the wet ESP make it an attractive alternative
to wet scrubbers, dry ESP's, or fabric filters. The moisture-conditioning
effect eliminates the problem of high resistivity and minimizes reentrain-
ment. No rapping mechanism is required and pressure drop is low. Its disad-
vantages are that it has a high potential for corrosion and scaling and it
requires a water treatment system.
Although the use of wet ESP's appears to be a viable technology for par-
ticulate control of sintering and scarfing operations, design data sre gen-
erally considered confidential because of the intense competition between
manufacturers.
Wherever components of a wet ESP are similar to those of a dry precipi-
tator, the operation and maintenance procedures discussed previously can be
considered generally applicable.
Plate-Type (Horizontal Flow) ESP's--
The effluent gas stream is usually preconditioned to reduce temperature
and achieve saturation. As it enters the inlet nozzle, the ges velocity
decreases as a result of the diverging cross section. At this point, addi-
tional water sprays may be used to effect good mixing of water, dust, and gas
and to ensure complete saturation of the gas before it enters the electro-
static field. In addition, baffles are often used to distribute the velocity
evenly across the inlet of the ESP.
Within the charging section, water is sprayed near the top of the plates
in the form of finely divided drops that become electrically charged and are
attracted to the plate. This gives the plates an even coating of water.
Solid particles, which are simultaneously charged, "migrate" to the plates,
where they become attached. Since the water film is moving downward by grav-
ity on both the collecting and discharge electrodes, the particles are cap-
tured in the water film, which drains from the bottom of the ESP as a slurry.
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
C-37
-------
Concentric-Plate Type ESP's--
When applied to scarfing operations, this ESP uses an integral tangential
prescribing inlet chamber followed by a vertical wetted-wall, concentric-ring
ESP chamber.
The concentric cylindrical collection electrodes are wetted by fluids
dispensed at the top surface of the collection electrode system. The discharge
electrodes consist of an expanded metal system with uniformly distributed
corona points formed on a mesh background. This type of discharge electrode
system is supposed to provide a combination of a high, nearly uniform electric
field associated with a parallel plate system and a nearly uniform corona
current density distribution associated with the closely spaced corona points
on the electrode system. Higher gas flows can be handled by the addition of
more concentric electrode systems and by increasing the length of each electrode.
Conventional Pipe-Type ESP--
The pipe-type configuration is preferred over the plate-type for scarfing
applications because it appears to distribute water better during the washing
cycle. Weirs are also used in some cases to ensure that the tubes are kept
clean.
The actual system consists of a number of vertical collecting pipes. In
the center of each is a discharge electrode (wire type), which is attached to
the upper framework and held taut by a cast iron weight at the bottom. The
lower steadying frame keeps the weights and wires in position.-
The upper frame is suspended from high-voltage insulators housed in
compartments on top of the ESP shell. Heating and ventilating systems help
to prevent moisture and dust from accumulating in the insulator compartments.
The washing system usually consists of internal nozzles located at. the
top of the pipes. At specified intervals (after approximately 20 cycles),
this system thoroughly washes the tubes. While the washing is taking place,
the louver damper to the exhaust fan is closed to prevent droplet carry-over.
APPENDIX C-ESP APPLICATIONS IK IRON AND STEEL INDUSTRY
C-38
-------
Pre- startup P rocedyres
Before initiating the startup of ESP's, all of the major items of equip-
ment, connecting pipes, lines, and auxiliaries must be inspected, cleaned,
and tested, as follows:
0 Check for adequate flow, leaks, and pressures in all lines.
0 Check orientation of nozzles,
0 Inspect drain system.
0 Check all clearances between piping and high-voltage system.
ESP Internals--
0 Inspect all high-voltage system connections to transformer-rectifier
(T-R) sets, bus ducts, and grid.
0 Check insulators for cleanliness, cracks, and chips.
0 Check to see that high-voltage bus ducts to insulator connection are
tight.
0 Check heaters and pressurizing fan to be sure they are in proper
operating condition.
0 Check seals on access doors and plates and make sure that the interlock
contact is good.
T-R Sets -
0 Check oil level.
0 Tighten electrical connections.
0 Check to see if ground switch is operating properly
Control Panels —
0 Check all fuses and make sure indicator lamps for all high-voltage
items, heaters, and blowers are functioning.
Startup Procedures
0 Turn heaters on.
0 Energize pressure blower.
0 Turn high-voltage circuit breaker on.
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
C-39
-------
c Activate spray system, allowing sufficient time to clear air from
lines.
0 Open damper to introduce effluent gases.
Shutdown Procedures
0 Close inlet damper.
0 Deenergize ESP.
0 Shut down spray system.
0 Allow heater and blower system to remain energized to prevent accumu-
lation of moisture on the high-voltage insulators.
Normal Operation
Heaters and blowers are energized during normal operation. The spray
system is always activated just prior to energizing the high-voltage system.
The gas flow is monitored by damper, and water pH must be monitored at the
waste discharge,
Inspection and Haintenance During Normal Operation
Actual inspection and maintenance practices are very system-specific,
and the manufacturer's instructions should be followed closely. Because
ESP's operate with very high voltage, the system's internals must be properly
grounded.
General inspection and maintenance practices ere briefly outlined below.
Mechanical Maintenance--
All internal components should be checked for alignment, excessive dust
buildup, tight bolts, structurally sound welds, and general structural inte-
grity of the cross bracing and other support members. Because the support
insulators perform such a vital function in electrostatic precipitation, the
structural support end of the high-voltage insulator in the high-voltage
housing must be thoroughly inspected for cracks, chips, etc.
Water System-
All pumps, internal spray nozzles, and related valving and piping should
be checked. Because nozzles are subject to plugging, they should be routinely
disassembled, cleaned, and/or replaced as necessary. The main water-pressure
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
C-40
-------
supply pumps and all pipe joints must be checked for leaks, and all couplings
must be checked for tightness. Nozzle orientation also should be checked to
ensure that the intended spray pattern is being achieved. Regular attention
also should be given to the recirculated water system to maintain the strict
water quality requirements necessary for successful wet ESP operation.
Electrical System-
General areas to check include the high-voltage control panel, the
heater and blower control panel, high-voltage insulators, heater system
thermostats, T-R sets, and all related electrical connections.
APPENDIX C-ESP APPLICATIONS IN IRON AND STEEL INDUSTRY
C-41
-------
REFERENCES FOR APPENDIX C
1. U.S. Environmental Protection Agency. Pollution Effects of Abnormal
Operations in Iron and Steel Making, Volume VI. Basic Oxygen Process,
Manual of Practice. EPA-600-2-78-118f, June 1978.
2. Szabo, M. F., and R. W. Gerstle, Operation and Maintenance of Participate
Control Devices on Selected Steel and Ferroalloy Processes, EPA-600/2-78-
037, March 1978.
3. Kaiser Steel Dedicates New Steelmaking Shop. Iron and Steel Engineer,
May 1979. p. 29
4. Rowe, A. D. et al. Waste Gas Cleaning Systems for Large Capacity Basic
Oxygen Furnaces. Iron and Steel Engineer, 1970, p. 74
5. Henschen, H. S. Wet vs. Dry Gas Cleaning in the Steel Industry. JAPCA
18(5) p 338, 1968.
6. U.S. Environmental Protection Agency Revised Standards for Basic Oxygen
Process Furnaces - Background Information for Proposed Standards.
(Drafts EIS) Office of Air Quality Planning and Standards, Research
Triangle Park, North Carolina. November 1980.
7. Wheeler, D. H. Fume Control in L-D Plants. JAPCA, 18(2}:98-107, February
1968.
8. Goldman, L. J., et al. Performance of BOF Emission Control Systems In:
Proceedings of Symposium on Iron and Steel Pollution Abatement Technology
for 1981. EPA-600/9-82-021, December 1981.
9. Iron and Steel Society of the AIME (American Institute of Mining,
Metallurgical and Petroleum Engineers) BOF Steelmaking. Warrendale,
Pennsylvania.
10. Southern Research Institute. A Manual of Electrostatic Precipitator
Technology, Part II. Application Areas, August 1970 pb-196-381.
11. Rennhack, R. K. A New Precipitator for Partially Burned BOF and Q-BOP
Waste Gas. Iron and Steel Engineer, September 1978, p.44.
12. Jahlin, R. and Coy, D. W. Engineering Study of Roof-Mounted Electric
Precipitator (REP) for Fugitive Emission Control on Two Basic Oxygen
Furnaces of 300-Ton Capacity In: Proceedings of Symposium on Iron and
Steel Pollution Abatement Technology for 1982 EPA-600/9-82-021, December
1982.
APPENDIX C-1SP APPLICATIONS IN IRON AND STEIL WDUStRY
C-42
-------
REFERENCES (continued)
13. Katz, J. Iron and Steel Applications. In: Proceedings of Operation and
Maintenance of Electrostatic Precipitators Air Pollution Control Associa-
tion, Pittsburgh, Pennsylvania. April 1978.
14, Bandi, W, R. and A. J. Pignoscco. Control of Combustion in Sinter Plant
Precipitators. Ironmaking Proceedings, Vol. 38 A.I.H.E., Warrendale,
Pennsylvania, 1979.
15. Arthur D. Little, Inc. Steel and the Environment, A Cost Impact Analysis,
May 1975.
16. U.S. Environmental Protection Agency Pollution Effects of Abnormal
Operations in Iron and Steel Making, Volume II. Sintering Manual of
Practice. EPA-60Q-2-78-lI8b, June 1978.
17. U.S. Environmental Protection Agency. Coke Oven Battery Stacks, Back-
ground Information for Proposed Standards, (Draft EIS) Office of Air
Quality Planning and Standards, Research Triangle Park, North Carolina.
May 1980.
18. Trenholn, A. R. Study of Coke Oven Battery Stack Emission Control Techno-
logy Vol I. Prepared by Midwest Research Institute, Kansas City, Missouri
for U.S. Environmental Protection Agency under EPA Contract-68-02-2609,
Task 5. (undated).
19. Hill, R. L. Electrostatic precipitation of Scarfer Fume,
Iron and Steel Engineer. August 1979 p. 38.
20. Midwest Research Institute Study of Coke Oven Battery Stack Emission
Control Technology. Final Report - Volume II. Control Methods. EPA
Contract No. 68-02-2606, Task No. 5. Prepared for Industrial Studies
Branch. EPA - ESED. Research Triangle Park, North Carolina. March
1979.
APPENDIX C-ESP APPLICATIONS IN WON AND STEEL INDUSTRY
C-43
-------
-------
APPENDIX D
ESP APPLICATIONS IN MUNICIPAL INCINERATORS
INTRODUCTION
The primary purpose of most municipal incinerators is to reduce the
volume of solid waste for easier disposal. By adding the appropriate equip-
ment, some facilities are able to use the heat released in the furnace to
produce steam for either heating or electrical generation and thus derive
secondary benefits from the process. Both ESP's and wet scrubbers are used
to control participate matter emissions from municipal solid-waste-reduction
incinerators. Although ESP's require a relatively large initial capital
expenditure compared with that for scrubbers, they have generally been the
preferred method for controlling particulate matter emissions because of the
high operating costs, water treatment, sludge disposal, and corrosion concerns
connected with the use of scrubbers.
The application of ESP's to this particulate emission source can be
quite difficult because of the typical wide variations in the as-fired heat
content, moisture content, and size of the refuse being burned as fuel that
must be faced daily. These variations make it difficult to maintain adequate
control of the combustion process, and good combustion is essential to the
proper operation of an ESP.
CHARACTERIZATION OF MUNICIPAL INCINERATOR OPERATION
Most incineration systems distribute the "fuel" onto a series of grates
to volatilize and combust the material. Incinerator sizes vary widely,
ranging in feed capacities from 50 to over 600 tons/day of solid waste,
Examples of municipal incinerators with and without heat recovery are shown
in Figures D-l and D-2, respectively. Several different grate designs are
used, but inclined reciprocating grates seem to be preferred. Usually these
APPENDIX 0-ESP APPLICATIONS IN MUNICIPAL INCINERATORS Q-l
-------
INCINERATOR
STOKER
UNIT
Figure D-l. Diagram of a 120 ton/day municipal incinerator equipped with
heat recovery and an ESP.
D-2
-------
CHARGING
HOPPER
COMBUSTION
CHAMBER TO
SPRAY
COOLING
OVERFIRE AIR
UNDERFIRE
AIR ZONES
ASH SI FT INGS
SLUICE
AIR-COOLED SILICON
CARBIDE WALL
ACCESS DOOR
BIFURCATED
DISCHARGE
CHUTE
Figure D-2. Example of Municipal Incinerator with no heat recovery.
-------
grates are "zoned" into three to five areas to provide good burning character-
istics. Generally, the first stoker zone is not supplied with underfire
combustion air. Refractory arches usually are positioned to reflect radiant
heat from the combustion process toward the incoming material to dry it and
to begin driving off the other volatile, combustible components. Materials
are generally brought up to ignition temperature in this first zone, and the
burning is not very vigorous.
The highest heat release in the furnace is provided in the next zone or
zones. Underfire grate air is supplied to increase the rate of combustion on
the grate. Usually 40 to 50 percent of the total air supplied for combustion
is supplied to the undergrate air system, and the balance is supplied by
overflre air. Many of the better-controlled incinerators supply approximate-
ly 100 to 150 percent excess air for combustion, or approximately the stoi-
chiometn'c requirement of combustion air to the grate. Because most munici-
pal solid wastes are high in volatile material and low in fixed carbon,
however, much of the material is volatilized in the active burning zone, and
will burn as it leaves the fuel bed or as it enters the secondary combustion
zone, where adequate overfire air is provided. Proper mixing is essential in
this zone, and the furnace design and placement of the overfire air nozzles
are such that it is usually provided.
The last stoker zones provide for carbon burnout and cooling of the ash
for later disposal. Depending on the design of the system, a small amount of
undergrate air may be directed to these zones; however, some designs do not
supply any air to these areas. It is desirable to cool the materials to the
extent possible to minimize damage to the ash-handling system and to reduce
entrainment of dust as the ash is discharged.
TYPES OF ESP'S UTILIZED IN MUNICIPAL INCINERATION APPLICATIONS
The ESP's used in this application typically have been either weighted-
wire rigid frame or rigid electrode designs. Generally, they are not large;
9
SCA's of 200 to 300 ft /I000 acfm are required to achieve an emission
standard of 0.08 gr/dscf, 12 percent C0~. More stringent emission standards
2
will generally require larger SCA's {i.e., up to approximately 450 ft /1000
acfm). These ESP's also tend to only be two fields deep from inlet to
APPENDIX D-ESP APPLICATIONS IN MUNICIPAL INCINERATORS
D-4
-------
outlet, although more recent designs may call for three or four fields to
improve performance and reliability. In spite of the typically lower SCA,
these ESP's are usually designed with, a very modest superficial velocity of
between 4.0 and 4.5 ft/s with more modern designs in the 3,0 to 3.5 ft/s
range. Because these values are on the lower end of the range that may be
applied, the reentrainment potential of the ESP is reduced.
OPERATING PROBLEMS THAT AFFECT ESP PERFORMANCE
Wet Fuel
Because of the nature of the fuel, several operating problems are possi-
ble that can affect the ESP performance. First, the fuel is usually wet.
Depending on the site and time of year, moisture content can vary from 20 to
60 percent by weight. Items such as food wastes, grass and tree clippings,
and other wet wastes add to the moisture content of the material. The large
quantity of paper products and plastics help offset this problem somewhat.
The net effect of the high moisture content is to lower the heat content per
pound of material charged. At a fixed charging rate, this in turn lowers the
overall heat release rate and the combustion temperature of the furnace. In
some locations this situation can be particularly troublesome, e.g., where
wet garbage arid snow may be mixed together to be burned,
Noncombustible^ Materials
Another problem is the substantial quantity of noncombustible materials
that may be charged to the incinerator; these include cans, other metal
materials, large or bulky wooden objects, rocks, dirt, etc. Although paper
products and plastics tend to have low ash contents, the overall "ash" con-
tent is typically 20 to 30 percent by weight. This ash content also contrib-
utes to a lower heating value, and when combined with the moisture effects,
heat contents varying from 3000 to 5000 Btu/lb are not unusual. Although
typical heat contents range from 3000 to 4000 Btu/lb, in some instances the
time/temperature relationship is not sufficient to produce complete combus-
tion.
APPENDIX D-E5P APPLICATIONS IN MUNICIPAL INCINERATORS
D-6
-------
Particle Size Distribution
Still another difficulty encountered is the variation in size distribu-
tion of fuel that is fed to these incinerators. As with most grate fire sys-
temSi good combustion relies on even distribution of properly sized material
on the grate. The underfire air will follow the path of least resistance.
If the distribution of material on the grate is uneven, the gas will channel
through the path of least resistance and the remainder of the fuel bed will
be "starved" for air. The underfire air is also required to protect grate
components, and failure to supply adequate cooling air can result in prema-
ture grate burnout. Items that usually provide the most difficulty are wash-
ing machines, hot-water heater tanks, discarded beds, bicycles, large cans,
large auto parts, tires, and large wooden items. These items are usually
more troublesome for the smaller incinerator than for the large because they
affect a larger percentage of the grate area. Because these items also may
be difficult to discharge from the incinerator, some operations, such as
cogeneration from waste materials, are more selective in their fuel selection
and use wastes generated by commercial and small industrial facilities that
are dry, free of. bulky material, and consist mostly of paper products and
plastics.
Incomplete Combustion
If the fuel quality is poor, then combustion may also be poor and result
in the generation of partially combusted material because of inadequate
time/temperature/turbulence considerations in the combustion zone. Such
materials are often high in carbon content, and they tend to condense to forr,
very fine but effective light scattering particles. The high carbon content
tends to lower the resistivity, and the finer particles become reentrained;
these particles may not be captured because the relatively small ESP's are
unable to handle large quantities of fine particles. Under normal circum-
stances, furnace zone temperatures usually are held to approximately 1400° to
1650°F to provide adequate combustion temperature. The temperature set point
is generally maintained by overfire/underfire air ratio and feed rate.
Changes in fuel quality affect these requirements, and failure to maintain
proper temperature will usually have an adverse effect on ESP performance.
APPENDIX D-ESP APPLICATIONS IN MUNICIPAL INCINERATORS
D-6
-------
Excess Air
Another area that is important to ESP performance is the quantity of
excess air that is carried by the flue gas. Excess air levels of 100 to 150
percent are generally desirable in this application to overcome problems of
mixing and to help assure complete combustion. This does not mean that lower
excess air quantities cannot be achieved, as some sources are capable of op-
erating at 50 to 75 percent. For those sources using heat recovery, however,
care must be taken to control flame impingement and localized reducing zones,
which may cause premature tube failure.
Increasing the amount of air in excess of stoichiometric requirements
generally improves combustion and raises combustion temperatures. At some
point, however, no real increase in combustion efficiency (or conversion to
C02 and H^O) results, and increasing the quantity of combustion air actually
decreases the flame temperature (because both the combustion products and the
excess air must be heated). This can also increase the flue gas volume
leaving the incinerator. Even after gases are cooled in an evaporative spray
chamber or passed through a heat exchanger, the result is usually a higher
gas volume to the ESP, which reduces the actual SCA and treatment time and
increases the superficial velocity and the opportunity for more particle
reentraintnent out of the ESP. Of course, the point of introduction of the
excess air will determine how much the excess air will affect combustion
efficiency. Careful attention to combustion air settings and to points of
inleakage will help minimize problems.
Temperature Effects
The gas stream leaving the incinerator must be cooled before it enters
the ESP. Generally, the acceptable temperature range is between 400° and
700°F. Outside of this range, corrosion and other stress-related problems
become a concern. The presence of chlorinated plastics, rubber, and other
compounds may produce hydrochloric acid. At low temperatures, the acid may
condense and attack the metal; at higher temperatures, the corrosive protec-
tion afforded by natural inhibitors is substantially reduced. Very often,
however, the low temperature acid-dewpoint is of most concern because local-
ized inleakage of air or localized cooling can accelerate the corrosion rate,
APPENDIX D-ESP APPLICATIONS IN MUNICIPAL INCINERATORS
D-?
-------
Both of these factors can lead to very rapid destruction of the ESP and other
support equipment. Rebuilding of the ESP may be necessary after 3 to 5 years
if careful attention is not given to maintaining temperature and providing
preventive maintenance.
Evaporative cooling is typically used for incinerators not using heat
recovery, A spray baffle system is used to provide contact between the flue
gases and water without excessive water droplet carryover. The baffle ar-
rangement also enables the capture of the larger particulate carried out of
the incinerator. Use of a temperature sensor/controller is generally recom-
mended to control the amount of evaporative cooling water that is needed,
For those incinerators using heat recovery (typically steam generation),
special design considerations are necessary to ensure complete combustion,
low slagging and fouling of heat exchange surfaces, and less erosion of boil-
er tubes. For furnaces equipped with water wall tubes, it is not unusual to
have furnace volumes 2 to 3.5 times larger than those used for solid fuels
(wood, coal) of equivalent heat release. This increased furnace volume
provides longer residence time and an increase in heat absorption before the
flue gas enters the convection heat transfer area. Slagging and flame im-
pingement on boiler tubes can cause heat-transfer problems and accelerate
corrosion of boiler tubes. These problems can be further aggravated by
localized reducing zones in the furnace. Thus, good mixing of sufficient
oxygen is important to minimize these problems. Although a higher percentage
of excess air will decrease the efficiency of the boiler somewhat, it is
usually worthwhile when maintenance considerations are taken into account.
Another consideration in furnaces designed for heat recovery is erosion of
the refractory and tubes. Generally, the designs call for much lower veloci-
ties than would be found in a typical boiler. Special tube designs or ar-
rangements may also be used to minimize tube erosion which can lead to tube
failure.
Corrosion Prevention
Use of corrosion resistant materials, proper insulation, prevention of
air inleakage, purging acid-laden gas directly after shutdown of the ESP, and
use of an indirect heating system during shutdown can extend the life of the
ESP from 3 to 5 years to 10 years or better.
APPENDIX D-ESP APPLICATIONS IN MUNICIPAL INCINERATORS
U-o
-------
Corten or mayari-R have been used in the past instead of mild steel for
the ESP shell and collecting electrodes. Although Corten has about twice as
much resistance to moisture corrosion as mild steel, it is not significantly
more resistant to acid corrosion than mild steel.
Three inches of fiberglass or mineral wool insulation covered by metal
lagging should be provided for the casing and hoppers to minimize corrosion.
A clean purge air system should be employed to provide clean air to the ESP
internals while it is still hot during shutdowns. This will remove free
chloride and sulfide ions.
A careful check of the casing should be made during shutdowns to determine
if there are any corrosion holes. This may be difficult because of a buildup
of dust, which should be removed before inspecting the casing.
To keep the ESP temperature above dewpoint during shutdowns, a small
auxiliary heater fired with No. 2 fuel oil can be installed and the heated
air blown into the isolated ESP at a temperature of about 350°F. A diagram
of such a system is shown in Figure D-3. This type of heating system can
also be used to keep the spray tower refractory warm during shutdowns and
subsequently minimize thermal shock to the refractory during the succeeding
startup.
Another method of keeping the ESP above dewpoint is through the use of a
heat-jacket type of shell for the ESP. Warm air is circulated between the
double steel casing wall by a small blower. This is similar to the type of
shell used on recovery boiler ESP's in the paper industry, and adds about 15
percent to the capital cost of the ESP.
Re s i s_t 1 v ityJEf feet s
The combination of good combustion and high moisture content in the flue
gas, either from moisture in the solid waste or from evaporative cooling,
leads to a dust that is relatively easy to collect. Resistivity data for
municipal incinerator fly ash is very limited; however, resistivity does not
appear to be a substantial problem. Some materials introduced into the
incinerators (e.g., dirt and sand) cen increase the resistivity of the dust;
however, wood and paper products usually produce an ash that is high in
alkali content. In ESP applications for wood and bark combustion, resistivity
APPENDIX D-ESP APPLICATIONS IN MUNICIPAL INCINERATORS
0-9
-------
STACK
FURNACE
_T
Figure D-3. Location of an auxiliary heating system to keep ESP
temperature above the dewpoint.'
(Permission granted by APCA.)
D-10
-------
has been found to be somewhat lower than desirable, but acceptable (10 to
O
10 ohm/cm) at typical operating temperature. This condition is expected to
apply to firing municipal solid waste as well. From a practical standpoint,
operating voltages and currents in these ESP's tend to reflect an acceptable
resistivity, as operating voltages and currents are moderately high. For
most ESP's in this application, however, low resistivity is probably of
greater concern than high resistivity.
Problems with low resistivity appear to worsen when combustion efficien-
cy diminishes or when excess air levels are allowed to become too high. As
previously discussed, poor combustion may generate finer particles with
higher carbon content. This combination tends to decrease the overall per-
formance of the ESP,because it is not generally sized to handle the higher
loading of fine particles. In addition, the higher carbon content tends to
decrease resistivity and increase reentrainment. Increasing the quantity of
flue gas by increasing the excess air may further aggravate the reentrainment
problem because the velocity through the ESP is higher. In addition, placing
high levels of carbon in ESP hoppers increases the possibility of hopper
fires. Maintaining and monitoring good combustion in the furnace is essen-
tial to proper ESP operation.
Hopper Pluggage
As in other applications, the hoppers should never be used for long-term
storage because the dust flows easily when warm, but not so well when cooled.
Continuous removal of the dust not only minimizes the risk of hopper fires,
it also tends to reduce the possibility of hopper pluggage that can create a
number of other problems (see Section 4). The dust tends to be hygroscopic
and, when cooled, can be extremely difficult to remove from the hopper, as it
forms clinker-like nodules. Hopper insulation is essential to reduce both
hopper pluggage and corrosion problems. In some circumstances, hopper heat-
ers, windbreaks, or enclosures may be necessary to prevent excessive cooling
of the dusts.
In conclusion, the application of ESP's to collect the fly ash from
municipal incineration does not, in principle, represent a particularly
difficult application. Although smaller ESP's have typically been used to
handle the lower-resistivity dusts that are expected from this application,
APPENDIX D-ESP APPLICATIONS IN MUNICIPAL INCINERATORS n .,
-------
newer designs are larger and sectionalized. Some corrosion problems still
must be addressed to improve continuous compliance aspects of these ESP's;
however, in those installations where adequate monitoring is provided and a
conscientious effort is made to maintain good combustion, problems are gener-
ally minimal. When the combustion process is not well maintained, however,
ESP performance tends to deteriorate noticeably. A good inspection and
preventive maintenance program 1s also essential to minimizing normal operat-
ing problems and the potential for corrosion problems in this application.
APPENDIX D-ESP APPLICATIONS IN MUNICIPAL INCINERATORS n , ?
-------
REFERENCES FOR APPENDIX D
1. Franconeri, P. Electrostatic Precipitator Corrosion on Incinerator
Applications. Presented at 68th Annual APCA Meeting, Boston,
Massachusetts, Oune 1975.
APPENDIX D-ESP APPLICATIONS IN MUNICIPAL INCINERATORS D-13
-------
-------
APPENDIX E
DATA SHEETS AND EXAMPLE CHECKLISTS
APPENDIX E - DATA SHEETS AND EXAMPLE CHECKLISTS
-------
PRECIPITATOR CONSOLE DATA SHEET
UNIT: LOAD
DATE; OPACITY
INLET TEMP.
OUTLET TEMP.
Gas 1 Flow
Legend 6 Rat in j
T.R.
PR1.
PR1 .
SEC.
SEC,
No.
V
A
KV
mA
^00
128
4S
1000
UPPER BOX
1B2
2B2
3B2
4B2
5B2
6B2
7B2
1B1
2B1
3B1
4B1
5BI
6BL
7BI
1A1
2A1
3A1
4A1
SAl
6A1
7A1
1A2
2A2
3A2
4A2
5A2
]
i
6A2
7A2
LOWER 10X
1B2
2B2
3B2
4B2
5B2
6B2
7B2
1B1
2B1
3B1
411
5B1
6B1
7B1
1A1
2A1
3A1
4A1
SAl
6A1
7A1
IA2
2A2
3A2
4A2
SA2
6A2
7A2
E-2
-------
CORONA ELECTRODE FAILURE AND CROWD REPORT FORM
PLANT
- UNIT NO,
DATE:
/ /
DATA BY;
TR or
BUS
sta ION
NUMBER
SYMPTOM
DATE
FOUND
TYPE
CAUSE
OF
GROUND
DESCRIP-
TION OF
GROUND
GRID
LOCATION
KIRE I
LETTER !
LANE
NO.
FAILURE
LOCATION
(see *
below)
SYMPTOM
TYPES
A,
B.
C.
Dead Ground - zero voltage, some current, no sparking.
Intermittent Ground - large fluctuation of voltage and
current, sparking.
LOK Power Operation - Steady state low voltage sparking.
CAUSE OF GROUND
A,
B.
C.
Failed corona
electrode (wire)
touching ground
Close clearance
or contact be-
tween high
voltage system
and ground
Ash bridge between
high voltage
systen and
ground
DESCRIPTION
Bl.
a2.
a3.
84.
aS.
a6.
87.
bl.
b2.
b3.
b4.
bS.
b6.
b7.
Cl.
c2.
c3.
c4.
Penciled break, wire surface away from break
smooth (sparking erosion)
Flat cross-sectional wire surface away from
break smooth (mech. fatigue)
Reduced and/or pitted cross-section, wire
surface away from break rough (corrosion)
Shroud worn through support angle.
Wire pulled out of top shroud.
Wire pulled out of bottom shroud.
Wire from another bus section.
Shifted collecting plates.
Weight out of guide.
Slack wire-weight tilted in guide.
Dimensioned wire-weight oissing.
Kinked wire.
Bowed collecting plates.
Shifted high voltage system.
High hopper flyash level.
Clinker between wire and plate.
Clinker between shroud and plate.
Contaminated Insulators.
E-3
-------
EXAMPLE FORM FOR KEEPING WIRE FAILURE RECORDS.1
GAS FLOW
I
GRID FOR BUS SECTIONS OF FIELDS 2, 4 6 6
Lane Number
15 10 IS 20
A*****O*********O****
jj.*«»»****». *********
£
D
£ E
* * * O>v O* * * * * * * *xs* * * * *
* * « * O Q******* *^O * * * *
«*«**«*•*.**.**«***«*
A.*****************.
***********
**. ********
•*•*****
** *****
- SUPPORT/RAPPER ROD
- WEIGHT GUIDE FRAME PIPE SUPPORT
E-4
-------
AUXILIARY PRECIPITATOR RAPPERS
OBSERVER:
DDDDDDDDDD
ocn aQr-^iOLn^rocsjr-
tOCJ CM CM CM CVJ CM CM CM ^
24A
23A
DDDDDDDDDD
O cr« co'^.^om^rncMj""
DDDDDDDDDD
o ^j* <**> ^*JT]
22A
21A
DDDDDDDDDD
OCh>OOI*^i£>"">*>"fr'Mir!
r^CJCMCMCSlCMI^JCMCMi^
DDDDDDDDDD
ooor-»i0ir>*3-«v>cMi—
cyr_(_l_^i_r_r_r_^-
34A
33A
ODDDDDDDDD
OO"*oQ r^iD tr> ^f-fOcsj*^
DDDDDDDDDD
OO^oo*^ ^OiO ^i-rocsi1"
32A
31A
DDDDDDDDDD
ooi cof^'UDin "^rocsj*™"
DDDDDDDDDD
C3fQI — \DlA«*fr)CM'~
44A
43A
DDDDDDDDDD
o cr>oof--«frp'ioJ^~
DDDDDDDDDD
ocr* ooi--«3i/'»«*-Mea'~
42A
41A
DDDDDDDDDD
otn oo f"*" *o i" «9" ef> *^ r
R
P
H
M
X
3
"O
O
3>
—I
O
"O
n
?o
1. Loud sound and full plunger movement
2. Paint sound or short plunger movement
3, No sound or plunger movement
GAS FLOW
-------
10
30
CD
10
CD
30
CD
10
CD
30
f
10
I
30
CZ
10
CZ
30
d
9 8
29 28
rn CD
35
9 8
CD a
28 28
a CD
34
9 8
CD a
29 28
cz! C
33
9 8
a cz
29 28
CD d
- 32
9 8
a d
29 28
! CD d
31
48 47
B CD CZ
48 4?
oCH C
7
rn i
27
CID 1
27
CD 1
CZI
27
CD
7
CD
27
CD
7
1 CD
27
1 CD
46
] CD
46
) CD
6
ID
26
6
ID
26
6
a
26
I
6
1
«:6
CD
6
CD
26
CD
45
CD
45
CD
54321
CD CD CD CD CD
25 24 23 22 21
rn rn rn rn r^
45
54321
CD CD CD CD CD
25 24 23 22 21
CZI CZI CD CD CD
44
54321
CD CD CD CD CD
25 24 23 22 21
1 1 1 1 1 I I 1 1 1
43
54321
CD CZI CD CD CD
25 24 23 22 21
CZ] CD CD CD CD
42
54321
1 II II II II 1
25 24 23 22 21
CD CD CD CD CD
41
44 43 42 41
CD CD CD CD
44 43 42 41
anna
E
D
C
B
A
B
D
EXAMPLE FORM FOR PRECIPITATOR PLUNGER MOVEMENT INSPECTION
E-6
-------
4
8
12
16
20
3
7
11
15
19
H
25
24
23
22
21
2
6
10
14
18
1
5
9
13
17
0
4321
35
8765
34
12 11 10 9
33
16 15 14 13
32
20 19 18 17
31
1. Strong vibration
2. Weak vibration or short time period
3. No vibration
Make a note of any loose or missing bolts, nuts, etc.
Two vibrators in each group operate at the same tine (f.e.,
U2, 3&4)
EXAMPLE FORM FOR PREC1P1TATOR VIBRATOR INSPECTIONS
E-7
-------
EXAMPLE FORM FOR PRECIPITATOR FLY ASH HOPPER INSPECTION
DATE;
TIME:
SHIFT:
FLY ASH HOPPER CHECK
OPERATOR: SHIFT FOREMAN: SHIFT ENGR:
Hopper
no.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Vacuum range
High
Low
Visual comments
Hopper
no.
20
21
22
23
24-
25
26
27
28
29
30
31
32
33
34
35
Vacuum range
High
Low
Visual comments
E-8
-------
Month/Year:
EXCESS OPACITY REPORT'
Station/Unit(s)
Opacity Limit: 30% (six-minute)
Opacity (%)
Day
Begin*
Cause
Corrective action taken
\
Initials
Start time of the six-minute exceedance (24-hour clock).
-------
ELECTROSTATIC PRECIPITATOH
SHIFT S DAILt OPERATION RECORD5
PIAKT
UNIT
ESP
BATE:
GAS TEMP.: A
«EL. HUMIDITY:
TlHE:
OPACITY: A_
10,,
/e
REVIEWED BY/DATEr
CROSS LOAD:
AMI. TEMP.!
T/R OOKTROL
sn HO,
FRIMAHY
VOiTS
PRIMARY
AKFS
SHIFt CHECBLTST
TritiifonB«r-Reetif ler
Central M«t«r Rttdtnga
PreelpllBtor Ash Hopper
L*v*l«
IUpp*r t Vibrator Controller
Operation
DAII.f CHECKLIST
HV BUB Duet NotBe
LociHteJ Spirkinft
Preclpitntor or Hoi IFF Upneti
SECONDARY
AMPS 1
CHECKED BY
(tHITTALS)
SECONDARY
AMPS 2
TIKE
SECONDARY
KVOLTS 1
SECONDARY
tVOLTS 2
CONBITtOfi (CHECK)
ACCEPTABLE
UNACCEPTABLE
SPARKS PKR
MINUTE
REMARKS
MATURE OF DEFICIENCY t
CORRECTIVE ACTION TAKEN
E-10
-------
ELECTROSTATIC flECTPITATOR
MEEKLY EXTERNAL IKSrECTIrtH RECORI
, 3
PLANT
UNIT
REV 1 WED BT/DATE:
RSP
INSPECTION fTEM
Check HV Transformer fill Level
i Temp,
Inspect T/R Control & Purge
Air Filter*
Check Accent Door Air tnleakage
Check Purge Air I Hester
SynreB Operation
Rapper Kyatea Settings Check
Vibrator Syulea Setting* Check
arUCKED Bt
(IMIT1ALS)
PAT«
COTOIT1f»H (CHECK)
ACCEPTAil.E
UNACCEPTABLE
NATURE OF IIEFlerENCJ t
cotRtennN ACTION TAKEN
Weekly Averofte M Fired Fuel An«1y»l»: X Moisture
S Aeh
X Sulfur
iTU/LB
Rapper/Vibritor Setting Record
Ripper ScttlngB-Ptevlout
Intensity
Freiuencjr
Rapper Settlngs-Hev
IntcnaltT
Frequertff
Vibrator Sect Ingn-Prevlouii
Intensity
Frequentf
VJhr«tor Setline«-Nev
tnteftff 1 1 y
rrmarncj
it
(IBITIALS)
DATE
HELD 1
riELD 2
FIELD 3
TIET-D 4
REMAUKS
E-ll
-------
ELECTROSTATIC PRECIPITATO*
QUARTERLT KTERKAl, tHSPECTTOH RECORD
PUUTT OHTT
ESP
REVIEWED
INSPECTION ITEM
titan Rapper. Vibrator, T/R
Control*
Rapper Svlteh Contact fntpcction
vibrator Switch Contact
Imped Jon
ChrcV Rapper Aimrmblf Elndinf;/
Hiaallgniaent
Ripper/Vibrator Boot S*al(
Inspection
Rapper Hunger Condition
Inspection
Check for Defective Xapprra
Check for Defective Vlbmtora
Check Vibrator Air Cap Setting
Check Inatruvent Calibration
IT
(INITIALS
PATt
CON11TTTOH (CHECK)
ACCEPTABLE
UNACCEPTAW.F
NATURE OF DEFICIiDCT 4
CORRECTIVE ACTION TAKEN
E-12
-------
ELECTROSTATIC FRRCimATOR
ANNUAL IlfTERNA!, INSPECTION RECORD
FLAHT
UNIT
ESP
iHsrucTioH ITRW
A. Transformer Enclosure:
Clean and Check Insulators,
Bunhlngn
Check Electrical Connect ion §
Check and Set Surge Arreotore
B. ItV Bue Ihict:
Insptct for Hun or Selling
CItan «nd Check Pott Innultlort
Clwck for IJDOM Connect Ion a
IT
(INITIALS)
DAT!
CffflDlTIOH
ACCEPTABLE
-
UNACCEPTABLf
NATURE OF TiEFlCtENC? t
COltECTIVE ACTIOJf TAKEN
E-13
-------
TWSPECTTDH ITIX
Repair l.ooie tat Clbova
C. Penthouie, BapperB,
VlhraLora:
Check Centering of Upper
Rapper Rod
Clean and Check Rapper
Insulators
Inspect for Ash Accumulations
Around Rode
Check Centering of Lower
Ripper Roid
Clieck Insulator Heater
Check for HalerMlr leakage
In Penthoupe
Inspect Roof Penetrationi for
Uiter Leakage
Clircli All 1IV ConnerrlonR
Clean nnd Clieck Support
Tnnul*tors
Clvfck Calltrt on Vibrator
Insulators
D. Collectlnj Surface
Anvil Re««:
n
UNIIlAtS
DATE
common
ACCEPTABLE
UHACCErTABiF
NATURE Of nr.FfCTF.HCV t,
CORRECTIVE ACTION TAKEN
E-14
-------
INSPECTION ITW
tnnpect Hanger Rods and Clip*
Reanve Packed A»h Behind
Anvil Bea>
Innpect Welds of Rods to
Anvil Se*B
E. Upper HT Frape *»t«rtlf :
Tnuppct Ueld* «t Hunger Pipe
re Frnae
Check. 1(7 Franc Support Bolt*
Inspect Welds «r Support
Angles to Be a™
Check Level and Square ttl Fr*ae
F. Lower IIT Frime AtilBblf
Check Weight-Guide Htnga
& Align
Check Level irtd Square
of Fraiw
Check Lower FriiK Twitting
r. . Siabillsatlen Tniulitorn:
BY
(INITIALS
DATE
CnHWTTOH
ACCEPTABLE
UHACCECTABLt
HATUIE OF DEFICIENCY *
CnRRECTIVE ACTION TAKEN
E-15
-------
INSPECTION ITEM
Clean and Check Iniulat.org)
H. Collecting Electrodes 1
Check Aah Buildups
Check Electrode Alignment
t Spacing
Check Plumb ind Square
of Conponentn
Tnnpect 'or Roving or Bellying
T. Di«chirge Electrode
AflteaMy :
Check Alh Buildup!
Check for Broken Electrodes
Check Alignment and Spicing
Check Weight* for Alignment
& Freed OB
J. Hopper Inspection
Check Isr A*h Buildup in
Upper Corners
BY
(INITIALS
RATE
OTHDITfOH
ACCEPTABLE
UNACCEPTABLE
NATURE or nrricrRHcv t
CORRECTTVE ACTION TAKEN
E-15
-------
INSPECTION tTEH
Check Tor Dchrln In notion
And Valve
Check Hopper Level Detectors
Check Hopper Vibrators
K. C»ner»I:
Inspect for Interior Corrosion
Check Safety-Key Interlocks
Inspect Electric*! Grounding
Sy»le»
Check Thermal Expansion ((
Required
Check .for Ash Buildup on V«ne»,
Ducts
Cherk TranRformpr Oil Dleletrlc
If
(INITIALS]
BATE
CONDITION
ACCEPTABLE
UNACCEFTA1LE
fMTURf: or nprictEHCT i
romircTivR ACTION TAKEN
REVIEWED BY;
Electrical Pore"«n:
H*ch«nle«l foreign:
Hatncentnce Supervliori
Operating Supervlior:
Malnlxtritlvc Siifter»!«ori
M«n*|«r:
r«
lion
Dtte
P«t«
D«le
Date
Oetf
E-17
-------
ESP BASELINE TEST INFORMATION
A. Process Conditions
Gas flow rate
Gas temperature - points A,B,C, etc.
Static pressure
Process level (load, I capacity)
Process feed rate(s)
Process feed descriptor(s)
Process product level(sj
Process product descn'ptor(s)
Conditioning agents (type and ppm)
B. ESP Operating Conditions
Gas temperature, ESP inlet(s)
Sas temperature, ESP outlet(s)
Primary voltage for each field
Primary current for each field
Secondary voltage for each field
Secondary current for each field
Spark rate for each field
Rapper frequency, plate
Rapper frequency, wire
Rapper duration, plate
Rapper duration, wire
Rapper intensity
C. Stack Test Results
Testing crew, group leader, date
Outlet participate concentration, average
Outlet participate concentration, individual runs
Inlet particulate concentration, average
Inlet particulate concentration, individual runs
Sampling times per run
Sampled gas volumes per run
Gas composition
Moisture content
Emission level, Ib/h
Emission level, Ib/process input parameter
Emission level, Ib/process output parameter
Effluent opacity (including presence or absence of puffing)
D. Applicable Regulations
Particulate
SO-
Opacity
E-18
-------
ELECTROSTATIC PRECIPITATOR BASEL IMF COMPARISON
Possible operating problems
I. ELECTRICAL
A. Particle resistivity
1. Peak voltage low (down
5-10 kV)
2. Rapping intensity
(increased)
3. temp changed (±50°F)
4. Spark rate increased
(±50 sparks/min)
5. Opacity high
6. Coal sulfur content low
B. Transformer-rectifier set problems
1. No secondary current
2. No penthouse purge
3. Voltage zero, current high
4. Opacity high
C. Insulator failure
1. Peak voltage low
2. Penthouse purge (not used)
3. Penthouse temp high (±20°F)
4. Opacity high
5. Cracks visible
Average baseline
Observed
(specify value) (specify value) Location'
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
E
E
E
E
E
E
N/7T
E
E
E
E
E
E
E
E
I
Abnormal
(check)
E is external, I is internal.
-------
ELECTROSTATIC PRECIPITATOR BASELINE COMPARISON (continued)
Average baseline Observed Abnormal
Possible operating problems (specify value) (specify value) Location3 (check)
I. ELECTRICAL (continued)
D. Broken discharge wires
1. Deposits on wires N/A N/A I
2. Violent meter fluctuating N/A N/A E
3. Hopper level indicator not N/A N/A E
used
4. Spark rate high (A50 sparks/ E
min)
5. Opacity high E
6. Broken discharge wires N/A N/A I
II. GAS FLOW
A. Excessive velocity
1. Flow rate high E
2. Voltages high, currents low ZHZHHIZZZZ E
3. Opacity high E
B. Nonuniform distribution
1. Flow rate increased E
2. Secondary currents nonparallel N/A N/A E
3. Hopper level differences I
on parallel branches
4. Rappers on distribution E or I
plates not used
m
rv> E is external, I is internal.
-------
ELECTROSTATIC PRECIPITATOR BASELINE COMPARISON (continued)
Average baseline Observed Abnormal
Possible operating problems (specify value) (specify value) Location3 (check)
III. MECHANICAL
A. Rapper problems
1. Puffs visible N/A N/A E
2. Peak voltage changes, second- E
ary current constant
3. Spark rate changed E
4. Low sulfur coal used E
5. Dust sticky N/A N/A E
B. Hopper solids removal
1. Broken discharge wires N/A N/A I
2. Mass loading probably N/A N/A E
increased
3. Nonuniform gas distribution N/A N/A E
4. Hoppers not emptied N/A N/A E
continuously
5. Level indicators not used N/A N/A E
6. Heaters not used N/A N/A E
7. Vibrators not used N/A N/A E
8. Hoppers not insulated N/A N/A E
9. Corrosion around outlet valves N/A N/A I
10. Hopper slope <60° N/A N/A E
11. Hoppers full or bridged N/A N/A I
C. Collection plate warpage and
malalignment
1. Change in air load N/A N/A E
2. Repeated hopper overflow N/A N/A E or I
3. Air inleakage N/A N/A E
7 4. Malalignment visible N/A N/A I
to
t—'
flE is external, I is internal.
-------
ELECTROSTATIC PRECIPITATOR BASELINE COMPARISON (continued)
Possible operating problems
IV. EFFLUENT CHARACTERISTICS
A. Mass loading increases
1. Opacity high
2. Inlet section, secondary
currents, low
3. Hopper unloading frequency
increased
Average baseline Observed Abnormal
(specify value) (specify value) Location3 (check)
E
E
E is external, I is internal.
-------
AIR LOAD TEST RECORD
PLANT
1JNTT
TIME: REVIEWED BY:
DATE: TESTED BY:
ESP:
ESP TEMP.:
CONTROL
MANUAL
AUTOMATIC
T/R SET NO.
FIELD ENERGIZED:
AMBIENT TEMP.: WEATHER COND.:
TRANSFORMER
PRIMARY
AMPS
VOLTS
PRECIPITATOR
MILLIAMPS
1
2
S/M
KV1
KV2
REMARKS
m
S/M: Sparks/Minute
KV1: Kllovolta/BuBhlng 1
-------
EXAMPLE BID EVALUATION FORMS FOR
- ELECTROSTATIC PRECIPITATORS
E-24
-------
BID EVALUATION FORM FOR ELECTROSTATIC PRECIPITATORS
.4
Bidder « 1 Bidder *2 Bidder g J Bidder
VENDOR
OPERATING AND PERFORMANCE DATA
Volume — CFM (S operatinf eondiuonj
Temperature — "Ft? operating rendition
lltiet loading — grams.'Cj (g optrmng
DUJI Bulk Dentil)'
Ou»r»nl*ed Efiicienc)- — Percen!
Oytlrt Loading — gr . in cy ft
Drop Across Prrripitaio;
including |AS dutnbyt^on de^icp^
Gas Velocity — ft.ittc
Trtatrnent
PBEClPtTATOR ARRANGEMENT
Number of Precipitaiors
Chambers (number) ^Frecipjtator
Fields (number and length) 'Precipilator
Cells !nyn>t«r) /Prtcipitsior
Bui tettioni (number! ,'Precipnaloi
C»i.r,j Material and Thickness n^chts
Cajing Design Pre*sur« (' ' Vi C )
(Check one. Positive O °r Nesaiive ~i
Number cl Hoppers -Precipitator
Hopper Material and Thitknes? (inches)
MiHrmym Hopper Valley Angle
Total Hopper Capacity (eybic fe*U/Precipi!*tor
Hopper Accessories (list rich itparateK )
Iniulltor Comptrtment Materiel «nd
Thickness (inches I
Penthoute Material and Thickness (inches)
Number of Iiuulnor Compinments Precipitator
Surface Area (iq ftJ/Pricipitator
Shell
Hoppers
Other
E-25
-------
BID EVALUATION FORM (continued)
Bidder el Bidder *2 Bidder *3 " Biddf
Internal Gas Dittnbution Device*
Quantity and Location rrenpjlilor
Material end Thieknei?
Number and Type Ripper?
Number, Type and Sue of Access Doori Prrr;p:t*
-------
BID EVALUATION FORM (continued) .
Bidder*! Bidder #2 Biddtr *3 Bidder * 4
Tr»njform*r R*cuHer Insulation fluid
Wivt Ferm of Huh Volttft
Number Hid Type Hlfh Voltage Switches
Key Interlock* fitf — *C
M»*imy»n Arnbiint Tempenlure for Tr»ruform«r
Rectifier Control Cibinels — "C
Power ConjympSion KVA 'Prfcipitaior
i Tr«niformer-rettiri*r
2 Happen
3 Insulator heiten >nd blowers
4 Hopper h*iters
3 Lights
t Other
Total
Ton! connected loid KVA Frecipilator
Povcr Diilributian
Individual breaker: each control cabinet
Central distribution panel
OTHEE ALXH1AHY CQUfFMENT OR SERVICES
Heat Lru uUtiors — Type it Thickjitu
Weather Enclosure — I Root and 'or Hopper, typo
Inlet and outlet TrtruiUoni — Material and Thickness
Aceeu FaciiitiM — Type & Location
Insulator Compartmint Blower System (number)
Mode! Study
Startup— Training — Te»l>n| SupervisiOh
Erection Supervuion
WEIGHTS
Total Prtcipnaicr Wtight Including Electrical
Equipment but Excluding Dusi Load
E-27
-------
BIDDING DATA LIST
FOR
ELECTROSTATIC PRECIPITATORS
DATE PURCHASER'S INQUIRY NUMBER.
Section A
General Information
1) Purchaser Address
i) User (if different from above)
Afldress ^___
2) Sue Location _____ .—_—_—_—
3) Individual, title and iddre&s to whom proposal u to be lent
i) Number of copies of proposal to
4) Date proposal is to be submitted —
5) Purpose of proposal budgetary or firm1!
6) Is formaJ proposal required, or priced letter"1-
7) Equipment delivery requirement date
8) Bid Basis: FOB shipping point; FOB cars destination, FOB shipping point FA to site, othe:
Section B.
Plant Process
1) Process to which precipitator will be applied
Furnace, boier, kiln, other,
Dctip data: Type
Make
Output: con tifiuoui riling.
peak rating
2) Description and analyses of raw materiaJ or fuel
3) Expected viriaiioni in raw material input to furnace, etc.
E-28
-------
4) Description ind riling of cxiitini collector equipment, if tny
Section C.
Operating Condition
1) Cat volume at precipitator inlet as measured by Pilot tube:
(»" Continuous rating, actual cfm <§> °F., psia.
P Peak rating, actual cfm @ °F., psia.
Moisture content in gas % by volume or % by weight
Volume fot which pieripilatoi efficiency guarantee is to be based, continuous or peak
2) Gas analysis, Orsat or calculated
3) Chemical analysis of dust or liquid to be collected; include specific gravity, bulk density of dust
with two values, one for volumetric capacity of dust hoppers and the other for structural design
a) Is a representative sample of dust available? (yes)
(no)
4) Particle size analysis
5) Dust load at precipitator inlet:
<§ Continuous rating, iciual pains/cubic foot.
@ Peak rating, actual pains/cubic foot
Conditions on which precipitsior performance guarantee is to be based, continuous or peak
6) Barometric pressure or elevation at plant site.
Section D
Performance
1) Collection ifficiency %.
Section E.
Layout Dnwinp
1) The precipitator will be installed generally in conforming* with the atuched drawing number _
2) Maximum permintible prtnure drop through equipment being supplied by bidder: - inches W.C.
3) Indicate tingle or multipk charaber requirement .
Section F.
DoajB Fatvm
The precipitator wiD have the following itructuzal dcagn features:
1} Operating preawn W.C. Nefttn* -Pootrvt.
E-29
-------
2} Design pressure W,C, Negative . Positive
3) Design temperiiure °F.
4) Ciiin| material ; Thickness _
5) Hopper mittrial ; Thickness
6) Minimum hopper valley angle ° from horizontal,
7) Type of bottom: hoppers (pyramid*], bunker) drag scraper,
wel
Specify ilorige capacity hours,
Seel ion C.
Auxiliary Equipment By Purchaser Bidder Description
Supporting steel ,
Access facilities
Transitran nozzles
Dyetwork & expansion joints ,—
Caies & dampers
Hopper duit vaJvesA conveyors
Control room .
Instrumentation
Fans & auxiliaries
Motor control center - ,
Other . __
Sect ion H.
Erection Scope
1) Erection By: pyrchaiet or bidder , i_
2) Erection luperviior services, separate quote or include in material price ^___
3) Start up and lest engineer sernots, separate quote or include in equipment price
4) Travel and subsistence costs for erection forces by purchaser or bidder
5) Erection period: Sttrtinf date Completion dale.
6) Site information:
Storage area; size in sq. ft. and distance from job site
Aviilabiliiy of closed storage area.
Freedom of crane area
Overhead obstacles
Distance utility souiotiaie from job rite
Is truck roadway and/or railroad right-of-way to storage area ; to job site.
if not, distance from unloading poinl lo storage area .
E-30
-------
7) SOOJM of erection mpontibility: Purchaier Bidder
Foundatk»Qi (pikt or alabi) - — .
Material unloidini to floraa* - .......
Material rehandiinno job tili . — - - —
Low voltage wiring - —
Inuktion (type, tic.) -
Ductwork, ptes, expaniion
joinii, ew. ____________________
Lighting -
Erection equipment: crane i,
welding rnachinei, etc. __________________
Erection facilities: field office,
change shanty A stnJtuy
Erectksn utiiitier. ts.witer, light - -
Field painting: (complete or
toudnip) - -
Electrical cubftation - -
Attach description of above it* mi if not covered by plant lUndardi.
8) Available electric power
For Freciprutof - VolU, - Fha-e, _ Cycle, _ KVA
For Erection - VolU, - Fhaie, _ Cycle, _ KVA
For Control! &
Instrumentation - Volu, - Phaae, _ Cycle, _ KVA
9) Plant standard* attached - „ _ ____
10) Other remaiki or corrunenti _
E-31
-------
REFERENCES FDR APPENDIX E
1. Ahern, A. J, A Coordinated Approach By a Central Group to Maintain
Precipitator Compliance. Presented at APCA Specialty Conference on
Operation and Maintenance Procedures For Gas Cleaning Equipment. Spon-
sored by IGCI and EEI. Pittsburgh, Pa. April 1980.
2. Commonwealth Edison Company. Particulate Emissions Compliance Procedures,
3. Carolina Power and Light Company. Operation and Maintenance Manual For
Electrostatic Precipitators. February 1981.
4. Industrial Gas Cleaning Institute, Inc. Bid Evaluation Form for Elec-
trostatic Precipitators. Publication No. E-P4, January 1968.
5. Industrial Gas Cleaning Institute, Inc. Information Required for the
Preparation of Bidding Specifications for Electrostatic Precipitators.
Publication No. E-P5, January 1973.
APPENDIX E - DATA SHEETS AND EXAMPLE CHECKLISTS r
-------
GLOSSARY OF TERMINOLOGY
GLOSSARY OF TERMINOLOGY G-l
-------
TERMINOLOGY FOR ELECTROSTATIC PAECIP1TATORS
I. ENVELOPING STRUCTURE
A Casing
I. For Rectangular Configuration
a Gastij"hi roof
b. Side V.all
c. End Wall
d. Hoppers and or botloms
e Gas inlet
(. Gas outlet
f. Dmdinf ukll (lasltghi)
h Load oeanriy nail inon-gastighu
i External roof (non-gaslight)
2 Fur Cylindrical Configurat'inr,
a Gas inlet
b Gas outlet
c. Hoppers 2nd or bottom
d. Head
e. Shell
B, Auxiliar) Items :
1. Insulator Carripijnrnem • Enclosure for ihc insuiasortsi
supporting the high * callage sv stem I mas contain one or
more insulators, but not enclosing the roof as a *hc*lel
2. Enclosures
2. Penthouse - A weatherproof, gas-iighi enclosure
o%er the precipitator lo contain the high \ohage
insulators.
b. Upper Weather Enclosure - A non gas-light
enclosure on the roof of the precipitator 10 shelter
equipment (T R seii, r2ppers. purge air fans, eic )
and maintenance personnel.
c. Lower Weather Enclosure - A non gas-light
enclosure at base of prccipitators to protect hoppers
from wind and or detrimental weather conditions.
3. Access Means
a. Doors - A hinged or detached cover prov ided with a
hand operated fastening device where accessibilu) is
required.
a. Control - A deuce mjulled in a duel to rejulaie the
gasfloui bv defree of closure, examples Buiierfh or
Mulu-Louver
b. halation - A dmce installed in a duel to isolate a
prtcipitaior chamber from procos gas
5. Sai'eiv Grounding Deuce • A deuce for physical!}
grounding tlsc h:g6 ^olta^c s\stsm prior to personnel
emennf Ihe precipitator. (The fncst cornmon typve
Cor,vi^£s of a conwucior. one end of uhich i> grounded
to the casing, the other and attached 10 the high voltage
system usmg an insulated operating Ir^'erl
6. Transport - An aeroCsnamicali; ctugntc ink] or
outlet dud connection to the precipitator. Transition;.
arc norrnalh incited as part of the prccipnator,
* G.'> D;sirihu[(f- Defers - ]r.:jrna! e]e-f-:> i- thf
transition or dtjnuori. to produce Ihe desired \dociu
contour at the inlel and outlet lace oi the precipitator.
example turmr.g sanes or pcrfcraied plate*
a. Ann-sneakigc baffles - Internal baffie ekments,
wuhifi ihc precipitator to pre\ent iht pas from
bypassing ihe »;ti>e fieid or causing hopper re-
ir.irainrTten',
b Turning \aties - Varies in ductwork or transition to
guide the gas and dust flow through the ductwork in
order to minimize pressure drop and lo conirol ihe
>elocit> and dust concentration conioun
c Gas distribution plate rapper - A de\tcr used to
prevent dun buildup on perforated plates.
II. COLLECTING SYSTEM
The grounded portion of ihe prtcipitaior lo which ihe
charged dust panicles art driven and 10 which the) adhere.
A. Collecting Surfaces
The individual elements which make up the collecting
system and which collective!} provide thr lotal surface
area of thr prectpiutorfoi Ihe deposition of dust panicles,
B. Collect,nj Surface Rapper
Ade%icr for imparting vibration or ihock to the collecting
surface to dislodge the deposited panicle* or dust
b. Boiled Plate - A cover provided »ith sufficient bolts ,
to insure light closure *here occanortil accessibiliti III. HIGH VOLTAGE SYSTEM
u required.
All pans of the preapiuior which *rc maintained it a high
4, Dampers electrical potential
GLOSSARY OF TERMINOLOGY
G-2
-------
A. High Voltage Structure
E. Chamber
The structural elements necessaryto support the discharge
electrodes in their relation to ihe collecting surface by
means of high voltage insulators. '
B. Discharge Electrode
Trie pan which is installed in the high voltage system to
perform the function of ionizing the gas and creating the
electric field. Typical configuraiions (see sketch) are:
Rigid Frame
Weighted Wire
Rigid Discharge Electrode
C. Discharge Elecirode Rapper
A device for imparting vibration or shock 10 the discharge
electrodes in order to dislodge dust accumulation.
D. High Voltage System Support Insulator
A deuce to physically support and electrically isolate the
high voltagt system from ground.
E. Rapper Insulator
A device to electrically ibOlate. discharge electrode rappers
yet transmit mechanically, forces nfccsb4r> !o crcaie
\ibrasion or shock in the high sokage 5} im-
mersed in mineral oil or S!licone nil
A sirtgk precipitator is an arrangement of collecting
surfaces and discharge electrodes contained within one
B Bus Section
The smallest ponmri of Ihe precipitaiqr which can be
independently dc-enirr^jfied (by aub-dAisson of Ihe high
voltage s;> stem and arrangement of support insulators.)
C. Field (In Depth)
A field is an arrangement of bus sections perpendicular to
gas flow, thai is energized by one or more high soilage
power supplies.
D. Cell (In Width)
A cell is in arrangement of bus sections parallel to gas
now.
Note: Number of cells wide limes number of fields deep
ec uals ihe loial number of bus sections
2. O;hcr
Older precipitaior installations rna> be
equipped >A:ih electronic iu^e or selenium
r™euf!erv ho^^xcr ih^M- t;,p-> are ob*o!c'.c
todiv
B. Impedartce Devices
I, Linear inductor or current limiting reactor retired ;o
»orl "ith SCR-iyps controllers.
2, A transformer with i specially designed high impedance
core and coiU.
3, Saiurable core reactor.
4, Resistor*.
C. Control Equipment • Consists Mainly Of:
a. High Voltage Power Supply Control Equipment
Eject rxal component* required to protect, monitor, and
Rfutite the power supplied to the precipilaior high
GLOSSARY OF TERMINOLOGY
6-3
-------
collage system. Regulating the prirnzr* solugc of the
high vohage :i*snsfc:rrr,LN-rrct;;,ef is «ccomp!,:hed b'.
one of ihe following deuces
I Saiurable Core Reactor
A variable impedance de\ice.
2, Variable *Auio-Tramforrner" control
3, Sjhcon-ConuoJk'd Rectifier (SCRj - Electronic
switch for \okagc regulation,
1 he above Cijti b.
1. Automatic pr»»cr suppl\ control • ^uiomanc
regulation o! hi£h ^oltapc f^o^er for change^ ITI
precipnaior operating condition* usili?in£ feed
: M.r.u^1 I'owi-r Sup,-!-. I'lir.:;^ - \Vn-jj.
precipiiator operating conditions as observed b%
pjant operators
b Au.xiliar) Control Equipment
Electrical component* required to pi nect. monitor, and
C'inito' :hc operation of precipiiaior rappers, heaters.
and othct associated equipment
D High Soilage Conductors
Conductor to transmit Ihe high \oltage from ihe
transformer-rectifier to ihe precipiiaior high voltage
system.
1 H.V. Bus
A conductor enclosed vnithin a grounded duct,
2. Cable
a. Oil-filled cable
b. Dry cable
E. Genera! Terms
I. Primar) Current
Currem in Ihe transformer primary as measured by an
AC »mmeier.
2. Primar}' Voliagt
The voltage as indicated b> AC voltmeter across ihe
primary of ihe iransformer.
3. Precipiiaior Current
The rectified 01 unidirectional auraje curren: to Ihe
prfcipaio: meis-'fes! h; u rr,i!l;arr,;icr in tin piJund Ie|
of the recitfie'
4. Precip^iator \ oltage
The a\rragf DC xol'.agc bciueen the high voliagc
svsiern and grounded udt of the precipiiator.
5- Spark
A distharge from the high lolugf syiirm lo ihe
grounded system, ietf-eRtingujshing and of short
Arc
A discharge o.' suastantial maj-niiude of the high
^oltbiT s}^rrm to !.hf grounded system, of rdaf,nel}
long duraiion and not tending 10 be immsflistel) self-
VI, GENERAL TERMS
A DL-M o' Misi Concentration
The ueighi of dusi or misi contained in a wnn of gas, t.g
pound? per thousand pounds of gas. grains per sciual
cubic loos ol gas. or grains per standard drj CUBIC fool ithe
temperature and pressure of the ga> must be specified sf
guen a> velum:]
B Collection Efficiency
The weight cf dust collected per unit iirnc divided b> ihe
ueijht of dust entering Ihe precipiiaior during the same
unii time expressed in percentage. The compulation is M
(Dust in)-(Dust out)
Efficienci, = X 100
{Dujiir)
C. Prenpiiaior Dimensiom
1, Effective Length
Totai length of collecting surface measured in ihe
direction cf gas f!o». Ltnph between fields is lo be
excluded.
2. Effective Height
Total height of collecting surface measured from top 10
bouom.
3 Effective Width
Toial ntmber of gas passages multiplied by ipacirtg
dimension of the collecting surfaces
4, Effraivr Cross-Seciioru! Area
Effective width limes effecit« height.
GLOSSARY OF TERMINOLOGY
G-4
-------
D. Precipitator Gas Velocity
A figure obtained fc> dividing ihe volume rai: of gas flo*
through the precipitttor by the effective cross-sectional area
of the preciprtatof- Cas velocity is feneraWy expressed in
icrms of ft..'tec, and n computed as follows:
Velocity z
Cat Volume (F!.!isec.)
Effective cross-KCtion area (Ft,!)
Effective cross-section is construed 10 be the effective field
height X width of gas passage X number of passages
E. Treatment Time
A figure, in seconds, obtained by dividing ihe effective length.
in feet, of a precipiutor by the predpitator gas velocity figure
calculated above.
F. Asptci Raiio
The ratio obtained bs dniding effective length of the
precipitate" &5 thetffcctiie hnghi.
G. Collecting Surface Area
The total flat projected area of collecting syrfsce exposed ic
ihe active electrical field (effective length x effective height * 2
s number of fas passages),
H Specific Collecting Area [SCAl
A figure obtained b> dnidmgioia! eiiective collecting surfctt
of ihe precipitaidf by $a» volume, expressed in thousands of
aciua! cubic feet pet minute.
1. Migration %'elocit>
^ FursiTH'ter ir ;h? Dtfutch-And^f^on cqustion u>ed to
determine the required size of an electrostatic prectpitator to
meet ipecified design condiuons Other terminology are ^~
vslue and precipitation rate. Values are generally stated in
terms ci H, mm or cm iec.
J. Current Dcnsiij
The amoum of stcondarv curreni per unit of ESP collecting
surface. Common units are ma fi.-'and nA cm:
K. Corona Power (KW)
The product of secondary cyrrent and secondarj voltage
Power density is gerverallv expressed in terms of: ID nans per
square foot of collecting surface, or (2) »ziis per 1000 ACFM
of gas flow.
L. Rapping Intensity
The "g" force measured a! various poinu on collecting or
discharge electrodes. Measured forces should be speciftrt as
longuudinal or transverse.
M.Oas Passage
Formed by swo adjacent ro»s of collecting surfaces'.
measured from collecting surface centerlme to collecting
surface cemerline.
N, Hopper Capacity
Total volumetric capacity of hoppers measured from a plane
10" below high voltage system or plates, whichever is lower.
GLOSSARY OF TERMINOLOGY
G-5
-------
men YOLrioi tvrron »»ULITO>
Gil**tt»Ct
• Ul «iCTI8* !H TTPICM.)
6-6
-------
M
u
a
w
J
I
I-
E
i
Q
IU
I-
X
Oi
UJ
6-7
-------
REFERENCES
1. Industrial Gas Cleaning Institute, Inc. Terminology for Electrostatic
Precipitators. Publication ED-1 January 1984 Revision.
GLOSSARY OF TERMINOLOGY G-8
------- |