EPRI
U.S. Environmental Industrial Environmental Research EPA-600/8-79-009
Protection Agency Laboratory April 1979 ^
Office of Research Research Triangle Park NC 27711
and Development
Electric Power 3412 Hillview Avenue P/N 3283
Research Institute Palo Alto CA 94303
Lime FGD Systems
Data Book
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
• 2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the SPECIAL REPORTS series. This series is
reserved for reports which are intended to meet the technical information needs
of specifically targeted user groups. Reports in this series include Problem Orient-
ed Reports, Research Application Reports, and Executive Summary Documents.
Typical of these reports include state-of-the-art analyses, technology assess-
ments, reports on the results of major research and development efforts, design
manuals, and user manuals.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/8-79-009
April 1979
Lime FGD Systems
Data Book
by
T.C. Ponder Jr., J.S. Hartman, H.M. Drake, R.P. Kleir,
J.S. Master, A.N. Patkar, R.D. Terns, and J.D. Tuttle
PEDCo Environmental, Inc.
11499 Chester Road
Cincinnati, Ohio 45246
Contract No. 68-02-2603
Tasks No. 5 and 35
Program Element No. EHE624A
EPA Project Officer: Warren D. Peters
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
EPRI Project Officer: Thomas A. Morasky
Electric Power Research Institute
3412 HiIIview Avenue
Palo Alto, CA 94303
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
PREFACE
To date, lime-based FGD systems account for approximately
11,000 of the 28,000 MW of FGD capacity installed or under con-
struction. As a result, a large information base is becoming
available, but up to now it has not been compiled in a format
that is readily accessible and usable by the utility industry.
The Lime FGD Systems Data book should permit a utility to anti-
cipate the performance, reliability, and maintenance character-
istics of alternative lime scrubbing system designs available,
as a function of site-specific variables. This information
should improve the quality of bid specifications as well as the
ability to judge the merit of alternative lime scrubbing system
proposals.
The objective of the Lime FGD Systems Data Book is to
provide the utility industry with 1) detailed guidelines about
design features, equipment specifications and selection criteria
of lime scrubbers, and 2) specific procedures to determine which
system design parameters are critical in confidently selecting a
lime slurry system. The book is designed to enable a utility
engineer to predict and/or specify scrubber system parameters
such as energy requirements, equipment and vessel sizes, system
efficiencies, equipment and subsystem redundancy needs, scrubber
waste characteristics, fresh water makeup, maintenance require-
ments, and system costs. Proper implementation of the informa-
tion in this manual will result in scrubbing systems having
increased reliability and decreased maintenance needs. In
addition, the Data Book describes the process chemistry involved
in lime scrubbing and highlights the interrelationship of pro-
cess chemistry with the proper selection of system components.
It is essential to understand this relationship to apply a
logical chemical engineering approach when integrating a lime
scrubbing process into a utility boiler system.
11
-------
CONTENTS
Preface
List of Tables
List of Figures
Conversions
Abbreviations
Acknowledgements
1.0 Introduction
1.1 General 1.1-1
1.2 Project Purpose and Scope 1.2-1
1.3 Description of Contents 1.3-1
2.0 Process Design
2.1 Introduction 2.1-1
2.2 Effects of Design Parameters on Process Design 2.2-1
2.3 Procedures for Calculating a Material Balance 2.3-1
2.4 Sludge Disposal 2.4-1
3.0 Process Control
3.1 Introduction 3.1-1
3.2 Practical Applications of Process Control 3.2-1
3.3 Lime Scrubber Control Subsystems 3.3-1
3.4 Basic Control Hardware 3.4-1
4.0 Equipment Design
4.1 Introduction 4.1-1
4.2 Recirculating Pumps 4.2-1
4.3 Other Process Pumps 4.3-1
4.4 Lime Unloading and Storage 4.4-1
4.5 Slurry Preparation 4.5«-l
4.6 Scrubber/Absorber 4.6-1
4.7 Mist Eliminator 4.7-1
4.8 Fans 4.8-1
4.9 Thickener/Clarifier 4.9-1
4.10 Mechanical Dewatering Equipment 4.10-1
4.11 Reheaters 4.11-1
4.12 Corrosion 4.12-1
4.13 Instrumentation 4.13-1
iii
-------
CONTENTS (continued)
5.0 Bid Request/Evaluation
5.1 Introduction 5.1-1
5.2 Design Basis 5.2-1
5.3 Guarantee Requests 5.3-1
5.4 Equipment and Instrumentation Summary 5.4-1
5.5 Bid Evaluation 5.5-1
Glossary G-l
iv
-------
LIST OF TABLES
Table Page
2.2-1 Factors That Influence Scrubber System Design
Outside the Battery Limits , 2.2-2
2.2-2 Factors That Influence Scrubber System Design
Inside the Battery Limits 2.2-3
2.2-3 Typical Analysis of Representative U.S. Coals
As Received , 2.2-4
2.2-4 Average Ash Constituents of Three Ranks of Coal , 2.2-12
2.2-5 Ash Analyses of Selected Coals 2.2-13
2.2-6 Constituents and Properties of Selected Coals 2.2-16
2.2-7 Composition of Various Grades of U.S. Coals 2.2-17
2.2-8 Classification of Coal by Rank 2.2-18
2.2-9 Coal Analyses and Sulfur Variability Over Various
Averaging Times 2.2-19
2.2-10 Process Approximations 2.2-51
2.3-1 Enthalpies of Various Gases 2.3-2
2.3-2 Molecular Weights Frequently Used in Material
Balance Calculations «• 2.3-3
2.3-3 Energy Requirement Calculations 2 .3-4
2.3-4 Design Information 2.3-17
2.3-5 Typical Pressure Drop Data 2.3-18
2.3-6 Lime Analysis • 2.3-18
2.3-7 Composition of the Available Lime for S02 Absorption ... 2.3-42
2.3-8 Waste Slurry Suspended Solids 2.3-48
2.3-9 Inlet Flue Gas to Scrubber (Stream 6) 2.3-49
2.3-10 Cleaned Flue Gas Composition (Stream 7) 2.3-53
v
-------
LIST OF TABLES (Continued)
Table
2.3-11 Design Boiler and Fuel Data for Bruce
Mansfield Nos. 1,2, and 3 2.3-59
2.3-12 Bruce Mansfield Scrubber and Absorber Data 2.3-62
2.3-13 Boiler and Fuel Data, Came Run No. 4 2.3-64
2.3-14 Pertinent Boiler and Fuel Data for Conesville No. 5 2.3-67
2.3-15 Material Balance for Conesville No. 5 2.3-70
2.3-16 FGD Tank Information for Conesville No. 5 2.3-72
2.3-17 Pertinent Boiler Data, Green River Plant 2.3-74
2.3-18 Fuel Data, Green River Plant 2.3-74
2.3-19 Green River Scrubber Data 2.3-77
2.3-20 Reaction Tank Data, Green River Plant 2.3-79
2.3-21 Makeup Water Requirements, Green River Plant 2.3-79
2.3-22 Boiler Data for Unit 6, Paddy's Run Station 2.3-81
2.3-23 Fuel Data for Unit 6, Paddy's Run Station 2.3-81
2.3-24 Absorber Data, Paddy!s Run 2.3-83
2.3-25 Carbide Lime Analysis, Paddy's Run 2.3-84
2.3-26 FGD Tank Data, Paddy's Run 2.3-85
2.3-27 Thickener and Vaccuum Filter Data, Paddy's Run 2.3-86
2.3-28 Water Requirement, Paddy's Run 2.3-87
2.3-29 Boiler Data, Phillips Power Station 2.3-88
2.3-30 Phillips Power Scrubber Data for Particulate
and FGD Scrubber Modules 2.3-89
2.3-31 Conversion Factors 2.3-92
2.4-1 Lime FGD Systems that Fixate Sludge 2.4-5
2.4-2 Flue-Gas-CleaninB and Sludge Disposal Practices for
Utility Scrubbers Using Throwaway Processes and Oper-
ational on November 1, 1977 2.4-17
vi
-------
LIST OF TABLES (Continued)
Table
2.4-4 Summary of the Revenue Requirements During the
30-Year Life of a New Power Plant for Various
FGD Waste Disposal Methods 2.4-22
3.4-1 Recommended Control Modes 3.4-5
4.2-1 Survey of Vendor Pump Specifications • ••- 4.2-14
4.2-2 Slurry recirculation Pumps Malfunctions and Causes 4.2-16
4.2-3 Scrubber slurry recirculation pump specifications—
Existing Facilities 4.2-17
4.3-1 Lime FGD Systems Pump Data—Lime Slurry Feed Pumps 4.3-2
4.3-2 Lime Scrubber Pump Data Thickener Supernatant Pumps .... 4.3-2
4.3-3 Lime Scrubber Pump Data—Thickener Underflow Pumps 4.3-7
4.4-1 Standard Sizes of Quicklime 4.4-2
4.4-2 Physical properties of Quicklime 4.4-4
4.4-3 Covered or Hopper Barges r 4.4-10
4.4-4 Existing Facility Design Specifications 4.4-24
4.5-1 Characteristics of Detention and Paste Slakers 4.5-15
4.5-2 Operating Characteristics of Lime Slaking Facilities ... 4.5-19
4.5-3 Approximate Agitator Motor Horsepower Required
for Lime Suspension
4.5-4 Stabilization and Storage Systems *•5~25
4.6-1 S02 Absorbers in Operational Lime FGD Systems 4.6-2
4.6-2 Venturi Throat Corrosion Spool Test Data 4.6-6
4.6-3 Scrubber Corrosion Spool Test Data Below and Above
the Mist Eliminator • *' 6~7
4.6-4 Conditions for the Corrosion Tests at the
Shawnee Station of TVA - 4.6-8
4.6-5 Chemical Analysis of Alloys 4.6-9
vii
-------
LIST OF TABLES (Continued)
Tables
4.6-6
4.6-7
4.6-8
4.6-9
4.6-10
4.6-11
4.6-12
4.9-1
4.9-2
4.9-3
4.9-4
4.11-1
4.11-2
4.12-1
4.13-2
4.13-3
4.13-4
4.13-5
4.13-6
5.4-la
5.4-2
Typical Characteristics of Epoxy Resins
Property Characteristics for Natural Rubber and
Neoprene Rubber
Properties of Type "H" and "L" Bricks
Summary of Operating Lime FGD Systems as of
January 1978
Chemical Analysis of the Sludge of a Lime-
Typical Size Distribution of a Lime -Generated Sludge . . .
Summary of the Characteristic Differences Between
the Two Classes of Colloids
Existing Thickener Facilities for Lime Scrubbing
FGD System ,
Corrosion of Reheater Tubes from Colbert Pilot
Plant After 3800 Hours
Solids Content Instrumentation
EPRI Lime FGD Systems Data Book Equipment List
EPRI Lime FGD System Data Book Venturi Scrubber
Page
4.6-11
4.6-12
4.6-13
4.6-14
4.6-16
4.6-18
4.6-34
4.9-6
4.9-18
4.9-19
4.9-30
4.11-25
4.11-26
4.12-14
4.13-6
4.13-8
4.13-11
4.13-17
4.13-24
5.4-2
5.4-4
viii
-------
LIST OF TABLES (Continued)
Tables Page
5.4-3 EPRI Lime FGD System Data Book Venturi
Recirculation Pump Specifications 5.4-6
5.4-4 EPRI Lime FGD System Data Book Venturi
Recirculation Tank Specifications 5.4-8
5.4-5 EPRI Lime FGD System Data Book Presaturator
Specifications 5.4-9
5.4-6 EPRI Lime FGD Systme Data Book Absorber Specifications.. 5.4-12
5.4-7a EPRI Lime FGD System Data Book Reaction Tank
Pump Specifications , 5.4-14
5.4-7b EPRI Lime FGD System Data Book Absorber
Recirculation Pump Specifications 5.4-15
5.4-8 EPRI Lime FGD System Data Book Absorber
Recirculation Tank Specifications 5.4-17
5.4-9 EPRI Lime FGD Systems Data Book Mist Eliminator 5.4-19
5.4-10 EPRI Lime FGD System Data Book Centrifugal Compressor
Specifications 5.4-22
5.4-11 EPRI Lime FGD System Data Book Damper Specifications ... 5.4-24
5.4-12 EPRI Lime FGD System Data Book Duct Work Specifications.. 5.4-26
5.4-13 EPRI Lime FGD System Data Book Booster Fan
Specifications 5.4-28
5.4-14 EPRI Lime FGD System Data Book Belt Conveyor
Specifications 5.4-30
5.4-15 EPRI Lime FGD System Data Book Slaker Specifications ... 5.4-32
5.4-16 EPRI Lime FGD System Data Book Lime Stabilization/
Storage Tank Specifications 5.4-34
5.4-17 EPRI Lime FGD System Data Book Lime Slurry
Pump Specifications 5.4-35
5.4-18 EPRI Lime FGD System Data Book Freshwater Pump
Specifications 5.4-37
5.4-19 EPRI Lime FGD System Data Book Thickener
Specifications 5.4-39
ix
-------
LIST OF TABLES (Continued)
Table
5.4-20 EPRI Lime FGD System Data Book Flocculant
Proportioning Pump Specifications 5.4-41
5.4-21 EPRI Lime FGD System Data Book Thickener
Underflow Pump Specifications , 5 .4-42
5.4-22 EPRI Lime FGD System Data Book Thickener
Overflow Pump Specifications 5.4-44
5.4-23 EPRI Lime FGD System Data Book Thickener
Overflow Tank Specifications 5.4-45
5.4-24 EPRI Lime FGD System Data Book Centrifugal
Separator Specifications * 5.4-47
5.4-25 EPRI Lime FGD System Data Book Vacuum
Filter Specifications 5.4-49
5.4-26 EPRI Lime FGD System Data Book Filtrate or
Centrate Pump Specifications 5.4-51
5.4-27 EPRI Lime FGD System Data Book Fixation Tank
Specifications 5.4-53
5.4-28 EPRI Lime FGD System Data Book Sludge Disposal
Specifications 5.4-55
5.4-29 EPRI Lime FGD System Data Book Belt
Conveyor Specifications 5,4-57
5.4-30 EPRI Lime FGD System Data Book Screw
Conveyor Specifications 5.4-59
5.4-31 EPRI Lime FGD System Data Book Pond
Water Return Pump Specifications 5.4-60
5.4-32 EPRI Lime FGD System Data Book pH
Instruments Specifications 5.4-63
5.4-33 EPRI Lime FGD System Data Book Liquid
Level Indicator Specifications 5.4-65
5.4-34 EPRI Lime FGD System Data Book Flowmeter
Specifications •. • 5.4-68
5.4-35 EPRI Lime FGD System Data Book S02
Analyzer Specifications 5.4-70
x
-------
LIST OF TABLES (Continued)
Table
5.4-36 EPRI Lime FGD System Data Book Pressure
Sensors Specifications , 5.4-71
5.4-37 EPRI Lime FGD System Data Book Temperature
Sensors and Controllers Specifications 5.4-73
5.4-74 EPRI Lime FGD System Data Book Control
Valve Specifications 5.4-74
5.5-1 Bid Evaluation for Slakers 5.5-2
5.5-2 Lime Slurry Pump 5.5-3
5.5-3 Bid Evaluation for S02 Absorber 5.5-4
5.5-4 Bid Evaluation for Mist Eliminatory 5.5-5
5.5-5 Bid Evaluation for Reheater 5.5-6
5.5-6 Bid Evaluation for Booster Fan 5.5-7
5.5-7 Bid Evaluation for Recirculation tank 5.5-8
5.5-8 Bid Evaluation f6r Thickener 5,5-9
xi
-------
LIST OF FIGURES
Figure Page
2.3-1 Additive vs. outlet requirements 2.3-5
2.3-2 Horsepower vs. outlet requirements 2.3-6
2.3-3 Recirculation pump energy requirement 2.3-7
2.3-4 Fan energy requirement 2.3-8
2.3-5 Reheat energy requirement 2.3-9
2.3-6 Psychrometric chart 2.3-10
2.3-7 S02 emission calculation 2.3-11
2.3-8 Lime requirement calculation 2.3-12
2.3-9 Saturated gas calculation , 2.3-13
2.3-10 Recirculation tank capacity calculation 2.3-14
2.3-11 Wet sludge calculation 2.3-15
2.3-12 500-MW model plant for example 1 2.3-16
2.3-13 Overall material balance 2.3-22
2.3-14 Boiler-furnace material balance ,, 2,3-23
2.3-15 S02 and particulate balance 2.3-24
2.3-16 Scrubber material balance 2.3-26
2.3-17 Detailed boiler-furnace material balance 2.3-27
2.3-18 Summary of S02 and particulate material balance 2,3-41
2.3-19 Summary of slurry preparation material balance 2.3-46
2.3-20 Pond material balance 2.3-48
2.3-21 Psychometric chart for example 1 2.3-52
2.3-22 Summary of scrubber material balance 2.3-56
2.3-23 Water balance summary 2.3-58
2.3-24 Simplified flow diagram of Bruce Mansfield 2.3-60
xii
-------
LIST OF FIGURES (Continued)
Figure Page
2.3-25 Flow diagram of Cane Run No. 4 2.3-65
2.3-26 Flow diagram of Conesville No. 4 2,3-69
2.3-27 Flow diagram of Green River Station 2.3-76
2.3-28 Flow diagram of Paddy's Run 2.3-82
2.3-29 Flow diagram of the dual-stage scrubber train
at Phillips 2.3-90
3.2-1 Feedback control loop in a stack gas reheat
control system 3.2-2
3.2-2 Improved feedback control in stack gas reheat
control system 3.2-5
3.2-3 Cascade control 3.2-6
3.2-4 Feed-forward control 3.2-8
3.2-5 Lime scrubber control system 3.2-11
3.3-1 Fuel-fired reheat control system 3.3-2
3.3-2a Hot water reheat control system: steam raises
hot water temperature 3.3-4
3.3-2b Hot water reheat control system: flow of water
charged by temperature controller 3.3-5
3.3-3 Shape of neutralization curve »• • • 3.3-7
3.3-4 Simple pH control 3.3-9
3.3-5 Cascade pH control « 3.3-11
3.3-6 Outlet S02 feedback control (limited) 3.3-13
3.3-7 Inlet S02 feed-forward control 3.3-14
3.3-8 Lime slurry solids content: density feedback 3.3-16
3.3-9 Lime slurry solids content: feed-forward control 3.3-17
3.3-10 Absorber solids content: thickener bleed control 3.3-19
3.3-11 Absorber solids content: thickener bleed control 3.3-20
xiii
-------
LIST OF FIGURES (Continued)
Figure
3.3-12 Thickener solids content: underflow control flow 3.3-21
3.3-13 Thickener solids content: recycle control 3.3-21
3.3-14 Thickener solids content: flocculant control 3.3-23
3.3-15 Absorber: thickener solids control system 3.3-24
3.3-16 Simple pressure control 3.3-26
3.3-17 Basic fan control subsystem 3.3-27
3.3-18 Module load balancing control 3.3-28
3.3-19 Flue gas flow: fan control 3.3-30
3.4-1 Theoretical controller response curves 3.4-4
3.4-2 Valve characteristics 3.4-7
4.2-2 Slurry recirculation pump detail assembly 4.2-3
4.2-3 Specific gravity of scrubber recirculation slurry 4.2-6
4.2-4 Types of impellers for recirculation pumps 4.2-9
4.2-5 Pump seal water flow 4.2-11
4.4-1 The angle of repose 4.4-4
4.4-2 Pressure-differential tank trailer 4.4-7
4.4-3 Hopper trailer with air-activated gravity
discharge hopper 4.4-9
4.4-4 Covered hopper car 4.4-9
4.4-5 Vacuum barge unloading • 4.4-11
4.4-6 Blower truck receiving bin 4.4-13
4.4-7 Vacuum railcar unloading 4.4-15
4.4-8 Mechanical conveyors 4.4-17
4.4-9 Positive-pressure pneumatic conveyor 4.4-18
4.4-10 Closed-loop pneumatic system 4,4-20
xiv
-------
LIST OF FIGURES (Continued)
Figure Pa8e
4.4-11 Feed bin design ........................................ 4.4-23
4.5-1 Lime slurry system — schematic flow ..................... 4.5-2
4.5-2 Volumetric belt-type feeder ..... ....................... 4.5-4
4.5-3 Oscillating hopper volumetric feeder ................... 4.5-5
4.5-4 Mechanical gravimetric feeder ......... .... ...... ....... 4.5-7
4 . 5-5 Gravimetric feeder ..................................... 4 . 5-9
4.5-6 Detention s laker ............. ......................... 4.5-13
4.5-7 Paste slaker ........................................... 4.5-14
4.5-8 Typical stabilization tank (simplified) ................ 4.5-21
4 . 5-9 Agitator baffle design ................................. 4 . 5-24
4.5-10 Bruce Mansfield stabilization and storage .............. 4.5-26
4.5-11 Conesville stabilization and storage ....... ............ 4.5-27
4.5-12 Phillips stabilization and storage ..................... 4.5-29
4.5-13 Green River slaking and stabilization system ........... 4.5-30
4.7-la Discontinuous horizontal chevron ziezae baffle ......... 4.7-4
4.7-lb Continuous horizontal chevron zigzag baffle ............ 4.7-4
4.7-2 The chevron impingement principle ...................... 4.7-5
4 . 7-3 Heil chevron mist eliminator ...................... ...... 4 • 7-6
4 . 7-4 Heil chevron mist eliminator performance ..... .......... 4 . 7-7
4.7-5 Verticle configuration of Humboldt Lamellar
separator in horizontal duct (looking downward)
marketed by Matsuzaka .................................. ^ . 7~°
it. 7-6 Hunters euro form mist eliminator ........... , ........... 4.7-11
4 . 7-7 Radial-vane mist eliminator ............................ 4 . 7-12
4.7-8 Mist eliminator designs showing the difference in
baffling angle and the number of passes .......... . ..... 4.7-15
xv
-------
LIST OF FKUJRF.S (Continued)
Figure
4.7-9 Schematic of two-, three-, and six-pass
chevron mist eliminators 4.7-16
4.7-10 Slanted mist eliminator for vertical gas flow 4.7-17
4.7-11 Cross section of mist eliminator with special
reentrainment-prevention features 4.7-19
4.7-12 Horizontal and vertical mist eliminators 4.7-21
4.7-13 Bulk separation systems 4,7-23
4.7-14 Koch flexitray wash tray 4.7-24
4.7-15 UOP trap-out tray 4.7-25
4.7-16 Droplet sizing photograph 4.7-31
4.7-17 Percent removal vs. mist droplet diameter obtained
by the application of the CFH method 4.7-32
4.8-1 Scrubber FD fan application 4.8-3
4.8-2 ID fan application 4.8-4
4.8-3 Wet fan application 4.8-7
4.9-1 Sequence of sedimentation in a cylinder 4.9-2
4.9-2 Thickener supported by a center column, with truss-
type rake arms and Thixo post plow blades , 4.9-4
4.9-3 Hinged rod-type designs with two types of
support structure t • 4.9-5
4.9-4 Calcium sulfite sludge 4.9-7
4.9-5 Calcium sulfate sludge 4.9-7
4.9-6 Center drive unit (EIMCO) 4.9-10
4.9-7 Plow blade designs shown with truss-type rake arms 4.9-13
4.9-8 Positive lifting device used on column-supported
thickeners 4•9~15
4.9-9 Standard tunnel system 4.9-16
xvi
-------
LIST OF FIGURES (Continued)
Figure
4.9-10 Underflow pumping arrangements through
center column 4.9-17
4.9-11 Effect of polymer on settling rate of
lime-generated sludge , 4.9-21
4.9-12 Typical automatic dry polymer feed system 4.9-23
4.9-13 Graphical analysis of interface settling curve 4.9-26
4.9-14 EIMCO swinglift thickener at Cane Run Power Station ,.., 4.9-32
4.9-15 Dorr-Oliver cable torq thickener at Conesville
Power Station 4.9-34
4.10-1 Solid bowl centrifuge 4.10-2
4.10-2 Cutaway view of a rotary-drum vacuum filter 4.10-9
4.10-3 Operating zones of vacuum filters 4.10-10
4.10-4 Flow sheet for continuous rotary-drum vacuum filtration.. 4.10-11
4.10-5 Drum and internal piping 4.10-13
4.10-6 Filtration rate vs. feed solids 4.10-17
4.10-7 Dry cake rate vs. cycle time 4.10-17
4.10-8 Filter leaf test apparatus 4.10-21
4.10-9 Cross section of a belt filter 4,10-24
4.10-10 Filter costs 4.10-25
4.10-11 Pox-0-Tec process 4.10-27
4.11-1 Reheat system location * 4.11-2
4.11-2 In-line reheat system 4.11-5
4.11-3a Direct-firing reheat system using in-line burner 4.11-7
4.11-3b Direct-firing reheat system using external
combustion chamber 4.11-7
4.11-4 Indirect hot air reheat system 4.11-9
xvii
-------
LIST OF FIGURES (Continued)
Figure Page
4.11-5 Bypass reheat system , 4.11-10
4.11-6 Reheat by exit gas recirculation , 4.11-10
4.11-7 Schematic of heat balance around downstream
system with indirect hot air reheater 4.11-12
4.11-8 Schematic of heat balance around downstream
system with indirect hot air reheater 4.11-14
4.11-9 Schematic of heat balance around downstream
system with direct combustion reheater ., 4.11-16
4.11-10 Effect of reheat temperature (for in-line reheat)
on ground level SC>2 concentration at various
scrubber efficiencies 4.11-20
4.12-1 Continuous weld and series of spot welds 4.12-9
4.12-2 EB insert ring, consumable flat ring, and standard
backing ring 4.12-10
4.12-3 Double butt weld, single fillet lap weld, and
double fillet lap weld 4,12-11
4.12-4 Examples of good and bad finishing techniques for welds.. 4.12-12
4.12-5 Complete full weld 4.12-13
4.13-1 Ideal S02 sampling system 4.13-10
4.13-2 Displacement level instrument control system 4.13-13
4.13-3 Flange-mounted differential-pressure level
control system 4.13-15
4.13-4 Quadrant-edged orifice plate 4,13-20
4.13-5 Typical magnetic flow meter 4 i13-21
xviii
-------
CONVERSIONS
The U.S. Environmental Protection Agency policy is to
express all measurements in metric units. Generally, however,
this report uses British units of measure which are still com-
monly used in the industry. For conversion to the metric sys-
tem, use the following conversions:
To convert from
°c*
oF
Btu
Btu/scf (70°F)
Btu/lb
ft
ft2
ftVlOOO acfm
ft2
acfm
scfm
gal
gpm/ft2
m3
gal/1000 scf
To
°F
°C
J
MJ/Nm3 (0°C)
kJ/kg
m
m2
m2/(m3/s)
m3
m3/h
Nm3/s
liter
1/min/m2
liter
liter/m3
Multiply
1.8 (°C)
0.556 (°
1.055 E
4.011 E
2.326 E
3.048 E
9.290 E
1.968 E
2.832 E
1.699 E
4.383 E
3.785 E
4.080 E
1.000 E
1.337 E
by
+ 32
F-32)
+ 03*
- 02
+ 00
- 01
- 02
- 01
- 02
+ 00
- 04
+ 00
+ 01
+ 03
- 01
xix
-------
CONVERSIONS (continued)
To convert from
To
Multiply by
gr
gr/ft3
gr/scf
hp
in. W. G.
Ib
Ib moles/hr
lb/106 Btu
lb/ft3
psi
ton
W/ft2
Metric prefixes
g
g/m3
g/Nm3
kw
Pa
kg
g moles/min
g/mJ
kg/m3
Pa
Mg
W/m2
giga (G) 109
mega (M) 106
kilo (k) 103
centi (c) 10~2
milli (m) 10~3
nano (n) 10~9
6.480 E - 02
2.288 E + 00
2.464 E + 00
7.457 E - 01
2.491 E + 02
4.536 E - 01
7.560 E + 00
4.298 E - 01
1.602 E + 01
6.895 E + 03
9.072 E - 01
1.076 E + 01
Abbreviations given on pages ix through xi.
E indicates the power of 10 by which the conversion
factor must be multiplied to obtain the correct value.
XX
-------
ABBREVIATIONS
Symbol
Unit
degree Celsius
degree Fahrenheit
actual cubic foot per minute
British thermal unit
British thermal unit per standard cubic foot
British thermal unit per pound
foot
square foot
acfm
Btu
Btu/scf
Btu/lb
ft
ft2
ft2/I000 acfm
ft3
gal
gpm
gal/1000 scfm
gpm/ft2
gr
gr/scf
gr/ft3
g
square foot per one thousand cubic feet per
minute
cubic foot
gallon
gallons per minute
gallon per one thousand standard cubic feet
gallons per minute per square foot
grain
grain per standard cubic foot
grain per cubic foot
gram
xxi
-------
ABBREVIATIONS (continued)
Symbol
Unit
g-moles
h
hp
in.
in. W. G.
J
kJ/kg
kW
Ib
lb/106 Btu
lb/ft3
Ib-moles
Ib-moles/h
Ib-moles/min
m2
m3
mg
MJ/Nm3
ng
Nm3/s
Pa
ppm
gram moles
hour
horsepower
inches
inch Water Gage
joule
kilojoule per kilogram
kilowatt
pound
pound per million British thermal units
pound per cubic foot
pound moles
pound moles per hour
pound moles per minute
square meter
cubic meter
milligram
megajoule per normal cubic meter
nanogram
normal cubic meter per second
pascal
parts per million
xxil
-------
ABBREVIATIONS (continued)
Symbol Unit
psi pound per square inch
s second
scfm standard cubic foot per minute
W/ft2 watt per square foot
xxiii
-------
ACKNOWLEDGEMENTS
This report has been jointly sponsored by the Electric
Power Research Institute (EPRI), Palo Alto, California, and the
U.S. Environmental Protection Agency (EPA), Industrial Environ-
mental Research Laboratory, Research Triangle Park, North
Carolina.
This report represents the combined efforts of many indi-
viduals. It is impractical to express appreciation to every
contributor who aided in the compilation of data for this book.
The authors, however, do wish to acknowledge the advice and
assistance of Thomas M. Morasky, the EPRI Project Manager, and
Warren D. Peters and J. David Mobley, the EPA Project Officers.
A. V. Slack and Dr. P. S. Lowell also reviewed and contributed
their expertise to this document. The authors would also like
to acknowledge the valuable assistance of the EPRI review com-
mittee, who so graciously contributed their time and effort.
Mr. Timothy W. Devitt and Mr. Thomas C. Ponder, Jr., served
as Project Directors for PEDCo Environmental, Inc., and Mr.
Russell P. Kleir and Mr. J. Scott Hartman served as Project
Managers.
Additionally, the assistance of several PEDCo Environmental
personnel aided in the completion of the book; they are H. M.
Drake, Dr. K. W. Hahn, M. P. Hartman, S. P. Kothari, J. S.
Master, A. N. Patkar, R. R. McKibben, R. D. Terns, and J. D.
Tuttle.
Finally, the authors wish to acknowledge the help afforded
by a great number of process and equipment suppliers, whose
assistance aided in the completeness of the text.
xx iv
-------
SECTION 1
INTRODUCTION
1.1 GENERAL
This data book represents the joint effort of the Electric
Power Research Institute and U.S. Environmental Protection
Agency. The project was initiated under fundings from the EPA
Industrial Environmental Research Laboratory (IERL), Research
Triangle Park, North Carolina, and the Electric Power Research
Institute (EPRI), Palo Alto, California as part of the Tech-
nology Transfer Program. This book is an assemblage of current
data on lime flue gas desulfurization (FGD) technology.
This manual serves to integrate and summarize the results
of extensive utility, architect-engineer, vendor, EPA, and EPRI
efforts in the development of lime scrubbing technology. Much
of the information contained herein is derived from the results
of research projects funded by EPRI and EPA. During the compil-
ation of this manual, review and suggestions regarding the tech-
nical content of this manual were provided by an advisory com-
mittee that consisted of representatives from the utilities,
Edison Electric Institute's Prime Movers Committee, architect-
engineering firms, and the EPA.
1.1-1
-------
1.2 PROJECT PURPOSE AND SCOPE
- -^ssr-nSSL TK^^eSus
disposal site ge lscha^e to the final sludge
1.2-1
-------
1.3 DESCRIPTION OF CONTENTS
The Lime FGD Systems Data Book is organized into five major
sections as discussed below:
1. Introduction
This section describes the general background, pur-
pose, scope, and organization of the manual.
2. Process Design
This section identifies the chemical process design
information associated with lime scrubbing systems and
how the various design parameters are interrelated.
In this section, emphasis is placed on understanding
the method of and need for calculating material bal-
ances for lime scrubbing systems as an aid to specify
and/or check process flow sheets for scrubber systems.
3. Control Systems
This section identifies the critical parameters re-
quired to design a scrubber control system and details
the field experience related to the control and opera-
tion of full scale scrubbers. Information contained
in this section will aid the utility design engineer
in specifying scrubber control systems that will
provide economical, safe, and stable operation of the
flue gas scrubbing process.
4. Equipment Design
In this section, characteristics of specific scrubber
equipment as well as design considerations and operat-
ing histories are presented. Existing lime-based FGD
systems are described along with operating experience,
maintenance practices, and corrective actions. This
information should help specify the individual equip-
ment items required in lime scrubbing systems.
5. Bid Preparation/Evaluation
This section provides guidance for the specification
and evaluation of lime scrubber system bids. This
guidance is intended to ensure that each bid received
from vendors for evaluation contains equipment of the
proper size and type, meaningful guarantees to meet
emission regulations under varying operating condi-
tions, a defined maintenance schedule, and a specified
degree of redundancy.
1.3-1
-------
Additionally a key word glossary of lime scrubbing terms
has been included at the end of the Data Book.
Numerous EPA publications and EPRI published reports pro-
vided inputs to the construction of this manual. Extensive
references follow each section.
1.3-2
-------
SECTION 2
PROCESS DESIGN
2.1 INTRODUCTION
This chapter of the Lime FGD Systems Data Book presents
discussion of many of the major factors that affect the design
of an FGD system.
All of these sections point toward the overall design of an
FGD system and enumerate a number of items that must be con-
sidered in any such design.
In Section 2.2, the effects of 19 design parameters on FGD
system design and their interrelation are discussed. Among
these are items such as coal properties, absorber type, the
effect of regulatory constraints, and redundancy.
In Section 2.3, a detailed procedure for calculating a
material balance for a specific plant site is given. The step-
by-step calculations and descriptions will enable a lime scrub-
ber system design engineer to calculate a material balance for a
specific plant. At the end of this section information on
operable lime FGD systems is presented that can be used for
comparison.
In Section 2.4, a review of the present status of FGD
sludge handling and disposal is presented. For more in-depth
information, the reader is referred to EPRI reports FP 671,
Volumes 1 through 4.
2.1-1
-------
CONTENTS
2.2 EFFECTS OF DESIGN PARAMETERS
2.2.1 Coal Properties
2.2.1.1 Sulfur Content and Type 2.2-1
2.2.1.2 Ash Content 22-9
2.2.1.3 Fly Ash 2!2-9
2.2.1.4 Chloride Content 2.2-11
2.2.1.5 Heating Value 2.2-14
2.2.1.6 Moisture Content 2.2-14
2.2.1.7 Combustion Variability 2.2-15
2.2.2 Boiler Characteristics 2.2-15
2.2.2.1 Type of Boiler 2.2-15
2.2.2.2 Size of Boiler 2.2-21
2.2.2.3 Age of Boiler/Air Leakage 2.2-21
2.2.2.4 Flue Gas Flow 2.2-21
2.2.2.5 Additional Control Equipment 2.2-21
2.2.2.6 Loading Characteristics 2.2-22
2.2.3 Flue Gas 2.2-22
2.2.3.1 Temperature 2.2-22
2.2.3.2 Flow 2.2-23
2.2.3.3 Dew Point 2.2-23
2.2.3.4 Particulate Loading 2.2-23
2.2.3.5 Particulate Alkalinity 2.2-24
2.2.4 Lime Properties 2.2-24
2.2.4.1 Calcium and Magnesium Contents 2.2-24
2.2.4.2 Impurities/Grits 2.2-25
2.2.4.3 Reactivity 2.2-26
2.2.4.4 Size 2.2-28
2.2.5 Makeup Water 2.2-29
2.2.5.1 Chemical Composition and Variability 2.2-29
2.2.5.2 Source 2.2-30
2.2-i
-------
CONTENTS (continued)
2.2.6 Site Conditions
2.2.6.1 Land Availability 2.2-31
2.2.6.2 Soil Permeability 2.2-31
2.2.6.3 Ambient Humidity 2.2-32
2.2.6.4 Rainfall 2.2-32
2.2.6.5 Climate 2.2-32
2.2.7 Regulations 2.2-33
2.2.7.1 SO2 Emission Standards 2.2-33
2.2.7.2 Particulate Standards 2.2-33
2.2.7.3 Plume Visibility Standards 2.2-33
2.2.7.4 Water and Land Requirements 2.2-34
2.2.8 Absorber Type 2.2-35
2.2.9 Waste Slurry Disposal 2.2-37
2.2.10 Redundancy 2.2-38
2.2.11 Method of Reheat 2.2-39
2.2.12 Degree of Instrumentation 2.2-41
2.2.13 Mist Eliminator Configuration 2.2-41
2.2.14 Losses Throughout the System 2.2-43
2.2.15 Process Layout 2.2-44
2.2.16 Materials of Construction 2.2-44
2.2.17 Chemistry 2.2-45
2.2.17.1 pH Gradient 2.2-45
2.2.17.2 Sulfite to Sulfate Oxidation 2.2-46
2.2.17.3 Chloride Balance 2.2-46
2.2.17.4 Liquid-to-gas Ratio 2.2-46
2.2.17.5 Point of Fresh Slurry Addition 2.2-47
2.2.17.6 Scaling 2.2-47
2.2.17.7 Oxidation of SO2 to SO3 2.2-49
2.2.17.8 NO Interferences 2.2-50
2.2.17.9 Stoichiometric Ratio 2.2-50
2.2.17.10 Lime Utilization 2.2-50
2.2.18 Process Approximations and Design Data 2.2-50
References 2.2-53
2.2-ii
-------
2.2 EFFECTS OF DESIGN PARAMETERS ON PROCESS DESIGN
This section discusses the major factors both inside and
outside the battery limits of the SO2 absorber that influence
the design of the FGD system. These factors are listed in
Tables 2.2-1 and 2.2-2.
This section summarizes information presented in the body
of this data book, providing a synopsis of items that may great-
ly affect FGD system operation. Individual sections containing
additional information are noted.
2.2.1 Coal Properties1
Properties of the coal fired in a utility boiler determine
whether, or to what degree, particulate and SO2 controls are
needed. Typical analyses of representative U.S. coals are
presented in Table 2.2-3. The power generation industry uses
coal from fields throughout the country. In general, the east-
ern fields contain anthracite and bituminous coals of medium and
high volatility while the western fields contain subbituminous
and lignitic coals.
In designing an FGD system, the user has two principal
means of evaluating the coal to be burned. First, he may obtain
detailed laboratory analyses of the chemical and physical prop-
erties of the coal and call upon combustion experts for evalua-
tion on the basis of experience with a similar fuel in a compar-
able combustion unit. If this type of information is lacking or
if laboratory test results indicate marginal performance, the
utility may undertake full-scale testing by burning a sample
load.
Full-scale testing over a minimum 1-week period, in con-
junction with use of laboratory data to predict combustion
performance, is the best available means of evaluating new coal
sources for a combustion unit. The short-term full-scale test,
however, may not disclose all combustion problems, since some
ash deposition problems occur only after a "conditioning" period
(up to a few months).
2.2.1.1 Sulfur Content and Type2 —
To meet the S02 emission constraints on utility boilers,
personnel at some coal firing installations have considered
burning low-sulfur coals. Following are some of the points to
be considered.
1. Anthracite is commonly low in sulfur, but this low-
volatility fuel is difficult to pulverize and burn and
is not suitable for a steam generator arranged to burn
high-sulfur, bituminous coal of medium to high vola-
tility.
2.2-1
-------
Table 2.2-1. FACTORS THAT INFLUENCE SCRUBBER SYSTEM DESIGN
OUTSIDE THE BATTERY LIMITS
Coal properties
Sulfur content, type
Ash content
Fly ash composition, particle size
Chloride content
Heating value
Moisture content
Composition variability
Boiler design
Type of boiler
Size of boiler
Age of boiler
Flue gas flow
Additional control equipment
Loading characteristics
Lime properites
Percent inert
Ca, Mg contents
Reactivity
Size
Site conditions
Land availability
Soil permeability
Ambient humidity
Rainfall
Climate
Regulations
SO2 emission/ambient standards
Particulate standards
Plume visibility standards
Water/land standards
Makeup water
Chemical composition
Source
Flue gas
Temperature
Flow
Dew point
Particulate loading
Particulate alkalinity
2.2-2
-------
Table 2.2-2. FACTORS THAT INFLUENCE SCRUBBER SYSTEM DESIGN
INSIDE THE BATTERY LIMITS
Absorber type
Waste slurry disposal scheme
Redundancy
Reheat amount/type
Degree of instrumentation
Mist eliminator configuration
Lime slaking completeness
Fugitive losses throughout
Process layout
Materials of construction
Makeup water distribution
Chemistry
pH gradient throughout
Sulfite to sulfate oxidation
Chloride balance
Liquid-to-gas ratio (L/G)
Point of fresh slurry addition
Scaling
Corrosion
Oxidation of SO2 to 803
NOX interference
Stoichiometry
Lime utilization
2.2-3
-------
Table 2.2-3.
TYPICAL ANALYSIS OF REPRESENTATIVE U.S COALS AS RECEIVED'
(values in-percent except as noted)
State and county
ALA., Jefferson
Walker
ARK., Franklin
COLO., El Paso
Las Animas
ILL., Franklin
Williamson
Sangamon
St. Clair
Peoria
Fulton
IND., Clay, Greene,
Vigo
Greene, Sullivan
Greene, Sullivan,
Gibson
Greene, Sullivan,
Rnox
IOWA, Appanoose,
Wayne
Marion
Monroe
Polk
Boone
KAN., Cherokee
Leavenworth
E. KY., Floyd,
Letcher, Pike
Perry, Breathitt,
Knott, Letcher
Harlan
Mining
district or
seam
Mary Lee
Mary Lee
Denning
Colo. Springs
Trinidad
Franklin
Williamson
Springfield
Belleville-Saunton
Peoria
Fulton
No. 3
No. 4
No. 5
No. 6
Mystic
Cherokee
Leavenworth
Elkhorn
Hazard No. 4
Harlan
Moisture
2.55
3.35
2.25
22.30
2.30
9.99
8.77
13.09
11.17
15.41
16.33
11.50
13.55
11.15
14.90
7.25
6.50
5.25
10.30
12.30
5.00
11.50
3.40
3.75
3.25
Volatile
matter
28.10
30.80
14.25
33.30
29.80
32.82
32.64
36.51
39.31
34.34
35.50
38.25
33.55
35.70
31.65
36.00
39.00
41.00
38.25
38.20
33.10
35.35
36.75
36.75
36.90
Fixed
carbon
58.40
52.80
74.00
38,25
58.70
49.27
51.41
41.14
39.20
38.52
37.01
40.45
45.40
42.65
46.15
47.50
46.75
46.25
39.65
43.80
52.90
39.95
55.85
55.30
55.95
Ash
10.95
13.05
9.50
6.15
9.20
7.92
7.18
9.26
10.32
11.73
11.16
9.80
7.50
10.50
7.30
9.25
7.75
7.50
11.80
5.70
9.00
13.20
4.00
4.20
3.90
Sulfur
1.00
0.70
1.90
0.40
0.50
1.03
1.10
3.77
4.22
2.97
2.89
4.55
0.94
4.18
2.20
3.75
5.00
5.25
5.00
4.75
4.65
4.20
0.75
0.70
0.85
Heating
value ,
Btu/lb
13,300
12,360
14,000'
8,625
13,780
11,857
12,177
10,935
11,223
10,422
10,220
11,550
11,740
11,370
11,325
11,500
10,200
11,750
10,500
10,500
12,930
10,900
14,000
13,755
13,960
Ni
•
to
(continued)
-------
Table 2.2-3. (continued).
State and county
W. KY., Union,
Webster, Hopkins,
Huhlenburg
HD., Allegany
HIGH., Saginaw
MO. , Adair
MONT., Carbon
Carbon
H. HEX, McKinley
Santa Fe
N.D. Host Middle
and Western
Counties
OHIO, Morgan,
Noble, Washington,
Harrison
Belmont
OKLA., Pittsburgh
PENN . , Luzerne t
Lackawanna
Dauphin, Schuyl-
kill, Carbon
Cambr ia
Cambria
Mining
district or
seam
Eastern Interior
Seam No. 9
Georges Creek
Saginaw
Bevier
Red Lodge
Bear Creek
San Juan
Cerillos
(General )
Mergs Creek
Pittsburgh No. 8
McAlester
Northern Coal Field
Southern Coal Field
Upper Killaning
Lower Killaning
Upper Freeport
Lower Freeport
Moisture
4.85
2.60
9.00
11.75
11.40
9.40
11.50
3.70
36.00
4.00
5.95
2.00
3.00
4.00
2.55
2.30
2.75
2.85
Volatile
matter
36.65
19.10
34.00
34.50
35.30
35.60
39.10
35.00
29.00
36.00
37.80
37.25
6.10
6.40
16.25
18.65
21.60
22.40
Fixed
carbon
49.50
71.35
53.20
40.70
42.80
45.60
42.60
49.50
28.00
48.50
46.80
56.25
82.00
80.50
71.90
72.45
67.40
67.05
Ash
9.00
6.95
3.80
13.05
10.50
9.40
6.80
11.80
7.00
11.50
9.45
4.50
8.90
9.10
9.30
6.60
8.25
7.70
Sulfur
3.30
1.20
1.05
4.80
1.70
2.40
0.70
1.00
0.65
4.25
4.20
0.75
0.70
0.90
' 2.10
1.44
1.45
1.65
Heating
value,
Btu/lb
12,490
14,135
12,750
11,150
9,900
10,700
11,300
12,800
6,600
12,250
12,055
13,500
13,000
12,800
13,865
14,400
13,930
13,960
10
•
to
I
en
(continued)
-------
Table 2.2-3. (continued).
State and county
Clearfield
Somerset
Westmoreland
Allegheny
TBNN. , Campbell
Bledsoe
TEXAS, Bowie S.W.
to La Salle
UTAH, Carbon
Summit
VA., Tazewell
Wise
MASH., Kittitas
Kittitas
Pierce
Pierce
W. VA., Monongalia,
Marion, Harrison
Fayette
Mercer
Kanawha, Fayette
Mingo
Mining
district or
seam
Lower Kittaning
Lower Freeport
Upper Kittaning
Lower Kittaning
Redstone
Upper Freeport
Jellico
Swanee
Lignite Fields
Castlegate
Hasatch
Pocahontas
Norton
Clealum (Cle Elum)
Roslyn
High Vol. Carbonado
Med. Vol. Carbonado
Fairmont
New River
Pocahontas
Kanawha
Thacker
Moisture
2.70
3.05
2.75
2.75
2.10
2.50
3.50
3.20
33.40
5.50
14.00
2.90
1.40
8.00
3.70
3.80
3.80
1.80
2.10
2.60
1.80
2.45
Volatile
matter
21.15
24.80
17.35
16.25
33.25
34.00
36.30
29.30
40.40
39.20
38.00
21.20
34.13
34.60
34.30
36.00
29.30
37.55
22.50
17.75
35.80
35.80
Fixed
carbon
67.85
65.20
71.00
73.00
53.55
54.50
52.90
59.70
17.20
47.80
43.00
71.50
58.47
44.70
48.60
51.20
49.90
54.15
72.20
75.00
55.70
56.40
Ash
8.30
6.95
8.90
8.00
11.10
9.00
7.30
7.80
9.00
7.50
5.00
4.40
6.00
12.70
13.40
9.00
17.00
6.50
3.20
4.65
6.70
5.35
Sulfur
1.85
1.60
1.40
1.70
2.45
2.25
1.60
0.85
1.10
0.60
1.40
0.55
0.82
0.45
0.30
0.50
0.50
2.20
0.65
0.65
0.90
0.95
Heating
value,
Btu/lb
13,940
14,025
13,810
13,990
13,140
13,400
13,630
13,500
7,600
12,500
10,700
14,550
14,250
11,410
12,250
13,400
11,500
13,850
14,860
14,635
13,500
14,100
to
ro
a\
-------
2. Almost all western subbituminous and lignitic coals
have sulfur contents of less than - 2 percent, and in
most the sulfur content is under 1 percent. These
western coals are easy to burn, with medium to high
volatility; in this respect they are similar to the
eastern and midwestern bituminous coals. Many western
coals, however, have other characteristics that are
markedly different from those of eastern fuels and
cause significant changes in operation. While there
is no typical western coal analysis, some of the major
features include the following:
Total moisture content often ranges from 25 to
over 30 percent.
Heating value is typically low, ranging from 6000
to 10,000 Btu/lb as received.
Ash content seldom exceeds 12 percent and usually
ranges from 5 to 8 percent. Sodium oxide in the
ash ranges from 0.5 to 8 percent or more. Potas-
sium and iron content in the ash are frequently
low, but calcium oxide content can exceed 25
percent.
Grindability is moderate, ranging from 40 to 70
Hardgrave Grindability Units.
3. The quantity of fuel required to sustain a given
output varies with heating value. Use of western
fuels generally requires more coal and may require
additional capability in the pulverizer and primary
air system.
4. Efficiency varies with moisture content and heating
value of the fuel and with the gas volume produced by
combustion. Use of western coal generally reduces
furnace/boiler efficiency.
5. Coals with higher moisture content and lower heating
value produce higher mass flow rates of flue gas,
leading to increased gas velocity and draft loss in
the convection passes. Requirements for fan power and
capacity are also higher.
6. Adjustment of the air heater may be required because
of moisture in the fuel. Primary air temperature may
be too low for coals with higher moisture content. In
addition, adjustment of the tempering air system may
be required for proper control of fuel systems with
variable moisture content.
2.2-7
-------
7. Additional facilities for handling and storage of fuel
and ash may be needed if tonnages are significantly
different from those now being used and generated.
All coals used as fuel in utility boilers contain sulfur in
amounts ranging from 0.5 percent to more than 5 percent. De-
pending on the heating value (Btu/lb) of the coal, this sulfur
content may produce SO2 in amounts ranging from 700 to 3500
parts per million (ppm) or from 0.8 to 8.5 Ib SO2/106 Btu heat
input. The current Federal New Source Performance Standards
(NSPS) allow a maximum of 1.2 Ib SO2/10G Btu of heat input.
Proposed changes in the NSPS may call for 85 percent SO2 removal
with a maximum allowable emission of 1.2 Ib SO2/106 Btu of heat
input and a maximum uncontrolled emission level of 0.2 Ib SO2/
106 Btu of heat input (i.e., boilers emitting more than the
amount would need FGD equipment).
To estimate the S02 emission (Ib SO2/10fi Btu), the follow-
ing equation may be used:
(2 x 104) x (coal wt.% sulfur) x (fractional conversion of sul-
fur in coal to S02)/(heating value in Btu/lb).
EXAMPLE: (2 x 104) x (3.5%) x (0.92 conversion)/(11,000 Btu/lb)
= 5.85 Ib S02/10G Btu.
If the conversion of sulfur to sulfur dioxide is unknown, use a
0.95 conversion factor for rough estimation.
Sulfur is normally present in the coal in three forms:
organic sulfur compounds, pyrites (primarily FeS2 ), and inor-
ganic sulfates. All the organic sulfur is liberated when the
coal is burned, but not all the inorganic and pyritic sulfur is
liberated. Some of it is removed as bottom ash or is included
in the fly ash. The inorganic sulfates may remain in the sul-
fate form; or the sulfates may be thermally decomposed to SO2
and oxygen. The main factors that affect the action of sulfates
are the type of sulfate (and hence its tendency to decompose)
and the effective temperature to which the sulfate is exposed
within the coal particle.
The coal must be tested to determine roughly what percen-
tage of the organic and inorganic sulfur is converted to sulfur
oxides, primarily SO2.
Typically 95 percent or more of sulfur in coal is converted
to SO2; about 0.5 to 1.0 percent may be converted to sulfur
trioxide (SO3), which reacts with water to form sulfuric acid
(H2SO4). If the flue gas is saturated with water and is then
cooled, H2SO4 tends to form a mist that is difficult to remove.
2.2-8
-------
In the material balance, it is assumed that all the sulfur
is oxidized to SO2 . The amount of SO2 liberated, in excess of
the emission regulation, must be reacted with lime slurry.
Assuming a 10 percent excess of lime in the absorbing slurry
(1.1 stoichiometric ratio), 0.96 Ib of lime (calcium oxide) is
needed to react with each pound of SO2.
2.2.1.2 Ash Content4'5—
Ash content of coal ranges from less than 4 to more than 17
percent. Some of the ash leaves the boiler with the flue gas as
fly ash, and some remains in the boiler and is removed as bottom
ash.
To determine the percentage of the ash that evolves as fly
ash one must test the coal. The amount that is converted to fly
ash is typically about 75 percent, although this may vary great-
ly with particular coals and with the type of boiler in which
the coal is fired. in cyclone-fired boilers, about 70 percent
of the ash is removed as bottom ash and 30 percent as fly ash;
in pulverized-coal-fired boilers, the proportions of bottom ash
and fly ash are reversed.
What is commonly termed coal ash has it origin in coal as
mineral matter that includes complex metal and organic sili-
cates, chlorides, carbonates, sulfides, sulfates, and phos-
phates. The principal elemental constituents are calcium,
aluminum, iron, silica, magnesium, sulfur, sodium, potassium,
and manganese. Most of the naturally occurring elements of the
periodic table also are present as minor and trace elements.
When coal is burned, the flame temperature generally ex-
ceeds 1650°C (3000°F); however, the expanding hot gas is rapidly
cooled by heat losses to the water walls and convection passes
of the boiler system. During combustion and subsequent cooling,
the mineral content of the coal is thermally decomposed, forming
fly ash, vapor, and slag. Fly ash is the gas-borne material
that flows through the boiler convection passes as discrete
particles. Vaporized mineral matter usually includes sodium and
potassium compounds, which can deposit on boiler tubes in the
lower temperature zones of the boiler system. Slag consists of
ash that is removed from wet-bottom or cyclone furnaces in a
molten state.
2.2.1.3 Fly Ash5'7'8'9—
Fly ash has four major impacts on FGD system design:
1. It can erode the piping and pumps.
2.2-9
-------
2. It can contribute to scaling or plugging within the
SO2 absorber and wet particulate scrubber, if one is
used.
3. If the fly ash has alkalinity value from its available
calcium oxide (CaO), sodium oxide (Na2O), and magne-
sium oxide (MgO) constituents, it may reduce the
requirements for lime makeup and fresh or recycled
water and thus reduce cost.
4. The chemically reactive part of the solids consists of
calcium sulfite and unreacted lime. When fly ash is
present in the slurry, it dilutes the reactive solids,
and a hold tank of larger volume may be required.
Before addressing each of these impacts, we consider brief-
ly the effect of the fly ash on particulate removal because this
largely determines the impact of particulates on the FGD system.
The fly ash carried in a flue gas stream can be removed by
an electrostatic precipitator (ESP), a fabric filter (baghouse),
and/or a wet scrubber. The choice of removal system is deter-
mined by the physical characteristics of the fly ash (especially
resistivity and particle size), the projected operating and
maintenance costs, and space or land constraints. Both ESP's
and baghouses can remove over 99 percent of the particulate
matter; venturi scrubbers can also remove over 99 percent if the
pressure drop is high enough,7
High fly ash resistivity limits the power input to an ESP
and hence the driving force for particle capture. Fly ash
resistivity is mainly a function of the sulfur content of the
coal, the gas temperature, and the chemical composition of the
fly ash. Precipitator manufacturers and others have developed
several indices to aid in design of precipitators for low-sulfur
coal applications, which generally produce fly ashes with high
resistivities.9
As an example of the impact of sulfur content on an ESP, a
precipitator operating at about 98 percent efficiency on a coal
with 2.5 percent sulfur can easily drop to below 90 percent
efficiency at 1 percent sulfur. The extent of efficiency de-
gradation is highly variable, depending largely on the plate
area of the installed precipitator and the plate rapping effi-
ciency. Performance degradation may be overcome by conditioning
the flue gas, operating at higher or lower temperatures, de-
rating the boiler, or installing more plate area.
Alternatively, operators sometimes remove both the parti-
culates and the S02 in the FGD absorber. In such cases, the
2.2-10
-------
presence of ash particulates requires a greater purge from the
absorber recirculation loop to maintain the total solids level.
This in turn reduces the concentration of calcium compounds in
total solids because the amount of calcium solids purged remains
constant. The fly ash concentration may accentuate scaling
tendencies.
As mentioned earlier, particulates that are not removed
prior to the wet scrubber or absorber can cause erosion, scal-
ing, and plugging and can reduce the demand for alkaline absor-
bent. Although erosion affects parts of the recirculating
slurry loop, scaling and plugging are usually accentuated in
areas of low flow.
Some types of fly ash offer a beneficial effect because of
their alkaline constituents. A typical chemical analysis of fly
ash from bituminous coal is as follows:
si°2 48% MgO 0.95%
A1203 21% Ti02 1.32%
Fe203 18% Na20 0.60%
Ca° 4% K20 1.40%
Some fly ashes, especially from lignite, contain high percen-
tages of CaO, MgO, and Na2O, and thus reduce the alkaline absor-
bent demand. The economic impact of these alkaline fly ashes
can be significant; a "typical" ash may have less than 6 percent
CaO, MgO, and Na2O total alkalinity, whereas lignite ash may
contain more than 27 percent CaO, MgO, and Na2O total alka-
linity. Examples of ash contituents are given in Tables 2.2-4
and 2.2-5.
2.2.1.4 Chloride Content—
The chloride content of coal is variable. As the coal is
burned, the chlorides are volatilized and carried out with the
flue gas stream. Experience has proved the need for considering
the corrosive properties of chlorides when selecting construc-
tion materials for scrubbers and absorbers. The use of 316L
stainless steel has sometimes been successful, but because the
300 series stainless steels are prone to chloride stress cor-
rosion cracking, they should be used in such environments only
with caution.10
The chloride level in the recirculating SO2 absorbent
slurry is controlled by the chloride content of the coal fired,
the allowable level of chlorides in discharges from the sludge
2.2-11
-------
to
to
I
Table 2.2-4. AVERAGE ASH CONSTITUENTS OF THREE RANKS OF COAL
BitusUnova states (average percent of ash constituents)
State
Alabama
Arkansas
Illinois
Indiana
•Iowa
Kansas
E. Kentucky
W. Kentucky
Maryland
Missouri
Ohio
Pennsylvania
Tennessee
Utah
Virginia
•.V. Virginia
S.N. Virginia
Ash
9.0
8.25
10.05
8.62
13.4
10.45
6.32
10.14
9.5
11.73
11.6
10.23
10.4
7.7
7.8
10.21
7.73
S
1.6
2.5
3.35
2.88
S.15
4.0
1.07
3.03
1.3
4.6
3.6
1.95
2.0
0.76
1.09
2.56
1.00
SiOj
43.7
24.8
45.5
46.9
34.3
38.2
46.2
47.86
51.65
42.2
31.6
45.43
47.7
51.4
45.6
41.20
50.86
A1203
26.4
19.75
19.1
22.78
13.95
16.35
27.5
23.05
30.35
15.8
22.9
27.55
36.32
15.1
27.8
26.11
30.89
Fe203
19.9
23.4
23.33
20.7
33.4
37.75
10.5
21.76
10.05
31.05
28.0
21.15
15.9
7.4
14.6
23.38
10.50
TiO,
1.18
0.95
0.95
1.08
0.85
0.65
1.43
1.20
1.4
0.7
1.0
1.05
1.19
0.96
1.34
1.16
1.52
P2°5
0.23
1.06
0.157
0.145
0.29
0.16
0.13
0.16
0.21
0.1
0.21
0.27
1.86
0.58
0.24
0.40
0.27
CaO
2.98
13.1
5.16
3.39
9.65
6.75
2.16
2.19
1.85
4.9
2.0
1.85
1.91
11.8
4.5
3.39
2.07
Hgo
1.29
4.9
0.89
0.88
1.25
0.55
1.04
0.92
0.65
0.65
0.69
0.55
1.25
3.3
1.5
0.85
0.81
Ka20
0.27
1.45
0.373
0.45
O.S
0.3
0.45
0.25
0.6
0.15
0.24
0.21
0.31
1.7
0.88
0.40
0.56
K20
2.36
1.25
1.62
2.43
1.2
1.0
1.86
2.37
2.55
2.1
1.5
1.95
2.68
0.6
2.1
1.62
1.74
so3
2.13
10.15
1.73
1.07
3.05
2.7
2.17
1.00
0.85
2.45
1.14
1.26
1.6
6.0
2.5
2.36
l.«7
Def.
temp . .
•P
2385
2175
2016
2214
1975
1970
2463
2034
2705
1978
2092
2377
2411
2166
2377
2331
2682
Soften
temp. ,
•r
2318
2270
2085
2325
2025
2025
2615
2352
2790
2028
2206
2456
2456
22SO
2485
2376
2638
Fluid
tenp. ,
•P
2490
2400
2290
2512
2165
220'1'
2710
2S01
2740
2295
2411
2S79
2610
2409
2623
2529
2737
Subbitusdnous states (average percent of ash constituents)
State
Montana
Hew Mexico
WyoeUng
Ash
12.6
10.53
10.4
S
0.59
1.13
1.2
SiOj
35.4
49.2
31. C
A1203
21.5
21.82
16.9
Fe203
5.31
13.76
9.7
TiO,
0.83
1.05
1.4
P2°5
0.41
0.06
0.36
CaO
13.46
6.38
20.1
MqO
4.63
2.0
4.6
Na2O
2.8
0.67
0.15
K20
0.67
0.58
0.55
so3
13.33
4.68
15.3
Oef .
temp . ,
•F
2355
2318
2450
Soften
temp. ,
•F
2435
2372
2510
fluid
temp. ,
•F
250S
2474
2630
Lignite state (average percent of ash constituents)
State
North
Dakota
Ash
11.8
S
0.98
Si02
26.3
A12°3
12.1
Fe203
6.85
TiO,
0.73
P2°5
0.21
CaO
21.1
MgO
6.4
Na,0
4.42
K20
0.33
SO,
20.6
Oef.
temp.,
•F
2180
Soften
temp . ,
•F
2237
Fluid
temp. ,
•F
2303
-------
Table 2.2-5. ASH ANALYSES OF SELECTED COALS6
Coal analysis
Btu/lb
Ash, %
Moisture, %
Sulfur, %
Ash analysis, %
Si02
A12°3
Ti02
Fe2°3
CaO
MgO
Na2°
K2°
so 3
Fouling
potential
Slagging
potential
Type of coal
Lignite
6500
10
40
0.8
28.4
11
0.4
14
18
5
7.0
0.7
19.8
High
Severe
Subbituminous
9086
10.6
27.2
1.0
34.2
15
0.8
12
18
4.5
0.3
0.3
17
Medium
High
Bituminous
10,290
18
10.4
5.1
44.6
18
0.6
18
5
1.2
1.35
1.9
5.0
High
Severe
2.2-13
-------
pond to adjacent bodies of water, and the use of cooling tower
blowdown. The intent of current water pollution legislation is
to enforce "zero discharge" from point sources; this means that
the utility FGD systems must operate in a "closed-loop" mode.
The result of closed-loop operation is that aqueous slurries
with very high chloride levels (5,000 to 10,000 ppm) could be
recirculated.*1 Saturation is not reached because of chloride
loss in the interstitial water in the calcium sulfate/sulfite
sludge.
In some FGD systems in Japan, the clear liquor purge rate
is controlled by the chloride content of the slurry, which may
necessitate excessive blowdown.
Although the effect of chlorides is the subject of ongoing
research, the design engineer must be aware of possible diffi-
culties caused by the chloride content of the coal, the impact
of "closed-loop" operation, and the chloride in cooling tower
blowdown.
2.2.1.5 Heating Value—
The heating value of various coals ranges from under 9000
Btu/lb to over 12,000 Btu/lb. Accordingly, to obtain a million
Btu of heat input in a boiler, the feed rate of coal may range
from under 83 Ib of coal with high heat content to over 111 Ib
of coal with low heat content. With the increased coal feed
rate, the resulting greater quantities of fly ash may call for
additional particulate collection capacity (retrofit), neces-
sitating additional capital expenditures and higher operating
and maintenance costs.
Use of coal with lower heat content also requires greater
capacity for coal handling, processing, and grinding if it
replaces a higher-heat-content coal for which the unit was
designed. Again, the capital, operating, and maintenance costs
rise.
2.2.1.6 Moisture Content—
In general, coals with a higher moisture content reduce
pulverizer capacity. An increase in moisture content also
entails the need for much warmer air and for more air to dry and
carry the fuel to the burners. Both surface and inherent mois-
ture are present in coal. Surface moisture is affected by
weather and by the methods of mining, fuel preparation, and
transportation (slurry pipelines); typical surface moisture is 8
to 10 percent of dry weight, with variations. Surface moisture
is independent of the type of coal. Inherent moisture is that
which is intimately associated with the coal in the particle
structure. Western coals, especially lignite and subbituminous
coal, are high in inherent moisture.
2.2-14
-------
Most utility and industrial combustion systems can handle
wet coal (up to 20% moisture content) on an intermittent basis.
Units designed to burn lignite and subbituminous coal must
handle total moisture contents as high as 40 percent. Signifi-
cant changes in moisture content will require modifications of
the air preheater system to provide higher air temperatures in
the pulverizer. Modifications of fuel drying systems are pos-
sible, but generally costly.
Since higher coal moisture contents necessitate the use of
more coal for the same net heat input, more air is required for
proper combustion. The greater air volume calls for larger SO2
absorbers and/or more absorbers to accommodate the additional
flue gas volume. Additionally, the greater the air volume, the
greater the makeup requirement for water. More water is ab-
sorbed as the air is adiabatically cooled in either the wet
particulate scrubber or the S02 absorber, even though the coal
supplies additional moisture. The increased water needed for
adiabatic cooling may be a major concern if freshwater supplies
are limited.
2.2.1.7 Combustion Variability—
As vegetable matter is transformed in stages from wood to
peat, lignite, subbituminous coal, bituminous coal, and anthra-
cite, its moisture content drops from 60 percent to 5 percent;
volatile matter decreases from 70 to less than 10 percent; and
fixed carbon increases from 20 to about 80 percent. Tables
2.2-6, 2.2-7, and 2.2-8 show some of the chemical and physical
properties of various coals.
The FGD system designer should be concerned with the varia-
bility of the sulfur, moisture, chloride, and heat contents of
the individual coals within a seam or from seam to seam. Table
2.2-9 shows the variability in sulfur content as sampled from
unit train coal deliveries. Note that the sulfur content can
vary by 50 percent from the long-term average if a single 3-h
sampling time is used to evaluate coal sulfur content. There-
fore, the designer should be aware not only of "average" values,
but also of the variability of these values over the averaging
time required by the applicable regulations.
2.2.2 Boiler Characteristics
2.2.2.1 Type of Boiler—
Three general types of coal combustion systems are used by
industries and utilities in this country: stokers, pulverized-
coal-fired units, and cyclone-fired units.
2.2-15
-------
Table 2.2-6. CONSTITUENTS AND PROPERTIES OF SELECTED COALS
State
Pa.
Pa.
Pa.
Ky.
Ohio
111.
Iowa
Colo.
Wyo.
N. Dak.
Analysis of Coals, as received
H2O
2.0
4.0
3.0
3.0
6.0
14
13.9
24
24
40
VKa
1.8
17
23.1
34.4
34.8
34.3
36.9
30.2
30
27.6
FCb
86.2
69
63.9
56.6
49.2
39.7
35.2
40.8
36
23.4
Ash
10
10
10
6.0
10
12
14
5
10
9.0
Sulfur
0.79
1.63
2.17
0.72
2.44
4.07
6.15
0.36
0.33
1.42
Btu
13070
13430
13600
13800
12450
10470
10244
9200
8450
6330
Coal
type
Anthracite
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Subbituminous
Subbituminous
Subbituminous
Lignite
Volatile material.
Fixed carbon.
2.2-16
-------
Table 2.2-7
COMPOSITION OF VARIOUS GRADES OF U.S. COALS"
(percent)
Fuel
classification
Mood
Peat
Lignite
Lignite
Subbituninoua C
Subbituninous B
SubbituBinous A
Bituminous high
volatile C
Bituminous high
volatile B
Bituminous high
volatile A
Bituminous
•ediun volatile
Bituminous
volatile
temiknthracite
Anthracite
Metaanthracite
Minnesota
N. Dakota
Texas
Wyoming
Wyoming
Wyoming
Colorado
Illinois
Pennsylvania
W. Virginia
H. Virginia
Arkansas
Pennsylvania
Rhode Island
Moisture
(as-received)
46.9
64.3
36.0
33.7
22.3
15.3
12.8
12.0
8.6
1.4
3.4
3.6
5.2
5.4
. 4.5
Volatile
matter
78.1
67.3
49.8
44.1
40.4
39.7
39.0
38.9
35.4
34.3
22.2
16.0
11.0
7.4
3.2
Fixed
Carbon
20.4
22.7
38.1
44.9
44.7
53.6
55.2
53.9
56.2
59.2
74.9
79.1
74.2
75.9
82.4
Ash
1.5
10.0
12.1
11.0
14.9
6.7
5.8
7.2
8.4
6.5
2.9
4.9
14.8
16.7
14.4
Sulfur
0.4
1.8
0.8
3.4
2.7
0.4
0.6
1.8
1.3
0.6
0.8
2.2
0.8
0.9
Hydrogen
6.0
5.3
4.0
4.6
4.1
5.2
5.2
5.0
4.8
5.2
4.9
4.8
3.4
2.6
0.5
Carbon
51.4
52.2
64.7
64.1
61.7
67.3
73.1
73.1
74.6
79.5
86.4
85.4
76.4
76.8
82.4
Nitrogen
0.1
1.8
1.9
1.2
1.3
1.9
0.9
1.5
1.5
1.4
1.6
1.5
0.5
0.8
0.1
Oxygen
41.0
30.3
15.5
18.3
14.6
16.2
14.6
12.6
8.9
6.1
3.6
2.6
2.7
2.3
1.7
to
«
NJ
I
-------
Table 2.2-8. CLASSIFICATION OF COAL BY RANK
&
Class
Anthracite
Bituminous
Subbituninous
Lignite
Group
1. Metaanthracite
2. Anthracite
3. Semianthracite0
1. Low volatile bituminous coal
Fixed Carbon
limits, «
(Dry, mineral-
matter-free
basis)
Equal to or
greater
than
98
92
86
78
2. Medium volatile bituminous coal 69
3. High volatile A bituminous coal
4. High volatile B bituminous coal
5. High volatile C bituminous coal
1. Subbituminous A coal
2. Subbituminous B coal
3. Subbituminous C coal
1. Lignite A
2. Lignite B
Less
than
98
92
86
78
69
Volatile matter
limits, %
(Dry, mineral-
matter-free
basis)
Greater
than
2
8
14
22
31
Equal to
or less
than
2
8
14
22
31
Calorific value
limits, Btu/lb
(Moist, b
mineral- matter-
free basis)
Equal to or
greater
than
14,000d
13,000d
11,500
10,500e
10,500
9,500
8,300
6,300
Less
than
14.000
13,000
11,500
11,500
10,500
9,500
8,300
6,300
Agglomerating
character
Nonagg loner at i ng
Commonly
agglomerating
Agglomerating
Nonagglomerating
N)
•
to
I
M
00
* This classification does not include a few coals, principally nonbanded varities, that have unusual physical and chemical
properties and that come within the limits of fixed carbon of calorific value of the high volatile bituminous and sub-
bituninous ranks. All of these coals either contain less than 4a% dry, mineral-matter-free fixed carbon or have more
than 15,500 moist, mineral-matter-free Btu/lb.
b Moist refers to the natural inherent moisture of the coal but not including visible water on its surface.
c If agglomerating, classify in low volatile group of the bituminous class.
d Coals having 59% or more fixed carbon on the dry, mineral-matter-free basis are classified according to fixed carbon,
regardless of calorific value.
8 There may be nonagglomerating varities in these groups of the bituminous class, and there are notable exceptions in
high volatile C bituminous group.
-------
Table 2.2-9.
COAL ANALYSES AND SULFUR VARIABILITY'
VARIOUS AVERAGING TIMES12
.OVER
Coal type
Eastern bituminous,
14% ash, 12,000
Btu/lb
Eastern bituminous,
14% ash, 12,000
Btu/lb
Western subbituminous
8% ash, 10,000
Btu/lb
Plant size,
MW
25
500
1000
25
500
1000
25
500
1000
Maximum average sulfur content, %
Long-term
7.00
7.00
7.00
3.50
3.50
3.50
0.80
0.80
. 0.80
Annual
7.36
7.23
7.22
3.68
3.62
3.61
0.84
0.83
0.83
30 days
8.27
7.79
7.75
4.13
3.89
3.87
0.96
0.90
0.89
1 day
9.36
8.88
8.78
4.68
4.44
4.39
1.12
1.05
1.03
3 h
9.73
9.23
9.19
4.86
4.61
4.59
1.18
1.10
1.09
ro
ISJ
I
Distribution from unit train sampling.
-------
In stokers, sized coal with a minimum of fines is burned on
or above a grate. Stoker designs include hand-fired units,
stationary grates, vibrating grates, spreader stokers, underfeed
stokers, and traveling grates. The large industrial stokers are
primarily traveling-grate and spreader units.
Stoker furnaces are limited in feed rate and generally are
used on units rated at less than 600 million Btu/h heat input.
Free-burning bituminous coal and lignite are commonly used.
Anthracite is generally unsatisfactory because it is a low-vola-
tility fuel and does not burn easily.
Pulverized-coal-fired units operate on the principle of
suspension burning. Coal is pulverized to about 200-mesh size
or finer and injected into the furnace pneumatically. These
furnaces are classified as dry-bottom or wet-bottom, depending
on whether the ash is removed in the solid or molten state. in
the modern direct-fired system, hot primary air is ducted to the
pulverizer, where the raw coal is dried and pulverized. The
mixture of hot air and pulverized coal is continuously conveyed
to the burners. The current maximum capacity of an individual
pulverized-coal burner is about 164 million Btu/h. Although a
unit may have as many as 70 burners, 16 to 30 is more common.
Cyclone-fired coal combustors burn coarse coal, approxi-
mately -4 mesh, in a horizontal cylinder into which part of the
combustion air is introduced tangentially to impart a whirling
or centrifugal motion to the coal. Combustion temperatures are
high, causing the ash to melt and adhere to the walls of the
cyclone furnace. The ash is removed as slag. Less than 2
percent of the utility coal combustion systems are cyclone
fired.
Various types of furnace/boiler configurations are avail-
able.
The type of boiler determines to a large extent the volume
of fly ash that the downstream units receive and to some extent
the volume of flue gas that the FGD systems must treat. Few
utility boilers use the stoker design; pulverized-coal-fired
boilers usually evolve more than twice as much fly ash from a
given coal as do the cyclone-fired boilers. This difference is
entirely due to boiler design characteristics and the size of
the coal particles.
Depending on the type and effectiveness of the particulate
removal device (ESP, baghouse, wet scrubber), FGD system desig-
ners may or may not need to be concerned about boiler type.
2.2-20
-------
2.2.2.2 Size of Boiler--
As the size of the boiler (measured by its power generating
capacity) increases, the volume of flue gas treated by an FGD
system also increases. Boiler size largely determines the size
and number of SO2 absorber units and the size of the whole FGD
system. A common conversion factor is 3000 acfm per MW of
generating capacity, assuming that air leakage is not excessive.
2.2.2.3 Age of Boiler/Air Leakage—
The FGD system must be designed to handle more gas in later
years. With respect to new FGD installations, the age of the
boiler has only one major effect on FGD system design: as the
boiler ages and leaks occur, the volume of flue gas the FGD unit
must handle will increase as a result of increased excess air
requirements or of increased air leakage. Estimates of the
expected increase in flue gas flow should be included in the
initial design, taking into consideration the design life expec-
tancy of the FGD system and the remaining life of the boiler.
Furthermore, an increase in particulate carryover usually accom-
panies an increase in gas volume.
2.2.2.4 Flue Gas Flow—
The primary factor in determining absorber size and cost is
the flue gas flow resulting from combustion of the coal. The
volatilized combustion products, NO , SO , chlorides, fly ash,
etc. , must be treated to remove the species that are generated
in excess of the allowable levels.
FGD system designers should obtain best estimates of all
the critical design parameters and consider the variability of
these parameters within the averaging time specified in the
emission permit. Provision for redundancy in critical units,
such as the SO2 absorber, should be viewed in light of current
and projected removal requirements and the projected increase in
flue gas and particulates resulting from increasing combustion
air requirements and air leakage as the boiler ages.
The design objective is installation of an FGD unit that
will remove the pollutants required while operating reliably, so
that boiler capacity reductions are minimized over the expected
life of the FGD unit.
2.2.2.5 Additional Control Equipment—
In addition to S02 removal, operations of utility boilers
must remove particulates to the mandated emission level. ESP's,
baghouses, and/or wet particulate scrubbers are used to remove a
major portion of the particulate loading. Improving particulate
removal, so that flue gas treated by the SO2 absorber is as
clean as possible, is often a goal of the FGD system designer.
When an ESP is used for particulate removal and upgrading of
2.2-21
-------
existing facilities is needed, the various alternatives include
conditioning the flue gas, reducing the temperature of the flue
gas stream, and installing additional plate area for particulate
collection.
Flue gas conditioning affects surface conductivity of the
particles by the injection of sulfuric acid, sulfamic acid,
sulfur trioxide, or sodium or iron salts to yield a more easily
collectible fly ash.
Reduced temperature sometimes improves ESP operation. The
gas temperature may be reduced by modifying the air preheater;
however, "cold end" corrosion due to condensation of sulfuric/
sulfurous acid might then occur in the preheater.
The most common method of upgrading precipitator perfor-
mance is to install additional ESP collection plate area. This
will permit adequate collection of the increased fly ash caused
by changes in coal ash or sulfur content, reduced capacity of
the existing ESP, or other factors.
The particulates not removed ahead of the FGD system can be
removed in a properly designed scrubber. Removing particulates
in the SO2 absorber can contaminate the lime slurry; however,
the particulates removed may act as a fixation aid in lime
sludge disposal. At some sites, fly ash that has been removed
ahead of the scrubber or the absorber is added to the sludge to
aid in fixation.
2.2.2.6 Loading Characteristics—
Boiler operation has a direct effect on the mode of opera-
tion of the FGD system. If the boiler handles peak loads, then
the FGD unit undergoes highly cyclical operation and the removal
cost per unit SO2 is higher than normal.
If the boiler handles base loads, the FGD unit can be
expected to operate at fairly steady-state conditions, which
would be ideal. Steady-state operation reduces the operating
and maintenance costs and allows for optimization of operating
practices.
2.2.3 Flue Gas
2.2.3.1 Temperature—
The inlet flue gas temperature determines the amount of
water that evaporates when the gas is cooled in either an SO2
absorber or a wet particulate scrubber. This has a major impact
on the overall FGD system water balance. The inlet gas tempera-
ture also affects decisions on whether the scrubber or the
absorber should be lined and what type of liner should be used.
2.2-22
-------
Selection of liner material is based on the ability to withstand
the flue gas temperature if water flow to the presaturator
section is interrupted. Instrumentation may reduce the proba-
bility of damage by activating a flue gas bypass, a backup water
deluge system, or a combination of both. The liner may be
protected if a bypass for the flue gas can be tripped either by
a temperature sensor set above the adiabatic gas temperature but
below the temperature of lining damage or by a flow sensor that
signals when the water supply is interrupted. As an alterna-
tive, a backup pumping system could be activated by a tempera-
ture excursion; the deluge effect might reduce the flue gas
temperature enough to prevent lining damage. In a combination
option, insufficient deluge (as indicated by the flue gas tem-
perature exiting the vessel), could trigger the bypass.
2.2.3.2 Flow--
The volume of flue gas to be handled per FGD train and the
desired gas velocity in the train largely determine the dimen-
sions of the particulate scrubber, if needed, and of the SO2
absorber. Furthermore, the availability of space can influence
the design. Peaking load of the boiler also affects the number
and size of FGD trains, depending on the turndown ratio of the
system.
2.2.3.3 Dew Point—
After particulates and S02 are removed, the flue gas stream
leaves the SO2 absorber through the mist eliminator for removal
of entrained liquid droplets, which contain both suspended and
dissolved solids. The gas stream is saturated with water vapor.
The stream passes through the ductwork at about 128°F. It may
be further cooled at this stage, causing condensation on the
equipment surfaces. The temperature at which condensation
begins is known as the "dew point." The condensate is a dilute
solution of sulfurous and sulfuric acid, which can cause major
corrosion in the ducting, stack, and equipment downstream from
the absorber.
Because corrosion has been widespread in FGD systems, flue
gas reheat systems have been installed in many units. Reheat
can be accomplished by: (1) bypassing a portion of the flue gas
stream (if S02 emission regulations allow), (2) installing
in-line, direct-fired systems or in-line, indirect steam or hot
water systems, or (3) heating outside air with direct or in-
direct systems for mixing with the cleaned flue gas stream.12
The subject is discussed further in Section 4.11 - Reheaters.
2.2.3.4 Particulate Loading—
The mode of firing the boiler determines the percentage of
the coal's total ash content that leaves with the flue gas as
fly ash. Fly ash in most FGD systems is removed by an ESP, a
2.2-23
-------
fabric filter, or a wet particulate scrubber. Removing parti-
culates prior to the SO2 absorption step reduces the solids
content of the absorber recycle stream.
2.2.3.5 Particulate Alkalinity—
Some coal fly ashes are alkaline, especially lignite ash
and occasionally subbituminous coal ash. A well-documented coal
source yielding alkaline ash is the Colstrip seam. Refer to
EPRI report entitled "Scrubbing Systems of Low Sulfur Alkaline
Ash Coals." The natural alkalinity of this particular fly ash,
which is approximately 20 percent CaO, reduces the amount of
lime that must be added to the absorbers at the Colstrip units
of Montana Power Co.14
By way of contrast, bituminous fly ashes normally contain
less than 5 percent alkalinity as CaO. Examples are presented
in Tables 2.2-4 and 2.2-5.
Since the alkalinity of the fly ash may yield definite
benefits by reducing absorbent costs, the designers should
investigate the properties of the fly ash and the potential
effects on the FGD system.
2.2.4 Lime Properties
Properties of lime are examined in detail in Sections 4.4
and 4.5, but their effects on system design are briefly dis-
cussed here.
2.2.4.1 Calcium and Magnesium Contents--
Lime is normally a product of calcination of limestone,
though some lime is produced by calcination of sea shells. It
consists primarily of the oxides of calcium and magnesium. On
the basis of their chemical analyses, limes may be classified
into three groups:
1. High-calcium quicklime, containing less than 5 percent
MgO.
2. Magnesian quicklime, containing 5 to 35 percent MgO.
3. Dolomitic quicklime, containing 35 to 40 percent MgO.
The characteristics of lime are discussed in much more detail in
Section 4.4. Work on lime-based S02 removal at Shawnee Station
of TVA and by Louisville Gas and Electric (LG&E), together with
other EPA research, indicates that 1000 to 4000 ppm of soluble
magnesium ions (Mg ) in the recirculating slurry can signifi-
cantly affect the ability of a lime absorbent slurry to remove
S02 from a flue gas stream. The magnesium ions (Mg ) increase
2.2-24
-------
the absorption capacity of liquid slurry and can depress sulfate
supersaturation.15 In a recent test at LG&E, SO2 removal in-
creased from about 83 percent wji^h calcitic lime to 95 percent
by addition of 3000 ppm of Mg to the lime slurry.15 When
soluble magnesium values exceeded 5000 ppm, only slight addi-
tional SO2 removal was observed. It should be noted that magne-
sium sulfite removes SO2 at least as well as calcium sulfite
(CaSO3) and that magnesium sulfite and bisulfite are much more
soluble in aqueous slurries than CaS03. This is important
because Mg reduces the tendency for scale formation. Addi-r
tional work in this area is reported by EPA, Kellogg, Dravo, and
Combustion Engineering.
A disadvantage is that Mg is more likely to be carried
over from the sludge pond in any aqueous discharge and may
pollute local waters.
R
Thiosorbic lime has received wide attention and is dis-
cussed in Section 4.4. Dravo has patented this magnesium-
promoted lime for use in increasing the SO2 removal capabilities
of FGD systems. The patent, however, is being challenged in
court. Naturally occurring magnesium has been cited as the
reason for reported increases in SO2 removal in several opera-
ting lime FGD systems. In successive runs in the same SO2
absorbers of a lime FGD system with constant liquid-to-gas (L/G)
ratio and identical lime addition rates, SO2 removal++effi-
ciencies reportedly were increased by the presence of Mg from
the 72 to 88 percent range to the 94 to 99 percent range on an
inlet flue gas stream of 3000 ppm of SO2.
The benefits of Mg in promoting S02 removal and in re-
ducing scaling tendencies are gaining acceptance in FGD tech-
nology and should be carefully considered in the design of a
lime FGD system.
2.2.4.2 Impurities/grits—
With high-quality chemical limes that have been thoroughly
calcined and have a loss-on-ignition of 1 to 1.5 percent or less
as CO2 , the total grit content that must be wasted will be only
1 to 2 percent of the weight of the lime. Grit losses, however,
may range up to 5 percent or more with lime of poor quality.
Included in the grit, as well as the carbonate core, are in-
soluble silicates and lesser amounts of aluminates, sulfates,
and ferrites, all of which are impurities occurring in the
limestone before it was calcined. When the grit is ejected from
the slaker, it resembles a mass of wet sand particles of a size
ranging from 1/4 in. to 100 mesh.
Degritting is performed to improve lime quality and to
reduce abrasion and wear on equipment. In extreme cases, cast-
iron centrifugal pumps have been worn out within a month when
pumping lime slurry that has not been degritted. With degrit-
ting, the same equipment can operate for 2 years or more.
2.2-25
-------
Degritting is performed in the dilution tank adjacent to
the slaking chamber. The slurry, or paste, is dispersed over a
weir into the dilution chamber and diluted by water sprays as it
passes. The heavier grit particles settle rapidly to the bottom
and are removed automatically by rakes that drag the grit up an
incline and out the chamber or to a classifier in the bottom of
the dilution chamber, where the grit is washed and a small
amount of slaked lime particles is recovered and mixed into the
diluted slurry. The washed grit is then disposed of manually or
automatically.
2.2.4.3 Reactivity—
This topic, covered in more detail in Section 4.5, Lime
Slurry Preparation, is summarized briefly here. The rate of
reaction of the lime when being slaked is a direct function of
its size, how well it was fired, and the percentage of grit.
The rate of reaction also controls the temperature of the lime
slaking and/or dilution equipment. It is most desirable to
complete slaking in 5 to I'O min while maintaining a slaker
temperature of approximately 200°F. Temperatures are higher if
the lime is more finely ground, because the higher surface area
of smaller particles leads to more rapid reaction. The slaking
time required for the size of lime particles selected should be
considered in design of the slaker and/or hold tanks in the lime
slurry preparation system.
The speed of slaking and the maximum temperature are in-
fluenced by how well the lime was calcined. The most important
test in determining optimum slaking is to measure reactivity of
the lime in water — specifically, how much the temperature
increases and in what length of time. In this test, a specific
weight of lime of a prescribed degree of fineness is added to a
specified volume of water at 77°F in a calorimeter; the tempera-
ture is measured at intermediate points and at completion of
hydration. This test is standardized by the American Water
Works Association (AWWA B202) and the ASTM in specification C
110 on Physical Tests of Lime.
Reactivity of the lime is classified on the basis of the
time needed to produce a temperature rise of 40°C (72°F), as
follows:
Time for
No. of minutes completion
High reactivity 3 or less 10 min or less
Intermediate reactivity 3 to 6 10 to 20 min
Low reactivity more than 6 20 min or more
Generally limes of high reactivity are soft-burned, i.e.,
calcined either at temperatures from 900° to 1000°C (1650° to
1850°F) or at temperatures from 1200° to 1300°C (2200° to 2400°F)
2.2-26
-------
for a short duration. The result is a reactive, porous lime of
lower density that slakes rapidly with a high temperature rise.
Limes of low reactivity are the converse, i.e., hard-burned,
denser, and heavier; they slake more slowly and evolve heat more
gradually, so that the temperature rise is appreciably less.
Dolomitic limes are inherently of low reactivity in varying
degrees, regardless of how they are calcined.
Slaking temperatures can be elevated artificially by using
more vigorous agitation and hot water for slaking, or by using
lime of finer particle size, such as pulverized lime. By such
measures the slaking rate may be increased so that a lime of
intermediate reactivity approximates the behavior of a highly
reactive lime. If these methods are applied to a highly reac-
tive lime, the slaking is extremely rapid, almost instantaneous,
so that the lime and water literally explode on contact. This
dangerous practice, potentially harmful to employees and equip-
ment, also produces slaked lime of poor quality. A complete
slaking time of 5 to 10 min is much more desirable. Conversely,
the efficiency of a high quality, reactive lime can be seriously
impaired by using too much cold water, especially with lime in
lumps or large pebbles that is inadequately mixed or agitated.
The resultant slurry may be coarse, fast settling, and incom-
pletely slaked. The slaking conditions, then, can enhance the
efficiency of a lime of mediocre (possibly poor) quality and can
impair the efficiency of a high-quality lime.
Two extreme conditions should be avoided:16
1. If the excess of slaking water is too great, and
particularly if the water is cold, an adverse reaction
called "drowning" occurs. The surface of the quick-
lime particle hydrates quickly, but the mass of hy-
drate formed impedes penetration of the water into the
center of the particle and delays rupture of the
particle into microparticles. The rise in temperature
is inhibited and slaking is delayed, causing coarser
hydrate particles and badly delayed or incomplete
hydration.
2. At the other extreme, adding insufficient water to the
lime, causes the hydrate to be "burned," because
temperatures are higher (250° to 500°F) than the
desired level, just below boiling. Too much water is
lost as steam, and unhydrated particles may remain.
The heat can be so intense that paint on the equipment
blisters or ignites and that dehydration of initially
hydrated lime particles occurs.
2.2-27
-------
The lime selected for an FGD system must be properly fired
and of good quality to ensure proper slaking. If the lime is
"overburned," then the surface area of the lime particles is not
reactive; and any reactive lime is trapped inside this encapsu-
lation, leading to a higher percentage of unreactive material or
grits. Possible loss of available lime should be considered in
design of a lime feeding system. Often systems are designed
with a feeding capacity from 10 to 20 percent greater than that
needed to allow for unreactive material.
The designer should know the average percentage of inerts
or insolubles and the range of these values for the grade of
lime selected; these variables affect the design of handling,
slaking, and storage equipment and the sizing of equipment for
removal and storage of inerts.
2.2.4.4 Size—
Lime is available in relatively standard sizes, as follows:
1. Lump lime - produced in vertical kilns in sizes rang-
ing from 8 in. diameter down to 2 to 3 in.
2. Crushed or pebble lime - the most common form, pro-
duced in most kiln types, ranging from about 2 to 1/4
in.
3. Granular lime - obtained from Fluo-solids kilns in
sizes ranging from 100 percent passing a No. 8 sieve
to 100 percent retained on a No. 80 sieve (a dustless
product).
4. Ground lime - obtained by grinding the larger-sized
materials or screening off the fines; typically almost
100 percent passes a No. 8 sieve and 40 to 60 percent
passes a No. 100 sieve.
5. Pulverized lime - obtained by more intense grinding
than that yeilding ground lime; nearly 100 percent
passes a No. 20 sieve and 85 to 95 percent passes a
No. 100 sieve.
6. Pelletized lime - made by compressing lime fines into
about 1-in. pellets or briquettes.
The need for these several sizes has evolved from the
process requirements of various systems. The most common lime
used in FGD systems is the crushed (pebble) size, which gives
good, controlled rates of reaction.
2.2-28
-------
2.2.5 Makeup Water
Water is lost from a scrubbing system in the form of water
vapor and entrained liquid particulate in the saturated flue
gas. It is also lost in disposal of sludge or gypsum byproduct.
(An EPRI report, "Lime/Limestone Scrubber Operation and Control
Study," gives further details.) The system uses water in
slaking, in dilution of lime slurry, in the mist eliminator
wash, in pump seals, and at other points.
The need for makeup water is directly proportional to such
uses in the system. Chemical composition of the makeup water is
important.
2.2.5.1 Chemical Composition and Variability16—
The quality of the water for slaking lime is more critical
than for any other scrubber use. Water should be of (or near)
potable quality. Waste or recycle process waters containing
sulfites and sulfates retard the slaking process and reduce the
quality of the resulting lime slurry. Poor slaking water causes
a larger average size of the slacked particles; the resulting
reduced surface area retards reactivity in a scrubber. In fact,
some of the lime does not hydrate and is wasted. It appears
that the lime precipitates the SO3 and SO4 ions as calcium
sulfite/sulfate, which coats the unreacted CaO particles and
prevents the complete penetration of water.16
Once the lime has been slaked, however, recycled or waste
process water can be used to dilute the thick lime slurry to the
desired consistency. The SO3 and SO4 ions produce little or no
effect on the quality of the diluted lime slurry.
In the S02 absorber (or the wet particulate scrubber, if
the flue gas passes through one prior to the S02 absorber),
water evaporates and the moisture content of the flue gas is
increased as the flue gas stream is cooled from about 300°F to
its saturation temperature of about 128°F.
The presence of Mg++ or sodium ions (Na++) in the S02
absorber makeup water is beneficial because the sulfite form of
both cations is much more water-soluble than the calcium sul-
fite/sulfate; the greater solubility should improve the SO2
removal and aid in reduction of scaling. The quality of water
required for slaking and dilution is discussed further in Sec-
tion 4.5.4.4.
The mist eliminator wash may be fresh, recycled, or waste
process water, or any combination of fresh and recycled water.
Although all fresh water would be ideal, closed-loop operation
often requires a mix. Continuous wash is often done with recy-
cled water, whereas high-volume, intermittent wash is done with
2.2-29
-------
fresh water. Since the trend in construction of mist elimina-
tors is toward the chevron baffle unit made of plastic or other
corrosion-resistant materials (e.g. Hastalloy), the chemical
makeup of the water normally has little effect except for forma-
tion of scale.17
Freshwater is recommended for pump seals because the heat
associated with pump operation could cause deposition of the
sulfate and other solids from recycled pond water and consequent
pump failure.
2.2.5.2 Source--
The following sources of makeup water are discussed below:
freshwater, sludge pond recycle water, waste process water,
cooling tower blowdown, and rainfall.
Freshwater may come from a river, a lake, a municipal water
system, or other source. For pump seals and mist eliminator
wash, freshwater is most desirable. Its use is limited, how-
ever, by concern for the system water balance and closed-loop
operation. Normally it is used in the most critical areas where
no other water is suitable, i.e., in lime slaking, in inter-
mittent mist eliminator wash,17 and as dilution for supernatant
liquid.
Sludge pond water is decanted from the calcium sulfite/
sulfate sludge either in a settling pond or in a filtration
step. This water is often a mixture of aqueous streams from all
points within the FGD system. The sulfite, magnesium, and
sodium ions in this water are ultimately beneficial to the SO2
absorption process. Undesirable ions include heavy metals
(because of water pollution concerns), sulfate (because it can
render the absorbent unreactive or increase scaling potential),
and chloride (because it can corrode materials). Sludge pond
recycle water is used wherever fresh water is not required.
Waste process water streams must be evaluated in terms of
availability, chemical composition, and variability of chemical
composition. If these streams contain undesirable chemical
constituents or if the reliability of stream supply is doubtful,
the design engineer may decide either not to use the stream or
to mix it with other streams. Water used in cooling bearings is
a common type of process water.
Cooling tower blowdown is the purge stream from units
designed to cool process streams. The concentration of dis-
solved and suspended solids is often the criterion for frequency
of blowdown. Whether blowdown should be used in the system and
how and where it should be used are determined by the kind of
cooling tower chemical treatment, if any, and by the possible
concentration of impurities in the feed water stream, such as
particulates that may be washed from the air by cooling tower
flow.
2.2-30
-------
2.2.6 Site Conditions
2.2.6.1 Land Availability—
Lime sludge disposal is a major concern in lime FGD systems
unless unlimited land is available. Section 2.4 gives a de-
tailed discussion of sludge disposal.
The solids content of untreated sludge is often in the 30
to 40 percent range, and its dewatering characteristics usually
are not good. The major constituents of the byproduct sludge
are calcium sulfate and calcium sulfite, although the fly ash
content may be significant.
Calcium sulfite crystals from lime systems are thin, frag-
ile platelets, usually occurring in clusters or rosettes, which
form an open structure with water filling the voids. The ro-
settes settle well but are difficult to dewater. In most cases,
calcium sulfite makes up 20 to 90 percent of the sulfur-contain-
ing solids, the remainder being calcium sulfate. The EPRI
report, "EPRI/Radian Particle Balance Concept Study," discusses
sulfite precipitates. Sludges with large amounts of calcium
sulfite generally are not suitable for landfill disposal without
additional treatment.18
When the sludge is thickened or filtered and subsequently
subjected to fixation, solids concentrations in the 70 to 80
percent range can be achieved. Thus the pond volume required to
contain untreated lime sludge can be many times as great as that
required for the product of sludge fixation. Fixed sludge may
KSKH - f°r landfill' whereas untreated sludge would
probably remain in the sludge state indefinitely.
Where available land for ponding is limited, sludge treat-
ment and/or fixation may be the only feasible methods of sludge
disposal. *
2.2.6.2 Soil Permeability—
The pollution potential of sludge liquor seeping into
groundwaters is governed by the mobility of the leaching waters;
mobility is limited by the coefficient of permeability of the
various media through which the leaching water must pass.
The permeation rate of leaching waters through the sludge
defines an upper limit to the amount of leachate entering the
subsoil. The amount of liquid and the degree of contamination
o± this liquid jointly determine the pollution potential of any
given waste disposal site.
Coal is the source of many contaminants, such as arsenic
(As), boron (B), lead (Pb), mercury (Hg), and selenium (Se),
which may condense on fly-ash particles or be scrubbed in the
cl 13 S O JT.D 6 3T •
2.2-31
-------
When major leachate products from untreated sludge were
compared with those of oxidized (gypsum) sludge, the calcium,
chloride, and magnesium levels were higher in leachate from the
untreated sludge; sulfate levels were comparable.
Permeability of the soil at sites proposed for storage of
lime sludges is a primary criterion in determining the effect of
possible leachates.
2.2.6.3 Ambient Humidity—
The average ambient humidity of the plant site affects the
FGD system from the standpoint of overall system water balance.
If ambient humidity is high enough to limit evaporation, and if
the lime sludge dewaters well, more water may be available for
recycle than is needed; periodic discharges into regulated
bodies of water may occur.
If the humidity is low, with ample evaporation, additional
freshwater is often needed. Abundant freshwater makeup may be
desirable for an FGD system if enough water is available.
With high average humidity, the combustion air requires
less water to reach saturation temperature, when it is cooled
adiabatically. This can strongly affect the water balance of
the system by reducing the makeup water requirement.
The major areas of impact of average humidity are large,
exposed volumes of process water (such as clarifiers, thicken-
ers, and the sludge pond) and the water balance, as affected by
the moisture in the combustion air. The effects of rainfall on
makeup requirements are related to the effects of ambient humid-
ity..
2.2.6.4 Rainfall--
The average annual rainfall can have a major impact on the
overall water balance of a utility FGD installation. Generally,
in an area where the rainfall is high, the evaporation rates are
low; and an excess of water may build up in the closed-loop
system. This might require periodic discharge of water (opening
the "closed loop") or forced evaporation.
Where rainfall is low and the evaporation rate is high, the
need for more makeup from outside sources should be expected.
2.2.6.5 Climate—
Climatic conditions should be considered in FGD system
design to anticipate problems that might affect operations
markedly. Average temperature, wind velocity, precipitation,
and other factors may influence the system-wide water balance
and other operating parameters.
2.2-32
-------
The major impact of climate will be on decisions relating
to enclosure of the unit and the insulation and/or heat tracing
of process lines. Critical equipment and control equipment
should always be protected.
2.2.7 Regulations
2.2.7.1 SO2 Emission Standards —
Under the current NSPS, emissions from large boilers are
limited to 1.2 Ib S02/106 Btu of heat input. A number of states
require an even lower maximum SO2 emission. Under proposed
revisions to the NSPS, SO2 emissions would be limited to a
maximum of 1.2 lb/106 Btu of heat input and uncontrolled SO2
emissions would be required to be reduced by 85 percent. The
percent reduction requirement would not apply if S02 emissions
o atro°sphere are less than 0.20 lb/106 Btu of heat in-
Currently lime and limestone systems constitute more than
90 percent of the operable FGD systems in the United States.
Approximately the same percentage of units under construction
are lime and limestone systems, and most units being planned by
utilities are of these types.
in Japan, about half the FGD installations use lime or
limestone as the SO2 absorbent. The SO2 removal capabilities of
these predominantly oil-fired units are mostly above 90 percent.
By the end of 1978, the revisions to the NSPS should be
complete and prospective FGD system operators will be better
able to determine the applicable Federal requirements.
2.2.7.2 Particulate Standards —
iK1?? 6current NSPS limit the outlet particulate emission to
0.1 lb/10 Btu of heat input. The proposed revision calls for a
maximum of 0.03 lb/106 Btu of heat input.
Complying with the proposed particulate standard may re.-
quire improved mist elimination because the entrained droplets
contain both suspended and dissolved solids. An unexpected
consequence of excessive entrainment can be that the particulate
level of effluent from the absorber may exceed that of the flue
gas stream entering the absorber. This is presently under study
by EPRI in a scrubber characterization project. Entrained
dissolved solids may contribute more to the total solids loading
than do the fly ash and other particulates.
2.2.7.3 Plume Visibility Standards —
In most states, statutes limit flue gas stream opacity to
20 percent (Ringelmann No. 1). The Ringelmann test is designed
2.2-33
-------
to reflect only particulate loading and not water vapor (the
measurement is taken after the water vapor disperses). Parti-
culate loading is controlled by the efficiency of the baghouse,
ESP, or wet scrubber prior to the SO2 absorber, by the ability
of the SO2 absorber to remove incoming particulates, and by the
ability of the mist eliminator to reduce carryover.
Where there is no particulate removal ahead of the SO2
absorber, its ability to remove particulates is critical, and
the unit must be designed to remove both SO2 and particulates.
Unless redundant absorber modules are available, there would
(where it is allowed) be no effective particulate control when
the absorber is partially or completely shut down.
In some states, such as West Virginia, the limit maybe as
low as 10 percent. The design engineer should be fully aware of
the particulate removal demands on the FGD system, as well as
the characteristics of the individual pieces of process equip-
ment, if plume visibility requirements are to be met.
An important restriction in a number of regions is that no
visible plume may extend beyond the property line. A very wet
plume, however, could well violate such an ordinance, and the
FGD system designer should be aware of this possible problem.
2.2.7.4 Water and Land Requirements—
Local, state, or Federal regulations relating to possible
water pollution or land use may have major impacts on the design
of an FGD system. In some high-density metropolitan areas, such
as the State of Massachusetts, even the onsite disposal of fly
ash is forbidden — ash must be removed to disposal sites in
nearby states. Disposal of S02 absorber sludge is also con-
trolled in other states. Land use regulations must be carefully
reviewed for possible constraints that may affect or dictate the
design criteria.
Recently, concern has grown about possible contamination of
substrata water by trace amounts of heavy metals, such as arsen-
ic, cadmium, and lead, that may be present in the aqueous over-
flow or leachate from sludge ponds. The leachate problem is
currently being investigated. In some cases, the use of pond
liners has been successfully tested.
Heavy metals are present in trace amounts in coal. They
are volatilized into the flue gas stream and may either condense
on the surface of particulates or be washed from the flue gas
stream in a particulate scrubber or an SO2 absorber.
2.2-34
-------
2.2.8 Absorber Type
The S02 absorbers used in lime FGD systems, described in
detail in Section 4.6, may be classified into five different
types: venturi, tray, packed bed, mobile bed, and spray.
The basic design of a venturi absorber is essentially the
same as that of a venturi particulate scrubber. The short
residence time for intimate contact of absorbing liquid with gas
and the cocurrent gas-liquid flow limit the mass transfer from
the gas to liquid. A large pressure drop increases gas/liquid
contact and therefore increases SO2 removal. Lengthening resi-
dence time to improve gas/liquid contact leads to higher power
requirements for flue gas fans and higher operating costs.
Chemico offered venturi absorbers in early lime FGD systems.15
Presently venturi SO2 absorbers are usually not recommended in
lime FGD applications.
A tray absorber consists of a vertical chamber with one or
more trays mounted transversely inside to provide multiple
countercurrent contact of gas and liquid. The gas velocities
should be limited to provide good gas/liquid contact but should
be well below flooding condition.21 At present, none of the
utility lime FGD systems uses a tray absorber, mainly because of
severe plugging and associated scaling caused by the excessively
long residence times of the reactant slurry in the scrubber.
Babcock and Wilcox have supplied several perforated tray ab-
sorbers for utility limestone FGD systems.
A packed absorber is a vertical column filled with packing,
in which there is continuous countercurrent contact of liquid
and gas. The maximum permissible liquid and gas rates are
determined by the flooding and liquid entrainment characteris-
tics of the packing.21 Packed absorbers have such good turndown
capability that little or no loss in SO2 removal efficiency
occurs at reduced gas loads. Packed absorbers are not widely
used in utility FGD systems because of plugging problems and a
high fire risk. Research-Cottrell has provided a packed tower
absorber at a utility limestone FGD installation.15
A mobile bed absorber provides a zone of mobile packing,
usually of plastic or glass spheres. In the countercurrent
operation of a mobile bed absorber, a highly turbulent action is
maintained at each mobile packing stage, ensuring both efficient
mass transfer and back mixing.22
Plastic spheres are more widely used than marbles (glass
spheres). Because of the lower mobility of marbles, the absor-
ber operation becomes less efficient and more susceptible to
plugging.
2.2-35
-------
Breakage of hollow plastic spheres has caused plugging in
some mobile bed absorbers. The use of solid foam or plastic
balls has eliminated the breakage and provided enough turbulence
to maintain the self-cleaning action inherent in Turbulent
Contract Absorbers (TCA). At the Conesville station of Columbus
and Southern Ohio Co., solid plastic spheres were used to re-
place hollow spheres in the mobile bed absorbers.23 Mobile bed
absorbers have been installed at several lime FGD facilities
(see Table 4.6-1).15 Combustion Engineering offers a marble bed
absorber in a lime FGD system. The American Air Filter Co. and
UOP (TCA) are leading vendors offering mobile bed absorbers with
plastic or foam spheres.15
There is a trend away from complicated absorbers to the
simplified types such as spray absorbers, which can be classi-
fied in three categories: countercurrent, crossflow, and cocur-
rent.24 The configuration of the tower may be vertical or
horizontal. In a spray absorber, the gas passes through ato-
mized, liquid absorbent droplets. The spray absorber has the
advantages of limited internal surface areas, reducing plugging
due to scaling. The "openess" of the tower reduces the pressure
drop and the energy requirement of the fan. Much of this ener-
gy, however, is incorporated into the slurry side to atomize the
slurry.
In a vertical spray absorber the gas/liquid flow is usually
countercurrent, with the gas entering at the bottom and exiting
at the top. The liquid slurry is introduced into the absorber
through spray nozzles located throughout the length of the
tower. Vertical spray tower absorbers have very low pressure
drops and good turndown.24 They can handle large jas volumes,
500,000 to 1,000,000 acfm in a single vessel.25 In currently
operating facilities, spray tower absorbers are very common in
limestone FGD systems, but not in lime FGD systems. The use of
spray tower absorbers is on the increase, however, in planned
lime FGD systems.15 Chemico, Combustion Engineering, Combustion
Equipment Associates, Pullman-Kellogg, and Peabody Engineering
are the leading vendors of spray absorbers.
Crossflow spray absorbers must be horizontal so that the
liquid is crosscurrent with respect to the gas flow. Crossflow
spray absorbers operate at lower pressure drop, and operation
costs may be lower than those of countercurrent spray absor-
bers.22 In addition to excellent turndown, the horizontal
configuration permits the vertical mist eliminator to be brought
in line with the gas flow without any of the complex ducting
required by vertical absorbers. However, in horizontal cross-
flow spray absorbers, the short contact time between the gas and
liquid spray must be compensated for by providing an ample
absorber volume.26
2.2-36
-------
Successful tests were carried out on a 160-MW prototype
unit of the horizontal crossflow spray absorber at Mohave Power
Station of Southern California Edison Co. Additional successful
tests were completed at the Four Corners station of Arizona
Public Service.15 At present no full-scale horizontal spray
absorber is in use at any lime FGD facility. One is being
constructed on the 825-MW Unit 3 of the Pennsylvania Power's
Bruce Mansfield station for the lime FGD system and is scheduled
for startup in April 1980. 15 The vendor offering the patented
horizontal crossflow (weir) absorber is Pullman-Kellogg.25
A cocurrent spray absorber with a vertical configuration is
in the development stage, as described in an EPRI report "Cocur-
rent Scrubber Evaluation TVA's Colbert Lime/Limestone Wet-Scrub-
bing Pilot Plant." EPRI is funding the evaluation of a 10-MW
cocurrent scrubber at the Shawnee Test facility in Paducah,
Kentucky . '
2.2.9 Waste Slurry Disposal
K Jhe byproduct generated by an FGD system depends on the
absorbent (lime), the flue gas characteristics, and the mode of
operation of the scrubber. The major constituents of the by-
products are calcium sulfate and calcium sulfite. The ratio of
sulfite to sulfate depends on the extent of oxidation, which is
in turn mainly a function of slurry liquid composition and 'the
4°ntent °f flue gas* This is discussed more fully
In general, byproduct sludges may contain less than 5 or as
much as 10 percent fly ash, depending on the amount of fly ash
removed from the flue gas prior to the scrubbing process.
... If the Primary constituent of the sludge is calcium sul-
tite, generally in the form of thin platelets, it is extremely
difficult to dewater, as explained in Section 2.2.4.18 More
information concerning sludge characteristics is given in an
EPRI report "Full Scale Scrubber Sludge Characterization
btudies. Solids content of sulfite scrubber sludge ranges from
Ib to 40 percent upon settling. Thickened and chemically fixed
sludge contains 40 to 70 percent solids. The solids content of
untreated calcium sulfite sludge is low because water is trapped
in the sludge by the platelet structure of the sulfite crystals.
This type of sludge affects FGD system design because more land
disposal area and makeup water are required. Except where
makeup water is a scarce commodity, a high water makeup rate
offers a side benefit in that it allows the use of more fresh-
water in the mist eliminator wash and other areas of FGD system.
2.2-37
-------
Increased oxidation of the sulfite portion of the sludge
can be accomplished by bubbling air into the solution in the
reaction tank at specified pH levels.27 Work by researchers at
the TVA's Shawnee plant has demonstrated almost complete conver-
sion of the sulfite to sulfate. The rod-shaped sulfate crystals
dewater much more readily.18 The effects of this on FGD system
design is that less land disposal area is needed and the sludge
may be used as a landfill. Less water would leave the system as
interstitial blowdown, which may increase concentration of
dissolved solids in the system.
The thickening ability of a sludge may be increased by
adding a settling aid. In a gravity thickener, a settling aid
(such as about 30 ppm of acrylamide copolymer) may increase
solids concentration in the thickener underflow and reduce
settling time. Depending on the sulfite/sulfate ratio and
crystal size, thickening may yield a slurry of 25 to 50 percent
solids. Vacuum filtration may yield a 45 to 70 percent solids
material. Again, the impact on FGD system design is to reduce
ponding requirements, improve the likelihood of this sludge
being used as landfill, and reduce possible water pollution
problems.
Chemical fixation of the thickened sludge can be accom-
plished by one of several proprietary processes, such as those
of IU Conversion Systems, Dravo, or Chemifix.18 The primary
criterion for fixation is the intimate mixing required for
proper disposal of the stabilizing agent and complete wetting of
fly ash. The resulting fixed sludge may have from 60 to 80
percent solids. Often this sludge is of landfill structural
quality. The effects of FGD system design include higher cost
for sludge handling and fixation, minimum possibility of water
pollution problems, the need for only minimum sludge pond area
because of the landfill properties of the sludge, and minimum
impact on the system water balance.
2.2.10 Redundancy
A designer's attitudes toward redundancy are influenced by
management policies regarding expenditures for installed spare
capacity and by the regulatory constraints on FGD system opera-
tion.
Operating an FGD system properly and reliably while accom-
plishing the mandated SO2 removal may well entail the installa-
tion of spares for the major process equipment, such as critical
pumps, absorbers, wet scrubbers (if any), and slakers. The
degree of redundancy is an engineering decision. The capital
expenditure required for a spare absorber module can be signifi-
cant. The number of spares that may be needed depends on the
2.2-38
-------
number of operating units, the reliability of each unit, and the
impact on overall operation if the unit is not online (e.g., if
an absorbent to other stages may be able to compensate). Fur-
thermore, allowance for in-process storage or surge tanks may
permit shutdown of some pieces of major process equipment for
maintenance or repairs.
A recent draft of the proposed NSPS for coal-fired utility
boilers recommends that exemptions for malfunctions not be
allowed under the proposed SO2 standards. Further, it concludes
that spare FGD modules should be installed and that when a
malfunction occurs where emission requirements cannot be met,
the operator should derate the boiler or shut it down. The
power gap would be closed by increasing the load on other gen-
erating units within the system or by the purchase of outside
power.
The FGD system designer should be aware of how to minimize
redundancy, but should consider both the legal and financial
ramifications of the need for redundancy.
2.2.11 Method of Reheat
There are basically six alternatives with respect to flue
gas reheating: no reheat, in-line reheat, direct-fired reheat,
indirect hot air reheat, flue gas bypass, and exit gas recycle
reheat. Much of this information on reheat is detailed in
Section 4.11. The reasons for reheat are to prevent corrosion
caused by condensation, to prevent a visible plume, and to
attain the desired plume rise; however, corrosion of the reheat
equipment is a significant problem. More information is given
in an EPRI report, "Stack Gas Reheat for Wet Flue Gas Desulfuri-
zation Systems."
Although the no-reheat option may seem to exert zero impact
on daily operating costs, there is a high probability that the
saturated flue gas stream will reach its dew point, with conden-
sation on material surfaces, and that the condensate will absorb
SO2 to form a corrosive, acidic liquid downstream from the SO2
absorber. Therefore, the design and cost implications are that
the ducting, the stack, and fan must be as corrosion-proof as
possible; that equipment will need intensive maintenance and
earlier replacement; and that daily operating cost will be lower
because no reheat cost is incurred. Design of the downstream
equipment for corrosion prevention/reduction is critical with
the no-reheat option.
In-line reheat often involves the use of steam inside banks
of tubes designed for maximum heat transfer to impart the tem-
perature increase required to prevent condensation. High-pres-
2.2-39
-------
sure hot water (250° to 350°F at 15 to 120 psig) may also be
used. The design considerations are all aimed at corrosion
resistance on the tube surfaces. Carryover of liquid from the
mist eliminator may lead to solids deposition on the tubes as
the liquid dries, and corrosion may take place beneath these
deposits. In addition to use of corrosion-resistant materials
of construction, soot blowers are sometimes used to reduce
solids deposits.
With in-line steam reheat, operators of some FGD systems
report better corrosion protection by leaving the steam on at
all times. Again, materials of construction are chosen on the
basis of projected flue gas stream conditions and chemical
composition. Decisions regarding materials of construction and
mode of operation (on/off or "always" on) are economic ones.
The economics are affected by whether the steam used for reheat
causes derating of the turbine or whether it comes from a separ-
ate steam boiler.
With direct-fired reheat, oil or gas is burned in the flue
gas stream. The cost of fuel needed for this practice is often
significantly higher than that of steam reheat, especially if
the steam requirement exceeds boiler capacity. Again, corrosion
of the exposed metal surface is of concern, although the exposed
area is much less. Air leakage greatly accelerates corrosion
rates; operators of some FGD systems believe that prevention of
leaks and rapid repair are critical to maintaining the reheat
equipment.
Because of corrosion of reheat equipment in the flue gas
stream, some designers have opted for a more expensive but more
reliable means of achieving the desired temperature increase.
In indirect hot air reheat a fan brings in outside air, which is
heated by direct or indirect means (gas/oil or steam) and is
then mixed with the flue gas stream. The volume of gas to the
stack is increased by the hot air volume, and the cost is
higher; however, maintenance problems and the possibility of
emergency shutdown because of reheat failure are greatly re-
duced.
Under the present NSPS, some systems can bypass a portion
of the flue gas stream from the S02 absorption train. This
stream can be used to reheat the portion of the flue gas stream
that has been cleaned. If this practice provides compliance
with the applicable emission standards, it is the least expen-
sive method of reheat. The draft of the proposed NSPS calls for
85 percent removal of SO2 over a 30-day period, with a maximum
allowable emission of 1.2 Ib SO2/106 Btu heat input. If the
emission level is as low as 0.2 Ib SO2/106 Btu, the percent
removal requirement will not apply.14 Should this proposal
become law, use of the bypass reheat option will be extremely
difficult.
2.2-40
-------
Reheat by recycling the exit gas involves removing cleaned,
heated flue gas from the stream before it enters the stack and
further heating this side stream to the temperature needed for
recycling to the main stream at the absorber outlet, where it
reheats the whole stream. in contrast to indirect hot air
reheat, the total gas flow is not increased and reheat is less
influenced by ambient air conditions. This mode of reheat has
not been installed on a utility boiler.
These reheat methods raise the temperature of the flue gas
stream in most systems from the adiabatic saturation tempera-
ture, about 125°F, to approximately 175°F. When heated gas
streams are mixed with the flue gas stream, air or recycled flue
gas must be at about 400°F to achieve reheat.
To prevent a visible plume, reheat of the flue gas usually
depends on the ambient temperature and relative humidity.
2-2.12 Degree of Instrumentation
The principal concern in instrumentation is pH control of
the absorbent slurry to achieve maximum S02 removal. To this
end, feedback instrument loops were initially used in many
systems; however, because of scaling of the probes, lime block-
age, breakage of probes, and many other problems, many operators
have resorted to manual grab samples for control of lime feed
and S02 emissions. AS a result, lime usage often is well over
tnat required, and SO2 emissions are higher than those provided
in design.
As operators gain understanding of scaling tendencies and
learn where pH probes may be placed for best control and main-
tenance, they are returning to automated control.
Sophisticated pH systems are needed to comply with increa-
singly strict S02 emission regulations. As a side benefit, lime
usage and overall control are improving. An advanced pH system
utilizes the feed forward/feedback concept. An alternative is
to measure the SO2 concentration directly, although instrument
problems have arisen in the severe service. The flue gas stream
can be monitored at the inlet for SO2 concentration and flow
volume and at the absorber exit to determine SO, removal.
A method being tried for control of the absorbent recycle
volume is to regulate pump output as the system gas flow load
fluctuates.
2.2.13 Mist Eliminator Configuration17
Detailed information concerning mist elimination is given
in Section 4.7.
2.2-41
-------
Mist eliminators are designed to remove entrained water
droplets, which carry both suspended and dissolved solids.
Efficient removal of the droplets reduces problems with parti-
culate emission and plume opacity; it also reduces the proba-
bility of corrosion associated with aqueous acidic solutions and
the likelihood of scaling caused by deposition of solids on the
metal surfaces of such parts as reheaters, fans, and ducts.
The variables in mist eliminator design include horizontal
or vertical configuration, the use of bulk separation and knock-
out devices, the number of passes, the number of stages, the
distance between stages, and the type of mist eliminator (con-
tinuous or discontinuous, chevron, or radial).
Horizontal mist eliminators have several disadvantages.
The flue gas stream flows upward through the apparatus, and the
droplets collected by the mist eliminator fall back through the
gas flow, increasing the possibility of reentrainment. They
also remain on the blade surface longer because of the action of
the gas flow, increasing the probability of scaling or plugging.
Difficulties also occur in the mist eliminator wash because all
wash falls back into the absorber and because wash cannot be
done longitudinally, which is the most effective way.
In vertical mist eliminators the gas flow is horizontal.
Hook sections are used on the baffles to prevent reentrainment,
and droplets are collected in a tray beneath the vanes. These
units may be operated at higher gas velocities than the hori-
zontal type without reentrainment. The vertical type is the
most widely used in Japanese FGD installations.
Bulk entrainment separators and knockout devices are de-
signed to remove large liquid droplets before the gas stream
enters the mist eliminator. Bulk entrainment separators often
take the form of an expansion area in which velocity is reduced
and the droplets are removed by gravity. Some knockout devices
permit the collection of wash water for reuse.
The number of passes indicates the number of direction
changes the gas stream must make before it exits the mist elimi-
nator. Often, the greater the number of passes, the greater the
efficiency; however, because of possible plugging, three passes
are most common in lime FGD units and provide good collection
efficiency with adequate washability.
Both one- and two-stage mist eliminators are common. A
single-stage unit often operates in conjunction with a bulk
entrainment separator. Two-stage units allow the use of more
wash water. The second stage is unwashed and collects mist from
the washing of the first stage as well as from normal entrain-
ment.
2.2-42
-------
The chevron baffle comes with continuous or discontinuous
baffles. Entrained liquid is forced to make abrupt changes in
direction, causing inertial impaction on baffle walls. Both
sharp and smooth vane bends are used, but sharp bends predomi-
nate. Continuous chevron design is often selected because it
provides greater strength at lower cost. Pressure drop is not a
consideration for either chevron design.17
The radial vane mist eliminator is a cyclonic separator
whose curved vanes redirect the gas stream from the vertical
path to the horizontal path, which is aimed at the vessel wall.
Heavier liquid droplets and solid particles are attracted toward
the vessel wall, where they impact and are collected. The
pressure drop of radial vane units is much greater than that of
the chevron type. For mist elimination, however, radial vane
units have not operated as well as chevron units.
Mist eliminators may be constructed of 316L stainless steel
(strong, rigid, corrosion-resistant, not temperature-sensitive,
but heavy and more expensive than other materials), fiberglass
reinforced polyester (FRP) (corrosion-resistant and light, but
pressure- and temperature-sensitive, may become brittle), or
Noryl thermoplastics (inexpensive and light, but temperature-
sensitive). Despite its drawbacks, FRP is the predominant
construction material for mist eliminators.
2.2.14 Losses Throughout the System
An often-overlooked design objective is to minimize process
losses. Concentration on this goal begins with lime delivery
and continues through slaking, slurry dilution, operation of the
reaction/hold tank beneath the SO2 absorber, and sludge handling
(until the sludge is delivered to the pond). This effort also
applies to minimizing water losses from the sludge pond.
As the lime is delivered, the air from pneumatic unloading
entrains lime as it exits the storage silo. An excellent way of
recovering this lime and reducing pollution potential is to have
two cyclone separators in series or a baghouse for particulate
removal. This arrangement can often remove over 99 percent of
the airborne solids. The recovered lime is returned to the
silo.
The lime delivery conveyor belt or screw conveyor system
should be covered, tightly sealed, and maintained at a slightly
negative pressure to avoid losses. The entrained lime is re-
covered and discharged into the lime storage silo.
The area around the slaker should have washdown facilities
from which the water goes into a collection sump, where large
2.2-43
-------
particles can be separated and the slurry can be recycled for
dilution. Washdown stations and slurry collection pumps also
are desirable around the slurry dilution tanks, the absorber
reaction or hold tanks, and major process pump areas.
2.2.15 Process Layout
Good process layout calls for a balance between minimizing
the distance between the major process operators and allowing
adequate space in the ground and overhead for process main-
tenance and instrumentation.
It is desirable to minimize the distance between lime
storage, lime slaking, and lime slurry preparation areas. The
distance from the makeup slurry tank to the reaction/hold tank
beneath the absorber is not critical. The distance from the SO2'
absorber and reaction/hold tank to the purge thickener should be
minimal, but it is not critical.
Although each process supplier has his own preference
regarding location of equipment, it is generally desirable to
locate the reaction/hold tank directly below the SO2 absorber.
Individual pumps supply slurry to each header system. Mist
eliminators are most often of the horizontal type, directly atop
the SO2 absorber.
2.2.16 Materials of Construction
Materials of construction and the mechanics of corrosion
attack are discussed in detail in Section 4.12.
The most common forms of corrosion in FGD systems are
crevice corrosion, intergranular corrosion, and erosion-corro-
sion. Early experience with FGD systems taught designers that
the particulate scrubber performs best if it is lined, since the
mixture of chlorides, sulfuric acid, sulfurous acid, and parti-
culates attacks any unprotected steel surface. The absorber can
be constructed of carbon steel with a liner or of alloy steel
without a liner. The debate over which is superior continues.
When a wet particulate scrubber is part of the FGD system
design, the materials of construction are a carbon steel for the
shell and 316 or 316L stainless steel for the venturi throat.
The internal liner, which completely covers the carbon steel
shell, may be of natural or neoprene rubber, flaked glass (such
as Heil 490 or Ceilcote 103 or 151), or Precrete. Natural and
neoprene rubber are the most popular liners, usually with excel-
lent operational records.
2.2-44
-------
The SO2 absorber should be constructed from corrosion-
resistant materials or carbon steel that is coated with plastic
material such as flaked glass or is rubber lined. Most common
are 316L stainless steel and neoprene-lined carbon steel absor-
bers. Hastelloy G, Inconel 625, or 316L stainless steel are
used for some exposed internal parts.
The mist eliminator should be made of FRP or stainless
steel (316L) for corrosion resistance.
The reheat equipment may be constructed of anything from
carbon steel to exotic alloy steel, depending on the system
supplier's policies, the flue gas composition, the expected
volume and composition of carryover from the mist eliminator,
and other factors.
The fans may be constructed of anything from carbon steel
to Inconel, depending on whether the fan is operated dry or wet,
the expected chemical composition of the gas and/or liquid, and
the system supplier's preference.
Pumps and recirculation and transfer piping may be made of
rubber-lined (natural or neoprene) carbon steel or alloy steel.
Lime slakers are primarily constructed of carbon steel.
Tanks may be carbon steel, FRP, rubber-lined carbon steel,
or 316 or 316L stainless steel.
Selection of a suitable stack lining material is still in
question, partly because everything used to date has been less
than successful. Recent hypotheses blame fluorides for the
failure of some plastic coatings. Sprayed-on coatings often
show pinholes where corrosion starts; the corrosion accelerates
until the whole lining fails or comes off in sheets or large
pieces.
2.2.17 Chemistry
2.2.17.1 pH Gradient--
As the lime slurry enters the SO2 absorber, the pH is often
in the 7.5 to 8.5 range. When the absorbent reacts with the
SO
2 '
the pH drops as the slurry becomes more acidic. The pH
may
be in the 4.5 to 6 range as the slurry leaves the absorber.
Johnstone reported in 1935 that the logarithm of the equi-
librium vapor pressure of SO2 over lime solution was inversely
proportional to the pH of the solution. This results in lower
SCS equilibrium vapor pressure at higher (more alkaline) pH.
Test work recently conducted at the Shawnee test facility of TVA
2.2-45
-------
showed higher SO2 removals at higher pH with constant L/G
ratios. The disadvantage of increasing the pH is that the
excess lime required for this mode of operation increases the
cost of operation and the tendency for scale formation.13
2.2.17.2 Sulfite to Sulfate Oxidation—
The calcium sulfite in the absorbent stream may react with
free oxygen and form calcium sulfate, which has no SO2 absorbing
value. The oxygen may result from high excess air used in
firing the coal in the power boiler, or it may enter as air
leakage (in the preheater or holes in ducting).
In FGD sytems design, the excess air for fuel firing should
be reduced to the lowest safe level; in system operation, any
holes that appear in the flue gas ducting should be repaired.
The sulfite purge from the absorber can be oxidized in a
separate reaction tank to convert the sulfite to sulfate, pro-
ducing a crystalline structure that may be dewatered much more
easily. This is done either in a separate tank or in a separate
scrubbing stage where low pH can be maintained to encourage
oxidation.
2.2.17.3 Chloride Balance—
The chloride level in the operating mode practiced at most
FGD systems is maintained through losses of chlorides in the
interstitial water in the calcium sulfite/sulfate sludge.
Theoretically, the chloride steady-state level in FGD system
operations may reach 30,000 ppm; reported values rarely exceed
6,000 ppm.
Only one U.S. unit controls absorber purge on the basis of
chloride content, whereas this is a common practice at many
Japanese units because of the adverse effect of chloride in
wallboard production. The major concern with chlorides is
corrosion of the wet particulate scrubber, and of exposed steel
surfaces on the absorber. The damage resulting from chloride
attack is well documented in U.S. FGD experience.
The chlorides originate in the coal and are volatilized as
the coal is burned. They are removed when they contact aqueous
streams. Chlorides also decrease the effect of magnesium and
should therefore be minimized when magnesium is added purposely
to a scrubbing system. Notwithstanding these factors, a calcium
chloride system has been used for S02 removal.
2.2.17.4 Liquid-to-gas Ratio—
The L/G ratio expresses the lime slurry flow (gallons) in
the absorber per 1000 acfm of flue gas flow at absorber con-
ditions. Increasing the L/G ratio enhances liquid mass transfer
2.2-46
-------
at the interface because of increased turbulence and better gas
distribution, which increases the mass transfer coefficient. A
high L/G ratio also provides more driving force because of the
lower sulfite buildup in the slurry per pass.
The L/G ratio also affects residence time of the gas stream
in the absorber, which in U.S. FGD units ranges from 3 to 9 s.
In mobile bed absorbers the mobile packing has a major
impact on the liquid/gas interface; at the Shawnee test unit of
TVA, mobile bed packing had more of an effect on SO2 removal
than did gas velocity or absorber residence time. With a mobile
bed absorber, the L/G approaches 80 gal overall. L/G ratios in
normal mobile bed absorber operation range from 30 to 50. This
range is usually well below flooding conditions in a mobile bed
absorber. (A flood condition occurs when enough slurry is held
up in the scrubber to prohibit the passage of flue gas through
the absorber without ensuring a pressure drop considerably
higher than designed.)
Because there is no flooding problem in spray towers, the
power required for pump operation may be a constraint. Liquid-
to-gas flows of 60 to 100 gal/1000 acfm have been used, although
normal operating values are from 60 to 80.
2.2.17.5 Point of Fresh Slurry Addition—
In the past, the diluted fresh lime slurry was added at
several places in the absorbent recirculation loop. The two
major points of addition were in the piping recirculating to the
absorber and in the reaction/hold tank beneath the absorber.
Because chemical reactions occurred in the piping, on the top
tray, in the first rows of balls or marbles in a mobile bed
absorber, and in other locations, the addition of fresh makeup
slurry in the recirculation piping has been discontinued.
All the major lime FGD system suppliers now add the makeup
lime slurry to the reaction/hold tank. Care must be taken in
slurry addition, since this can directly affect the precipitated
particle size. An EPRI report, "EPRI/Radian Particle Balance
Concept Study," discusses in detail the proper point of lime
feed addition.
2.2.17.6 Scaling—
Early lime FGD systems were generally operated without
attention to the basic methods of scale-prevention: high L/G,
high crystal content in the slurry, and adequate retention time
outside the scrubber. After discovery of massive deposits of
hard sulfate scale and also some deposits of softer sulfite
scale, steps were taken to control scaling. Deposits also occur
at the wet/dry interface within the wet particulate scrubber or
the S02 absorbing vessel. These deposits are mixtures of fly
ash, sludge, and sometimes soluble salts.
2.2-47
-------
Scaling occurs by three basic mechanisms: (1) by nucle-
ation on equipment surfaces with subsequent growth transforma-
tion of soft deposits (sulfite) to gypsum (sulfate) scale, (2)
by pH excursions into the lower ranges, and (3) by physical
drying.
Precipitation of calcium sulfate normally occurs on exis-
ting gypsum crystals, which provide an excellent nucleation
site. A higher slurry solids content and smaller slurry par-
ticle size increase the number of nucleation sites in the slurry
(as opposed to the internal surface area of the scrubbing ves-
sel) and decrease the likelihood that gypsum scale forms on the
scrubber internals. Therefore, a slurry solids content of 15
percent provides the system with more resistance to scale forma-
tion (depending on type of solids) than a slurry solids content
of 5 percent.
The effect of slurry particle size distribution can be even
more important than slurry solids content. When lime slaking
occurs, for example, the calcium hydroxide crystals produced
range 'in size from 1 to about 5 pm. The corresponding mean
particle size of scrubber slurry when using lime reagent will
range from 10 to 40 (jm, depending on the operating conditions.28
Scale formation caused by pH excursions has been less
pronounced with lime reagent than with limestone reagent because
the pH is at a higher level when lime is used. The degree of
calcium sulfite oxidation to calcium sulfate is reportedly
reduced when the pH level is increased. The tendency of slurry
to form high levels of dissolved calcium and sulfate ions is
thus suppressed by use of lime reagent at higher pH levels.
When a pH excursion to 4 to 4.5 occurs, regardless of the type
of reagent, severe and rapid formation of calcium sulfate scale
can arise. At one unit severe scale formation occurred on the
bottom level of perforated distribution trays as a result of a
pH variation initiated by a switch to high-sulfur coal. The pH
of the absorber slurry is at its lowest level when the slurry
reaches the bottom level of distribution trays in the absor-
ber.28
Soft deposits of scrubber slurry tend to form in all re-
gions of separated gas flow and in quiescent zones inside the
absorber vessel where flow is insufficient. The characteristics
of soft deposits are usually as follows:
0 A large fraction of the deposit consists of calcium
salts, including calcium sulfite.
0 The wet surface of the soft deposit is exposed to a
gas stream containing S02 .
2.2-48
-------
Very low rates of movement of liquid occur in the
interstices between the mechanically deposited soft
solids.
Under the conditions specified, it is only a question of
time before the soft deposit is cemented together by calcium
sulfate precipitation into a mass of hard, chemically bonded
scale. The SO2 absorbed into the liquid on the wet, exposed
surface of the deposit reduces the pH of the interstitial liquid
at the surface. This low pH liquid can diffuse into the soft
deposit, dissolving calcium sulfite and/or calcium carbonate
solids along the way. Oxygen dissolves into the liquid and
reheats with sulfite or bisulfite to form sulfate ion. The
dissolved calcium ion concentration continues to build up,
because of the low rates of movement in the interstitial liquid,
until the liquid is eventually supersaturated with calcium
sulfate. Finally, calcium sulfate precititates on the existing
solids in the soft deposit, bridging the gaps and "gluing" the
entire structure together with hard gypsum scale.
This slow but continuous precipitation of calcium sulfate
causes a transformation to hard scale in a matter of days or
weeks, depending on the reagent type, dissolved solids content
in the slurry, and specific conditions (i.e., pH level and SO2
and 02 concentration in the gas stream) at the location of the
soft deposit.
Scale caused by physical drying can occur at a wet/dry
interface upstream from the scrubber and also as a result of
repeated outages, during which the scrubber internals dry out.
in £>oth cases, the mechanism involves repeated cycles of wetting
and drying, during which the physical deposition of dissolved
salts during the drying cycle helps bond the normally soft
deposit into a much harder deposit. In an absorber operation,
the exact location of the wet/dry interface is primarily a
•£K t°n gas flow rate; the interface can move back and forth
with boiler load, thus creating the wetting and drying cycles
necessary for the formation of dried scale.
2.2.17.7 Oxidation of S02 to SO3--
As coal is burned, sulfur in the coal is volatilized and
oxidized to form SO2, and about 3 percent of the SO2 is further
oxidized to sulfur trioxide (SO3). Some iron compounds and
vanadium compounds act as catalysts for the conversion of SO2 to
S03 in the presence of free oxygen (excess air) at temperatures
that occur within a boiler. About 60 percent of the S03 is
absorbed in the wet particulate scrubber and the SO2 absorber,
according to work done at the Shawnee test facility of the
Tennessee Valley Authority (TVA).
2.2-49
-------
2.2.17.8 NO Interferences—
In one major case, Wood River Unit 4 of Illinois Power, NO2
interference was noted. When the FGD unit was put on-stream, a
visible plume remained. Since the FGD system removed more than
99 percent of the particulates and since the plume had a charac-
teristic brownish tint, NO2 was surmised to be the culprit. The
absence of particulate eliminated the masking effect that nor-
mally made the N02 emission unnoticeable.
2.2.17.9 Stoichiometric Ratio—
Newer lime systems use 1.10 to 1.15 moles of lime per mole
of SO2 removed, although some systems go below 1.10. A range
for most lime-based FGD units in the United States is 1.05 to
1.30 moles of lime per mole of S02 removed.
The higher the Stoichiometric ratio, the higher the opera-
ting cost, because more usable lime may be lost. A higher
Stoichiometric ratio may be required if there are wide varia-
tions in the inlet SO2 loading, so that SO2 removal may remain
at or above the mandated levels. The higher Stoichiometric
ratio increases the SO2 removal efficiency of the scrubbers.
2.2.17.10 Lime Utilization—
If the Stoichiometric ratio is based on S02 removed, the
utilization of lime is the inverse of the ratio. As the
Stoichiometric ratio increases, the utilization of lime de-
creases. Utilization is a function of mass transfer efficiency
and degree of reaction in the reaction tank. Good utilization
can be attained with good mass transfer, sufficient recycle tank
residence time, and extraction of the purge from the slurry
recycle lime. Some lime-based FGD units in the United States
report 88 to 99 percent utilization of lime.
2.2.18 Process Approximations and Design Data
This section gives information in tabular forms for process
approximations. It often is unnecessary to calculate a value
when all that is required is a rough estimate, e.g., an estimate
of the acreage required for disposal of sludge for a power
generating unit of a given size. Table 2.2-10, summarizes
approximate process values for coals of low, medium, and high
sulfur content.
2.2-50
-------
Table 2.2-10. PROCESS APPROXIMATIONS
Coal analysis
Moisture, %
Volatile matter, %
Fixed carbon, %
Ash, %
Sulfur, %
Heating value, Btu/lb
Ib SO- basis
dscf flue gas/lb S02
acf 300°F flue
gas/lb SO2
Ib CaO required/lb SO-
(1.1 times stoichio-
metric requirement)
Ib dry sludge
produced/lb SO-
Makeup water,
gal/lb SO-
Ton of wet coal basis:
dscf flue gas/ton
acf 300°F flue
flue gas/fcon
Ib S02/ton
Ib CaO required/ton
Ib dry sludge
produced/ton
Makeup water,' gal/ton
Sulfur content of coal, %
0.6 3.4 5.0
25.7
34.5
32.0
7.2
0.6
8250
9485
16,830
0.96
2.35
4.9
227,700
403,950
24
23.1
56.4
119
4.18
37.64
46.54
8.24
3.40
12,920
2388
3700
0.96
2.35
1.36
325,600
503,200
136
131
320
186
3.97
38.89
40.65
11.47
5.02
12,230
1535
2372
0.96
2.35
0.8
308,200
476,400
200
192
470
161
(continued)
2.2-51
-------
Table 2.2-10. (continued)
MW basis
dscfm flue gas/MW
acfm 300°F flue
gas/MW
Ib S02/h per MW
Ib CaO/h per MW
Ib sludge (dry)/yr
per MW
Makeup water gal/min
per MW
Million Btu basis (input)
dscf flue gas/million
Btu
acf 300°F flue gas/
million Btu
Ib S02/million Btu
Ib CaO/million Btu
Ib (dry) sludae
produced/million Btu
Makeuo water, gal/
million Btu
Sulfur content of
0.6 3.4
2300
4080
14.4
13.8
265,000
1.2
13,800
24,480
1.45
1.40
3.4
7.2
2100
3246
52.8
51.0
980,000
1.1
12,600
19,475
5.26
5.05
12.4
6.6
coal, %
5.0
2100
3246
82.2
78.6
1,500,000
1.1
12,600
19,475
8.2
7.9
19.3
6.6
2.2-52
-------
REFERENCES
1. PEDCo Environmental, Inc. Analysis of the Effect of Coal
Properties on Furnace/Boiler Combustion Characteristics.
April 1975, pp. 1-1 through 1-4.
2. PEDCo Environmental, Inc. Analysis of the Effect of Coal
Properties on Furnace/Boiler Combustion Characteristics.
April 1975, pp. 1-5 through 1-7.
3. Combustion Engineering Handbook, 1978, pp. 24-27.
4. PEDCo Environmental, Inc. Analysis of the Effect of Coal
Properties on Furnace/Boiler Combustion Characteristics.
April 1975, pp. 4-20 through 422.
5. Szabo, J.F., and R.w. Gerstle. Operation and Maintenance
of Particulate Control Devices on Coal-fired Utility
Boilers. EPA-600/2-77-129, July 1977, PP; 2-24 through
£* ^ *3 \J *
6. PEDCo Environmental, Inc. Analysis of the Effect of Coal
Properties on Furnace/Boiler Combustion Characteristics.
April 1975, pp. B-3, 4, 7, 8.
7. Szabo, M.F., and R.w. Gerstle. Operation and Maintenance
of Particulate Control Devices on Coal-fired Utility
Boilers. EPA-600/2-77-129, July 1977, pp. 2-49 through
8. Szabo, M.F., and R.W. Gerstle. Operation and Maintenance
of Particulate Control Devices on Coal-fired Utility Boil-
ers. EPA-600/2-77-219, July 1977, pp. 2-26 through 2-45.
9. PEDCo Environmental, Inc. Analysis of the Effect of Coal
Properties on Furnace/Boiler Combustion Characteristics.
April 1975, pp. 4-26 through 4-30.
10. Devitt, T., R. Gerstle, L. Gibbs, S. Hartman, R. Klier, and
B. Laseke. Flue Gas Desulfurization Systems Capabilities
for Coal-fired Steam Generators. PEDCo Environmental,
Inc., Contract No. 68-02-2603, November 1977, pp. 3-33,
-38, -39, -44, -45, and -82.
2.2-53
-------
11. Bechtel Corporation. Flue Gas Desulfurization Implications
of S02 Removal Requirements, Coal Properties, and Reheat
(Draft). July 1977.
12. Gibbs, L.L., D.S. Forste, and Y.M. Shah. Particulate and
Sulfur Dioxide Emission Control Costs for Large Coal-fired
Boilers, PEDCo Environmental, Inc. EPA Contract No. 68-02-
2535, Task No. 2, October 1977.
13. Battelle Columbus Laboratories. Stack Gas Reheat for Wet
Flue Gas Desulfurization System. EPRI Report, November
1976.
14. Flue Gas Desulfurization Using Fly Ash Derived from Western
Coals. EPA-600/7-77-07b, July 1977, p. 4.
15. Laseke, B.A. Environmental Protection Agency Utility FGD
Survey, December 1977-January 1978. EPA-600/7-78-051s,
March 1978.
16. Lime Handling, Application, and Storage. National Lime
Association, 1949. pp. 49, 50, and 56.
17. Conkle, H.M., H.S. Rosenberg, and S.T. DeNova. Guidelines
for the Design of Mist Eliminators for Lime/Limestone
Scrubbing Systems. EPRI Report No. FP-327, pp. 2, 50, 58
61, 56, 72, and 74.
18. Michael J. Baker, Inc. State-of-the-Art of the Flue Gas
Desulfurization Sludge Fixation. EPRI Report No. FP-671,
January 1978.
19. Disposal and Use of Byproducts from FGD Processes. in:
FGD Symposium. EPA-650/273/038, May 1973.
20. Federal Register, Part V. Department of the Interior,
September 19, 1978.
21. Treybal, R.E. Mass Transfer Operations. Ch. 6. McGraw-
Hill Book Co., New York, 1968.
22. Bethea, R.M. Air Pollution Technology. Van Nostrand
Reinhold Co., 1978.
23. Melia, M., M. Smith, W. Fischer, and B. Laseke. Environ-
mental Protection Agency Utility FGD Survey, June-July
1978. EPA Contract No. 68-01-4147, PEDCo Environmental,
Inc., Cincinnati, Ohio.
2.2-54
-------
24. Slack, A.V. Design Considerations in Lime-Limestone Scrub-
bing. Paper presented at Second Pacific Chemical Engi-
neering Congress, Denver, August 1977.
25. Bendor, F., J. Englick, and A. Saleom. Flue Gas Desulfuri-
zation in Venturi Scrubbers and Spray Towers. Paper pre-
sented at Second Pacific Chemical Engineering Congress,
Denver, August 1977.
26. Edward, W.M., and P. Huang. Mass Transfer in the Kellogg-
Weir Air Control System. Paper presented at National AIChE
Meeting, Houston, March 1977.
27. Haas, J.C., and W. Lombardi. Landfill Disposal of Flue Gas
Desulfurization Sludge. Combustion Engineering Power
Systems. Paper presented at NCA/BCR Conference and Expo
III, October 19-21, 1976.
28. Jones, D.G., A.V. Slack, andK.S. Campbell. Lime/Limestone
Scrubber Operation and Control Study. EPRI No. RP630-2,
April 1978.
2.2-55
-------
CONTENTS
2.3 PROCEDURES FOR CALCULATING A MATERIAL BALANCE
Page
2.3.1 Introduction 2 3-1
2.3.2 Process Description 2 3-1
2.3.3 Sample Calculations 2 3-21
2.3.4 Boiler Furnace Material Balance (Step 1) 2.3-25
2.3.4.1 Ash Balance 2 3-25
2.3.4.2 Carbon Balance 2*3-28
2.3.4.3 Chloride Balance 2*3-29
2.3.4.4 Theoretical Amount of Air Required for
Combustion 2 3-29
2.3.4.5 Moisture Balance 2!3-31
2.3.4.6 Nitrogen Balance 2*3-34
2.3.4.7 Oxygen Balance 2!3-36
2.3.5 S02 and Particulate Material Balance (Step 2) 2.3-38
2.3.6 Slurry Preparation Material Balance/Lime
Requirement (Step 3) 2.3-40
2.3.6.1 Slaker 2.3-42
2.3.6.2 Slurry Feed Tank 2!3-44
2.3.7 Scrubber Material Balance (Step 4) 2.3-45
2.3.7.1 Waste Slurry Calculation 2.3-45
2.3.8 Water Balance (Step 5) 2.3-55
2.3.9 Information on Several Operable Lime Absorber
Systems on Utility Boilers 2.3-57
2.3.9.1 Bruce Mansfield Station, Pennsylvania
Power Co. 2.3-57
2.3.9.2 Cane Run No. 4, Louisville Gas and Electric 2."3-63
2.3.9.3 Conesville No. 5, Columbus and Southern Ohio
Electric Co. 2.3-67
2.3-i
-------
CONTENTS (Continued)
2.3.9.4 Green River Station, Kentucky Utilities 2.3-73
2^3!9.5 Paddy's Run Station, Louisville Gas and
Electric 2.3-80
2.3.9.6 Phillips Power Station, Duquesne Light Co. 2.3-86
2.3.10 Conversion Factors 2.3-91
References 2.3-100
2.3-ii
-------
2.3 PROCEDURES FOR CALCULATING A MATERIAL BALANCE
2.3.1 Introduction
This section is intended to aid a design engineer in making
a material balance for a particular plant site. A set of con-
ditions is specified, and a material balance is developed in
stepwise fashion. Details concerning process approximations,
estimating techniques, and design data are given in Tables 2.3-1
through 2.3-3, and Figures 2.3-1 through 2.3-11.
The process selected for this illustration, and general
aspects of the process material balance are described in Section
2.3.2. Efforts were made to simulate operating plants in selec-
ting these data. Some of the process chemistries are simplified
for the purpose of illustration. A full sample calculation is
shown in Section 2.2.3.3.
2.3.2 Process Description
A flow diagram for the example lime scrubbing FGD process
(Example 1) is presented in Figure 2.3-12. In this example,
420,000 Ib/h of coal is fired to generate approximately 500 MW
(gross) of electricity. Although an FGD system of this size
usually consists of several modules, it is assumed in these
stream calculations that all the modules are combined.
The major components of coal are carbon, oxygen, nitrogen,
hydrogen, sulfur, free moisture, and ash. A minor, but impor-
tant, component is chloride because of the corrosion effect. A
typical coal analysis is presented in Table 2.3-4. The high
heating value (HHV) of this coal is 11,150 Btu/lb; ash consists
of 80 percent fly ash and 20 percent bottom ash.
The mechanisms of coal combustion are complex, and discus-
sion is beyond the scope of this section. For simplicity, it is
assumed that the degree of combustion is 100 percent and that
all carbon is oxidized to CO2 (no CO) and all sulfur to SO2 by
the following reactions:
C + O2 -* C02 t (1)
S + O2 -» S02 t (2)
Other reactions in the boiler furnace are as follows:
H2 + 1/202 -> H20 (g) (3)
H2 + C12 -» 2HC1 (4)
2.3-1
-------
Table 2.3-1. ENTHALPIES OF VARIOUS GASES
(Btu/lb of gas except as noted)
Temp., °F
100
150
200
250
300
350
400
450
500
550
600
700
co2
5.8
17.6
29.3
40.3
51.3
63.1
74.9
87.0
99.1
111.8
124.5
150.2
N2
6.4
20.6
34.8
47.7
59.8
73.3
84.9
97.5
110.1
122.9
135.6
161.4
H20a
17.8
40.3
62.7
85.5
108.2
131.3
154.3
177.7
201.0
224.8
248.7
297.1
°2
8.8
19.8
30.9
42.1
53.4
64.8
76.2
87.8
99.5
111.3
123.2
147.2
Air
9.6
21.6
33.6
45.7
57.8
70.0
82.1
94.4
106.7
119.2
131.6
156.7
a The enthalpies tabulated for H20 represent a gaseous system,
and enthalpies do not include the latent heat of vaporization,
It is recommended that the latent heat of vaporization at
60°F (1059 Btu/lb) be used where necessary.
2.3-2
-------
Table 2.3-2. MOLECULAR WEIGHTS FREQUENTLY USED
IN MATERIAL BALANCE CALCULATIONS
Name
Air
Calcium
Calcium carbonate (limestone)
Calcium hydroxide
Calcium oxide (lime)
Calcium sulfate
Calcium sulfate (dilhydrate)
Calcium sulfite
Calcium sulfite
Carbon dioxide
Hydrogen
Magnesium
Magnesium hydroxide
Magnesium oxide
Magnesium sulfate
(heptahydrate)
Magnesium sulfite
(hexahydrate)
Magnesium sulfite (trihydrate)
Nitrogen
Oxygen
Sulfur
Sulfur dioxide
Water
Formula
Ca
CaC03
Ca(OH)2
CaO
CaS04
CaS04-2H20
CaS03
CaS03-l/2H26
CO 2
H2
Mg
Mg(OH)2
MgO
MgS04-7H2O
MgSO~ • 6H«0
MgSO, • 3H~O
N2
°2
S
so2
H_0
Molecular weight
28.85
40.08
100.09
74.10
56.08
136.14
172.18
120.14
129.15
44.01
2.02
24.31
58.33
40.31
246.51
212.46
158.40
28.01
32.00
32.06
64.06
18.02
2.3-3
-------
Table 2.3-3 . ENERGY REQUIREMENT CALCULATIONS
C = Specific heat, Btu/(lb)(°F)
P = Energy required, kilowatts
E = Heat energy, Btu
H = Head, ft
s
L/G = ratio of liquor flow to flue gas rate, gal/1000
acf at the outlet
IP. = air flow rate at the inlet of reheat section,
Ib/min
AP = Pressure drop through FGD system, in. H2O
Q = Gas flow rate at the outlet of scrubber, acfm
AT = Degree of reheat, °F
1. Slurry recirculation pumps (70% pump efficiency assumed)
P = 0.000269 x HS x (L/G) x
= H (L/G) Q x (2.69 x 10~7)
2. Flue gas fans
p = 0.0002617 x AP x Q (assuming 80% efficiency)
3. Reheat of scrubber flue gas
E = 0.0.1757 m C.Q AT
2.3-4
-------
NJ
t
U>
I
Ul
2.0
1.5
<0
~ 1.0
c/)
O
0.5
OR MAG/I TMF pH 7
0.6
1.2
1.8
2.5
Ib S02/million Btu
Figure 2.3-1. Additive vs. outlet requirements.
Courtesy of Combustion Engineering.
-------
to
V
CTl
2.5
Ib/SOp/million Btu
Figure 2.3-2. Horsepower vs. outlet requirements
Courtesy of Combustion Engineering.
-------
4000 -
3000 -
-------
20,000 -
15,000 -
10,000 -
5,000 -
PLANT CAPACITY = 1000 MW
20 30 40
AP, in. H20
Figure 2.3-4. Fan energy requirement.
60
2.3-8
-------
3
-(-»
CO
2.0
1.5
1.0
0.5
PLANT CAPACITY = 1000 MW
10 20 30 40 50 60
Figure 2.3-5. Reheat energy requirement.
Note: Flue gas volume at reheat is assumed to be 21,000 acf/MW.
2.3-9
-------
NJ
V
!*•
TEMPERATURE, V.
Figure 2.3-6. Psychrometric chart.
-------
% S IN FUEL BY WEIGHT
(Q
fD
ro
CO
O
ro
fO
en
in
O
a
o>
o
c
O)
o
3
S02 EMISSIONS, Ib/million Btu heat input (coal)
i
REFERENCE LINE
S02 EMISSIONS, Ib/million Btu heat input (oil)
-------
Ctr-
3
.»->
00
c
o
2--
o
£ 4Hr
o
LU
ec.
LU
CO
o ::
co
o
t—t
CO
(/I
CM
o
CO
8-::
40-T-
36--
32--
28--
24--
o
LIME 20-
STOICHIOMETRIC
RATIO
8--
4— -
3--
2--
1--
O-1-
-M
CO
O
to
o
Figure 2.3-8. Lime requirement calculation.
2.3-12
-------
9- -
B--
7--
4- .
J- •
2--
ZOO
TEMP., °F
- -1
- -1
- -4
•3 LU
:D
oo
o
t LO
'* C\4
10
(C
o
- -7
•i-10
Figure 2.3-9. Saturated gas calculation.
Note:
nneSnrthCriPl"a",m^ J? ^ ! ™ ' ^ > bUt' 1l
on both the left and right ordinates for each case
2.3-13
-------
I
in
tsj
If 1 I 10s- -
1110
l«±
Figure 2.3-10. Recirculation tank capacity calculation.
2.3-14
-------
Figure 2.3-11. Wet sludge calculation.
2.3-15
-------
to
•
U)
I
COAL
H.V.=11,150 Btu/lb
ASH
S
°2
N,
H-0
14.0%
3.5%
7.2%
1.2*
61.18%
8.5%
0.02%
Figure 2.3-12. 500-MW model plant for Example 1.
-------
Table 2.3-4. DESIGN INFORMATION
Plant capacity
Coal
Excess air:
(70°F, 80%
relative
humidity)
Lime:
Scrubber:
Consumption
High heating value
Sulfur content
Oxygen content
Hydrogen content
Nitrogen content
Carbon content
Chloride content
Moisture content
fcontent
Ash V
Lbottom ash/fly ash
To air heater
To boiler-furnace
Utilization
Stoichiometric ratio
L/G ratio
Inlet gas temperature
Sulfite-to-sulfate oxidation
Reheating
Maximum
emission
(NSPS)
Solids:
Particulate
so2
Slaker
Slurry feed tank
Recycle tank
500 MW
420,000 Ib/h
11,150 Btu/lb
3.5%
7.3%
4.3%
1.2%
61.18%
0.02%
8.5%
14.0%
20 lb/80 Ib
40%
20%
91%
- 1.10
40 gal/1000 acf
285°F
20%
40°F
0.1 Ib/million Btu
1.2 Ib/million Btu
35%
20%
14%
2.3-17
-------
In these reactions, some of the hydrogen in the coal is consumed
to form hydrochloric acid. At steady state, this acid is com-
pletely absorbed in the scrubbing solution and removed with
interstitial water in the waste sludge.
In some coals, calcium in the ash reacts with sulfur diox-
ide or sulfur trioxide; this effect is neglected here.
Air at 70°F and 80 percent relative humidity (annual aver-
age) is supplied to the boiler furnace through an air heater.
The excess amounts of air supplied to the furnace (Stream 3) and
to the air heater (Stream 2) are 20 percent and 40 percent of
the theoretical air requirement for coal combustion, respec-
tively. Twenty percent excess air from Stream 2 is assumed to
leak to Stream 5 at the air heater.
The fly ash in gas Stream 5 is primarily removed in an ESP,
and some fly ash removal occurs in the scrubber section. The
maximum particulate emission must be in compliance with the NSPS
promulgated by EPA in December 1971, which is 0.1 Ib/million Btu
heat input. The resulting gas (Stream 6) enters the SO2 absor-
ber at 285°F/ and the SO2 is removed by a countercurrent lime
scrubbing process. The current NSPS limitation for SO2 is 1.2
Ib/million Btu heat input.
The temperature of the saturated flue gas from the absorber
(Stream 7) is increased 40°F in a reheater. It is assumed that
there is no mist carryover in the gas. The cleaned and reheated
flue gas (Stream 8) is discharged to the atmosphere through a
stack. in calculating gas flow rate the ideal gas laws are
assumed.
Typical pressure drop data are presented in Table 2.3-5.
Table 2.3-5. TYPICAL PRESSURE DROP DATA
Equipment
S02 absorber
(mobile bed)
(spray tower)
Demister
Reheater
Duct work
Pressure drop, in. H2O
10-12
5-6
2
3
3
2.3-18
-------
A 35 percent lime slurry is prepared in a slaker using
fresh water and lime. A typical lime analysis is presented in
Table 2.3-6. Sixty percent of the inert materials is removed
while the slurry is in the slaker. in this example, lime is
used in 10 percent excess of that required by the stoichiometry.
Table 2.3-6. LIME ANALYSIS
(wt. percent)
CaO
MgO
Inert
Total
92
3
5
100
in the slaker, the hydration reactions of lime containing
MgO as an impurity are expressed by the following:
CaO + H20 -> Ca(OH)2
MgO + H20 -> Mg(OH)2
This 35 percent slurry (specific gravity = 1.24) is then
diluted to 20 percent (specific gravity = 1.12) with a portion
of recovered water (Stream 13) in a slurry feed tank, then
pumped to a recycle tank, which maintains 14 percent solids
concentration (specific gravity = 1.09).
Slurry from the recycle tank is recirculated through the
absorber for removal of SO2, and a portion of the slurry (to be
determined by the material balance) is purged to a pond for
disposal (Stream 20).
The liquid-to-gas (L/G) ratio in the absorber normally
ranges from 40 to 60 gal of slurry per 1000 acf of gas, depen-
ding on the sulfur content, and the absorber type and effi-
ciency. In this example, the L/G is 40. The spent slurry from
the absorber returns to the recycle tank together with wash
water from the mist eliminator (Stream 11) and the recycled wash
water (stream 15). The flow rate in the mist eliminator wash
(Stream 15) is 300 gal/min or 150,000 Ib/h. The recycled liquor
from the pond is added to the recycle tank.
In the scrubber, both calcium and magnesium ions react with
S02 . The S02 reacts with dissolved alkalines such as Mg** more
rapidly than with suspended alkaline solids such as Ca(OH),
This follows the reaction:
2.3-19
-------
2SO2 + 2H2O + Mg++ •* [Mg + 2HSO3](dissolved solids) +2H+
(5)
Some of the Mg(OH)2 (assuming 5 percent in this example) forms
MgCO3 by the reaction with CO2 in the flue gas:
Mg (OH) 2 + CO2 -> MgCO3 (S ) + H2 O ( 6 )
The calcium ions tend to regenerate the bisulfite ions in the
same manner as in double alkali FGD systems. Calcium hydroxide,
Ca(OH)2/ reacts with magnesium bisulfite, Mg(HSO3)2 to regener-
ate the magnesium sulfite Mg(SO3), and form calcium sulfite,
CaSO3. At steady state, MgSO3 (dissolved solids) as well as
CaCl2 is removed from the system with the interstitial water in
the waste sludge. Most of the SO2 is ultimately removed by
reaction with Ca(OH)2. The overall reaction is:
Ca(OH)2 + S02 •» CaS03-l/2H20 + 1/2 H2O) (7)
Some of the calcium sulfite formed in Reaction (7) is
oxidized to sulfate with the oxygen in the flue gas. The degree
of oxidation in this example is 20 percent. The overall reac-
tion is expressed by:
Ca(OH)2 + S02 + 1/202 + H20 -» CaS04-2H20 (8)
The excess hydrated lime, Ca(OH)2, reacts with C02 in the
flue gas and is discharged to the pond as calcium carbonate
(CaC03) according to the reaction:
Ca(OH)2 (excess) + C02(g) -> CaCO3 * + H20 (9)
Therefore, the solids in the waste (Stream 20) are CaSO3*l/2
H20, ash, CaCO3, grit, and crystals of CaSO3-l/2H2O and
CaSO4-2H2O. The amount of dissolved gas in the waste stream is
neglected.
A slipstream of recycle slurry is sent to the pond (Stream
20); 50 percent suspended solids settles in the pond. The water
is recovered in Stream 12 and and reused in Streams 13, 14, and
15. The makeup water, which includes pump seal water and equip-
ment wash water, is supplied through Stream 11.
The design information discussed in the section is sum-
marized in Table 2.3-4.
2.3-20
-------
2.3.3 Sample Calculations
The overall material balance includes the boiler furnace
system and the FGD system, as shown in Figure 2.3-13. Inputs to
the combined systems consist of coal, air, lime, and water.
Outputs from these systems consist of ash (bottom and a portion
of fly ash), waste sludge, and cleaned flue gas. Only the
amount of coal consumption is known, however, and the rest of
the information must be calculated from a material balance for
each major component of the overall system. These material
balance calculations are performed in five steps, as follows:
The boiler-furnace (Step 1)
SO2 and particulate removal (Step 2)
Slurry preparation/lime requirement (Step 3)
Scrubber system (Step 4)
Overall water balance (Step 5)
The material balance of a boiler-furnace (Step 1) is repre-
sented in Figure 2.3-14 and the elements are summarized below.
Input:
1. Weight of coal charged
2. Weight of dry air supplied
3. Weight of moisture in air and coal supplied
Output:
1. Weight of dry gaseous products
2. Weight of water vapor in gaseous products
3. Weight of refuse (bottom and fly ash)*
Coal is charged at 420,000 Ib/h. The rest of the items,
however, must be calculated from the design information (Table
2.3-4) and the coal combustion properties discussed in Section
2.3.2. This calculation requires a component material balance.
The minimum amounts of fly ash and SO2 (Step 2) to be
removed are determined by the current NSPS limitation. Eighty
percent of overall particulate (fly ash)* removal is achieved in
an ESP, and 20 percent in the SO2 scrubber. From the input-
output data, the minimum acceptable efficiencies of the ESP and
the SO2 scrubber can be calculated. The material balance of
this step involves only SO2 and particulate, and the irrelevant
inputs to the FGD system are eliminated from Figure 2.3-15 for
simplicity.
* Since no mist carryover is assumed, fly ash and particulate
are essentially the same.
2.3-21
-------
COAL-
AIR-
BOILER-FURNACE
SYSTEM
ASH
CLEANED FLUE GAS
FGD
SYSTEM
LIME
•WATER
WASTE SLUDGE
Figure 2.3-13. Overall material balance.
2.3-22
-------
COAL-
BOILER-FURNACE
i
AIR PREHEATER
21
•AIR
GAS
BOTTOM ASH
Figure 2.3-14. Boiler-furnace material balance.
2.3-23
-------
CLEANED FLUE GAS'
|so2
IPARTICULATE
FLUE GAS 5
jso2
] PARTICULATE
1 I
fc.
^
ESP
[22
|7
FGD
SYSTEM
1
— 1
FLY
ASH
S02 AND FLY ASH AS
SLUDGE AND DISSOLVED
SOLIDS
Figure 2.3-15. SCL and particulate balance.
2.3-24
-------
The theoretical lime requirement (Step 3) depends on the
amount of SO2 to be removed from the FGD system, since removal
of one mole SO2 requires one mole of alkalinity (eigher CaO or
MgO). The actual supply of lime will be 110 percent of the
theoretical lime requirement (1.1 stoichiometric ratio) as
discussed in Section 2.3.2. Water required for slurry prepara-
tion can be calculated in this step.
Step 4 gives an overall material balance of the scrubber
system, as shown in Figure 2.3-16. Each component of waste
sludge is calculated with the process chemistry described in
Section 2.3.2. At steady state, the dissolved solids and the
chloride are analyzed based on the assumption that these species
are removed with the interstitial water in the waste sludge.
This determines the maximum concentrations of these species in
the scrubbing system.
The temperature of the flue gas leaving the scrubber is
determined by assuming adiabatic saturation. Use of a psychro-
metric chart for this purpose will be discussed. The final
conditions of the cleaned flue gas leaving the FGD system and
the makeup water requirement to this scrubber system are calcu-
lated in this step. The overall water balance is determined in
the next step.
Step 5, the overall water balance, includes the waste
treatment system, the distribution of recovered water, and the
water inputs and outputs to the combined system.
Following the calculation of each of these steps is a
summary in table or flow chart form. Stream numbers indicated
in the calculations are from Figure 2.3-12.
2.3.4 Boiler Furnace Material Balance (Step 1)
This step calculates the input and output data for coal
combustion in the boiler-furnace, as shown in Figure 2.3-17.
All calculations are based on 420,000 Ib of coal charged per
hour. The figures in the coal analysis are significant to 0.01
percent, which corresponds to 42 Ib/h. For hydrogen, this
represents about 40 mol/h; for chlorine and sulfur, about 1
mol/h. In this example, therefore, results are calculated to
the nearest 10 Ib/h and the nearest 0.1 mol/h.
2.3.4.1 Ash Balance—
This is a straightforward calculation from the ash content
of coax and the coal charged:
2.3-25
-------
FLUE GAS 6
FROM ESP **
LIME 17
SLURRY *"
CLEANED
FLUE GAS
SCRUBBER
SYSTEM
_ MAKEUP
WATER
20
WASTE SLUDGE
(IN A POND)
Figure 2.3-16. Scrubber material balance.
2.3-26
-------
U)
ro
ASH
co2
HC1
NITROGEN
OXYGEN
soz
WATER
TOTAL
COAL J
HHVa« 11
ASH
CARBON
CHLORIDE
HYDROGEN
NITROGEN
OXYGEN
SULFUR
WATER
TOTAL
aHHV = HIGH
Ib-mol/h
21,393.4
2.37
114,626.1
5070.4
458.5
13.814.7
41!
BOILER
FURNACE
3
Ib/h
47,000
941,600
86.4
3,211,100
162,200
29,370
248,900
4,640,300
20* EXCESS A
4,232,100 Ib
,150 etu/ib JDRY AIR = 4,179,61
1 UATFP - «;? «;nn it.
14.001 21
61.18
0.02
4.30 '
1.20 BOTTOM ASH
7-30 H.800 Ib/h
3.50
8.50
100. OOX
HEATING VALUE
ASH
co2
HC1
NITROGEN
OXYGEN
so2
WATER
TOTAL
J
ECONOMIZER
i
ID
AIR HEATER
rn
2 40t EXCESS AIR 0 70"F
1,111,300 acfm
DO Ib/h 5| ( 4.9J/,bOO Ib/h
fh I JDRY AIR • 4,875,200 ib/h
f ^ WATER • 61,300 Ib/h
Ib-mol/h
21,393.4
2.37
133,700
10,140.8
458.5
14.301.1
179.996.2
Ib/h
47,000
941.600
86.4
3.745.500
324,500
29,370
257.700
5,3451700
Figure 2.3-17. Detailed boiler-furnace material balance.
-------
Total ash = coal charged x ash content of coal
n nnn lb coal 14 Ib ash
= 420,000 H x 100 lb coal
= 58,000 Ib/h,
which consists of 80 percent fly ash and 20 percent bottom ash.
Thus,
Fly ash = total ash x fly ash content
lb ash 80 lb fly ash
= 58,800 ±-5 x 10Q lb *sh
= 47,000 Ib/h,
Bottom ash (Stream 21) = total ash - fly ash
= (58,800 - 47,000) Ib/h
= 11,800 Ib/h.
Ash Balance Summary
input Output
in coal 58,800 Ib/h Fly ash 47,000 Ib/h
Bottom ash 11,800 Ib/h
Total 58,800 Ib/h Total 58,800 Ib/h
2.3.4.2 Carbon Balance— .
The purpose of the carbon balance is to calculate the
amount of C02 generated. The moles of carbon in coal are calcu-
lated as:
Mol carbon = coal charge x carbon content of coal
mol carbon
x mol. wt.* in coal
lb coal 61.18 lb carbon
= 420,000 g x 100 lb coal
Tb-mol carbon
x 12.011 lb carbon
= 21,393.4 Ib-mol/h of carbon
This value is the same as the amount of CO^ formed since 1 mol
carbon forms 1 mol CO2 . Therefore, the weight of C02 formed is
* mol. wt. = molecular weight
2.3-28,
-------
Weight of C02 = lb-mol CO? mol. wt. of COo
2 h x mol C02
= 21,393.4 lb"mo1 CO? x 44.009 Ib CO,
Ib-mol CO2
= 941,600 Ib C02/h
2.3.4.3 Chloride Balance—
Chloride in coal combines with hydrogen by Reaction (4)
[H2 + C12 -» 2HC1]. The amount of HC1 formed is calculated
thus :
Amount of HC1 = coal charged x chloride content of coal, as
C12
x 2 mol HC1
mol. wt. of chloride (C12)
= 420,000 lb coal x 0'02 lb Cl,
h * 100 lb coal
x 2 Ib-mol HC1
2 x 35.452 lb C12
=2.4 Ib-mol HCl/h
or x 36ih:;»rWrr = 9°-° "> nci/h
2.3.4.4 Theoretical Amount of Air Required for Combustion--
The weight of air theoretically required for complete
combustion depends on the chemical composition of the fuel and
the stoichiometric relations involved in combustion. Since the
one element in common for all combustion reactions is oxygen,
the mass flow rate of air required for combustion must be calcu-
lated from an oxygen balance. The oxygen already in the coal is
assumed to be used in the combustion process.
Calculating the oxygen requirements for combustible consti-
tuents of coal entails the requirement for carbon and sulfur.
Carbon in coal was calculated at 21,393.4 Ib-mol/h carbon.
Since 1 mol oxygen is required per mol carbon by Reaction (1)
[C + O2 -» C02], oxygen required is 21,393.4 Ib-mol/h.
The moles of sulfur present in the coal at this feed rate
are calculated as follows:
Mol sulfur = coal charged x sulfur content of coal
x mol sulfur
mol. wt. of sulfur
= 420,000 lb coal x 3.5 lb sulfur Ib-mol sulfur
h 100 lb coal x 32.06 lb sulfur
= 458.5 Ib-mol sulfur/h
2.3-29
-------
Since 1 mol oxygen is required per mol sulfur (S + O2 » SO2 ),
the oxygen required is also 458.5 Ib-mol/h.
Some of the hydrogen in coal forms hydrochloric acid by
Reaction (4), and the available hydrogen in coal for Reaction
(3) is
Mol hydrogen = coal charged x (H2 content - C12 content)
mol hydrogen
x mol. wt. of hydrogen
n nnn lb coal 4.3 lb H9 - 0.02 lb Cl,
= 420,000 H x 100 lb coal
Ib-mol hydrogen
x 2.016 lb hydrogen
= 8916.7 Ib-mol/h available hydrogen
Since 1/2 mol oxygen is required per mol of hydrogen by Reaction
(3), oxygen required is 4458.3 Ib-mol/h. Water formed by this
reaction, however, is 8916.7 Ib-mol/h.
The oxygen content of coal is used in the combustion pro-
cess as follows:
Oxygen in coal = coal charge x oxygen content of coal
mol oxygen
x mol. wt. of oxygen
_ 420 000 lb coal 7.3 lb oxygen
— <±/:u,uuu ^ A 100 lb coal
Ib-mol oxygen
x 31.998 lb oxygen
= 958.2 Ib-mol/h oxygen
Because of this oxygen contained in coal, less air is required
for combustion reactions. Therefore, the theoretical oxygen
required for combustion of 420,000 Ib/h of coal is the sum of
the oxygen required for carbon, sulfur, and hydrogen oxidations
minus the amount of oxygen in coal.
Theoretical O2 = (21,393.4 + 458.5 + 4458.3 - 958.2) -
requirement
= 25,352 Ib-mol/h.
Theoretical air requirements are calculated as follows.
Since dry air contains 21 mol percent oxygen,* the theoretical
amount of air required for combustion of 420,000 Ib/h of coal is
2.3-30
-------
Theoretical dry air = 100 mol dry air
requirement m01 oxY9en * 21 mol 02
_ 25 oc-> Ib-mol 00 100 Ib-mol air
' h x 21 Ib-mol 02
= 120,723.8 Ib-mol/h dry air
The amounts of excess air to be supplied to the boiler-
furance and to the air heater are assumed to be 20 percent and
40 percent of the theoretical air requirement, respectively
(Section 2.3.2 and Table 2.3-6).
Thus,
Dry air supplied ., ......
to boiler-furnace theoretical air 120 mol air
in Stream 3 = requirement A 100 mol air
= 120 723 8 lb"mo1 dry air 120 Ib-mol air
h x 100 Ib-mol air
= 144,868.6 Ib-mol/h dry air
= 4,179,600 Ib/h of dry air
Dry air supplied
to air heater _ theoretical air 140 mol air
in Stream 2 requirement x 100 mol air
= 120 7?^ ft Ib-mol dry air 140 Ib-mol air
h X 100 Ib-mol air
= 169,013.3 Ib-mol/h dry air
= 4,876,200 Ib/h dry air
In the air heater, balance air between Streams 2 and 3
leaks to Stream 5. Thus,
Air leaks to Stream 5 = (Air suPPlied (Air supplied
to Stream 2) to Stream 3)
= (4,876,200 - 4,179,600) Ib/h dry air
= 696,600 Ib/h dry air
or = 24,144.7 Ib-mol/h dry air
2.3.4.5 Moisture Balance—
To complete the material balance, one must know the weight
of moisture in the gaseous products, which comes from free
moisture in coal, from available hydrogen in coal, and from
moisture introduced with the air.
* A more accurate value is 20.99 percent [International Critical
Tables, Vol. 1, p. 393 (1926)]. The 0.04 percent of C02, H2 ,
and rare gases may be ignored in combustions calculations.
Therefore, a commonly used air composition is 70 percent N2
and 21 percent 02. The weight of 1 mol of air is 28.851.
2.3-31
-------
Free moisture _ Coal charged x moisture content of coal
in coal mol water
x mol. wt. of water
_ 4?0 ooo lb coal 8.5 Ib water
- 420,000 h x
Ib-mol water
X 18.015 lb water
= 1981.7 Ib-mol/h of water
The moisture from available hydrogen in coal has been
calculated in the oxygen requirements; the value is 8916.7
Ib-mol/h water since 1 mol of H2 forms 1 mol of water.
In the calculation of moisture introduced with dry air,
year-round average condition of air is assumed to be 70°F and 80
percent relative humidity (Section 2.3.2). Since the weight of
dry air is known, the moisture content of air can be computed
from the molal humidity, H (lb water per lb dry air), expressed
as :
mol. wt. of water - ^A
-
~ weight of 1 mol 5
e A
where P = partial pressure of water vapor, psi
f\
and P = total pressure (14.696 psi)
The partial pressure of water vapor in air can be calculated
from the relative humidity, HR, defined as the ratio of the
partial pressure of the water vapor to the vapor pressure of
water at air temperature. By definition,
P.
H = 100
-------
and the humidity, H, from Eq. 1 is
_ 28.015 Ib water _ 0.290 psi _
28.851 Ib dry air X 14.696 psi - 0.290 psi
= 0.01257 Ib water/lb dry air.
The moisture introduced in air to the boiler-furnace (Stream 3)
is calculated as follows:
*
lb-mo1 water
s 52,500 Ib/h of water x IB^ ^ ^&r
= 2916.3 Ib-mol/h of water
The moisture in air to the air heater (Stream 2) is calculated
as follows:
™»n = Molal humidity x weight of dry air
Stream 2 -'in Stream 2
= 0.01257 a . 4,875,200
« 6!, 300 Ib/h of water ,
= 3402.7 Ib-mol/h of water
The moisture in the gaseous products (Stream 4) is calculated as
follows:
Moisture in _ Free moisture . SK?r\50m Moisture
Stream 4 ~ in coal + available hydrogen + ± stream
in coal
= (1981.7 + 8916.7 + 2916.3) "mo water
= 13,814.7 Ib-mol/h water
or = 248,900 Ib/h water
Since it is assumed that some of the air supplied to the
air heater leaks to gas Stream 5, the moisture content in Stream
5 increases accordingly, thus:
2.3-33
-------
Moisture in. _ Moisture in + Moisture in
Stream 5 ~ Stream 4 Stream 2
= (13,814.7 + 3402.7 - 2916.3)
= 14,301.1 Ib-mol/h water
or = 257,700 Ib/h water
Moisture in
Stream 3
Ib-mol water
Moisture Balance Summary, Boiler Furnace
Input
Ib/h
Free moisture
in coal
(1981.7 Ib-mol/h) 37,700
From available H2
in coal
(8916.7 Ib-mol/h) 160,700
Moisture introduced
with dry air 52,500
Total
248,900
Output
Total
Ib/h
Moisture in
gaseous products 248,900
248,900
Moisture Balance Summary, Air Heater
Input
Ib/h
Moisture in gaseous
products 248,900
Moisture introduced
with dry air 61,300
Total
310,200
Output Ib/h
Moisture in flue
gas 257,700
Moisture in dry
air leaving 52,500
Total 310,200
2.3.4.6 Nitrogen Balance—
Nitrogen balance of the boiler-furnace is represented by an
input consisting of the nitrogen in coal charged plus the nitro-
gen in air supplied in Stream 3 and an output consisting of the
nitrogen in gaseous products (Stream 4). Thus,
2.3-34
-------
Nitrogen in = Coal ch d x Nitrogen content
coal charged ^ of coal
= 420 000 lb coal v 1-2 Ib nitrogen
420,000 h 3. 10Q lb coal
= 5040 Ib/h nitrogen
Ib-mol nitrogen _ ,7Q Q Ib-mol nitrogen
or x 28.014 lb nitrogen *•''*•'* h
Nitrogen in air _ Dry air in 79 mol nitrogen
supplied in Stream 3 Stream 3 x 100 mol dry air
= 144,868.6 - ar
= 114,446,2 Ib-mol/h nitrogen
or = 3,206,100 Ib/h nitrogen
Nitrogen in gaseous _ Nitrogen in Nitrogen in air
products (Stream 4) coal charged in Stream 3
= (179.9 + 114,446.2) Ib-mol/h nitrogen
= 114,626.1 Ib-mol/h nitrogen
or x 28.014 lb nitrogen
Ib-mol nitrogen
= 3,211,100 Ib/h nitrogen
Nitrogen _ nitrogen in + dry air leaks 79 mol nitrogen
Stream 5 Stream 4 to Stream 5 100 mol dry air
= 114,626.1 " k " + 24,144.7
79 Ib-mol nitrogen
100 Ib-mol dry air
= 133,700 Ib-mol/h nitrogen
Nitrogen Balance Summary
Input Ib/h Output Ib/h
Nitrogen in coal 5,000 Nitrogen in
gaseous products 3,211,100
Nitrogen in air
supplied 3,206,100
Total 3,211,100 Total 3,211,100
2.3-35
-------
2.3.4.7 Oxygen Balance—
In the oxygen balance of a boiler-furnace, the input con-
sists of free oxygen in coal charged, oxygen in dry air supplied
for combustion, oxygen in free moisture in the coal, and oxygen
in moisture supplied with the air. The output consists of free
oxygen in gaseous products, oxygen in the CO2 and S02 in gaseous
products, and oxygen in the moisture in gaseous products. In
the calculation of amounts of free oxygen in the gaseous pro-
ducts in Streams 4 and 5, the procedure using these input-output
data is not simple. An alternative method is to use the excess
oxygen supplied, since this is the only free oxygen in the
gaseous products. Thus,
Oxygen in gaseous _ Excess oxygen supplied
products Stream 4 in Stream 3
= Theoretical oxygen required
in the boiler-furnace
20 mol oxygen
100 mol theoretical oxygen
lb-mol 20 mol
= 5070.4 Ib-mol/h oxygen
31.998 Ib oxygen = 162/200 lb/h oxygen
0 lb-mol oxygen
Oxygen in flue gas (Stream 5) is the sum of the free oxygen in
the gaseous products (Stream 4) and the oxygen from the dry air
leaks to Stream 5 in the air heater. Therefore,
Oxygen in Stream 5 = Oxygen in Stream 4
Dry air leaks 21 mol oxygen
+ to Stream 5 100 mol dry air
= 5070.4 Ib-mol^oxygen
+ 24,144.7 Ib-mol^dry air
21 lb-mol oxygen
x 100 lb-mol dry air
= 10,140.8 Ib-mol/h oxygen
31.998 Ib oxygen = 324 500 lb/h oxy
or x lb-mol oxygen
Air at 70 °F in Stream 2 consists of 169,013.3 Ib-mol/h dry
air and 3402.7 Ib-mol/h water. Therefore, the volumetric flow
rate is
2.3-36
-------
-
-------
Output lb/h
Weight of dry air leaving 4,179,600
Weight of moisture in air leaving 52,500
Weight of flue gas leaving including
ash and moisture 5,345,700
Total 9,577,800
Figure 2.3-17 presents the complete summary of Step 1, the
boiler-furance material balance, including air heater balance.
Component balance is also shown.
2.3.5 SOo and Particulate Material Balance (Step 2)
In this material balance, the input (Stream 5) consists of
the weight of fly ash (47,000 lb/h) and the weight of SO2
(29,370 lb/h). The output consists of the weight of particulate
in the cleaned flue gas, the weight of fly ash removed by ESP
(Stream 22), the weight of fly ash removed with sludge, the
weight of S02 in the cleaned flue gas, and the weight of SO2
removed by the scrubber as suspended and dissolved solids.
Since the input values are known, calculation begins with
the first output item, the weight of particulate in the cleaned
flue gas. Under the current NSPS regulation, the maximum al-
lowable particulate emission is 0.1 Ib particulate per million
Btu heat input. The maximum heat input, Q, into this plant is
Q = Coal charged x High heating value of coal
= 420,000 lb/h coal x 11,150 Btu/lb coal
= 4683 x 106 Btu/h.
Therefore, the maximum allowable particulate emission, Ep/ from
this plant is
v = 0.1 Ib particulate 46g3 x IQ6 Btu/h heat input
^P 106 Btu heat input
= 468 lb/h particulate
In calculating the weight of fly ash removed by ESP (Stream
22), we consider the minimum allowable fly ash to be removed
from this plant. Thus,
2.3-38
-------
Minimum allowable _ (Total fly ash _ (Maximum rate of
fly ash removal input) ~ particulate emission)
= 47,000 lb £1y ash - 468 lb Particulate
h h
= 46,532 Ib/h fly ash
By the assumption made in Section 2.3.2, the ESP removes 80
percent of the required minimum fly ash removal. Thus,
Fly ash removed _ Minimum allowable 80 lb fly ash removed
by ESP fly ash removal x 100 lb fly ash entering
80 lb fly ash removed
100 lb fly ash entering
= 46,532 lb fly ash x 80 lb fly ash removed
= 37,226 lb fly ash/h
and the minimum ESP efficiency, n^or,' is
CiOir
- Fly ash removed by ESP , 0/
nESP ~ Total fly ash input /0
_ 37,226 Ib/h fly ash
~ 47,000 Ib/h fly ash x 1UU/0
= 79.21%.
The weight of fly ash removed with sludge is represented by
the following balance:
Fly ash removed _ Minimum allowable Fly ash removed
with sludge fly ash removal ~ by ESP
= (46,532 - 37,226) Ib/h fly ash
= 9306 Ib/h fly ash
The minimum allowable particulate removal efficiency, r\ , is
Minimum allowable fly ash removal , nrv>/
np " Total fly ash input x 1UU/0
- 46,532 Ib/h fly ash
~ 47,000 Ib/h fly ash x iuu/0
= 99.00%
Next we calculate the weight of SO3 in the cleaned flue
gas. Under the current NSPS regulation, the maximum allowable
S02 emission is 1.2 lb S02/million Btu heat input. Since the
maximum heat input, Q, is 4683 x million Btu/h, the maximum
allowable SO2 emission, Ec_ , from this plant is
2
2.3-39
-------
E = — 1.2 Ib S02 46Q3 1()6 Btu/h heat
SO2 106 Btu heat input
= 5620 Ib/h SO2
The final calculation in the SO2 and particulate material
balance is the final element of output, the weight of SO2 re-
moved by the scrubber.
SO2 removed by _ cn inr».+ Maximum allowable
scrubber ~ 2 inpuu " SO2 emission
= (29,370 - 5620) Ib/h SO2
= 23,750 Ib/h SO2
" 37°'76 lb-mol/h so*
Therefore, the minimum allowable SO2 removal efficiency, nso ,
SO, removed by scrubber lnn°/
nS02 = S02 input x 100/°
23.750 Ib/h SO,
~ 29,370 Ib/h SO2
The S02 and particulate material balance is summarized in
Figure 2.3-18.
2.3.6 Slurry Preparation Material Balance/Lime Requirement
(Step 3)
As discussed earlier, the theoretical lime requirement
depends on the amount of SO2 and HCl to be removed from the FGD
system (scrubber). In Step 2, this is calculated as 370.76
lb-mol/h S02 and 2.4 lb-mol/h HCl. Since 1 mol SO2 requires 1
mol alkalinity (as CaO and MgO), and 1 mol HCl requires 1/2 mol
alkalinity, the theoretical alkalinity requirement is 370.76 +
1.2 = 371.96 lb-mol/h. The actual alkalinity supplied is 110
percent of the theoretical value. Thus,
noy
Actual alkalinity = Theoretical alkalinity x ±QO%
_ 371 96 lb-mol 110*
- 371.96 h X 1QO%
= 409.2 lb-mol alkalinity/h,
which is supplied from lime containing 92 percent CaO, 3 percent
MgO, and 5 percent grit, as shown in Table 2.3-5.
2.3-40
-------
CLEANED FLUE GAS
JS02 = 5620 Ib/h
JPARTICULATE = 468 ib/h
U)
FLUE GAS 5
S02 = 29,370 Ib/h *"
PARTICULATE = 47,000 Ib/h
ESP
nESp= 79.21%
i
6
S02 = 29,370 Ib/h
FLY ASH = 9774 Ib/h
so2
SCRUBBER
n^ = 80.9%
22
FLY ASH
37,226 Ib/h
SLUDGE
S02 = 23,750 Ib/h
FLY ASH = 9306 Ib/h
Figure 2.3-18. Summary of SOp and particulate material balance.
-------
Part of the excess alkalinity (MgO + CaO) will react with
the CO2. Since the calcium salts are less soluble, the alkalin-
ity will appear as MgCO3. The exact amount of MgO that will
react cannot be determined because of lack of information about
how much CO2 is dissolved in the slurry recycle stream, the
amount of Mg(OH)2 formed during the slaking process, the amount
of MgSO3 formed during the scrubbing, and other unknown vari-
ables. The composition of the available lime for SO2 absorption
is summarized in Table 2.3-7.
Table 2.3-7. COMPOSITION OF THE AVAILABLE LIME
FOR SO2 ABSORPTION
(Basis: 100 lb lime supplied)
Component
CaO
MgO
Total
Molecular
weight
56.079
40.309
Weight,
lb
92
3.00
95.00
Ib-mol
1.6405
0.0744
1.7149
Mol percent
95.662
4.338
100.000
im
The amount of CaO in the lime supplied is
CaO in lime = Actual alkalinity x -, ^.p"" I" tt""?. . -
•* 1.7149 mol alkalinity
= 409 2 x 100 lb lime
tu^.z x ]_7149 mol alkalinity
= 23,869 lb/h lime,
which contains 1190 lb/h grit by balance.
Lime Requirement Summary (Stream 9):
wt%
lb/h
mol/h
CaO
MgO
Grit
92
3
5
21,950
720
1,190
391.4
17.8
Total
100
23,860
409.2
2.3.6.1 Slaker—
CaO and MgO are hydra ted to Ca(OH)2 and Mg(OH)2 in the
slaker, respectively. Hydration also removes 60 percent of the
grit from the lime supplied. Thus,
2.3-42
-------
Grit removed _ Total grit in 60 Ib grit
(Stream 23) lime supplied x 100 Ib grit
= 1190 lb/h grit x
= 710 lb/h grit
Since 1 mol CaO or MgO forms 1 mol Ca(OH)2 or Mg(OH)2, the
resulting solids from the slaker consist of:
CatOH) = Total Ca° in x mol. wt. of Ca(OHU
lz lime supplied Ib-mol CaO
= 391.4 Ib-mol/h CaO x 74.094 Ib Ca(OH)?
Ib-mol CaO
= 29,000 lb/h Ca(OH>2
Ma(OH) = Total M9° in x mol! wt. of Mg(OH),
yv lz lime supplied Ib-mol MgO
= 17.8 Ib-mol/h MgO x 58^24 "> Mq(OH)?
Ib-mol MgO
= 1040 lb/h Mg(OH)2,
Grit _ Total grit in _ Grit removed
in slurry lime supplied ~ in slaker
= (1190 - 710) lb/h grit
=480 lb/h grit
Therefore,
Total weight of solids = Ca(OH)2 + Mg(OH)2 + Grit in slurry
= (29,000 + 1040 + 480) lb/h
= 30,520 lb/h solids,
which consists of 35 percent of the slurry from the slaker.
The water content of the slaker slurry is
Water content _ Total weight of 65 Ib water
in slaker slurry solids x 35 Ib solids
- 30,520 lb/h solids x
= 56,680 lb/h water
Therefore, total water requirement (Stream 10) in the slaker is
2.3-43
-------
Total water _ Water required + water content in
requirement ~ for hydration slaker slurry
... __ , . . mol. wt. of water
= (Hydration Hydration
for CaO for MgO)
Water content in
slaker slurry
- /-^cn A + 17 Q> lb-mol water 18.015 Ib water
- (391.4 + i/.9) h x lb_mol water
+ 56,680 Ib water
= 64,050 Ib/h water
It is often convenient to convert the mass flow rate unit to
volumetric flow rate. The conversion factor is
gal _ Ib solution v ft3 7.48 gal h
min ~ H62.4 Ib x sp. gr. ft? 60 min
Ib solution 0.00200
h sp. gr.
Material Balance Summary, Slaker
Input Ib/h output lb/h
CaO 21,950 Ca(OH)2 29,000*
MgO 720 Mg(OH)2 1,040*
Grit 1,146 Grit in slurry 480
Subtotal (Stream 9) 23,816 Water 56,680
Subtotal (Stream 16) 87,200
Water (Stream 10) 64,050 Grit removed
(Stream 23) 710
Total 87,866 Total 87,910
2.3.6.2 Slurry Feed Tank—
Thirty five percent lime slurry is diluted to 20 percent
with a portion of recovered water in a slurry feed tank, as
discussed in Section 2.3.2.
Since 20 percent lime slurry contains the same amount of
solids as the 35 percent slurry, the weight of the 20 percent
slurry is
* Values in the output calculations were rounded; thus, the to-
tals of input and output are not equal.
2.3-44
-------
Weight of 20% _ Weight of solids 100 Ib 20% slurry
slurry in 35% slurry x 20 Ib solids
= 30,520 Ib/h solids x 10° ** 2°%
20 Ib solids
= 152,600 Ib/h 20% slurry
The water content of the 20 percent slurry is 122,080 Ib/h water
by balance (152,600 - 30,520).
The amount of recovered water required for dilution (Stream
13) is
Recovered water _ Weight of 20% Weight of 35%
requirement slurry ~ slurry
= (152,600 - 87,200) Ib/h water
= 65,400 Ib/h water
Material Balance Summary, Slurry Feed Tank
Ib/h Output (Stream 17) Ib/h
Ca(OH)2 29,000 Ca(OH)2 29,000
Mg(OH)2 1,040 Mg(OH)2 1,040
Grit 480 Grit 480
Total solids 30,520 Total solids 30,520
Water in Water in
35% slurry 56,680 25% slurry 122,080
Subtotal (Stream 16) 87,200
Recovered water
(Stream 13) 65,400
Total 152,600 Total 152,600
A complete summary of the slurry preparation material balance is
presented in Figure 2.3-19.
2.3.7 Scrubber Material Balance (Step 4)
2.3.7.1 Waste Slurry Calculation—
Input data consist of the following:
2.3-45
-------
to
•
U)
I
LIME 23,830 Ib/h
C CaO 21,930 Ib/h
1 MgO 715 Ib/h
(GRIT 1190 Ib/h
FRESH WATER ~^-
63,960 Ib/h
\
-^
9
SLAKER
16 35% SLURRY _
87,080 Ib/ti
23 <
Ca(OH)2 28,970
Mg(OH)2 1030
GRIT 480
WATER 56,600
GRIT
710 Ib/h fra
13)65,320
SLURRY
FEED
TANK
„
17
20% SLURRY
152,400 Ib/f
lnit\ 9H Q7n
RECOVERED WATER
Ib/h
Mg(OH)2 1030
GRIT 480
WATER 121,920
Figure 2.3-19. Summary of slurry preparation material balance.
-------
Available alkalinity
for SQ2 removal Other
Ca(OH)2 = 391.4 Ib-mol/h Grit = 480 Ib/h
Mg(OH)? = 17.8 Ib-mol/h Mg(OH)2 for
forming MgC03 = 0.89 Ib-mol/h
Total = 409.2 Ib-mol/h Ash = 936 Ib/h
Since total SO2 + HC1/2 to be removed is 372.0 Ib-mol/h, the
excess alkalinity is 409.2 - 372.0 =37.2 Ib-mol/h.
Sulfite formation:
(1) MgSO3 formed by Reaction 5 as dissolved solids is 17.8
Ib-mol/h
(2) CaSO3 formed by Reaction 7 = 354.2 Ib-mol/h
Twenty percent oxidation of sulfite to sulfate:
(1) CaSO4 formed by Reaction 8 = 354.2 Ib-mol/h
20 Ib-mol CaSOA
100 Ib-mol CaSO
= 70.84 Ib-mol/h
(2) CaSO4-2H20 crystal formed = 70.84 Ib-mol/h
(12,200 Ib/h)
(3) CaSO3 left in the product = 354.2 - 70.84
= 283.36 Ib-mol/h
(4) CaS03'l/2H2O crystal formed = 283.36 Ib-mol/h
(36,600 Ib/h)
(5) Oxygen required by the reaction 8 = 1/2 x 70.84
Ib-mol/h = 35.42 Ib-mol/h = 570 Ib/h
MgCO3 formation by Reaction 6:
(1) Mg(OH)2 available for Reaction 6 = 0.89 Ib-mol/h
(2) MgCO3 formed =0.89 Ib-mol/h (75 Ib/h)
(3) C02 consumed =0.89 Ib-mol/h (40 Ib/h)
Limestone formation by Reaction 9:
(1) Excess Ca(OH)2 =37.2 Ib-mol/h
(2) CaC03 formed =37.2 Ib-mol/h (3720 Ib/h)
(3) CO2 consumed = 37.2 Ib-mol/h (1640 Ib/h)
The solids content of the waste slurry is summarized in
Table 2.3-8.
2.3-47
-------
Table 2.3-8
WASTE SLURRY SUSPENDED SOLIDS
Component
Ash
Grit
CaCO3
CaS03*l/2H2O
CaSO4-2H2O
MgCO3
Total
Mass flow rate,
Ib/h
9,306
480
3,720
36,600
12,200 '
74
62,380
Composition, %
(dry basis)
14.92
0.77
5.95
58.68
19.56
0.12
100.00
Since the solids content in waste Stream 20 is 14 percent,
total waste slurry flow rate is
62,400 Ib/h suspended solids x 14
= 445,500 Ib/h (or = 819 gal/min),
and the water content is 383,100 Ib/h by balance.
The waste slurry containing 14 percent suspended slurry is
introduced into the pond, where the sludge containing 50 percent
suspended solids settles and the balance water is recycled to
the FGD system.
The resulting material balance is represented in Figure
2.3-20.
14% suspended solids
445,500 Ib/h (819 gpm)
- 383,100 Ib/h
ds - 62,400.1b/h
Blowdown
1
24
Pond*
Suspended solids = 62,400 Ib/h (50»)
'H-,0 '- •-- --
Total
= 62,400 Ib/h
= 124,800 Ib/h
12
Recovered water
370,700 Ib/h
gpm)
Figure 2.3-20. Pond material balance.
Stream 24 blowdown is assumed to be zero during normal
operation, although some blowdown may be necessary on an inter-
mittent basis to remove chlorides from the system.
2.3-48
-------
It should be noted that the interstitial water in the
sludge contains MgSO3 and CaCl2 as dissolved species and remains
on the bottom of the pond. The hydrogen chloride introduced
into the system will react with the alkalinity. Consequently,
the 2.4 Ib-mol of hydrogen chloride produced will react to form
1.2 Ib-mol of calcium chloride (130 Ib/h CaCl2). At steady
state, all the calcium chloride introduced (130 Ib/h) to the FGD
system and 17.8 Ib-mol/h MgS03 (or 1860 Ib/h MgSO3) are removed
with the interstitial water. The concentrations of the dis-
solved species are as follows:
Concentration
of
MgSO3
_ amount of MgSQ., in the interstitial water
weight of solution less suspended solids
- I860 Ib/h MgSOi 6
~ 383,100 Ib/h solution x 10 ppm
= 4,855 ppm MgS03 as dissolved solids
Concentration
of
HC1
_ amount of CaCl^ in the interstitial water
weight of solution less suspended solids
130 Ib/h HC1
383,100 Ib/h solution
= 349 ppm HC1
x 106 ppm
Inlet flue gas to the scrubber—The composition of the
inlet gas to the scrubber (Stream 6) is the same as the gas
leaving the air heater except that 79.21 percent (37,226 Ib/h)
of the fly ash is removed. The result is presented in Table
2.3-9.
Table 2.3-9. INLET FLUE GAS TO SCRUBBER (STREAM 6)
Components
Ash
CO2
HC1
N2
02
SO2
H20
Total
Flow rate
Ib/h
9,774
941,600
130
3,745,500
324,500
29,370
257,700
5,308,570
Ib-mol/h
21,393.4
1.2
133,700.0
10,140.8
458.5
14,301.1
179,995.0
Composition,
wt. %
18.68
0.00
74.30
6.44
0.58
100.00
Dry basis
2.3-49
-------
The volumetric flow rate, V , is
Vj
179,996.2 Ib-mol flue gas (460 + 285)°F 359 ft,
VG ~ h (460 + 32)°F Ib-mol
h 4039.2 in. H;>O
60 mm (4039.2 + 20) in. H2O
= 1,623,000 acfm at 285°F
when the gauge pressure of the gas is 20 in. H2O. The molal
humidity of gas, H, is
_ weight of water
~ weight of dry gas
= 257.700 Ib water/h
5,308,500 Ib/h flue gas - 257,700 Ib/h water - 9774 iD/n ash
= 0.05112 Ib H2O/lb dry gas
Outlet gas from the scrubber (Stream 7)—In the scrubber,
some S02 (23,750/h Ib SO2 ) and ash (9306 Ib/h) are removed by
the scrubbing slurry, and some oxygen is consumed to oxidize
sulfite to sulfate. Carbon dioxide is also consumed by Reac-
tions 6 and 9. Hydrogen chloride is completely consumed by the
excess alkalinity to form calcium chloride.
Therefore, the resulting gas compositions (dry basis) are
as follows:
Ash = 9774 Ib/h - 9306 Ib/h = 468 Ib/h ash
CO2 = 941,600 Ib/h - 40 Ib/h consumed by Reaction 6 - 1630
consumed by Reaction 9
= 939,930 Ib/h CO2 (21,357.7 Ib-mol/h CO2)
N2 = 3,745,500 Ib/h (133,700 Ib-mol/h N2)
°z = °2 in inlet flue gas - O2 consumed in Reaction 8
= 324,500 Ib/h 02 - 570 Ib/h O2
= 323,930 Ib/h O2 (10,123.4 Ib-mol/h O2)
SO2 = S02 in inlet flue gas - SO2 removed in the scrubber
= 29,370 Ib/h S02 - 23,750 Ib/h SO2
= 5620 Ib/h SO2 (87.74 Ib-mol/h SO2)
Total = 5,015,400 Ib/h flue gas (165,268.84 Ib-mol/h)
(dry basis)
Flue gas temperature leaving the scrubber (Stream 7)—Use
of a psychrometric chart permits rapid estimation of the humidi-
ty and the temperature of the air leaving a scrubbing system.
2.3-50
-------
When unsaturated air is introduced into a scrubbing system,
water will evaporate into the air under adiabatic conditions at
constant pressure. The wet-bulb temperature remains constant
throughout the period of vaporization.
If evaporation continues until the air is saturated with
water vapor, the final temperature of the gas will be the same
as its initial wet-bulb temperature (dew point).
For example, Figure 2.3-21 shows that the air (point A) is
at a temperature of 285°F with humidity of 0.05112 Ib water per
Ib dry air. As vaporization takes place, the molal humidity of
air increases but the wet-bulb temperature remains constant.
The dry-bulb temperature must correspondingly decrease along the
125°F wet-bulb temperature line (adiabatic cooling line).
Therefore, the air leaving the scrubbing system is saturated at
a temperature of 125°F. As illustrated here, the psychrometric
chart is prepared for the "Air-Water" system.
For gas containing carbon dioxide, a line is established
between the lines X = O and X = 1.0, depending on the mole
fraction (Xt ) of the carbon dioxide in the gas. Then a line
parallel to this is projected from point A, representing a
dry-bulb temperature of 285°F and a humidity of 0.05112, to the
saturation curve. The intersection, B, is at a temperature of
126°F, and the molal humidity at this point is 0.09896 Ib H20/lb
dry air.
When a more accurate value is needed, it can be calculated
from the water properties and the definition of the saturation
humidity, H ,
5
18.015 Ib water PA
s ~ 28.851 Ib dry air x P - Pa (Eg- 3)
A
where PA = vapor pressure of water
and P = total pressure
At the saturation temperature of 126°F, P = 2.01046 psi.
Therefore, A
H 18.015 Ib water 2.01046 psi
s = 28.851 Ib dry air (14.696 - 2.01046) psi
= 0.09896 Ib water/lb dry air
The amount of water vapor in the outlet gas is
2.3-51
-------
ADIABATIC
COOLING LINE
s-
(O
-o
JD
o
evj
0,09896
~ 0.05112
TEMPERATURE, °F
Figure 2.3-21. Psychrometric chart for Example 1 (not to scale).
2.3-52
-------
0.09896 lb water
Ib dry gas
x 5,015,400 Ib/h dry gas = 496,300 Ib/h water
or = 27,550.6 Ib-mol
water
and the total mass flow rate of the gas is
5,015,400 Ib/h dry gas + 496,300 Ib/h water = 5,511,700
Ib/h gas.
The volumetric gas flow rate is
{165,268.84 + 27,550.6) Ib-mol (460 + 126)°F
h (460 +• 32)°F
h „ 359 fta „ 4039.2 in. H->Q
x
60 min Ib-mol (4039.2 + 6) in. H2O
= 1,372,100 acfm at 126°F
The composition of the cleaned flue gas leaving the scrub-
ber (Stream 7) is summarized in Table 2.3-10.
Table 2.3-10. CLEANED FLUE GAS COMPOSITION (STREAM 7)
Component
Ash
C02
N2
°2
S02
H2O
Total
Mass flow rate,
Ib/h
468
939,900
3,745,500
323,900
5,620
496,300
5,511,700
Composition,
wt. %
0.0085
17.05
67.96
5.88
0.1020
9.00
100.00
Gas leaving reheater (Stream 8) --Gas flow rate:
4039.2 in. water
- 1 372 100
- j., J/-S,-LUU
x (460 + 166)°F
x °
126)°F (4039.2 + 3) in. water
= 1,464,700 acfm at 166°F
This gas contains 445 ppm or 5620 Ib/h SO2 .
Relative humidity:
When the gas is heated, the dry-bulb temperature increases,
but the molal humidity of the gas remains constant. Therefore,
the gas property moves along the dotted line from B to C in
2.3-53
-------
Figure 2.3-21. Point C represents the gas leaving the reheater.
The relative humidity at this point can be read from the figure
or calculated by Eg. 1, where P is the same as the vapor pres-
sure at 126°F, and P is the vapor pressure at 166°F. There-
fore,
PA at 126°F _ -. .0 2.01046 psi
-, ™o/ _ -. nr.0/ v .
HR = 100% X Pa at 166°F - 10°% X 5.4916 psi
f\
=36.6 percent
1,372,100 acfm -*7 Reheater -»8 1,464,700 acfm
at 126°F at 166°F
H = 100% H = 36.6%
Scrubber slurry flow requirement (Stream 19)—The L/G ratio
of the scrubber in this process is given as 40 gal slurry per
1000 ft3 gas. Since the gas flow rate from the scrubber is
1,372,100, the slurry flow rate, Lg, is
40 gal 1,372,100 ft3 9.065 Ib 60 min
s ~ 1000 ft3 min gas h
= 29,851,400 Ib/h of slurry (or = 54,880 gal/min),
which contains 14 percent suspended solids (or 29,851,400 x 0.14
= 4,179,200 Ib/h) and 86 percent water (or 25,672,200 Ib/h).
From the flow diagram, Figure 2.3-12, the total slurry flow
rate in Stream 18 is the sum of Streams 19 and 20, i.e.:
Stream 18 = Stream 19 + Stream 20
= 29,851,400 Ib/h slurry + 445,500 Ib/h waste
slurry
= 30,296,900 Ib/h slurry (55,700 gal/min),
which contains 4,241,600 Ib/h suspended solids (14%) and
26,055,300 Ib/h water (excluding water in the crystals).
Water requirement—The total water requirement to the
scrubber and the recycle tank can be calculated from the overall
material balance of the above equipment. Therefore, in Figure
2.3-12, the sum of inputs by Streams 6, 11, 14, 15, and 17
should be balanced with that of outputs by Streams 7 and 20.
Streams (6 + 11 + 14 + 15 + 17) = Streams (7+20) (Eq. 4)
2.3-54
-------
where the flow rates are
Stream 6 = 5,308,500 lb/h,
Stream 17 = 152,600 lb/h,
Stream 7 = 5,511,700 lb/h,
Stream 20 = 445,500 lb/h.
Streams (11 + 14 + 15) are the total water requirement to the
scrubber and recycle tank. From equation 4,
Streams (11 + 14 + 15) = Streams (7+20-6-17)
= 5,511,700 + 445,500 - 5,308,500 -
152,600
= 496,100.
Since Stream 15 is set at 150,000 lb/h for mist eliminator
washing water from the recovered water,*
Streams (11 + 14) = (496,100 - 150,000) lb/h
= 346,100 lb/h
The flow rate in Stream 14 can be calculated from the recovered
water balance in Step 5. The result, however, is presented in
Figure 2.3-22 for convenience.
2.3.8 Water Balance (Step 5)
Water input to the absorber and the recycle tank system
consists of the following:
Stream 6: Water in the flue gas to the scrubber (257,700
lb/h)
Stream 11: Fresh makeup water
Stream 14: Recovered water to the recycle tank
Stream 15: Demister wash water from the recovered water to
the scrubber-absorber (150,000 lb/h)
Stream 17: Water in the feed slurry (122,080 lb/h)
Water output from the scrubber and the recycle tank system
is as follows:
Stream 7: Water carried out by flue gas (496,300 lb/h)
Stream 20: Water in the waste slurry (383,100 lb/h)
Recovered water to the recycle tank (Stream 14) is calcu-
lated as follows:
Stream 14 = Stream 12 - (Strearts 13 + 15)
= 320,700 lb/h - (65,400 lb/h + 150,000 lb/h)
= 105,300 lb/h (or 211 gal/min)
* Note that the recovered water contains 4855 ppm MgSO, and 235
ppm HCl as dissolved species. 3
2.3-55
-------
WASH AND
MAKEUP WATER
240.900 Ib/h (482 gal/mln) 11^
RECOVERED WATER
100,000 Ib/h (300 gal /mi n) 15
5,308,500 Ib/h 6 __
1.623,000 acfm P <
ASH 9.774 Ib/h
S02 29,370 Ib/h
C02 941,600 Ib/h
N2 3,745,500 Ib/h
02 324.500 Ib/h
HjO 257,700 Ib/h J
RECOVERED WAT
>85°F
ER 14 __
105.400 Ib/h (211 gal/mTn
201 SLURRY FEED 17 _
152,400 Ib/h (27
/SOLIDS • 30.500
\H^O • 121,900 Ib
2 gal/mTn
Ib/h \
/h /
7
SCRUBBER
ABSORBER
1
1 RECYCLE
TANK
i
18
1
.« 19
5,511.700 Ib/h
ASH 468 Ib/h
SO? 5620 Ib/h
COo 939,900 Ib/h
N2 3,745,500 Ib/h
Oo 323,900 Ib/h
H20 496.300 Ib/h
L/G • 40 gal/1000 acf
29,851,400 Ib/h (54,880gal/m1n)
/SOLIDS - 4,179,200 Ib/h \
l,H20 • 25,672,200 Ib/h y
— 20 »• UA1TF 1IIIRRY
30.296.900 Ib/h
(55.700 gal/min)
/SOLIDS -4,241,600 Ib/h
\H0 -26,055.300 Ib/h
445,500 Ib/h (819 gal/mln)
SOLIDS • 62.400 Ib/h \
,0 • 383.100 Ib/h )
Figure 2.3-22. Summary of scrubber material balance.
2.3-56
-------
Fresh makeup water (including pump seals) is the water in
Stream 11. The sum of Streams 11 and 14 is 346,100 Ib/h.
Stream 14 is known to be 105,300 Ib/h. The calculation there-
fore is,
Stream 11 = Streams (11 + 14) - Stream 14
= (346,100 - 105,300) Ib/h
= 240,800 Ib/h fresh water (482 gal/min)
The total freshwater requirement consists of freshwater for
the slaker (Stream 10) plus fresh makeup water (Stream 11).
Thus,
= 65,000 Ib/h + 240,800 Ib/h
= 304,800 Ib/h of freshwater (or = 610 gal/min)
The results are summarized in Figure 2.3-23.
2.3.9 Information on Several Operable Lime Absorber Systems
on Utility Boilers
This section describes six operating lime absorber systems
in use on utility boilers. Following an introduction to each
plant is a description of the FGD system and its major com-
ponents, together with a flow diagram. This information is
representative of the lime processes and equipment now in use.
2.3.9.1 Bruce Mansfield Station, Pennsylvania Power Co.—
Introduction—The Bruce Mansfield plant of Pennsylvania
power is a 2700-MW, three-boiler, coal-fired facility located on
the Ohio River in Shippingport, Pennsylvania. This facility was
built by Pennsylvania Power Co., which is acting on its own
behalf and as an agent for the other participating companies:
Cleveland Electric Illuminating Co., Duquesne Light Co., Ohio
Edison Co., and Toledo Edison Co.
Bruce Mansfield Nos. 1, 2, and 3 are once-through, super-
critical steam generators that fire 333 ton/h of coal and gener-
ate approximately 6.5 million Ib/h (each) of steam at 3785 psig
and 1005°F.
The units are rated at 825 MW each. The emission control
equipment is designed to meet state emission regulations of 0.6
Ib S02 per million Btu of heat input and 0.0175 gr/scf of parti-
culate when burning 11,900 Btu/lb coal having average ash and
sulfur contents of 12.5 and 4.7 percent, respectively. Addi-
tional design-related information is presented in Table 2.3-11.
FGD system—The following describes the FGD system for
Units 1 and 2 at Bruce Mansfield. Figure 2.3-24 is a flow
diagram of the FGD system showing the FGD equipment and con-
necting mass flows.
2.3-57
-------
to
I
en
oo
FRESH MAKEUP
UATCD 1 TWn IIHTNC
MAI tK ^ INlLUUlnQ
PUMP SEALS)
WATER IN THE
FLUE GAS
1
1
f
7
49<
11
! 240.900
lb/h
(482 gal/m1n
i
| ^
1
1
1
1
6
257,700 lb/h
I
1
1
1
1
1
1
1
ABSORBER
1
RECYCLE
TANK
24
0 lb/h
i.300 lb/h
1
1
15
150,000 lb/h
(300 gal/rain)
14
105,400 lb/h
(211 gal/min)
W
121 900 lb/h
20 383,100 lb/h
POND
12
320,700 lb/h
(641 gal/min)
WATER OUT WITH
FLUE GAS
13
55.300 lb/hf
SLURRY ,,
FFFO ^ I0 SLAKER
TANK 56,600
lb/h
10 64.000 lb/h
(128 gal/min)
FRESH
WATER
Figure 2.3-23. Water balance summary.
-------
Table 2.3-11. DESIGN BOILER AND FUEL DATA FOR
BRUCE MANSFIELD NOS. 1, 2, AND 3
Data
Boiler manufacturer
Year placed in service
Gross plant rating/unit
Net plant rating/unit
Heat rate at 825 MW
Heat input/unit
•
Boiler load range, % of capacity
Flue gas flow rate (full load) /unit
Sulfur content of coal
Ash content of coal
Heating value of coal
Item
Foster-Wheeler
No. 1 (1976),
No. 2 (1977),
No. 3 (in construction)
917 MW
825 MW
10,000 Btu/kWh
8,055 million Btu/h
25-100
3,308,000 acfm
4.7%
12.5%
11,900 Btu/lb
2.3-59
-------
to
•
CO
I
TO ABSORBERS 2-6
TO ABSORBERS 2-6
CALCILOX
FEED
HOPPERS
GRIT FROM
SLAKERS ~] \
SLUDGE
PREPARATION
TANK
SLAKER
TRANSFER
TANK
FLUE GAS
TO TRAIN 1
SETTLING BASIN
— BACKWASH
-) FROM
( OTHER
(ABSORBERS
FROM
OTHER
ABSORBERS
FLOCCULENT
- — " • I —
SCRUBBER
MIST
ELIMINATOR
WATER
DISTRIBUTION
BOX
-TO OHIO RIVER
ABSORBER
MIST ELIMINATOR
WATER
LOW
DISSOLVED
SOLIDS POND
TO
ASH HANDLING
SYSTEM
PUMP SEAL HATER
HIGH
DISSOLVED
SOLIDS POND
WASTE DISPOSAL SYSTEM BYPASS
LITTLE BLUE RUN
DISPOSAL AREA
TO BOTTOM-ASH SYSTEM
•*TO LIME AREA FLUSH
-••TO BOTTOM ASH SYSTEM
— TO SCRUBBER/ABSORBER AREA FLUSH
--TO THICKENER AREA FLUSH
Figure 2.3-24. Simplified flow diagram of Bruce Mansfield
-------
Unit 1 began full commercial operation on June 1, 1976 and
Unit 2, on October 1, 1977. Six separate venturi scrubber and
venturi SO2 absorber trains service each boiler. The system was
designed and furnished by Chemico Air Pollution Control Company
The system utilizes the Dravo Corporation's thiosorbic lime* as
the S02 absorbent. Both units are guaranteed to remove 92 per-
cent of the sulfur dioxide, and 99.8 percent of the particulate
matter from the flue gas.
Unit 3, now under construction, will be equipped with a
Kellogg/ Weir absorber, a horizontal, cross-flow spray unit that
can be operated at a flue gas velocity above 20 ft/s. The FGD
system will operate at 25 to 100 percent boiler capacity and no
bypass will be provided. Continued operation is ensured by an
extra absorbing stage on each module and by one entire spare
module. Estimated startup is April 1980.
Scrubber--Each scrubbing train consists of a variable-
throat, a particulate scrubbing venturi, a 9000-hp induced-draft
(ID) fan, and a fixed-throat venturi absorber in series. There
are six scrubber trains per boiler. The variable-throat venturi
removes most of the particulates. Additional particulate re-
moval is accomplished in the absorber. Sulfur dioxide is ab-
sorbed in both the particulate scrubber and the absorber. The
scrubber vessels are 35.5 ft in diameter and 52 ft high, with a
"plumb-bob" arrangement for the variable-throat mechanism. The
absorber vessels are 34 ft in diameter and 51.5 ft high. The
scrubber and absorber vessels are lined with polyester flake-
glass material. The ID fan housing is lined with rubber and the
fan rotors are made of Inconel 625. Information regarding the
venturi-scrubber and absorber is given in Table 2.3-12.
FGD system tanks—The FGD tank system consists of the
slaker transfer, lime recycle, distribution box, transfer,
underflow, and sludge preparation tanks.
Lime is fed directly by belts from day silos into two lime
slakers having maximum lime-slurry capacities of 100 gal/min.
The slaked lime is pumped to the 36-ft-diameter slaker transfer
tank, where it is retained for about half an hour. This tank
helps complete the lime slaking. The slurry is pumped to the
12-ft-diameter lime recycle tank located 3000 ft from the
slakers. The recycle tank supplies fresh slurry to each of the
scrubber and absorber vessels via individual branches. A por-
tion of the spent slurry is bled to the distribution box—a
mixing tank with zero-time retention. Here the slurry is mixed
with flocculant and fly ash slurry from the boiler prior to
entering the 200-ft-diameter thickener. Overflow from the
thickener goes to the transfer tank for reuse in the fly ash
* Dravo's patented lime contains 6 to 12 percent magnesium
oxide.
oxide.
2.3-61
-------
Table 2.3-12. BRUCE MANSFIELD SCRUBBER AND ABSORBER DATA
Data
Item
Particulate scrubber
Manufacturer
Type
L/G ratio
Scrubber pressure drop
across throat
Dimensions
S0_ absorber
Manufacturer
Type
L/G ratio
Absorber pressure drop
across throat
Dimensions
Chemico Air Pollution
Control Company
Variable throat venturi
40 gal/1000 acfm
25 in. KLO
35.5 ft dia. , 52 ft high
Chemico Air Pollution
Control Company
Fixed-throat venturi
40 gal/1000 acfm
4-7 in. H2O
34 ft dia., 51.5 ft high
2.3-62
-------
recovery system. Sludge from the thickener (30% solids) is
pumped to the adjacent 10r500-gal underflow tank, where the
sludge is pumped to the waste disposal system or to the onsite
ponds in emergency cases.
At the waste disposal area, two 100-ton hoppers (12 ft dia.
x 24 ft high) distribute Calcilox to the 176,000-gal sludge
preparation tanks (35 ft dia. x 35 ft high) before the sludge is
pumped to the Little Blue Run disposal site.
Reheat—Three absorber outlets combine into a single reheat
chamber. For each boiler there are two 25-ft-diameter reheat
chambers. The flue gas is reheated by three fuel-oil burners
from about 126°F to the designed 165°F before it exits through
the stack. Pennsylvania Power Company has yet to use the re-
heaters because of duct vibrations caused by resonance. The
reheaters are being redesigned to correct the resonance problem.
Plume buoyancy appears to be sufficient without reheat, but
atmospheric conditions sometimes cause condensation and precipi-
tation of moisture from the plume. Predominating winds often
cause liquid fallout to occur in Shippingport. The fallout
occurs as a clear liquid, but leaves a film upon drying.
Mist eliminator—Excessive mist carryover has occurred
during scrubber operations. The mist eliminators were designed
for 1 gr/scf liquid carryover, but plant personnel estimate
actual carryover at 3 gr/scf. The horizontal mist eliminators
installed as part of the vessels are also designed to operate at
gas velocities of 8 to 10 ft/s. Pennsylvania Power is investi-
gating installation of vertical mist eliminators in the ducting
downstream of the scrubber vessels. Because of duct diameter
and space restrictions, however, such an arrangement might
generate flow velocities as high as 50 ft/s. A trial vertical
mist eliminator was installed, but it collapsed as a result of
structural failure caused by high flue gas velocity.
Water system—The system is not being operated in a closed
loop as designed because water is being retained at the sludge
disposal site and not recycled to the process. Makeup water
from the disposal pond is not needed since fresh water is added
to the system in the fan sprays and during lime slaking. Plant
operators believe closed-loop operation is possible, but concen-
trated efforts to resolve the system's mechanical problems have
allowed insufficient time to demonstrate this possibility. In
the event they do not operate in a closed loop, the plant has a
permit to discharge water to the Ohio River.
2.3.9-2 Cane Run No. 4, Louisville Gas and Electric—4'5
Introduction—The Cane Run Power Station, located in Louis-
ville/ Kentucky, is operated by the Louisville Gas and Electric
Company (LG&E). The plant has six electric power steam genera-
ting units providing a total steam turbine net generating capa-
city of 992 MW.
2.3-63
-------
Unit No. 4 is a coal-fired boiler with a continuous net
generating capacity of 178 MW. Boiler and fuel data are pre-
sented in Table 2.3-13.
Table 2.3-13. BOILER AND FUEL DATA, CANE RUN NO. 4
Boiler capacity, MW (net)
Maximum generating capacity, MW
Unit heat rate, Btu/kWh
Sulfur content of coal, %
Ash content of coal, %
Heating value of coal, Btu/lb
178
190
10,030
3.5 - 4.0 (avg.)
11 - 12 (avg.)
11,500
The emission control system for this unit consists of an
ESP in front of a wet scrubbing system. The ESP provides pri-
mary particulate control; the wet scrubber provides additional
particulate removal and primary SO2 control.
FGD system—The FGD system consists of two identical paral-
lel scrubbing trains designed and installed by the American Air
Filter (AAF) Company. The wet scrubbing system was put into
operation in August 1976. The system uses calcium hydroxide
sludge (carbide lime), a waste byproduct generated in a nearby
acetylene plant. The system is guaranteed to remove 90 percent
of the sulfur dioxide and 99 percent of the particulate matter
from the flue gas.
The flow diagram (Figure 2.3-25) shows the FGD equipment
and connecting mass flows.
Absorber—Because the FGD system was not providing the
required SO2 removal, LG&E and American Air Filter modified the
unit to ensure its compliance with Federal and county standards.
As determined by LG&E tests, the following absorber modifi-
cations were made to achieve the required SO2 removal:
0 Increase of the L/G from 35 to 60.
0 Reduction of pressure drop by adding turning vanes in
the elbow and absorber base.
0 Installation of additional spray headers above the
mobile bed.
0 Replacement of the existing spray nozzles with ceramic
ones.
2.3-64
-------
to
•
U)
I
nut
MS
FLOttuuirr
TO POM
raw
HATH
mum
TO CIUMCLI
TMU
Figure 2.3-25. Flow diagram of Can Run No. 4.
-------
0 Proposed addition of underbed sprays to help circulate
the mobile bed balls.
The modules have three stages, with 1 ft of balls on each
tray and 5 ft between stages. The 1-1/4-in.-diameter balls are
made of polyurethane. The absorber is lined with Precrete and
Placite 4005.
The above-mentioned modifications enabled the system to
exceed the requirement for 85 percent SO2 removal (Jefferson
County) and to hold emissions below the Federal standard (1.2
lb/106 Btu). Test results indicated 86 to 89 percent SO2 re-
moval efficiency.
Quencher—Preceding each absorber is a quencher, which
consists of a wetted-wall conical frustum section in the duct.
Within the throat, several large-diameter injector nozzles are
located tangential to the flow of flue gas, along with an in-
ternal spray header flow that parallels the gas stream in the
center of the duct. These nozzles inject slurry into the gas
stream to ensure thorough wetting of the flue gases before they
enter the absorber. This minimizes possible scale formation
caused by the dry flue gas reaching the absorber internals.
Reaction tank—The reaction tank is constructed of rein-
forced concrete and is divided into three compartments. Slurry
flows from one compartment to the next through an opening in the
bottom of the separating walls and, in emergencies, over weirs
at the top of each compartment wall.
Reheat—Direct oil-fired heaters have been installed at the
base of the stack in conjunction with the turning vanes to help
correct the acid liquid carryover, which was attacking the
stack. The reheat, which adds 40° to 50°F, will increase the
exit flue gas temperature to 170° to 180°F.
ID fans—The booster fans are Buffalo Forge double-width,
double-inlet units rated at 367,000 acfm at 325°F. The fans and
fluid drive are powered by Reliance Electric induction motors
rated at 1250 hp at approximately 720 rpm. Because these boos-
ter fans have insufficient capacity to overcome the pressure
drop in the FGD system, the maximum boiler output has been
limited to 150 to 155 MW. Certain modifications have been made
to adjust the pressure drop. Turning vanes have been added in
the flooded elbow area, at the base of the absorber, above the
demister, and at the base of the stack. The radial vane demis-
ter has also been replaced by a chevron type. The booster fans
are designed to handle a pressure drop of 13 in. H20 through the
FGD system.
2.3-66
-------
Water removal and water system—Slurry is taken from the
bottom of the reaction tank feed section and pumped to the
85-ft-diameter thickener, in which flocculant is added to aid
settling. Sludge is removed from the bottom of the thickener
and pumped to the pond. Liquid from the upper level flows into
the thickener return tank and is punped to the return section of
the reaction tanks. Makeup water from the pond is added to the
thickener return tank, and an emergency overflow is provided
from the return tank to the pond.
2.3.9.3 Conesville No. 5, Columbus and Southern Ohio Electric
Co.6 —
Introduction—The Conesville Power Station is located on
the Muskingum River near Coshocton in northeast Ohio. The plant
has a present capacity of 1644 MW (design), and an addition with
capacity of 411 MW is under construction. Units 1, 2, and 3
have a combined capacity of 433 MW and share a common stack.
Unit 4 is rated at 800 MW, and Unit 5 at 411 MW. Unit 6, cur-
rently under construction, will also be rated at 411 MW. Units
4, 5, and 6 each have a separate'stack.
Boiler 5 is a dry-bottom, pulverized-coal-fired Combustion
Engineering unit, installed in 1976. Forty percent of the coal
is delivered by conveyor from a coal mine complex 7 miles away.
The remainder is hauled by truck from southeast Ohio. Boiler
and fuel data are presented in Table 2.3-14.
FGD system—The system at Conesville No. 5 consists of a
Research-Cottrell, cold-side ESP, followed by two Universal Oil
products (UOP) SO2 absorber modules in parallel. The ESP is
designed for 99.65 percent particulate removal, and the turbu-
lent contact absorbers are designed for 89.6 percent S02 re-
moval. The system is designed for an outlet SO2 loading of 1.0
Ib/million Btu of heat input. Boiler ID fans are located im-
mediately downstream from the ESP.
Table 2.3-14.
PERTINENT BOILER AND FUEL DATA FOR
CONESVILLE NO. 5
Boiler capacity, MW (rated)
Plant capacity, MW (design)
Boiler manufacturer
Sulfur content of coal, %
Ash content of coal, %
Heating value of coal, Btu/lb
441
1644
Combustion Engineering
4.2 to 5.1 (avg.)
12 to 19 (avg.)
10,300 to 11,220
2.3-67
-------
The flow diagram (Figure 2.3-26) shows the FGD equipment
and connecting mass flows. The material balance for this plant
is given in Table 2.3-15. The following paragraphs describe the
system.
Absorber—Following the ID fans, the flue gas enters two
parallel TCA absorbing trains. Each absorber is capable of
handling 60 percent of the flue gas flow. A presaturator sec-
tion, constructed of Carpenter 20, lowers the flue gas tempera-
ture from 286° to 125°F and provides some initial SO2 removal.
The gas then enters the neoprene-lined, carbon steel absorber
modules, where two stages of 1.5-in. plastic balls provide a
contacting surface for the lime slurry and the flue gas. The
stages are approximately 5 ft apart. The lower stage is com-
partmented to maintain uniform ball depth. Following each
absorber module, the flue gas passes through a fiberglass en-
trainment separator and two horizontal banks of chevron-type
mist eliminator. The bottom of the trap-out tray is washed
intermittently, and the lower mist eliminator is washed continu-
ally with recycled pond water. The flue gas from the parallel
absorber trains then enters the 800-ft, Ceilcote-lined stack.
Following the boiler ID fan, there is bypass breeching around
the entire scrubber loop. Each module can be bypassed inde-
pendently. No stack gas reheat is currently used, although it
is possible that reheat will be added in the future.
FGD system tanks—Dravo thiosorbic lime from Maysville,
Kentucky, is used in the UOP scrubber modules at a stoichio-
metric ratio of 1.1. The calcined, pelletized lime has a nomi-
nal particle diameter of 1.75 in., an MgO content of 3 to 8
percent, and a CaO content of 90 to 95 percent. The lime slaker
discharges the slurry (20 percent solids) into an agitated lime
slurry sump, where it is retained for 5 minutes before being
pumped to the lime slurry storage tank, which handles surge
requirements of the absorption system. Slurry is transferred
from the storage tank to the TCA recycle tanks by variable-speed
pumps, which respond to changing SO2 concentrations and boiler
loads by means of a pH monitor. The scrubbing liquor contains
about 7 to 12 percent solids and is recirculated by four pumps
(one standby), each rated at 12,000 gal/min. The pH at the
scrubber outlet is 5.8, and pH in the recycle tank is approxi-
mately 6.8.
A bleed stream of spent reaction products is withdrawn
continuously from the recycle tank and pumped to the thickener
system.
Wash water tanks supply the presaturator and demister.
Information regarding the FGD tanks is given in Table 2.3-16.
2.3-68
-------
to
•
U)
4 MAKE UP WATER
FLUE GAS "~CO
FROM ESPS
LIME
CaO 9H
MgO 5i
INERTS 41
SLAKER FEED BIN
SERVICE HATER
WASTE SLURRY TO
IUCS SLUDGE
TREATMENT
24 GRIT
Note: S.W. SEAL WATER INPUT TO PUMPS
Figure 2.3-26. Flow diagram of Conesville No. 5.
-------
Table 2. 3-15. MATERIAL BALANCE FOR CONESVILLE NO. 5
Description
(Gas)
Mass flow rate
103 lb/h
acfm 103
Temp., »c (°F)
Gauge press.
(in. H2O)
Fly ash, 103 Ib/h
SG-2 103,lb/h
C07. N2, 02
103 lb/h
HjO 103 lb/h
1
Total qas
for cleaning
4,440
1,394
147 (296)
8.0
.194
39
4,160
240
la and Ib
Gas for scrubber
A, B
2,220
697
147 (296)
8.0
.097
19.5
2,080
120
2a and 2b
Gas from scrubber
A, B
2,280
583
52 (126)
0.8
.077
2
2,078
200
3
Flue gas at
outlet battery limit
4,560
1,166
52 (126)
0.8
.154
4
4,156
400
M
•
U)
I
Description
(Water)
Haas
flow
rate
103 lb/h
gal/nin
Specific gravity
Tenp., -
•C v A, B
93.5
187
1.0
27 (80)
6a and 6b
Return
wash water
500
1000
1.0
52 (126)
7a and 7b
Wash water
pump discharged
597.5
1,195
1.0
52 (125)
8a and 8b
Deniater and
Wash water to
Scrubber A, B
545
1,090
1.0
52 (125)
9a and 9b
Makeup water
to presat.
52.5
105
1.0
52 (125)
lOa and lOb
Return water
to prenat.
82.5
165
1.0
38 (100)
Ua and lib
Total water
to presat.
135
270
1.0
33 (100)
12
Service water
to systejn
158.5
317 .
1.0
amb
13
Service water
to pump seals
64.5
129
1.0
amb
14
Service water
to slakers
94
188
1.0
amb
15
Return water
to slaker
158.3
317
1.0
38 (100)
16
Return water
from thickener
394.5
789
1.0
38 (100)
17
Return water to
draw-off pump
75
150
1.0
38 (100)
(continued)
-------
Table 2.3-15 (continued)
to
•
oo
Description
(Slurry)
Naaa 103 Ib/h
flow
rate gal/min
Solids, %
CaSOj • 1/2 H,0.
103 Ib/h 2
CaS04 . 2 H-0,
103 Ib/h z
MgS04.
103 Ib/h
Specific gravity
Te«p. , «c (T)
IB
Lime
to system
34.1
-
-
-
-
-
-
19a and 19b
SJurry to
Recycle Tank A, B
143.2
261
14.76
-
-
1.7
1.1
amb
20a and 20b
Recycle slurry
to Scrubber A, B
20.540
38,000
13
1,926
642
423
1.08
52 1126)
21a and 21b
Spent slurry
to draw-off pond
281.4
521
13
26
9
6
1.08
52 (126)
22
Thickener
feed »lurry
641.8
1200
11.57
52.8
17.6
13
1.08
52 (126)
23
Haste «ludqe
for disposal
251.3
420
29.5
52.8
17.6
5
1.22
38 (100)
24
Grit
4.2
-
-
-
-
-
-
Notei amb - ambient:
-------
Table 2.3-16. FGD TANK INFORMATION FOR CONESVILLE NO. 5.
Item
Unit size
Capacity
Material of
construction
Slurry transfer
tank
16 ft dia. , 10 ft high
15,825 gal
Fiberglass (FRP)
Slurry storage
tank
20 ft dia. , 22 ft high
51,700 gal
Carbon steel
Recycle tank
45 ft dia. , 28 ft high
332,930 gal
Carbon steel
Wash water tank
8 ft dia. , 10 ft high
3,600 gal
Fiberglass (FRP) steel
shell
NJ
-------
Waste system—The thickener is 100 ft in diameter and 14 ft
deep in the center. The reaction product slurry is fed continu-
ously from the recycle tanks and concentrated to an underflow
composition of approximately 40 percent solids. This underflow
is cycled to IU Conversion Systems, Inc., (IUCS) fixation facil-
ities, where it is further thickened, vacuum-filtered, and mixed
with a blend of dry fly ash and lime to form a 73 percent solid
substance (IUCS Poz-o-tec). The product is currently being
discharged to a 3500-acre-foot diked pond.
The wastewater pond received ash sluice water, cooling
water blowdown, and water from the sludge treatment plant. This
system is not operating in the closed-loop mode at present.
2.3.9.4 Green River Station, Kentucky Utilities7'8—
Introduction—The Green River Station of Kentucky Utilities
(KU) is located on the Green River in Central Kentucky, approxi-
mately 5 miles north of Central City. American Air Filter (AAF)
designed and installed a tail-end wet lime scrubbing system on
Boilers 1, 2, and 3. These boilers service two turbines rated
at 32 MW (gross) each. The station operates a total of four
steam turbines with a combined generating capacity of 242 MW.
All three boilers are dry-bottom, pulverized-coal-fired units
manufactured by Babcock and Wilcox, and were installed in 1949
and 1950. They are used for peak loads and are normally opera-
ted 5 days a week, with one or more of the boilers often at
reduced load. Present plans do not call for retirement of these
units.
Heat rate is approximately 13,250 Btu/net kWh per unit.
The boilers burn primarily a high-sulfur western Kentucky coal
that has an average heating value of 10,800 Btu/lb, a sulfur
content of 3.8 to 4.0 percent, and an ash content of 13 to 14
percent. Boiler and fuel data are presented in Tables 2.3-17
and 2.3-18.
FGD system—The FGD system at Green River was started up on
a half-load basis in September 1975. It continued to operate at
half load until March 1976, when operation began at full-load
capacity in a closed-water-loop mode. The FGD system uses
slaked lime for primary S02 removal. Sulfur dioxide removal
efficiency usually averages more than 90 percent, well above the
guaranteed design efficiency (80%).
Primary particulate removal is provided by Western Precipi-
tator multicyclones designed to operate at an efficiency of 85
percent. A variable-throat venturi scrubber designed to operate
at an overall efficiency of 99.7 percent provides additional
particulate removal. Under full-load conditions, the maximum
2.3-73
-------
Table 2.3-17. PERTINENT BOILER DATA, GREEN RIVER PLANT
Boiler data
Boiler
Boiler manufacturer
Year placed in service
Total generating capacity
Maximum heat input
Heat rate per unit
Percent excess air required
Percent boiler efficiency
Item
Nos. 1, 2, and 3
Babcock and Wilcox
1949, 1950
64 MW
848 106 Btu/h
13,250 Btu/net kWh
25
80
Table 2.3-18. FUEL DATA, GREEN RIVER PLANT
Fuel data
Type (primary)
Analysis
Heating value
Fuel consumption
High-sulfur western
Kentucky coal
3.9 percent sulfur
13.4 percent ash
12.1 percent total
moisture
10,800 Btu/lb
1/416 io6 short
tons/wk
2.3-74
-------
allowable particulate and sulfur dioxide emissions are 0,097
lb/106 Btu and 1.67 lb/106 Btu of heat input, respectively.
Actual emissions are unknown at this time because air leakage in
the boilers has caused monitoring to be indecisive. The flow
diagram in Figure 2.3-27 shows FGD equipment and connecting mass
flows on Green River Boilers 1, 2, and 3. The following para-
graphs describe the system.
Scrubber train—Flue gas desulfurization and particulate
removal systems are combined in a single scrubber module de-
signed to handle a maximum of 360,000 acfm of flue gas at 300°F.
This scrubber module contains a mobile bed contactor for SO2
removal and a variable-throat, flooded-elbow venturi for fly ash
removal.
The absorber is 20 by 20 ft, and 22.5 ft high, and is
constructed of mild steel with a 3/4-in.-thick, acid proof
Precrete lining. The internals consist of a mobile bed with 10
compartments. The mobile bed stage contains approximately
175,000 to 190,000 solid 1.25-in.-diameter balls, packed to a
maximum thickness of 2 ft (16 in. at rest). The balls are made
of polyvinyl chloride and polyethylene. Underbed dampers are
used to adjust for reduced removal requirements during periods
of low steam demand.
The variable-throat venturi scrubber is constructed of mild
steel with a stainless steel throat and acidproof brick lining.
It has a 100-in. throat opening with a 94-in. plug.
The original mist eliminator of stainless steel construc-
tion is a single-stage, spin-vane type positioned 27.5 ft above
the scrubber bed. The mist eliminator depth and vane spacing
are 3 ft. Gas flows through the demister first horizontally,
then vertically; pressure drop is 2 in. H2O. Kentucky Utilities
is attempting to optimize the operation of the mist eliminator.
If they fail, they intend to replace it with a standard design,
chevron-type mist eliminator.
Data regarding the scrubber are presented in Table 2.3-19.
FGD system tanks—The FGD system has a slaker tank, a
mix/hold tank, and a reaction tank.
A storage bin equipped with a vibrating bottom and an 8-in.
screw conveyor discharges lime at a rate of 2 tons/h into the
covered, agitated slaking tank. Two agitated slake tanks were
installed for this purpose, one serving as a backup. Each has a
liquid volume capacity of 1680 gal, and each is equpped with a
10-hp, 84-rpm agitator.
2.3-75
-------
303,000 acfm P 116 °F
WET
SCRUBBER
STACK
*MIST ELIMINATOR
WASH WATER 50 gal/min
BOOSTER
FAN
RECYCLE 10% SOLIDS Ca (OH)2 109 Ib/min
11,800 gal/min
Cax SOX (APP.) 7,380 Ib/min
REACTANT ADDITION
20% SOLIDS
53 gal/min HjO 6 Ib/min GRIT
111 Ib/min Ca (OH)2
LIME STORAGE
25 gal/min PUMP SEALSI
10 gal/min FAN
BEARING COOLING
REACTION/
(TANK,
Ca (OH)2
Ca S0_
228 gpm BLEEDJO_POND,
9 Ib/min
190 Ib/min
105! SOLIDS
-€7
RECYCLE
SLAKING WATER
3 gal/min
DILUTION WATER
50 gal/min
HOLD TANK
*REFER TO TABLE 5
Figure 2.3-27. Flow diagram of Green River Station.
2.3-76
-------
Table 2.3-19. GREEN RIVER SCRUBBER DATA
Data
Item
Particulate scrubber
Type
Manufacturer
Scrubber pressure drop
(in. H20)
Dimensions
Material of construction
Shell
Shell lining
No. of stages
Nozzle size
No. of nozzles
SO absorber
Type
Manufacturer
L/G ratio, gal/1000 ft3
Absorber pressure drop
(in. H20)
Dimensions
Material of construction
Shell
Shell lining
Internals
Type
No. of stages
Packing thickness
Variable throat venturi
at flooded elbow
American Air Filter
7
Throat 8 ft 4 in.,
plug, 7 ft 10 in.
Mild steel with
stainless steel throat
Acidproof Precrete
1.5-in.orifice,
1360 gal/min in venturi
690 gal/min in damper
12
Mobile bed contactor
American Air Filter
39.5
4
20 by 20 ft, and 27.5 ft high
Mild steel
0.25-in. acidproof Precrete
Mobile bed (ping-pong balls)
1 (10 compartments)
1.33 ft at rest/
20 ft in use
2.3-77
-------
Slurry is discharged from the slaker to a mix/hold tank.
This tank, which has a liquid volume capacity of 1980 gal, is
also agitated (5-hp, 45-rpm agitator). The fresh scrubbing
slurry (20 percent solids) is then pumped to the return section
of the reaction tank installed beneath the scrubbing module.
The reaction tank, constructed of acidproof concrete, is 72
by 24 ft, and 24 ft high. Two partitions form three individual
compartments, each agitated and connected by underflow openings.
The total liquid capacity of each compartment is 100,000 gal.
Total retention time in the reaction tank is 21 min (7 min per
compartment). The function of each compartment is descibed
below:
0 The return section of the reaction tank system re-
ceives the reaction products and collected fly ash
discharged from the scrubbing module. Fresh lime
slurry, fresh makeup water (cleaned river water), and
pond return water are also supplied to the system at
this point.
0 The recycle/discharge section of the reaction tank
feeds both the venturi scrubber and the mobile bed
contactor with recycled scrubbing solution. Bleed
pumps remove the scrubbing wastes from this section of
the reaction tank so that a 10 percent solids slurry
will be maintained in the tank. The bleed stream is
discharged to a settling pond, and clear water is
returned from the pond to the return section of the
reaction tank.
0 One additional section, situated between the return
and recycle sections of the reaction tank, was in-
stalled only for the purpose of providing additional
checks on the process chemistry.
Table 2.3-20 summarizes reaction tank data.
2.3-78
-------
Table 2.3-20. REACTION TANK DATA, GREEN RIVER PLANT
Data
Item
Materials of construction
Configuration/dimensions
Capacity, gal
Retention time
Covered
Agitators
Materials of construction
Horsepower
Acidproof concrete
Rectangular - 3 compartments
24 by 24 ft, and 24 ft high
103,680 per compartment
7 min per compartment;
21 min total
No
1 agitator per compartment
Rubber-lined agitator
50 hp, 45 rpm, 84-in.-dia.
turbine agitator
Disposal and water system—R^-t-i ™ products and collected
particulate matter are pumped to an impervious, clay-lined pond
located on the plant site approximately 0.8 mile from the scrub-
bing module. The pond capacity is 148 acre-ft at a depth of 20
ft. Calculated life expectancy of this pond is approximately 9
years, but it is expandable to 414 acre-ft, which would yield
another 20 years. The closed-loop operation returns clarified
pond water to the reaction tank. Treated river water is used as
makeup and is introduced into the reaction tank, lime slaking
tank, and demister, as well as the various pump seals and fan
bearings. The makeup water requirements are listed in Table
2.3-21.
Table 2.3-21.
MAKEUP WATER REQUIREMENTS, GREEN RIVER PLANT
(gallons/minute)
Lime slaking
Slaker
Mix/hold tank
Pump seals
Fan bearing cooling
Demister wash water
Total
3
50
25
10
50
138
2.3-79
-------
2.3.9.5 Paddy's Run Station, Louisville Gas and Electric9'10—
Introduction—The Paddy's Run Station of Louisville Gas and
Electric Company (LG&E) is used primarily to meet peak loads.
This station has six generators, but only the boiler on Unit 6
is retrofitted with an FGD system. This dry-bottom, pulverized-
coal-fired boiler was installed by Foster-Wheeler in 1951. It
is rated at 65 MW, but runs at 71 to 72 MW at full load. The
heat rate ranges from 13,000 to 13,500 Btu/kWh. The boiler
burns Peabody high-sulfur coal, which has an average heating
value of 12,400 Btu/lb, ash content of 15 percent, and a sulfur
content of 3.7 percent. Boiler and fuel data are presented in
Tables 2.3-22 and 2.3-23.
FGD system—The FGD system, a lime scrubber designed by
Combustion Engineering, Inc., was put 'into operation in April
1973. The system uses calcium hydroxide sludge (carbide lime),
a waste byproduct generated in a nearby acetylene plant. The
slurried mixture of this carbide lime constitutes the replenish-
ing fresh scrubbing slurry. The system meets the required >85
percent SO2 removal efficiency, and sometimes operates at effi-
ciencies greater than 99 percent.
A Research-Cottrell ESP, which operates at 99 percent
efficiency, provides primary particulate removal and keeps the
boiler in compliance with the maximum allowable rate of 0.1
lb/106 Btu of heat input. Continuous monitoring equipment shows
that Unit 6 is also in compliance with regulations limiting
atmospheric emission of S02 to 1.2 lb/106 Btu of heat input.
The flow diagram in Figure 2.3-28 shows Unit 7 FGD equip-
ment and connecting mass flows. The following paragraphs de-
scribe the system.
Absorber—The FGD system consists of two identical absorber
modules, each sized to handle 175,000 acfm at 350°F. The absor-
bers are constructed of a mild steel coating with a fiberglass
reinforced polyester (FRP) flake lining 1/2 in. thick. Internal
supports are Type 316 stainless steel. Each absorber contains
two beds of 1-in. glass marbles. The packing thickness of each
bed is 3 in. The thickness of the layer is controlled by the
height of the overflow pots.
Atop the absorbers are two-stage chevron mist eliminators
followed by gas reheaters. The modules are 17 by 18 ft, and 50
ft high. Table 2.3-24 presents additional absorber data.
FGD systems tanks—The FGD system has an additive slurry
tank, a reaction tank, and a reaction surge tank. All three are
constructed of mild steel.
2.3-80
-------
Table 2.3-22. BOILER DATA FOR UNIT 6,
PADDY'S RUN STATION
Boiler data
Item
Boiler manufacturer
Year placed in service
Unit rating
Unit rating at full load
Maximum heat input
Maximum continuous heat input
Maximum flue gas rate
Percent excess air required
Heat rate
Foster-Wheeler
1951
65 MW (nameplate),
70 MW (maximum continuous, net)
71 to 72 MW
910 million Btu/h
810 million Btu/h
400,000 acfm at 325°F
25 to 30
13,000 to 13,500 Btu/kWh
Table 2.3-23. FUEL DATA FOR UNIT 6,
PADDY'S RUN STATION
Fuel data
Heating value, Btu/lb
Fuel consumption at maximum
heat, input, Ib/h
Fuel consumption at maximum
continuous heat input, Ib/h
Item
Peabody high sulfur coal
14 percent ash,
3.7 percent sulfur
12,400
73,400
65,300
2.3-81
-------
to
00
GAS
REHEAT
*WASH
WATER ~
FLUE GAS
325,000 acfm-
at 335°F
*WASH
— >
— >
126°F 67
ABSORBER
50gal/m1n(TOTALh 6750
WAIhK
gal/min (TOTAL*
i
*
GAS
REHEAT
126°F
ABSORBER
8100 gal/min(TOTAL)
LIME STOICHIOMETRIC
RATIO (IN ABSORBERS):
1.04 TO 1.05
I 1350gal/min
*WATER 400 ga
170 gal/min
10% SOLIDS
FLOCCULANT
LIME
lOOlb/ton OF
DRY SLUDGE
1350 gal/min "
10% SOLIDS
REACTION
SURGE
TANK
4800 Ib/h DRY
1 Ca (OH)2CARBIDE LIME
ADDITIVE
SLURRY
TANK
SLURRY
20-30% SOLIDS
20-25%
SOLIDS
FILTER
*" CAKE
*REFER TO TABLE 7
2 FILTERS
lOton/h EA. SOLIDS
Figure 2.3-28. Flow diagram of Paddy's Run.
-------
Table 2.3-24. ABSORBER DATA, PADDY'S RUN
Type
L/G ratio, ,
gal/1000 ft at 126°F
Gas velocity through absorber,
ft/s
Material of construction
Shell
Internals
Internals
Type
Number of stages
Type and size of packing
Packing thickness per stage
Material of construction
Packing
Supports
Absorber train pressure drop
AP across each marble bed (in. H O)
AP across tower (in. HO)
AP across demister (in. H?O)
AP due to ductwork (in. H_0)
Total AP (in. HO)
,^...^^^^^^MI^MMM^MBHlH^HiMHHHMHHHMtaH^MH
Tower
15 to 18/stage
8 to 12
Mild steel with an
FRP flake lining
2-1/2 in. thick
316 stainless
Marble bed
2
Glass marbles - 1 in,
3 in.
Glass
316 stainless
5.5 to 6
11 to 12
1.5
3 to 4
15 to 18
2.3-83
-------
Stockpiled carbide lime is diluted from a 50 percent solids
mixture to a 20 to 30 percent solids mixture before it is fed to
the additive slurry tank, which serves as a holding tank.
During full-load operations, surge capacity is 2-1/2 hours.
"Black lime," a byproduct from the lime kiln scrubber of an
acetylene plant, is added to the carbide lime in the additive
slurry tank. The analysis of black lime is essentially the same
as that of carbide lime, except that the magnesium oxide (MgO)
content is higher (2 to 4 percent) and the lime contains less
Ca(OH)2 and more CaCO3. The carbide lime analysis is presented
in Table 2.3-25.
Table 2.3-25.
CARBIDE LIME ANALYSIS, PADDY'S RUN
(percent)
Solids analysis
Ca(OH)2
CaCO3
SiO2
C
S
MgO
Cl
90-92
3-8
2-2.5
0.3
0.03
<0.1
trace
Mass flows of carbide lime slurry from the additive tank,
slurry from absorber/ water from the thickener, filtrate from
the vacuum filter, and makeup water are fed into the reaction
tank, where they are mixed by mechanical agitators. Under
full-load conditions, the mixture is retained for 20 minutes
before it is pumped to the reaction surge tank. The slurry is
retained 3 minutes in the reaction surge tank, then sent to the
absorber spray nozzles., The location of the reaction surge tank
downstream from the reaction tank ensures proper mixing and
precipitation of scale before the slurry is used in the absor-
bers. Table 2.3-26 presents FGD tank data.
Thickener—Slurry bled from the absorber into the thickener
has a solids content of 9.5 to 10.5 percent. Lime is added at
this point to stabilize the sludge. The lime consumed for this
purpose amounts to about 100 Ib/ton of dry sludge solids genera-
ted. Flocculant is injected into the thickener to aid settling
2.3-84
-------
Table 2.3-26. FGD TANK DATA, PADDY'S RUN
ro
•
CO
Item
Size and capacity
Retention time at
full load
Solids concentration.
Material of construc-
tion
Additive slurry
tank
8 ft dia., 17 ft high
(6,400 gal)
2-1/2 h
20 to 30
Mild steel
Reaction surge
tank
20 ft dia., 15 ft high
(35,200 gal)
3 min
10
Mild steel
Reaction
tank
48 ft dia., 17 ft high
(210,000 gal)
20 min
10
Mild steel
00
Ul
-------
Table 2.3-27. THICKENER AND VACUUM FILTER DATA, PADDY'S RUN
Thickener
Number
Dimensions and capacity
Solids concentration, %
Retention time at full load
Material of construction
Rotary vacuum filter
Number
Cloth area/filter
Capacity
Solids concentration, %
Precoat
50 ft dia . , 14 ft high
(205,500 gal)
10 in
20-24 out
4.3 h
Mild steel
150 ft /filter
10 tons/h {wet cake)
20-24 in
35-45 out
none
2.3-86
-------
by maintaining a concentration of 4 to 7 ppm Mass
to - 4-r
overflow water to the reaction tank. Table 2 3-27 presents
regarding the thickener and the vacuum filter. Presents
weir water is returned to the thickener for
Table 2.3-28. WATER REQUIREMENT, PADDY'S RUN
(gallons per minute)
Dilution of carbide lime feed to additive tank
and replenish losses'via
350
Demister wash water
Pump seals
Other
T°tal
390-400
2.3.9.6 Phillips Power Station, Duquesne Liaht Co ii' 12*13/14
Introduction-The Phillips Power Station of' the Duquesne
Lxght Company is located on the Ohio River in Allegheny Coun?y
Pennsylvania 20 miles northwest of Pittsburgh The plant
consists of six generating units having a total gross continues
generating capacity of 408 MW. The net station capacity is 111
m W^T™ ^ scrubber modules and the absorbed module are
No6 is the largene.rat0rS *" cycii»<3 > base-load units.
e
unt No6 s the lar. .
?2J M£ All tS hi ^9e generator< having a net capacity of
143 MW. All the boilers, manufactured by Foster-Wheeler are
dry-bottom, pulvenzed-coal-fired units. Y The first unit was
installed in 1942, and the sixth unit in 1956.
sis, is 18.2 percent, and sulcnt i 215
Boiler data are tabulated in Table 2.3-29.
2.3-87
-------
Table 2.3-29. BOILER DATA, PHILLIPS POWER STATION
Boiler data
Item
Boiler capacity, MW (gross)
Units 1,2
3,4,5
6
Boiler manufacturer
Years placed in service
Maximum continuous generating
capacity, MW (net)
Maximum heat input, 106 Btu/h
Heat rate, Btu/kWh (net)
Maximum flue gas rate, acfm
at 360°F
35
65
148
Foster-Wheeler
1942-1956
373
4,463
11,900
1,650,000
FGD system—The lime FGD system at the Phillips Power
Station was the first in the United States. The Phillips scrub-
ber system began operation in July 1973. Primary particulate
emission control is provided by mechanical collectors followed
by an ESP on each boiler. Downstream, final particulate control
is achieved by four parallel Chemico venturi scrubbers with
design efficiencies of 99 percent. These scrubbers also remove
approximately 50 percent of the SO2 entering the system. Emis-
sions of S02 from one scrubber are further controlled by a
Chemico second-stage venturi absorber, with lime as the absor-
bent. Performance guarantee tests indicate that the SO2 removal
efficiency of this two-stage train has averaged about 90 per-
cent. The system's overall SO2 removal efficiency is 50 to 60
percent.
The maximum particulate emission is limited to 0.08 lb/106
Btu of heat input. Particulate emissions from the unit are in
compliance with that regulation.
Atmospheric emissions of S02 are limited to 0.6 lb/106 Btu
of heat input. Present SO2 emissions from the single-stage
scrubbing system, using high-calcium lime, exceed this limit.
Bringing the system into compliance will require an SO2 removal
efficiency of 83 percent.
2.3-88
-------
Extensive tests conducted from October through December
1975 showed that the necessary SO> removal could be achieved
with the existing single-stage scrubber trains by using thio-
sorbic lime. Although original plans were to achieve compliance
with dual-stage scrubbers, Duquesne Light has notified the
regulatory agencies that it will operate the existing single-
stage scrubbing system with Dravo's Thiosorbic lime.
The flow diagram in Figure 2.3-29 shows FGD equipment and
connecting mass flows of the dual-stage scrubber train. The
following paragraph describes the system.
Scrubber train—All four scrubber trains are equipped with
Chemico variable-throat venturi scrubbers for removal of fly ash
and SO2. The fourth train has an added second venturi for
increased SO2 absorption. The system was designed to handle a
total gas volume of 2,190,000 acfm with all four trains in
service. The cleaned gases exiting the trains enter a common
wet duct (also lined with Ceilcote) that leads to a 340-ft,
acid-resistant, brick-lined, concrete stack. A 316L stainless
steel section of the duct preceding the stack is equipped with a
direct oil-fired reheater unit that can raise stack gas tem-
perature in the range of 110° to 120°F by as much as 30°F.
Normal reheat is about 20°F. Information regarding the scrub-
bers is given in Table 2.3-30.
Table 2.3-30. PHILLIPS POWER SCRUBBER DATA FOR
PARTICULATE AND FGD SCRUBBER MODULES
Superficial gas
velocity, ft/s
Module size
Equipment internals
Material of construction
Shell
Internals
40
40 ft dia., 50 ft high
Venturi
Mild steel, Ceilcote liner
Some 316L stainless, Ceilcote liner
ID fans—Gases leaving the scrubbers enter booster ID far^s
equipped with freshwater sprays to remove any accumulation of
solids resulting from scrubber carryover. The ID fan housings
are lined with 1/4-in. thick natural rubber. Wheel material is
Carpenter 20 Cb 3, a stainless steel containing niobium and
tantalum. The fan shaft is 316L stainless steel. The spray
nozzles have been relocated and replaced by a new type (Bete fog
nozzle No. TF16FC), and a Neoprene-29 rubber coating has been
applied to the 316L SS fan hubs. Each fan is driven by a
5500-hp motor. A closed system supplies cooling water to the
fan bearings.
2.3-89
-------
FLUE GAS
FLUE GAS TO ABSORBER
DEHISTER
WASH MATER
MAKEUP
WATER
TO THICKENER
to
vo
o
TO THICKENER
TROUGH
GRIT
LIME
SLURRY
PUMP
Finure 2.3-29.
clow diaqram of the dual-stage scrubber train at Phillips.
The complete scrubbing system includes four trains, three
have first stages with mist eliminators instead of second stages
-------
Sludge disposal— Phillips Power Station will be using a new
ESS, -- «
or ne
900-gal/Min 100-hp pumps taking suction Irom the
provides water for the fan
. —^^^..j,. j. u po-uviaes water for the fan ^nrav
pump seal, demister spray, instrumentation flush, reagent mixing
tank, and the emergency water supply for the scrubber! *
2.3.10 Conversion Factors
;rsi\nS\h^0nrePpo
-------
Table 2.3-31. CONVERSION FACTORS
Multiply
Atmospheres (atm)
Barrel - oil (bbl)
British thermal units (Btu)
Btu/minute (Btu/min)
Centimeters of mercury (cm Hg)
Cubic centimeters (cm )
By
76.0
29.92
33.90
1.0333
14.70
1.013
42
0.2520
777.5
3.927 x 10"4
107.5
2.928 x 10~4
12.96
0.02356
0.01757
17.57
0.01316
0.4461
136.0
27.85
0.1934
3.531 x 10"5
6.102 x 10~2
1.308 x 10~6
2.642 x 10"4
To obtain
cm Hg
in. Hg
ft H-O
2
kg/cm
lb/in.2
bars
gal - oil
kg - cal
ft - Ib
hph
kg - m
kWh
ft - Ib/s
hp
kW
W
atm
ft H2O
kg/m2
lb/ft2
lb/in.2
ft3
in.
yd3
gal
(continued)
2.3-92
-------
Table 2.3-31. (continued)
Multiply
— -
Cubic feet (ft )
Cubic feet/minute (ft /min)
Cubic feet/second (ft /s)
Cubic meters (m )
Feet (ft)
Feet of water (ft
Foot-pounds (ft-lb)
(continued)
~'
By
2.832 x 104
1728
0.02832
0. 03704
7.48052
28.32
472.0
0.1247
0.4720
62.4
0.646317
448. 831
35.31
61.023
1.308
264.2
30.48
0.3048
0.02950
0.8826
0.03048
62.43
0.4335
1.286 x 10~3
5.050 x 10"7
To obtain
cm
in.3
m
yd3
gal
liters
cm /s
gal/s
liters/s
Ib H20/min
million
gal/day
gal/min
ft3
in.3
ya3
gal
cm
m
atm H~0
in. H20
kg/cm
lb/ft2
lb/in.2
Btu
hph
2.3-93
-------
Table 2.3-31. (continued;
Multiply
Foot-pounds
Foot-pounds/minute (ft-lb/min)
Foot-pounds/second (ft-lb/s)
Gallons (gal)
Gallons water (gal H20)
Gallons/minute (gal/min)
Gallons H20/minute (gal H2O/min)
Grams (g)
By
3.241 x 10~4
0.1383
3.766 x 10~7
1.286 x 10~3
0.01667
3.030 x 10~5
3.241 x 10~4
2.260 x 10"5
7.717 x 10~2
1.818 x 10~3
1.945 x 10"2
1.356 x 10~3
3785
0.1337
231
3.785 x 10"3
4.95 x 10~3
3.785
8.3453
2.228 x 10~3
0.06308
8.0208
6.0086
2.205 x 10~3
To obtain
kg - cal
kgm
kWh
Btu/min
ft-lb/s
hp
kg-cal/min
kW
Btu/min
hp
kg-cal/min
kW
3
cm
ft
3
in.
m3
yd3
liters
Ib H20
ft3/h
liters/s
ft3/h
tons H2O/d
Ib
(continued)
2.3-94
-------
Table 2.3-31. (continued)
Multiply
Grams/cubic centimeter g/cm
Grams/liter (g/liter)
Horsepower (hp)
Horsepower (boiler)
Horsepower-hours (hph)
Inches (in.)
inches of mercury (in. Hg)
By
62.43
0.03613
8.345
0.062427
42.44
33,000
550
1.014
10.70
0.7457
745.7
33.479
9.803
2547
1.98 x 104
641.7
2.737 x 105
0.7457
2.540
0.03342
1.133
0.03453
70.73
0.4912
To obtain
lb/ft3
lb/in.3
lb/1000 ge
lb/ft3
Btu/min
ft-lb/min
ft-lb/s
hp (metric
kg-cal/mir
kW
W
Btu/h
kW
Btu
ft-lb
kg-cal
kg-m
kWh
cm
a tin
ft H90
2
kg/cm
lb/ft2
lb/in.2
(continued)
2.3-95
-------
Table 2.3-31. (continued)
Multiply
Inches of water
Kilograms (kg)
2
Kg/cm
Kilowatts (kW)
Kilowatthours (kWh)
Li v.ers
By
0.002458
0.07355
0.002540
5.202
0.03613
2.205
1.102 x 10~3
0.9678
32.81
28.96
2048
14.22
56.92
4.425 x 104
737.6
1.341
14.34
103
3415
2.655 x 104
1.341
850.5
3.671 x 105
103
0.03531
To obtain
a tin
in. Hg
kg/cm
lb/ft2
lb/in.2
Ib
tons (short)
atm
ft H20
in. Hg
lb/ft2
lb/in.2
Btu/min
ft-lb/min
ft-lb/s
hp
kg-cal/min
W
Btu
ft-lb
hph
kg-cal
kg-m
cm
ft3
(continued)
2.3-96
-------
Table 2.3-31. (continued)
================================
Multiply
Meters (m)
Meters/minute (m/min)
Meters/second (m/s)
Pounds (Ib)
=================================
By
61.02
ID'3
1.308 x 10~3
0.2642
100
3.281
39.37
ID'3
103
1.094
1.667
3.281
0.05468
0.06
0.03723
196.8
3.281
3.6
0.06
16
256
7000
0.005
453.5924
1.21528
14.5833
=========================
To obtain
in 3
m
ya3
gal
cm
ft
in.
km
mm
yd
cm/s
ft/min
ft/s
km/h
mph
ft/min
ft/s
km/h
km/min
oz
drams
gr
tons (short)
g
Ib (troy)
oz (troy)
(continued)
2.3-97
-------
Table 2.3-31. (continued)
..
Multiply
Pounds of water (lb/H2O)
Pounds of water/minute
(Ib H20/min)
Pounds/cubic foot (Ib/ft )
2
Pounds/square inch (Ib/in. )
Temp (°C)+273
Temp (°C)+17.78
Temp (°F)+460
Temp (°F)-32
Tons (long)
By
0.01602
27.68
0.1198
2.670 x 10~4
0.01602
16.02
5.787 x 10~4
0.06804
2.307
2.036
0.07031
1
1.8
1
5/9
1015
2240
1.12000
To obtain
ftJ
m
gal
ft3/s
g/m3
kg/m3
Ib/in.3
atm
ft H2O
in. Hg
kg/cm
abs. temp (°c)
temp (°p)
abs. temp (°F)
temp ( ° c )
kg
Ib
tons (short)
(continued)
2.3-98
-------
Table 2.3-31. (continued)
Multiply
Tons (metric)
Tons (short)
Tons of water/24 h
By
103
2205
2000
32000
907.18486
2430.56
0.39237
29156.56
0.90718
83.333
0.16643
1.3349
To obtain
kg
Ib
Ib
oz
kg
Ib (troy)
tons (long)
oz (troy)
tons (metric)
Ib of water
gal/min
ft3/n
2.3-99
-------
REFERENCES
1. 1976 Generation Planbook. Pacemaker Plants/Mansfield.
Designing Large Central Stations to Meet Environmental
Standards, McGraw-Hill, Inc. pp. 25-34.
2. Laseke, B.A. EPA Utility FGD Survey: December 1977 -
January 1978. EPA 600/7-78-051a, March 1978.
3. Durker, K.R. Survey of Pennsylvania Power's Bruce Mans-
field Power Generating and Flue Gas Desulfurization System.
U.S. Environmental Protection Agency, Research Triangle
Park, July 14, 1977.
4. Laseke, B.A. EPA Utility FGD Survey: December 1977 -
January 1978. EPA 600/7-78-051a, March 1978.
5. American Air Filter - Better Air is our Business. Cane Run
FGD Plant Description. 1976. pp. 1-63.
6. Laseke, B.A. EPA Utility FGD Survey: December 1977 -
January 1978. EPA 600/7-78-051a, March 1978.
7. Beard, J.B. Scrubber Experience at the Kentucky Utilities
Company Green River Power Station. Environmental Technolo-
gist Kentucky Utilities Company, 1976.
8. Laseke, B.A. EPA Utility FGD Survey: December 1977 -
January 1978. EPA 600/7-78-051a, March 1978.
9. VanNess, R.P., and J. Jonakin. Paddy's Run No. 6 S02 Re-
moval System - a status report. Presented at the 12th
annual Purdue University Air Quality Conference, Louisville
Gas and Electric Co., Indianapolis, Indiana, November 8,
1973.
10. Laseke, B.A. EPA Utility FGD Survey: December 1977 -
Janauary 1978. EPA 600/7-78-051a, March 1978.
2.3-100
-------
CONTENTS
2.4 SLUDGE DISPOSAL
Page
2.4.1 Introduction
£. * ft~*i
2.4.2 Environmental and Land-use Impacts 2.4-1
2.4.3 Disposal Practices 2 4_4
2.4.3.1 Ponding
2.4.3.2 Landfilling -> ± \
2.4.3.3 Chemical Fixation 248
2.4.3.4 Alternative Disposal Methods 2!4-14
2.4.4 Economics
2.4—16
References 2.4-23
Bibliography 2 4-25
2.4-i
-------
2.4 SLUDGE DISPOSAL
2.4.1 Introduction
Lime FGD systems reduce the quantity of air pollution from
coal combustion; however, they also produce a large amount of
sludge that can create a solid waste and/or water pollution
problem. This section is concerned with the techniques of
sludge disposal, the associated environmental impacts in terms
of land use and water pollution, and the reported costs of
disposal. The major studies on FGD sludge disposal have been
reviewed giving particular attention to environmental impacts,
disposal practices, and disposal costs.
Analyses show that FGD sludge can contain toxic trace
elements; therefore one environmental concern is that sludge
leachate is a possible source of groundwater contamination.
Another concern stems from the large amounts of sludge generated
and the extensive land required for its disposal. This is a
serious problem in areas where land is at a premium. Land
reclamation is still another related concern.
Several sludge disposal methods are available and currently
in use. However, limited knowledge regarding the degree of
protection provided versus cost makes selection of the proper
technology a difficult task. Absence of direct U.S. Environ-
mental Protection Agency (EPA) regulations adds to selection
problems.
The assessment indicates that ponding or landfilling of raw
sludge in a disposal site lined with some impermeable material
may provide adequate environmental protection against leaching.
However, questions regarding life expectancy of lining materials
and the effect this type of disposal may have on land reclama-
tion are still unanswered. Chemical fixation of the sludge
before disposal appears to be the best technique available, in
that it provides permanent protection, reduces disposal volume,
and facilitates reclamation.
Alternative sludge disposal methods such as use of the
sludge in byproducts show promise, but they have not been suffi-
ciently developed to permit full-scale utilization.
2.4.2 Environmental and Land-use Impacts
Coal-fired electric generating stations commonly dispose of
their solid wastes by ponding or landfilling. Pollution of
ground or surface waters is a potential hazard, especially when
toxic trace elements are present, and coal ash and sludge from
FGD systems normally contain varying amounts of trace elements.
2.4-1
-------
The composition of the waste product (sludge) that is
discharged from the FGD system is a function of the particular
FGD process, ash removal practice, and coal composition. Up to
1.4 million tons/yr of fly ash and sludge can be produced by a
1000-MW plant, which means tremendous quantities of waste
material must be disposed of. The total quantity could amount
to as much as three times the tonnage of fly ash normally re-
moved from a power station.1
It is difficult to predict accurately the amount of sludge
that will be produced by an FGD system because numerous vari-
ables affect the quantity and quality of scrubber sludge. Major
variables include FGD efficiency, the amount of coal fired,
impurities in the coal, sulfur and ash content of the coal, the
amount of lime used, and particulate removal efficiency.2
Disposing of sludge is a difficult task not only because of
the massive quantities involved but also because of its gener-
ally poor chemical and physical properties. The physical pro-
perties of FGD sludge can affect handling and disposal methods
and also any possible future land use. Primary sludge consti-
tuents that affect the chemical and physical properties are
water, fly ash, calcium sulfate, and/or sulfite. Raw sludge
drained from an FGD system can contain as much as 85 to 95
percent water; thus, water is the major component in the volume
of waste to be disposed of and affects its physical properties.
Fly ash is a source of trace elements, and calcium sulfite can
create dewatering difficulty and produce undesirable physical
characteristics.
Generally, FGD sludge components from lime FGD systems
consist of calcium sulfite hemihydrate, calcium sulfate dihy-
drate, fly ash, and unreacted absorbent. The relative amounts
of each depend on many factors, including the kind and amount of
fuel burned, the efficiency of sulfur dioxide and particulate
removal, the purity of the lime, and the boiler type and opera-
ting practices. Sulfate sludges are more easily dewatered than
sulfite sludges and thus result in smaller volumes to be
handled. Generally, the higher the water content in the sludge,
the less desirable are its physical characteristics. Further-
more, sulfate sludges are nonthixotropic, whereas sulfite
sludges are thixotropic. Thixotropic materials will reliquify
upon agitation, which affects structural properties and sub-
sequent land utilization.2
Chemical characteristics of FGD sludge are a function of
elements in the coal, the scrubber absorbent, and operating
parameters of the system. Toxic trace elements are of great
concern in sludge disposal. Most trace elements in FGD sludge
originate in the coal and are carried to the sludge by the fly
ash and combustion gases. Even if the practice is to remove fly
2.4-2
-------
ash prior to scrubbing, some fly ash and trace elements are
inevitably carried over to the sludge. The absorbent and the
scrubbing water are also minor sources of trace elements.
The major toxic trace elements found in fly ash and FGD
?in ^ a/e mercur?' 2inC' arsenic' le*d, and selenium. Because
FGD sludge contains trace elements, its disposal presents a
potential hazard to both ground and surface walers; how much of
a hazard depends on the solid characteristics, weather, topo-
graphy, and proximity of ground and surface waters.
n . overflow of sludge disposal ponds and runoff from sludge
landfills can pollute nearby surface waters (lakes, streams?
rivers), thereby providing pathways through which trace element^
and dissolved solids can enter into waters from which drinking
water supplies may be withdrawn. Although the effects should be
* Potential
Because raw FGD sludge is very permeable, sludge liquid
(liquor) or other liquids can pass through, and possibly pollute
the groundwater by leaching. The permeability of FGD sludge is
a measurement of the rate at which water can pass through the
ma Tn^s / ette. sPrubber sludge has a permeability of 1
-------
the public and to wildlife. Sludge that has been sufficiently
dewatered and is nonthixotropic could be reclaimed and revege-
tated to produce an area adequate for recreation or building.2
Weathering also has a detrimental effect on the physical
properties of sludge. The EPA has conducted research in this
area as well.4 Sludge disposal areas are also possible sources
of fugitive dust. This problem is confined to specific areas,
however, and can be alleviated through proper site management.5
2.4.3 Disposal Practices
Lime FGD systems produce large amounts of sludge. Unde-
sirable physical and chemical properties, the presence of trace
elements that may be toxic, and the volume of the sludge have
created environmental concerns related to disposal. These
concerns include not only extensive land use, but also preven-
tion or limitation of the release of pollutants to surface or
groundwaters.
Although FGD sludge could be used to produce gypsum for use
in wallboard or portland cement, most utility power plants will
dispose of the sludge. Sludge utilization is not economically
attractive at the present time.6
Coal-fired power plants in the United States have been
disposing of coal ash by ponding or landfilling for many years
and have extended these conventional ash disposal practices to
the disposal of FGD sludge. However, several utilities chemi-
cally treat or fix the wastes before disposal. Fixation im-
proves the structural properties of the waste and tends to
decrease leaching. It is possible that the EPA will eventually
require sludge fixation before landfilling or other methods of
disposal.
Groundwater pollution due to leachate migration or runoff,
the physical strength of the sludge, and land use are the major
environmental problems associated with sludge disposal. At
present, no Federal criteria specifically apply to FGD sludge
disposal, but the Resource Conservation and Recovery Act of
1976, which was signed into law in October 1976, requires that
the EPA establish regulations or guidelines for disposal of
wastes from air pollution control systems such as FGD.7 Guide-
lines for FGD sludge and coal ash disposal have been prepared
and are under reivew. The EPA has been directing efforts since
mid-1975 toward preparation of technically supportable documents
that can be used to set FGD waste disposal guidelines.8
The EPA has indicated that disposal of raw sludge is un-
acceptable.2 In September 1975, the EPA declared "permanent
land disposal of raw (unfixated) sludge to be environmentally
unsound because it indefinitely degrades large quantities of
2.4-4,
-------
land.' Eventually, however, disposal of raw FGD sludge may be
? 1 ?vC
-------
Typically, sludge is pumped or trucked to a pond where it
settles to 35 to 45 percent solids.9 Supernatant (sludge
liquor) can be recycled to the scrubbing system. The use of
ponds for sludge disposal requires the availability of suffi-
cient land near the power plant.
The techniques of pond construction and operation are well
established. Wet disposal virtually precludes the fugitive-dust
problems that sometimes occur in dry disposal operations.
Holding volume can be increased by building up the sides of the
disposal pond. At the end of its useful life, the disposal site
can be left undisturbed for a hydraulic head to form over the
matrix.
Although ponding appears to eliminate the complications of
other disposal techniques, it has disadvantages. Compared with
other methods, ponding requires a larger volume to hold the
sludge; this volume may or may not be available. Also, because
the sludge may remain fluid, concerns may arise about the even-
tual abandonment of the site and the possible removal of large
areas of land from any future use.
Ponding techniques were developed with little or no concern
for their environmental effects. Groundwater contamination and
overflow into surface waters are now important considerations.
Possible pollutants from ponding of sludge include soluble toxic
species, quantities of species not considered toxic, and exces-
sive suspended and dissolved solids. The chemical oxygen demand
of sulfite sludge is also a potential problem.
The water above and in the sludge provides a force for
percolation and the contamination of any existing underground
aquifer. Pond liners can be used to form a barrier between the
sludge and the aquifers. Both clay and synthetic materials have
been used as liners. Potential problems include the unknowns,
such as what effect the sludge has on a liner and what the
actual life expectancy of a liner is. In view of these uncer-
tainties, a system designed to monitor the effectiveness of the
liner should be incorporated into the disposal site.1
2.4.3.2 Landfilling—
When limited land availability and/or economic considera-
tions make ponding unfeasible for FGD sludge disposal, landfil-
ling of dried sludge may prove practicable. Like ponding,
landfilling has historically been used by utilities for disposal
of coal ash. The major difference between the landfilling of
ash and sludge landfilling is that the ash is often collected
dry, whereas sludge must be processed into a dry state suitable
for landfilling.
One operating lime FGD system, at the Paddy's Run No. 6
unit of Louisville Gas and Electric, disposes of slurry by
2.4-a,
-------
landfilling. This method offers several advantages. For exam-
ple, dry disposal by landfilling largely precludes reclamation
problems associated with ponding. It also eliminates the need
for dams or dikes, which may be required for disposal by pond-
ing- This method of disposal is also more efficient in terms of
total land requirements because much of the water has been
removed. Rainfall onto a dry disposal area can be treated as
runoff, whereas with ponding, it creates the possibility of
runoff and adds to the supernatant. Leaching is believed to be
less of a problem with landfills because the amount of water in
and above the sludge has been reduced. The primary disadvan-
tages of dry sludge disposal by landfilling are the handling and
processing steps added to the disposal system in order to con-
vert the sludge into a dry form.
The simplest method of reducing the water content of FGD
sludge is to add dry solids such as fly ash, if available. To
produce a drier product, it is usually necessary to apply fur-
ther treatment. Popular methods include interim ponding, clari-
fication, centrifugation, and vacuum filtration. Interim ponds
or clarifiers are often primary dewatering devices. An interim
pond provides temporary storage as well as clarification and
sludge settling. Clarifiers achieve the same result. These
methods are not adequate to produce a sludge dry enough for
landfilling, except at several western facilities, where an arid
climate, low-sulfur fuels, and other factors make interim pond-
ing alone a feasible disposal technique. Vacuum filtration or
centrifugation can be applied in conjunction with these primary
dewatering techniques to achieve a higher sludge solids content.
Centrifuges produce concentrated sludge and good clarification.
They have achieved up to 75 percent solids content in TVA tests,
but their primary drawback is high power consumption. Vacuum
filtration is used for further dewatering at Paddy's Run No. 6.2
This method has achieved solids contents of 55 to 70 percent.6
Dewatering of the sludge to a minimum of at least 50 percent is
necessary to ensure handling capabilities.
Landfilled dried sludge is believed to pose less of an
environmental hazard than ponded raw sludge. Dried sludge,
however, has the potential of rewatering when exposed to rain-
fall; s° tne possibility of groundwater and surface water pollu-
tion', although reduced, is not eliminated. When water from
rainfall or subsurface flow contacts the sludge, it can create
leachate problems if allowed to percolate through the sludge.
Additionally, if the runoff from such a landfill is permitted to
seep through nearby land, leachate can pollute nearby streams or
the groundwater surface through leaching.5
Properly designed and managed landfills can minimize pollu-
tion potential. Drains upstream of the landfill can prevent
subsurface flow into a landfill, and a landfill liner, such as
that discussed for ponding, can trap leachates at the bottom.
2.4-7
-------
Covering the landfilled material can greatly reduce the quantity
of leachate while affording protection against the possibility
of rewatering and preventing surface leaching. Landfilling
dried sludge appears to be more environmentally acceptable than
ponding in that it reduces the volume of material to be disposed
of and does not necessarily preclude future productive use of
the site.
2.4.3.3 Chemical Fixation—
Sludge fixation, although not a disposal technique in
itself, is a means of physically and chemically stabilizing FGD
sludge to reduce its pollution potential and make its disposal
more environmentally acceptable. A nontoxic, nonleachable,
load-supportive material is the desired end-product. Table
2.4-2 lists the installations that fixate FGD sludge. Fixation
has been described as an encapsulation process because chemical
and physical changes that provide a barrier against pollutant
migration are effected in the sludge.3
Table 2.4-2. LIME FGD SYSTEMS THAT FIXATE SLUDGE
Utility
Plant
Process
Columbus and Southern Ohio
Electric
Duquesne Light
Pennsylvania Power
Conesville 5
Elrama
Phillips
Bruce Mansfield
1 and 2
IUCS
IUCS
IUCS
Dravo
Only Dravo Corporation and IU Conversion Systems, Inc.,
offer sludge fixation processes that are considered well enough
developed and tested to be commercially viable. In general,
these fixation processes reduce the solubility of major chemical
species by a factor of two or four, reduce the permeability by
at least an order of magnitude, and improve the structural
properties of the sludge. The environmental effects of disposal
of fixated sludge can be less than those from other available
sludge disposal methods, but the additional cost of chemical
fixation must be weighed against the benefits.
Unless proper procedures are followed in disposal of fix-
ated sludge, the potential for chemical pollution still exists.
For example, if rainwater is allowed to percolate through the
sludge, unbound chemical species can be leached out.5 Tests
have also indicated that fixation does not appear to improve
leachate quality with respect to trace metals. In fact, some of
2.4-8
-------
the trace elements in the leachate may even be due to the fixa-
tion chemicals. ' Even though fixation does not appear to reduce
leachate concentrations of trace elements, it can reduce the
concentration of manor chemical species by 25 to 50 percent
reduce the permeability of the material, and allow more effil
cient (volume) disposal.5 C-LJ-J.
Bravo system— The Dravo Corporation process is being used
in connection with a full-scale FGD system at the Bruce Mans-
field Power station of Pennsylvania Power Co. in Shippingport,
Pennsylvania. Experience was also gained on the Dravo sludge
treatment system during a 2-yr demonstration program at the
Phillips Generating Station of Duquesne Light in South Height,
Pennsylvania. ' '
.0™.haS been involved with FGD systems since the early
1970 's. Its research led to the development of sludge fixation
processes that are based on an additive called Calcilox1* (also
developed by Dravo). Calcilox is a hardening agent derived from
blast furnace slag. when added to FGD sludge, it effects
changes in the sludge that result in an end product that alle-
viates some of the concerns associated with the disposal of raw,
untreated scrubber sludge. The product is more physical^
stable, stronger, and less permeable than untreated wastes
Three different disposal variations of the sludge fixation
process are avail able from Dravo, each involving the addition of
Calcilox The full impoundment method is used at the Bruce
Mansfield Station Scrubber sludge, Calcilox, and hydrated lime
are mixed, and the slurry is piped to a final disposal pond,
where the mixture cures or stabilizes. Excess water or super1
natant is pumped back to the scrubbing system to be recycled.
The correct amount of Calcilox addition is determined through
testing, so that the sludge will possess an unconfined compres-
sive strength of 4.5 tons/ft* after 30 days. The curing time is
not particularly important in the full impoundment method. At
Mansfield, the slurry will cure beneath the supernatant to form
a stabilized mass. The site can eventually be used for light
industrial development or for recreational purposes.10
If dry handling methods are preferred, Dravo offers two
such methods— interim ponding and mechanical dewatering In-
terim ponding was demonstrated at the Phillips Power Station of
Duquesne Light Company with the method, a mixture of sludge
and Calcilox is pumped to small curing ponds. After it has
cured, the sludge is excavated and moved to a landfill.
When dry fly ash is available (i.e., collected before the
scrubbing system) the sludge can be dewatered by mechanical
means, then mixed with Calcilox and fly ash. The resultant
mixture can be sent directly to a landfill After a curing
period of 5 to 6 days, the fixated sludge would be spread on the
landfill. This is Dravo 's mechanical dewatering process and il
has not yet been used at any full-scale installation
2.4-9
-------
The water content of FGD sludge is slightly reduced by
stabilization with Calcilox and depends on both the original
water content of the sludge and the amount of Calcilox added.
The solids content of the sludge increases only slightly when
interim ponding or full impoundment is used because little
dewatering is involved. Thus, the required disposal volume is
approximately unchanged with Calcilox stabilization.2
Sludge fixated with Calcilox will not reslurry unless it is
subjected to severe remolding. Even if it does reslurry, the
sludge will harden again if left undisturbed. There appears to
be no tendency to reslurry after the slurry has been cured for
about 90 days.2 Sludge disposal areas can support loads; thus
they can utlimately be used for development. If the full im-
poundment method is used, the final site can be used, possibly
as a landfill, depending on drainage provisions.2
Treatment of sludge with Calcilox improves the environ-
mental acceptability of its disposal. Leachate quantity is
reduced, and tests indicate that permeability is reduced by at
least a factor of 10. Permeability coefficients normally range
from 10~4 to 10~5 cm/s for raw sludges and 10~5 to ip""6 cm/s for
Calcilox-treated sludge, and values as low as 10~8 cm/s have
been attained.10 Dravo has proposed using sludge treated with
extra Calcilox (to ensure the lowest possible permeability) as a
liner for the disposal area.
The full impoundment method requires a larger volume for
disposal than simple dry landfilling. Dikes and/or dams may be
necessary to contain the sludge during the settling and curing
processes. Other disadvantages of this method include an in-
ability to monitor the quality of the sludge as it cures and an
increased amount of leachate during pond operation (because of
the pool of supernatant above the curing sludge).
An advantage of the Dravo full impoundment system, in
addition to reduced leachate and lower permeability, is that it
does not require dry fly ash, which is necessary for most dry
disposal systems. Also, the disposal area can be reclaimed for
building development or used as a lake. It is not known, how-
ever, what effect fixated sludge would have on the water quality
of a recreational lake reclaimed from a disposal site.
Compared with the ponding of raw FGD sludge, Dravo's in-
terim ponding system allows for somewhat more efficient land
use. Based on density, however, the volume of the disposal area
is still one and a half times that necessary for dry landfilling
of untreated sludge.2 This method also has the potential advan-
tage of not requiring dry fly ash. Final disposal is dry; thus,
dams or dikes are unnecessary. This reduces costs and elimi-
nates a potential problem area. Environmental and land reclama-
tion advantages discussed for the full impoundment method apply
to the interim ponding technique as well.2
2.4-10
-------
The processes of curing, excavation, and final disposal
occur in sequence and somewhat complicate disposal by adding
handling steps. This system uses a series of steps, and prob-
lems in any one of these steps can disrupt operation of the
system.
In summary, Dravo offers either wet or dry disposal sys-
tems. No dry fly ash is required, and tests show the processes
are applicable to sludge with various sulfite and sulfate con-
tents . Dravo has not indicated any uses of the sludge fixated
with Calcilox other than for landfilling or as a liner.7
Further investigation appears to be appropriate regarding
changes in properties (if any) of Calcilox-fixated sludge over a
period of years. Furthermore, because the only facility that
produces Calcilox is in Pennsylvania, application of the process
might be geographically, and therefore economically, limited.
IU Conversion Systems—IU Conversion Systems, Inc. (IUCS)
is another vendor with full-scale commercial experience in the
stabilization of FGD sludge. The company markets a physico-
chemical fixation system called Poz-O-Tec , which it claims
produces a sludge that is ecologically acceptable. This system
has been proven through full-scale operation.10
The technology developed by IUCS over the past 25 years
utilizes pozzolanic (cementitious) reaction principles. The
poz-O-Tec process, which was developed about 8 years ago, pro-
vides a method of including FGD sludge in a chemically stabi-
lized matrix. The sludge is trapped and encapsulated within a
matrix, which is hard and relatively impermeable.11 The treat-
ment system involves sludge dewatering and the addition of lime,
dry fly asn' and additives.
IUCS systems have been used or demonstrated at various
locations across the country. The largest system is at the
Conesville Station of Columbus and Southern Ohio Electric Com-
pany. It uses the Poz-O-Tec process on sludge from two units
generating a total of 800 MW.
An IUCS interim processing plant has been operating at the
Elrama station of Duquesne Light Company since November 1976.
This system treats wastes from an FGD system that now treats 200
of the total 500 MW. A full-scale facility under construction
will be capable of treating all the wastes generated when the
FGD system is completed. An IUCS system has been constructed at
Duquesne's Phillips Station. These two systems handle approxi-
mately 700,000 and 450,000 tons/yr, respectively.2
IUCS gained experience by operating a pilot plant at the
Mohave Station of Southern California Edison Company. In this
demonstration project, Poz-O-Tec was used as landfill, as a base
2.4-11
-------
course for parking lots and roads, as an aggregate for concrete,
and finally, as a land base upon which a condominium was con-
structed.
G.W. Carson and Co., the predecessor to IUCS, also demon-
strated the process and possible uses for the fixed wastes.
Poz-O-Tec was used to prepare a base course of a parking area
and as a pond liner for fly ash disposal ponds at two generating
stations.l °
Sludge disposal using the IUCS process involves three
steps: dewatering, mixing, and placement or disposal of the
fixated sludge. Drum vacuum filters are generally used to
dewater the sludge. If necessary, they may be preceded by
thickeners. Centrifuging is an alternative method of dewater-
ing. After the moisture content has been reduced, the sludge is
thoroughly mixed with dry fly ash, lime, and an additive. The
end product can be hauled by truck, rail, barge, or conveyor
belt to a disposal site or to wherever it will be used.10
The chemical reactions involved in the Poz-O-Tec process
are similar to those occurring in portland cement, but they
proceed at a slower rate. Sludge particles and fly ash are
bound together in a rigid, physically stable matrix that will
not reliquify.
Physical characteristics of Poz-O-Tec-treated sludge vary,
depending on the raw scrubber sludge and the degree of treat-
ment. The consistency of the end product can be made to re-
semble dirt, sand, or solid rock. The IUCS process increases
the density of the sludge, resulting in a smaller volume to be
disposed of. Compared with that produced by the combined pond-
ing of raw scrubber sludge and fly ash, the disposal volume can
be halved by the Poz-O-Tec process. Compared with the separate
disposal of ponded raw sludge and landfilled fly ash, the volume
savings is approximately 15 percent.2
Strength and compressibility are also improved by the
Poz-O-Tec process. The fixated sludge is very incompressible,
and landfilled material will support normal foundation loads.
The material gains strength with curing and shows no tendency to
reslurry upon reworking or exposure to water. The properties
have been compared with that of a "low-strength concrete-like
material." The chemical bonding of the sludge particles im-
proves stability and preserves the desirable physical proper-
ties.2
Stabilizing FGD sludge with the IUCS process decreases land
requirements and improves reclamation potential. The use of
this stabilization process also decreases the quantity of lea-
chate and improves the quality because the chemical bonding
creates less soluble species. To date, trace element leaching
2.4-12
-------
data have been inconclusive because of low concentrations and
variable results. Further testing is being carried out by the
U.S. Army Corps of Engineers at its Waterways Experiment Sta-
tion.2'12
Permeability is another important factor in determining
environmental accep_tability_. Tests on Poz-O-Tec material have
shown values of 10 6 to 10~7 cm/s with curing. Thus, the IUCS
process reduces permeabilities of scrubber sludge to levels of
100 to 1000 times lower than those of raw sludges. The physical
encapsulation process that occurs in the IUCS process limits
water contact. Improved leachate and low permeability reduce
the mass of leached material per unit of time from 200 to 2000
times less than that in unstabilized scrubber sludge.2
Sludge fixated by the IUCS process can be landfilled or
disposed of in a quarry, mine, or ravine. Sludge produced at
the Columbus and Southern Ohio Electric Company's Conesville
Station is being disposed of on flatlands. Over the normal life
expectancy of the plant, a 100-ft hill, which will eventually be
reclaimed by placement of topsoil and revegetation, will be
created.10
If the fixated sludge is to be used as a byproduct rather
than simply disposed of, IUCS modifies the process with more
additives to further increase the physical stability of the
material. Pond liners and road bases for public highways and
parking lots have been constructed with Poz-O-Tec. Other by-
product possibilities have included use as a synthetic aggregate
for concrete blocks and as a subbase for a warehouse.2
Although the IUCS process can be used to stabilize sludge
from any calcium-based (lime, limestone, dual alkali) FGD sys-
tem, applicability depends on the availability of dry fly ash.
A minimum of 10 percent fly ash is required for the process.
Although it has not been demonstrated, IUCS claims the Poz-O-Tec
process can be applied at plants where fly ash is collected wet.
According to IUCS, substitutes for fly ash can also be used, and
they are currently conducting research to determine possible
substitutes. The IUCS system could be applied to almost all of
the existing or planned utility FGD systems because most are
calcium-based and have available fly ash.
Scrubber sludge fixation by the Poz-O-Tec process offers
several advantages when compared with other disposal techniques.
Dams or dikes, which may be needed for ponding of raw or treated
sludge, are unnecessary. Because it is a dry disposal tech-
nique, more efficient land use (disposal volume) is realized.
Furthermore, the environmental impact is decreased because a
separate disposal site is no longer needed for fly ash.
2.4-13
-------
Sludge treated by the Poz-O-Tec system exhibits greater
strength and density and possesses lower permeabilities (by
several orders of magnitude) than raw FGD sludge. This low
permeability can eliminate the need for liners because the
material can be equal to or even less permeable than standard
requirements for landfill liners. Reduced leachate potential
and the good physical properties of this material simplify
eventual reclamation of the disposal site. Successful demon-
strations of byproduct utilization also increase the attract-
iveness of the Poz-O-Tec process.
2.4.3.4 Alternative Disposal Methods—
Ponding and landfilling are the established sludge disposal
methods. They will probably continue to be the primary sludge
disposal techniques even though other techniques have been used
and continue to be investigated. Alternative methods such as
mine filling and ocean disposal appear to offer benefits, but
could adversely affect the environment. Current EPA programs
for investigating alternative disposal methods are designed to
provide answers to the major questions.
Mine disposal—Coal mine disposal of FGD waste has long
interested engineers because of existing railroad links between
coal mines and power plants and because of the need for material
to fill the empty areas left by mining. Only recently, however,
have studies been undertaken concerning the technical, environ-
mental, and economic factors connected with mine disposal. An
initial review suggests that two types of mines are best suited
for FGD waste disposal: active, surface-area coal mines located
between the Rocky Mountains and Appalachia and active, room-and-
pillar, underground coal mines of the East, including Appala-
chia. Unit 2 of the Milton R. Young Station of Minnkota Power
Cooperative in North Dakota is currently depositing flue-gas
cleaning wastes in a surface lignite mine; and utilities in Ohio
and Minnesota have considered mine disposal projects, but have
not adopted them.13
Mine disposal of FGD wastes could increase the total dis-
solved solids in waters recharged by leachate from the disposal
site. When part of the overburden is to be placed in the mined-
out strip before the deposit of FGD wastes, the wastes might
remain above the groundwater table, with the result that there
would be less likelihood of pollution from the leachate.13
Ocean disposal—Arthur D. Little is currently studying
ocean disposal of FGD waste for the EPA. Lack of land for
sludge disposal sites is one reason why many power plants in the
Northeast cannot fire coal. The same installations, however,
often have access to the ocean; and if ocean disposal were
proven environmentally acceptable, they might convert from
burning oil to firing coal. Still, until better data about the
technical, environmental, and economic aspects of ocean dumping
are available, disposal of sulfite-rich FGD wastes on the Conti-
nental Shelf or in the deep ocean appears unadvisable.13
2.4-14
-------
Various environmental problems could result from ocean
disposal of FGD wastes. Fine-grained, untreated FGD wastes
could "pave" over the coarse-grained sand particles that cover
the ocean floor and are most conducive to marine life on the
Continental Shelf. In addition, the settling and resuspension
of wastes could expose various marine organisms to harmful
concentrations of suspended sediments. The effects of sulfite
toxicity, oxidation, and dissolution on the marine environmental
could also prove detrimental. Finally, both treated and un-
treated wastes may have trace-element concentrations that exceed
acceptable levels for marine life.13
Sludge utilization—Utilization of FGD sludge appears to
hold limited promise. Although various applications and by-
products are known to be technically feasible, few, if any, can
be economically justified in the United States at present. As a
result, disposal will be practiced by an overwhelming majority
of utilities with FGD scrubber sludge.
Primary examples of FGD sludge utilization were discussed
earlier in connection with the IUCS Poz-o-Tec process. Sludge
treated by this process has been used as a base for highways,
parking lots, a warehouse, and also a pond liner. Unfortu-
nately, the overall demand for sludge for these uses is not
expected to exceed that which could be supplied. Nevertheless,
diversion of some of the waste into useful end products could
lengthen the useful life of a disposal site.2
Other uses of scrubber sludge are in the production of
construction materials and in agricultural and chemical re-
covery. One of the principal byproducts of lime and limestone
FGD systems can be calcium sulfate (gypsum). In Japan, gypsum
from FGD systems is used extensively in the production of wall-
board and portland cement. This type of process has been demon-
strated in the United States, but its widespread use is doubtful
because this country has large, natural sources of dry gypsum,
whereas Japan has little or none.9
Other construction materials that can be derived from FGD
sludge include brick, aerated and poured concrete, and mineral
wool. The technology for producing these products was developed
at the Coal Research Bureau at West Virginia University. These
sludge-derived products have shown properties comparable to
their natural counterparts.9
Agricultural applications of scrubber sludge are also being
investigated. These include soil amendment and fertilization.
The Coal Research Bureau has shown that the calcium in sludge
can be beneficial to plant growth and that sludge could possibly
be used to adjust pH of the soil. The TVA is conducting re-
search for the EPA on the production of fertilizer from lime/
limestone scrubbing wastes. Thus far, only two commercial
2.4-15
-------
applications of FGD sludge for agricultural purposes are known.2
This usage of sludge is expected to be limited and will depend
largely on local conditions.
A final consideration is the utilization of FGD sludge for
chemical recovery. An EPA program now underway is investigating
the conversion of scrubber sludge to hydrogen sulfide (from
which elemental sulfur can be derived) and calcium carbonate
(limestone). Several different processes and methods have been
suggested, but none has been used on a full-scale basis. Most
would result in elemental sulfur or some sulfur compound, and
extensive sludge processing would be required for any chemical
recovery. Element sulfur, however, is abundant and relatively
inexpensive, so there is little economic incentive to develop a
recovery plant.2
Sludge utilization has been proven to be technically fea-
sible. Limited markets may be found, but the prospects for
widespread use of large amounts of sludge appear dim. Lack of
economic competitiveness is the primary constraint. As a re-
sult, most FGD sludge will probably be disposed of by ponding or
landfilling.
2.4.4 Economics
Table 2.4-3 lists the flue-gas cleaning and sludge-disposal
practices of all utility scrubbers that were operating in Novem-
ber 1977 and using throwaway processes. Although the informa-
tion is now somewhat outdated (e.g., the Will County 1 Station
of Commonwealth Edison and the St. Clair 6 Station of Detroit
Edison are no longer operational), it still gives a fair indica-
tion of how utilities deal with FGD sludge. Ponding remains the
most common form of disposal.2
The costs of disposal can vary greatly. They depend on the
quantity and quality of the sludge, the nature and location of
the disposal site, and associated design, material, labor, and
delivery considerations. These factors may differ widely from
site to site.
An EPA-sponsored symposium on flue gas desulfurization,
held November 8 though 11, 1977, provided some useful informa-
tion on sludge disposal costs. Two presentations of particular
value on sludge disposal techniques used by utilities and their
costs are discussed in the following paragraphs.
Boston and Martin gave a presentation regarding the opera-
tion of the Conesville Generating Station of Columbus and
Southern Ohio Electric.14 This generating station uses the IUCS
sludge disposal technology. Unit 5, a 400-MW boiler, fires coal
with a sulfur content of 4.5 percent. Unit 6, which began
operation in June 1978, is similar to Unit 5. The lime-based
-------
Table 2.4-3. FLUE-GAS-CLEANING AND SLUDGE-DISPOSAL PRACTICES FOR UTILITY SCRUBBERS
USING THROWAWAY PROCESSES AND OPERATIONAL ON NOVEMBER 1. 19772
utility
Arizona
Public Service
Arizona
Public Service
Columbus i
Southern Ohio
Electric
Commonwealth
Ediaond
Detroit
Edisond
Duquesne
Light
Duquesne
Light
Indianapolis
Power t Light
Kansas City
Power 4 Light
Kansas City
Power t Light
Station name
and unit number
Choi la 1
Four
Corners 1. 2, 3
Cones vi lie 5, 6
Hill County 1
St. Clair 6
Elraraa
Phillips
Petersburg 3
Hawthorn 3 , 4
La Cygne 1
MW equiv.
oil scrubber (s)
115
575
S06
140
163C
500
400
515
180
840
Fly ash removal
Mechanical,
R-C flooded disc
Chemico
venturi
ESP
ESP, B4W
venturi
Mechanical ,
ESP, Lurgi
Venturi, and
Peabody spray
tower
Mechanical ,
ESP
Mechanical ,
ESP
ESP
C-E marble bed
Btw venturi
SO, removal
R-C packed tower,
unpacked tower
b
UOP-TCA
BsW 2-stage
perforated plate
b
Chemico
venturi
Chemico
venturi
UOP-TCA
b
BSW 2-stage
countercurrent
tray
Reagent
Limestone
Alkaline ash,
lime
Lime0
Limestone
Limestone
Lime
Lime
Limestone
Lime
Limestone
e
| Oxidatio
u
p Thickene
•
•
•
•
•
•
ilter j
Vacuum f
«
n*
1 Centrifu
i
tr>
c
V
UJ
•
J £
R
•O 0
C
II >.
, . — 1
"3
rH
-H tl
» E
•a -i
c
ti -a
-• c
ID ig
*
-
Commercial
V)
•
fixation
[ DRAVO
_..
.
Ultimate
disposal
j Pond
(LTndf ill
_ i
>
-. 1 -
•
i -..
•
i-.
•
*\
(continued)
-------
Table 2.4-3. (continued)
Utility
Kansas
Power t Light
Kentucky
Utilities
Louisville
Gas t Electric
Louisville
Gas ( Electric
""Minnesota
Power t Light
Minnesota
Power t Light
Minnkota
Power Coop-
erative
Montana
Dakota
Utilities
Montana
Power
Northern
States Power
Pacific Power
t Light
Station name
Lawrence 4 , 5
Green
River 1, 2
Cane Run 4
Paddy ' s Run 6
Aurora 1, 2
Clay
Boswell 3
Milton R.
Young 2
Lewis and
Clark
Colstrip 1, 2
Sherburne 1, 2
Dave Johnston 4
MW equiv.
525
64
190
65
116
350
450
50
716
1500
330
C-E rod venturi
Mechanical ,
AAF venturi
ESP
ESP
Elbair spray
impingement
Elbair spray
impingement
ESP
Mechanical
CEA venturi
C-E rod venturi
Chemico venturi
C-E spray tower
AAF mobile bed
AAF mobile bed
C-E marble bed
b
b
ADL/CEA
spray tower
R-C flooded disc
CEA countercurrent
spray, KOCH tray
C-E single-stage
marble bed
b
Limestone
Lime
Carbide sludge
Carbide sludge
Alkaline ash
Alkaline ash
Alkaline
ash and lime
Limestone
Alkaline
ash, lime
Alkaline
ash, lime
Alkaline
ash, lime
cidatio
o
•
lickene
H
•
•
•
•
•
14
0
*-4
I
U
•
•
tl
sntrifu
o
TJ
8.
sttling
«
•
•
3 £
n
3 «
U V.
CO m
•
«J
£
•H
c
W tJ
CQ <3
i-i
a c
-t c
u -
U 4.
0 n
P >
O"
U
U)
•
i
L
!
i
j
s
<
•4
j
§
$
Qi
•
•
•
•
U "
AJ «
•0 •
|
—* '
2 t
T3
C .
S.1-
i
•
•
•
•
.
^
9
1
L
n
•4
3
indf ill
-1
•
to
00
(continued)
-------
Table 2.4-3. (continued)
Utility
Pennsylvania
Power
Public Service
of Colorado
Public Service
of Colorado
Public Service
of Colorado
South Carolina
Public Service
Authority
Southwest
Public
Services
Springfield
City Utilities
Texas
Utilities
Texas
Utilities
TVA
Station name
and unit number
Bruce Mansfield
1, 2
Arapahoe 4
Cherokee 1, 3, 4
Valjnont 5
Winyah 2
Harrington 1
Southwest 1
Martin Lake 1
Monticello 3
Widows Creek 8
MW equiv.
oil scrubber (s)
1650
100
600
80
140
350
200
750
750
550
Fly ash removal
Chemico venturi
Mechanical, ESP,
and UOP-TCA
Mechanical, ESP,
and UOP-TCA
Mechanical, ESP,
and UOP-TCA
ESP
ESP
ESP
ESP
ESP
ESP
SO, removal
Chemico venturi
b
b
b
B4W venturi
C-E marble bed
UOP-TCA
R-C spray tower
Chemico spray tower
TVA venturi.
qrid tower
absorber
Reagent
Lime
Alkaline ash
Alkaline ash
Alkaline ash
Limestone
Waste CaC02
slurry
Limestone
Limestone
Limestone
Limestone
0
•H
4J
a
Tl
X
O
II
c
V
-H
0
0
V
4J
tw
3
i
u
>
V
1
n
u
c
«
O
a
•o
c
&
U*
C
-H
f+
JJ
+J
cn
JZ
-H
> .C
B
•o a
e
*-* 1-1
ffl »UI
j:
B
4-4
•1 V
> s
*o *~*
c
wo
-< c
m «
a
u
it
o
u
in
M
§
4J
a
X
•"
0
a
|
a
Tl
C
O
0,
0
•
_
iH
4
a
a
•a
^
•H
*W
Tl
C
a Settling pond, a pond not used for final disposal.
b Where there is no SO2 removal device indicated there is some incidental SO2 removal that results in
503/504 sludge from the scrubbing process.
c Lime contains 4 to 6% MgO.
d These units are no longer operating as SO2 removal systems.
e One half of rat(>d capacity is being scrubbed.
-------
FGD system that cleans the flue gases produces a sludge that is
30 percent solids. This sludge slurry is sent to a primary
thickener, a secondary thickener (if needed), then pumped to the
vacuum filters. After the sludge is filtered, it is thoroughly
mixed with lime and fly ash in a pug mill to achieve fixation.
The mixture is then conveyed to the disposal area, which is part
of the existing ash pond.
The sludge fixation process equipment is owned by IUCS and
leased to Columbus and Southern Ohio Electric Company (C&SOE),
whose personnel operate and maintain the system. This arrange-
ment resulted in minimal capital costs, but annual operating
costs are relatively high. Preparation of the disposal site and
work connected with the ash system have accounted for the only
capital costs incurred by C&SOE to date. This has amounted to
$1,639,000 for both units. Annual operating costs for Unit 5,
including the IUCS fee, are $2,928,000 or 1.63 mills/kWh. The
annual cost for both units is expected to be $3,271,000 or 0.91
mills/kWh.14
The waste disposal system at the Bruce Mansfield Power
Plant was the topic of another presentation at the FGD Symposium
that the EPA sponsored in November 1977.15 Dravo Corporation
designed and constructed the waste disposal system for both
units at this power plant. The full impoundment method using
Calcilox additive was decided upon, with 100 percent redundancy
specified for the sludge treatment and transport systems.
After being thickened to 25 to 35 percent solids, the
sludge is pumped to a mixing tank for the addition of Calcilox.
From the mixing tank the mixture must be pumped approximately 7
miles to the disposal site, a 1400-acre valley that has been
made into a reservoir by construction of a 400-ft impoundment
dam. The system has been designed as a closed-loop operation,
and runoff and supernatant can be returned to the FGD system.
Monitoring wells surrounding the disposal site are used to check
groundwater quality.
The capital cost of the Dravo disposal system for both
units is reported to be $90,000,000 or $54.5/kW. This value is
in reasonable agreement with the $50.70/kW reported to FPC. The
operating cost is reported to be $3.81 per ton of slurry solids,
which is equivalent to approximately 0.55 mills/kWh, a value
considerably larger than the 0.04 mills/kWh reported to FPC. AS
at Conesville, the annual operating costs should decrease with
the addition of the second unit.15
Although FGD sludge is being fixated at both the Bruce
Mansfield and Conesville plants, costs differ significantly.
This is partially due to site conditions and the difference in
disposal techniques, but the primary reason involves the leasing
(Conesville) versus ownership (Bruce Mansfield) arrangement.
Actual costs over a projected 30-yr life span would probably be
comparable.
F 2.4-20
-------
A recent EPA report offers some broad generalizations about
the costs of disposing of FGD waste.16 This report calculated
the lifetime revenue requirements for new 200-, 500-, and 1500-
MW power plants using various disposal methods. These calcula-
tions are presented in Table 2.4-4. It should be stressed that
the estimates assume a limestone, rather than lime, scrubbing
process and do not include the cost of the process. Only the
costs of the disposal alternatives are estimated.
2.4-21
-------
Table 2.4-4. SUMMARY OF THE REVENUE REQUIREMENTS DURING THE
30-YEAR LIFE OF A NEW POWER PLANT** FOR VARIOUS
FGD-WASTE-DISPOSAL METHODS16
Caseb
Untreated
200 MW
500 MW
1500 MW
Dravo
200 MW
500 MW
1500 MW
IUCS
200 MW
500 MW
1500 MW
Chemf ix
200 MW
500 MW
1500 MW
Total
actual
lifetime
revenue
requirement, $
58,750,000
97,757,800
203,309,200
94,392,200
175,764,900
375,002,700
89,013,000
131,224,200
254,498,000
111,241,300
167,942,300
333,190,900
Lifetime
average
unit revenue
requirements ,
mills/kWh
2.30
1.53
1.06
3.70
2.76
1.96
3.49
2.06
1.33
3.36
2.63
1.74
Total
present-
worth
lifetime
revenue
requirement, $c
20,204,800
33,612,100
69,819,400
33,368,200
62,052,600
133,456,200
30,584,100
45,381,700
88,798,600
38,655,100
59,099,300
119,154,500
Levelized
unit revenue
requirements ,3
mills/kWh
2.03
1.35
0.94
3.36
2.50
1.79
3.08
1.83
1.19
3.38
2.38
1.60
to
I
to
to
Basis
Over previously defined power plant operating profile. 30-yr life: 7000 h for first
10 yr; 5000 h for next 5 yr; 3500 h for next 5 yr; 1500 h for next 10 yr.
Midwest plant location, 1980 operating costs.
Constant labor cost assumed over life of project.
New plants, coal analysis (wt.%): 3.5% S (dry), 16% ash, fly ash removed with S02
to meet New Source Performance Standards.
Discounted at 10% to initial year.
Equivalent to discounted process cost over life of power plant.
-------
REFERENCES
1. Rossoff, J., and R.c. Rossi. Mid-Term Report Study of
Disposal of Byproducts from Nonregenerable Flue Gas Desul-
furization Systems, Volume II: Technical Discussion. U S
Environmental Protection Agency, Washington, D.C., December
j. / x y / j •
2. Michael Baker, Jr., inc. State-of-the-Art of FGD Sludge
Fixation. Research Project 786-1, Task 3 (Draft). Elec-
tric Power Research Institute, Palo Alto, California
August 1977.
3. Mahloch, J.L., D.E. Averett, and j.j. Bartos, Jr. Pollu-
tant Potential of Raw and Chemically Fixed Hazardous Indus-
trial Wastes and Flue Gas Desulfurization Sludges Interim
Report. EPA-600/2-76-182.
4. Radian Corporation and Southern California Edison Company.
The Environmental Effects of Trace Elements in the Pond
Disposal of Ash and Flue Gas Desulfurization Sludge.
Electric Power Research Institute, Palo Alto, California,
September 1975.
5. Rosoff, J., R.E. Rossi, et al. Disposal of Byproducts from
Nonregenerable Flue Gas Desulfurization Systems: Second
Progress Report. EPA-600/ 7-77-052, May 1977.
6. Radian Corporation. Evaluation of Lime/Limestone Sludge
Disposal Options. EPA-450/3-74-016, November 1973.
7
Smith, C.L. Sludge Disposal by Stabilization - Why? In:
Proceedings of the Second Pacific Chemical Engineering
Congress, Inter-American Confederation of Chemical Engi-
neering, Asian Pacific Confederation of Chemical Engi-
neering, Denver, August 28-31, 1977.
8. Jones, J.W., and T.G. Brna. Environmental Management of
Effluents and Solids Wastes from Steam Electric Generating
Plants. In: National Conference on the Interagency Ener-
gy/Environment Research and Development Program, Washing-
ton, D.C., June 6-7, 1977.
9. Radian Corporation. Disposal of Lime/Limestone Sludges
EPA-68-02-0046, September 10, 1973.
10. Radian Corporation (Draft). Byproduct/Waste Disposal for
Flue Gas Cleaning Processes. EPRI RP 786-2, Electric Power
Research Institute, Palo Alto, California, March 29 1977
2.4-23
-------
11. Taub, S.I. Treatment of Concentrated Waste Water to Pro-
duce Landfill Material. In: International Pollution
Engineering Exposition and Congress, Anaheim, California,
November 10, 1976.
12. FGD Quarterly Report, Vol. 1, No. 2. U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina,
August 1977.
13. Jones, J.W. Disposal of Flue-gas-cleaning Wastes. Chemi-
cal Engineering, 84 (4):79-85, February 14, 1977.
14. Boston, D.L., and J.E. Martin. Full-scale FGD Waste Dis-
posal at the Columbus and Southern Ohio Electric's Cones-
ville Station. In: Symposium on Flue Gas Desulfurization,
Hollywood, Florida, November 8-11, 1977.
15. Lobdell, L.W., E.H. Rothfuss, Jr., and K.H. Workman.
Eighteen Months of Operation Waste Disposal System, Bruce
Mansfield Power Plant, Pennsylvania Power Company. In:
Symposium on Flue Gas Desulfurization, Hollywood, Florida,
November 8-11, 1977.
16. Barrier, J.W., H.L. Faucett, and L.J. Benson. Economics of
Disposal of Lime/Limestone Scrubbing Wastes: Untreated and
Chemically Treated Wastes. EPA-600/7-78-023a, February
1978.
2.4-24
-------
BIBLIOGRAPHY
Jones, J.W. Research and Development for Control of Waste anrt
Water Pollution from Flue Gas Cleaning Systems Inf EPA Sy^-
posium on Flue Gas Desulfurization, New Orleans, March 8-11,
Leo, P.P., and J. Rossoff. Control of Waste and Water Pollution
from Power Plant FliiP. nac m <*=,„-:— o.._.. _. . _ wj.j.uuj.wii
, . .
Coal Ash and FGD Sludge for the New England ESECA Candidates
-s EEcld
istration, Washington, December
Jones, J.W. Chemical Engineering, February 14, 1977, pp. 79-86.
FGD sludge Disposal Process Cost Assessment. Federal Energy
Administration, Washington, D.C., February 1978. *"««*
2.4-25
-------
SECTION 3
PROCESS CONTROL
3.1 INTRODUCTION
This section of the Data Book deals with the basic science
of process control and its application to the design and opera-
tion of a lime scrubbing FGD system. When used in this section,
the word "control" refers to process control and not to instru-
mentation hardware, which is discussed in Section 4.13. Process
control is not chemical control, which concerns the conditions
that cause scale formation or that affect SO2 removal. Chemical
control is discussed as part of the process chemistry in Section
£ •
In a continuous operation, such as a lime scrubber, process
control is required for safe and stable operation. The primary
goal of the control system in a lime scrubbing system is to
ensure that sulfur dioxide emissions from the scrubber meet the
emission limits. This section of the Data Book will discuss
techniques of controlling the following "variable's to meet the
emission limits:
0 Lime feed
0 Solids
0 Flue gas fluctuations
in addition to emission control, pH and lime feed control must
be designed to reduce excess chemical use. Solids and PH con-
trol are used to prevent scaling and plugging. The quantity of
reheat is controlled to reduce energy consumption while yielding
adequate emission dispersal. Liquid levels are controlled to
prevent tanks and vessels from overflowing. Flocculants, which
are fed to thickeners to improve clarification, must also be
controlled.
Process control has evolved into a distinct engineering
area, with a conceptual approach and a language that differs
from that of other engineering disciplines. Section 3.2 of this
Data Book illustrates, primarily through example, this approach
and some of its specialized terms.
Section 3.3 examines several of the major subsystems of a
lime scrubber control system to illustrate the practical appli-
cation of the approach to the control problems of a scrubber
3.1-1
-------
The section describes control techniques being used, or that
have been proposed, to solve such problems as control of reheat
and lime feed rate.
The controller and the control valve, which are common to
every control system, are discussed in Section 3.4. The action
of a proportional, integral, and derivative (PID) controller is
described, followed by a discussion of the linearity of the
control element (usually a control valve), which is a frequent
cause of poor control system performance.
3.1-2.
-------
CONTENTS
3.2 PRACTICAL APPLICATION OF PROCESS CONTROL
Page
3.2.1 Basic Control
3.2-1
3.2.2 Control Improvements
•J • ^ "^Tf
3.2.2.1 Avoidance of Component Errors q ? A
I'l'l'l El^na^°n °5 DIsturbances, Cascade Control 3^6
3.2.2.3 Elimination of Disturbances, Feed Forward 3.2-6
3.2.3 Control Systems
3 * 2 ™ y
Bibliography
3.2-i
-------
3.2 PRACTICAL APPLICATION OF PROCESS CONTROL
3.2.1 Basic Control
A study of process control is not concerned with the physi-
cal hardware of a plant, but with plant variables, which are
measurable properties of an operating process. A tank of water,
for example, has properties that include temperature, liquid
level, and total weight, any of which can be used as a control
variable. A flowing stream of slurry has the properties of
density, pressure, solids content, and flow rate; and any or all
of these could be a variable for use in process control.
The objective in the control of a continuous process is to
maintain variables at the desired values; the basic building
block to accomplish this is the control loop. There are two
types of loop: a feed-forward loop and a feedback loop. The
term "control loop" implies a circle, and a loop consists of six
components interconnected to form a continuous path. A simple
feedback control loop is shown in Figure 3.2-1. The six compo-
nents of the loop are:
1. A sensor, which is a device that measures the value of
a controlled variable. In Figure 3.2-1, the control-
led variable is the temperature of the stack gas; the
sensor is a thermocouple. The sensor in any control
loop is connected, either mechanically or with a
pneumatic or electronic signal, to:
2. A comparator. This is a mechanical or electrical
mechanism that compares the value of the controlled
variable with a set point. The set point is a mechan-
ical connection or a signal that defines the desired
value of the controlled variable. The difference
between the actual and the desired value, which may be
either positive or negative, is the error that exists
at any instant in time. The comparator is always
connected to, and is physically located, in the same
housing as,
3. A controller. A process controller is a computing
device that performs a mathematical manipulation based
on the value of the error. As a result of the mathe-
matical manipulation, a controller generates a signal,
known as the controller output. The output signal is
connected to:
4. A control element. This is usually a control valve
that moves mechanically in response to the output
signal from the controller. In the example of stack
gas reheat, the control valve opens or closes slightly
as the output signal increases or decreases. The
control element must modify:
3.2-1
-------
REHEATED GAS
TO STACK
SENSOR
(THE THERMOCOUPLE)
STEAM TRAP
o-—
/—X COMPARATOR AND
-J )TEMPERATURE CONTROLLER
.STEAM
CONTROL VALVE
SCRUBBED
FLUE GAS
Figure 3.2-1. Feedback control loop in a
stack gas reheat control system.
3.2-2
-------
5. The manipulated variable. In this example, the flow
rate of steam changes as the control valve opens or
closes. In some manner, the manipulated variable must
affect the controlled variable. The connecting link
between the two is:
6. The process. A portion of the plant hardware and a
flowing stream of fluid are integral components of
every control loop. In this example, a change in
steam flow rate changes steam coil pressure and,
consequently, steam condensing temperature. This in
turn changes the flow rate of heat through the walls
of the steam coil, thereby changing stack gas tempera-
ture, which is the measured variable. The circular
path is completed.
A feedback control loop is therefore a complex interaction
of mechanical and mathematical components. If the loop is to
operate properly, each component must be compatible with the
others, and each must be properly designed to accomplish its
intended function. Even if the loop is mechanically correct,
however, it may or may not be adequate to handle a specific
control problem. Two concepts provide a route to evaluate the -
probable performance of a control loop and to indicate whether
or not control improvements are indicated.
The first is the concept of disturbance. Any condition of
operation either within or external to the control loop that
will cause an unintentional change to the controlled variable is
said to be a disturbance to that loop. In the example of stack
qas reheat, at least two conditions could cause substantial
changes to reheat temperature. An increase in stack gas flow
rate will cause a drop in temperature, as will a drop in steam
pressure. Either of these conditions would constitute a distur-
bance to the control loop.
The second concept is that of time lag, which is usually
negligible in the control of electric power variables, but is a
major problem in chemical process control. In the example of
stack gas reheat, a time lag is created by the heat-sink effect
of the condensate and metal in the steam coil. The more massive
the coil, the longer the time lag will be. In pH control, as
much as 15 min can elapse following an increase in lime slurry
feed rate before a change is noted in the reading of the pH
sensing instrument in the recycle tank. On the other hand, some
loops have a short time lag; flow rate of a liquid changes
almost immediately whenever the control valve position is
changed. Most loops have time lags between these extremes, and
these can be estimated. Time lag in a loop is estimated in many
instances by the volume of process fluid that is affected by the
manipulated variable. To change the pH of a large volume, such
as the body of liquid in a scrubber, requires more time than to
3.2-3
-------
change the pH in a flowing stream in a pipeline. Time lag can
also be related to volume or mass of the manipulated stream.
If a feedback control loop has a short time lag (fractions
of a second to seconds), it can usually provide adequate control
even if disturbances are large (a liquid flow control loop
operates well even with large changes in upstream pressure).
Conversely, if disturbances are slight, feedback control usually
operates fairly well even though the time lag is long (a few
seconds or longer). (Feedback control of scrubber pH is ade-
quate if nothing disturbs the system.) Given large or frequent
disturbances and a long time lag, feedback control cannot main-
tain good control, and control improvements are usually neces-
sary.
3.2.2 Control Improvements
3.2.2.1 Avoidance of Component Errors—
The feedback loop must be properly designed. The first
step should be an examination of each of the loop components to
ensure that each is compatible with the others and able to
accomplish its intended function. Problems of compatability
include not only the more obvious mechanical, electric, and
range-selection matches, but also the problem of nonlinearity,
which is emphasized throughout this section of the Data Book.
The greatest problem of hardware inadequacy will be found
in the loop sensor. In most instances, these inadequacies are
due to the inability of the sensor to measure the controlled
variable often because of poor installation or low-cost equip-
ment. In some instances, however, the design is inadequate to
measure accurately the actual value of the controlled variable.
For instance, in the example of stack gas reheat control, Figure
3.2-2 shows a basic revision that has been made to the sensor to
measure the controlled variable more accurately. Temperature
can vary significantly from point to point across a large duct.
A single thermocouple will probably not measure the true tem-
perature. The average reading of four thermocouples, each in a
different part of the duct, is probably more representative of
actual temperature.
3.2.2.2 Elimination of Disturbances, Cascade Control—
In Figure 3.2-3, the output of the temperature controller
does not directly manipulate the steam control valve, but it is
used instead to adjust the set point of a second controller. In
a separate control loop, the second controller adjusts the
control valve to maintain the flow rate of steam. In this
manner, a control loop is separated into two parts, and the
technique is known as a cascade. Although there are other
reasons for using cascade control, the purpose in this example
is to eliminate disturbances that would otherwise be created by
3.2-4
-------
TO STACK
STEAM TRAP
Y
SENSORS
(FOUR
THERMOCOUPLES
o
o- —
o- ----
AVERAGING
INSTRUMENT
TEMPERATURE
CONTROLLER
_ STEAM
CONTROL VALVE
SCRUBBED FLUE GAS
Figure 3.2-2. Improved feedback control in stack
gas reheat control system.
3.2-5
-------
TO STACK
SENSOR
(THERMOCOUPLE)
STEAM TRAP
TEMPERATURE
CONTROLLER
O
FLOW CONTROLLER
STEAM
CONTROL VALVE
FLOW RATE
SENSOR
Figure 3.2-3. Cascade control.
3.2-6
-------
thereby in the temperatr of he flue * gl?
3.2.2.3 Elimination of Disturbances, Feed Forward—
™
known as the "feed forward «? Th^ i through an instrument
control differs from that of feelack^in"? v* °f • feed;forward
spect. In feedback, a disturbs^ 1 ^ Y lmP°rtant re-
error, which is then corrected F^H f* pe™ltted to caus^ an
ulated variable before ? the Irror h.t ^orward c^ges the manip-
an external condition that if not Cn«,n U^d *by resP°ndin9 to
a disturbance to the control loop TT/* f°^' would
into the control loop as an ea?ai n« i- f°rward 1S
controller and the
t^e
flow rate of stack gas we« T
adjusted in direct P^oportTon rate lt±e
lation were conducted accuratelv no It™ i ^ calcu-
controlled variable of the fee^ack loon ^It re/Ult ±n the
ratio adjustment is not ent!S!y acc^at. ^ ^"^^^
controller is still -in +*<* "*-«••«• e A y accurate, the temperature
anf to haUe other dJst*ban°e°sP accoI"Plish "«»1 correction
neutrall^d o d
the advantage of not requiring absolut'* £°™ard also has
-S
3.2-7
-------
TO STACK
FROM GAS
FLOW SENSOR
SENSOR
THERMOCOUPLE
STEAM TRAP
n_
o—
TEMPERATURE
CONTROLLER
—O—
FEED FORWARD
INSTRUMENT
STEAM
CONTROL VALVE
SCRUBBED
FLUE GAS
Figure 3.2-4. Feed-forward control
3.2-8
-------
3.2.3 Control Systems
In contrast to a control loop, where the objective is to
maintain one variable at a fixed value, an integrated contro?
system has the objective of maintaining the overall outpS? from
the plant within limits. in a scrubber, the "output" is flSe
gas discharged to the stack, and an integrated control system
includes all adjustments to all variables that influence the
quality of this discharge. ««iw.e t-ne
A control system therefore consists of many loops, con-
taining every adjustment needed to accomplish the plant objec-
tive. Routine manual adjustments by the plant ope?ator are an
integral part of the control system. if he changes a valve or a
controller set point to maintain a variable at I desired valSe
he is functioning as a controller in a control loop. Systeii
design includes the definition and instrumentation of Ihell
manual, or ''open," loops in. addition to the "closed" loops
implemented with automatic instrumentation. j-oopfa
Variables that must be controlled to form a complete con-
trol system can be divided into five classifications:
In most ^rubbers, only two
gas S02
™« + v u posslbly' temperature. To date, SO*
content has been controlled, if at all, only with an
^ermediate Process variables. The principal inter-
?S ?£LVa v,16 U Variable which has an impact on
the flue gas characteristics) is scrubber pH, which is
usually instrumented.
Auxiliary variables. Operations outside the main
Exarn^L /,eam rneg,Uire coordination with the process
Examples are solids content of thickened sludge and
e
whhot H * These can be controlled
with both closed and open loops.
4. inventory variables. Liquid level in process vessels
^
caused by process changes or
5. Limiting variables. For plant and process safety
certain variables must not exceed given limits Most
plants will contain loops to override other loops
3.2-9
-------
thereby protecting the plant during abnormal operating
conditions. Scrubber vessels have water deluge sys-
tems to prevent scrubber lining failures during recir-
culation pump failures.
Except for limiting variables, which usually develop as
details of subsystems, the basic loops of all classifications
can be shown on such a diagram as Figure 3.2-5. Without de-
fining details or specifying whether control is to be an open or
closed loop, this figure can show the essential requirements for
effective control of the scrubber. An important characteristic
of a control system is that every flowing stream of material
that enters or leaves the scrubber will be a part of at least
one control loop; some variable of each stream will be either
manipulated or controlled. A thorough control system design
will be based on a drawing of this sort that will in turn define
the instrumentation needed to execute each loop.
In the following section, some of the subsystems shown on
Figure 3.2-5 will be individually examined; however, although
they may be studied separately, in no instance does any sub-
system stand alone. Every control loop is either disturbed by
the action of others or creates a disturbance in others. The
approach of control system design is to use these interrelation-
ships as a guide to the complexity required in the instrumenta-
tion of each subsystem.
3.2-10
-------
STACK
ro
i
FRESH
MATER
SO^ CONCENTRATION
FRESH
LfATTP
^-THICKENER ^ -.-WATER INVENTORY
REHEAT
CONTROL LOOP
SLUDGE
SOLIDS
Figure 3.2-5. Lime scrubber control system.
-------
BIBLIOGRAPHY
Axelby, G. The Gap—Form and Future. IEEE Transactions on
Automatic Control, April 1964, pp. 125-126.
Bode, H.W. Feedback—The History of an Idea. In: Selected
Papers on Mathematical Trends in Control Theory, Dover, New
York, 1964. pp. 106-123.
Cohn, N. State of the Automatic Control Art in the Electric
Power Industry of the United States. Proc. of the JACC, 1965.
pp. 110-123.
Dorf, R.C. Modern Control Systems. Addison-Wesley Pub. Co.,
Reading, Massachusetts, 1966.
Dorf, R.C. Time-Domain Analysis and Design of Control Systems.
Addison-Wesley, Reading, Massachusetts, 1965.
Lime/Limestone Scrubber Operation and Control Study. EPRI
(RP630-2), April 1978.
Marks, L.S. Standard Handbooks for Mechanical Engineers.
McGraw-Hill, 7th ed., 1967. pp. 16.33 ff.
Maxwell, J.c. In: Governors Proceedings of the Royal Society
of London, 16, 1868. In: Selected Papers on Mathematical
Trends in Control Theory, Dover, New York, 1964. pp. 270-283.
Newton, G., L. Gould, and J. Daiser. Analytical Design of
Linear Feedback Controls. Wiley, New York, 1957.
Proposed standards and Terms for Feedback Control Systems, Part
2. A.I.E.E. Committee Report, Elec. Eng. 70, 1951.
Popov, P.E. The Dynamics of Automatic Control Systems.
Gostekhizdat, Moscow, 1956. Addison-Wesley, Reading, Massa-
chusetts, 1962.
Vyshnegradskii, I.A. On Controllers of Direct Action. Izv, SPB
Technology, Inst., 1877.
3.2-12
-------
CONTENTS
3.3 LIME SCRUBBER CONTROL SUBSYSTEMS
Page
3.3.1 introduction 3.3-1
3.3.2 Reheat Control 3.3-1
3.3.2.1 Fuel-fired Reheat 3.3-1
3.3.2.2 Hot Water Reheat 3.3-1
3.3.2.3 Steam Reheat 3.3-3
3.3.3 Lime Feed Control 3.3-6
3.3.3.1 Nonlinearity 3.3-7
3.3.3.2 Time Lag 3.3-7
3.3.3.3 Simple pH Control 3.3-9
3.3.3.4 Cascade pH Control 3.3-9
3.3.3-5 Outlet SO2 Feedback Control 3.3-9
3.3.3.6 Inlet S02 Feed-forward Control 3.3-12
3.3.3.7 Summary of Lime Feed Rate Control 3.3-14
3.3.4 Slurry Solids Control 3.3-14
3.3.4.1 Lime Slurry Feed Control 3.3-16
3.3.4.2 Absorber/Scrubber Solids Control 3.3-16
3.3.4.3 Thickener Solids Control 3.3-21
3.3.4.4 Flocculant Control 3.3-21
3.3.4.5 Solids Control Summary 3.3-24
3.3.5 Gas Flow Control 3.3-24
3.3.5.1 Pressure Control 3.3-26
3.3.5.2 Flow Control 3.3-26
3.3.5.3 Boiler Safety 3.3-30
3.3.5.4 Fan Control 3.3-30
Bibliography 3.3-33
3.3-i
-------
3.3 LIME SCRUBBER CONTROL SYSTEM
3.3.1 Introduction
In this section, four of the principal control subsystems
of a typical scrubber unit are examined. The subsystems control
reheat, lime feed rate, slurry solids (including waste dis-
posal), and gas flow.
3.3.2 Reheat Control
Several types of reheat systems are being used in lime
scrubbers. Fuel-fired reheat, either direct or indirect; steam
reheat, direct or indirect; and hot water reheat must be con-
trolled differently. This section will discuss some of the
control techniques for each technique.
In addition, temperature sensing is also important irres-
pective of the type of reheat used. In Figure 3.2-1, a single
sensor is used to measure the temperature of the reheated flue
gas. This type of measurement is adequate if the measuring
location is isokinetic and system accuracy and reliability are
not important. To improve gas temperature measurement and to
ensure the continuous monitoring if a single thermocouple fails,
however, the multiple sensor shown in Figure 3,2-2 should be
used.
3.3.2.1 Fuel-fired Reheat—
Gas- or oil-fired, direct or indirect, the control systems
are similar for fuel-fired reheaters, and they are the simplest
of all reheat controls. As shown in Figure 3.3-1, changes in
stack gas temperature vary the fuel supply to the burner.
Although it will not be discussed in detail here, the burner
system requires several control loops. As the fuel supply
changes, the amount of air must be varied to maintain proper
combustion. In addition, several limiting loops are required to
stop the fuel flow in case of flameout or combustion blower
failure. Multiple burners may be required in flue gas systems
to obtain adequate turndown. A single burner normally has a
turndown capacity of 3:1.
3.3.2.2 Hot Water Reheat—
Two control posibilities are available to control hot water
reheat systems. The temperature of the water can be changed or
the heat transfer coefficient and log mean temperature differ-
ence can be changed in the heat exchanger by changing the flow
rate. In either control loop, the quantity of energy in the hot
water supply system must be capable of reheating the gas to the
maximum desired temperature at full gas flow.
3.3-1
-------
SENSOR
r>—
BURNER
1
j
0 TEMPERATURE
CONTROLLER
t
FUEL SUPPLY
DUCT
Figure 3.3-1. Fuel-fired reheat control system.
3.3-2-
-------
Temperature control—One scheme for controlling the temper-
ature of the hot water is shown in Figure 3.3-2a. In this case,
steam is used to raise the temperature of the hot water as
required by the temperature controller. Other variations of
this system are possible.
The advantage of controlling the water temperature is that
it results in a short time lag between gas flow fluctuations and
the resulting change in heat input from the reheat medium, hot
water. The disadvantage of this technique is that moderating
the temperature may be hard to accomplish. Hot water (conden-
sate) may be obtained from the turbine discharge.
Flow control--The other technique for moderating the flue
gas temperature is shown in Figure 3.3-2b. In this system, the
flow of water is changed by the temperature controller. As the
velocity of water in the heat exchanger is increased, the outlet
water temperature rises since the higher flow rate requires less
temperature drop in the water to transfer the same amount of
heat to the flue gas. This causes the log mean temperature
difference to rise, which, in effect, raises the temperature of
the flue gas. The advantage of this type of system is its
simplicity. There is no major impact on the turbine system as
long as the supply of hot water is greater than needed. The
disadvantage of this system is the cost of the pump and the
horsepower required to operate it. Energy requirements for
pumping are fairly constant since the control valve absorbs the
energy when the flow is reduced.
3.3.2.3 Steam Reheat—
To control the degree of flue gas reheat using a steam
reheater coil, flow through the coil is changed. A steam re-
heater has two zones, one that has a high heat transfer rate
where the steam is condensing, and a second that has a much
lower rate where the hot condensate transfers heat to the flue
gas. Care must be taken in the design of the coil to assure
that the heat transfer to the gas is relatively uniform across
the gas flow. If this is not accomplished, all the condensing
sections of the steam coil will be on one side, and the rate of
heat transfer will reach a maximum and then drop sharply. This
phenomenon is called heat blinding; it may cause inadequate
temperature rise in the flue gas. The advantage of this system
is that the control requirements are well understood since steam
flow control is quite common.
Steam is also utilized in reheater systems in which air is
heated by the steam and then mixed with the flue gas.
3.3.3 Lime Feed Control
In a lime scrubber system, control of lime feed is abso-
lutely essential. Lime feed rate is one of the factors that
3.3-3
-------
TO STACK
(CZX^
SENSORS
O— —
o
'o
0
X-
s
EXCHANGE
^xr-^:
AVERAGING
INSTRUMENT
- y^~X
/ \_
V /
X — S
"1
1
1
1
1
,
1
1
) TEMPERATURE
CONTROLLER
WATER i
BLEED SUPPLY J
^C. TANK J^
_ COOL WATER ''f3
I
_ HOT WATER r^^
[ ^-^
ivLt ^STEAM SUPPLY
X **
\
1 '
RECIRCULATION STFAM \
DUMP OltMTI
PUMP SPARGER
DUCT
Figure 3.3-2a. Hot water reheat control system: steam
raises hot water temperature.
3.3-4
-------
TO STACK
HEAT
EXCHANGER
CENSORS o
O--
o
o
AVERAGING
INSTRUMENT
HOT WATER SUPPLY
COOL WATER RETURN
I
TEMPERATURE
( ) CONTROLLER
I
I
RECIRCULATION
PUMP
Figure 3.3-2b. Hot water reheat control system: flow
of water charged by temperature controller.
3.3-5
-------
determines the outlet SO2 concentration in the flue gas. Proper
pH control through proper lime addition can prevent scaling and
reduce corrosion, thereby improving the mechanical performance
of the scrubbing system.
All systems now operating control the lime feed rate with a
feedback loop, and all use pH as the controlled variable. Most
are unable to handle large changes in gas flow rate or inlet SO2
concentration automatically and therefore require considerable
operator attention.
Many of the systems in operation report problems with lime
feed control. There are a number of reasons for the poor per-
formance of these systems. Mechanical problems have been numer-
ous, especially those created by poor design of pH-measuring
electrode stations. Details are discussed in the Instrumenta-
tion Section (4.13) of this Data Book. From the standpoint of
process control, however, the major problems are created by two
characteristics of the process itself:
The response of pH to a change in lime feed rate is
extremely nonlinear, and the shape of the pH curve
changes as the chemical composition of the slurry
changes.
By itself, pH is an inadequate variable on which to
base lime feed rate. Optimal pH varies with the
chemical composition of both the slurry and the flue
gas.
These limitations necessitate complex control systems to
control lime feed rate dependably. The following paragraphs
describe the two process conditions that have major impacts on
lime feed control. In addition, methods for controlling lime
feed rates are shown with the disadvantages and advantages of
each.
3.3.3.1 Nonlinearity—
The shape of the titration curve of an acid-base neutra-
lization is shown in Figure 3.3-3. This graph shows the re-
sponse of pH as increasing quantities of base are added to an
acidic solution. The curve does not have uniform slope.
Greater quantities of a base are needed to change the pH of a
solution from 4 to 5 than to change it from 5 to 6.
A standard controller is a linear device. It is adjusted
to add a certain quantity of lime to correct a pH change of a
certain magnitude; if the change is doubled, the rate of lime
addition is also doubled. This is not the best response for pH
control.
3.3-6
-------
10
9
8
o
6
5
4
3
INCREASING AMOUNT OF BASE
Figure 3.3-3. Shape of neutralization curve.
3,3-7
-------
If the acid-base neutralization is buffered, as it is in a
lime scrubber, the titration curve is slightly flatter though
less regular. There are "plateaus" where the mixture absorbs
quantities of lime, and there is little or no change in pH. As
the degree of buffering changes, the shape of the curve changes.
Buffered solution in a scrubber, however, is no more amenable to
standard linear control than is an unbuffered pure solution.
During the last several years, nonlinear controllers speci-
fically designed for control of pH have been produced by several
manufacturers. These controllers partially correct the mismatch
by supplying two bands with different amplifications. Within a
pH band near the set point, little amplification of the control-
ler output signal occurs. As pH change increases, larger ampli-
fication makes a larger change to lime feed rate to drive the pH
back more rapidly into the acceptable range. Controllers of
this type are in use in scrubbers. While not a complete answer
to the lime feed rate control problem, they are more suited to
this application than are standard linear instruments.
3.3.3.2 Time Lag—
Since the reaction of lime with acidic constituents in the
scrubber slurry is not instantaneous, tanks are normally in-
cluded in the process to hold the scrubber slurry and lime
mixture from 5 to 15 minutes before it is recycled to the scrub-
bing vessels. The best-controlled scrubber systems use baffled
or overflow chambers in the slurry recirculation tank. In some
systems, the important control point for pH is the slurry as it
is repumped; therefore, a time lag of several minutes is in-
herent in the process. However, this is not necessarily the
case. At Conesville, for example, the pH sample is taken from
the scrubber effluent before it reaches the reaction tank.
Therefore, the response is relatively rapid. The return pH
depends on the amount of lime added, which in turn depends on
the amount of SO2 to be absorbed. At Paddy's Run, for example,
return pH is 9 to 10; at Shawnee, it is 6.5 to 7.5.
3.3.3.3 Simple pH Control—
As shown in Figure 3.3-4, pH can be used to control lime
feed in a scrubbing system. The primary advantage of this
system is that it is simple. In many systems, the scrubbing
liquor has an ability to absorb sufficient SO2 so that this type
of control may be adequate. A major disadvantage is low lime
utilization. In some cases the lime feed rate is set using a
material balance at the highest S02 rate; therefore, the pH
controller only detects major high and low SO2 changes. in
addition, major SO2 variations in flue gas are only detected
after impacts on scrubber chemistry, scaling, and corrosion.
This disadvantage prevents consistent process control.
3.3-8
-------
LIME SLURRY FEED
LIME SLURRY
CONTROL VALVE
SLURRY FROM
SCRUBBER
O-Q
pH CONTROLLER
CO
pH PROBE
TO RECIRCULATION
PUMP
Figure 3.3-4. Simple pH control.
3.3-9
-------
3.3.3.4 Cascade pH Control—
Several of the newer lime scrubbing FGD systems are being
designed with either a closed- or an open-loop cascade control
system (Figure 3.3-5). Two pH controllers are used. The lime
feed pH is regulated by a secondary loop maintaining the pH in
the lime slurry mix tank. The primary pH controller, measuring
at the point where the recirculating slurry is returned to the
absorber, readjusts the set point of the secondary controller to
compensate for the varying offset. By greatly reducing time
lag, control is significantly improved.
In some systems, the primary controller is open loop. Only
a pH recorder is supplied, and the operator becomes the control-
ler, periodically readjusting the secondary instrument.
3.3.3.5 Outlet S02 Feedback Control--
The optimal feed rate of lime to a scrubber depends not
only on pH, but also on inlet SO2 concentration and the concen-
tration of other chemical elements. The pH set point that will
give adquate SO2 removal must be determined. This depends
mainly on where the pH sample is taken and on the mass transfer
capability of the scrubber (some excess lime may be necessary to
offset inadequate scrubber capability). Inlet SO2 concentration
may affect the latter factor, by requiring more excess when the
inlet S02 is high. Gas flow is also a factor if all the scrub-
bers are left operating at low load so mass transfer will be
improved. Magnesium in the lime will also require a different
pH set point. Most of these do not vary much in practice,
however, so a given set point may be adequate for an extended
period of time.
However, the use of outlet S02 concentration as a control-
led variable for lime feed rate is a more desirable concept,
since this is the variable that defines the quality of the
"product" of the scrubbing process. The eventual direct control
of a scrubber to maintain constant SO2 content in the outlet gas
is likely with SO2 limits and emission averaging times proposed
in the New Source Performance Standards. The primary advantage
of this system is that the response of lime feed to a measured
error in SO2 content is more nearly linear than the pH. The
time lag for direct SO2 control is about the same as pH control,
since it is set primarily by transport time through the hold
tank. The loop sensor used in direct SO2 control is an SO2
analyzer. The performance of these instruments and sampling
trains in some of the newer lime scrubbers has been rather poor.
Application of this control method, however, will require
pH control. Although pH can be varied to obtain the best S02
removal, it must be maintained within a range that will cause
neither corrosion nor scale formation. Instrumentation can be
3.3-10
-------
SLURRY FROM
SCRUBBER
LIME SLURRY
FEED
Co
•
u>
I
SECONDARY
pH CONTROLLER D
p—O O
PRIMARY
pH CONTROLLER
I
MIXING
COMPARTMENT
oU
HOLD
COMPARTMENT
CO
TO SCRUBBER
Figure 3.3-5. Cascade pH control.
-------
used to limit a feedback loop, as shown in Figure 3.3-6. The
time-synchronized recorder charts show the action that would be
expected. The lime feed rate would be controlled by the SO2
loop, providing the pH remains within limits. If pH reaches its
high limit, the high-limit pH controller would take over control
of the valve, preventing further addition of lime. Similarly,
the low-limit pH controller would prevent the S02 controller
from closing the lime feed valve too far. It should be greatly
emphasized that SO2 feedback control cannot be widely used until
wet SO2 analyzers operate reliably.
3.3.3.6 Inlet SO2 Feed-forward Control—
Feed-forward control of the lime feed rate, using measure-
ment of the inlet SO2 concentration, has also been suggested as
a possible means to control improvement, and the basic instru-
mentation has been included in some scrubber designs. Feed-
forward systems are the preferred method of control of lime feed
rate in scrubbers in other countries, especially Japan.
In this system, the flow rate and S02 content of the gas
entering the scrubber would be measured. Instruments measuring
the SO2 content of a dry gas have shown more reliable operation
in the field than have wet gas stream SO2 analyzers. From these
measurements, an instrumented calculation of the mass flow rate
of S02 can be made. The lime feed rate should be in proportion
to the quantity of SO2 entering the scrubber, and a feed rate
controller would be set accordingly.
The primary advantage of this system is that it responds to
SO2 and gas changes. However, although it appears to be simple
and basically sound, it suffers from three process disadvan-
tages, which may or may not be important in a specific appli-
cation:
0 Time lag is not eliminated since the principal lag is
in the hold tank, which is part of the manipulated
stream.
0 Proper operation presumes a constant efficiency of
utilization of the lime under all conditions of opera-
tion.
0 Outlet SO2 emissions do not change the lime feed rate.
Feed forward of the inlet SO2 concentration would probably
operate well if "trimmed," but trimming must be limited by
slurry pH. A schematic of this arrangement is shown in Figure
3.3-7.
3.3-12
-------
so2
ANALYZER
" V
HIGH LIMIT I
pH CONTROLLERS^
V - T0
J STACK
UJ
*~ ~~\
\
\
i
\
1 LOW LIMIT 1
J^i pH CONTROLLER J^
INTERNAL RESET
FEEDBACK ^
o
pH MEASURING
INSTRUMENT
SELECTING V. _/•*
EQUIPMENT x—
LIME FEED
VALVE
TIME _
•T2
A
pH
RECORD
so2
CONTROLLER
._!
HIGH LIMIT
pH
SET POINT
LOW LIMIT
PH
SET POINT
so2
SET POINT
Figure 3.3-6. Outlet S02 feedback control (limited)
3.3-13
-------
^LINEARIZERS
MULTIPLYING
INSTRUMENT
AUXILIARY
I INSTRUMENT
CONTROLLERS
AUXILIARY
INSTRUMENTS
Figure 3.3-7.
ADDING
INSTRUMENT
LIME FEED
CONTROL
ELEMENT
Inlet S02 feed-forward control
3.3-14
-------
3.3.3.7 Conclusions--
If the operability of wet SO2 analyzers improves, they
should be added to the system. New control systems should have
provisions for future feedback control. With the emphasis on
stricter emission limits and shorter emission averaging times,
outlet SO2 emissions will be monitored by the EPA. The in-
centive to develop workable SO2 monitoring equipment will exist.
However, to date, simple pH control has proved to be acceptable,
and it reduces the amount of instrumentation that can cause
problems. It is possible that such control causes more lime
usage than is absolutely necessary, but this hypothesis is
difficult to prove. In any event, the excess merely gives
higher SO2 removal than the regulation requires.
3.3.4 slurry Solids Control
In a lime slurry system, there are three areas where the
solids content of the slurry is controlled: the lime slurry
feed, the absorber recirculation loop, and the thickener under-
flow. Although solids content can vary without being critical
to the operation of the absorber, the use of a consistent lime
slurry can reduce plugging and deposits in the absorber, reduce
waste volume from the thickener, and improve pH control.
To avoid fouling of the sensor, each control area uses
indirect measurements of the slurry density by magnetic and
nuclear density sensors. These devices measure the absorptive
properties of the slurry and correlate them to the solids con-
tent of the slurry. Although the correlation is not exact, this
type of measurement has proved highly accurate for most appli-
cations .
3.3.4.1 Lime Slurry Feed Control—
As shown in Figure 3.3-8, lime slurry concentration can be
controlled by measuring the density of the slurry leaving the
stabilization tank. A simple feedback system uses freshwater
makeup or thickener overflow to reduce or increase the solids
content as required. The advantage of this system is simpli-
city; the disadvantage is the control time lag.
If a gravimetric feeder is used for lime feed, then a
feed-forward system with slurry density feedback trim should be
used. This system is illustrated in Figure 3.3-9. The advan-
tage of this system is more uniform control; the disadvantage is
system complexity and added cost.
3 3.4.2 Absorber/Scrubber Solids Control—
To prevent a buildup of the absorption products, the solids
content of the recirculation slurry must be controlled. Al-
though there are several theories on concentrated slurries and
diluted slurries, which will not be discussed here, the solids
3.3-15
-------
MAKEUP WATER OR THICKENER OVERFLOW
CONTROL VALVE
LIME STABILIZATION TANK
O
CONTROLLER
I
J
DENSITY SENSOR
TO SCRUBBERS
Figure 3.3-8. Lime slurry solids control: density feedback.
3. 3-16
-------
FRESHWATER MAKEUP OR THICKENER OVERFLOW
LIME SILO
GRAVIMETRIC
CONTROL VALVE
GRAVIMETRIC
CENSOR
\
\
SLAKER
GRAVIMETRIC
CONTROLLER
DENSITY
CONTROLLER
DENSITY
SENSOR
STABILIZATION
TANK
ABSORBER!
LIME
PUMP
Figure 3.3-9. Lime slurry solids content:
feed-forward control.
3.3-17
-------
content must be controlled to minimize plugging and buildup and
reduce the load on solids concentrating equipment such as the
thickener. The simplest system for controlling the solids
content in the absorber loop is a density sensor and a bleed to
the thickener or pond. In this system, which is shown in Figure
3.3-10, a level controller adds more fresh-water or thickener
overflow as the slurry is bled off. This dilutes the slurry to
the required extent. As the solids content decreases, water
makeup and the thickener bleed are reduced.
Although such a system has the advantage of simplicity, it
is limited by severe wear on control valves. One method of
solving the problem of the eroding control valves is to add a
variable-speed pump to the bleed line, as shown in Figure
3.3-11. As the solids content rises, solids would be purged to
the thickener, and additional water would be used to lower the
solids content. With this system good solids control is pos-
sible. The disadvantages are system complexity and the greater
cost of the additional pumps.
3.3.4.3 Thickener Solids Control—
This section discusses techniques for both constant and
intermittent solids control in the outlet and overflow of the
thickener when the feed to the thickener varies both in solids
content and quantity.
One approach for thickener solids control is shown in
Figure 3.3-12. In this case, a control valve is used to vary
the underflow rate to maintain solids content in the thickener.
The advantage of this system is uniform solids content in the
underflow. The disadvantages are nonuniform underflow flow
rates, lack of control of the overflow, and excessive wear of
the control valve. A no-flow situation is not tolerable, and a
certain amount of control is sacrificed to maintain a minimum
underflow.
Another method of solids control is shown in Figure 3.3-13.
In this system, the underflow is constantly recycled to the
thickener with a bleed stream going to the downstream sludge
treatment. The advantage of this system is uniform pump opera-
tion and more uniform solids level in the underflow without the
problems of maintaining a minimum underflow. However, again a
control value is required, which will be worn by abrasive
slurry. Other disadvantages of the system are the need for an
oversized pump and the pump's excessive energy requirements.
3.3.4.4 Flocculant Control—
Provisions should be made in the solids control system for
flocculant addition. Although it is expensive to allow for some
oversize in thickeners, thickener optimization is difficult with
3.3-18
-------
FRESHWATER MAKEUP
OR THICKENER OVERFLOW
LEVEL
CONTROLLER
U)
i
H
vo
DENSITY
CONTROLLER
THICKENER
BLEED
CONTROL
VALVE
RECIRCULATION
PUMP
Figure 3.3-10. Absorber solids content:
thickener bleed control.
-------
OJ
•
U)
I
LEVEL
CONTROLLER
a
\
CONTROL VALVE
LEVEL
SENSOR
ABSORBER
DENSITY
SENSOR
DENSITY
-{ J CONTROLLER
in
11
11
11
TO THICKENER
RECIRCULATION
PUMP
VARIABLE-SPEED
BLEED PUMP
Figure 3.3-11. Absorber solids content: variable-speed pump control.
-------
THICKENER
L
DENSITY
CONTROLLER
o-
TO POND
-*-
CONTROL VALVE OR VACUUM
SENSOR
Figure 3.3-12. Thickener solids content:
underflow control flow.
RECYCLE
THICKENER
DENSITY
CONTROLLER
CONTROL
VALVE
DENSITY
SENSOR
CONTROL VALVE
Figure 3.3-13. Thickener solids content-
recycle control.
3.3-21
-------
the current state of the art. Flocculants can be used to solve
solids content problems in systems in which the thickener is
undersized or the chemical properties of the coal and lime
change and cause thickening problems. A typical flocculant
control system is shown in Figure 3.3-14. The flocculant feed
is varied with the underflow solids content. This system has a
slow response because settling rate changes are slow compared
with gas flow variations. True regulated control is probably
not possible. Flocculant addition should be manually adjusted
to allow the solids content of the underflow to remain within
certain limits.
One of the previously discussed control methods should be
used to control the precise solids content of the thickener
underflow.
3.3.4.5 Solids Control Summary—-
Solids content in a scrubbing system cannot be controlled
adequately by the individual loops discussed above. Since the
solids content of the scrubber and thickener are interrelated,
the solids should be controlled as a system. In a plant in
which the water loop is closed, a system such as that shown in
Figure 3.3-15 is feasible. In this system, the bleed from the
absorber is restricted with an orifice plate, the thickener acts
as a surge for the variations in solids loading on the system,
and a variable-speed pump is used to control the solids content
of the thickener underflow.
Although the system has numerous limitations, which have
been discussed above, it does incorporate most of the best
features of a satisfactory system. Although solids content in
the scrubber will vary with large shifts in gas flow or compo-
sition, some variation will not be detrimental to SO2 absorp-
tion. Most of the wear problems are solved by the orifice plate
and the variable-speed pump. The danger of line plugging is
solved since the system is continuous. Manual recycle may be
required during startup and shutdown of the thickener when
solids production in the absorber is low.
An alternative arrangement would be to use variable-speed
pumps for both thickener feed and underflow, since the slightly
increased cost would definitely increase reliability and con-
trollability.
3.3.5 Gas Flow Control
A lime scrubbing system usually needs distribution of stack
gas flow. Gas volume is controlled by the boiler. The scrub-
bing system must respond and not affect boiler operation. if
multiple scrubbing units are used, a dependable system to bal-
ance gas flow rates in the parallel units must be provided. in
3.3-22
-------
FLOCCULANT FEED
THICKENER
VARIABLE-SPEED
FLOCCULANT PUMP
V
DENSITY
CONTROLLER
TO POND
OR VACUUM
DENSITY SENSOR
Figure 3.3-14. Thickener solids content:
flocculant control.
3.3-23
-------
U)
•
U)
FRESHWATER MAKEUP OR THICKENER OVERFLOW
LEVEL CONTROLLER
DENSITY CONTROLLER
SCRUBBER
ORIFICE
PLATE
RECIRCULATION
PUMP THICKENER ' L
MANUAL RECYCLE
DENSITY
CONTROLLER
-O
I />
W
POND
VARIABLE-SPEED
PUMP
DENSITY
SENSOR
Figure 3.3-15. Absorber: thickener solids control system.
-------
a retrofit installation of a scrubbing system, coordination
between the boiler and scrubbing controls is necessary. Even in
new units, the control coordination used for retrofit scrubbers
is needed, since the boiler and the scrubber are most often
designed as separate units and frequently have separate control
rooms.
3.3.5.1 Pressure Control—
The simplest method of controlling gas volume through each
scrubbing vessel is the use of pressure control. A constant
pressure is maintained at the inlet of the absorber. This
pressure is designed so that it does not impair the operation of
the boiler by increasing the back pressure on the induced-draft
fans of the boiler.
It has been demonstrated that simple feedback control, as
shown in Figure 3.3-16, responds too slowly to maintain an
adequately consistent pressure. A surge of pressure occurs with
each change of boiler firing rate. A connection from the boiler
control system is necessary, usually in the form of a feed-
forward signal from the boiler combustion controls. The basic
agreement is shown in Figure 3.3-17. Feed forward alone is
insufficient. However, since no two fans or dampers will re-
spond identically, feedback correction is necessary to prevent
variations in pressure with changes in load. The instrumenta-
tion is similar to that used to balance the firing rate of
parallel boilers.
3.3.5.2 Flow Control—
In multiple module scrubbing systems, it would be helpful
to control the gas volume to each scrubbing module precisely.
This requires one fan per module. However, flow is difficult to
measure because of nonisokinetic flows in short duct runs. Long
duct runs, which achieve the isokinetic conditions, are not
economical. A possible system for flow control is shown in
Figure 3.3-18. Because of the inaccuracies of flow measurement,
the damper control is tripped with pressure control. Boiler
feed-forward control is also fed into the system. The advantage
of this system is that the flow rates for each module can be
accurately controlled. The disadvantages of this system are
complexity and flow measurement inaccuracy. Precise flow rates
through each scrubber module are not required if the system can
meet the removal efficiency at full load. At lower gas loads
excess SO2 would be removed. Properly implemented, such a
system would allow some of the modules to be baseloaded and
others to vary according to changes in boiler firing.
3.3.5.3 Boiler Safety—
The most difficult problems of gas flow control arise in
the protection of the boiler-scrubber system from explosion or
3.3-25
-------
FLUE GAS
FROM BOILER
PRESSURE
/^"VONTROLLER
PRESSURE
DAMPER SENSOR
ACTUATOR
1
SCRUBBER FAN
SCRUBBER
Figure 3.3-16. Simple pressure control.
3.3-26
-------
FLUE GAS FROM
BOILER
O PRESSURE
CONTROLLER
DAMPER
ACTUATOR
5 FROM BOILER
COMBUSTION CONTROL
INSTRUMENTATION
FEED-FORWARD
INSTRUMENT
I
PRESSURE
SENSOR
SCRUBBER
SCRUBBER
FAN
Figure 3.3-17. Basic fan control subsystem.
3.3-27
-------
FLUE GAS
FROM
BOILER
_ FEED-FORWARD
T SIGNAL FROM
I BOILER
CO
*
U)
I
K>
00
FLOW CONTROLLERS
PRESSURE
CONTROLLER
LINEARIZERS
TO OTHER
MODULES
PRESSURE
CONTROLLER
TO OTHER
*MODULES
FLOW
SENSOR
DAMPER
•//•///
w
r -r r
i i •
1 i SCRUBBER
DAMPER |
ACTUATOR |
PRESSURED
SENSOR rS
^
T
- }
/
FLOW
SENSOR
DAMPER
/•/•/•/•/
\l
DAMP
ACTU/l
I —
l__
PP
c
^
\~
r r
i i r
J i i
ER | 1
TOR |
>* — - I
.ESSUREr -1
>ENSORrLi
T
i
SCRUBBER
- }
Figure 3.3-18. Module load balancing control.
-------
implosion damage on trip-out of the boiler or on loss of a
scrubber fan. When the boiler shuts down, there is either a
sudden increase or decrease in gas flow rate, depending on the
safety requirements of the boiler. Although interlocks will be
used to achieve simultaneous shutdown of the boiler and the
scrubber, pressure or vacuum surge can develop in the boiler and
duct work if the dynamic response to this condition is unfavor-
able. Similarly, loss of a scrubber fan will produce a pressure
surge in the opposite direction that will trip the boiler, but
may also create potentially damaging surges.
If only part of the flue gas is passed through the scrubber
and the remainder is bypassed through a damper, connection of
the bypass damper and the scrubber fans to the boiler flame
safeguard system is an acceptable solution. In the event of an
emergency boiler shutdown, the scrubber fan is shut down and the
damper is opened. If the scrubber fan fails, the damper is
opened and an operator can conduct a more orderly shutdown of
the boiler. Guidelines for this interconnection are being
prepared by insurance standards organizations.
If the scrubber is not bypassed, however, the solution is
much more complex. It has been found that conventional steady-
state engineering analysis is inadequate to deal with this
problem. Unusual flow conditions are created by flame collapse
on unit shutdown, which can reverse flue gas flow direction and
cause substantial implosion damage if outside air is not admit-
ted into the unit. At the present time, each scrubber instal-
lation requires individual mathematical dynamic simulation
studies to predict the effect of boiler failures and to aid in
the design of dampers and interlocked trip sequences to reduce
the possibility of boiler damage.
3.3.5.4 Fan Control—
Another possible method of flow control through individual
modules is the use of variable-speed fans. A typical control
loop is shown as Figure 3.3-19. This system uses the pressure
at the scrubber inlet and a feed-forward signal to regulate fan
speed. The advantages of this system include some reduction in
energy consumption in the fan and better flow control than
simple pressure control allows. The disadvantages include slow
control response and higher capital costs. The slow response is
due to the inertial of a large fan, which responds slowly to
control signals.
There have been wide differences in fan location among
operating lime scrubbers. Individual fans have been provided
with each module, but in some units one fan serves several
modules. Fans have been located upstream and downstream from
the scrubber and also between the scrubbing modules.
3.3-29
-------
BOILER
FLUE
GAS
BOILER COMBUSTION
CONTROL INSTRUMENTATION
PRESSURE
CONTROLLER
/
/
/
/ PRESSURE
/ SENSOR
SCRUBBER
VARIABLE-SPEED
SCRUBBER FAN
Figure 3.3-19. Flue gas flow: fan control.
3.3-30
-------
BIBLIOGRAPHY
Athans, M. , and P. Falb. Optimal Controls. McGraw-Hill New
York, 1966.
Del Toro, V., and S. Parker. Principles of Control Systems
Engineering. McGraw-Hill, New York, 1960.
Doennebrink, F., and J. Russell. LEM Stabilization and Control
System. AIAA Guidance and Control Conference, August 1965.
Dorf, R.C. Modern Control Systems. Addison-Wesley Publishing
Co., Reading, Massachusetts, 1966.
Dorf, R.C. Time-Domain Analysis and Design of Control System
Addison-Wesley, Reading, Massachusetts, 1965.
Franklin, G. Introduction to Modern Control Theory Holden-
Day, San Francisco, 1967.
Kuo, B.C. Automatic Control Systems. Prentice-Hall, Englewood
Cliffs, New Jersey, 1962.
Lime/Limestone Scrubber Operation and Control Studv EPRI
(RP630-2), April 1978.
Marks, L.S. Standard Handbook for Mechanical Engineering.
McGraw-Hill, 7th ed., 1967. pp. 16-33.
Thaler, G.J., and R.G. Brown. Analysis and Design of Feedback
Control Systems. McGraw-Hill, 2nd ed., New York, 1960.
Truxal, J.G. Automatic Feedback Control System Synthesis.
McGraw-Hill, New York, 1955.
3.3-31
-------
CONTENTS
3.4 BASIC CONTROL HARDWARE
Page
3.4.1 Introduction 3.4-1
3.4.2 The Controller 3.4-1
3.4.3 Control Modes 3.4-3
3.4.4 Control Element Characteristics 3.4-4
Bibliography 3.4-11
3.4-i
-------
3.4 BASIC CONTROL HARDWARE
3.4.1 Introduction
In most control loops, two interrelated items of instrumen-
tation combine to regulate the flow rate of a manipulated stream
of process material. This section of the Data Book describes
the interrelationship of the controller and the control element
(control valve) in greater detail than was previously presented,
defines some of their specialized features, and emphasizes
problems of compatibility that could reduce the performance of a
control loop.
3.4.2 The Controller
A process controller is an analog (mechanical) computer
that performs a continuous mathematical manipulation based on
the error (change) that exists between the controlled variable
and the controller set point. Most texts on process control use
advanced and specialized mathematics to describe the action of a
controller. A working description, however, can be expressed
using conventional mathematics.
Generally a feedback control instrument, whether electronic
or pneumatic, is based on the principle of the proportional,
integral, and derivative (PID) controller. The controller
continuously readjusts its output signal using an equation that
has this general form:
outputt = ^ et + k2Jedt + k3 jjj|
where e. = error at time "t"
k1,k2,k3 = adjustable controller constants
This equation states the output is the sum of
1. The error existing at the instant multiplied by kj .
2. An integral term that is the sum of all the errors,
both positive and negative, that existed during the
time period when the controller was first placed in
operation and between time t multiplied by k2,
3. A derivative term that represents the speed with which
the error is increasing or decreasing at a measured
instant multiplied by k3.
In many controllers the value of ka is expressed as its
reciprocal, the proportional band, PB. The constants k2 and k3
3.4-1
-------
are divided by kx so that an adjustment of proportional band
automatically changes the integral and derivative constants.
Thus the controller form is
3 dt
OUtpUt , = -- - pg
where
kj = 1/PB
K2 = k2/PB
K3 = K3/PB
When a technician adjusts the knobs, levers, or push-but-
tons on a control instrument to tune the control loop to obtain
best operation, he is adjusting the values of the three tuning
constants K2 , K3 , and PB. Proportional band is expressed in
the units "percent of scale." With a small proportional band (5
to 20 percent), a large change in output occurs with even a
small error. With a large proportional band (500 to 800 per-
cent), output changes very little upon detection of an error.
The integral adjustment K2 sets the relative importance of
the integral mode in modifying controller output. Integral is
synonymous with reset, which was coined years ago as an adver-
tising term. The integral or reset adjustment is expressed in
the unit "repeats per minute," or sometimes by its reciprocal,
"minutes per repeat" or an equivalent, "seconds per repeat."
The units relate to a standardized test that can be used to
measure the value of the constant. With a large "repeats per
minute" (0.5 to 20), a continuing error rapidly changes the
controller output; this is described as "fast" reset. With a
small "repeats per minute" (0.02 to 0.1), the effect of the
integral mode in the equation is much reduced, and this is
described as "slow" reset.
Derivative adjustment is also expressed in "repeats per
minute" or its reciprocal, also relating to an empirical test
procedure. In most scrubber applications, the derivative mode
will be adjusted between 0.1 and 10 repeats per minute. This
mode will usually have less importance than the integral mode in
modifying controller output.
3.4.3 Control Modes
Most control loops do not require the use of all three
modes- therefore, controllers are built with one or two modes
omitted A one-mode controller contains only proportional
control" action; constants K2 and K3 are set equal to zero. A
controller of this type may be required in some feed forward or
multiinput control systems where only proportional or ratio
action is suitable. One-mode control may also be used in simple
control loops where accuracy is unimportant, such as control of
3.4-2
-------
liquid level in a tank. In most cases, however, the dampening
effect of the integral mode is needed in scrubber feedback
loops.
By far the majority of control problems are handled best by
a two-mode controller, which contains only proportional and
integral modes (K3 = O). The integral mode causes changes in
output to occur more slowly, resulting in fewer surges in the
manipulated variable. Of more importance, tuning the integral
mode allows the response rate of the controller to match the
response rate of the process. With integral mode, control
following a disturbance is restored more rapidly and accurately
than with proportional control alone.
Only a small percentage of processes require a three-mode
controller. While extremely useful in some loops, derivatives
cannot be used in others. In contrast to the delaying action of
the integral mode, a derivative ±s designed to overreact to
small errors. For example, if the pH of the slurry in a scrub-
ber begins to drop, immediate control action will restore pH
more quickly. In this process with a long time lag, the deri-
vative mode responds neither to how far pH has dropped, nor to
how long it has been away from set point, but to the rate at
which it is dropping. Derivative control would add an extra
volume of lime to halt the decrease. When pH stops dropping,
the action of the derivative mode would cease. By that time,
however, the integral mode, acting more slowly, would be gradu-
ally increasing the lime feed rate. Eventually the pH begins to
rise, and the derivative mode would act in opposition to the
integral mode to prevent overshoot. Properly tuned, derivative
control substantially reduces the severity of errors caused by
process disturbances. One manufacturer uses the copyrighted
name "Pre-Act" to describe its brand of three-mode controllers;
the name accurately describes the apparent action. The deriva-
tive mode, however, cannot distinguish between a genuine change
in the measured variable and a short-term transient change, such
as those occurring normally in a flowing stream of material in a
pipeline. Derivative control can therefore only be used in
loops where a fairly large volume of process fluid is in contact
with the sensor, and where the sensor provides a steady signal,
free from electronic or process noise. In a lime scrubber,
control of pH is a potential application.
Figure 3.4-1 illustrates the response of various control-
lers in restoring control following a sudden load change or
disturbance. While this figure is theoretical, recorded curves
of this type are obtained in actual plant loops.
In Table 3.4-1, recommended control modes are shown for the
type of controls found in lime scrubbing systems.
3.4-3
-------
ONE-MODE
TWO-MODE
RESET TOO
FAST
TWO-MODE
RESET TOO
SLOW
TWO-MODE
RESET
OPTIMUM
THREE-MODE
DISTURBANCE
TIME
Figure 3.4-1. Theoretical controller response curves,
3.4-4
-------
Table 3.4-1. RECOMMENDED CONTROL MODES
pH
Level
Solids
Gas volume
Reheat
Proportional
X
X
X
X
X
Integral
X
X
(x)a
X
Derivative
X
a Not required if proportional control is based on actual pressure
vs. flow rate measurements.
3.4.4 Control Element Characteristics
A detail of control system design closely related to the
action of a controller is the response characteristic of a
control valve or other control element. The proportional band
of a controller is always adjusted to change the controller
output by a definite amount in order to correct an error of a
certain magnitude. This ratio is dictated by the response of
the process to a certain change in the manipulated variable. It
follows, therefore, that for ease of design the manipulated
variable should change in direct linear ratio with the control-
ler output. In other words, an incremental change in controller
output should produce an incremental change in the manipulated
variable, regardless of the initial value of the manipulated
variable. If this is not the case and if the controller is
tuned when the process is operating at full load, the controller
will apparently be out of tune when the process is dropped to
half load. In many operating control loops, the controller can
be tuned to operate well only over a narrow range of operating
flows. Above or below this range, the proportional band of the
controller must be readjusted for the loop to operate properly.
The fault is not with the controller but with the operating
characteristic of the control element. it should be remembered
that if one element of a control loop is nonlinear, the loop is
nonlinear (for example, pH sensors behave nonlinearly).
Linear loops are not inherently better than nonlinear
loops, although they are much easier to design. Stability is
easier to maintain in linear systems. Some variables can be
easily transformed to linear forms. Thus, sensing pressure
drop, AP, in an orifice plate requires only a square root func-
tion to produce flow rate. A digital controller accomplishes
this with ease.
3.4-5
-------
Some control valves are manufactured with a linear control
characteristic. At a constant pressure drop, valves of this
type produce a flow rate that is directly proportional to the
value of the actuating signal (Figure 3.4-2a). A control loop
that includes a valve of this type will provide accurate control
under any loading conditions, providing all other conditions
retain a linear relationship and pressure drop through the
control valve remains constant.
In most instances, however, other conditions do not remain
constant; valve friction increases and pump discharge pressure
decreases as the flow rate increases. Therefore, pressure drop
across the control valve may decrease significantly as flow rate
increases. A valve with a linear characteristic does not pro-
duce a flow rate proportional to controller output if valve
pressure drop varies as flow rate changes.
In a practical development some years ago, control valves
were designed so the controller output signal would be propor-
tional to the logarithm of the flow rate passing through the
valve at constant pressure drop (Figure 3.4-2b). Valves of this
type are said to have an equal percentage control characteris-
tic. They were designed to match more closely the pattern of
pressure and flow created by conventional chemical engineering
design practices, using a centrifugal pump to deliver the mani-
pulated stream. In some loops, for instance, the equal percen-
tage valve characteristic is a very close match to operating
conditions, and a wide range of operating rates can be accommo-
dated without necessitating retuning of the controller.
In many control loops, however, neither a linear nor an
equal percentage valve characteristic is sufficient. To obtain
best control, these loops ought to be supplied with individually
designed valves. However, valves with special characteristics
are not manufactured, since there are other ways to control the
system. In loops where the manipulated variable is always
expected to operate within a narrow range of flow rates, it is
advisable to accept the mismatch and choose between the two
commercially-available characteristics. In general, if the
configuration of the loop is such that little change in pressure
drop will occur with a change in flow, the linear characteristic
is better. In loops with long piping runs, where pressure drop
due to friction is significant, the equal percentage character-
istic is better. A rule of thumb for various types of conven-
tional control loops is shown in Figure 3.4-3. Since flow of
heat is not proportional to flow of heating medium, equal per-
centage valves are better for almost all temperature control
loops. In fact, in no loop is flow of the manipulated variable
in exact linear ratio to its effect on the controlled variable,
except in direct flow rate control loops.
3.4-6
-------
100
% OF
CONTROL
VALVE
ACTUATING
SIGNAL
25 50 75
% OF FLOW AT CONSTANT VALVE A P
100
a. Linear control characteristic.
100
% OF
CONTROL
VALVE
ACTUATING
SIGNAL
10 20 30 50
% OF FLOW AT CONSTANT VALVE A P
b. Equal percentage control characteristic.
100
Figure 3.4-2. Valve characteristics.
3.4-7
-------
OJ
00
CONTROLLED VARIABLE
FLOW, SQUARE-ROOT
FLOW, LINEARIZED
PRESSURE
LIQUID LEVEL
TEMPERATURE
PERCENT OF TOTAL SYSTEM PRESSURE
DROP ABSORBED BY THE CONTROL VALVE
AT DESIGN FLOW RATE.
10 20 30 40 50 60 70 80
1 | ! 1 ' 1 1 1
1 1 1 1 1 | 1 1
l 1 1 • 1 i 1 1
'////////l
l i 1 i 1 1 | l
1 II
//////////////// /j
i i i i i i i l
i ' i i 1 i 1 i
//////////////// //////JJA
i i i i i i i i
i i i i i 1 1 i
////////////i
i i 1 i 1 i i i
i i i i I i 1 i
LLLZ 1.LJ ////////////////////////////////////,
1 1 1 1 1 I I |
90 100
1 i
! 1
I i
1
< i
1
i i
i i
1
i i
i
i
'/////////l
I 1
LEGEND:
- EQUAL PERCENTAGE CHARACTERISTIC
I - LINEAR CHARACTERISTIC
REFERENCE: PUBLICATION, FOXBORO
COMPANY, 1974
Figure 3.4-3. Selection of control valve characteristics.
Source: Foxboro Co., 1974.
-------
a
If the loop control must operate properly over a wide range
of flow rates, several alternatives can be used. If a digital
computer is being used, rather than a conventional analog in-
strument, the proportional band is simply readjusted as con-
troller output changes. This is one of the advantages of digi-
tal process control.
With analog control, correction is possible through the use
of cascade control. As described previously, in cascade control
a flow loop is added to the manipulated variable, and the output
of the primary controller manipulates the set point of the
secondary flow controller. if the flow rate sensor produces a
signal that is linear with flow, the primary controller "sees" a
linear characteristic. Good control is achieved over a wide
range. Cascade control does not eliminate the mismatch; it
merely transfers it to the flow control loop. However, control
of flow rate can be less than optimal without causing an error
in the much-slower primary loop.
An alternative method is to add a cam-type valve positioner
to the control valve. This device is interposed between the
controller and the control valve actuator. An internal cam is
cut to a slope that modifies the signal in a manner that
achieves, for a specific loop, the desired linear flow rate.
While less expensive in initial cost, this method requires the
services of an experienced field technician to determine and
produce the proper cam configuration. This method was widely
used at one time but is now rarely employed because of the wider
use of digital control.
The greatest problems with mismatch in valve characteristic
will occur when valves with an inherently nonlinear characteris-
tic are used. Valves such as butterfly, ball, knife, gate, and
pinch are desirable in other respects, but they perform very
poorly when used without supplemental devices in control appli-
cations. To obtain wide range control with these valves, it is
essential that cam-type valve positioners be used. The appli-
cation of positioners to these valves does not necessitate
extensive field work, since vendors can usually supply precut
cams that convert the valves to the equivalent of linear or
equal percentage control characteristics. However, none of
these valves should be used as control valves in erosive slurry
systems.
If variable-speed pumps are used instead of valves, as the
control element, the same characteristic principles apply
Variable-speed centrifugal pumps may have an extremely nonlinear
response and may require special mechanical or electronic acces-
sory devices to achieve adequate control. On the other hand
positive displacement pumps with variable drive motors most
often have very good control characteristics
3.4-9
-------
BIBLIOGRAPHY
Chestnut, H. Systems Engineering Tools. Wiley, New York, 1965;
p. 613 .
Clark, R.N. Introduction to Automatic Control Systems. Wiley,
New York, 1962.
Doeblin, E.G. Dynamic Analysis and Feedback Control. McGraw-
Hill, New York, 1962.
Dorf, R.C. Modern Control Systems. Addison-Wesley Publishing
Co., Reading, Massachusetts, 1966.
Electro-Craft Corporation. Motomatic Speed Control. Hopkins,
Minnesota, 1964.
Horowitz, I.M. Fundamental Theory of Automatic Linear Feedback
Control System. IRE Trans, on Automatic Control, Dec. 1959.
Kuo, B.C. Automatic Control Systems. Prentice-Hall, Englewood
Cliffs, New Jersey, 1962.
Lime/Limestone Scrubber Operation and Control Study. Report
prepared by Southern California Edison Co., EPRI (RP630-2),
April 1978.
Marks, L.S. Standard Handbook for Mechanical Engineering.
McGraw-Hill, 7th ed., 1967. pp. 16-33.
Schultz, W.C., and V.C. Rideout. Control Systems Performance
Measures: Past, Present, and Future. IRE Trans, on Automatic
Control, February 1961.
3.4-10
-------
REFERENCES
1. Foxboro Company. 1974.
3.4-11
-------
SECTION 4
EQUIPMENT DESIGN
4.1 INTRODUCTION
This section includes design information, to supplement
that with which a design engineer should be familiar and to
assist in determining the type of equipment needed for a parti-
cular FGD system. The following topics are discussed:
Section Equipment
4-2 Recirculating Pumps
4-3 Other Process Pumps
4-4 Lime Unloading and Storage
4-5 Slurry Preparation
4•6 Scrubber/Absorber
4-7 Mist Eliminator
4-8 Fans
4•9 Thickener/Clarifier
4-10 Mechanical Dewatering Equipment
4-ll Reheaters
4-12 Corrosion
4-13 Instrumentation
The characteristics of the equipment, design considerations
and criteria, materials of construction, and a review of the
equipment used at various lime FGD system installations are
presented. Specific areas of concern such as the causes of
equipment failure (insofar as they are known) are reviewed for
operational installations.
The emphasis of this section is on data that must be con-
sidered in order to design the best operational lime FGD system
for an individual site.
4.1-1
-------
CONTENTS
4.2 RECIRCULATION PUMPS
4.2.1 Introduction
4.2.2 Design Criteria
4.2.2.1 Service Description
4.2.2.2 Flow/Head
4.2.2.3 Net Positive Suction Head
4.2.2.4 Pump Efficiency and Energy Requirements
4.2.2.5 Impellers
4.2.2.6 Drives
4.2.2.7 Seals
4.2.2.8 Materials of Construction
4.2.2.9 Vendor Specifications
4.2.2.10 Auxiliary Design Considerations
4.2.3 Performance Histories
References
4.2-i
-------
4.2 RECIRCULATION PUMPS1'2
4.2.1 Introduction
The purpose of this section is to supplement the design
engineer's basic knowledge by an analysis of recirculation pump
design. Emphasis is placed on features that are unique to the
design of slurry recirculation pumps used in lime FGD systems.
The recirculation pumps are the largest pumps in a lime FGD
system (Figure 4.2-1), with capacities ranging from 5000 to
15,000 gal/min. They receive the slurry directly from the
bottom of the scrubber or from a reaction/hold tank. The dis-
charge slurry is continuously recirculated through the absorber.
Normally, a portion of the recirculation stream is bled to the
solids disposal system. Occasionally, the pumps may have vari-
able speed drive to allow liquid flow control, but as a rule the
pumps operate at constant speed and supply constant liquid flow,
even though the flue gas flow may be variable.
A typical slurry pump has many features (Figure 4.2-2) that
set it apart from the typical centrifugal pump used for clear
liquids. Wall thicknesses of wetted-end parts (casing, impel-
ler, etc.) are greater than in conventional centrifugal pumps.
The cutwater, or volute tongue (the point on the casing at which
the discharge nozzle diverges from the casing), is less pro-
nounced in order to minimize the effects of abrasion. Flow
passages through both the casing and impeller are large enough
to permit solids to pass without clogging the pump. Since the
gap between the impeller face and suction liner will increase as
wear occurs, the rotating assembly of the slurry pump must be
capable of axial adjustments to maintain the manufacturer's
recommended clearance. This is critical if design heads, capa-
cities, and efficiencies are to be maintained. Other special-
ized features include extra-large stuffing-boxes, replaceable
shaft sleeves, and impeller back-vanes that act to keep solids
away from the stuffing box. Although the impeller back-vanes
also reduce axial thrusts by lowering stuffing-box pressures,
these vanes can wear considerably in abrasive services. Hence,
both the radial and the axial-thrust bearings on the slurry pump
are heavier than those on standard centrifugal pumps.3
Because recirculation pumps handle abrasive slurry, their
design involves special considerations, many of them related to
the selection of materials. Recirculation pumps are available
in a variety of materials of construction to handle the abra-
sion, corrosion, and impact requirements of the solids-handling
application.
4.2-1
-------
Figure 4.2-1. Scrubber slurry recirculation pump.
Source: A-S-H Corporation
4.2.2
-------
NJ
I
,11*1-1 'PIA. i>TD. COAftSE fHRD
f TAPPCO HOLEi,
EOJALUY
WR! NCH USEf ro«
REMOVING mctu.cn
Figure 4.2-2. Slurry recirculation pump detail assembly (A-S-H, DG95-pump)
(continued)
-------
Legend for Figure 4.2-2
No. Name
1 Shell half, suet. side(S.S
2 Shell half, eng. side (C.S.
3 Shell half liner, s.s.
4 Shell half liner, e.s.
5 Cap screw, 1-1/4 in.
6 Nut, hex., 1-1/4 in.
7 Spacer
8 Bell, suction side (S.S.)
9 Throat liner
10 Side liner
11 Spacer ring
12 Car screw, fl. hd.
3/4 in.
13 Stud, 1 in.
14 Washer, 1 in.
15 Nut, hex., 1 in.
16 Bolt, hex. hd., 1-1/4 in.
17 Washer, f1., 1-1/4 in. -
18 Stuffing box
19 Stud, 3/4 in. (stuffing
box/shell)
20 Washer, fl., 3/4 in.
21 Nut, hex., 3/4 in.
22 Lantern ring
23 Packing ring, gland
24 Gland half
25 Stud, 3/4 in. (gland/
stuffing box)
26 Cap screw, hex., 1/2 in.
27 Nut, hex., 1/2 in.
28 Cap screw, hex, 1-1/4 in.
29 Washer, fl., 1-1/8 in.
30 Washer, fl., 1-1/4 in.
31 Impeller
32 Impeller clamp plate
33 Shaft sleeve
34 Jackscrew, sq. hd.,
1 in.
35 Stud, 1-1/4 in.
36 Name plate
37 Shaft
38 Shaft spanner wrench
\
39 Bearing housing
40 Bearing housing cap
No. Name
•) 41 Bearing housing cover
) 42 Bearing housing cover
43 Gar lock seal
44 Hydraulic packing
45 Hydraulic packing
46 Adaptor sleeve
47 Roller bearing
48 Bearing lock washer
49 Bearing lock nut
50 Bearing spacer
51 Roller thrust bearing
52 Split spacer
53 Thrust collar
54 Spring
55 Spring retaining ring
56 Bearing retaining ring
57 Socket hd. cap screw
58 Lock washer, 1/2 in.
59 Adjusting plug pin
60 Adjusting plug
61 Adjusting plug cover
62 Retaining chain assembly
63 Locking pin
64 Locking pin nut,
1-5/8 in.
65 Jam nut, hex.,
1-5/8 in.
66 Cap screw, 1 in.
(cap/hsg.)
67 Washer, lock 1-in. std.
68 Oil gauge
69 Pipe plug, 1-1/2 in.
70 Air vent
71 Service ell, 1/8 in.
x 45 degrees
72 Pipe clip
73 Grease fitting No. 1610
74 Washer, 5/8 in.
75 Cap screw, 5/8 in.
76 Sq. hd. jackscrew, 1 in.
77 Flinger
78 Warning tag (not shown
on BRG housing cap)
79 Direction arrow (not
shown on BRG housing
cap)
Figure 4.2-2 (continued)
4.2-4
-------
4.2.2 Design Criteria
4.2.2.1 Service Description--
In order to select recirculation pumps properly, a compre-
hensive service description must be developed. This necessi-
tates detailed analysis of the parameters described below:
Composition—The fluid to be pumped is a slurry containing
many solid and dissolved species. The major solids are lime,
fly ash, calcium sulfite (CaSO3•1/2H2O), and calcium sulfate
(CaSO4-2 H2O), all of which are erosive. Solids levels normally
range from 5 to 20 percent by weight. The dissolved species
include calcium, magnesium, sodium, sulfite, sulfate, chloride,
and carbonate ions, together with the ion pairs, such as hydro-
gen and hydroxide ions. Before specifying materials of con-
struction for a recirculation pump system, the designer must
know the chemical analysis of the specific slurry. This is
particularly important with closed-loop operation, since species
present only in trace amounts, such as chloride ions, can build
up to critical levels of 1000 ppm or more and dictate the use of
highly corrosion-resistant materials. In addition, the nature,
concentration, and size distribution of the solids should be
known. Information about all these important elements is neces-
sary to determine abrasion-corrosion resistance and the mechani-
cal strength required of the pump.
p_H—The slurry pH at the inlet to the absorber is usually
controlled to between 7 and 9, whereas the pH at the absorber
outlet ranges from 5 to 6. The recirculation pump is normally
located after the reaction tank. Therefore, it will be exposed
to a pH of about 7 or more. On systems without a reaction tank,
such as Bruce Mansfield, or where there is a pump both before
and after the reaction tank, such as Paddy's Run, the pH at the
recirculation pump inlet will be lower. Thus, the pump location
determines the pH to which the pump material will be exposed.
The pH values are an important factor in the selection of pump
materials.
Specific Gravity—The specific gravity of the recirculating
slurry is usually between 1.05 and 1.14. Figure 4.2-3 is a
graphic representation of specific gravity as a function of the
solids content of the slurry. In systems that incorporate
automatic solids control, the specific gravity of the slurry is
relatively constant. Many systems, however, do not control
solids content, and the specific gravity varies over a wide
range.
Viscosity—Knowing details of the rheology of the slurry
makes it possible to evaluate the reduction in pump performance
due to the viscosity of the mixture and the added slip between
the fluid and the solid particles as the mixture accelerates
through the pump impeller. This slip is greater in mixtures
with higher settling velocities.
4.2-5
-------
1.3
1.2
NJ
I
0*1
o
« 1.1
o
LU
Q.
CO
OH
1.0
0
10
I
I
15 20
SOLIDS, WEIGHT %
25
30
35
Figure 4.2-3. Specific gravity of scrubber recirculation slurry.
-------
Ash—The slurry may contain significant amounts of fly ash,
depending on the coal and the amount of flue gas pretreatment.
The amount and type of ash must be determined prior to equipment
specification, because certain types of ash increase the erosive
action of a slurry.
Gas Entrainment—It is possible that the recirculating
slurry could contain entrained flue gas as it exits from the
scrubber. Gas entrainment can result in the slurry in the pump
ranging from all liquid to essentially all gas. These variable
conditions, if allowed to exist, can cause shaft deflection,
which may result in bearing failure and abnormal packing wear.
Gas entrainment also reduces the liquid flow, which can reduce
SO2 absorption. Adequate phase separation in the scrubber will
prevent this problem.
4.2.2.2 Flow/Head—
Low-speed operation is one of the most important wear-re-
ducing features of a slurry pump. Pump abrasive wear increases
proportionally to the third power of rpm. Impeller tip speed of
rubber-lined pumps is limited to 3500 to 4500 ft/min. This
limits the rpm of the recirculation pumps with rubber linings to
400 to 600 rpm, which corresponds to a maximum discharge head of
about 100 ft.
Total liquid flow rate required for the absorber is deter-
mined by the design L/G ratio. The normal recirculation flow
rate range, corresponding to the rpm and head limitations, is
6000 to 10,000 gal/min for a rubber-lined pump. The number of
pumps required per scrubber, other than spares, is determined by
a technical and economic analysis of alternatives over the
specified range.
4.2.2.3 Net Positive Suction Head (NPSH)—
It is necessary to differentiate between available NPSH,
absolute suction head, and required NPSH. The available NPSH,
which is a characteristic of the system in which a centrifugal
pump works, represents the difference between the existing
absolute suction head and the vapor pressure of the slurry at
the operating temperature. The absolute suction head is the
algebraic sum of the suction pressure, static head, and the
frictional loss in suction line at a given capacity. With a
given static pressure at the suction side and a specific slurry
temperature, the available NPSH is reduced with increasing
capacities by the friction losses in the suction piping. The
design engineer must specify the available NPSH to the pump
manufacturers.
In a pump, the pressure at any point in the suction line
must never be reduced to the vapor pressure of the liquid be-
cause of the danger of cavitation. The required NPSH, which is
4.2-7
-------
a function of the pump design, represents the minimum required
margin between the suction head and vapor pressure at a given
capacity. The factors determining required NPSH include suction
area of the impeller, shape and number of impeller vanes, im-
peller velocity, and the impeller eye area (the annulus between
hub and vane walls). The required NPSH, which increases basic-
ally as the square of the capacity, must be obtained from the
pump manufacturer. Most slurry pumps require an NPSH of 15 to
30 ft.
The available NPSH must be determined accurately. More
pump troubles result from incorrect determination of available
NPSH than from any other single cause. As the available NPSH
for a given pump decreases, its capacity and efficiency de-
crease, and a low-suction pressure develops at the pump inlet.
The pressure decreases until a vacuum is created and the liquid
flashes to vapor (if the pressure is lower than the liquid vapor
pressure). This condition, which can lead to cavitation damage,
must be avoided by ensuring that the available NPSH is greater
than the required NPSH.
4.2.2.4 Pump Efficiency and Energy Requirements--4'5
The selection of efficiency should not be left entirely to
the pump manufacturer. He should be given data regarding energy
costs, the service conditions, and flow and head requirements.
An energy cost of 3<:/kWh and a penalty of $1000 per additional
horsepower can be specified as the basis for preliminary compar-
isons with the most efficient pump. Variations in size and
efficiency result from each manufacturer's effort to choose,
from his standard line of pumps, the one that most closely meets
the required conditions. Hence, the specifications should not
be so restrictive as to cause exclusion of high-efficiency
pumps. Finally, when the efficiency penalty is less than 10
percent, the pump with lower speed should be selected because
the increased pump life will compensate for the slightly higher
operating costs.
4.2.2.5 Impellers—
The impeller consists of a number of vanes open, semiopen,
or shrouded. The shrouded (closed-type) impeller has shrouds on
both sides to enclose the liquid passages. The closed- or
semiopen-type impeller is generally more efficient and is used
for service with abrasive slurry (See Figure 4.2.4). Closed
impellers experience less loss in efficiency than do open im-
pellers with the same widening of face clearance between the
impeller and the casing wall. An accelerated wear test of open
and closed impellers of otherwise identical geometry showed that
when the clearance of both impellers opened to 0.050 in., the
efficiency of the open impeller dropped by 28 percent, whereas
that of the closed impeller fell only by 14 percent.6 It has
4.2-8
-------
(a) OPEN IMPELLERS: IMPELLERS ON THE RIGHT ARE STRENGTHENED BY
PARTIAL SHROUDS.
(b) SEMIOPEN IMPELLER
(c) CLOSED IMPELLER
Figure 4.2-4. Types of impellers for recirculation pumps.
4.2-9
-------
already been pointed out that low-rpm operation is of the utmost
importance in reducing wear. For highly abrasive applications,
therefore, a range of acceptable pump speeds should first be
determined. Then, the pump speed should be altered over this
range and a maximum-size impeller selected to obtain the capa-
city required at the given head. In addition to reduced parts
wear, the advantages of a full-sized impeller are a slight gain
in efficiency and ready availability of replacement impellers,
which are made of Ni-Hard or chromium iron or are rubber-lined.
4.2.2.6 Drives—
Slurry pumps operate at relatively low speeds, from 400 to
600 rpm. Since the motors are either 1800 or 1200 rpm, some
type of speed reducer must be used. The most common way of
driving a lime slurry pump is by using a V-belt drive with a
fixed ratio, which has the advantages of flexibility and low
cost. For applications above 300 hp, however, gear reducers
should be considered. V-belt drives can be overhead-mounted or
side-mounted on horizontal pumps. Since it is difficult to
determine friction values of certain slurries for which data are
not readily available, it is advisable to use V-belt drives with
variable-pitch diameters. Without increasing the initial pur-
chase cost to a great extent, these drives simplify balancing of
the system at startup, allow the pump to meet future changes in
flow rate and head, reduce deterioration of pump performance due
to wear, and allow correction to initial system design for a
particular slurry.
4.2.2.7 Seals--
Horizontal-type centrifugal slurry pumps have a shaft
passing through the pump casing, which must be sealed to prevent
leakage. Mechanical seals, which are used for clean liquids,
are not suitable for slurries. Packed stuffing boxes have
customarily been used to seal the shafts since they cost less,
allow faster repair, and usually last longer in abrasive ser-
vice.
A continuous flow of clear water should be introduced into
a lantern ring at an intermediate position in the packing
(Figure 4.2-5). This flush water prevents abrasive solids from
entering the critical stuffing box and shaft sleeve area thereby
greatly extending the life of the packing and the sleeve.
Because abrasive solids may enter the packing during an FGD
system shutdown or upset, the pump should be designed with a
shaft sleeve of hardened alloy. Even under the best operations,
abrasive slurry may enter the packing.
The flush water entering the stuffing box will flow either
past the packing into the process or out the stuffing box. The
volume of flush water that mixes with the recirculating slurry
4.2-10
-------
PUMP SEAL
.WATER INLET
INTO THE
PROCESS
INTO THE I
PROCESS' />
Figure 4.2-5. Pump seal water flow.
4.2-11
-------
may be sufficient to affect the scrubber system water balance
(Section 2.2.3.3, Sample Calculations). As a housekeeping
measure, a large-diameter drain line should be provided to carry
the leakage from the stuffing box to the sump. For closed-loop
systems, the sump water should be pumped to the thickener.
The flush water supply system must be external to the
scrubber system and reliable enough to deliver a minimum quanti-
ty and pressure at all times. Process water can be considered
for use if the suspended particulate matter is less than 40 pm
in size and has a maximum of 1000 ppm by weight.7 The required
clarity may be achieved with the addition of a filter in the
water line. If thickener or pond overflow is to be used as
flush water, use of flocculant is necessary (Section 4.9.3.2,
Use of Flocculant). In most slurry pump designs, the pump
impeller will have back-vanes, which remove slurry from the
stuffing box region. This makes the pressure in the stuffing
box assembly essentially the same as the suction pressure to the
pump, so the sealing fluid need be supplied at a pressure only 5
to 10 psi above the suction pressure. On designs without such
back-vanes or with an excessively worn pump, the pressure in the
stuffing box region may rise to the discharge pressure. Then it
is necessary to supply the sealing fluid at 5 to 10 psi above
the discharge pressure. As a precautionary measure, an alarm
may be installed on the seal water feed line to indicate low
pressure or low flow of seal water.
4.2.2.8 Materials of Construction—
Since the pump parts in contact with the slurry are sub-
jected to abrasive-corrosive action, the "wetted" parts must be
constructed of a corrosion-resistant material that is either
harder than the slurry solids or resilient. The pump casings
and impellers are typically made of hardened iron or steel, such
as Ni-Hard, or of carbon steel lined with rubber. Since pump
manufacturers will not accept responsibility for selection of
material for these wetted parts, the design engineer must know
which materials are suitable.
Rubber—Molded rubber is the material specified most often
for wetted parts of lime slurry pumps. Both natural and syn-
thetic rubbers of about 1/4-in. thickness are used.
Although rubber is resilient in abrasive service and resis-
tant to corrosion, the use of rubber parts has some disadvan-
tages. One is the limitation of tip speed, as mentioned ear-
lier, and the resultant limitation of head to about 100 ft. in
addition, the entry of tramp metal (welding rods, bolts) into
the pump can destroy the rubber impeller and lining. At one
facility, strainers were installed in the pump section to pro-
tect the lining. When the strainers plugged, however, the pump
cavitated, stripping the lining from the casing. The strainers
4.2-12
-------
have been removed from the recirculation piping. Once the
impeller or lining rubber is damaged, it cannot be repaired and
must be replaced with new factory-supplied parts. Hence if
strainers are to be used, they should be accompanied by effi-
cient cleaning devices.
Hard Iron--Ni-Hard is a cast iron containing nickel (4%)
and chrome (1.4 to 3.5%). it is a very hard, brittle material
(550 to 650 Brinell) that can be finished only by grinding.
Ni-Hard has been used successfully in scrubber applications
where good pH control is achieved. It should not be subjected
to pH below 4. Ni-Hard is superior to rubber in that it allows
higher heads and is not as vulnerable to damage by tramp metal.
Alloys—Alloy-20, which contains nickel (35%); chromium
(20%); copper (3.5%); and molybdenum (2.5%), was applied unsuc-
cessfully on a lime slurry pump at one installation. Although
this alloy offers good corrosion resistance under many appli-
cations, the eroding action of the slurry removed the passi-
vating film and allowed corrosion to proceed at a high rate.
The lining and impeller failed in 3 months. For details on
other materials tested at this site, see Section 4.2.3, Perfor-
mance Histories—Phillips.8'9'10
Ceramics—Ceramic lining is being used more frequently in
vessels and piping. Although some manufacturers produce ceramic
pump liners, not one is currently in service on scrubber slurry
pumps. Ceramics provide excellent resistance to corrosion and
abrasion, but they are brittle and subject to shock failure.
4.2.2.9 Vendor Specifications—
As part of the Data Book project, several pump vendors were
asked to bid on a typical recirculation pump. Table 4.2-1 shows
results of this survey, in which each vendor was requested to
specify a pump for operation under the following lime scrubber
conditions:
Flow: 9000 gal/min max., 8200 gal/min normal
Head: 100 ft max.
Service: Lime slurry, CaSO3-1/2H2O, CaSO4-2H20, fly ash
Slurry specific gravity: 1.05 to 1.14
Temperature: 120° to 135°F
Solids: 5 to 10 percent
Size of solids: -100 mesh
Available NPSH: 30 ft
It is noted that pump efficiencies vary from 71 to 84 percent,
whereas pump speeds vary from 470 to 800 rpm. Assuming a mini-
mum required efficiency of 75 percent and a maximum allowable
speed of 500 rpm, preliminary comparison would lead to the first
two vendors. For further details on evaluation procedures, see
Bid Evaluation Section 5.5.
4.2-13
-------
Table 4.2-1. SURVEY OF VENDOR PUMP SPECIFICATIONS
(Flow 9000; Net Head, 100)
to
I
Characteristics
Pump characteristics
Pump
Pump dimensions, in.
Impeller eye area, in.
Efficiency, %
Pump speed, rpm
BHP @ design
BHP max.
Construction
Lining
Casing
Impeller
Shaft
Near rings
Drive
HP
Allis
Chalmers
SRL-C
16 x 14 x 34
189
85
510
305
350
Nat. rubber
Cast iron
Rubber
SS
No
Belt
350
Allan
Sherman Hoff
D-G- 9-5
16 x 16 x 39
201
76
470
300
342
Hypalon rubber
Cast iron
Rubber
316 SS
NR
Belt
400
Worthington
12 R 265
14 x 12 x 36
148.5
77
800
300
340
Nat. rubber
Cast iron
Steel
SS
NR
Belt
350
Ingersoll
Rand
400 CIR
16 x 12 x 27.94
154
72
720
HR
368
Nat. rubber
Carbon steel
Steel
SS
NR
Belt
400
Naqle
16 XR
15 x 33
NR
76
NR
NR
339
Nat. rubber
Cast iron
Cast iron.
Hastelloy
18
NR
Belt
350
Denver
Equipment Co.
NR
16 x 14 x 28
182.65
71
670
365
392
Nat. rubber
Cast iron
Cast iron
Alloy steel
SS
Belt
400
NR » Not Reported.
-------
4.2.2.10 Auxiliary Design Considerations-
Maintenance—Recirculation pumps are often located in a
limited space with difficult access, especially in retrofit
installations. Since this large equipment must be dismantled
periodically (typically at least every 18 months for inspec-
tion), the system design should facilitate maintenance. A
winch-and-trolley system for moving heavy parts and ample space
for dismantled components will simplify repairs. Some of the
common malfunctions of slurry pumps and their most probable
causes are listed in Table 4.2-2.
Housekeeping—Often the recirculation pump area is the most
unsightly part of a scrubber facility. A constant stream of
seal water, slurry, and oil leaks from the pumps, even in well-
maintained systems. Therefore, the pump area should be designed
for easy cleaning, with such features as sloping floors, wide
(24-in.) floor trenches, and a good supply of water.
Expansion—Expansion joints on pump inlet and outlet piping
are common sources of operating problems. When the joints fail,
they can leak slurry under pressure (discharge side) or bleed
air into the pump under vacuum (suction side). Proper specifi-
cation, installation, and maintenance of expansion joints are
important to pump performance.
Pump washouts—Since the circulating fluid is a slurry,
solids will settle out of the liquid whenever the flow is
stopped. If solids settle out in the pump, the pump impeller
and lining can be damaged on startup. For this reason, a flush
system that purges the pump with fresh water whenever the system
becomes inoperative for extended periods is recommended.
Spare pumps—In order to achieve reliable operations, it is
a practice to have spare equipment for the critical components
of a lime FGD system. The degree of redundance (the number of
spare pumps) for sluury recirculation pumps varies from 40 to
100 percent for the number of operating pumps. The number of
spare pumps per scrubber should be specified by the design
engineer. The stagnant slurry upstream from inlet and outlet
isolation valves may lead to plugging. This can be minimized by
keeping the spare pumps drained and by washing them frequently.
4.2.3 Performance Histories
Table 4.2-3 lists specifications of recirculation pumps
installed at FGD facilities. The following is a brief summary
of performance histories of lime slurry recirculation pumps
4.2-15
-------
Table 4.2-2. SLURRY RECIRCULATION PUMPsH/12
MALFUNCTIONS AND CAUSES
Malfunction
Causes
Pump develops less head and con-
sumes less power over its whole
working range, while efficiency
remains unaltered
Head falls off rapidly with an
increase flow rate while shutoff
head is unchanged
Flow rate is lower than rated by
a constant amount at any given
head
Head, capacity, efficiency, and
horsepower are all lower over the
entire range
Head and efficiency are reduced,
but horsepower is unchanged
Capacity drops off abruptly as
head is reduced
NPSH requirements are higher
at all flow rates
Deformed impeller casting,
rotational speed lower than
specified, undersized im-
peller
Reduced throat area of the
volute, reduced area be-
tween diffuser vanes
Worn wearing rings for closed
impellers, worn wearplate or
vanes for semiopen impeller
Excessive clearance in the
wearing rings, or between the
vanes and wearplates
Rough waterways in the impeller
or casing (because of rust,
scale, etc.)
Insufficient NPSH
Worn seal rings, rough
waterways
4.2-16
-------
Table 4.2-3. SCRUBBER SLURRY RECIRCULATION PUMP
SPECIFICATIONS - EXISTING FACILITIES5'7'10'11'12'13
10
M
^J
Location
Phillips
Duquesne Light
Green River
Kentucky
Utilities
Conesville
-Columbus t
Southern Ohio
Electric
Paddy's Run
Louisville. Gas
* Electric
Cane Pun
Louisville Gas
t Electric
Bruce Mansfield
Pennsylvania
Power Company
Elrama
Duquesne Light
Plant
rating,
MW
410
64
400
65
180
835
510
NO.
in
3
10
3
6
12
12
10
•
Pump
vendor
I-R
I-R
A-S-H
JV-C
Denver
A-S-H
I-R
Pump
model
12 x 22 IP
400c
DC5- 9- 5
NR
NR
DG-9-5
DG-9-5
12 X 22 LP
Im
pum
Flow,
gal/roin
9,000
5,900
9,544
6,000
5,800
11,000
9,500
9,000
Hvidual
T capacity
Head,
ft
100
115
<)0
140
100
95
101
100
Motor,
hp
350
300
400
450
300
son
350
Solids,
I
12
13
9
10
10
NR
12
Pump
speed ,
rpm
1185
695
450
'1000
1000
-500
1185
Pump
size, in.
14 x 12 x 18
NR
16 x 16 x 39
12 x 16
NR
16 x 16 x 39
14 x 12 x 16
Materials of
construction
Stellited 317L SS
Rubber- lined
using HiCrome (2811
impellers
Rubber-lined
Hi-Hard
Rubber- lined
Rubber- lined
Rubber- lined
NR - Not reported.
-------
Phillips8'9'10—At Phillips Station of Duquesne Power and
Light Co., Ingersoll-Rand pumps with Carpenter 20 casing and
impellers were originally used. The Carpenter 20 parts were
found unsuitable for scrubber slurry service. As a result of
erosion, the impellers and wear rings required replacement every
3 to 6 months.
Phillips has undertaken an extensive program to test a
number of alloys. It has tested such materials as Alloy 20,
317L SS, 26 percent CrFe, CD4MCu (a high-chrome, high-nickel
alloy), titanium, Carborundum, and TAPCO iron in various combi-
nations for construction of impeller and wear rings. In addi-
tion, several impellers were rebuilt with the wear areas hard-
surfaced with Stellite (Haynes No. 6). As a result of this
material testing program, Phillips has eliminated the less
promising materials (titanium, TAPCO iron, and 26% CrFe), and
has achieved definite improvements in service life with the
stellited 317L SS impellers and Carborundum wear rings. At
present, these materials have given over 4000 hours of service
without any wear. Tests are currently in progress on several
rebuilt CD4MCu impellers hard-faced with plasma spray coatings.
Green River8'13--Inqersoll-Rand pumps with rubber-lined
impellers and casing were originally installed at the Green
River Station of Kentucky Utilities.
The rubber has repeatedly peeled from the impellers and the
lining was destroyed after only 4 months of service. Ingersoll-
Rand is changing from a two-piece to a one-piece impeller
design. Green River is experimenting with high-chrome (28%)
metal impellers. Estimated life is 1 year.
Paddy's Run8'14—Allis-Chalmers1 pumps with Ni-Hard casing
and impellers were installed at Paddy's Run Station of Louis-
ville Gas & Electric Co.
Paddy's Run reports no failures of pump materials after
12,000 operating hours. There is slight evidence of erosion,
but no evidence of corrosion on the impeller or lining. This
operation maintains a pH of about 6 at the scrubber outlet.
Bruce Mansfield8'15—A-S-H pumps with rubber impellers and
lining were installed at the Bruce Mansfield Station of Pennsyl-
vania Power Co.
The plant reports no failure due to wear or corrosion.
Impellers and liners have been replaced because of damage from
miscellaneous material (welding rods) going through the pump.
Because the pumps are located under the scrubbers, with no
facilities for hoisting the parts, repair is difficult.
4.2-18
-------
Conesville8'16—A-S-H pumps with rubber impellers and
lining were installed at the Conesville No. 5 Station of Colum-
bus of Southern Ohio Electric Co.
Conesville reports no failures due to wear or corrosion
It has replaced the rubber lining because of damage from pieces
of pipe going through the pump. There are five pumps per
module, one of which was designed to be spare. In recent opera-
tions, only three pumps are used at full load.
Elrama8'9'1°--Ingersoll-Rand pumps with Alloy 20 casing and
impellers were originally installed at the Elrama Station of
Duquesne Light Co.
Elrama has had severe pump problems similar to those at
Phillips. At this station, the utility experimented with rub-
ber-lined pumps. The first set of rubber-lined pumps, supplied
by Ingersoll-Rand, failed after approximately 1000 hours. The
manufacturer has indicated that the lining failure on the im-
pellers was due to a faulty two-piece design and aged rubber.
The new rubber-lined pumps, now being tested, are supplied by
Worman.
Cane_Run8'14--At the Cane Run Station of Louisville Gas &
Electric Co., Joy Denver pumps with rubber-lined impellers and
casings are used. The main problem has been leakage from the
packing gland, which needs replacement every 3 months. The
rubber lining has not been replaced for about 2 years, though it
has shown some erosion.
4.2-19
-------
REFERENCES
1. Karassik, I.S., et al. Pump Handbook. McGraw-Hill Book
Co., New York, New York, 1976.
2. Neerken, R.F. Selecting the Right Pump. Chemical Engi-
neering, April 3, 1978, pp. 87-98.
3. Dalstad, J.I. Slurry Pump Selection and Application.
Chemical Engineering, April 25, 1977, pp. 101-106.
4. Dublin, J.H. Select Pumps to Cut Energy Cost. Chemical
Engineering, January 17, 1977, pp. 137-139.
5. Reynolds, J.A. Saving Energy and Costs in Pumping Systems.
Chemical Engineering, January 5, 1976, pp. 135-138.
6. Doolin, J.H. Pumping Abrasive Fluids. Plant Engineering,
November 1972.
7. Private communication with J.H. Wilhelm, EIMCo Process
Machinery Division of Envirotech, October 1977.
8. Laseke, B.A., Jr. EPA Utility FGD Survey: December 1977 -
January 1978. U.S. EPA-600/7-78-051, Industrial Environ-
mental Research Laboratory, March 1978.
9. O'Hara, R.D., and R.L. Nelson. Operating Experience at
Phillips and Elrama Flue Gas Desulfurization Facilities.
Second Pacific Chemical Engineering Congress, September,
1977 pp. 308-315.
10. Private communication with R.D. O'Hara and J. Mallone,
Duquesne Light Company, February 1978.
11. Yedidah, S. Diagnosing Troubles of Centrifugal Pumps -
Part II. Chemical Engineering, November 21, 1977, pp.
193-199.
12. Yedidah, S. Diagnosing Troubles of Centrifugal Pumps -
Part I. Chemical Engineering, October 24, 1977, pp. 125-
128.
13. Private communication with J. Beard and V. Anderson, Ken-
tucky Utilities, February 1978.
4.2-20
-------
14. Private communication with R. Van Ness, Louisville Gas and
Electric, February 1978.
15. Private communication with D. Boston, Columbus and Southern
Ohio Electric, February 1978.
16. Private communication with R. Forsythe, Pennsylvania Power
Company, February 1978.
4.2-21
-------
CONTENTS
4.3 OTHER PUMPS
4.3.1 Introduction 4.3-1
4.3.2 Reagent Feed Pumps 4.3-1
4.3.2.1 Service Description 4.3-1
4.3.2.2 Typical Characteristics 4.3-3
4.3.2.3 Special Design Considerations 4.3-3
4.3.2.4 Performance Histories 4.3-3
4.3.3 Thickener Supernatant Pumps 4.3-3
4.3.3.1 Service Description 4.3-3
4.3.3.2 Typical Characteristics 4*.3-4
4.3.3.3 Special Design Characteristics 4.3-4
4.3.3.4 Performance Histories 4.3-4
4.3.4 Thickener Underflow Pumps 4.3-6
4.3.4.1 Service Description 4.3-6
4.3.4.2 Typical Characteristics 4!3-6
4.3.4.3 Special Design Considerations 4.3-6
4.3.4.4 Performance Histories 4.3-6
4.3.5 Pond Water Return Pumps 4.3-8
4.3.5.1 Service Description 4.3-8
4.3.5.2 Typical Characteristics 4!3-8
4.3.5.3 Special Design Considerations 4.3-8
References 4.3-9
4.3-i
-------
4.3 OTHER PUMPS
4.3.1 Introduction
This section presents design information on pumps, other
than slurry recirculating pumps, that perform important func-
tions in the lime scrubber systems by pumping slurry feed,
thickener supernatant liquid, thickener underflow, pond water,
and fresh water to the system.
4.3.2 Reagent Feed Pumps
In lime FGD systems, the lime solids are mixed with water
in a slaker to form a slurry that mainly contains suspended
calcium hydroxide [Ca(OH)2]. The lime slurry is continuously
pumped to the scrubber system, as required, to replenish the
calcium ions that are discharged to the thickener as calcium
sulfite and calcium sulfate. The reagent slurry is usually
stored in tanks and transferred to the slurry recycle tanks by
the slurry feed pumps.
Lime slurry is less erosive than scrubber recirculation
slurry because of the absence of calcium sulfite, calcium sul-
fate, and fly ash. If grits are present, however, they will
increase erosion because of their hardness and nonuniform par-
ticle size. Though pumping of lime slurry is not as severe a
service, the basic design philosophy will be the same as for the
recirculation pumps. The salient features of lime slurry pumps
are discussed in this section.
4.3.2.1 Service Description--
The lime as received contains tramp materials, such as
rocks, metal, and wood. Proper design and operation of the
screening process and the slaker should remove these impurities
so that they do not hamper operation of the feed pumps (Section
4.5, Slurry Preparation).
The slakers usually use freshwater. Hence, the lime slurry
is not subject to buildup of corrosive ions, such as chlorides,
that occurs in closed-loop systems. In addition, because the pH
of the lime slurry is always highly alkaline, the slurry will
not cause acid corrosion. Typical lime slurry feed conditions
are presented below:
pH 12
Solids, wt. percent 15
Solids Ca(OH)2, Mg(OH)2
Temperature, °F 120
Specific gravity 1.1
4.3-1
-------
Table 4.3-1. LIME FGD SYSTEMS PUMP DATA - LIME SLURRY FEED PUMPS
,tk
•
OJ
Location/Utility
Conesville No. 5
Columbus &
Southern Ohio
Elrama
Duquesne
Light
Phillips
Duquesne Light
Green River
Kentucky Utilities
Cane Run
Louisville Gas
and Electric
Paddy's Run
Louisville Gas
and Electric
Bruce Mansfield
Nos. 1 and 2
Pennsylvania
Power
Manu-
facturer
Galigher
Goulds
Morris
Goulds
Morris
I-R
Joy Denver
Worth ington
Joy Denver
Model
3-VRA-200
1-1/2 JC-14
1-1/2 JC-14
40 CIR
NR
ER-3729-2-1/2RO!
SRL-2
Flow,
gal/min, each
145-320
90
120
90
200
1 100
300
Head,
ft
72
105
127
58
75
60
98
Pump
speed,
rpm
675
NR
NR
1550
1800
1800
875
Drive
Hydraulic
Direct
Direct
Belt
Belt
Direct
Belt
Materials
Casing
Rubber
lined
Cast
iron
Cast
iron
Rubber
lined
Cast
iron
Cast
iron
Rubber
lined
Impeller
Rubber
lined
Cast
iron
Cast
iron
Rubber
lined
Cast
iron
Cast
iron
Rubber
lined
Motor.
hp
20
15
15
5
10
5
125
No. of
pumps
3
6
5
2
2
2
4
NR - Not reported.
-------
4.3.2.2 Typical Characteristics—
The lime slurry pumps are usually of centrifugal type with
belt or direct drive. The impeller and casing may be plain cast
iron, neoprene rubber, or hard-iron lined with natural rubber.
The rubber lining is desirable if the grits removal system is
not very effective.
4.3.2.3 Special Design Considerations—
Since slurry feed pumps are located near the base of the
lime slurry tank, they are subjected to occasional dousing with
lime slurry; therefore, the motor and drive housing should be
enclosed in watertight casings to shield them from lime slurry.
Operation of the entire scrubber system depends on a con-
stant supply of lime slurry. Design of the lime slurry supply
system should, therefore, provide enough redundance to ensure
scrubber reliability.
4.3.2.4 Performance Histories—
Design data for slurry feed pumps at existing lime FGD
systems are presented in Table 4.3-1. Following is a summary of
reported operating experiences:x 5
No major problems are reported in pump operations at Cones-
ville, Green River, Phillips, and Elrama stations. The presence
of excessive grits in the lime, however, has recently been a
source of concern at the Conesville No. 5 unit. At Cane Run and
Paddy's Run, the packing glands were initially too tight and had
to be replaced. Recent operations have been trouble free.
Bruce Mansfield No. 1 unit required some modifications to
the system. The rubber liners in the lime slurry feed pumps
were damaged by cavitation. Baffle plates were installed over
the pump suction opening in the scrubber vessels. This solved
the problem, but several modifications of the baffles were
required before one was found that could withstand the stresses
and corrosive atmosphere.
4.3.3 Thickener Supernatant Pumps
Waste slurry is usually pumped to a thickener, where clari-
fied liquid overflows and is pumped back to the scrubber system.
4.3.3.1 Service Description—
Although most solids are removed in the clarifier, the
overflow may occasionally contain some suspended solids. The pH
of the liquid is dependent on good pH control in the scrubber.
Although the range of pH is normally 7 to 8, under upset condi-
tions in the scrubber the pH may go down to 2. since corrosive
ions such as chlorides may build up in the liquid in a closed-
loop operation, the main design consideration for wetted parts
of the pumps is corrosion.
4.3-3
-------
Following are typical properties of thickener overflow
liquid:
pH 7 to 8 (normal)
Solids, wt. percent 1
Temperature, °F 130
Specific gravity 1.01
4.3.3.2 Typical Characteristics--
The thickener supernatant pumps are usually the centrifugal
type with direct drive. The impeller and casing are commonly
316L SS, though at Cane Run and Paddy's Run pumps with rubber
impellers and rubber-lined casings are used.
4.3.3.3 Special Design Considerations—
As with the lime feed pumps, operation of the scrubber
system is dependent on continuous functioning of the thickener
supernatant pumps. Systems should be designed with sufficient
redundance to insure scrubber reliability.
Since these pumps are usually located outdoors in a remote
area, instrumentation is needed for monitoring pump operation
from the scrubber control room. In addition, protection is
recommended at those locations where pumps may freeze when the
scrubber is shut down.
4.3.3.4 Performance Histories--
Design data for thickener supernatant pumps at existing
lime FGD systems are given in Table 4.3-2. Following is a
summary of reported operating experiences:
At Conesville No. 5 the pumps have been operating satisfac-
torily with no reported problems.
At Elrama and Phillips the pumps were originally made of
cast iron, which suffered severe corrosion because of the pH
fluctuations caused by inadequate pH control. The present
material (CD4M Cu) has given satisfactory service. The pumps at
Cane Run and Paddy's Run were originally underdesigned. No
problems have been reported after the capacity of the pumps
increased.
Bruce Mansfield Nos. 1 and 2 pumps initially had vibration
problems. The fiberglass reinforced plastic (FRP) piping was
replaced by steel at the discharge end to reduce the noise
level. In addition, an insufficient net positive suction head
(NPSH) caused air entrainment and cavitation. The reclaimed
water tank was raised by about 4 ft to alleviate this problem.
4.3-4
-------
Table 4.3-2. LIME SCRUBBER PUMP DATA THICKENER SUPERNATANT PUMPS
ui
I
Location/Utility
Cones vi lie Ho. 5
Columbus I
Southern Ohio
Elraraa
Duquesne
Light
Phillips
Duquesne
Light
Cane Run
Louisville Gas
and Electric
Paddy'* Run
Louisville Gas
and Electric
Bruce Mansfield
NOB. 1 and 2
Pennsylvania
Power
Manu-
facturer
Goulds
Morris
Goulds
Morris
Goulds
Morris
Goulds
Morris
Allis
Chalmer
Goulds
Morris
Model
NR
3175
3175
3196
912
3175
Flow,
gal/rain, each
1080
16SO
1650
600
250-300
6760
Head,
ft
240
180
ISO
100
120
140
Pump
speed , rpm
NR
NR
NR
1800
1600
1180
Drive
Direct
Direct
Direct
Direct
Direct
Direct
Materials
Casing
316
CD4-M Cu
CD4-M Cu
Rubber
lined
Rubber
lined
316L
Impeller
316
CD4-H Cu
CD4-M Cu
Rubber
lined
Rubber
lined
316L
Motor,
hp
125
125
125
25
30
300
No. of
pumps
2
3
3
2
3
4
HR - Not reported.
-------
4.3.4 Thickener Underflow Pumps
The underflow from the thickener is pumped either to a
dewatering system or to a sludge pond, which is often located
several thousand feet from the thickener. Since the service is
severe, proper specifications for these pumps are critical.
4.3.4.1 Service Description—
The thickener underflow is a thick slurry containing all
the solids species present in the scrubber (calcium sulfate,
calcium sulfite, fly ash, lime, etc.)- If a centrifugal pump is
to be specified, the concentration of solids in the thickener
underflow will be limited to about 40 percent maximum because
higher concentrations could not be pumped with a centrifugal
pump without causing nonuniform flow. For thickener underflow
containing more than 40 percent solids, positive displacement
pumps are commonly used. Scrubber pH control is important so
that these pumps will not be subjected, even intermittently, to
high concentrations of corrosive ions. Because the slurry does
contain high concentrations of abrasive solids, the main design
consideration is erosion.
Following are typical thickener underflow characteristics:
pH 8 .to 10
Solids, wt. percent 40
Solids Fly ash, calcium sulfate,
calcium sulfite
Temperature, °F 130
4.3.4.2 Typical Characteristics—
The thickener underflow pumps are either centrifugal or
positive displacement type, with belt drive. For the centri-
fugal pumps, the impeller and casing lining material is rubber.
If positive displacement pumps are used, it is recommended that
the rotors be rubber lined and high alloys be used for stators.
The positive displacement pumps should be designed to have a
nonpulsating uniform flow.
4.3.4.3 Special Design Considerations—
The factors that influence the thickener overflow pumps
also apply to the underflow pumps:
0 Redundancy to insure scrubber reliability
0 Controls for remote operation
° Protection against damage by freezing (if required)
4.3.4.4 Performance Histories—
The design data on thickener underflow pumps at existing
lime FGD systems are given in Table 4.3-3._ The following is a
summary of reported operating experiences.1"5
4.3-6
-------
Table 4.3-3. LIME SCRUBBER PUMP DATA - THICKENER UNDERFLOW PUMPS
Ul
I
Location/Utility
Conesville No. 5
Columbus &
Southern Ohio
Elrama
Duquesne
Light
Phillips
Duquesne
Light
Cane Run
Louisville Gas
and Electric
Paddy's Run
Louisville Gas
_ and Electric
Bruce Mansfield
Pennsylvania
Power
Manu-
facturer
Galigher
Allen-
Shermanhoff
Allen-
Shermanhof f
Robbins
Myers
Allen-
Shermanhof f
Joy Denver
Model
NR
A- 6- 5
A- 6- 5
2XNG12f
GDI
AA-6-5
SRL-C
Flow,
gal/min,each
460
200
200
200
150
1500
Head,
ft
100
65
65
115
120
70
Pump
speed, rpm
NR
NR
NR
1800
1800
700
Drive
Hydraulic
Belt
Belt
Variable
Belt
Belt
Materials
Casing
Rubber
lined
31 6L
316L
Neoprene
rubber
Rubber
lined
Rubber
lined
Impeller
Rubber
lined
Rubber
lined
lubber
lined
Hi-A
alloy,
stator
Rubber
lined
Rubber
lined
Motor .
hp
40
25
25
20
5
75
• " •_•-•• - — • i »
No. of
pumps
2
3
3
2
2
4
NR - Not reported.
-------
At Conesville No. 5 the centrifugal, rubber-lined pumps
have been operating very well. At Elrama, the centrifugal pumps
have enough capacity to handle the additional load from two or
more thickeners. The Moyno (positive displacement) pumps at
Cane Run have given excellent service. The thickener underflow
line was replaced with one of larger diameter at Paddy's Run to
reduce frictional losses. At Bruce Mansfield Nos. 1 and 2, the
rubber lining of the centrifugal pumps is replaced about once a
year.
4.3.5 Pond Water Return Pumps
Most of the lime FGD systems pump the pond water back to
the system, which may be several thousand feet from the pond.
4.3.5.1 Service Description--
This water is similar to overflow water from the thickener.
The only major problem is caused by buildup of corrosive ions
such as chlorides in a closed-loop operation.
Typical characteristics of pond return water are presented
below:
pH 6 to 8
Solids, wt. percent 0
Temperature, °F 70
4.3.5.2 Typical Characteristics—
The pond water pumps are centrifugal pumps with alloy steel
impellers. The head developed by these pumps is generally about
200 ft H20.
4.3.5.3 Special Design Considerations—
These pumps also are usually located outdoors, several
thousand feet from the scrubber control room. They therefore
require equipment for remote control and for protection of
material in freezing weather.
4.3-8
-------
REFERENCES
Personal communication with D. Boston, Columbus and
Southern Ohio Electric, February 1978.
Personal communication with J. Beard and V. Anderson,
Kentucky Utilities, February 1978.
Personal communication with R. O'Hara and J. Mahone,
Duquesne Light Company, February 1978.
Personal communication with R. Vanness, Louisville Gas and
Electric, February 1978.
Personal communication with R. Forsythe and W. Norrocks,
Pennsylvania Power Company, February 1978.
4.3-9
-------
CONTENTS
4.4 LIME UNLOADING AND STORAGE CONTENTS
Page
4.4.1 Characteristics of Lime 4.4-1
4.4.2 Lime Transportation 4.4-6
4.4.2.1 Truck Shipments 4.4-6
4.4.2.2 Rail Shipments 4.4-8
4.4.2.3 Barge Shipments 4.4-10
4.4.3 Unloading Design Criteria 4.4-12
4.4.3.1 System Design 4.4-12
4.4.3.2 Dry Lime Conveying 4.4-16
4.4.3.3 Lime Storage and Feed Bins 4.4-19
4.4.4 Existing Facilities 4.4-22
4.4.4.1 Receiving 4.4-22
4.4.4.2 Storage 4.4-22
4.4.4.3 Conveying 4.4-22
4.4.4.4 Performance History 4.4-25
Appendix 1 4.4-26
References 4.4-28
4.4-i
-------
4.4 LIME UNLOADING AND STORAGE
This section describes the design of lime unloading and
storage facilities used in lime slurry scrubbing systems. It
begins with bulk receipt and ends at the inlet to the lime feed
mechanism, which itself is discussed in Section 4.5, Lime Slurry
Preparation.
4.4.1 Characteristics of Lime
Lime is manufactured by heating crushed limestone to high
temperatures (1652° to 2192°F) . The process is known as cal-
cining.
heat
CaC03 -* CaO + CO2 t (Eq. 4.4-1)
limestone quicklime + carbon dioxide
form
Lime manufacturers also react quicklime with water to
hydrated lime, primarily calcium hydroxide [Ca(OH)2].
CaO + H20 -> Ca(OH)2 * (Eq. 4.4-2)
quicklime water slaked lime
(calcium hydroxide)
This is a bulk material that is easier to store and handle, but
since it is more costly than quicklime, it is rarely used by
large installations. About 4 Ib of hydrated lime are needed to
neutralize the same amount of acid neutralized by 3 Ib of quick-
lime.
The National Lime Association, a lime manufacturing indus-
try cooperative, recognizes six standard size classifications of
quicklime (Table 4.4-1 ). These classifications are not binding
to individual producers, however, who may adopt other defini-
tions or sell other sizes. A "nonstandard" size, in fact,
consisting of pebble lime screened or crushed to eliminate
particles larger than about 1 in., is used by many utility lime
scrubber installations.
The chemical composition of quicklime is determined pri-
marily by the composition of the limestone rock from which it
was made.1 The most variable chemical constituent is the ele-
ment magnesium. Geologic processes have caused carbonate rocks
to be formed most often either with little magnesium (calcite),
or with approximately equal molar amounts of calcium and magne-
sium carbonates (dolomite). Most commercial limes are therefore
either "high-calcium," containing less than 5 percent MgO, or
"dolomitic" lime, containing 35 to 45 percent MgO by weight.
Lime made from a mixture of the two rocks, or from deposits
4.4-1
-------
Table 4.4-1. STANDARD SIZES OF QUICKLIME
Size
Description
1. Lump lime
2. Crushed or pebble
lime
3. Granular lime
4. Ground lime
5. Pulverized lime
6. Pelletized lime
Maximum size 8-in. diameter
2 to 2-1/4 in. diameter
Product from fluidized-bed kilns;
size range = 100% passing No. 8
sieve to 100% retained on No. 80
sieve
Obtained by grinding or screening
larger sizes; size range = ~100%
passing No. 8 sieve to 40-60%
passing No. 100 sieve
Obtained by more intense grinding
than for ground lime: size
range = -100%
passing No. 20 sieve to 85-95%
passing No. 100 sieve
Obtained by compressing quicklime
fines into ~l-inch pellets or
briquettes
4.4-2
-------
with an uncommon magnesium content, has recently become popular,
and the term "magnesian lime" has been accepted to define a
product containing from 5 to 35 percent MgO by weight. The
trade name "Thiosorbic" has also been coined to describe a
magnesian lime containing between 5 to 10 percent MgO. Calcitic
lime contains 0 to 5 percent MgO.
In addition to containing calcium and magnesium oxides,
quicklimes consist of 1 to 10 percent of material that will not
react with water. Generally called "grit," this material has
two distinct portions: (1) sand and fused particles of iron and
aluminum oxides or silicates, and (2) calcium and magnesium
carbonates that were not converted into lime by calcining.
Table 4.4-2 lists the physical properties of good quality
quicklime. The angle of repose (see Figure 4.4-1), it should be
noted, varies with particle size distribution; a high proportion
of fine particles increases the angle. Particle size is impor-
tant in that pulverized quicklimes do not flow from a hopper as
readily as coarser grades.
Quicklime does not corrode ordinary construction materials
such as carbon steel, concrete, and most plastics. Nor is it
especially abrasive, but it will cause a moderate amount of
mechanical wear in bins and conveying equipment. Although
quicklime is incombustible, high temperatures can develop if it
accidentally contacts water or chemicals containing water of
hydration.
Quicklime is hazardous and can burn the skin; it is parti-
cularly damaging to the eyes and, if dust is inhaled, to the
throat and lungs. In areas where quicklime dust may be preva-
lent, workers should wear a lightweight filter mask and tight-
fitting safety glasses. Additional protection is required to
prevent contact with the skin, particularly in hot weather when
workers are perspiring. Besides eye protection and respirators,
workers exposed to quicklime dust should wear proper clothing:
a long-sleeved shirt with sleeves and collar buttoned; trousers
with legs down over shoes or boots, head protection, and gloves.
It is also advisable to apply a protective cream to exposed
parts of the body, particularly the neck, face, and wrists.
First aid treatment is given in Appendix 1, at the end of Sec-
tion 4.4.
Although quicklime is hazardous, hydrated lime presents no
danger. Its dust is irritating if inhaled, but causes no last-
ing damage. Hydrated lime will not burn the skin, does not
react with water, and will not reach high temperatures unless
mixed with strong acids. Dry hydrated lime is a light, fluffy
powder, not abrasive or corrosive to ordinary materials. Its
most troublesome characteristic is its angle of repose. If the
material is aerated and dry, it may flow almost like liquid,
4.4-3
-------
Table 4.4-2. PHYSICAL PROPERTIES OF QUICKLIME'
Specific gravity
Bulk density (pebble lime)
Specific heat @ 25°C
Angle of repose (pebble lime)
3.2 to 3.4
55 to 60 lb/ff
0.20 Btu/lb
50° to 55°
0 - THE ANGLE OF REPOSE
\\V//A\\Y//A\\Y/7A\W///U \ Y7/X \ W//XV\ V/M\\ W/A\\\
GROUND
Figure 4.4-1. The angle of repose.
4.4-4
-------
flooding through feeders and spilling over the edges of equip-
ment. If the hydrated lime is compacted and slightly damp, the
angle of repose may be as much as 80°, inhibiting steady flow
from a hopper.
Before lime is fed to a lime scrubbing FGD system, it is
slaked with water to form a slurry (Equation 4.4-2).
Any magnesium present in the quicklime will exist as a
relatively stable double oxide, and two reactions can occur when
water is added:
CaO'MgO + H2O -> Ca(OH)2-MgO (Eq. 4.4-3)
CaOMgO + 2H20 -> Ca(OH)2 -Mg(OH)2 (Eq. 4.4-4)
All these reactions liberate large quantities of heat. A
significant fraction of the energy added to quicklime by the
calcining process is released in the reaction of quicklime and
water.
All the reaction products have a very low solubility in
water. Reaction products containing magnesium are even less
soluble than limestone or gypsum.
Only under proper conditions, involving the right tempera-
ture, water-to-lime ratio, and degree of agitation, will quick-
lime react completely with water; under other conditions, a
particle of quicklime will become surrounded by a layer of
reaction products, thereby excluding water from contact with
material in the center of the particle. Unreacted quicklime
will therefore appear in the lime slurry and can cause erosive
damage to slurry handling equipment.
The mechanism by which hydration reactions can be carried
to completion has been studied extensively, primarily to allow
users to optimize the "slaking" reaction, the reaction of quick-
lime in an excess of water. According to one theory, at optimum
slaking conditions heat from the reaction converts water into
steam at the surface of a quicklime pebble. The steam's expan-
sion, plus agitation of the mixture, causes reaction products to
be carried away from the surface of the pebble as they form,
thereby exposing fresh surfaces for further reaction. This is
further discussed in Section 4.5.
Electron micrographs have shown that hydrate structures in
properly prepared lime slurry using either quicklime or hydrated
lime are mostly in the form of very small needles, less than 1
(jm in length. Such a slurry exposes a very large surface area
to subsequent reactions, and is efficient in neutralizing acids.
As particles increase in size, their reactivity drops rapidly in
proportion to their surface area, even in concentrated acid.
4.4-5
-------
4.4.2 Lime Transportation
Lime is transported from the manufacturers to the users in
various ways. When very small quantities are involved, the most
convenient method is to ship it in bags. In the quantities
required for lime scrubbers, it is bought in bulk and trans-
ported in trucks, trains, or barges designed for handling bulk
solids.
4.4.2.1 Truck Shipments--
The truck trailer most often used for transporting lime is
a pressure-differential tank trailer or "blower truck" (Figure
4.4-2), which is manufactured by several companies. They are
available with capacities of 12 to 25 tons. The trailer is
virtually self-unloading, requiring only that the receiving bin
be equipped to accept the pressurizing air and to control dust.
The trailer is built as a tank with hopper bottoms. The hoppers
connect through valves to a short manifold. For unloading, the
manifold is connected with a hose to the customer's 4-in. pipe
leading to a storage bin. The trailer is then pressurized to
about 15 psi by a motor-driven compressor mounted on the truck.
Additional air from the compressor is sent through the 4-in.
pipe to the storage bin, and the valve is opened between one
hopper and the manifold. Lime is blown from the tank into the
manifold and is pneumatically conveyed through the pipeline into
the bin. Tank pressure is adjusted to maintain an even flow of
lime. When one hopper has emptied, the other hoppers in turn
are opened into the manifold. When all hoppers are empty, the
compressed air in the trailer blows into the storage bin.
Pressure-differential trailers are usually equipped with aera-
tion pads in the hoppers so they can also be used for hydrated
lime. Pads are not usually needed to unload quicklime. The
largest size of pebble lime that can be handled is 1-1/4 in.,
but a 1-in. top size is preferred to prevent plugging during
unloading. Pebble lime can be blown as much as 100 ft verti-
cally and 150 ft on a combined vertical and horizontal run.
Greater distances are possible, but would involve excessive
unloading times.
The self-loading and self-unloading hopper trailer is a
more sophisticated version of the pressure-differential trailer.
Although most often used to ship more expensive chemicals,
trailers of this type are occasionally used for lime. They
contain rotary feeders, pneumatic conveyors, and dust collection
equipment. They require no supplemental equipment to transport
bulk solids, nor do they need a bin dust collection mechanism.
Air-activated, gravity-discharge hopper trailers are also
used to transport lime, particularly finer products such as
pulverized quicklime and hydrated lime that are not normally
used in scrubber facilities. These trailers have a single
4.4-6
-------
ENGINE-DRIVEN
COMPRESSOR
CONVEYING AIR
LOADING HATCHES
HOSE CONNECTION
BUTTERFLY VALVE
Figure 4.4-2. Pressure-differential tank trailer.
4.4-7
-------
outlet connection (Figure 4.4-3), and blow-through airlock
feeders regulate the unloading rate. A low-power fan mounted on
the trailer supplies air to aeration pads in the bottom of the
tank to fluidize the lime and cause it to flow by gravity to the
outlet. This type of trailer cannot elevate lime into a storage
bin; the customer must supply a mechanical or pneumatic conveyor
for this purpose.
4.4.2.2 Rail Shipments—
Most rail shipments of lime are made in standard covered
hopper cars (Figure 4.4-4). Originally designed to handle
cement, these cars are used for any solid material that must be
protected from rain. They can hold about 100 tons of quicklime.
The cars are built with two to four separate rectangular com-
partments, each with at least one loading hatch, and a sloping
hopper bottom closed with a slide gate. Other than brackets for
attaching vibrators, there are no special features to simplify
unloading; all unloading equipment must be supplied by the
customer. The costs involved and the car's ready availability
account for its wide use.
Unloading quicklime from hopper cars is difficult. Often
it is dumped into an undertrack hopper, from which it is trans-
ported by conveyor into storage bins. Alternatively, vacuum
unloading attachments are available that clamp onto the dis-
charge gate and suck lime from the hopper into a storage or
transfer tank. With either system, unloading is a dusty and
relatively hazardous job. Avalanches inside a partly unloaded
hopper can cause lime to spill from open hatches. Lime tends to
hang in corners and must be knocked down with poles from the top
of the car. Fabric chutes are often used during gravity un-
loading to minimize dusting, but they are not very effective.
Placement of the gates makes it difficult to attach suction
equipment; if the car has worn springs, it may have to be jacked
up before the attachments will fit. The unloading crew may also
have to use a car puller to position the car so that all hoppers
can be unloaded.
Railroad cars that are much easier to unload are in wide
use, but are available only by lease arrangements with their
owners. Unless the car can be kept busy, the cost may be too
high to warrant its use. Two cars that are the rail equivalents
of the truck trailers described earlier are also in use for lime
hauling. One of these is a pressure-differential tank car,
which consists of three or four cylindrical pressure vessels
with hopper bottoms permanently mounted on a flat car. Each
tank holds about 15 tons of quicklime. An aeration pad and an
outlet pipe are installed in the bottom of each hopper. The
unloading operation is identical to that described for the
pressure-differential truck trailer, except that air pressures
are higher and air must be supplied from a stationary source at
4.4-8
-------
LOADING HATCHES
AERATION PADS
ENGINE-DRIVEN BLOWER
ROTARY FEEDER
Figure 4.4-3. Hopper trailer with air-activated,
gravity-discharge hopper.
Figure 4.4-4. Covered hopper car.
4.4-9
-------
the unloading station. With this car, lime can be blown over
longer distances and into higher storage bins than can be
achieved with the blower truck.
The air-activated, gravity-discharge hopper car hauls up to
45 tons of lime. This car is essentially a single large bin
built into a boxcar framework with aeration pads built into the
hopper-shaped bottom. In unloading, air from an external source
causes the lime to flow from two outlets located halfway along
the car length and on opposite sides of the car. The outlets
clear the rails by at least a foot, and dustproof unloading
chutes can be easily attached. Aeration of the lime during
unloading prevents avalanches and hangups; one man can easily
handle the unloading operation.
4.4.2.3 Barge Shipments—
Barge transport of lime is common
truck shipments, in fact, are unloaded
points for distribution to users who
waterways. Operators of a large lime
cated, may realize substantial savings
lime in barge quantities.
in this country. Many
from barges at central
are not near navigable
scrubber, properly lo-
by directly purchasing
The craft used for lime transport is a hopper or covered
barge, built with a separate hold inside the framing of the
hull. The deck is waterproof and is fitted with waterproof
hatch covers. For use on inland waterways, hopper barges have
shallow drafts and are approximately square ended, with a long
bow rake and a shorter stern rake. Three standard sizes are in
use, built to fit efficiently into standard river locks (Table
4.4-3). Similar vessels are built for marine service.
Table 4.4-3. COVERED OR HOPPER BARGES
Barge
type
Pittsburgh
Jumbo
Large
Length, ft
175
195
240
Width, ft
26
35
40-50
Approximate
pebble lime
capacity, tons
800
1200
2000
Barge transport requires substantial investment in dock
facilities that cannot usually be justified for lime handling
alone. A typical barge-unloading installation with vacuum gear
is shown in Figure 4.4-5. Most scrubber operators using barged
lime add the equipment to existing coal docks; separate equip-
ment is usually needed to prevent intermixing of products and to
keep the lime dry.
4.4-10
-------
AIR EXHAUST
FLEXIBLE
..HOSE
S
^TELESCOPIC PIPE EXTENDS
' FOR LOW WATER LEVELS
CONVEYOR TO
STORAGE
VACUUM
EXHAUSTER
SIX-HOSE
MANIFOLD
Figure 4.4-5. Vacuum barge unloading.
4.4-11
-------
4.4.3 Unloading Design Criteria
4.4.3.1 System Design--
It is essential that any plant using bulk lime should
include facilities to receive shipments by blower truck. Even
if rail or barge is to be the principal mode of lime transport,
truck shipments will probably be needed to supplement the sup-
ply, or to compensate for late deliveries. At least one lime
storage bin, with a capacity of at least 1500 ft3, should be
fitted with inlet piping, a dust collector, and a pressure
relief device.
The pneumatic pipe through which lime is transferred to the
bin should be ordinary 4-in. carbon steel pipe, preferably with
no more than one 90° change of direction. To reduce flow resis-
tance and minimize wear, the pipe should be bent on a 3- to 4-ft
radius. Lime should first be blown vertically and then hori-
zontally (Figure 4.4-6). Greatest wear and most blockages will
occur in pipe bends ahead of vertical runs. Total length of the
piping should not exceed 150 ft, and total change of elevation
should not exceed 100 ft. The bottom end of the vertical pipe
section should be 4 ft from the ground and end with a 150-lb
flange.
The dust collector on the bin may be a small cyclone col-
lector, but better operation is obtained with a bag filter.
Since a blower truck delivers about 750 ft3 of air per minute
and an air-to-cloth ratio of 2.0 is recommended for this ser-
vice, a cloth area of 375 ft2 is required.
A pressure-relief device is needed to prevent excessive
pressure in the bin. A manhole with hinges and gaskets is
frequently used. Pressure relief is needed especially at the
end of an unloading operation, when all the compressed air in
the trailer is exhausted into the bin. The final blast of air
serves to clear lime from the transfer line, but usually over-
loads the capacity of the dust collector. If dust from the
final blast is troublesome, the cloth area of the filter must be
increased somewhat to about 650 ft2.
Blower truck facilities are the most elementary facilities
suitable for lime unloading. In initial scrubber design, it is
wise to allow for future use of rail or barge delivery. The
ability to accept larger loads greatly increases the number of
possible suppliers and thus may effect considerable cost
savings.
Rail deliveries in standard covered hopper cars can be
unloaded either by an undertrack hopper or a vacuum-unloading
installation, depending on climate and topography. if the
region is hilly and if a spur track can be built into the side
of a hill, an undertrack hopper can often be designed with
4.4-12
-------
PRESSURE RELIEVING
MANHOLE COVER-
4-IN. CS PIPE BENT
4-FT RADIUS
AIR
EXHAUST
DUST
COLLECTOR
•-I-1- — "!-!-'
Figure 4.4-6. Blower truck receiving bin.
4.4-13
-------
sufficient drainage and air circulation to keep the lime and
equipment dry. In a flat area with a high water table and
frequent humid weather, the vacuum unloading system, although
often more expensive to construct, is usually the better choice.
Gravity unloading of rail cars requires a spacious,
weatherproof building. At least 12 ft of clearance is required
above the car, and building heat is desirable to dehumidify the
air during wet weather. The building must be well ventilated to
capture lime dust, and the air should be kept dry during un-
loading by a recirculating system that ventilates through a dust
filter.
The undertrack hopper is usually fitted with a screw con-
veyor or a drag conveyor to deliver lime to a bucket elevator
for transfer to a storage bin. For hillside unloading, the
bucket elevator is sometimes omitted; lime is transferred di-
rectly to storage with an inclined screw or drag conveyor.
Conveyors must be tightly enclosed and weatherproofed, and often
must also be contained in a building or tunnel. Belt conveyors
are usually unsatisfactory, since lime is lost into the conveyor
housing. Pneumatic conveyors are rarely used to empty under-
track hoppers since they are easily choked unless separate
feeders are provided.
If hopper cars are to be unloaded under vacuum, equipment
is usually bought as a complete system from a single supplier.
Several U.S. manufacturers, including Sprout-Waldron, Buell,
Fuller, and Ducon, offer this equipment. With any of the sys-
tems shown in Figure 4.4-7, adapter devices are attached to the
car outlet gates, and lime is sucked through metal hoses and
steel piping into a receiver. The latter consists of a vacuum
tank, a dust collector, and often a cyclone to separate larger
particles from the air stream. Vacuum is supplied by a large
suction pump or exhauster; the unloading rate is regulated by
bleeding in a stream of air at the hopper car adapter. The
receiver is mounted above a storage bin, and lime is continuous-
ly emptied from the vacuum tank through an air-lock mechanism.
Other conveyors may distribute the lime to several bins. in
vacuum unloading, particles move at high velocity (3000 to 7500
ft/min) and there is some degradation of pebble size. Air
volume ranges from 10 to 20 fts/lb of lime. There is usually
less than 5 psi of vacuum at the exhauster suction. A building
or shed is needed to allow unloading in wet weather, but vacuum
unloading creates less dust than gravity unloading and ventila-
tion requirements are therefore less severe.
Gravity-unloading hopper cars can be accommodated with
either an undertrack hopper or a vacuum unloading system if the
required adapters are purchased.
4.4-14
-------
CYCLONE AND
DUST COLLECTOR
HOPPER-BOTTOM CAR"
UNLOADING ATTACHMENT
VACUUM
EXHAUSTER
DRAG OR
SCREW CONVEYOR
Figure 4.4-7. Vacuum railcar unloading.
4.4-15
-------
4.4.3.2 Dry Lime Conveying—
Until now we have dealt solely with unloading, which is
simply a specialized application of conveying. Many lime scrub-
bing FGD systems, however, include conveyors other than unload-
ing devices to transport lime into feed bins or from transfer
bins into storage.
The most difficult in-plant conveying of dry lime occurs
when the scrubber is a considerable distance from the unloading
point. In such cases, a first thought is to prepare lime slurry
near the unloading station and pump the finished slurry to the
scrubber. This is usually impractical, since slurry that re-
mains in a long pipeline will settle out and eventually plug the
line. Use of water to rinse the line free of slurry causes
formation of scale, which also eventually plugs the line. As a
rule, slurry should be prepared within 200 ft of the point of
use; if the lime unloading point is more distant, dry lime
should be conveyed to the slurry preparation area. A pneumatic
conveyor can transport lime at least 1000 ft, and several con-
veyors in series will cover longer distances.
Most in-plant conveying, however, entails simple elevation
of the lime from a storage bin into a smaller feed bin. Figure
4.4-8 shows an application in which a screw conveyor accepts
lime from one of several storage bins and discharges it to a
bucket elevator that carries it into a feed bin. A simple
combination of mechanical devices can move lime from storage at.
less initial cost and with less power consumption than that
entailed with a pneumatic conveyor. Mechanical conveying re-
quires careful arrangment of bins and equipment, which would
preferably be aligned in a single straight row. Each change of
direction usually requires another conveyor. Since pan or drag
conveyors do not plug as readily as screw conveyors, they are
preferred for long runs and for handling coarse grades of quick-
lime. Belt conveyors generally lose too much material into the
housing to be considered suitable for quicklime service.
As conveying distances or elevations increase, or if con-
veyance involves several changes of direction or multiple points
of delivery, the economic advantage of pneumatic conveying
increases rapidly. Unlike those of a mechanical conveyor, the
basic components of a pneumatic system are similar regardless of
distance or elevation. They differ only in the length of piping
and the size of the compressor and motor. Figure 4.4-9 shows a
simple dry lime transfer by the pneumatic conveying principle.
This illustration also shows one of the two basic types of
pneumatic conveyors, usually called the positive-pressure type.
Lime is blown up the inclined pipe by the force of air from the
compressor. The other basic conveyor type is the vacuum, or
negative-pressure, system already referred to. It sucks lime
through the pipe by means of a vacuum exhauster attached to the
4.4-16
-------
STORAGE BINS-
I
H
^J
ROTARY
FEEDER
BUCKET
ELEVATOR
SCREW CONVEYOR
0
SLURRY STABILIZATION
TANK
Figure 4.4-8. Mechanical conveyors.
-------
I
M
00
FILTER
SILENCER
STORAGE BINS
\-CONVEYOR
ADAPTER
DUST COLLECTOR AND
COLLECTING BIN
SLURRY STABILIZATION
TANK
AIR COMPRESSOR
Figure 4.4-9. Positive-pressure pneumatic conveyor.
-------
dust collector. The positive-pressure system has the advantages
of atmospheric pressure in the receiving bin and a slight reduc-
tion in air flow rate due to higher air density. The negative-
pressure system causes no condensation of moisture inside the
pipeline and offers an equipment package with all system com-
ponents close together. Both systems have the disadvantage of
using air on a once-through basis, bringing large quantities of
humid outdoor air into contact with the lime.
Another arrangement, usually called a closed-loop system,
is shown in Figure 4.4-10. A single charge of air is recir-
culated from the compressor to the conveyor and back. A minimum
amount of fresh air is drawn in, and the original charge remains
dry. The closed loop can be constructed as either a positive-
pressure or a vacuum system. This depends on the location of
the loop vent, where atmospheric pressure is maintained. In
Figure 4.4-10 (an example of a positive-pressure system) the
loop vent is located at the feed bin to which lime is delivered.
The closed-loop system offers the advantages of both the pres-
sure and the vacuum conveyor, plus elimination of moisture. Its
disadvantages include higher piping cost and greater power
consumption.
The pneumatic conveyor with the lowest initial cost is a
vacuum system in which the vacuum is produced by steam eductors.
However, these should not be used to handle materials such as
quicklime that are hazardous when contacted with water.
4.4.3.3 Lime Storage and Feed Bins—
In designing lime storage bins, the main consideration
should be capacity. Failure to provide ample storage is usually
the result of the assumption that lime deliveries will be con-
stant and unvarying. In practice, truck shipments are fre-
quently delayed by bad weather or breakdowns, and rail or barge
shipments are even less predictable. Minimum bulk storage
capacity for a constantly operating industrial facility is
generally considered to be either 150 percent of a plant's
normal shipment size, or capacity for 7 days' usage, whichever
is larger. Better practice in a lime scrubber is to provide
twice this volume, since lime is often transported on a less-
dependable schedule than other, more-expensive, bulk chemicals.
The storage unit most often used for lime is a steel silo
with a cone bottom. Cylindrical vessels hold the most material
for a given weight of steel, but square, rectangular, and hexa-
gonal bins are also used, and can be clustered to share common
walls and thus save ground space. Concrete storage bins have
been used in large installations and are often less expensive
than steel bins. The schedule for commissioning a scrubber may
preclude use of concrete, however, since concrete must "cure"
for several months after construction before being used to store
quicklime.
4.4-19
-------
VENT
DUST
FILTER
DUST COLLECTOR
AND
— COLLECTING BIN
I
ro
o
ROTARY
FEEDER
CONVEYOR
ADAPTER
SLURRY STABILIZATION
TANK
Figure 4.4-10. Closed-loop pneumatic system.
-------
Lime storage bins must be weatherproof and airtight to
prevent absorption of water and carbon dioxide from the atmos-
phere. Attempts to use corrugated metal silos to store lime
have been unsuccessful, since these silos are not sufficiently
weatherproof for this service.
Storage bins must be fitted with a cone-shaped or hopper-
shaped bottom to allow an even first-in/first-out flow of lime.
Steel is most often used for the hopper section, even in con-
crete bins, to reduce sliding friction. In a steel cone free
from fabrication edges, quicklime usually slides on a 45-deg
slope. Where the hopper is made of a rough material such as
concrete, a steeper angle is necessary; a pitch of 3:4 is usu-
ally satisfactory. Storage bins for hydrated lime require
60-deg cone bottoms.
The number of storage bins and their relative size and
proportion are determined by construction economy. Economy
usually increases with bin height and diameter, to the point
where the bin becomes so large its initial transportation cost
would be excessive. A diameter of 12 ft is often the most
economical, with a maximum height of 40 ft. A bin of these
dimensions will hold about 100 tons of quicklime.
The interior of a bin should not be painted. Protection
against corrosion is unnecessary, notwithstanding the fact that
abrasion from quicklime will remove any paint that is applied.
Abrasion will polish the metal of a bin, especially near the
outlet spout, but will not grind away sharp corners or weld
spatter. Before a bin is placed in service, the area near the
outlet should be ground smooth. It is undesirable to grind or
polish the upper vertical walls of a bin. Roughness in the
vertical surfaces may help to support the bin contents and
therefore minimize packing in the hopper section.
With pebble quicklime, special bin unloading attachments
are not usually required. As a precaution, attachments for
portable vibrators are sometimes installed on the hopper sec-
tion. Storage of pulverized quicklime usually justifies a
permanently installed vibrator; in a large bin, a static type of
antipacking device, such as that which discharges a volume of
high-pressure air to dislodge a mass of lime, may be desirable.
For hydrated lime, antipacking and flow assistance attachments
are virtually a necessity. There is a wide variety of devices;
no single type has proved superior for use with hydrated lime.
A lime storage bin may be connected directly to a lime
feeder that meters a flow of lime into a slaker. Frequently,
however, lime is transferred from a storage bin into a smaller
feed bin at a higher elevation. Feed bins are usually more
carefully designed than storage bins, since smooth flow to the
lime feeder is important in achieving trouble-free performance
of the lime slurry preparation equipment.
4.4-21
-------
Lime feed bins are often designed to hold enough lime to
permit either 10 or 26 hours of scrubber operation at maximum
rate, so that they can be routinely filled once a shift or once
a day. The hopper usually has a 60-deg slope; in addition,
offset hoppers are often used. The problem with concentric
hoppers on cylindrical bins is that "arches" or "domes" can form
as material is withdrawn (Figure 4.4-11). If an arch forms, the
operator must break it to restore flow. Formation of arches can
be prevented by constructing the hopper as an unsymmetrical cone
in which one edge is vertical (Figure 4.4-11). The offset
construction is more expensive than a concentric design and
wastes space. Use of offset hoppers on large storage bins
usually cannot be justified; however, the extra expense may be
warranted for one or two relatively small feed bins.
A rotary or slide-type shutoff gate is needed at the bottom
of a feed bin, or any bin that connects to a lime feeder, to
permit maintenance of the feeder without emptying the bin. As a
service to the customer, dry feeder manufacturers will usually
resell a suitable gate or one may be purchased separately.
4.4.4 Existing Facilities2"6
Data from utility lime scrubber systems are summarized in
Table 4.4-4 and detailed below.
4.4.4.1 Receiving—
Four plants receive Thiosorbic quicklime in 1-3/4 in. top
size. One plant, Bruce Mansfield, receives the material by
barge;6 the other three receive it by truck.
The two plants in Louisville, Kentucky, Cane Run and
Paddy's Run, purchase hydrated lime from a carbide plant. Since
the material arrives in a 30-percent solids slurry, these plants
do not have dry lime handling equipment. Green River is the
only plant currently receiving lime in rail cars. Note that the
top size of Thiosorbic lime is larger (approximately 1-3/4 in.)
than that of the other types (approximately 3/8 to 3/4 in.).
4.4.4.2 Storage—
All facilities store lime in carbon steel silos except
Conesville, where concrete is used. Two plants, Phillips and
Elrama, are constructing three carbon steel silos. Storage
capacity on full-scale plants ranges from 10 to 30 days. Green
River, which has a 10-day supply, has shut down occasionally
because of lime shortages.
4.4.4.3 Conveying—
Belt and pneumatic systems are both used in existing facil-
ities. The belts are used on larger sizes of lime with rela-
tively long runs; pneumatic conveyors are used for the shorter
runs.
4.4-22
-------
(a) SCHEMATIC OF ARCH
FORMATION IN A
CONCENTRIC HOPPER
(b) OFFSET HOPPER
CONSTRUCTION
Figure 4.4-11. Feed bin design.
4.4-23
-------
Table 4.4-4. EXISTING FACILITY DESIGN SPECIFICATIONS
i
ro
Plant
Receiving
Type of lime
Top size. In.
Normal method
Size or receipt, tons
Storage
Capacity of silos,
tons (each)
Number of silos
Storage capacity at
design flow days
Material of construction
Conveying
Type
Distance, ft
Elevation change, ft
Conesv L 1 lo
No. 5
Qulckline,
Th iosorbic
1-3/4
Truck
25
11,500
2
*0C
Concrete
Belt
300
ISO
Four Corners
No. 5A*
Quicklime,
high calcium
3/8
Truck
25
300
1
30
C.S.
Pneumatic,
blower
NR
100
Bruce Mansfield
Not. 1 and 2
Quicklime,
Thiosorhlc
1-1/2
Barge
1000-1500
4500
4
12*
C.S.
Belt
1500
-100
Mohave
IA'
Quicklime,
high calcium
3/8
Truck
25
30
1
14
C.S.
Pneumatic ,
Mover
80
-50
Green River
No*. 1, 3. and 3
Quicklime,
high calcium
3/4
Rallcar
70
500
1
10
C.S.
Pneumatic,
vacuum
75
73
Cane Run
No. 4
Hydrated
line
NA
Barge
NA
NA
NA
NA
NA
NA
NA
NA
Paddys
Run
No. 6
Hydrated
lime
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Phillips
Quicklime.
Thiosorbic
1-3/4
Truck
25
NRb
3
NR
C.S.
Pneumatic,
blower
70
NR
Elrama
Quicklime,
Thiosorbic
1-3/4
Truck
25
b
NR
3
NR
C.S.
Pneumatic,
blower
70
NR
* rco «r*tM no lonver operational.
Currently building *ilo*.
c 30 day* when Unit 6 start* up.
d IS day* when Unit 3 atart* up.
N*. - Wot reported.
HA - Hot applicable.
-------
4.4.4.4 Performance History--
No problems have been reported in lime handling systems
4.4-25
-------
APPENDIX 1
First Aid Treatments for Calcium Oxide (Lime) Splashes7
Splashes of the skin
1. Flood the splashed surface thoroughly with large quantities
of running water and continue for at least 10 minutes, or
until satisfied that no chemical remains in contact with
the skin. Removal of splashes with solvents, solutions,
and chemicals known to be insoluble in water will be facil-1
itated by the use of soap.
2. Remove all contaminated clothing, taking care not to con-
taminate yourself in the process.
3. If the situation warrants it, arrange for transport to
hospital or refer for medical advice to the nearest doctor.
Provide information to accompany the casualty on the chemi-
cal reponsible and brief details of the first aid treatment
given.
Ingestion
1. If the chemical has been confined to the mouth give large
quantities of water as a mouth wash. Ensure the mouth wash
is not swallowed.
2. If the chemical has been swallowed give copious drinks of
water or milk to dilute it in the stomach.
3. Do not induce vomiting.
4. Arrange for transport to hospital. Provide information to
accompany the casualty on the chemical swallowed with brief
details of the treatment given and if possible an estimate
of the quantity and concentration of the chemical consumed.
Splashes of the eye
1. Flood the eye thoroughly with large quantities of gently
running water either from a tap or from one of the eyewash-
bottles provided and continue for at least 10 minutes.
4.4-26
-------
2. Ensure the water bathes the eyeball by gently prying open
the eyelids and keeping them apart until the treatment is
completed.
3. All eye injuries from chemicals require medical advice.
Arrange transport to hospital and supply information to
accompany the casualty on the chemical responsible and
brief details of the treatment already given.
4.4-27
-------
REFERENCES
1. Lime Handling, Application, and Storage. Bulletin 213, National
Lime Association, Washington, D.C. Third Edition. May 1976
2. Personal communication with D. Boston, Columbus and Southern OK •
Electric, February 1978. nic
3. Personal communication with J. Beard and V. Anderson, Kentucky
Utilities, February 1978. y
4. Personal communication with R. O'Hara and J. Mahone, Duquesne
Light Company, February 1978.
5. Personal communication with R. VanNess, Louisville Gas and
Electric, February 1978.
6. Personal communication with R. Forsythe and W. Norrocks, Penn-
sylvania Power Company, February 1978.
7. Hazards in the Chemical Laboratory. G.D. Muir, ed. The Chemical
Society, London, England. Second Edition. 1976.
4.4-28
-------
CONTENTS
4.5 LIME SLURRY PREPARATION
4.5.1 Introduction
4.5.2 Dry Lime Feeding
4.5.2.1 Introduction
4.5.2.2 Volumetric and Gravimetric Feeders
4.5.2.3 Available Equipment
4.5.2.4 Existing Facilities
4.5.3 Lime Slurrying (Slaking)
4.5.3.1 Introduction
4.5.3.2 Service Description
4.5.3.3 Available Equipment
4.5.3.4 Water Requirements for Slaking
4.5.3.5 Existing Facilities
4.5.4 Slurry Stabilization and Storage
4.5.4.1
4.5.4.2
4.5.4.3
References
Bibliography
Glossary
Introduction
Service Description
Existing Facilities
4.5-3
4.5-3
4.5-6
4.5-9
4.5-10
4.5-10
4.5-10
4.5-12
4.5-17
4.5-18
4.5-18
4.5-18
4.5-18
4.5-23
4.5-31
4.5-32
4.5-33
4.5-i
-------
4.5 LIME SLURRY PREPARATION
4.5.1 Introduction
This section of the Lime FGD Systems Data Book presents
design information on the process from the dry lime feeder
through the slurry preparation system to the lime slurry feed
pump. The receipt, storage, and conveying of dry lime through
the inlet of the dry lime feeder mechanism are discussed in
Section 4.4, the slurry feed pump in Section 4.3.
Preparation of lime slurry involves three steps:
1. Feeding dry lime to the system.
2. Reacting lime with water (slaking).
3. Stabilizing and storing the lime slurry.
These functions must operate as an integrated system to
supply lime to the FGD system. The slurry is added to the
scrubbing system at a rate normally controlled by the pH of the
absorber liquor, and the system should also be designed to
maintain a readily available supply of slurry in storage tanks.
Consequently, when the slurry in the storage tank falls to a
specified level the necessary equipment must be activated to
produce slurry at a rate at least equivalent to that of maximum
usage and preferably at an even higher rate to assure a suffi-
cient inventory of lime slurry that will allow time for any
maintenance of lime feeding equipment. The slurry feed rate
control system can be designed for manual operation, which would
require regular attention by an operator, or could be controlled
semiautomatically, relieving the operator of much of the re-
sponsibility.
A suggested arrangement of equipment and instrumentation in
a semiautomatic lime slurry preparation system is shown in
Figure 4.5-1. It includes the following recommended design
features:
0 Quick lime storage above the feeder.
A slaker discharging directly into the stabilization
tank.
0 A single baffled tank to accomplish stabilization and
storage.
Each of these recommended features will be discussed in depth in
subsequent subsections.
4.5-1
-------
CO-
1
-•c — i
T SLURRY
- V-
OH-
-H{J
t
BACKPRESSURE
VALVE
STABILIZATION
AND STORAGE
Figure 4.5-1. Lime slurry system—schematic flow.
Source: Beals, J.L., Handling hydrated lime slurries.
Wallace & Tiernan, April 1976.
4.5-2
-------
4.5.2 Dry Lime Feeding
4.5.2.1 Introduction—
The first step in slurry preparation is to deliver dry lime
from a bin at a constant rate. Most lime scrubbing FGD systems
use dry feeders for this purpose.
Dry feeders are not manufactured specifically for the
feeding of lime; they are used in other industries for a wide
variety of granular and powdered solids. Several suppliers
manufacture dry feeders of varying specifications and cost.
Generally, models suitable for preparation of lime slurry have
an accuracy of between 1 and 5 percent by weight.
Not all commercially available equipment is suitable for
use in lime slurry preparation. The types used for bulk con-
veying of solids such as coal or crushed stone are unsuitable
because they do not provide the consistent flow needed to ensure
uniform slurry quality. Feeders suitable for precision-blending
applications are unnecessarily accurate and inordinately expen-
sive for use with lime.
4.5.2.2 Volumetric and Gravimetric Feeders—
There are two methods of measuring the quantity of lime
discharged from a dry feeder that provide sufficient accuracy at
a reasonable cost—the volumetric method and the gravimetric
method.
Volumetric feeders deliver a constant volume (ft3/h) of
material, regardless of its bulk density. When a volumetric
feeder is used, the feeding mechanism is manually adjusted to a
fixed position. If lime flows uniformly to the feeding mechan-
ism and if the lime is of consistent quality (e.g., size distri-
bution), a consistent weight of lime will be discharged within a
measured accuracy of approximately 5 percent. Examples of both
volumetric belt-type feeders, whose belt speed and the position
of the feed regulating gate determine the volume of material
fed, and vibrating volumetric feeders, in which electromagnetic
vibration delivers a steady lime feed, are shown in Figures
4.5-2 and 4.5-3, respectively. Other types of volumetric feeder
are also available. Volumetric feeders require continual super-
vision because any blockages that develop in the feeder throat
or in the lower section of the feed bin will not clear automati-
cally- Volumetric feeders usually require a smaller capital
investment than gravimetric feeders.
Gravimetric feeders deliver a constant weight (Ib/h) of
material, thus compensating for variation in the lime quality as
well as voids in the stream of solids. They are accurate to
approximately 3 percent by weight. This improvement over the
4.5-3
-------
FEED SECTION
STATIONARY^7
DECK
FEED
BELT
Figure 4.5-2. Volumetric belt-type feeder.
Source: Wallace & Tiernan Division, Pennwalt Corp.
4.5-4
-------
OSCILLATING
HOPPER
SCRAPER
Figure 4.5-3. Oscillating hopper volumetric feeder.
Source: Wallace & Tiernan Division, Pennwalt Corp.
4.5-5
-------
specification of volumetric feeders is not significant and
should not be a criterion for selection of a gravimetric type
feeder over a volumetric type. However, gravimetric feeders are
more reliable since their feeding mechanism is adjusted auto-
matically by instrumentation operating in conjunction with a
monitor that measures the weight of lime being fed. Thus, when
a blockage begins to form in a gravimetric feeder the internal
control system adjusts the feeding mechanism in such a manner as
to increase the flow. This action in itself minimizes the
occurrence of major blockages. If one should develop, the
internal control of a gravimetric feeder would detect the low
feed rate and sound an alarm. Thus the need for continual
operator attention is reduced.
A gravimetric feeder manufactured by Wallace & Tiernan is
shown in Figure 4.5-4. Contained xn this particular machine is
a weighdeck, beneath the conveyor belt, which measures the
weight of lime carried on a section of the belt. Instrumenta-
tion adjusts a vertical gate so that only a set weight of
material is allowed out of the feed bin and into the slaker.
This same equipment, with a manually adjusted vertical gate and
without a weighdeck, can function as a volumetric feeder.
The choice between a gravimetric and a volumetric feeder
is, of course, a complex one, but one of the more important
considerations is the design of associated equipment. If the
lime feed bin, the slaker, and other components of the lime
slurry preparation system are designed for operation without
continual supervision, the benefits of a gravimetric feeder
would probably outweigh the increased capital investment.
However, when an operator is employed full time on a slurry
system, there would seem to be no significant advantages from
the use of a gravimetric feeder.
Although salesmen may encourage engineers to specify the
feeder and slaker as a single item, the price often is signifi-
cantly lower if they are bought separately. There are no tech-
nical reasons why the separate components should come from the
same manufacturer as a single package.
4.5.2.3 Available Equipment—
Although many different feeding mechanisms are available in
both gravimetric and volumetric feeders, only four have proved
successful for handling lime:
1. The belt conveyor with a mechanical gate to regulate
loading.
2. An oscillating hopper, which has a reciprocating
mechanism that pushes ribbon-like layers of material
from a fixed tray.
4.5-6
-------
TRACTION
ROLL
FLEXURES
WEIGHBELT
Figure 4.5-4. Mechanical gravimetric feeder.
Source: Wallace & Tiernan Division, Pennwalt Corp.
4.5-7
-------
3. A screw conveyor, which incorporates a rotating helix
to move the material.
4. A vibrating device, in which the vibrations are in-
duced either electromagnetically or mechanically.
An example of the belt-type feed mechanism is shown in Figure
4.5-4. This type of equipment is available from most of the
larger manufacturers of intermediate-accuracy feeders, including
Acrison, B-I-F, Merrick, Wallace & Tiernan, and Weightometer.
The belt-type feeder is used most often in large lime slurry
preparation installations. Most companies make several versions
of this type of feeder, but the vertical-gate model shown in
Figure 4.5-4 is usually recommended for quicklime service.
Other versions use rotary gates, or fixed gates, with automati-
cally-adjusted belt speeds and may offer some advantages in lime
slurry preparation systems. In the conveyor feeder unit illus-
trated, the feed rate is dictated by the speed of the belt,
which is determined by the drive motor and motor gearing.
Variable-speed drive motors are available, but are not usually
needed. Some manufacturers make the same feeder with different
belt widths in order to expand the use of a single basic design
and create a wide range of feeding capacities. Feeders of this
type can handle feed rates ranging from 1 Ib/h to 200,000 Ib/h.
For a given belt size and speed, however, turndown ratio from
the gate alone is only about 10 to 1. Changing the feed range
usually requires only a change of gears and occasionally minor
modifications to the weighdeck.
Oscillating hopper (Figure 4.5-2) and screw conveyor
feeders have an inherent disadvantage when used in a slurry
preparation system. The usual configuration would place the
feeder directly above a lime slaker; therefore, water vapor
rising from the slaker could react with lime at the mouth of the
screw or on the tray of the oscillating hopper. During normal
continuous operation, the small amount of reaction is not espe^
cially troublesome, but when the feeder is stopped, the lime
exposed to moist air reacts to form hard deposits, which could
interfere with the operation of moving parts. During inter-
mittent operation, these feeders require frequent cleaning and
maintenance. Therefore, screw conveyor and oscillating hopper
feeders are built primarily as small- or medium-capacity units,
best suited for use in plants where the flow of lime is not
intermittently stopped and restarted. B-I-F, Acrison, and
Wallace & Tiernan manufacture screw-conveyor-type feeders.
Oscillating hopper units are constructed primarily by foreign
manufacturers.
Figure 4.5-5 shows a gravimetric feeder made by B-I-F, with
a vibratory feed mechanism and a weigh belt. The intensity of
vibration is regulated by a "control wedge" of hard rubber
suspended from a scale beam. If the amount of lime on the weigh
4.5-8
-------
belt is less than specified, the control wedge drops, trans-
mitting more vibrations to the feed plate. Several companies
make vibratory feeders in both volumetric and gravimetric ver-
sions. Syntron Division of the FMC Corporation is a major
manufacturer of volumetric vibratory feeders. Low initial cost
is the principal advantage of this mechanism. Its main disad-
vantage is noise that, in large units, may reach a level that
requires special sound proofing.
SCALE BEAM , HOPPER
CONTROL WEDGE -
TRANSMISSION ENDLESS BELT
Figure 4.5-5. Gravimetric feeder.
Source: B-I-F, a unit of General Signal Corp.
4.5.2.4 Existing Facilities—
Of the 10 plants operating lime FGD systems during early
1978, six feed dry quicklime into slurry preparation systems and
two buy hydrated lime as a byproduct from local chemical plants.
These two plants, Paddy's Run and Cane Run, do not have slurry
preparation systems of the type described here. No data are
available on two other plants, Hawthorne No. 3 and Hawthorne No.
4.
The Phillips, Elrama, and both Bruce Mansfield plants use
belt-type gravimetric feeders. In all these units, each feeder
is so arranged that it can feed only one slaker. Phillips and
Elrama both have five feeders, one of which is a spare. At
Bruce Mansfield, four Merrick feeders are in use, each of which
is capable of feeding quicklime at 50,000 Ib/h and is actually
operated to feed about 48,000 Ib/h. There are no spare feeders.
Conesville uses five screw conveyors, each delivering about
8000 Ib/h into individual slakers. Green River employs one
conveyor screw to deliver about 8000 Ib/h into either of two
slaking tanks.1 Neither the Conesville nor Green River system
provides for variation in the rate of lime feed to the slaker.
The systems are designed for a specific and constant feed rate.
4.5-9
-------
All of these systems are operated intermittently. At
Phillips and Elrama, switching on and off is done manually. At
Green River and Conesville, semiautomatic controls are used, and
similar controls are being installed at Bruce Mansfield.
4.5.3 Lime Slurrying (Slaking)
4.5.3.1 Introduction--
Lime slurry may be obtained by three processes: slaking
quicklime, purchasing dry, previously slaked lime, or purchasing
slurry from a producer. Louisville Gas and Electric has elected
the third option for both the Paddy's Run and Cane Run FGD
systems. Most of the operating units have elected the first
option, and this will be discussed in detail.
When quicklime (CaO) is added to water, a vigorous reaction
occurs with the evolution of heat. The result is a lime slurry
suspension of calcium hydroxide. The process itself is called
slaking.
CaO + H2O -> Ca(OH)2 + heat
The objective of lime slaking is to produce a smooth, creamy
mixture of water and very small particles of calcium hydroxide.
This mixture should contain no unreacted quicklime and no large
calcium hydroxide crystals. A lime slaker mixes regulated
streams of quicklime and water under the appropriate agitation
and temperature conditions to disperse soft hydrated particles
as they form. Dispersion must be rapid enough to prevent local-
ized overheating and rapid crystal growth of the hydroxide;
however, the mixture must be held in the slaker long enough to
permit complete reaction.
4.5.3.2 Service Description--
Quicklime of any type or size distribution can be processed
in a properly designed slaker. The speed at which various
grades and sizes of quicklime can be slaked varies considerably
and directly influences the rate at which a slaker can produce
lime slurry. A highly porous lime slakes more rapidly than a
"hard-burned," nonporous material. Finely divided particles
slake more rapidly than large lumps, and high-calcium limes
slake more rapidly than magnesian limes. The American Society
for Testing and Materials (ASTM) has developed testing methods
to determine the relative slaking rates of quicklimes. It is
not possible to predict the rate at which a slaker can produce a
good slurry unless the lime has been tested by standard methods
such as ASTM C25-72 and ASTM C110-716. The rated capacities of
commercial slakers are based on lime that slakes rapidly.
Actual slurry production rates are often lower, especially when
slaking pebble lime with high magnesium content.
4.5-10
-------
Because no technology has been developed that will auto-
matically produce slurry of optimum characteristics, operation
of a slaker is often described as an "art." The operator must
manipulate certain variables, depending upon the characteristics
of the slaker. At present, two types of slaker are available:
detention slakers and paste slakers. With a "detention" slaker
the temperature of the mixture is monitored in order to estab-
lish a range of about 50°F over which the slaker produces slurry
of adequate quality. The temperature may be controlled by
varying the amount of water added.
With a "paste" slaker, the torque of the mixer shaft is
monitored as a guide to optimum conditions, since measuring the
temperature of a thick paste is difficult. Detention and paste
slakers, the types used most commonly with FGD systems, are
described under Available Equipment, Section 4.5.3.3.
The range of slaking conditions is limited in any type of
slaker by the occurrence of "drowning" or "burning," conditions
that produce low-quality slurry. "Drowning" is caused by the
presence of too much water in proportion to lime and is indi-
cated by a sharp drop in slaking temperature when the flow of
water is increased slightly. The result is that the calcium
hydroxide being produced adheres to the quicklime particles.
Each quicklime particle is surrounded by a layer of relatively
impervious and unreactive calcium hydroxide. At the other
extreme, "burning" also produces particles of unreacted quick-
lime surrounded by a hard impervious layer of hydrate. In
burning, insufficient water causes localized overheating at the
surface of the quicklime particles. The very high temperatures
that develop cause formation of unreactive oxide and hydrate
crystals. Burning usually produces steam, which removes water
from the mixture and causes a further increase in temperature.
Burning is therefore indicated by a rapid rise in slaking tem-
perature when flow of slaking water is decreased slightly. If
not controlled, it will cause serious overheating that may
damage the equipment.
A lime slaker must include provisions for removing grit
from the slurry. Grit consists of sand and similar impurities,
plus the carbonate cores of quicklime pebbles that were not
calcined during lime manufacture. Good-quality quicklime con-
tains 1 to 2 percent grit; poor grades may contain up to 5
percent. The grit is usually discarded in a sludge pond or
landfill- Grit particles that remain in the slurry cause abra-
sive damage to slurry handling equipment. Properly slaked
slurry of pure hydrated lime can be considered nonabrasive,
since the lime particles are soft, lightweight solids that
cannot scratch or erode metals.
4.5-11
-------
4.5.3.3 Available Equipment—
Equipment used for lime slaking in a scrubbing system
should be specifically designed for this service.
A detention slaker is shown in Figure 4.5-6. Quicklime and
water are fed to the slaker in specific proportions in order to
produce a slurry containing 20 to 30 percent solids. The mix-
ture is agitated with a high-speed propeller mixer. From the
agitated chamber, slurry flows into a quiet section where grit
settles out. Degritted slurry is then diluted with additional
water and flows to a stabilization tank. Grit is continuously
removed from the quiet section by means of a mechanical scraper.
It is rinsed with a small stream of water and discarded. In a
detention slaker, water is added to each chamber; the amount of
water is manually proportioned by trial and error to achieve the
best results with the lime being used. Slurry usually is re-
tained in a detention slaker 20 to 30 minutes at a temperature
of about 167°F.
The paste slaker operates on the pug mill principle,
kneading a thick mixture of lime and water. Feeds are propor-
tioned to produce a putty-like mixture containing about 40 to 50
percent solids. The mixture is blended in a narrow trough by
paddles that rotate on horizontal shafts. The thick slurry
continuously overflows the end of the trough into a dilution
chamber, where more water is added and grit is separated,
rinsed, and discarded. A typical paste slaker is shown in
Figure 4.5-7. Slurry is retained in a paste slaker for only 5
to 10 min. The slaking temperature is usually about 185° to
194°F.
Neither the detention slaker nor the paste slaker has
proved superior for all applications. Both types have their
proponents and both offer certain advantages. Characteristics
of these slakers are listed in Table 4.5-1. In general, the
detention slaker is a more flexible unit that can be tailored to
match exactly the requirements of a specific installation. it
is simple in design, but bulky. The paste slaker is more com-
plex mechanically. It is much smaller and only available in
certain sizes. The freshwater requirements of the two slakers
are markedly different, and this is discussed in Section
4.5.4.4.
Both high temperature and long retention time are required
to slake some poor-quality grades of quicklime. It is uneconom-
ical to slake poor-quality quicklime with paste slakers if it is
necessary to operate several slakers at reduced capacity in
order to provide adequate retention time. In a detention
slaker, poor-quality quicklime can be processed by heating the
feed water, thereby providing both high temperature and longer
retention time. Preheated water can also be used with a paste
slaker, but the technique is less effective since the lower
water-to-lime ratio limits the amount of preheat that can be
used without causing "burning."
4.5-12
-------
CLASSIFICATION
COMPARTMENT
WASHED GRIT
DISCHARGE
RECIPROCATING
RAKES
MILK OF LIME
OUTLET
AGITATOR DRIVE
FEED INLET (LIQUID)
LIME INLET
SLAKING
AGITATOR COMPARTMENT
Figure 4.5-6. Detention slaker.
Source: Dorr-Oliver Corp.
4.5-13
-------
QUICKLIME
en
TORQUE-CONTROLLED WATER VALVE
OUST SHIELD
WATER SPRAYV
SLAKING WATER
'"'fli' ii :ii'[i''
•\HA r* x M . «A *
SLAKING COMPARTMENT^ DILUTION CHAMBER '
SLURRY DISCHARGE SECTION
DISCHARGE PORT-
CLASSIFIER
GRIT DISCHARGE
WATER FOR GRIT WASHING
GRIT ELEVATOR
Figure 4.5-7. Paste slaker.
Source: Wallace & Tiernan Division, Pennwalt, Corp.
-------
Table 4.5-1. CHARACTERISTICS OF DETENTION AND PASTE SLAKERS
Size of equipment for
equal capacity
Detention alaker
Large equipment
Paste Blaker
Smaller equipment
Size of available equipment
Quality of quicklime
Lime size range
Mechanical complexity
Hater-to-lime ratio
Ventilation requirements
Safety controls
Slurry quality
Initial cost
Can be built in any size
High-quality or poor-
quality quicklime
Can handle any size up
to 2-in. lumps
Very simple
Needs more slaking water
Natural draft ventilation
may be sufficient
Should have high-temper-
ature alarm. Low-
temperature alarm is
advisable
Larger particles
Less costly
Available only in standard
sizes, with capacities
limited to about 8000 lb
quicklime/h
High-quality quicklime
Should be limited to particles
smaller than 3/4 in.
Moderately complex
Needs less slaking water
Needs a powered ventilation
system
Should have high torque alarm
and a safety shutdown system.
Needs torque controls to
regulate addition of water
Usually slightly smaller
particles
More costly
-------
Slakers must be ventilated to prevent condensation of vapor
inside the dry feeder. If it is not allowed to escape, the
vapor condenses and reacts with lime dust to form hard deposits
on the slaker surface. A slaker ventilation system must remove
the hot, humid air. However, it should also be designed to
prevent discharge of lime dust into the atmosphere.
As the slaking temperature increases, the amount of mois-
ture evaporating from the slaking mixture increases; therefore,
paste slakers generally create more problems from condensation
than do detention units. Smaller detention slakers are simply
vented into the immediate area via a small scrubbing column,
through which a portion of the cool slaking water is admitted.
Only a very small draft of air is needed; natural draft is often
sufficient to protect the feeder. Less moisture is released ih
the work area; however, if a small stream of compressed air is
used to discharge the humidified air through plastic ductwork to
the outside of the building, the feeder area should be protected
from the moisture.
Because they produce greater amounts of moisture, paste
slakers usually require some type of forced ventilation.
Wallace & Tiernan supplies a water-operated eductor that uses a
portion of the dilution water to exhaust a small amount of air
from the slaking chamber. The same device is used with some
brands of detention slakers.
Most ventilation systems supplied with commercial slakers
are poorly suited for installations where the slaker is operated
intermittently. When the slaker is shut down and flow of water
is stopped, ventilation systems that use water eductors also
stop functioning, even though water continues to evaporate from
the slaking mixture. Condensation within the feeder usually
occurs on each shutdown. With a belt-type feeder, this condi-
tion is usually tolerable. With vibrating feeders, screw con-
veyors, or oscillating hoppers, however, condensation can cause
maintenance difficulties.
Although lime slurries are most often prepared by detention
and paste slakers, other methods of reacting water with quick-
lime are available. One method is "hydration," or "dry hydra-
tion," the system used by lime manufacturers to produce commer-
cial hydrated lime. Hydration units are small chemical plants
that react lime with steam in closed vessels and remove the
hydrated product by air classification. The resulting powder is
mixed with water to form a slurry. Dry hydration is most eco-
nomical if the heat of the reaction can be profitably recovered.
In many installations, the heat of reaction is used to preheat
air for feeding to a lime kiln. This process has been proposed
for use in large lime scrubbing FGD system because it allows
preparation of the slurry with recycled water, which cannot be
used directly in conventional lime slaking.
4.5-16
-------
Another potential lime slurry preparation system is the
"ball mill slaker," which wet grinds the quicklime. The resul-
ting slurry consists of calcium hydroxide, grit, and scale
compounds finely ground and suspended in water. The advantage
of this process is that close chemical control is unnecessary.
Any type of water can be used, the slaking temperature is unim-
portant, and even the poorest grades of lime can be treated.
Any limestone that might be in the feed material is also present
in the slurry and is available alkali in the FGD absorber.
Disadvantages of the ball mill include high initial cost, very
high operating costs, formation of an abrasive slurry, and
operation so noisy that the equipment is usually housed in a
separate building.
Some manufacturers offer general purpose mixing equipment
as a substitute for a lime slaker. Simple tanks equipped with
agitators have been used occasionally for slaking, on the as-
sumption that it is merely a mixing process. With this equip-
ment, quicklime and water are fed into the tank and the mixture
is stirred briskly. Although a "batch slaker" of this type is a
simple device, it invariably produces a poor quality slurry.
Even with high-energy agitation, slaking may not be uniform.
Hard, crystalline lime particles are formed, slaking is usually
incomplete, and part of the lime is lost as a hard scale that
forms in the tank. The slurry is usually very erosive and
reacts slowly in the FGD absorber. This type of system is much
larger than a continuous slaker and requires more operating and
maintenance labor.
4.5.3.4 Water Requirements for Slaking—
When quicklime is slaked in either a detention or a paste
slaker, the quality of the lime is affected by the slaking water
used. impurities, such as those found in recycled water from a
lime scrubbing FGD system, reduce the slaking rate and cause the
production of large, dense particles of partially hydrated lime.
The slurry may be more abrasive and thereby increases the main-
tenance requirement for the FGD system. Slaking water should be
free of high concentrations of ions such as carbonates, bicar-
bonates, sulfates, or phosphates that will precipitate in the
presence of calcium and cause a scaling problem. Similarly,
other metal ions that will precipitate in the presence of hy-
droxide ions are also objectionable. High concentrations of
chlorides in slaking water do not appear to be detrimental to
the slaking process; however, high concentrations of chlorides
may increase the degree of equipment corrosion.
Opinions differ regarding the use of recycled scrubber
water for slurry dilution. if the slaker is operated properly
and if complete slaking of quicklime has been achieved, dilution
with recycled water probably may be satisfactory, and only
minimal use of freshwater would be recommended. A portion of
the lime will react with the dissolved sulfates and sulfites in
the recycled water, causing precipitates to form on a propor-
tional amount of suspended lime.
4.5-17
-------
The slaked lime is usually diluted to a concentration of 10
to 25 percent calcium hydroxide by weight. Dilution is required
so that the slurry can be pumped successfully with centrifugal
pumps and fed into scrubbing equipment through control valves.
4.5.3.5 Existing Facilities—
Table 4.5-2 presents data from six existing plants and the
now-terminated experimental scrubber at the Four Corners Power
Station. Updated information is presented in "EPA Utility FGD
Survey" prepared bimonthly by PEDCo Environmental, Inc., for the
U.S. Environmental Protection Agency (EPA Contract Number 68-01-
4147, Task No. 3).
4.5.4 Slurry Stabilization and Storage
4.5.4.1 Introduction—
Slurry storage not only provides a surge volume between the
slaker and the FGD process, but also allows time to "stabilize"
the slurry. Addition of dilution water to a concentrated slurry
causes a series of chemical reactions between the lime and the
minerals dissolved in the water, such as alkaline-earth salts,
chlorides, sulfates, phosphates, etc. The reactions, which are
normally completed in less than 15 minutes in a slurry prepara-
tion system, cause the formation of hard, insoluble, crystalline
materials. The primary function of slurry storage is to hold
freshly diluted slurry until these scale-forming reactions are
complete. The slurry is then said to be stabilized. If no more
water is added, and if the slurry does not absorb carbon dioxide
from the air, no further scale formation reactions will occur.
Maintenance expenses therefore can be greatly reduced by allow-
ing sufficient stabilizing time for the slurry before it passes
through pumps, small-diameter piping, and/or control valves.
4.5.4.2 Service Description—
A critical design point in a slaker installation is the
conveying of diluted slurry to the stabilization tank. Since
crystalline scale compounds are formed most rapidly during the
first minute after slurry dilution, slurry should be transferred
to the stabilization tank by the fastest possible method.
Ideally, the slaker would be located directly above the stabili-
zation tank and the slurry would simply drop through a large
chute into the tank. If horizontal movement of the slurry is
required, open troughs that are easily removable for cleaning
are preferable to piping. In no case should slurry be pumped
directly from a slaker into a storage tank, since pump failures
and plugging of the pipes could be excessive.
In a well-designed system, a large stabilization tank is
used so that most of the scale compounds are present as a sus-
pension. As further scale compounds are formed, they adhere to
the suspended crystals, which increase in size and eventually
settle to the bottom of the tank as a loosely compacted sludge.
4.5-18
-------
Table 4.5-2. OPERATING CHARACTERISTICS OF LIME SLAKING FACILITIES1'2' ''
l
M
vo
Plant Name
Unit capacity, MM
Slaker type
Quicklime type*
Quicklime size,
in. (top site)
Number of feeders
Number of slakers
Feeder type
Normal lime rate
each feeder,
Ib/h
Hater uaed for
slaking
Slaking
temperature, *C(*I
Slaker alarms
Slaker shutdown
controls
Mater used for
dilution
final slurry
concentration,
t solids
Grit removal
Bruce
Mansfield
835
Detention
Thiosorbic
1-3/4
4
4
Gravimetric
48,000
Fresh
93 (199)
n
NO
NO
Recycled
10
Yes
Four
Corners
160
Detention
High
calcium
3/8
1
1
Not
reported
Not
reported
Fresh
82 (100)
Yes
No
Fresh
25
Yes
Phillips
410
Paste
High-calcium
dolomite
1/2
4
4
Gravimetric
8000
Fresh
60 (140)
Yes
Yes
Fresh
14
Yes
Elrama
510
Paste
Thiosorbic
1-1/2
4
4
Gravimetric
8000
Fresh
60 (140)
Yes
Yes
Fresh
14
Yes
Conesville
400
Paste
Thiosorbic
1-3/4
S
5
Screw
conveyor
8000
Fresh
74 (165)
No
No
Recycled
Approx. IS
Yes
Green
River
64
Mix
tanks
High
calcium
3/4
1
2
Screw
conveyor
4000
Fresh
55 (131)
No
No
Fresh
20-25
Yes
••fer to Section 4.4.1 for definitions of quicklime types.
' High-calcium lime usedt various dolomitic limes experimented
with to prevent scaling.
-------
In a small tank, there would be proportionately fewer suspended
crystals and a larger proportion of the scale-forming compounds
would attach to crystals adhering to the walls of the tank, thus
increasing the formation of a hard scale.
Figure 4.5-8 depicts a typical stabilization tank with a
baffle installed to prevent short-circuiting of diluted slurry
to the FGD system. A dilution tank with two chambers or two
separate, staged tanks may improve the quality of the slurry fed
to the FGD absorber. In such a slurry handling system, the
first tank would be very large. Slurry from this tank might
flow by gravity into a second tank. Fully stabilized slurry
would then be pumped from the second tank to the scrubber sys-
tem.
Solids tend to accumulate in the corners of square tanks or
along the perimeter in the bottom of flat-bottomed circular
tanks, and they may become compacted into a concrete-like mass.
Such deposits usually are left in place when the tanks are
cleaned. if agitation is efficient, however, reduction in the
useful volume of the tanks as a result of these deposits is not
likely to exceed 2 to 5 percent.
Lime slurry tanks should be covered to prevent excessive
absorption of ambient carbon dioxide, which increases scale
potential by the formation of calcium carbonate. If transfer
tanks are used, it is best to connect them and the storage tanks
with vent piping in order to minimize the amount of fresh atmos-
pheric air drawn into the system during slurry transfer. A
slurry stabilization and storage tank should be fitted with vent
piping if the air near the tank is likely to contain more than
normal atmospheric quantities of carbon dioxide.
Although proper design will reduce the frequency of
cleaning, stabilization and storage tanks must be shut down and
emptied periodically so that deposits can be removed. Instal-
lation of side-entering manholes will simplify this procedure.
Proper design of the stabilization and storage tank system
avoids the need for redundant slakers and lime feeders, thereby
decreasing initial costs and simplifying operation. Providing
duplicate stabilization tanks permits easy maintenance. When
both tanks are filled with slurry, the slaker may be shut down
and cleaned without interrupting scrubber operation. If any
part of the equipment is to be duplicated, it should be the
stabilization tank system. Slakers and feeders usually can be
repaired quickly, provided spare parts are stocked. Care should
be taken in the design of a lime system to allow for a large
inventory of finished slurry so that scrubber operation can
continue while the feeding and slaking equipment is down for
maintenance.
4.5-20
-------
FRESHLY DILUTED
SLURRY FROM
SLAKER
LIME SLURRY
RECIRCULATION
RETURN PIPING
FIXED BAFFLE
TO LIME SLURRY
~~*~ SUPPLY PUMPS
TANK VOLUME TO ALLOW
FOR AT LEAST 30 min
RETENTION TIME TO
MINIMIZE EFFECT OF
SHORT CIRCUITING OF
UNSTABILIZED SLURRY
Figure 4.5-8. Typical stabilization tank (simplified),
4.5-21
-------
Stabilization tanks are often constructed from carbon steel
and in some cases from fiberglass reinforced plastic.2 The
tanks should be about as deep as they are wide since shallower
tanks require more energy to achieve thorough agitation. Except
for use in tanks with capacities of less than 1000 gal, agita-
tors should be vertical-shaft, top-mounted units located axially
within the vessel. Agitators with internal bearings should not
be used, nor should any system that uses water for sealing or
lubrication. Turbine impellers are best for slurry agitation,
usually with motors connected to the shafts through speed-reduc-
tion gears.3 High-speed agitation is not needed for lime slur-
ry, because well-prepared lime particles settle very slowly.
Table 4.5-3 provides a rule-of-thumb guide for estimating the
horsepower requirement for an agitator motor.
Table 4.5-3. APPROXIMATE AGITATOR MOTOR HORSEPOWER
REQUIRED FOR LIME SUSPENSION3
Lime slurry solids
concentration
1 Ib/gal (17%)
2 Ib/gal (34%)
3 Ib/gal (51%)
Horsepower
per 1000 gal
of slurry agitated
0.25
0.50
1.0
Applies to tanks of 3000 to 15,000 gal capacity. Increase
horsepower by 50 percent for tanks of 1000 to 2500 gal
capacity. Applies to tanks with depth approximately
equal to diameter, containing four fixed baffles.
Modified from Preparation and Handling of Lime Slurries,
Wallace & Tiernan Division, Pennwe.lt Corporation.
The data on agitator motor horsepower given in Table 4.5-3
are too low for holding either particles of grit or large crys-
tals of scale in suspension. The heavier particles will accumu-
late at the bottom of the tank and must be removed during peri-
odic cleanings. Many engineers try to reduce the quantity of
heavy, abrasive material that enters the circulation pump and
passes through the control valve into the reaction tank. Others
prefer to increase agitation to hold heavy particles in suspen-
sion. They are willing to accept an increase in the abrasive-
ness of the slurry in exchange for reducing the frequency of
tank cleaning.
4.5-22
-------
Power consumption and turbine speed must be greatly in-
creased, perhaps doubled, if the tanks are not fitted with fixed
baffles. As shown in Figure 4.5-9, baffles consist of two to
four vertical plates, each about one-twelfth the diameter of the
tank, mounted on a framework that supports the plates away from
the sides and bottom of the tank at a distance of about one-half
the baffle width. Baffles, which break up the circular motion
of the slurry, should not be attached directly to the sides and
bottom of the tank because solid deposits will form behind them,
decreasing the effective volume of the tank and hampering slurry
agitation. Three baffles in a tank generally are sufficient.
More than four do not improve the agitation further.
A few operators further treat the lime slurry after stabil-
ization by pumping stabilized slurry into a final storage tank
through a classifier that separates heavier particles of grit
and scale.
4.5.4.3 Existing Facilities—
Table 4.5-4 summarizes the data on stabilization and stor-
age systems at operational lime FGD facilities. The available
data indicate that some provision for stabilization and removal
of heavier particles is included as part of the design in most
installations. The experimental system that was operated at the
Four Corners station of Arizona Public Service Company appears
to have been the most advanced.1'4 The slakers produced a lime
slurry, which was discharged into a dilution tank, from which
the heavier solids were later removed in a thickener. Clarified
slurry was pumped to the scrubber while underflow was removed to
a sludge disposal facility.
At Bruce Mansfield, (Figure 4.5-8) each of the three
slakers delivers about 900 gal/min of 10 percent lime slurry
into a 36-ft-diameter transfer tank, which provides almost an
hour of retention time at the maximum slurry production rate.5
The transfer tank is fitted with special underflow pumps to
remove accumulated solids. Slurry is transferred intermittently
from the transfer tank into a smaller recycle tank that allows
less than 10 min retention. A schematic of this system is shown
in Figure 4.5-10.
Data from the Conesville station are the most complete,2 as
hown in Figure 4.5-11. There the slurry from all slakers is
discharged into a 15,900-gal tank, which provides about 30
inutes of retention time at maximum slurry production rate.
Blurry at approximately 15 percent solids is pumped inter-
ittently with a pump designed to deliver 660 gal/min into a
large storage tank of 51,700-gal capacity. The slurry is then
umped as required to the scrubber with another set of pumps in
recycle loop designed to deliver up to 522 gal/min. Plant
rators rep0rt a usage of 483 gal/min of slurry at 60 percent
boiler load- Agitators are fitted with motors equivalent to
about 3 hp/1000 gal of slurry. Actual operating horsepower has
t been measured. The tanks have no baffles.
4.5-23
-------
D/24
TANK WALL
"D"
TANK DIAMETER
|« »| D/12
r I
D/12-U-^l
H
D/24
BAFFLE.
* TANK \ BANGLE IRON OR
D/24 BOTTOM-^ METAL ROD SUPPORTS
or (TYP.)
greater
ELEVATION
TWO-BAFFLE ARRANGEMENT
THREE-BAFFLE ARRANGEMENT
D/24
or
greater
FOUR-BAFFLE ARRANGEMENT
PLAN LOCATION
OF BAFFLES
Figure 4.5-9. Agitator baffle design.
4.5-24
-------
Table 4.5-4. STABILIZATION AND STORAGE SYSTEMS1'2'5'6'7
Ul
NJ
(Jl
Slurry from s lakers, maximum
feed rate
Slurry composition, % solids
Stabilization tank capacity,
gallons
Stabilization tank, agitator
motor horsepower
Storage tank capacity, gal
Storage tank, agitator motor
horsepower
Bruce
Mansfield
2,700
10
150.000
a
8,500
a
Four
Corners
20
25
a
a
a
a
Phillips
397
14
13,000
7.5
NA
NA
Elrama
297
14
13,000
7.5
NA
NA
Coneaville
517
Approx.
15
15,900
5
51,700
15
Green
River
61
20 to 25
1980
5
NA
NA
a Data not reported.
NA - Not applicable.
-------
SLURRY
FROM
SLAKERS
2700 gal /mi n
10% SOL IPS
SLAKER
TRANSFER
TANK
0
-------
SLURRY
FROM
SLAKERS
gal /mi n
517
HIN.
145
15% SOLIDS
,5 hp
0*0
J^
•
m
i
SLURRY TRANSFER TANK
15,900 gal
16 ft 5 in. DIAMETER
BY 10 ft DEPTH
(FRP)
gal/min
PUMP SEALS ADD
5 gal/min WATER
MAX. MIN.
522 150
15 hp
SLURRY STORAGE TANK
51,700 gal
20 ft DIAMETER
BY 22 ft DEPTH
(CARBON STEEL)
gal/min
MAX. MIN.
522 150
SLURRY TO
PUMPS
660 gal/min
(INTERMITTENT)
TRANSFER
PUMPS
Figure <.5-ll. Conesville stabilization and storage system.
-------
At Phillips and Elrama, all slakers discharge to a common
tank. Figure 4.5-12 presents a diagram of the stabilization and
storage system of the lime FGD facility at Phillips Power Sta-
tion, which is almost identical to the one at Elrama Power
Station. Slurry is pumped only once, directly to the scrubbers.
The original system was designed with a second tank, but this is
not now in use. Because of lower capital costs, the SO2 scrub-
bing systems at both the stations have switched from dual-stage
scrubbing using high-calcium lime to single-stage scrubbing
using magnesium modified lime.6'7
At Green River (Figure 4.5-13), slurry from the two slaking
tanks is dropped by gravity into a 1980-gal mix/hold tank fitted
with a 5-hp agitator.1
4.5-28
-------
LIME SLURRY
TO SCRUBBER
56.5 gal/mln
WATER |
20% SOLIDS
138.3 gal/mln
HYDRATED
LIME SLURRY
LIME
SLURRY
TANK
7.5 hp
15%
SOLIDS
TRANSFER TANK
15 ft 8 in.DIAMETER
-CX3-
LIME SLURRY
PUMPS
Figure 4.5-12. Phillips stabilization and storage system.
4.5-29
-------
BATCH FEEDS
10
J
SPARE
GRAVITY
DISCHARGE
OkD
MIX/HOLD TANK
1980 gal
265 ft3
WATER <25 gal/min
LIME 4000 Ib/h
ACTIVE
LIME
SLAKING
TANKS
(1 ACTIVE; 1 SPARE)
EACH 1680 gal
225 ft3
WATER
\
20% SLURRY
TO PUMPS
Figure 4.5-13. Green River slaking and stabilization system.
4.5-30
-------
REFERENCES
1. Laseke, B.A. Survey of Flue Gas Desulfurization Systems:
Green River Station, Kentucky Utilities. EPA-600/7-78-
048e, March 1978.
2. Private communication with D. Boston, Columbus and Southern
Ohio Electric Company, February 1978.
3. pennwalk Corporation. Preparation and Handling of Lime
Slurries. Wallace & Tiernan Division, 1976.
4. Alexander, W. , et al. Results of the 170-MW Test Modules
Program, Mohave Generating Station. In: EPA Flue Gas
Desulfurization Symposium, New Orleans, Louisiana, March
8-11, 1976.
5. Private communication with R. Forsythe and W. Norrocks,
Pennsylvania Power Company, February 1978.
6. Private communication with R. O'Hara and J. Mahone,
Duquesene Light Company, February 1978.
7. pernick. S.L. Elrama and Phillips Power Stations Lime
Scrubbing Facilities. In: EPA Flue Gas Desulfurization
Symposium, New Orleans, Louisiana, March 8-11, 1976.
4.5-31
-------
BIBLIOGRAPHY
Baker, K.J., and R.W. Jordan. Effect of Dissolved Solids in S02
Scrubber Water Used for Lime Slaking. In: WWEMA Industrial
Pollution Conference, April 1975.
Kunesh, C.J. The Calcination and Slaking of Quicklime. In:
International Water Conference, October 1976.
Laseke, B.A. EPA Utility FGD Survey, December 1977 - January
1978. EPA-600/7-78-051a, March 1978.
Pennwalt Corporation. Preparation and Handling of Lime Slur-
ries. Wallace & Tiernan Division.
4.5-32
-------
GLOSSARY
burning: Use of insufficient water during a quicklime slaking
operation resulting in a partially slaked, unreactive lime
slurry.
dolomitic: Indicates the presence of approximately equal molar
amounts of calcium and magnesium in a limestone, quicklime,
or hydrated lime.
drowning: Use of too much water or too little agitation during
a quicklime slaking operation resulting in a poorly hy-
drated lime.
feeder: A mechanical device that regulates rate of flow of bulk
solids. Also known as dry feeder or dry chemical feeder.
gravimetric: Indicates a measurement on the basis of mass.
hard-burned: Indicates a quicklime manufactured at conditions
resulting in low reactivity toward water.
high-calcium: Indicates the. presence of less than 5 percent
magnesium in a limestone, quicklime, or hydrated lime.
hydrated lime: The material resulting from the reaction between
quicklime and water, consisting primarily of calcium hy-
droxide or a mixture of calcium hydroxide and magnesium
oxide and/or hydroxide. Also known as lime hydrate or
slaked lime.
hydration: the process of reacting quicklime with water to
produce hydrated lime, usually in the form of a dry powder.
lime: A caustic infusible solid that consists of calcium oxide
together with magnesia, that is obtained by calcining
limestone.
limestone: A sedimentary rock consisting mainly of calcium
carbonate or mixture of calcium carbonate and magnesium
carbonate.
lime slurry: A more or less viscous slurry formed by slaking
quicklime with excess water or by addition of water to
hydrated lime. Also known as milk of lime.
4.5-33
-------
magnesian: Indicates the presence of from 5 to 35 percent
magnesium in a limestone, quicklime, or hydrated lime.
quicklime: The product of the calcination of limestone, com-
posed primarily of calcium oxide if high-calcium limestone
is used or of approximately equal molar amounts of calcium
oxide and magnesium oxide if dolomitic limestone is used.
scale: Insoluble or slightly soluble inorganic materials, often
crystalline, formed by the reaction of lime with impurities
in water or with atmospheric constituents.
slaker: Mechanical equipment designed to produce the slaking
reaction.
slaking: The process of allowing quicklime to react with water
to produce hydrated lime. In popular usage, slaking indi-
cates use of excess water under conditions of close chemi-
cal control to produce a hydrated lime slurry or paste.
soft-burned: Indicates a quicklime manufactured at conditions
resulting in high reactivity towards water.
stabilization: The process of holding a freshly prepared lime
slurry until all chemical reactions between slurry constit-
uents have approached equilibrium.
volumetric: Indicates a measurement on the basis of volume.
4.5-34
-------
CONTENTS
4.6 SCRUBBER/ABSORBER
4.6.1 Introduction
4.6.2 Industrial Scrubber-absorbers
4.6.2.1 Tray Absorbers
4.6.2.2 Packed Absorbers
4.6.2.3 Moving Bed Absorbers
4.6.2.4 Venturi Scrubber-absorbers
4.6.2.5 Spray Scrubber-absorbers 4.6-4
4.6.3 Materials for Construction of Scrubbers 4.6-5
4.6.3.1 Introduction 4.6-5
4.6.3.2 Steel and Alloys 4.6-5
4.6.3.3 Coatings 4.6-10
4.6.3.4 Rubber Liners 4.6-15
4.6.3.5 Brick 4.6-17
4.6.3.6 Conclusions 4.6-19
4.6.3.7 Existing Facilities 4.6-19
4.6.4 Mechanical Design 4.6-21
4.6.4.1 Introduction 4.6-21
4.6.5 Scale Formation 4.6-22
4.6.5.1 Types of Scale 4.6-23
4.6.5.2 Problems Resulting from Scale Formation 4.6-25
4.6.5.3 Techniques to Prevent Scale Formation 4.6-26
4.6.5.4 Effects of Various Factors on Scaling 4.6-27
4.6.6 Process Design Variables 4.6-28
4.6.6.1 Stoichiometry 4.6-28
4.6.6.2 L/G Ratio 4.6-29
4.6.6.3 pH 4.6-29
4.6.6.4 Increased Gas Velocity 4.6-30
4.6.6.5 Liquid Distribution 4!6-30
4.6.6.6 Water Balance 4]6-31
4.6.6.7 Interfacial Area 4!6-33
4.6-i
-------
CONTENTS (continued)
Page
4.6.7 Existing Facilities 4.6-33
4.6.7.1 Louisville Gas and Electric, Cane Run Unit 4 4.6-33
4.6.7.2 Louisville Gas and Electric, Paddy's Run
Unit 6 4.6-35
4.6.7.3 Kentucky Utilities, Green River Power
Station 4.6-35
4.6.7.4 Duquesne Light, Phillips 4.6-35
4.6.7.5 Duquesne Light, Elrama 4.6-35
4.6.7.6 Columbus and Southern Ohio Electric,
Conesville Unit 5 4.6-36
4.6.7.7 Pennsylvania Power, Bruce Mansfield Units
1 and 2 4.6-36
References 4.6-37
4.6-ii
-------
4.6 SCRUBBER/ABSORBER
4.6.1 Introduction
In this section, the term "scrubber" is used for the device
performing particulate removal as its major function, whereas
the term "absorber" is used to describe the device that is
primarily designed to remove SO2. The principal unit operation
involved in a lime FGD system is gas absorption by chemical
reaction. The SO2 in the flue gas is absorbed by a lime slurry,
which reacts with it chemically. The purpose of the equipment
used for the gas-liquid operation (the absorber) is to provide
intimate contact of the two fluids in order to facilitate inter-
phase mass transfer of SO2 . The rate of mass transfer is di-
rectly dependent on the interfacial surface area (the surface
exposed between the two phases), hence the nature and degree of
dispersion of one fluid into the other is of prime importance.
The equipment may be classified according to whether it dis-
perses either the gas or the liquid; however, the most widely
used types of absorbers are classified by the type of internals.
Each type of absorber is discussed with respect to the
salient design features, advantages, disadvantages, and the
vendors supplying that particular type. Table 4.6-1 gives a
summary of existing S02 absorbers in the operational lime FGD
systems. The degree to which the mass transfer characteristics
of an absorber can be utilized and its associated cost will
determine the applicability of the absorber for a specific SO2
removal requirement. Major factors that determine the operating
cost are pressure drop and the L/G ratio. Flexibility in the
design, which is the ability of the absorber to retain its S02
removal efficiency at reduced gas flow rates, is also a major
consideration in selection of an absorber.
4.6.2 Industrial Scrubber-absorbers1'2
4.6.2.1 Tray Absorbers —
A tray absorber consists of a vertical tower with one or
more trays mounted transversely inside. Gas comes in at the
bottom of the tower; passes upward through perforations, valves,
slots, or other openings in the tray; then bubbles through the
liquid to form a froth; disengages from the froth; and passes on
to the next tray above. The liquid enters at the top and flows
downward by gravity. On its way, it flows across each tray and
through a downspout to the tray below. The overall effect is a
multiple countercurrent contact of gas and liquid, although each
tray is characterized by a crossflow of the two. On each tray
the fluids are brought into intimate contact, interphase dif-
fusion occurs, and the fluids are separated.
4.6-1
-------
Table 4.6-1.
S02 ABSORBERS IN OPERATIONAL LIME FGD SYSTEMS
Unit/Utility
Green River
No*. 1, 2, J,
Kentucky Utilities
Bruce HantCield
Nos. 1, 2
Pennsylvania Cover
Cane Run Ho. 4
Louisville GtE
Paddy's Run Ho. 6
Louisville GfcE
Elraaa
Duquesne Power
Phillips
Duquesne Power
Cone* vi lie No. S
ColuMbus ii
Southern
SOj Absorber
Vendor
uner ican
Air
Filter
Che* ico
American
Air
Filter
Combustion
Engineering
Che Hi co
Che* i co
UOP
Type
Moving
Venturl
Moving bed
Moving bed
v*nturi
Venturi
Hoving bed
L/C, at 120T.
qal/1000 id
14
60
Si - 65
15 - 18
40
40
50
Ap, in. H,O
4
16
4
12
It
16
6
Internal*
Solid
spheres,
spray
nozzles
Plunb bob
Solid
spheres,
1-1/4-Ln. (dia.l
spxsy nozzle*
lUrblet,
1 in. Idia.l ,
3-in. Ideep)
bed
Upper con*
brill notile,
spr«y nozzles
Upper cone,
bull nozzle.
sprny noziles
Hollow
spheres,
1 in. (dia.l:
2- in. (deep)
bed
Dimensions, ft
20 x 20 x 27.5
JS Idia.) « 50
20 i 20 x 27.5
17 « It x 50
30 (dia.l x tO
45 (dia.) i 50
Hater 1st*
of construction
Absorbers
Mild
steel,
J/4-ln.
acid-proof
lining
Mild
steel,
polyester
Ilakeglase
lining
Mild
steel,
3/4-ln.
acid proof
lining
Mild
steel.
J-l/J-in.
thick
flake lining
Mild
steel,
Cellccted
Hlld
steel,
Ceilcoted
Mild
steel ,
rubber
lined
Internal*
PVC
balls,
c*ruic
nonlas
3l*L
ss
Ceilcotad
Polyurethan*
balls.
c«r*«lc
noztlea
Class,
lit SS
support*
UtL
SS,
Ceilcoted
3161
SS.
Ceilcoted
Heoprene bell
No. of modules
per unit
One for all the
three units
6
2
2
5
5
2
a\
l
-------
The number of theoretical trays in a tower is dependent
only upon material balance and equilibrium considerations. The
tray efficiency, and therefore the number of actual trays, is
determined by the mechanical desion and the operating condi-
tions. The diameter of the tower is principally determined by
the quantities of liquid and gas flowing through the tower per
unit time. Once the number of theoretical trays is determined,
optimization of absorber design is based on several opposing
factors described below.
Deep pools of liquid on the trays lead to high tray effi-
ciencies because of long contact time, but also lead to high
pressure drop per tray and a possibility of flooding, a condi-
tion in which liquid may fill the tower resulting in high liquid
carryover by the effluent gas and slugs of foam. High gas
velocities, within limits, provide good vapor-liquid contact
through excellence of dispersion, but lead to excessive entrain-
ment and high pressure drop. The general design procedure
involves selection of design configurations, based on experience
followed by calculations to ensure that the pressure drop and
the flexibility are satisfactory.
At present, none of the utility lime FGD systems uses a
tray absorber, primarily because of the severe plugging problems
associated with lime slurry handling through a close tortuous
path. The tray absorbers are, however, extensively used in
other industrial boiler FGD systems, such as those using sodium-
based alkali absorption, most of which are supplied by Koch
Engineering and FMC Corporation. Babcock and Wilcox has also
supplied tray absorbers at some of the utility limestone FGD
systems.
4.6.2.2 Packed Absorbers—
packed towers, used for continuous countercurrent contact
of liquid and gas, are vertical columns filled with packing.
The liquid trickles down through the packed bed, thus forming a
film of large surface area to contact the gas. The gas stream
to be cleaned typically flows upward through the packing. The
desirable properties for tower packings are larger specific
packing surface (the surface area per unit volume of packed
space), high fractional void volume, low density but high struc-
tural strength, chemical inertness to the fluids being pro-
cessed, and low cost.
packed tower absorbers are also not used at any of the
utility lime FGD systems because of their vulnerability to
plugging- The plugging problem has been alleviated in a modifi-
cation of the packed absorber, the moving bed absorber described
below. In industrial boiler FGD systems using sodium and
ammonia absorption, packed absorbers are offered by the Ceilcote
Company and Chemico, respectively.
4.6-3
-------
4.6.2.3 Moving Bed Absorbers—
Moving bed absorbers provide a zone of mobile packing,
usually plastic or glass spheres, where gas and liquid can
intimately mix. The absorber shell holds the perforated plate
on which the movable packing is placed. Gas passes upward
through the packing while liquid is sprayed up from the bottom
through the perforated plate, and/or down on top of the moving
bed. Because of the high gas velocity, the packing material
moves around constantly when the scrubber is operating. This
movement makes the bed turbulent and keeps the packing clean
The pressure drop of a moving bed is typically 2.8 to 5.9 in*
H2O per stage.
The major vendors offering moving bed absorbers for utility
FGD systems are American Air Filter and UOP. Combustion Engi-
neering has discontinued this type of absorber only recently.
4.6.2.4 Venturi Scrubber-absorbers—
Venturi scrubber-absorbers are towers with spray devices
that utilize a moving gas stream to atomize liquid drops, and
then accelerate these drops through the throat of a venturi
High gas velocity is used to produce a high relative velocity
between gas and liquid, which promotes particle collection The
scrubbers usually have a variable throat, whereas the absorbers
often have a fixed throat. High pressure drop venturi scrubbers
can collect particles with high efficiency; however, mass trans-
fer characteristics are limited because of the cocurrent nature
of the gas-liquid flow.
Two notable modifications to the conventional converging-
diverging design of the Venturis are annular orifice and rod
bank towers. in the annular orifice tower, which has the con-
verging section and the throat, gas impinges on a movable disc
while liquid flows cocurrently down the walls of the converging
section. In the rod bank tower, gas and liquid flow cocurrently
through the throat across several runs of rods, which usually
have adjustable spacing.
Chemico offered venturi scrubber-absorbers in the early
utility FGD systems. Combustion Equipment Associates still
offers this type of scrubber-absorber in utility FGD systems.
4.6.2.5 Spray Scrubber-absorbers—
A spray scrubber utilizes spray nozzles for liquid droplet
atomization. The sprays are directed such that the gas passes
upward through the descending atomized liquid droplets. If the
tower is vertical, the relative velocity between the droplets
and the gas is eventually the terminal settling velocity of the
droplets. Most droplets eventually hit the walls in a tall
4.6-4
-------
tower. Spray absorbers can also be used in horizontal configur-
ation. The flow of gas and liquid is, then, crosscurrent. An
EPRI report on the evaluation of a horizontal scrubber and
application in a 1-MW pilot plant at the Colbert Station of the
Tennessee Valley Authority will be published in late 1978.
Cocurrent horizontal and vertical absorbers are also being
investigated.
Spray towers are used for both particle collection (scrub-
bers) and mass transfer (absorbers). They generally have low
pressure drop and high liquid flow rate and are the least expen-
sive type of absorber in terms of capital expense. Particle
collection is limited by the terminal settling velocity and
diameter of the spray droplets.
Chemico, Combustion Engineering, Combustion Equipment
Associates, M.W. Kellogs, and Peabody Engineering are the lead-
ing FGD vendors who offer spray absorbers preceded by an ESP for
particulate collection.
4.6.3 Materials for Construction of Scrubbers
NOTE: Much of the information contained within Section 4.12 of
this data book is also pertinent to the following discus-
sion.
4.6.3.1 Introduction—
The choice of materials for the construction of scrubbers
and abosrbers is complex and depends on many variables, which
include the planned life of the unit, operation of the unit,
economic considerations, safety considerations, and the unit
location and environment.
The type of corrosion varies depending on the location in
the scrubber or absorber. For example, the venturi throat is
susceptible to high abrasion and hence suffers from erosion-
corrosion, whereas general corrosion is a major problem down-
stream from the mist eliminator. Operating conditions at parti-
cular locations in the module are an important factor in
material selection for a scrubber or absorber.
4 6.3.2 Steel and Alloys3'4'5--
Spool tests have been performed in several FGD systems by
olacing the spools in various locations in the scrubber. The
test data given in Tables 4.6-2 and 4.6-3 were reported by
Tennessee Valley Authority.3 The test conditions and the
analyses of different types of steel and other alloys are given
n Tables 4.6-4 and 4.6-5 respectively.
4.6-5
-------
Table 4.6-2. VENTURI THROAT CORROSION SPOOL TEST DATA'
Specimens:
Round, machined-edge spool pieces 2 in. O.D. x
23/64 in. I.D. x approximately 11 gauge. Single
air annealed cross weld. Insulated Teflon
separators.
Temperature: 80° to 170°F
Duration of test: 2370 hours
Material
Allegheny Metal 6X
Allegheny Metal 29-4
Mild steel, ASTM-285
Climax 18-2
Hastelloy C-276
Haynes Alloy 6B
Inconel 625
Jessop 700
Multimet
Nitronic 50M
Stainless T-216
Stainless T-316L
(2.3% Mo)
Stainless T-316L
(2.8% Mo)
Stainless T-317L
U.S.S. Cor -Ten A
Zirconium 702
Corrosion rate,
mils/yr
20
14
>1855
32
44
9
29
31
26
16
12
12
10
9
>2120
6
Based on general corrosion attack.
4.6-6
-------
Table 4.6-3. SCRUBBER CORROSION SPOOL TEST DATA BELOW
AND ABOVE THE MIST ELIMINATOR
Specimens: Round, machined-edge spool pieces 2 in. O.D.
x 23/64 in. I.D. x approx. 11 gauge. Single air-
annealed cross weld. Insulated Teflon separators
Temperature: 80° to 170°F
Material
Allegheny Metal 6X
Allegheny Metal 29-Y
Mild steel, ASTM-285
Climax 18-2
Hastelloy C-276
Hastelloy G
Haynes Alloy 6B
Inconel 625
Jessop 700
Multimet
Nitronic 50M
Stainless T-316L
(2.3% Mo)
Stainless T-316L
(2. 8% Mo)
U.S.S. Cor-Ten A
Zirconium 702
Below the mist eliminator3
Corrosion rate,a
mils/yr
<0.05
0.05-0.49
171
0.05-0.49
<0.05
0.05-0.49
-
0.05-0.49
—
<0.05
0.05-0.49
0.05-0. 49
<0.05
170
<0.05
Pitting,
mils
_d
_
5
10
—
—
9
_
7
—
4
_
_
_
—
Above the mist eliminator
Corrosion rate,0
mils/yr
<0.05
0.05-0.49
26
<0.05
<0.05
0.05-0.49
<0.05
<0.05
<0.05
<0.05
1.0
1.0
0.05-0.49
43
<0.05
Pitting,
mils
17
_
5
6
_
—
—
_
—
—
-
_
—
11
—
Test duration 220 h, 5000 ppm MgO added to the slurry
"'Test duration 1490 h, 5000 ppm MgO added to the slurry
•^
'Based on general corrosion attack.
Indicates negligible corrosion rate or pitting.
-------
Table 4.6-4. CONDITIONS FOR THE CORROSION TESTS
AT THE SHAWNEE STATION OF TVA3
Gas temperature, °F
Gas velocity, ft/s
Gas flow rate, 1000
acfm at 330°F
Test spool location
Venturi Before mist After mist
throat eliminator eliminator
80-170 125-130 125-130
40-100 4.5-9.4 4.5-9.
15-30 15-30 15-30
4
Gas composition
Component
S02
co2
°2
H2°
HC1
N2
Fly ash, gr/scf
% by volume
0.2-0.4
10-18
5-15
8-15
0.01
74
2-7
4.6-8
-------
Table 4.6-5. CHEMICAL ANALYSIS OF ALLOYS'
Alloys
Al 6Xa
Al 29-4a
Climax 18-2a
Cor-Ten Aa
Hascelloy C-276a
Hastelloy G3
Haynes 6B*
Inconel 625
Jessop 700
Mild steel A-28S
Multimet
Nltronic SOU
Type 216 SS*
Type 316L SS
Type 316L SS*
Type 316L SSa
Zirconium 702
Chemical analysis, I
C
0.027
0.004
0.016
0.11
0.002
0.02
0.96
O.lb
0.03
0.35b
0.2b
0.06b
0.069
0.020
0.025
0.022
0.015
Cr
20.32
29.3
18. 44
0.66
15.87
21.72
29.75
20-23
21.00
18-22.5
21
19.54
17.1
18.0
18.61
c
Ni
24.17
0.12
0.39
0.33
Bal.
Bal.
2.13
Bal.
25.00
18-22
14
6.77
13.8
13.9
13.62
Fe
Bal.
Bal.
Bal.
Bal.
5.96
18.68
2.36
5.00b
Bal.
Bal.
Bal.
Bal.
Bal.
Bal.
Bal.
Bal.
c
Cu
0.21
0.36
1.77
0. 35b
0.07
0.05
0.45
Ho
6.42
3.95
2.08
16.32
6.69
1.08
8-10
4.5
2.75-3.75
1.5-3.0
2.31
2.3
2.77
3.16
Mn
1.46
0.10
0.4
0.39
0.49
1.30
1.40
0.5b
1.70
0.80b
6
8.21
1.30
1.38
1.62
Si
0.56
0.05
0.39
0.44
<0.01
0.34
0.36
0.5b
0.50
1.00b
0.23
0.49
0.54
0.60
P
0.023
0.013
0.098
0.012
0.021
0.015b
0.05b
0.04b
0.023
0.016
0.011
0.021
S
0.004
0.013
0.026
0.010
0.011
0.015b
0.05b
0.03b
0.005
0.016
0.012
0.009
Others
N, 0.03
N, 0.010
N, 0.013; Ti, 0.33
Co, 1.84; W, 3.51; V, 0.25
Co, 1.57; Cb + Ta, 2.13; U, 0.54
Co, Bal.; U, 4.30
Co, 1.0b; Cb + Ta. 3.15-4.15; Al , 0.4b; Ti, 0.4b
Cb, 0.30
Cb, 0.75-1.5; Co, 18-22, N, 0.1-0.2; W, 2-3
N, 0.2-0.4; Cb, 0.1-0.3; V, 0.1-0.3
N, 0.358
B, 0.0008; Cb, 0.02; Co, 0.17; N, 0.065; Al, 0.012; TI, 0.004
H, 0.05; Hf, <0.10; Zr + Hf, >99.2
I
VO
Analysis was supplied with the material.
Maximum.
Cr + Fe, 0.10Z by weight.
-------
In the venturi throat, the greatest attack on the specimens
was due to erosion-corrosion. The high velocity of the lime
slurry, containing fly ash, SO2, CO2, and HC1, accounted for the
high rates of deterioration. Specimens of Cor-Ten A and mild
steel, which were completely destroyed, had penetration rates
greater than 1850 mils/yr. The most promising alloys in the
order of decreasing resistance to erosion-corrosion were Zir-
conium 702, Waynes 6B, Type 317L, AL 29-4, and AL 6X.
In the recirculation tank, the corrosion rates of mild
steel and Cor-Ten A were 35 and 26 mils/yr. Attack on the other
alloys was negligible. During this particular series of tests,
the attack on mild steel and Cor-Ten A varied greatly in the
scrubber tower.3 In the earlier tests, corrosion of the speci-
mens exposed in the top of the tower was greater than it was for
specimens exposed near the middle and bottom. However, during
the fourth series, corrosion was less for the specimen near the
mist eliminator. The installation of an automatic spray system
for washing the mist eliminator also washed the test spools. At
other test locations,4'5 where the stainless steel specimens
were coated with deposits of solids, pitting occurred more
frequently.
4.6.3.3 Coatings6'13--
Many types of coating are available for use in FGD systems.
The following list shows the basic types of resin that can be
used in a scrubber.
0 Bituminous
0 Chlorinated rubber
0 Coal tar epoxy
0 Epoxy
0 Polyester
0 Polyurethane
0 Vinyl ester
0 Furan
0 Phenolic
Of these resins, polyester, bituminous, epoxy, vinyl ester,
and furan are the most common ones found in utility FGD systems.
Furan, polyester, epoxy, and vinyl ester resins can be
applied as a coating by themselves, with glass flakes, or with a
fabric mat. Glass flakes are added to the resins to reduce
permeability, add strength, and minimize the possibility of
pinholes in the coating. A fabric mat is used with a resin to
increase the strength of a coating. Tables 4.6-6 through 4.6-9
show some physical characteristics of furan, epoxy, vinyl ester,
and polyester resins, respectively, with and without glass
flakes or a fabric mat.6
4.6-10
-------
Table 4.6-6. TYPICAL CHARACTERISTICS OF FURAN RESINS"6
en
I
Properties
Tensile strength,
psi
Coefficient of
expansion,
in./in./°F
Barcol hardness
Temperature re-
sistance, °F
Flexural strength,
psi
Abrasion resistance,
Taber Wear Index
Resin
1,200
2.0 x 10~5
NRa
350
3,800
NR
Flake glass
1,250
1.4 x 10~5
28
125
2,660
83
Fabric reinforced
8,150
1.5 x 10~5
20
125
19,850
57
NR - Not reported.
-------
Table 4.6-7. TYPICAL CHARACTERISTICS OF EPOXY RESINS
I
H1
ro
Properties
Tensile strength,
psi
Coefficient of
expansion,
Barcol hardness
Temperature re-
sistance, °F
Flexural strength,
psi
Abrasion resistance
Taber Wear Index
Resin
1,800
3.0 x 10~5
NRa
175
3,800
NR
Flake glass
3,350
1.5 x 10~5
40
160
6,735
129
Fabric reinforced
3,400
1.9 x 10~5
45
180
9,500
140
NR - Not reported.
-------
Table 4.6-8. TYPICAL CHARACTERISTICS OF VINYL ESTER RESINS6
Properties
Tensile strength,
psi
Coefficient of
expansion.
Barcol hardness
Temperature re-
sistance, °F
Flexural strength,
psi
Abrasion resistance,
Taber Wear Index
Resin
2,300
1.6 x 10~5
NRa
180
4,200
NR
Flake glass
2,300
1.5 x 10~5
38
160
6,000
167
Fabric reinforced
6,700
1.5 x 10~5
50
160
10,500
185
NR - Not reported.
-------
Table 4.6-9. TYPICAL CHARACTERISTICS OF POLYESTER RESINS6
Properties
Tensile strength,
psi
Coefficient of
expansion,
Barcol hardness
Temperature re-
sistance, °F
Flexural strength,
psi
Abrasion resistance
Resin
2,300
1.9 x 10~5
NRa
225
4,800
NR
Taber Wear Index 1
Flake glass
2,050
1.5 x 10"5
42
160
6,100
177
Fabric reinforced
6,600
1.5 x 10~5
52
160
12,200
187
NR - Not reported.
-------
Polyester coatings have some excellent characteristics
required of scrubber liners but have had only fair results in
the field. Polyester resins have excellent resistance to acid
and good resistance to heat and abrasion. However, there have
been reported failures of the polyester coating in the scrubbers
and ducts at one utility where the polyester decreased in hard-
ness, lost its adhesive properties, and several blisters
formed.
Vinyl resins have been improved to the point where vinyl
esters have better properties than polyesters. An unreinforced
sprayable vinyl ester resin is reported to be able to withstand
temperatures up to 400°F continuously, to have superior abrasion
resistance, and to resist acids. vinyl ester coalings have been
apP^6^ J? staclls/ ducts' and scrubbers and have adequately
handled the scrubbing system environment. auequauej.y
Epoxy resin coatings have done well in a pilot plant study.
They have less resistance to acids than do other resins, but
adhere to metals better and have a higher tensile strength.
They have good elastic properties. = uj_ cny L.H .
Furan resins have low temperature limit, tensile strength
and abrasion resistance, and are brittle and shrink when applied
over a large area. Furan resins do have a superior resistance
to acids and are very strong when reinforced with a fabric ma?
in a pilot plant test, furan resins did well in Si areas Lceot
immediately above the mist eliminator. except
Precrete, an inexpensive coating, has been used by Kentucky
Utilities and Louisville Gas and Electric Company to prevent
corrosion of stacks ducts, and scrubbers of their FGD sysllms?
Precrete dissolves at a constant rate, so a thick layer can be
applied and its life expectancy can be predicted. However
recoating requires extensive downtimes.
4.6.3.4 Rubber Liners14 —
There are a number of types of rubber liner, but natural
and neoprene rubber liners are most commonly used in a scrubber
or absorber. Natural rubber is softer, more resilient and
has more tear resistance than neoprene rubber; however, neoprene
provides more corrosion resistance and can withstand higher
mr 1^16 4'6-10 Sh°WS S°me ° *
both materials.
characteriscs
Neoprene and natural rubber liners have been tested KV
Bechtel and TVA in the scrubber of an FGD system ? The results
show natural rubber is superior. The natural liner withstood
the design scrubber environment, and there were no sians o?
general corrosion or erosion. Neoprene rubber liners did show
wear from erosion in the area where the flue gases entered the
scrubber. The neoprene liner also formed some blisters after ?
years of operation. *-*-<=*. 0
4.6-15
-------
Table 4.6-10. PROPERTY CHARACTERISTICS FOR NATURAL RUBBER AND NEOPRENE RUBBER9
Property
Hardness range (share
11 A") a
Tensile strength, psi
Maximum elongation/ %
Abrasion resistance
Maximum ambient tempera-
ture allowable °F
Resilience
Aging resistance
Flame resistance
Tear resistance
Natural rubber
40 - 100
4500
900
Excellent
160
Excellent
Good
Poor
Excellent
Neoprene rubber
30 - 90
3500
1000
Very good
225
Very good
Excellent
Good
Good
Indicates values for soft rubber. Values run higher for hard rubber,
-------
Rubber liners do have disadvantages. They are susceptible
to adhesion losses, mechanical damage, wear due to abrasion, and
fire. Overheating can cause adhesion losses and substrate
exposure to the corrosive environment. Rubber liners can be
torn or cut; this may be caused by material in the flue gases
during operation, installation, or when other equipment is
installed or removed. Natural rubber can withstand abrasion
better than neoprene rubber, but neither can withstand the
abrasion in the venturi throat. Rubber liners are not flame
resistant, so extreme care must be exercised when welding near
them.
4.6.3.5 Brick12'14—
There are many types of brick: those most commonly used in
FGD systems are red shale, fire clay, and silicon carbide. Each
of these bricks has limitations that restrict its use. Red
shale should be used where minimum permeation of liquor through
the brick is required and thermal shock is not a factor Fire
clay should be used where thermal shock is a factor and minimum
permeation is not required. Silicon carbide brick should be
used where high abrasion resistance is required.
Red shale brick is a type "L" and fire clay is a type "H»
brick under the "Specification for Chemical Resistant Masonry
Units," ASTM C279. Typical properties meeting type "H" and "L"
bricks are shown in Table 4.6-11.
In the venturi throat, silicon carbide brick in conjunction
with furan resin mortar should prove to be a suitable construc-
tion material. It can withstand the abrasion due to particulate
matter in the flue gases.
Fire clay brick can be used above the mist eliminator and
at the inlet to the absorber. In the absorber inlet, the slurry
from the sprays does not contact the gases, and above the mist
eliminator the mist in the flue gases is minimal; therefore,
fire clay brick with a furan resin mortar is recommended.
In the main section of the absorber, red shale brick could
be used. This section is normally in contact with the slurry
and the temperature of the flue gases is reduced. A furan
resin should be used as the mortar lining.
Corrosion-resistant brick alone will not protect the scrub-
ber shell from corroding. An impervious membrane must be ap-
plied between the scrubber shell and brick. The purpose of the
brick is to protect the membrane from abrasion and excessive
heat. The membranes are made from vinyl resins, natural and
synthetic rubbers, or asphaltic materials.
4.6-17
-------
I
t-1
oo
Table 4.6-11. PROPERTIES OF TYPE "H" AND "L" BRICKS
14
Designation
Type "H"
Type "L"
Minimum modules of
rupture (brick flatwise) ,
psi
Average of
five bricks
1250
2500
Individual
1000
2000
Maximum water absorption
by 2-h boiling test,
%
Average of
five bricks
6.0
1.0
Individual
7.0
1.5
-------
4.6.3.6 Conclusions--
Spool tests and actual field data have shown that T-316L
stainless steel can withstand corrosion and high temperatures
and can handle relatively high chloride and sulfuric acid con-
centrations. It should be noted, however, that stainless steels
are susceptible to stress corrosion in chloride environments.
The T-316L stainless steel is the cheapest metal that can with-
stand the abrasive, acidic environment of the scrubber-absorber
section for the expected life of the lime FGD system.
Precrete is an inexpensive coating that acts as an imper-
meable body. Because of its high corrosion rate, however, a
thick layer of precrete must be applied. Since it is so inex-
pensive, the precrete is still a viable option for use as a
liner. It also has an added benefit in that the life expectancy
of the coating can be predicted. However, relining can take
considerable time. This must be allowed for in the planned
availability of the system, either by reducing demand on the
system or by building in some redundancy.
4.6.3.7 Existing Facilities15"19—
Paddy's Run Station15'l6
Unit for: Boiler 6
Owned by: Louisville Gas and Electric Company
The FGD system at the Paddy's Run plant was supplied by
Combustion Engineering. The scrubber shell is made of mild
steel. An 80-mil coat of Ceilcote 156 (flakeglass) was applied
to the scrubber internals. This lining has eroded in areas of
high abrasion, where the flue gases and where the lime slurry
enter the scrubber. When the Ceilcote lining wears away, it is
patched with the same coating.
Cane Run Station15'l6
Unit for:Boiler 4
Owned by: Louisville Gas and Electric Company
The lime FGD system for Boiler 4 at the Cane Run plant was
supplied by American Air Filter. The scrubber shell is made of
mild steel and was initially coated with a Ceilcote lining.
When the Ceilcote lining failed in the lower portion of the
scrubber, it was replaced with a 2-in. coat of precrete. Pre-
crete was selected as the coating material since it is imper-
meable. Precrete was applied in a thick coat to allow for the
expected high failure rate. Above the mist eliminator to the
top of tne scrubber, the Ceilcote coating was replaced with
Plasite 4005.
4.6-19
-------
Green River Station16'x 7
Units for: Boilers 1, 2, and 3
Owned by: Kentucky Utilities
The Green River FGD system is a variable venturi throat
scrubber. The venturi is built of T-316 stainless steel and
lined with acid brick. The scrubbing module is built of T-316
stainless steel and was initially coated with a carboline liner.
This was replaced when the liner flaked off. This flaking was
caused by improper sandblasting. There was too much material
remaining on the steel for the coating to adhere to it after
sandblasting. In 1977, American Air Filter applied a precrete
liner to the scrubber.
Conesville Station16'l8
Unit for: Boiler 5
Owned by: Columbus and Southern Ohio Company
The FGD system for Boiler 5 consists of two identical
scrubbing modules. The scrubber shell is made of mild steel and
lined with 1/4 in. of neoprene rubber.
A major fire in scrubbing module 5A in December 1976 caused
extensive damage to the internal components of the scrubber; the
neoprene liner was destroyed. The module has since been rebuilt
and was put into service in May 1978. Columbus and Southern
Ohio has had problems with the adhesion of the liner to the
steel due to poor application procedures.
Phillips Station and Elrama Station16'i9
Units for: Two boilers at the Elrama Station
Boilers 1 through 5 at the Phillips Station
Owned by: Duquesne Light and Power Company
The FGD systems at the Phillips Station and Elrama Station
are identical. The venturi scrubbers and absorbers are built of
mild steel. The venturi throat is lined with 316L stainless
steel and no corrosion problems have occurred in this area. The
scrubber was initially coated with the flake glass resin Ceil-
cote 103. The Ceilcote 103 coating did a good job of preventing
corrosion when high calcium lime was used. The high calcium
lime produced a 1/4- to 1/2-in. scale buildup on the walls of
the scrubbers. The scale deposit, in addition to the Ceilcote
103 coating, prevented corrosion. The lime scrubbing system has
switched from high calcium lime to Thiosorbic lime. The new
lime removed the scale deposits and corroded the Ceilcote 103
coating. Since this coating could not withstand the scrubber
environment, Carboline 505 AR coating was applied.
4.6-20
-------
4.6.4 Mechanical Design
4.6.4.1 Introduction--
General characteristics of each scrubber type are discussed
in Section 4.6.1. in this section, we will discuss special
features that are often overlooked during scrubber design
Consideration of these features will help during troubleshooting
and maintenance periods. *
Ease of cleaning out the scrubber/absorber—Depending on
the type of scrubber/absorber, scale/mud deposits occur at
various locations. Deposits are especially a problem at the
wet-dry interface since-they dry out very quickly. This even-
tually leads to plugging. The spray tower is much less vulner-
able to plugging compared with the packed tower or venturi
scrubber. Thus, in the selection of a sizable system, the ease
with which it can be cleaned free of scale is an important
consideration.
Access to scrubber/absorber internals—P^TO-H- cleaning of
lime scrubbers may be required as much as every other month.
Each cleaning with minor maintenance can require several man-
hours . Deposits can occur in and around the throat of the
venturi scrubber, which can result in a higher pressure drop
through the system. Higher pressure drop decreases the amount
of gas that can be scrubbed to such an extent that the genera-
ting capability of the power plant is reduced and a cleaning
outage becomes necessary. Thus, if consideration is given to
easy access when the scrubber is designed, many man-hours can be
saved and prolonged outage can be avoided.
Manholes can be installed at each stage of the scrubber for
easy access during the maintenance period. Similarly side
doors should be located in the reaction tank so maintenance
personnel can get in with jack hammers. Tons of deposits may
have to be removed per maintenance period, hence doors should be
large enough for easy removal of such a quantity of mud.
pump suction line—slurry flow from the scrubber to the
thickener can be directed in two different ways: (1) by instal-
ling a pump to bleed the slurry from the point near the bottom
of the reaction tank, or (2) by taking a slip stream from the
recycled slurry stream. The first option is recommended for
lime scrubbing systems because the pump can be designed to carry
forward big chunks of solids that have settled at the bottom of
the reaction tanks. If the second option were followed, biq
chunks of solids would build up in the reaction tank and cause
operational problems and increase the maintenance time durina
scrubber shutdown. Furthermore, chunks of suspended solids
could easily be carried away to the spray area and clog the
nozzles.
4.6-21
-------
Drain line— In time solids build up in the reaction tank as
in the other portion of the lime scrubbing system. During the
scrubber maintenance period, the reaction tank should be com-
pletely drained and cleaned. It is therefore necessary to
install drain lines to empty the reaction tank.
View plates— View plates can be useful if installed at each
?ha6-5 Scale Formation
lime scrubbing systems can be defined as the
e
of slurry solids to adhere to the surfaces. Scaling
Contbutin5 to poor reliability in the
operain ii
and » ? of full-scale scrubbing systems. Scrubber internals
scalin ellminator surfaces are areas most susceptible to
scalin
^?!! °f the main trouble sP°ts for scaling is the point at
cn the gas passage walls (duct or scrubber) change from a dry
wet condition. Deposited mud tends to dry out at that point
ana eventually becomes very hard, difficult to remove material
•Lne usual remedy is to use a soot blower or wash water lance to
oiow or wash away these deposits into the scrubber. To remedy
scaling in other portions of the scrubber, however, is much more
aitticult and complicated. In these cases the best solution is
«> prevent mud deposits by chemistry control.
Scaling is a very complex phenomenon with many interrelated
iactors affecting it. It is beyond the scope of this book to
give a complete treatise on the. subject of scale formation
This section deals with various types and causes of scale forma-
tion and prevention measures. The impacts of various factors on
scale formation will be discussed concisely.
4.6-22
-------
4.6.5.1 Types of Scale—
Scaling in lime scrubbing systems can be caused by the
reaction products calcium sulfite, calcium sulfate, and calcium
carbonate. Of these, sulfate scaling is normally the most
difficult to control; however, sulfite and carbonate scale must
also be considered in system design.
In a typical lime scrubbing system, a large amount of
slurry enters the scrubber at the top, flows downward in contact
with the gas, passes to a reaction tank (where fresh absorbent
is added), and returns to the scrubber. A bleed stream flows
from the reaction tank to a thickener or waste pond, where the
product solids settle out and the supernatant liquor is returned
to the scrubber system.
in such an arrangement, the slurry circulating through the
scrubber generally contains crystals of both calcium sulfite and
calcium sulfate. Sulfite is formed in the scrubber and goes in
and out of solution as the PH changes. Sulfate, which is not
affected much by PH, forms both in the scrubber and the reaction
tank by oxidation of sulfite and crystallizes whenever the
super saturation gets so high that the solution can no longer
hold it. The crystallization occurs preferably in the reaction
tank, where it ordinarily does no harm, but it can occur in the
scrubber and: either plug gas flow openings or form masses that
eventually drop off and plug the liquor circulation system.
Calcium sulfite scaling—Calcium sulfite (CaSO,-1/2H2O)
scale is a soft, relatively soluble scale that can bl removed
from scrubber internal surfaces by a water jet arrangement.
This scale is formed in the scrubber under certain pH condi-
tions. These conditions are apparent when one considers the
sulfite-bisulfite equilibrium and compares the relative solu-
bilities of the corresponding calcium salts.20 Extremely sol-
uble bisulfite in solution changes to relatively insoluble
sulfite when the solution pH shifts from 4 to 10. When SO, is
absorbed, the scrubber solution is usually between pH 4 and 6-
therefore, the predominant species is bisulfite. Crystalliza-
tion of calcium sulfite occurs when the pH is suddenly raised
either in localized areas or in a reaction tank (Figure 4 6-1)
These calcium sulfite crystals then may attach to surfaces and
form scale.
Calcium sulfate scalincr--Calcium sulfate (CaSO4-2H,O) scale
is a hard, relatively insoluble scale that cannot be removed
from scrubber internal surfaces except by hammering and chip-
ping. This scale is formed in the system as a result of oxida-
tion in the scrubber, reaction tank, and thickener. Unlike
sulfite, pH gradient in the scrubber does not help hold sulfates
in solution.
4.6-23
-------
LIME
FEED"
mm,
M
SCRUBBER/
ABSORBER
REACTION TANK
Cd
^D
_J
(/)
UJ
I
o
o
UJ
OL
BLEED STREAM
PURGE
Figure 4.6-1. Arrangement of the reaction tank with respect
to scrubber/absorber and thickener.
4.6-24
-------
Calcium sulfate will begin to precipitate whenever its
saturation limit is exceeded, or, in other words, whenever the
relative saturation is greater t£an 1.5.20_ The ratio of the
products of the activities of Ca and SO4" to the solubility
product constant (K ) as a measure of the degree of saturation
is termed the relative saturation (RS). For further discussion
of chemical activities, the reader is referred to any standard
thermodynamics text or Perry's Handbook of Chemical Engineering
Fifth Edition (pp. 4-54).
SP (CaS04-2H20)
when:
RS < 1.0 solution is subsaturated;
RS = 1.0 solution is saturated;
RS > 1.0 solution is supersaturated.
Calcium carbonate scaling—Calcium carbonate (CaCO,) scale
is a soft and easy to remove scale. This, together with calcium
sulfite scale accumulates especially downstream from sudden
expansions and along the walls where irrigation is low.
It has been shown in small-scale tests that a high-pH lime
slurry fed to the scrubber can react with CO2 in the gas with
resulting scaling of calcium carbonate on scrubber surfaces
The net situation is obscured by the fact that high pH can also
cause sulfite crystallization. Because of this, it is not clear
whether carbonate scaling is a significant problem or not.
4.6.5.2 Problems Resulting From Scale Formation-
Scale formation can require a scrubber shutdown when
screens, piping, nozzles, packing material, mist eliminator
blades, or liquid distribution grids plug up with so much scale
that pressure differentials increase and flow rate capacities
are reduced. Scale formation can also occur in instruments and
sensor lines such as pH sample taps, pressure differential
sensors, level indicators, pressure gauges, and gas sampling
taps to the extent that the scrubbing system cannot be operated
as a reliable means of control.
Plugging due to scale formation can occur suddenly during
an upset condition or can accumulate and build up over a lonq
interval of operating time. Whenever scrubbers are shut down
for periods longer than a few hours, and the soft accumulations
are not immediately washed away, the soft scale begins to dry
and forms a much harder scale. Scale can cause accelerated
4.6-25
-------
corrosion either by concentrating electrochemical attack beneath
a layer of scale deposited on metallic surfaces or by damaging
protective coatings when scale chunks fall off—thus leading
indirectly to accelerated corrosion at the point of damage.
Even stainless steel material can be severely damaged by stress-
corrosion attack and pitting underneath scale deposits, espe-
cially if the scrubber slurry contains a high concentration of
chloride in solution from chloride in the coal or makeup water.
Scale formation also can significantly influence gas flow
distribution, especially in the mist eliminator area where a
uniform gas distribution is critical for preventing high local
velocities and subsequent carrythrough of solids and liquids.
4.6.5.3 Techniques to Prevent Scale Formation—
The remedy for sulfite scaling is to keep the entering
slurry pH at a level of 9 or less.21 The actual critical level
is not exactly known, and it varies with the type of scrubber.
There are some limitations as to how much the return pH can be
controlled. It should be noted that the elevation of pH in the
reaction tank is due to addition of the makeup lime, a quantity
that cannot be varied very much if it is desired to keep the
addition near the stoichiometric amount. The actual pH depends
on such factors as: S02 content of the inlet gas, L/G ratio,
delay time in the reaction tank, and absorbent feed ratio.
Sulfite oxidation to sulfate can be reduced by covering the
reaction tank, reducing the flue gas oxygen content, and re-
ducing the delay time in the reaction tank. Calcium sulfate
scaling can be minimized by circulating calcium sulfate seed
crystals, which act as nucleation sites forming homogeneous
precipitation of calcium sulfate. Sufficient seed crystal
concentration should be maintained by controlling the percent
solids content of the slurry circulated within the system. The
larger the surface area of the preexisting crystals, the more
the chance of precipitation occurring on the crystals rather
than on the scrubber system internals.
At the lime scrubbing installations in the United States,
practice has varied widely in regard to sulfate crystal concen-
tration. In some cases, where the fly ash makes up part of the
total solids, the upper limit for solids is usually considered
to be about 15 percent because of the increasing viscosity at
higher levels.
For scale prevention, scale inhibiting agents have been
used with some success. A study was conducted by Nalco Chemical
Company for Southern California Edison, regarding scale inhibi-
tors.22 The most effective scale inhibiting agent was an or-
ganic polymer consisting of 52 percent polyolester and 48 per-
cent polyamide dispersant (acrylonitrile), at a 300-ppm dosage
4.6-26
-------
level. Other organic polymers including sodium lignosulfonate
were much less effective. It should be noted that the use of
scale-inhibiting agents tends to reduce the effectiveness of
flocculant materials required in some cases for the proper
operation of thickeners.
The principle involved in this is that some organic com-
pounds attach themselves to the surface of calcium sulfate
lattices and prevent the bonding between crystals of calcium
sulfate. The combined actions of various factors mentioned
earlier have not been fully understood to the extent that the
beneficial effects of scale-inhibiting agents can be predicted.
4.6.5.4 Effects of Various Factors on Scaling—
For scrubber/absorber design it is very important to under-
stand the effects of various factors on scaling. These factors
are discussed below.
1. Recycle of gypsum (CaSO4-2H9O) crystals. Sulfate
scaling can be minimized by circulating gypsum seed
crystals up to about 5 percent by weight. The larger
the surface area of the preexisting crystals, the
lower the scaling will be on the scrubber intervals.
As mentioned earlier, the total solids content should
be approximately 10 to 15 percent.
2. pH; Sulfite scaling can be suppressed by keeping the
pH of the slurry returned to the scrubber to 9 or
less. The actual pH is dependent on other variables.
Hence, there are some limitations as to how much the
return pH can be controlled.
3. Degree of oxidation; Sulfate scale is formed in the
system because of oxidation of calcium sulfite in the
scrubber, reaction tank, and thickener. Therefore,
the lower the oxidation of the sulfite scale, the less
chance of scaling.
4. Degree of loop closing; For the definition of closed-
loop operation, refer to the Material Balance sectidn
in this book. Addition of fresh water reduces the
scaling potential.
5. L/G: The higher the L/C ratio, the lower will be the
scaling potential. The L/G ratio is dependent on the
type of scrubber. However, use of high L/G is a good
way to reduce scaling and achieve high SO2 removal
efficiency. These advantages have to be weighed
against higher pumping costs and the possibility of
flooding at high L/G ratios.
4.6-27
-------
6. Residence time in reaction tank: The solution leaving
the slurry is supersaturated with gypsum even at high
L/G and solids content. Thus, residence time in the
reaction tank should be high enough for the super-
saturation to dissipate, otherwise the sulfate-rich
slurry will pass the critical supersaturation level,
beyond which scaling will occur when it is recycled to
the absorber.
7. Presence of cations such as Mg and Na : The pres-
en^e of hi^gh levels of soluble cations such as Mg ,
Na , and K j.n the scr_ubber slurry reduces the quanti-
ties of SO3 and SO4 available for scale formation.
Significant amounts of soluble cations can be intro-
duced into the system with the reagents or fly ash.
Magnesium in particular is of special interest because
it reduces calcium sulfate supersaturation (by forming
soluble neutral complexes with sulfate ion) and also
promotes S02 removal.
4.6.6 Process Design Variables
Several of the primary process design variables that will
affect SO2 removal capability of the absorber are discussed.
This is not to imply only these factors will control the absorp-
tion of SO2; however, these are major design items in most
installations.
4.6.6.1 Stoichiometry—
Stoichiometry is defined as the number of moles of lime
required to react with 1 mole of SO2 . Theoretically, 1 mole of
CaO will be required to produce 1 mole of Ca(OH)2, which in turn
will remove 1 mole of S02 . Thus, the Stoichiometry of lime FGD
systems is 1 mole of CaO/mole SO2 removed. The actual lime
consumption for most lime-based FGD units is 1.05 to 1.30 mole
of lime per mole of SO2 removed. The higher the lime consump-
tion, the higher the operating cost, because more usable lime
may be lost. A number of Japanese units report lime consump-
tions below 1.0, and it is believed that the Japanese designs
use less excess of lime.
If more restrictive regulations were required for an exist-
ing unit, there are two ways in which somewhat higher SO2 re-
movals could be obtained: by means of a higher lime consump-
tion, and/or by using higher L/G ratio. This approach, which
can give only marginal improvement, depends on the inherent
design of the system with respect to the flexibility of avail-
able equipment.
4.6-28
-------
4.6.6.2 L/G Ratio—
The ratio of lime slurry flow in the absorber to the flue
gas flow, expressed in gal/1000 acf, is termed L/G. For a given
set of system variables, there is a minimum value of L/G that is
required to achieve the desired SO2 absorption. The minimum L/G
can be computed from equilibrium relationships. In practice,
the FGD system must operate with an L/G value more than the
minimum since equilibrium conditions are never achieved.
In moving bed absorbers the upper limit on the value of L/G
is set by the flooding condition, which is an L/G of approxi-
mately 80 gal/1000 acf. Spray towers do not have flooding
problems. However, the power required for pump operation is the
constraint. Normal L/G values vary from 30 to 50 gal/1000 acf
for moving bed absorbers, and from 60 to 80 gal/1000 acf for
spray towers.
The gas velocity through the absorber should allow a cer-
tain residence time for the gas stream. In U.S. FGD systems,
this ranges from 3 to 9 seconds. This factor should be con-
sidered when computing the operating L/G. Other major variables
that have an impact on gas-liquid interface conditions are the
type, size, and total height of the packings used to induce
turbulence in the moving bed absorbers.
4.6.6.3 pH—
As the lime slurry enters the S02 absorber, the pH often
ranges from 7.5 to 8.5. When the absorbent reacts with the SO2,
the pH of the slurry becomes more acidic. The pH of the slurry
as it exits the absorber may range from 4.5 to 6.0.
Johnson6 discovered in 1935 that the equilibrium vapor
pressure of SO2 over lime slurry is inversely proportional to
the slurry pH, resulting in a lower SO2 equilibrium vapor pres-
sure at a higher (more alkaline) pH. Test work recently con-
ducted at TVA's Shawnee test facility23 demonstrated greater SO2
removal at higher slurry pH with constant L/G ratios. The
limitation to this approach is that the excess lime required for
this operating mode increases the cost of operation, and the
tendency toward scale formation.
The pH used as the desired absorbent slurry control point
depends on the L/G ratio in the absorber, the inlet SO2 concen-
tration, and the required SO2 removal. In general, however, a
pH range of 8.0 to 8.5 may be expected.
As the absorbent is utilized, the pH of the slurry is
affected by the conversion of calcium hydroxide to calcium
sulfite/sulfate. The absorption of SO2 makes the resulting
slurry pH less alkaline. As noted, the exiting slurry pH may
4.6-29
-------
range from 4.5 to 6.0. As the pH of the slurry becomes more
acidic, the SO2 absorbing properties of the lime slurry are
reduced. Therefore it may be seen that, as the S02-rich flue
gas contacts the lime slurry, the rate of SO2 absorption in-
creases as the flue gas stream encounters more fresh absorbent.
Early lime FGD systems operated in the supersaturation
range of calcium sulfate without there being any awareness of
the necessity of controlling this variable. When massive depos-
its of hard sulfate scale and also some softer sulfite scale
were discovered, steps were taken to learn to control this
problem. One of the four basic mechanisms of scale formation is
pH excursion. Scale formation problems have been less pro-
nounced with lime absorbent systems, when compared with lime-
stone absorbent systems, because the pH is often maintained at a
higher (more alkaline) level.
The degree of calcium sulfite oxidation to calcium sulfate
is reportedly reduced when the pH level is increased. The
tendency to form high levels of dissolved calcium and sulfate
ions in the slurry is thus suppressed using lime reagent at
higher pH control levels. Whenever a pH control excursion
occurs, and the pH of the absorber slurry drops below 4.0 to
4.5, regardless of the type of reagent (lime or limestone),
severe and rapid formation of calcium sulfate scale can occur.
As scale forms in the absorber, concentration gradients
and/or differential aeration cells are established between the
particles trapped beneath the deposit and those outside. The
natural corrosive characteristics of the more acidic calcium
sulfite and sulfate tend to attack the absorber exposed surface.
The combination of the natural corrosiveness of the medium and
the concentration cells resulted in severe corrosion of early
installations.
4.6.6.4 Increased Gas Velocity—
The most common flue gas velocities encountered in absorber
design range from 7 to 10 ft/s. The EPRI is currently testing a
cocurrent scrubber at the Shawnee test facility. Gas velocity
up to 30 ft/s will be tested. Tray towers often are in the
range of 7 to 8 ft/s, and the velocity in TCA, spray tower, and
venturi absorbers most often ranges from 8 to 10 ft/s.
4.6.6.5 Liquid Distribution—
For best SO2 absorption, the intimate contact between the
SO2 molecules in the flue gas and the droplets of lime-based
absorbent is critical. The smaller the droplet of absorbent
slurry, the better the contact and the absorption. To achieve
this contact in spray towers, finer spray nozzles are employed.
In some cases, high pressure pumps are used to obtain finer
4.6-30
-------
atomization of the slurry. In other cases, impinger plates are
used to physically reduce the size of droplets exiting the
nozzle. In a venturi, the pressure drop across its throat
causes the liquid droplet to break up into many finer droplets.
The venturi principle is also used in packed- and moving-bed
absorbers. As the gas flow is forced between the spheres in the
bed, many small venturi effects occur. The wetted sides of the
spheres serve as an area of mass transfer.
In a spray tower, the droplet size must be controlled by
nozzle type, line pressure, and use of impinger plates. In a
mobile-bed absorber, the size droplet is not so critical since
the action of the gas flowing up through the packing or balls
causes the breakdown of the droplet size.
4.6.6.6 Water Balance--
Three external factors will impact the water balance of an
FGD system: the ambient humidity, the rainfall of the area, and
the climate. These three items will determine how the water
lost in the adiabatic cooling of the flue gas is replaced to
maintain a closed system. For greater detail of water balance
please read EPRI report24 FP 627 entitled "Lime/Limestone Scrub-
ber Operation and Control Study."
As the gas stream is cooled, water is absorbed by the gas.
This is the primary point of water loss throughout the FGD
system.
When one considers individual FGD plant sites from a design
standpoint, specific climatic conditions should be included to
avoid unanticipated problems that might have a major impact on
operations. The average temperature, wind velocity, precipita-
tion, and other items should be considered to determine their
possible effects on the system-wide water balance.
The quantity of the water used at the various points
throughout the FGD system is an important consideration. It is
desirable that the FGD system operate in the closed-loop mode,
i>e., makeup water should not exceed that lost in the flue gas
and the sludge. Some major uses of water in FGD systems are as
foll°ws:
1. For wet particulate scrubber makeup
2. For slaking of lime
3. For dilution of the lime slurry and/or additional
makeup for the SO2 absorber
4.6-31
-------
4. As mist eliminator wash
5. As pump seal water
All five of these use points may impact the scrubber/absorber
operation. All the above uses are quantified for six specific
scrubbers in EPRI report25 FP 627 entitled "Lime/Limestone
Scrubber Operation and Control Study."
Water used in the wet particulate scrubber is not required
to be of the best quality. Often recycled water from the sludge
pond is used.
The water required for slaking the lime is much more criti-
cal. It is desirable to use water of (or near) potable quality.
Waste or recycle process waters containing sulfites and sulfates
retard the slaking process—not only is more time needed to
complete the slaking step, but the quality of the resulting lime
slurry is impaired. The lime particle size increases and the
surface area shrinks, which in turn retards. In fact, some of
the lime does not hydrate at all and is wasted. The only ex-
planation is that the lime precipitates the S03 and SO4 ions as
calcium sulfite-sulfate, which coats the unreacted CaO particles
and prevents complete water penetration into the lime particles.
Once the lime has been slaked, however, recycled or waste
process water can be used to dilute the thick lime slurry to the
desired consistency. There is little or no effect by the SO3
and S04 ions on the quality of the diluted lime slurry produced.
The chloride ion in dilution water in reasonable amounts appears
at present to exert a minimal deleterious effect on the result-
ing slurry.
Mist eliminator wash may be freshwater, recycled water, or
any combination of fresh and recycled water. Ideally, all
freshwater would be used; however, to attain closed-loop opera-
tion, a mix is frequently required. Continuous wash is often
accomplished using recycled water. Freshwater normally is used
for high volume, intermittent wash. Since the trend in con-
struction of mist eliminators is toward the zigzag baffle or the
continuous chevron type of mist eliminator made of plastic mate-
rials, and because the mist eliminators are largely corrosion
resistant, the chemical makeup of the water normally has little
effect except for the possible buildup of scale when sulfate
saturated water is used.
Pump seal water is most often fresh water. Concern about
the chemical constituents, suspended solids, and sulfate satura-
tion of the recycled pond water dictates that recycled water be
used only with great caution. A major concern is that in the
presence of the heat associated with pump operation, the sulfate
and other solids in the recycled pond water would deposit and
cause pump failure.
4.6-32
-------
The sources of makeup water are as follows:
1. Freshwater
2. Sludge pond recycle water
3. Wastewater from other plant processes
4. Cooling tower blowdown
The nature of the cooling tower chemical treatment, impurities
that may become concentrated in the feed water stream and
particulates that may be washed from the air by cooling 'tower
flow, all must be considered in the decision as to whether to
utilize this water stream.
4.6.6.7 Interfacial Area—
The interfacial area may be defined as that area in which
the absorbent slurry contacts the flue gas stream. This will be
affected by the L/G ratio, the gas velocity, the slurry droplet
size, liquid distribution, and the type of absorber An ade-
quate contact area is required for the desired SO2 removal from
the flue gas stream. The impacts of the L/G ratio the eras
velocity, and the liquid distribution have been discussed as
they affect SO2 removal design for an absorber.
In a TCA, the size of balls or marbles used and the depth
of the contact bed are the two critical items. In a sprav
tower, the height (length) of the tower, droplet size, nozzle
pressure drop, spacing of sprays, and coalescence of droplets
are the critical design points. The height, at a given gas
velocity, gives the residence or contact time when SO, may be
removed from the gas stream. Internal sources of gas turbulence
such as grids must be considered in absorber design.
4.6.7 Existing Facilities
Table 4.6-12 indicates the operating experience of existing
lime FGD systems in terms of months of operation. A summary of
the performance of these systems is presented below.
4.6.7.1 Louisville Gas and Electric, Cane Run Unit 4**—
Following startup of Cane Run Unit 4, 75 to 80 percent of
the S02 was removed at full load. To increase the SO* removal
a new spray header system was installed above the mobile bed
sprayer to improve the L/G ratio. This resulted in superior
contact and an S02 removal rate above 85 percent. The sprav
nozzles were changed from plastic to ceramic to resist crackina
caused by expansion. The scrubber is now running consistently
above 90 percent S02 removal. This system has never experienced
scaling or plugging problems. *periencea
4.6-33
-------
Table 4.6-12. SUMMARY OF OPERATING LIME FGD SYSTEMS AS OF JANUARY 1978
a\
UJ
Utility name
Pennsylvania Power
Pennsylvania Power
Louisville Gas t Electric
Columbus t Southern
Duquesne Light
Kentucky Utilities
Kansas City Power t Light
Kansas City Power t Light
Louisville Gas t Electric
Duquesne Light
Montana Power
Montana Power
Minnkota Power Co-Op.
TVA
TVA
Process/generating units
Lime scrubbing
Bruce Mansfield No. 1
Bruce Mansfield No. 2
Cane Run No. 4
Conesville No. 5
Elrama Power Station
Green River Nos. 1, 2. and 3
Hawthorn No. 3
Hawthorn No. 4
Paddy's Run No. 6
Phillips Power Station
FGD/MW
825
825
178
400
S10
64
140
100
65
410
3517
Line/alkaline fly-ash scrubbing
Colstrip No. 1
Colstrip No. 2
Milton R. Voung No. 2
Lime/limestone scrubbing
Shawnee No. IDA
Shawnee No. 10B
360
360
450
1170
10
10
20
Startup
4-76
7-77
8-76
1-77
10-75
9-75
11-72
8-72
4-73
7-73
11-75
7-76
9-77
4-72
Experience, mo
21
£
17
12
27
28
£2
65
57
54
349
26
18
4
48
69
69
138
-------
4.6.7.2 Louisville Gas & Electric, Paddy's Run Unit 627—
Initial startup of Paddy's Run Unit 6 took place on April
5, 1973. A 7-hour shutdown was required when a marble bed
support plate broke, and the malfunction and repair of the dual
strainer switch in the bottom of the scrubber module caused two
more outages. During the beginning of 1976, the scrubber
achieved 99 percent SO2 removal. Tests run using calcitic lime,
instead of the usual carbide lime, resulted in scaling from the
increased oxidation level of the calcitic lime. When magnesium
hydroxide [Mg(OH)2] was added to the lime, this problem was
eliminated.
4.6.7.3 Kentucky Utilities, Green River Power Station28'29—
Serious plugging problems were observed at Green River
following startup on September 13, 1975. Hard gypsum scale
plugged the lower mobile beds, and the spray nozzles also exper-
ienced plugging problems. To remedy this, the oxygen content of
the flue gas was reduced by minimizing air leakage into the
system. Thus, the oxidation of sulfite to sulfate was pre-
vented, and the pH was lowered enough to prevent the precipi-
tation of gypsum. The scrubber balls were also replaced with
larger ones to reduce migration. To eliminate pitting that
occurred behind the Carboline stack liner, the liner was re-
placed with Precrete G-8 and metal backup plates were welded to
the pitted portions of the stack.
4.6.7.4 Duquesne Light, Phillips27'30'31—
Partial startup at Phillips station occurred July 1973, and
full startup took place on March 17, 1975. High calcium lime
caused deposit buildup around the throat dampers and lower cone;
deposits also formed in and around the spray nozzles. This
problem was partially alleviated by closing alternate nozzles,
thus producing higher velocities in the other nozzles. Tests
indicated that using lime with higher MgO content also reduced
the accumulation of scale. In one test the use of 8 to 10
percent MgO lime almost totally eliminated deposits and resulted
in an increased SO2 removal rate of 83 percent. During Sep-
tember 1977, when the system was running at higher capacity, the
SO2 removal rate dropped to 50 percent.
4.6.7-5 Duquesne Light Elrama27—
The main problem encountered following startup at the
Elrama station was a poor SO2 removal rate. In addition, there
were problems with a bleed valve leak. Dravo Thiosorbic lime is
now used, as it is at Phillips station, for increased S02 re-
moval .
4.6-35
-------
4.6.7.6 Columbus and Southern Ohio Electric, Conesville Unit
52 7 __
Because the lining of scrubbing Unit 5A was destroyed by
fire prior to startup at Conesville Unit 5, the scrubber had to
be relined before it was put into service. Scrubbing Unit 5B
started operating on February 13, 1977, using Dravo Thiosorbic
lime with an MgO content of 3 to 8 percent.
Problems encountered were a carryover of scrubbing liquid
into the mist eliminator and poor velocity distribution through
TCA beds. When the flow rate was low, some plugging occurred.
Scaling and buildup of deposits inside the scrubbers continue to
be a problem.
4.6.7.7 Pennsylvania Power, Bruce Mansfield Units 1 and
22 7 / 32 '33__
Both systems at Bruce Mansfield have experienced problems
with corrosion, scale, and stack liner failures. It was hoped
both systems could remove the required amount of SO2 using five
of the six scrubbing trains, but all six were needed because the
flue gas flow was greater than expected. Scale formation,
plugging, and acid corrosion resulted when the pH of the recir-
culating slurry was controlled manually (because of poor auto-
matic pH control). When the pH monitors were relocated, how-
ever, this problem was eliminated.
4.6-36
-------
REFERENCES
1. Treybal, R.E., Mass Transfer Operations, 2nd ed. McGraw-
Hill, New York, 1968.
2. Perry, R.H., and C.H. Chillon. Chemical Engineer's Hand-
book, 5th ed. McGraw-Hill, New York, 1978.
3. Crow, G.L. Corrosion Tests Conducted in Prototype Scrubber
Systems. In: Corrosion Problems in Air Pollution Control
Equipment Symposium, Atlanta, 1978.
4. Kopecki, E.S., and C.E. McDaniel. Corrosion Minimized/
Efficiency Enhanced in Wet Limestone Scrubbing, Power Eng.
80, No. 4, April 1976.
5. corrosion Properties of an SO2-Wet-Limestone Scrubbing
System. International Corporation Forum, Toronto, Canada,
April 1975.
6. Atlas Minerals and Chemicals Division Glass Flake Systems
Bulletins 4-1100, 4-1200, 4-1300, and 4-1400; Fabric Rein-
forced Systems Bulletins 4-2100, 4-2200, 4-2300, and 4-
2400.
7. Lewis, E.C.,'M.P. Stengel, and P.G. Maurin. Performance of
TP-316L SS and Other Materials in Electric Utility Flue Gas
Wet Scrubbers. In: Corrosion Problems in Air Pollution
Control Equipment Symposium, Atlanta, 1978.
8. corrosioneering, Inc. Specifications/Information Corrosion
Resistant Linings and Coatings. B-100.
9. Heil Process Equipment Company. Regiline Corrosion-Resis-
tant Coatings and Linings for the Power Industry, BM-403.
10. Dudick Corrosion Proof, Inc. Application Guide, Tech Data
Sheets.
Johnson, R.S., Jr. Materials Performance in a Flue Gas
Particulate and Desulfurization System. In: Corrosion
problems in Air Pollution Control Equipment Symposium,
Atlanta, 1978.
4.6-37
-------
12. Boova, A.A. Chemical Resistant Masonary, Flake and Fabric
Reinforced Linings for Pollution Control Equipment. In:
Corrosion Problems in Air Pollution Control Equipment
Symposium, Atlanta, 1978.
13. Singleton, W.T., Jr. Protective Coatings Formulated From
Vinyl Ester Resins for the Air Pollution Control Industry.
In: Corrosion problems in Air Pollution Control Equipment
Symposium, Atlanta, 1978.
14. Fontana, M.G., and N.D. Greene. Corrosion Engineering.
McGraw-Hill Book Company, New York, 1967. pp. 157-193.
15. Private communication with R. Vanness, Louisville Gas and
Electric, February 1978.
16. Laseke, B., Jr. EPA Utility FGD Survey, December 1977 to
January 1978, and April to May 1978 update.
17. Private communication with J. Beard and V. Anderson, Ken-
tucky Utilities, February 1978.
18. Private communication with D. Boston, Columbus and Southern
Ohio Electric, February 1978.
19. Private communication with R. O'Hara and J. Mahone,
Duquesne Light Company, February 1978.
20. Gogineni, M.R., and P.A. Maurin. Sulfur Oxides Removal by
Wet Scrubbing - Application to Utility Boilers. Presented
at Frontiers of Power Technology Conference, Stillwater,
Oklahoma, October 1-2, 1975.
21. Uchida S., C.Y. Wen, and W.J. McMichael. Role of Holding
Tank in Lime and Limestone Slurry Sulfur Dioxide Scrubbing.
Ind. Eng. Chem., Process Des. Dev. , Vol. 15, No. 1, 1976.
22. Shah, I.S. Paper Presented at Second International Sympo-
sium for Lime/Limestone Wet Scrubbing, New Orleans, La.,
November 8-12, 1971, (Proceedings issued by EPA, APTD-1161.
1, 345.)
23. Borgwardt, R.H. Pilot studies related to Unsaturated
Operation of Lime and Limestone Scrubbers. Presented at
Symposium on Flue Gas Desulfurization - Atlanta, November
1974. (Proceedings issued by EPA, EPA-650/2-74-126-a.)
24. Lime/Limestone Scrubber Operation and Control Study. EPRI
report FP 627.
25. EPRI report on water balance for 6 systems.
4.6-38
-------
26. Van Ness, R.P. Louisville Gas and Electric Company Scrub-
ber Experiences and Plans. Paper presented at FGD Sympo-
sium, Hollywood, Florida, November 8-11, 1977.
27. Laseke, B.A., Environmental Protection Agency Utility FGD
Survey: December 1977 - January 1978. EPA-600/7-78-051a
March 1975.
28. Beard, J.B. Scrubber Experience at the Kentucky Utilities
Company Green River Power Station. Paper presented at FGD
Symposium, Hollywood, Florida, November 8-11, 1977.
29. Laseke, B.A. Survey of Flue Gas Desulfurization Systems:
Green River Station, Kentucky Utilities. EPA-600/7-78-
048e. March 1978.
30. Nelson, R.L., and R.E. O'Hara. Operating Experiences at
the Phillips and Elrama Flue Gas Desulfurization Facili-
ties. Paper presented at the Second Pacific Chemical
Engineering Congress, Denver, August 28-31, 1977.
31. Knight, R.G., and S.L. Pernick. Duquesne Light Company
Elrama and Phillips Power Stations Lime Scrubbing Facili-
tl6S1l ,J.??er Presented at the FGD Symposium, New Orleans,
March 1976.
32. Laseke B.A. Survey of Flue Gas Desulfurization Systems:
Bruce Mansfield Station Pennsylvania Power Company (Draft).
EPA™ D o ~ (}£ ""£ DU j *
33. Private communication with R. Forsythe and W. Norrocks,
Pennsylvania Power Company, February 1978.
4.6-39
-------
CONTENTS
4.7 MIST ELIMINATORS
4.7.1 Introduction
4.7.2 Description and Function
4.7.3 Types of Mist Eliminator
4.7.3.1 Continuous or Discontinuous Chevron Baffle
4.7.3.2 Chevron Mist Eliminator Design and Per-
formance 4.7-3
4.7.3.3 Radial Vane 4.7-10
4.7.4 Mist Eliminator Design Factors 4.7-13
4.7.4.1 Mist Eliminator Shape 4.7-13
4.7.4.2 Number of Passes 4.7-14
4.7.4.3 Spacing Between Vanes 4.7-14
4.7.4.4 Slanted Mist Eliminator 4.7-14
4.7.4.5 Materials of Construction 4.7-18
4.7.4.6 Special Drainage Devices 4.7-18
4.7.5 Mist Eliminator Equipment Design Considerations 4.7-18
4.7.5.1 Mist Eliminator Configuration (Horizontal
vs. Vertical Gas Flow) 4.7-20
4.7.5.2 Use of Bulk Separation, Wash Tray, and
Knock-out Devices 4.7-22
4.7.5.3 Freeboard Distance 4.7-22
4.7.5.4 Number of Stages 4.7-26
4.7.5.5 Distance Between Stages 4.7-26
4.7.5.6 Passage Geometry 4.7-26
4.7.5.7 Miscellaneous 4.7_26
4 7.6 Mist Eliminator Wash System 4.7_27
4.7.6.1 Wash Water Type 4.7-27
4.7.6.2 Wash Direction 4.7-28
4.7.6.3 Wash Duration and Water Quantity 4.7-28
4.7.6.4 Wash Water Pressure 4.7-28
4.7-i
-------
CONTENTS (continued)
4.7.7 General Factors 4.7-29
4.7.7.1 Mist Eliminator Pressure Drop 4.7-29
4.7.7.2 Overall Collection Efficiency 4.7-29
4.7.7.3 Mist Loading and Particle-size Distribution 4.7-30
4.7.8 Existing Facilities (Experience and Equipment
Detail) 4.7-30
4.7.8.1 Phillips Station (Duquesne Light Company) 4.7-36
4.7.8.2 Paddy's Run 6 (Louisville Gas and Electric
Company) 4.7-36
4.7.8.3 Elrama Station (Duquesne Light Company) 4.7-36
4.7.8.4 Green River Station (Kentucky Utilities) 4.7-36
4.7.8.5 Bruce Mansfield (Pennsylvania Power Company) 4.7-37
4.7.8.6 Conesville (Columbus and Southern Ohio
Electric Company) 4.7-33
4.7.8.7 Cane Run (Louisville Gas and Electric
Company) 4.7-38
4.7.8.8 Hawthorn (Kansas City Power and Light
Company) 4.7-38
4.7.8.9 Colstrip (Montana Power Company) 4.7-38
4.7.9 Recommendations 4.7-38
References 4.7-40
4.7-ii
-------
4.7 MIST ELIMINATORS
4.7.1 Introduction
In any wet scrubbing system, small drops of liquid are
formed and carried out of the scrubber with the gas. A well-
designed mist elimination device is therefore necessary to
prevent plume rain and mist entrainment that would cause corro-
sion and scaling of downstream equipment. In addition, if a gas
reheater is required to evaporate the droplets, an efficient
mist eliminator can substantially reduce the reheating cost by
minimizing the amount of moisture that must be evaporated before
a temperature rise in the flue gas is obtained.
4.7.2 Description and Function
A mist eliminator is defined in Guidelines for the Design
of Mist Eliminators for Lime/Limestone Scrubbing Systems1 as:
"A device employed to collect, remove, and return to the scrub-
bing liquor the slurry droplets entrained with the desulfurized
flue gas exiting the scrubber or absorber."
The most common device is a set of baffles or slats set in
such a way as to impart a zigzag flow to the gas over distances
ranging from a few inches to a foot. Mist drops are removed by
impaction on surfaces that change the direction of gas flow.
Cyclonic flow, which causes the entrained moisture to impact on
ductwork surfaces, is also effective in mist removal.
The mist drops generally contain both suspended and dis-
solved solids. The suspended solids are derived from particu-
lates collected by the scrubber, lime particles introduced into
the scrubbing liquid, and/or products of chemical reactions
occurring within the scrubber. Similarly, dissolved solids come
from impurities in the gas, lime introduced into the scrubber
liquid, and/or products of reaction.
Mist carryover can cause a variety of problems, both within
the air pollution control system and in the ambient atmosphere
In cases where an induced-draft fan is used, drops can collect
in the fan. These drops tend to deposit solids causing failure
of the blades, housing, or supporting structures as a result of
excessive vibration or corrosion problems. Solids can also be
deposited in the reheater, ductwork, and stack; and as was the
case at Bruce Mansfield (Pennsylvania Power), the deposits can
break off in chunks and be blown out of the stack. Reheater
plugging and corrosion is a very common experience at lime
scrubbing installations, and problems with the reheaters can
usually be traced directly to inefficient mist eliminators.
Measures taken to reduce reheater plugging include use of effi-
cient mist eliminators upstream of the reheaters and use of soot
4.7-1
-------
blowers in the reheater housing. Entrained mist drops that
reach the stack can cause problems in the surrounding area as a
result of "rainout" of liquid drops.
The major mist eliminator problems encountered in lime
scrubber applications are plugging, scaling, and reentrainment/
carryover problems in downstream equipment. A soft, mudlike
deposit and/or scale can accumulate on the mist eliminator in
the course of time, unless it is sufficiently sprayed with wash
water of reliable quality. If solids build up to the point
where the collector is completely blocked and "blow holes"
develop, the result will be increased pressure drop across the
mist eliminator, increased wear and erosion in the blow-hole
areas, and drastic reduction in overall efficiency.
Solid deposits of calcium sulfite and unreacted lime can
occur in lime systems when the solid carryover from the scrubber
is trapped in the mist eliminator. More serious, however, is
formation of sulfate scale, which results from oxidation of the
sulfite solution collected on the mist eliminators. Scaling can
also occur as the result of absorption of residual flue gas SO2
by the unreacted lime on wetted surfaces. High stoichiometric
ratios of lime (poor lime utilization) compound the SO2 absorp-
tion problem, since larger quantities of unreacted lime are then
carried over to the mist eliminator. Surface irregularities
formed by the crystalline scale increase the potential for mud
accumulation and decrease the effectiveness of washing opera-
tions .
The failure of mechanical parts is another cause of break-
down. In some cases collector blades, especially fiberglass
ones, can in time become embrittled. Entrained solids combined
with forces of high-pressure wash-water sprays can deform,
shatter, or break the blades. Partial plugging of the mist
eliminator can also increase pressure drop, which in turn some-
times causes the blades to collapse. In early systems, stress
corrosion cracking occurred in mist eliminators constructed from
316L stainless steel. Newer systems, such as the now-terminated
Mohave unit, use Incoloy 825.
In installations where the scrubber exit gas is reheated, a
high-efficiency mist eliminator is a very important part of the
scrubbing system. The reheat energy requirement increases as
mist carryover increases. Increasing mist carryover also leads
to the collection of entrained substances on the heat exchange
surfaces of the reheater, eventually causing plugging and/or
corrosion. A high-efficiency mist eliminator is also important,
however, even when the gas is not reheated. It is usually
required to eliminate mist carryover through the stack.
4.7-2
-------
4.7.3 Types of Mist Eliminator
A number of designs are available to remove liquid and
solid particulates from gas streams, including the wire mesh,
tube bank, gull wing, and electrostatic precipitator (ESP)
types. For lime scrubbing operations, however, the two most
relatively successful designs in use are the chevron baffle and
the radial vane. These two types alone are discussed in this
report.
4.7.3.1 Continuous or Discontinuous Chevron Baffle—
The chevron baffle can consist of either continuous or
discontinuous zigzag baffles (Figure 4.7-1). The baffle uses
the inertial impaction collection mechanism, whereby the gas
stream with its entrained liquid droplets is forced to make
abrupt changes in direction. When the stream changes direction,
droplets impinge on the baffle walls, coalesce, and drain from
the mist eliminator blade (Figure 4.7-2). In Figure 4.7-2(a)
the chevron is positioned horizontally (vertical gas flow);
hence drops fall as shown into the scrubbing system. If the
chevron were positioned vertically (horizontal gas flow), drops
would fall vertically along the mist eliminator (Figure
4.7-2(b)). This configuration allows wash water to be easily
isolated from the scrubber system.
Although the chevron mist eliminator is simple, its collec-
tion efficiency when dealing with moderate to large droplets is
excellent. Its low pressure drop and wide-open construction
make it a popular choice in lime scrubbing operations, where the
high solids content of the slurry would readily cause plugging
in other eliminator types.
4.7.3.2 Chevron Mist Eliminator Design and Performance—
Heil chevron mist eliminator2 '3— The Heil design is shown
in Figure 4.7-3 (vertical configuration); its removal efficiency
curve is shown in Figure 4.7-4. There are no holes in the
blades for mounting in a Heil assembly and consequently bypass
leaks through hole clearances are avoided. Heil blades can be
used in the vertical or horizontal position. Their assemblies
and modules come in standard sizes, but can also be custom
fabricated.
Matsuzaka mist eliminator (Japan)4—The Matsuzaka Co.
manufactures and markets the Humboldt Wedag "Lamellar" separator
(Figure 4.7-5). Because of its higher cost, it can compete with
conventional chevron baffles only when superior performance is
required (for droplets of less than 30 Kim). One of its advan-
tages over chevrons is that maldistribution of gas has a much
less adverse effect on collection efficiency because its troughs
prevent reentrainment when installed at an angle or in a hori-
zontal duct, even if a disproportionately large amount of mist
4.7-3
-------
\\NN\\\\
III
t
GAS DIRECTION
Figure 4.7-l(a). Discontinuous horizontal chevron
zigzag baffle.
\
t
GAS DIRECTION
Figure 4.7-l(b). Continuous horizontal chevron
zigzag baffle.
4.7-4
-------
(a) HORIZONTAL CONFIGURATION
DROPLET
IMPINGEMENT
AND COALESCENCE
LARGE FALLING
DROPLETS
MIST-
LADEN
GASES
MIST
LADEN GASES
(b) VERTICAL CONFIGURATION
SIDE
DROPLET IMPINGEMENT
AND COALESCENCE
TOP
LARGE DROPLETS
FALL
Figure 4.7-2. The chevron impingement principled
4.7-5
-------
GAS
DIRECTION
Figure 4.7-3. Heil chevron mist eliminator.2
(Top view of ductwork)
Courtesy: Heil Process Equipment Co.
4.7-6
-------
GAS VELOCITY =
4.83 ft/s
GAS VELOCITY =10.1 ft/s
"~GAS VELOCITY = 12.8 ft/s
GAS VELOCITY =19.4 ft/s
REMOVAL EFFICIENCY
HEIL THREE-BEND PLASTIC BLADES
AT A 1.25-In. SEPARATION
AUGUST 22, 1976
C.F.H. RESEARCH LABORATORIES
PROJECT 905
FRANK W. HOFFMAN
95
45 55 65
DROPLET DIAMETER , ym
Figure 4.7-4. Heil chevron mist Gliirinator performance.'
4.7-7
-------
LAMELLAR
ELEMENT X
DUCT-
NEGATIVE
PRESSURE £ONE
r© POSITIVE
, m PRESSURE ZONE
LIQUIDJI
FGAS + LIQUID
WORKING PRINCIPLE OF HUMBOLDT LAMELLAR
SEPARATOR (TRANSECTION OF SEPARATOR)
Q.
O
o;
o
oo
oo
CCL
a.
0.60
0.40
0.30
0.20
0.15
0.10
0.080
0.060
0.050
I r
I i r
300 400 600800 1000
GAS FACE VELOCITY, ft/niln
Figure 4.7-5.
Vertical configuration of Humboldt Lamellar separator
in horizontal duct (looking downward) marketed by Matsuzaka.^
4.7-8
-------
is collected at one point on the cross section. For this rea-
son, a design with a collection trough is more likely to perform
as expected, since uniform gas distribution is seldom attained
with any mist eliminator.
A main problem in lime scrubbing application is clogging at
the bottom of the trough. In the early installations, the first
row, or "bank," remained clear while the second one plugged. To
alleviate this problem, the banks were separated and a spray
installed between them. The baffles in the first bank were also
set 2 in. apart instead of 1-1/4 in., and the troughs were made
larger.
The present "standard" design consists of three banks with
a continuous wash on the first, a 30-s intermittent wash on the
second, and no wash on the third. Fresh water is used on the
first, bank, but recycled liquor can be used on the others.
About two-thirds of the wash liquid is applied on the top of
each bank. The vertical trough is mounted with a considerable
slope, and a special spray nozzle keeps it clean.
The usual gas velocity is 23 ft/s; pressure drop through
the three banks is approximately 2 in. H2O. The expected per-
formance is for 99 percent removal of the 15- to 20-pm drops.
NGK mist eliminator (Japan)—This design is licensed from
Euroform (Aachen, Germany) and is similar to that offered by the
Heil Company. The configuration is a shallow S-curve with
relatively small "hooks" attached to the surfaces (similar to
Figure 4.7-3).
The NGK eliminator has had trouble with deposits in the
vertical "pocket" channels. The preferred nominal velocity
range at the mist eliminator for lime scrubbing application is
19.5 to 26 ft/s. An NGK mist eliminator in the horizontal
position (vertical gas flow) removes drops as small as 30 ym,
and in the vertical position (horizontal gas flow) it removes
drops down to 15 pm. In the vertical position, it can accept
higher inlet loadings without reentrainment. Practically all
units supplied by NGK are of the vertical type. The company
puts considerable emphasis on turning vanes to achieve uniform
gas distribution over the mist eliminator cross section. Pres-
sure drop through the mist eliminator is approximately 1 in.
H o. Its efficiency guarantee is usually based on outlet solids
loading.
Munters Euroform mist eliminator5— Various designs are
available for this mist eliminator, which is made under license
from NGK (Japan). The main difference between them is in blade
spacing. The various models available include the T-8, T-
271(K), T-271(M), T-71, TS-5/2, and T-100. Letters or numbers
referring to these models do not have a specific meaning; how-
ever, each one has a particular characteristic and application.
4.7-9
-------
Model T-8 is a coarse separator for vertical gas flow. Model
1-271 is used in applications requiring fine droplet removal.
It, too, is used with vertical gas flow. Letters "K" and "M"
after T-271 indicate, respectively, plastic and metal materials
of construction. Model T-71 is a half-sized version of T-271.
Model TS-5/2 is employed with horizontal gas flow. Models in
the T-100 series are made of various plastics. The "20" in
Model No. T-120 indicates that it has 20-mm spacing. Similarly,
T-125 has 25-mm spacing.
Figure 4.7-6 shows the configuration of a section and the
pressure drop curve for the T-271 type. The wash procedure with
this design involves a fresh water wash at the upper eliminator,
with preceding washes using recycled liquor. In the case of the
vertical flow eliminators, wash rates range from 0.5 to 0.75
gal/min per ft2 of eliminator surface using a coarse, full-cone
spray at 35 to 40 psi and for a period of 6 to 12 min/h. These
rates prevent buildup on nozzles and surfaces. The wash water
can be reused elsewhere in the process. This occurs only in
horizontal gas flow mist eliminators where the drainage is
collected.
Sprays are normally applied to both the upper and lower
surfaces in the primary eliminator and can be used on the bottom
face in the upper (secondary) eliminator. The T-8 is normally
used for the primary, and the T-271 model for the secondary
eliminator for use in scrubbing systems with vertical gas flow.
The highest efficiency eliminators are horizontal flow
types and are used where overall design permits. The gas velo-
city through horizontal flow mist eliminators is higher. The
same spray procedures are used, except that spray volumes are
1.5 to 2 gal/min per ft2 at the rate of 1 to 3 min every 10 min.
This increased spray rate is necessary because increased velo-
city and droplet removal efficiency of the eliminators would
otherwise cause a noticeable drying effect and subsequent crus-
ting.
4.7.3.3 Radial Vane6—
The radial vane is a cyclonic separator. One of the limi-
tations of this type of mist eliminator is that it always has to
be installed in a vertical duct. As shown in Figure 4.7-7,
curved vanes redirect the gas stream from the vertical into a
horizontal, spinning flow toward the vessel wall. Heavier
liquid and solid particles in the mist are first accelerated (by
the reduced cross-sectional area) and then directed to the
vessel walls, where they are collected. The collected mist
drains back into the scrubber, and cleaned gases exit to the
stack. The primary collection mechanism is inertial impaction.
Though considerably more expensive than the chevron mist
4.7-10
-------
DROPLET IMPINGEMENT-
DROPLET COLLECTION
DROPLET RUN OFF
\
6
Figure 4.7-6. Munters Euroform mist eliminator.5
Courtesy Munters Corp.
4.7-11
-------
GAS FLOW
Figure 4.7-7. Radial-vane mist eliminator.6
Courtesy Koch Engineering Co.
4.7-12
-------
eliminator in regard to both capital and operating costs, radial-
vane devices are claimed to have far superior collection effi-
ciency and greater washability. The pressure drop through the
radial-vane mist eliminator is considerably higher compared with
that of the chevron type; it ranges from 2 to 6 in. H2O during
operation. To reduce solids loading, the radial-vane eliminator
in some designs is preceded by an impingement tray whose under-
side surface is washed continuously with fresh water. The mist
eliminator itself, however, is not washed when the tray is used.
The center and rim of the scrubber vessel are blocked out to
increase the gas velocity. It should be noted that only two
radial-vane mist eliminators have been installed on utility
scrubbers, but they have not operated successfully (See Section
4.7.8) and are being replaced with chevrons.
4.7.4 Mist Eliminator Design Factors
The design of a complete mist elimination system is com-
plex, in that several conflicting objectives must be considered.
The desire for high collection efficiency and methods to reduce
reentrainment must be weighed against washability or suscepti-
bility to plugging and high pressure drops. Design considera-
tions must also include such factors as scrubber system design
and operating conditions, construction, scrubbing media, solids
content of the slurry, and sulfur content of the coal. Included
in the design are the following four broad areas, all of which
require specifications:
0 Mist eliminator construction
0 Mist eliminator equipment design
0 Mist eliminator wash system
0 General design factors
Each area is discussed in Sections 4.7.4 through 4.7.7.
Discussion is confined to continuous or discontinuous baffle-
type mist eliminators, since these types are preferred for lime
scrubbing applications. The following important factors in mist
eliminator specifications are discussed first:
0 Mist eliminator shape
0 Number of passes
0 Spacing between vanes
0 Slanted mist eliminator
0 Materials of construction
0 Special drainage devices
4 7.4.1 Mist Eliminator Shape—
The shape of the mist eliminator is determined by three
factors: (1) whether vane design is continuous or discon-
tinuous, (2) whether the continuous-vane type has sharp or
rounded bends, and (3) the angle between the vanes. The con-
tinuous chevron-shaped mist eliminator is employed to a far
4.7-13
-------
greater extent than the baffle type in lime scrubbing applica-
tions. Its main advantages are greater strength and lower cost.
Pressure drop is not a main consideration, neither for continu-
ous nor for discontinuous chevron eliminators since the designs
are similar.
Both sharp and smooth vane bends are employed in scrubbing
operations but sharp-angled collectors predominate. Figure
4.7-8 illustrates the difference between the S- and the Z-shaped
bends in the three-pass, continuous chevron mist eliminator.
Sharp-angle bends provide greater collection efficiency, but
they also have a greater tendency for reentrainment and plug-
ging. Figure 4.7-8 also shows a 120-deg bend and a 90-deg bend
chevron mist eliminator. The lower angle design causes more
sudden gas direction changes and resultant greater primary
collection efficiency.
4.7.4.2 Number of Passes—
The number of passes in the mist eliminator corresponds to
the number of direction changes the gas stream must make before
it exits. Normally, the greater the number of passes there are,
the greater the collection efficiency. Because of the high-
solids environment of a lime scrubber, however, the more passes
there are, the more likely it is that plugging will occur.
Figure 4.7-9 shows a two-pass (V-shaped), a three-pass (Z- or
S-shaped), and a multiple-pass chevron mist eliminator. Three-
pass collectors, most commonly used in the lime and limestone
systems, provide good collection efficiency (>90 %) with ade-
quate washability.
4.7.4.3 Spacing Between Vanes—
The spacing between individual chevron blades is an impor-
tant factor in mist eliminator design. The closer the spacing
is arranged, the better the collection but the greater the
potential for plugging. Single-stage mist eliminator spacing
ranges from 1.5 to 3 in.; if a second stage is used, the spacing
is usually the same as that for the first stage. It can, how-
ever, be reduced to as low as 7/8 to 1 in. to provide higher
collection efficiency.
4.7.4.4 Slanted Mist Eliminator7—
Figure 4.7-10 shows a Combustion Engineering design for a
slanted mist eliminator (A-shaped, two stages, and two passes).
It has better drainage than the conventional type, which helps
reduce plugging and intermittent problems.
4.7-14
-------
GAS DIRECTION
CROSS SECTION OF THREE-PASS, 120-deg BEND CHEVRON
MIST ELIMINATOR
90 deg
GAS DIRECTION
CROSS SECTION OF THREE-PASS, 90-deg BEND CHEVRON
MIST ELIMINATOR
Figure 4.7-8. Mist eliminator designs showing the,differences
in baffle angle and the number of passes.1
4.7-15
-------
GAS
DIRECTION
n =2
TWO-PASS
THREE-PASS
SIX-PASS
Figure 4.7-9. Schematic of two-, three-, and six-pass
chevron mist eliminators.1
4.7-16
-------
CHEVRON VANES
WASHER LANCE
BULK ENTRAINMENT SEPARATOR (BES)
Figure 4.7-10. Slanted mist eliminator for vertical gas flow.
4.7-17
-------
4.7.4.5 Materials of Construction—
Stainless steel and plastics are frequently used in mist
eliminator construction. Stainless steel, usually 316L, is
strong and rigid and therefore allows high-pressure washing. It
also has acceptable corrosion resistance in low-chloride sys-
tems, provides a smooth collection surface (which helps limit
scale formation), and is insensitive to sudden temperature
increases. It is heavier and more expensive than other mater-
ials available, however.
Another material frequently used is Noryl (thermoplastic
resin). Noryl is relatively inexpensive and very light, but can
be damaged easily during temperature excursions. It weighs 5
lb/ft2 of face area. Fiberglass reinforced plastic (FRP) weighs
5 to 7 lb/ft2 of face area and stainless steel (316L), 22 lb/ft2
of face area. *
When used to construct mist eliminators, FRP and polypropy-
lene have excellent corrosion resistance, similar to that of
Noryl. Nevertheless, these materials are not strong enough and
can become embrittled after long exposure to scrubbing slurry.
They also have temperature limitations. During temperature
excursions the risk of fire is quite great with these materials.
Despite the various drawbacks, FRP has become the predominant
material of construction.
4.7.4.6 Special Drainage Devices—
Special drainage features, such as hooks and pockets, have
been applied to lime scrubbing systems. Figure 4.7-11 shows
three examples: the Heil, Peerless, and Matsuzaka designs.
Hooks, placed in the gas flow, trap drainage and prevent reen-
trainment. Pockets, however, can easily become plugged and, for
this reason, have been more widely employed in horizontal gas
flow systems and/or in open-loop operations, where the elimina-
tor can be more thoroughly washed. The only large-scale appli-
cations of this pocket mist eliminator by U.S. utilities have
been at the Mohave Station of Southern California Edison (ver-
tical configuration) and the Colstrip Station of Montana Power
(horizontal configuration).
4.7.5 Mist Eliminator Equipment Design Considerations
The following important topics in mist eliminator equipment
design are discussed in this section:
0 Mist eliminator configuration (horizontal vs. vertical
gas flow)
0 Use of bulk separation and knock-out devices
0 Freeboard distance
0 Number of stages
4.7-18
-------
GAS FLOW
HEIL CHEVRON MIST ELIMINATOR WITH COLLECTION HOOKS
PLAN VIEW
GAS FLOW
PEERLESS CHEVRON MIST ELIMINATOR WITH COLLECTION POCKETS
PLAN VIEW
GAS FLOW
i>
MATSUZAKA MIST ELIMINATOR WITH COLLECTION POCKETS
PLAN VIEW
Figure 4.7-11. Cross section of mist eliminator with
special reentrainment-prevention features.
4.7-19
-------
0 Distance between stages
0 Passage geometry
0 Miscellaneous
4.7.5.1 Mist Eliminator Configuration (Horizontal vs. Vertical
Gas Flow)—
Figure 4.7-12 shows horizontal and vertical configurations.
In the vertical gas flow configuration, the gas flow opposes the
path of drainage. Before a collected mist droplet falls from
the mist eliminator blade, it must overcome drag forces exerted
by the gas stream. The balancing of drag and gravitational
forces results in a longer residence time of droplets on the
blade, which increases the chance of scaling, plugging, and
reentrainment. This is one of the disadvantages of the vertical
gas flow configuration.
A second disadvantage is the water balance constraint on
mist eliminator washing. In closed-loop operations, the avail-
able quantity of fresh wash water is limited by water balance
requirements, since the water is returned to the scrubber. This
limitation has often resulted in plugging and eventual shutdown
of the scrubber system for mist eliminator cleaning. Another
problem with the horizontal configuration is the limitation in
wash water direction. The most effective washing should occur
if water is admitted longitudinally along the length of the
vane. The horizontal configuration admits wash water from the
top face and/or bottom face of the mist eliminator only.
The use of mist eliminators in a vertical configuration
with horizontal gas flow is nearly universal in the Japanese FGD
industry; the mist eliminator units are normally installed in a
separate chamber after the scrubber. A main reason for the
efficient removal in the vertical configuration is that captured
liquid flows continuously to a collection area and is not al-
lowed to "pile up" only to be reentrained in the flue gas.
Because reentrainment is more difficult, the vertical configura-
tion can be operated at higher velocities.
The Japanese use the vertical configuration, even though it
has greater capital cost, because the higher elimination effi-
ciency effected by this configuration reduces the load on the
reheater. As discussed earlier, another reason for the higher
mist elimination efficiency achieved by the Japanese is their
wide use of hooks and pockets.
In North America, researchers at TVA, Riley Stoker, and
Ontario Hydro have recommended vertical mist eliminator systems.
Weir horizontal scrubbers readily employ them since the scrubbed
gas exits from the scrubber in a horizontal flow.
4.7-20
-------
HORIZONTAL CONFIGURATION (VERTICAL GAS FLOW)
t
GAS DIRECTION
VERTICAL CONFIGURATION (HORIZONTAL GAS FLOW)
SIDE
GAS DIRECTION
TOP
Figure 4.7-12. Horizontal and vertical mist eliminators.
4.7-21
-------
4.7.5.2 Use of Bulk Separation, Wash Tray, and Knock-out
Devices--
Bulk separators, wash trays, and knock-out devices, are
designed to remove most large liquid droplets from the flue gas
before the stream passes through the mist eliminator. In some
cases the devices are designed to allow continuous recirculation
of wash water.
A bulk separation device can consist of a single row of
baffle vanes (equivalent to a single-pass mist eliminator) with
relatively wide spacing between them, or a flow configuration
resulting in an abrupt change in flow direction (either 90 deg
or 180 deg) (Figure 4.7-13). Because these devices are low-
plugging separators with low pressure drop, they appear to have
only marginal value and are not employed at most sites.
A number of tray designs provide varying amounts of gas-
liquid contact at specific degrees of turndown. The most, ele-
mentary design is the "wash" tray, or impingement tray, which
employs a horizontal sieve deck to allow gas to contact cross-
flowing liquid. This tray is usually suitable for design condi-
tions, but is not capable of effective operation at turndown
conditions greater than 4 to 1. The tray will "weep" (i.e.,
bleed excessive liquid through the holes before it gets across
the tray) at low gas flows or "jet" (blow liquid off the tray
and prevent the downcomer from sealing) at high gas flows.
Holes in wash trays usually have a diameter as small as 1/4 in.,
which produces a high risk of plugging.
The knock-out devices shown in Figures 4.7-14 and 4.7-15
remove large liquid droplets while providing a means to recycle
the wash water. Recirculating this relatively clean water
offers several advantages. First, it allows the wash water flow
rate to be increased significantly, and second, it permits
flexibility in washing operations, wash water treatment, addi-
tion of scaling inhibitors, etc. The Koch Flexitray wash tray
(Figure 4.7-14) is a discrete stage that effects intimate
liquid-gas mass transfer.8 The Koch Flexitray is a valve tray
that employs a floating cap that adjusts itself above the open-
ing to maintain satisfactory tray hydraulics over wide varia-
tions in gas and/or liquid flows. Since knock-out devices are
complicated and have high pressure drops, they are not used at
most installations.
4.7.5.3 Freeboard Distance—
Freeboard distance is the distance between the end of the
absorption section and the mist eliminator. It varies widely
among installations, ranging from 4 ft to more than 20 ft. in
the freeboard area, entrained particles can coalesce and return
to the scrubber solution by gravity before encountering the mist
4.7-22
-------
to
A
t
BAFFLE SLATS
NKN
\
t
90-DEGREE
GAS DIRECTION CHANGE
\
ISO-DEGREE
GAS DIRECTION CHANGE
Figure 4.7-13. Bulk separation systems.
1
-------
I
to
Figure 4.7-14. Koch Flexitray wash tray.
1
-------
to
01
Figure 4.7-15. UOP trap-out tray.
-------
eliminator. Most particles that will settle are usually removed
in the first 8 to 10 ft; the additional freeboard is not effec-
tive in removing the smaller entrained particles before they
contact the mist eliminator.
4.7.5.4 Number of Stages—
Opinions vary regarding the use of one or two stages in a
vertical mist eliminator (horizontal gas flow). Both systems
are used in lime scrubbing operations.
Single-stage mist eliminator efficiency is increased with
the use of a bulk separation device, but spray volume and pres-
sure must be limited to reduce mist generation by the washing
operation.
Although the two-stage system is more expensive and compli-
cated, it has some advantages over the single-stage design. The
first stage of a two-stage system can be rigorously washed from
the front as well as from the back. Mist generated in the
washing operation is collected in the normally unwashed second
stage. A greater quantity of wash water, higher pressure, and
greater duration of washing are also possible. There is also
greater flexibility in designing a two-stage system.
4.7.5.5 Distance Between Stages—
Early designs of two-stage mist eliminators in horizontal
ducts provided less than 1 ft between stages. With such short
distances, severe plugging occurred frequently. Designers have
subsequently achieved higher collection efficiency by high-
volume washing of the first stages and including enough spacing
between stages to allow entrained liquid drops to settle out
before they contact the second stage. The optimum distance
between stages is approximately 6 ft. This also allows suffi-
cient space for personnel to walk between stages during cleaning
periods.
4.7.5.6 Passage Geometry--
Superficial gas velocity plays an important role in the
effectiveness of primary collection and reentrainment. In an
attempt to adjust the velocity of the gas passing through the
mist eliminator, some scrubbers have an enlarged cross-sectional
housing in their design. Because of the uneven gas flow distri-
bution caused by such an expansion, however, substantial gain in
collection efficiency has not been achieved.
4.7.5.7 Miscellaneous—
Some thought should be given in mist eliminator design to
features that will be useful during cleaning or replacement of
4.7-26
-------
the units. These could include a rectangular walkway, 2 by 6
ft, that would provide easy access and prevent maintenance
personnel from having to stand on the eliminator, where their
feet could be trapped between the blades. Standing on a cor-
roded or embrittled mist eliminator could also cause it to
collapse. Whatever the material of construction, mist elimi-
nator sections should be light enough to be lifted easily by two
people. Lightweight mist eliminators are easy to install and
clean during maintenance periods.
Also of the utmost importance is the ability of the mist
eliminator to operate satisfactorily under turndown conditions.
Reduced gas flow through the eliminator will reduce the velocity
of the gas, and the mist eliminator's efficiency will decrease
to a very low level. Modular design of the eliminator is one
way to solve this problem.19
4.7.6 Mist Eliminator Wash System
Design of the mist eliminator wash system has advanced
greatly since the magnitude of the scaling and plugging problems
first became evident. The following factors are important in
the specification of a complete mist eliminator wash system:
1. Wash water type
2. Wash water direction (front, top, and/or back)
3. Wash direction and quantity of water (intermittent vs.
continuous)
4. Wash water pressure
5. Type of coverage (total vs. partial)
4.7.6.1 Wash Water Type2'7—
Since the main purpose of mist eliminator washing is to
clean off accumulated scale and mud deposits, fresh water is
naturally preferred. For closed-loop lime scrubbing, 100 per-
cent fresh-water washing is usually impossible. The normal
procedure in lime scrubbing systems is to introduce all makeup
fresh water (in excess of that required for pump seals and lime
slaking) through the mist eliminator wash system. Additional
wash water, which is sometimes required, is usually obtained by
recycling clear water from the thickener or from the sludge pond
overflow. The disadvantage of using this liquor is that it is
already saturated with calcium sulfate. If lime carried up into
the mist eliminator reacts with residual sulfur dioxide in the
gas to form more calcium sulfate, then precipitation and scaling
can occur. This sulfate-saturated recycle liquor, however, can
be diluted with fresh water and chemically treated with soda ash
to remove the calcium ion. In some cases, scrubber slurry can
be used on the mist eliminator front-wash system, while fresh
water or diluted recycle liquor from the thickener or sludge
pond overflow can be used on the back-wash system.
4.7-27
-------
4.7.6.2 Wash Direction—
The direction of wash water flow depends on the mist elimi-
nator configuration and on the number of stages. With a hori-
zontal configuration, washing is possible only from the bottom
and top of the column. Using a vertical configuration, wash
water can be directed horizontally from the front and/or back
and vertically from the top. In all cases it has been found
that washing in a direction countercurrent to the gas flow, or
from the top in a vertical mist eliminator, generates large
quantities of mist. A second-stage mist eliminator is therefore
desirable when long-duration washing in those directions is
planned. When a single-stage mist eliminator is actually speci-
fied, countercurrent washing is normally limited to short dura-
tion at high pressure and high volume, i.e., deluge or flush
washing.
4.7.6.3 Wash Duration and Water Quantity—
Difficulty with scaling and plugging in the mist eliminator
section of the scrubbing system has been primarily associated
with the circulating wash system. At present, suppliers and
operators are designing systems for washing from different
directions using different wash durations and wash water pres-
sures .
In lime scrubbing facilities, where slurries are normally
composed of 5 to 15 percent solids and reagent utilizations are
greater than 85 percent, plugging problems are less severe than
in those systems that use limestone. An intermittent, short-
duration spray is usually sufficient to keep the mist eliminator
of a lime scrubbing facility relatively clean and operational.
An alternative scheme is to place the mist eliminator wash
system in a separate, closed-loop mode. This is possible on
vertical gas flow mist eliminators when devices such as the Koch
Flexitray or UOP trap-out tray (Figures 4.7-14 and 4.7-15) are
employed to collect the wash water for recycle. Horizontal gas
flow mist eliminators can easily be placed in a closed-loop
mode. Use of these devices can significantly increase the total
quantity of wash water available, allowing when necessary the
use of continuous, high-volume sprays for two-stage systems.
However, a purge to the scrubber must be made to prevent super-
saturation of the wash liquor.
4.7.6.4 Wash Water Pressure2'7—
Wash water pressures, which are important in the overall
wash system, vary widely; both lime and limestone scrubbing
systems use an intermittent flush spray at pressures ranging
from 20 to 40 psig. Where fresh-water input is restricted, a
4.7-28
-------
moderately high-pressure, short-duration wash spray will usually
maintain relatively clean collectors while conserving fresh
water.
Incorporation of high-pressure washing procedures must be
included in the original design. If not, certain design modifi-
cations must be made, such as replacing plastic blades with
stainless steel blades or blades constructed of thicker, rein-
forced material, so they can withstand the additional stress of
high-pressure washing.
4.7.7 General Factors
Several other factors remain outside the above specifi-
cation categories:
0 Mist eliminator operating pressure drop (for deter-
mination of pluggage)
0 Overall collection efficiency
0 Mist loading and particle size distribution
4.7.7.1 Mist Eliminator Pressure Drop—
Lime scrubbing installations experience less than 1 in. H20
pressure drop in their mist eliminators when they are clean. As
scale and mud deposits accumulate, however, pressure drop and
reentrainment increase. A TVA publication reports one experi-
ment in which a AP rise from 0.1 to 1 in. H20 corresponded to a
50 percent blockage of the free area between the mist eliminator
blades. In another experiment, the AP across the mist elimina-
tor rose to over 5 in. H2O during the course of one long run.
inspection showed that the mist eliminator was almost completely
plugged by a soft, mudlike accumulation of entrained solids.10
This pressure drop can be used as a control mechanism to deter-
mine the degree of pluggage present in an operating mist elimi-
nator.
4.7.7.2 Overall Collection Efficiency—
Overall collection efficiency of a mist eliminator depends
on several factors, including particle size, distribution of the
mist, pressure drop, gas velocity, type and design of the elimi-
nator, and the quantity of mud and scale on its surfaces.
Overall collection efficiency is easily obtainable from measure-
ments of similar equipment in liquid-only applications.1 Data
on actual lime or limestone scrubbing applications are either
not easily obtained or they are not currently available. Be-
cause of the added complexity solids give to sampling and data
interpretation, FGD system operators have not been able to
measure inlet and outlet particle size distribution and solid
and liquid particulate loadings. Consequently, evaluation and
4.7-29
-------
improvement of mist eliminator design have been based on other,
less-precise criteria, e.g., in-line reheater tube plugging,
stack corrosion, scaling, and/or corrosion of induced-draft
fans. This area deserves additional study. Quick and accurate
methods are needed to measure mist loadings and particle-size
distribution.
4.7.7.3 Mist Loading and Particle-size Distribution—
Mist loading and particle-size distribution should be
measured accurately to determine the removal efficiency of a
mist eliminator. Determination of mist eliminator removal
efficiency is important to evaluate the effectiveness of design
modifications and parameters. The effect of gas velocity, L/G
ratio, percent solids of the slurry, and other operational
variables cannot be adequately evaluated unless mist loading and
particle-size distribution are known. Until recently no simple,
accurate method of measuring mist loadings and particle sizes
had been available for lime scrubbing applications. Now it
appears that CFH Research Laboratories has developed such a
procedure, based on microphotographic methods.11 Direct photo-
graphs of drops in their natural environment are made using a
proprietary optical system. Figure 4.7-16 shows such a droplet-
sizing photograph. The photographs are electronically scanned
and the information fed directly to a minicomputer, which pro-
vides a statistical analysis (with accuracies of 10%) of the
particle loadings, density, size distribution, and/or veloci-
ties. Both laboratory testing and direct field studies of
nozzles, spray towers, cooling towers, mist eliminators, etc.,
may be possible with this system.11 Figure 4.7-17 is a graph
drawn from data collected by this method.
4-7.8 Existing Facilities (Experience and Equipment Detail)
This section summarizes operational and design experience
with mist eliminators in lime systems and provides background
information on the scrubbing facilities. Data from operational
plants show that most lime scrubbing FGD systems can apparently
operate the mist eliminators successfully with only intermittent
mist eliminator washing and manual washing during shutdown
periods. Both relatively high- and low-sulfur coals and dif-
ferent types of scrubbers are used in the systems without sig-
nificantly affecting mist eliminator operation. Design factors
for mist eliminators on lime scrubbing systems are summarized in
Table 4.7-1.
Most mist eliminators used in lime systems are three- or
four-pass, 90-deg bend chevron mist eliminators with vanes made
of reinforced plastic and spaced 1 to 3 in. apart. The single-
stage mist-eliminators are housed in a nonexpanded vertical duct
(top of scrubber) and placed 4 to 20 ft above the last absorbing
4.7-30
-------
Figure 4.7-16. Droplet sizing photograph.
Courtesy CFH Research Laboratories.
4.7-31
-------
I
UJ
ro
100
98
96
94
5 92
o
I 90
u
0)
86
84
82
80
0
10
20
30 40 50 60
MIST DROPLET DIAMETER, ym
70
80
Figure 4.7-17. Percent removal vs. mist droplet diameter obtained by
the application of the CFH method.11
-------
Table 4.7-1. DESIGN FACTORS FOR MIST ELIMINATORS FOR LIME SCRUBBING SYSTEMS.
£>.
•
I
Plant
number
1
2
3
4
5
6
7
8
9
10
•^— —
Plant
name
Phillips
Paddy • s
Run
Elrama
Green
River
1 and 2
Bruce Hans-
field 1
Bruce Mans-
field 2
Conesville 5
Can* Run
Hawthorn
Colttrip
1 and 2
Shawnee
Superficial
gas velocity,
ft/s
9-11
10
25-30
10
13.7 (calcu-
lated)
10
10
7.5
9.4
ME type
Chevron
(Non-
cont inuous)
Chevron
AAF radial
vane
Chevron
Chevron
Chevron
Chevron
Heil Chevron
Chevron
ME shape
Z- shaped
90° sharp-
anqlr
bend
Z-shaped
120'
Curved
vane
Z-shaped
Z-shaped
90' sharp-
a ncj 1 e be nd s
Rounded
Z-shaped
Smooth 120°
with hooks
Z-shaped
90° sharp-
anqle bends
Numbrr
of
passes
3
3
1
4
3
3
2
4
3
Spacing
between
vanes, in.
! i
1.5 to 2
2.5-3.0 at
inlet, 4.0
at outlet
1-1.25
2
1-1.5
3
1
3.5-4
Number
of
stages
1
2
1
1
2
2
2
1
1
Distance
between
•tages ,
ft
NA
NA
NA
4.5
6
1.25
NA
NA
(continued)
-------
Table 4.7-1 (continued)
CO
Plant
number
1
2
)
4
5
6
7
8
9
10
Plant
name
Phillips
Paddy ' s Run
Elrama
Green River
Bruce Mansfield 1
Bruce Mansfield 2
Conesville 5
Cane Run 4
Hawthorn
Colstrip 1 and 2
Shawnee
Type of
knock-out
tray
None
None
None
None
None
OOP trap-
out tray
None
None
—
-
Material
of
construction
FRP
FRP
FRP
Turning vanes stain-
less steel, outside
collection area
coated mild steel
NA
Stainless steel
Stainless steel
FRP
PVC
316 SS
Total
AT,
in. H2O
4
1
4
2. 1
0.5-0.75
1.6 - 1st stage
0.3 - 2nd stage
0.5 - 1.2
1.2
0.5
0.5
Col lector
combination
Horizontal
Horizontal slanted
inverted V-shape
Horizontal
Horizontal
Horizontal
Horizontal
Horizontal slanted
inverted V-shape
Horizontal
Horizontal
Distance between
last absorption
stage and HE
4-5 ft
5 ft
4-5 ft
10-15 ft
NA
10 ft total. 6 ft froix
absorber nozzle to tray.
4 to 5 ft from tray to «E
10 ft
13 ft. 5 ft between spray
nozzle and wish tray. 1 ft
between wash tray and ME
4 ft
(continued)
-------
Table 4.1-1 (continued1)
i
u>
Ul
Plant
number
1
2
3
4
5
6
7
8
9
10
Plant
name
Phillips
Paddy's Run
F. 1 r ama
Green River
1 and 2
Bruce Mansfield 1
Bruce Mansfield 2
Conesville 5
Cane Run 4
Hawthorn
Colstrip 1 and 2
Shawn ee
Wash
water
type
River water
River water
River water
treated
Recycle water
from thickener
overflow
River water
River water
River plus
thickener
Underwash :
pond overflow
and river water.
Topwaah: river
water. Hash
tray: fresh
water in closed-
loop system.
River water
Wash
duration
5 min every 2 h
10-15 min every 8 h
Continuous
Intermittent
Continuous both
stages
2 min every 5 min
1 to 1.5 h every 3
days (apx.)
Underwash: continu-
ous
Topwash: 24 min
every 24 h
Intermittent
Wash water
pressure ,
psiq
Underwash: 15-20
Topwash: 40
40-65
50
30
86
70
100 to 120 manual
wash; 70 with wash
lance
40-50
Topwash: high pres-
sure. Underwash:
low pressure.
Wash rate
(per module) ,
g a 1 /m i n
Topwash: 60-70
Underwash: 125
80-200
45
60
1st stage: 500
2nd stage: 500
Tray wash: 90
40
2000
Topwash: 390
Underwash: 30
fcopwash: 100 every 80
min
Underwash: 750 every 4 h
NA - Not available.
-------
stage in the nontilted horizontal configuration. Superficial
gas velocities ranged from 7.5 to 21.6 ft/s with most ranging
from 9 to 14 ft/s.
4.7.8.1 Phillips Station (Duquesne Light Company)12--
The mist eliminator at Phillips Station, supplied by
Chemico, started operation in 1973. The system has an internal
mist eliminator within the venturi before the wet scrubber
induced-draft fans. The system also has a large knock-out
chamber and mist eliminator after the fan. The internal mist
eliminator plugged frequently before the system was changed to a
continuous wash under the mist eliminator. Although this re-
duced the pluggage or scaling problem, it is doubtful that the
mist eliminator is effective. Better washing reduces buildup on
the induced-draft fan. Therefore, a good internal mist elimi-
nator is not required. Effective mist elimination probably
occurs in the large external mist eliminator, since few problems
have been reported and the mist eliminators are only washed
daily.
4.7.8.2 Paddy's Run 6 (Louisville Gas and Electric Company)13
The mist eliminator at Paddy's Run 6 was supplied on a
marble-bed absorber by Combustion Engineering. It has been in
operation for the last 4-1/2 years and has remained clean and
pluggage-free. Its trouble-free operation may be attributed to
the use of carbide lime. When a change was made from carbide
lime to high-calcium commercial grade lime, scaling and plugging
occurred.
4.7.8.3 Elrama Station (Duquesne Light Company)14—
The mist eliminator on a venturi scrubber at Elrama Station
was supplied by Chemico and has been in operation since August
1975. It, too, was redesigned by Gibbs & Hills, Inc., in 1975.
Less scaling and plugging have been reported since redesign,
though it is doubtful whether the mist eliminator is effective
This system, like that of Phillips, has large external mist
eliminators in knock-out chambers.
4.7.8.4 Green River Station (Kentucky Utilities)15—
The radial-vane mist eliminator, on a mobile-bed absorber,
was supplied by American Air Filter (AAF). Severe scaling and
plugging of the scrubbers system's downstream equipment were
caused by slurry carryover and low efficiency of the mist elimi-
nator. Some modifications are proposed to improve the elimi-
nator's performance. However, at the time of writing, the
modifications had not yet been carried out.
4.7-36
-------
4.7.8.5 Bruce Mansfield (Pennsylvania Power Company)16—
The mist eliminators on venturi scrubbers at Bruce Mans-
field Station were supplied by Chemico. All have been in opera-
tion since June 1976.
One of the mist eliminators on the scrubber became plugged
very soon after the operation began, probably because gas flow
through the scrubber was much higher than designed. The
material plugging the mist eliminator hardened to a point where
it could not be removed and the entire unit was replaced. These
mist eliminators are normally cleaned by intermittent spraying
with recycled water. To prevent the problem of plugging, a
system was installed that permitted the mist eliminator to be
flooded with large volumes of water in the event of excessive
pressure drop across the device.
Despite this new wash water piping, as of this writing, the
mist eliminators are still plugging with hard scale, and peri-
odically they must be manually cleaned. Even at design flow,
these units are not meeting the design criteria of 1 g/ft3
carryover. Tests have indicated that mist carryover from the
mist eliminator is from 2 to 3 g/ft3.
In an attempt to solve this problem, a section of vertical
mist eliminator (horizontal gas flow) was installed in the
outlet within the second-stage venturi (absorber) vessel, ad-
jacent to the outlet opening. It was intended to provide a
large area, and therefore a velocity of about 20 ft/s. The
strength of the structure was not sufficient to withstand the
forces of excessive turbulence that occur in this area of the
absorber resulting in the failure of this mist eliminator. It
was then decided to install a smaller section of the vertical
mist eliminator farther back in the outlet ductwork, where the
velocity is about 50 ft/s. Within a few minutes of operation,
the mist eliminator module collapsed and was scattered in small
pieces throughout the ductwork. The manufacturer supplied a
new, heavier module and the supervision to install it. This
mist eliminator was put in service May 23, 1977. During June
1977, Chemico conducted model studies on both the horizontal and
vertical mist eliminators. Full-sized mist eliminator sections
were used in the module studies; information gained from the
study provided the operating company with valuable design cri-
teria. The study revealed that pressure drops in excess of 0.75
to 1 in- H2° allowed excessive carryover from the horizontal
mist eliminators. When pressure drop was maintained at 0.5 in.
H2O or less' there was practically no carryover of entrained
water.
4.7-37
-------
4.7.8.6 Conesville (Columbus and Southern Ohio Electric Com-
pany)17--
The mist eliminator on a turbulent contact absorber at
Conesville Station was supplied by UOP. It has been in opera-
tion since July 1976. No operational problems are reported and
the unit has remained clean and plug-free. Thiosorbic lime is
used in this system.
4.7.8.7 Cane Run (Louisville Gas and Electric Company)13—
The radial vane mist eliminator on a mobile-bed scrubber at
Cane Run was supplied by AAF. The scrubber system started-up in
August 1976. Large pressure drop through the eliminator cut
holes through it and caused mist carryover. In addition, slurry
carryover caused scaling and plugging. The scrubber was shut
down April 18, 1977, together with the boiler, for a projected
2-month overhaul, during which the mist eliminator was removed
by cutting an 18-in. hole in its top section. This section was
replaced with two banks of chevron baffles and an associated
spray washing system. Since startup in July 1977, the chevron
mist eliminators have operated quite well and have remained very
clean.
4.7.8.8 Hawthorn (Kansas City Power & Light Company)18—
The mist eliminator on a marble-bed absorber at Hawthorn
was supplied by Combustion Engineering. The scrubber system
(limestone) started up in August, 1972. Since January 1977,
this system has been operated in a lime scrubbing mode. After
switching to low-sulfur coal burning, the utility has had no
problem with the mist eliminator.
4.7.8.9 Colstrip (Montana Power Company)19—
The mist eliminators at Colstrip were supplied by Heil
Process Equipment Company. The eliminator in Unit 1 was damaged
by temperature excursions and had to be replaced. The new
eliminator in Unit 1 has been in operation since October 1975,
that in Unit 2 since May 1976. Washing with 78 percent tray
pond return water and 22 percent river water has kept the mist
eliminators scale- and plug-free.
4.7.9 Recommendations
In view of the above discussion, some recommendations for
mist eliminator design and operation are summarized below:
1. The continuous chevron is better than the noncon-
tinuous chevron because of its greater strength and
relatively lower cost.
4.7-38
-------
2. The closer the spacing, the better the mist collection
efficiency and the greater the tendency for plugging.
First-stage spacing may be from 1.5 to 3 in. The
second-stage spacing can be as narrow as 7/8 to 1 in.
3. Special features such as hooks and pockets are desir-
able in vertical mist eliminators to decrease reen-
trainment in systems with proper scrubber chemistry to
reduce plugging NO scaling.
X
4. When compared with the horizontal configuration,
vertical configuration (horizontal gas flow) has the
advantages of higher efficiency at high loadings and
better washability; however, the capital cost is
higher.
5. Freeboard distance should normally be 4 to 6 ft.
6. Bulk separation and knock-out devices are recommended.
Proper lime utilization, scrubber gas velocity, and
proper chemistry may reduce the need for bulk separa-
tion.
7. Intermittent high-pressure, high-velocity wash is
better than continuous wash.
8. Noryl, FRP, and polypropylene as materials of con-
struction are relatively lightweight and inexpensive
compared with stainless steel. However, the protec-
tion provided by stainless steel during temperature
excursions should be weighed against the cost. Stain-
less steel (316L) should not be used in high-chloride
(greater than 2300 ppm) systems.
4.7-39
-------
REFERENCES
1. Conkle, H. N., H. S. Rosenberg, and S. T. DiNovo. Guide-
lines for the Design of Mist Eliminators for Lime/Limestone
Scrubbing System. EPRI FP-327 (Research Project 209),
Battelle Columbus Laboratories, December 1976.
2. Heilex-EB, Mist Eliminator Blades. Bulletin B-922-1, Heil
Process Equipment Company, Cleveland.
3. Private communication with R. Centa, J. Gavin, and J.
Jackson, Heil Process Equipment Company, Cleveland.
4. Humboldt Lamellar Mist Separator. Bulletin, Matsuzaka
Company (America), Inc., Oak Brook, Illinois.
5. Euroform Mist Eliminators. Bulletin ETS-1, The Munters
Corporation, Ft. Myers, Florida.
6. Wet Scrubbing Systems for Air Pollution Control. Bulletin
KPCZ, Koch Engineering Company, New York.
7. Green, K., and J. R. Martin. Conversion of the Lawrence
No. 4 Flue Gas Desulfurization System. In: Symposium on
FGD, Hollywood, Florida, November 8-11, 1977.
8. Private communication with R. Schwartz, Koch Engineering
Company, New York.
9. Private communication with T. Moraski, EPRI, Palo Alto,
California.
10. Schult, J. J., et al. Performance of Entrainment Separa-
tion in Slurry Scrubbing Process. Bulletin Y-93, National
Fertilizer Development Center, Tennessee Valley Authority,
Muscle Shoals, Alabama, June 1975.
11. Private communication with F. Hoffman, CFH Research Labora-
tories, Springfield, Massachusetts.
12. Private communication with J. Malone, Phillips Station,
Duguesne Light Company, South Height, Pennsylvania.
4.7-40
-------
13. Private communication with R. Van Ness, Paddy's Run Sta-
tion/ Louisville Gas and Electric Company, Louisville.
14. Private communication with F. Bork, Elrama Station,
Duquesne Light Company, Elrama, Pennsylvania.
15. Private communication with J. Reisenger, Green River Sta-
tion, Kentucky Utilities, Central City, Kentucky.
16. The Wet Scrubber Newsletter. The Mcllvaine Company, No.
38, August 31, 1977, p. 7.
17. private communication with D. Boston, Conesville Station,
Columbus and Southern Ohio Electric Company.
18. PEDCo Environmental, Inc. FGD Summary Report. EPA 68-01-
4147.
19. Grimm, C., et al. Particulate and SO2 Removal at the
Colstrip Station of the Montana Power Co. In: 2nd Pacific
Chemical Engineering Congress, Denver, August 1977.
4.7-41
-------
CONTENTS
4.8 FANS
4.8.1 Introduction
4.8.2 Service Description
4.8.3 Design Parameters
4.8.3.1 Location
4.8.3.2 Temperature Increase
4.8.3.3 Wet Fan
4.8.4 Materials of Construction
4.8.5 Existing Facilities
4.8.5.1
4.8.5.2
4.8.5.3
4.8.5.4
4.8.5.5
4.8.5.6
References
Bibliography
Cane Run Power Station
Green River Power Station
Conesville Power Station
Paddy's Run Power Station
Phillips Power Station
Bruce Mansfield Power Station
4.8-2
4.8-5
4.8-6
4.8-6
4.8-8
4.8-8
4.8-9
4.8-9
4.8-10
4.8-10
4.8-11
4.8-13
4.8-14
4.8-i
-------
4.8 FANS
4.8.1 Introduction
Fans are used to drive gas through lime scrubbing FGD
systems. These fans are radial-flow (or centrifugal) types, in
which the gas flow is at right angles to the axis of motor
rotation. Fan operation can be wet or dry. Dry fans operate at
temperatures higher than the dew point of the species present in
the flue gas. Wet fans operate in a flue gas atmosphere satura-
ted with water.
In this report, the fans are called forced-draft (FD),
induced-draft (ID), or ID booster fans, depending on the loca-
tion and the role of the fan in the lime scrubbing system. The
term "forced draft" is used when air or flue gases flowing in a
scrubber system are maintained at pressures above atmospheric
pressure (this fan is the induced-draft fan of the boiler sys-
tem) .
When air or flue gas flows in a unit under the influence of
a progressively decreasing pressure below atmospheric pressure,
the system is said to be operating under induced draft. This
term is used to describe a booster fan installed downstream from
the reheater.
Since the fan application in a scrubber system is very
similar to that in a boiler system, power plant engineers are
familiar with the performance and the mechanical design of these
fans. Thus, only operational and process design features unique
to a lime scrubbing FGD system will be presented in this sec-
tion.
4.8.2 Service Description
Fans are installed in the system to handle gases coming
from a boiler or a scrubber. The gas flow rates range from
300,000 to 500,000 acfm, and the temperature ranges either from
300° to 340°F or from 110° to 200°F, depending on the fan's
location in the scrubbing system. The gas may contain moisture,
particulates, sulfur dioxide, and/or acid mist. Particulates
are generally abrasive and if wet tend to form scale or build up
on the fan blades.
When a fan is upstream from a scrubber or downstream from a
reheater, the gas temperature is higher than the water satura-
tion temperature. Therefore, water will not condense on the fan
and the operation is dry. When a fan is located between two
scrubbers or between a scrubber and a reheater, entrained and
condensed water cause the fan operation to be wet. This type of
4.8-1
-------
operation may require wash sprays on the fan to eliminate the
solids buildup. In a high-sulfur coal application, the fan wash
will have a tendency to absorb SO2, turning the water acidic.
Care must be taken, therefore, in the selection of materials of
construction or lining for these wet fans.1"3
Because very few applications permit fans to operate con-
tinuously at the same pressure and volume, some convenient means
of volume control through the fans is needed to maintain scrub-
ber and boiler load requirements. This is commonly achieved
with a variable-inlet vane or damper(s) as well as variable
speed controls on the fans themselves.
The centrifugal action of a fan imparts static pressure to
the gas. The diverging shape of the scrolls (curved portion of
the fan housing) also converts a portion of the velocity head
into static pressure. Although the normal static pressure
requirement is approximately 20 in. H2O, scrubber system design-
ers commonly add 15 to 25 percent to the net static pressure
requirement when purchasing a fan as a safety requirement to
allow for buildup of deposits in ductwork and the inherent
inaccuracies of the calculation.
4.8.3 Design Parameters
4.8.3.1 Location--
A scrubber FD fan in a lime FGD system (Figure 4.8-1)
delivers hot gas to a scrubber. The gas temperature is higher
than the acid dew point and there is no corrosion problem;
however, unless a precipitator is located upstream (as shown in
Figure 4.8-1), the gas contains abrasive fly ash, which may
cause erosion in the fan. The gas flow rates are usually high
because of the high temperature, which requires a larger capa-
city fan and a higher operating cost. A system with a scrubber
FD fan operates under pressure as described in Section 4.8.1.
Therefore, any leaks that develop are easily noticeable by
corrosion products on the metal surface or discoloration of the
metal surface.
A system with an ID fan (Figure 4.8-2), on the other hand,
operates slightly below atmospheric pressure. If leaks develop
in this system, air may be drawn in, causing undesirable oxida-
tion of the scrubbing solution; such leaks will also increase
the volume of air that must be handled. Unlike the leakages in
the scrubber FD fan system, this type of leak is very difficult
to detect. Generally the ID fan handles less gas volume than
the FD fan because of the cooler gas temperature even though it
is usually located after the reheater. Its operation is con-
sidered dry though gas from the reheater may contain acid mist
4.8-2
-------
00
i
PARTICULATE
REMOVAL
DEVICE
FLUE
REHEATER
GAS
SCRUBBER FD FAN
QUENCHL
Figure 4.8-1. Scrubber FD fan application.
-------
FLUE GAS
r
REHEATER
DEMISTER
ID FAN
oo
t
Figure 4.8-2. ID fan application.
-------
carry-overs and particulates, which could cause corrosion,
erosion, and scaling problems in the fan. Such problems nor-
mally originate from poorly designed or operated mist elimi-
nators.
4.8.3.2 Temperature Increase—
The adiabatic compression process of a centrifugal fan will
increase the gas outlet temperature according to the equation:
P1
or
T _ rn I =_*. 1 R/Cn
-1-2 - -"-I \T> I P
> (it)
where Subscripts 1 and 2 denote the inlet and outlet conditions,
respectively, and
T = temperature (T)
p = pressure (M/Lt2)
V = cp/cv (dimensionless)
R = the gas constant (ML2/t2 T mole)
Cp = heat capacity at constant pressure per unit mass (L2/t2 T)
Cv = heat capacity of constant volume per unit mass (L2/t2 T)
The dimensions L = length, t = time, T = temperature, M = mass
are given in brackets.
Suppose for a scrubber ID fan:
T! = 166°F
P2 = 14.69 psia
AP = 20 in. H2O =0.72 psia
and C = 7.4186 Btu/lb-mol-°R
then P? = P2 - AP
= 14.69 - 0.72 psia
= 13.97 psia,
/14 69 osia\ 1-987/7.4186
* T - M66 + 46Q°m I PSia \
and T2 - ^-LDO T *±°u K) i, _ g_ . i
= 634.5°R \xj.y/ psia/
= 174.5°F
4.8-5
-------
Therefore, the theoretical temperature increase in the ID fan is:
AT (ideal) = T2 - Tx = 174.5 - 166
= 8.5°F
For a scrubber FD fan:
T! = 285°F
P! = 14.69 psia
AP = 20 in. H20
and C = 7.4186 Btu/lb-mol-°R
then pf = Pt + AP
= 14.69 psia + 0.72 psia
= 15.41 psia
/IR AI ^o-ia\ 1-987/7.4186
and T2 = (285 + 460°F) (TT'rf Psia)
= 754.6°R ^14-69 PSla'
= 294.6°F
Therefore, the theoretical temperature increase in a scrubber FD
fan is:
AT (ideal) = T2 - Tx = 294.6 - 285
9.6°F
The actual temperature increase in the fan operation is a little
higher because the fan efficiency is less than 1 (~0.8-0.9).
The temperature increase due to the ID fan operation is an
added advantage because it will lower the reheater duty.
4.8.3.3 Wet Fan—
A typical wet fan application is shown in Figure 4.8-3.
The gas from the prescrubber is saturated with water and con-
tains acid mists and abrasive carry-overs. Even though water
sprays are installed to clean the fan internals, corrosion,
erosion, and scaling can present continual problems. Buildup of
solids on the fan blades causes severe corrosion and rotor
imbalance, which in turn lead to higher noise levels, excessive
vibration, and finally to fan failure. Because of these limita-
tions, wet fans are usually not designed into the scrubbing
system. However, if they are designed into the system, protec-
tion against corrosion is essential.
4.8.4 Materials of Construction
Fans in dry operation are often constructed of carbon steel
and seldom experience serious problems since they are not sub-
jected to the intense corrosive conditions normally found in a
4.8-6
-------
FLUE GAS
PRESCRUBBER
oo
I
SPRAY
WATER
WET FAN
REHEATER
STACK
ABSORBER
Figure 4.8-3. Wet fan application.
ID with respect to prescrubber.
FD with respect to absorber.
-------
wet fan. Care must be taken, however, to minimize the abrasive
effects of the particulates found in flue gases.
In wet-fan operations, high levels of acidity (pH of about
2.0) and chloride ions caused pitting attack and stress-corro-
sion cracks.3'4'6 Thus, wet fans require special materials of
construction. Alloy steels are the most common materials used
to reduce corrosion of impellers and shafts. For example, a
wet-type fan at Phillips Power Station has a Carpenter 20 Cb 3
impeller and 316L shaft, shaft shrouds and sleeves, inlet dam-
pers, and stiffener bars.3'4 However, it should be noted that
stainless steels are susceptible to chloride stress-corrosion
cracking and care should be taken in using these materials in
such environments.
Protective coatings such as rubber or resin on standard fan
housings and impellers of mild carbon steel -construction have
proved satisfactory in some applications. The increasing use of
rubber (about 1/4-in. thick) for coating fan impellers and
housings deserves special attention. Rubber is one of the least
porous of materials and, when vulcanized to the metal, surrounds
and protects it from corrosive gases or fumes. It also can
withstand the high stresses set up in the fan and is flexible
enough to resist cracking. However, bonding failures have been
a problem when the coating has not been applied properly. As
with any lining, temperature excursions can be determined.
Most fan damage or failure is caused by separation of the
coating from the metal or by damage to the coating by solid
debris. A protective coating material should possess good
bonding properties and flexibility under fan operating condi-
tions since different flexibilities between the coating and the
metal can also cause lining failure. Although many coating
materials have been applied to protect blades from corrosion,3'4
no coating material that provides both good bonding and flexi-
bility for the impellers of a wet fan is presently available.
Care during welding and correct choice of materials are
also important for wet-fan construction, since welds are very
vulnerable to galvanic corrosion.
4.8.5 Existing Facilities
4.8.5.1 Cane Run Power Station—
Design information from this facility is the most complete.
Flue gases from the existing boiler ID fans at Cane Run Unit 4
flow through ID booster fans into the scrubbing system. Guillo-
tine isolation dampers are provided at the inlets of each of the
two booster fans and at the outlet of each contactor module.
The booster fans are needed to overcome the additional pressure
drop caused by the contactor modules.
4.8-8
-------
American Standard fluid drives permit the booster fans to
follow boiler load; the booster fans will maintain a constant
gas pressure to the scrubber and neutralize the additional
pressure drop the scrubber system places on the existing ID
fans. The booster fans are Buffalo Forge double-width, double-
inlet units rated at 367,000 acfm at 325°F. The fan bearings
are water-cooled, Dodge sleeve units. The fans have fluid
drives driven by 1250-hp Reliance Electric induction motors
rated at approximately 720 rpm. The fluid drives are American
Standard, size No. 427, Class 4 units with No. 1004 shells and
tube coolers. The coolers are water-to-oil, American Standard,
type BCP heat exchangers. The fluid drive oil pump is a con-
stant displacement gear pump.
Resistance temperature detectors on the fan, motor, and
fluid drive bearings provide a signal to temperature monitors.
Annunciator warning is set at 260°F.
A vibration element located on the fan outboard bearing
provides input to a vibration monitor. Vibration in inches per
second is read on the monitor. Contacts in the monitor provide
annunciator warnings of high vibration (0.2 in./s) and for motor
cutoff on excessive vibration (0.3 in./s). No scrubber fan
problems have been reported.
4.8.5.2 Green River Power Station—
At Green River Power Station, the flue gases from each
boiler are coupled into a series of mechanical collectors where
primary particulate removal takes place. The flue gas is then
drawn from the existing breeching through a guillotine-type
isolation damper and associated ductwork by the scrubber fan.
This fan is a Buffalo Forge double-inlet unit rated at 360,000
acfm at 300°F. It is constructed of mild steel and driven by a
1500-hp Allis Chalmer motor at 890 rpm. This fan generates a
pressure of 18 in. H2O and maintains a pressure of 0 in. H20
upstream from the fan. This ensures that there is no back
pressure on the boilers.
In an early operation, this fan was unbalanced and experi-
enced excessive vibration, which required system shutdown for
balancing.
4.8.5.3 Conesville Power Station—
Conesville Unit 5 has two air-foil type fans located after
the ESP and before the scrubber. Guillotine dampers are pro-
vided at the inlets and outlets of the fans, and a louver damper
permits a bypass.
4.8-9
-------
The fans, from the Green Fuel Economizer Company, are rated
at 850,000 acfm at 286°F. They are constructed of mild steel
and driven by 7000-hp motors at 900 rpm. They generate pres-
sures up to 46 in. H20.
This plant experienced occasional motor fluid coupling
problems.
4.8.5.4 Paddy's Run Power Station--
Paddy's Run Unit 6 has two 1500-hp ID booster fans after
the reheater, each having a 175,000 acfm capacity at 350°F.
These fans had no operating problems; design details are not
available.
4.8.5.5 Phillips Power Station—
The lime scrubbing FGD facility at Phillips Station is a
two-stage venturi scrubber system (designed by Chemico Corp.,
N.Y.), one stage to remove particulates and the other stage to
remove SO2.2 5 A wet-type fan was installed between the two
scrubbers* handling about 500,000 acfm flue gas at 340°F.
Sprays were provided to remove solids buildup on the fan blades
and to wash off acids resulting from scrubber carryover.
The fan housings are lined with 1/4-in.-thick natural
rubber. The wheel material is Carpenter 20 Cb 3, a special
stainless steel containing niobium and tantalum. The fan shaft
is 316L stainless steel. Each fan is driven by a 4160-V elec-
tric motor rated at 5500 hp at approximately 1200 rpm. A closed
system supplies cooling water to the fan bearings.
The outlet gas temperature from the first-stage scrubber is
normally in the 110° to 120°F range. At 175°F, a control valve
automatically opens to admit emergency cooling water to the
upper cone of the venturi. Additional temperature rise auto-
matically shuts down the fan and closes the isolation dampers.
Many fan problems occurred during the trial run. The first
problem was stress on the fan blades. In order to determine the
condition of the fans, Structural Dynamics Research Corporation
(SDRC) conducted a series of strain gauge tests. Its results
indicated that the yield strength was exceeded in several por-
tions of the fan blades and that a degree of metal deformation
was taking place. After additional testing, SDRC recommended
the installation of doubler (reinforcing) plates on each of the
blades to reduce the stress to acceptable levels. After actual
In the Chemico pilot-plant study, an ID fan was used down-
stream from the two scrubbers after a reheater. In the plant
application, however, space limitation dictated the wet-type
fan installation.
4.8-10
-------
exposure tests with stress-welded specimens in the fan atmos-
phere, it was also recommended that the doubler plates be welded
with Inconel 112 rod, rather than with rods of Carpenter 20, 4
NIA, or 8 N12, which were used earlier. Doubler plates were
installed on all fans by the end of 1974 and the results were
satisfactory.
Frequent inspection of the fans revealed significant pit-
ting attack under the fly ash/sludge deposits on the back of the
fan blades and numerous cracks from chloride stress corrosion
around the blade welding spots. Despite the addition of lime
for SO2 removal and pH control, the pH of the fan spray water
dropped from about 6.5 to 2 across the fan. The spray water
seemed to remove more SO2 than was expected, turning spray water
into acid mist. In a trial installation, caustic was added to
the fan spray water, but because large quantities were required
to obtain a pH of 4, the trial was terminated. Installation of
six new Bete Fog type nozzles (No. TF16FC) ahead of the fan gave
good results in removing deposits and reducing pitting attack on
the blades.
At the recommendation of Franklin Institute Research Labor-
atory (Philadelphia, Pa.), the blades were coated with an epoxy-
based, acid-resistant material, Coroline 505AB (Ceilcote Co.,
Berea, Ohio). During operation, the fan blades and the coating
material experienced different stresses, resulting in the fail-
ure of the coating. Prospective fan coating materials were then
tested by exposing coated test plates to the fan atmosphere.
Those tested showed polyurethane rubber to be the most suitable
coating, and one of the fan hubs was coated with that material
for further testing. Bonding failure prevented a full evalua-
tion.
Test specimens of Inconel 625 exposed to the fan atmosphere
showed no indication of corrosion.
4.8.5.6 Bruce Mansfield Power Station—
The design of the Bruce Mansfield FGD system is similar to
that of the Phillips system in that wet fans are installed
between two-venturi scrubbers.1 There are six fans in service,
manufactured by Green Fuel Economizer Co. These are airfoil,
radial-tip units rated at 558,000 acfm with a pressure of 75 in
H20 at 118°F. They are driven by 13.2-kV electric motors rated
at 9000 hp and 1300 rpm. The fan blades are made of Inconel and
shafts are fabricated of carbon steel, clad with Carpenter 20.
The carbon steel fan housing and scrolls are lined with rubber.
problems were experienced with these fans due to high
levels of acidity (pH of about 2) and rubber lining damage
resulting from chips of scale being carried through the system.6
4.8-11
-------
The exposed carbon steel corroded, and the fan scrolls were
replaced with Inconel. Some pitting of the carbon steel hubs
has also occurred as the result of seal leakage.
4.8-:.2
-------
REFERENCES
1. Designing Large Central Stations to Meet Environmental
Standards. Generation Planbook, 1976, pp. 25-34.
2. Isaacs, G.A. Survey of the Flue Gas Desulfurization System
at the Phillips Power Station, Duquesne Light Company. EPA
68-02-01321, June 16, 1975.
3. pernick, S.L., Jr., and R.G. Knight. Duquesne Light Co.,
Phillips Power Station Lime Scrubbing Facility. In:
Environmental Protection Agency Flue Gas Desulfurization
Symposium, Atlanta, November 4-7, 1974.
4. Pernick, S.L., Jr., and R.G. Knight. In: 69th Annual
Meeting of APCA. Boston, June 15-20, 1975.
5. pernick, S.L., Jr., and R.G. Knight. Duguesne Light Com-
pany, Elrama and Phillips Power Stations Lime Scrubbing
Facilities. In: Environmental Protection Agency FGD
Symposium, New Orleans, March 8-11, 1976.
6. Durkee, K.R. Survey of Pennsylvania Power's Bruce Mans-
field Power Generating and Flue Gas Desulfurization System.
Meeting Report, July 14, 1977.
4.8-13
-------
BIBLIOGRAPHY
Steam/Its Generation and Use. Babcock & Wilcox Co., 38th edi-
tion, 1975.
4.8-14
-------
CONTENTS
4.9 THICKENER/CLARIFIER
Page
4.9.1 Introduction 4.9-1
4.9.2 Service Description 4.9-3
4.9.2.1 Slurry pH 4.9-6
4.9.2.2 Calcium Sulfite and Sulfate Sludge
Characteristics 4.9-6
4.9.3 Design Parameters 4.9-8
4.9.3.1 Mechanical Design 4.9-8
4.9.3.2 Use of Flocculant 4.9-17
4.9.3.3 Sizing Criteria 4.9-21
4.9.4 Materials of Construction 4.9-27
4.9.5 Operating Procedure 4.9-27
4.9.6 Existing Facilities 4.9-28
References 4.9-34
Bibliography 4.9-36
4.9-i
-------
4.9 THICKENER/CLARIFIER
This section of the Lime FGD Systems Data Book presents
design information on the thickener/clarifier. The scope of
this section begins at the thickener/clarifier slurry feed
piping and ends at the clarified overflow collection weir and
the suction to the underflow sludge pumps.
4.9.1 Introduction
A thickener/clarifier is a sedimentation device that con-
centrates a slurry under gravity, so that the settled solids may
be disposed of and the clarified liquid recycled. This process
dates back to 1906.
The product of the sedimentation process usually dictates
the terminology. When the primary object is to produce clear
liquid from a dilute suspension, the process is called "clarifi-
cation," and the unit is known as a "clarifier.» When recovery
of the settled solids in the form of a concentrated slurry is of
prime importance, however, the terms "thickening" and "thick-
ener" are commonly used. Although both these operations are
equally important in lime scrubbing systems, the terms thicken-
ing and thickener are used in this discussion.
A general concept of sedimentation may be obtained by
observing some finely divided solids in water in a graduated
cylinder. At the start, the solids are uniformly dispersed
throughout the cylinder [Figure 4.9-1(1)]. When the sedimen-
tation process begins, solid particles begin to sink. The
concentration in the top of the liquid decreases; that in the
bottom increases. A concentration gradient forms and can be
divided into the zones shown in Figure 4.9-1(2). Zone A is a
clear supernatant. Zone B is the slurry zone. Zone C is a
transitional phase between the slurry and the concentrated
sludge, and Zone D is the sludge itself.
The solids in the second and third drawings settle at a
constant rate and the sludge portion (Zone D) increases as the
result of the accumulation of settled particles. The particles
in Zone D continue a slow compaction. This zone is referred to
as a "compression zone."
When the A-B interface finally joins the B-C interface (at
the critical point), Zones B and C disappear [Figure 4.9-1(4)]
A further settling takes place through the "compaction" of the
settled solids, the liquid being pressed out of the floe and out
Of the interstitial spaces between the floe [Figure 4.9-1(5)]
4.9-1
-------
_-_-. g -.-^
"g.ffmyx?v
CRITICAL
POINT
(1)
(2)
(3)
(4)
(5)
10
i
CLARIFIED LIQUOR
F>>>:-3 SLURRY
TRANSITION LAYER
COMPRESSION ZONE
COMPACTION ZONE
Figure 4.9-1. Sequence of sedimentation in a cylinder.
-------
All the operations described take place in the thickener of
a lime slurry scrubbing system. The degree of compression,
compaction, and overflow clarification depends on design vari-
ables, particularly the surface areas and sidewall depth, which
are discussed in this section.
4.9.2 Service Description
The most common type of thickener is circular in design
with a center drive unit for a rake mechanism. This is the type
used by lime scrubbing facilities and is described in this
section. When space is limited, an alternative design called
the high-capacity thickener may be"used. The Lamella plate-type
thickener is an example of this. It is used successfully in the
phosphate industry. This type of thickener was tested at the
Phillips Power Station and the TVA Shawnee facility.
A typical thickener (Figure 4.9-2) consists of a large
circular holding tank with a central vertical shaft. The shaft
is supported either by a center column or by a bridge (Figure
4.9-3). Two long, radial rake arms with the option of two short
ones extend from the lower end of the vertical shaft. Plow
blades are mounted on the arms at an oblique angle and have a
clearance of 1.5 to 3 in. from the bottom of the tank. They can
be arranged so that the bottom is swept either once or twice
during each revolution. The bottom of the tank is usually at a
grade of between 1:12 to 1.75:12 from the center. The settled
sludge forms a blanket on the bottom of the thickener tank and
is pushed gently toward the central discharge outlet. Center
scrapers clear the discharge trench and move the solid deposits
-toward the underflow discharge point.
In normal operations, scrubber slurry is fed through the
feedwell into the thickener at a concentration of 7 to 15 per-
cent solids and at a rate of up to approximately 1000 gal/min,
depending on tank size. The sludge is usually discharged as
underflow at a concentration of 20 to 45 percent solids; if
sulfate is the predominant species, the underflow may be as high
as 60 percent solids.
The use of flocculant can reduce solids concentration in
•the thickener overflow, which would contain 50 to 100 ppm sus-
pended solids without flocculant and 10 to 70 ppm with it.
Higher feed solids concentration, however, will result in higher
suspended solids in the overflow, which is returned to the
scrubbing system for reuse.
in this application, the thickener has a diameter of 50 to
100 ft (though it could be as much as 300 to 400 ft), and a
sidewall height from 8 to 14 ft.
4.9-3
-------
tin
INDICATOR
LAUNDER
ID
I
CENTER DRIVE
UNIT AND
LIFTING DEVICE
DRIVE MOTOR AND
GEAR ASSEMBLY
HIGH PRESSURE BACK-
FLUSHING WATER LINE
CENTER SCRAPERS
WALKWAY
FEED
OVERFLOW
-TORQUE AND
RAKE ARMS
-THIXO POST
PLOW BLADES
DISCHARGE TRENCH
Figure 4.9-2. Thickener supported by a center column, with
truss-type rake arms and Thixo post plow blades.
-------
TORQUE ARM
SOLUTION LEVEL
RAKE ARM
(NORMAL POSITION)
a) COLUMN-SUPPORTED CABLE TORQ THICKENER
(DORR-OLIVER, INC.)
TORQUE ARM
SOLUTION LEVEL
BRIDGE
RAKE ARM
(RAISED POSITION)
b) BRIDGE-SUPPORTED SWING LIFT THICKENER
(EIMCO)
Figure 4.9-3. Hinged rod-type designs with two
types of support structure.
4.9-5
-------
The rake arms are driven by a 2- to 5-hp motor with a worm
gear connection at a period of 10 to 20 rev/min.
4.9.2.1 Slurry pH—
Slurry introduced into a thickener contains dissolved ionic
species such as calcium, sulfite, sulfate, hydrogen, hydroxide,
and chloride ions. Slurry pH ranges between 4 and 11, according
to the predominant ionic species.
4.9.2.2 Calcium Sulfite and Sulfate Sludge Characteristics—
An analysis of major solids in the waste sludge at the
Conesville Station is given in Table 4.9-1. The sulfite/sulfate
ratio can vary between facilities depending on the degree of
sulfite oxidation.
Table 4.9-1. CHEMICAL ANALYSIS OF THE SLUDGE
(DRY BASIS) OF A LIME-SCRUBBING SYSTEM
(percent)
Fly ash
CaS03-l/2 H2O
CaSC-4-2 H2O
CaCO3
MgC03
Total
0.26
72.00
23.95
3.53
0.26
100.00
Without forced oxidation and large quantities of excess air
in the system, the scrubber slurry solids will tend to be mostly
sulfite for high-sulfur coals. If, however, the plant practices
forced oxidation, the sulfate species will predominate. Al-
though some limestone facilities in the United States practice
forced oxidation, the technique has only been employed in the
lime FGD system at the Shawnee steam plant of the Tennessee
Valley Authority.1
Calcium sulfite crystals occur as extremely thin, fragile
platelets about 10 to 100 pm across and 0.1 to 0.5 pm thick
(Figure 4.9-4). They usually appear in clusters, or "rosettes,"
which are stronger than single crystals. The clusters can be
composed of as few as 5 to as many as 100 or more platelets.
They form an open structure, with water filling the voids. As a
result, sulfite sludge is not as easily compacted by settling as
sulfate sludge and typically generates a 35 percent solids
underflow. The water retention of sludges containing sulfite,
as well as the crystalline structure, results in thixotropic
behavior. When shear force is applied, thixotropic material
4.9-6
-------
Figure 4.9-4. Calcium swlfite sludge.
Figure 4.9-5. Calcium sulfate sludge,
Source: Dravo Corp.
4.9-7
-------
becomes less viscous and flows readily. It should be noted,
however, that the term "thixotropic" is not used accurately
here, since thixotropic materials become stable when the shear-
ing force is removed. Flue gas desulfurization sludges tend to
remain in the liquid-plastic form.
The crystalline structure of calcium sulfate is diamond or
rhomboid shaped (Figure 4.9-5) and varies in size from 1 to 100
|jm along each edge. Calcium sulfate crystals have better set-
tling and dewatering characteristics compared with sulfite
crystals. They retain less water and, as a result, high-sulfate
sludges (having more than 60 percent sulfate) are nonthixotropic
and will settle rapidly to produce up to 60 percent solids
underflow.
4.9.3 Design Parameters
A thickener consists of several pieces, not all of which
are supplied by all thickener vendors. Parts that are normally
purchased from thickener vendors include the center drive unit,
torque control and alarm system, rake arm mechanism, lifting
device, lift indicator, feedwell and internal piping, anchorage
ring or bottom flange for structures with center-column support,
and the walkway.
Vendors also provide design specifications. They are not,
however, normally responsible for the design and construction of
the tank, nor do they provide items such as external piping, the
electrical conduit and wiring, weirs, baffles, and overflow
launders. The walkway beam supplied by thickener vendors is
usually designed to carry only the walkway itself, live loads,
and the handrail. Additional weight for piping or external
equipment is excluded.
4.9.3.1 Mechanical Design--
The major components of a thickener include the center
drive unit, feedwell, rake arm mechanism, and flow arrangements.
Design considerations for components and associated auxiliary
equipment are discussed in this section.
Center drive unit—The center drive unit is the workhorse
of the rotating rake arm mechanism. Choosing the right one from
several available designs depends on the size of the thickener
and the torque requirements of the thickened solids. The center
drive unit consists of a primary worm gear motor, an intermedi-
ate worm gear reduction box with torque control and alarm sys-
tem, and a main spur gear (Figure 4.9-6).
4.9-8
-------
SECONDARY
SPUR GEAR
ELECTRO-MECHANICAL
TORQUE CONTROL
AND ALARM SYSTEM
CAST STEEL CASING
DUAL BEARINGS
Figure 4.9-6. Center drive unit (EIMCO).
4.9-9
-------
when the rake arms stir the sludge, it resists the movement
and the torque is transferred to the center drive unit. The
normal running torque requirement of a thickener is empirically
expressed as:
I = KD2 (Eq. 4.9-1)
where T = torque, ft-lb
D = diameter of a thickener, ft.
K = constant.
In lime scrubbing systems, K would range from 15 to 20 ft-lb/
ft2. For example, the original design torque at Conesville
Power Station (Columbus and Southern Ohio Electric Company) was
150,000 ft-lb for a 100-ft-diameter thickener (t = 15 x 1002
ft-lb). When ordering a center drive unit, the design engineer
should specify peak torque in addition to running torque; peak
torque is generally twice the running torque, but must not
approach the yield strength of the metal.
During the thickener operation, the settled solids may bed
in around the rakes and cause excessive torque (see Section
4.9.5 for further details). A torque overload control is,
therefore, an important auxiliary for the protection of the
drive unit. Three types of control are available: hydraulic,
electromechanical, and electrical. Each can be adjusted to
sound an alarm and stop the equipment at a predetermined load
limit.2 At the Cane Run Station, for example, the overload
control sets off an audible alarm at 80 percent of maximum
design running torque and cuts off the motor at 90 percent.3
One or more torque control units can be used simultaneously and
coupled with a visual torque readout.
Feedwell—The feedwell, a crucial component in the tank
center, quiets the incoming flow before it enters the tank.
Current designs dissipate high inlet velocity by creating small
eddies and a radially uniform flow.
In a well-designed feedwell, solids settle rapidly with
minimal influence from turbulence.
Rake arm mechanism—The main function of the rake arms is
to move the settled solids to a central discharge point.
Bridge-supported and column-supported designs generally employ
two long arms with the option of two short ones, the latter
added when necessary to rake the inner area.
Rake arms must be strong enough for the required torque,
although their design also depends on the nature of the solids.
They come in two main types: a truss-type and a rod-type.
4.9-10
-------
The truss-type arm is a conventional design (Figure 4.9-7a)
that also acts as a torque arm. It is either bolted or welded
to the center column and turned by a center drive unit. Blades
are bolted to the bottom chord of the truss structure. When
lime-generated sludges reach high density and high viscosity,
the blades and the rake arm structure tend to form the sludge
into an immobile "donut" or "island." Proper blade angles and
extension of the blades below the raking arms by means of ver-
tical posts (e.g., Thixo posts, Figure 4.9-7b) help eliminate
this phenomenon by leaving fewer structural members to move
through the sludge in the critical center area. In larger
thickeners (300- to 400-ft diameter), the center depth would be
excessive if a single slope were employed. The double sloping
design (Figure 4.9-7c) eliminates this problem. This design
leaves the truss somewhat above the thickest sludge, which forms
near the center of the tank.
The rod type of rake arm has a hinged design (Figure
4 9_3). The arm is simply a long pipe pivoted at the center
column either by a single tilted pin (Swing Lift, EIMCO) or by
two pins, one horizontal and one vertical (Cable Torq, Dorr-
Oliver). The rake arms are suspended and dragged by cables from
torque arms, which rotate 30 to 45 degrees ahead of them. The
interesting feature of this design is that the pivot pins allow
the rake arms to swing over heavy deposits of coarse material.
This design has further advantages over the truss-type in that
it also offers lighter dead load, less drag, and less tendency
for scale buildup in the rake arms. Heavy deposits, however,
usually occur near the center of the thickener, and only a
minimum amount of improvement is experienced at this point
whatever the type of construction.
The size of the rod-type thickener is limited to 150 ft,
since even a small lifting action near the center creates exces-
sive swing at the tip of the rake arms. The lack of positive
control of the lifting and lowering action is another problem
with this design.
Lifting device—Truss-type thickeners are installed with
some type of device for lifting the arm. This device permits
continuous operation without excessive torque by lifting the
rakes over coarse, settled solids and lowering them as these
solids are removed. Hinged rake arm designs do not require a
separate lifting device, because of the unique design mentioned
earlier. It should be noted, however, that this design does not
have positive control of the lifting action.
The truss type of lift can be activated automatically by
the torque encountered by the arms, either by means of an am-
meter on tne mecnanism drive motor, or by a mechanical torque
indicator. When excessive torque is experienced, the entire
4.9-11
-------
(a) CONVENTIONAL DESIGN
(b) THIXO POST DESIGN
(c) DOUBLE SLOPE DESIGN
Figure 4.9-7. Plow blade designs shown with truss-type rake arms.
4.9-12
-------
mechanism is raised evenly. One lift of this type is shown in
Figure 4.9-8. Lifting devices have been employed with up to 5
ft of lift, although 1 ft to 2 ft is more common. Indicators
show the degree of arm lifting (Figure 4.9-2). Some vendors
also offer a lift height sensing device for remote indication of
rake arm location.
Flow arrangements — In a thickener, the three primary flow
arrangements are as follows:
(A) Feed - Slurry suspensions are usually fed to the tank
through a pipe under the walkway (Figure 4.9-2). In some
designs, the influent pipe runs under the tank and the feed
enters through the center column (Figure 4.9-9).
(B) Overflow - Clarified effluent is usually handled by a
peripheral launder at the upper tank edge. V-notch over-
flow weirs help to distribute the internal flow patterns
evenly within the tank (Figure 4.9-2).
(C) Underflow - Buried or inaccessible sludge lines incur the
lowest installation cost of all underflow arrangements.
Usually several of these lines are installed (vendors
recommend a minimum of two), with one kept as a spare in
case of plugging. A high-pressure water or air line should
be available to clear the plugging.
One underflow design consists of a tunnel (Figure 4.9-9) to
allow complete accessibility to the discharge point, valves, and
associated piping. Although the most costly approach, it
handles the maximum underflow solids concentrations with minimum
maintenance problems.
Center-column pumping (Figure 4.9-10) lowers installation
costs and can be used where ground conditions preclude a tunnel.
Outlet pipes from the discharge trench end in one central verti-
cal pipe, which passes up inside the center column, along the
walkway, to a pump situated outside the tank perimeter (Figure
4.9-10a). A similar system (Figure 4.9-10b) consists of a
double pipe, which permits removal of the inner pipe for clean-
ing- Another system has a submersible pump near the bottom of
-the center column (Figure 4.9-10c). with this arrangement, long
suction lines are avoided, and underflow concentrations ap-
proaching those of the tunnel system can be handled. Although
jt is more expensive, the system will pump higher solids concen-
trations with less probability of plugging or loss of prime. In
-the column pumping operation, however, spare pumps and a pump
hoist are recommended.
4.9-13
-------
MAIN GEAR
AND BASE OF
DRIVE UNIT
COLLECTOR RING
ASSEMBLY
SERVICE PLATFORM
CENTER COLUMN
LIFTING MOTOR
BEARING PLATE
DRIVE CAGE
LIFTING SCREW
LIFTING CAGE
RAKING ARMS CONNECT
TO LOWER CAGE
Figure 4.9-8. Positive lifting device used on
column-supported thickeners.
4.9-14
-------
FEED
ARMS
CONE SCRAPER
DISCHARGE CONE
DRIVE HEAD
FEEDWELL
FEED
UNDERFLOW
WALKWAY
WATER LEVEL
CENTER COLUMN
TUNNEL
UNDERFLOW
DISCHARGE
Figure 4.9-9. Standard tunnel system.
4.9-15
-------
HIGH-PRESSURE
WATER LINE
vo
I
M
O1
UNDERFLOW DISCHARGE LINE
LOCATED ON TOP OF WALKWAY
DRIVEHEAD
HATER LEVEL
CAGE AND
ARMS
UNDERFLOW TO
CENTRIFUGAL PUMP
UNDERFLOW LINES
CENTER
COLUMN
UNDERFLOW
TOP OF DRIVE UNIT PUMP
OUTER PIPE CENTER
COLUMN
REMOVABLE
UNDERFLOW
PIPE
* STEEL
CENTER
COLUMN
ARM
CONE SCRAPER
-DISCHARGE CONE
/J7!
DISCHARGE TRENCH
a) SINGLE PIPE
b) DOUBLE PIPE
c) SUBMERSIBLE PUMP
Figure 4.9-10. Underflow pumping arrangements through center column.
-------
4.9.3.2 Use of Flocculant—
Many of the solid particles in the slurry are dispersed as
a stable colloidal sol. Because the particles are so small
(Table 4.9-2), their surface area is large in relation to their
mass. As a result, surface phenomena predominate and control
the behavior of the sol.4'5
Table 4.9-2. TYPICAL SIZE DISTRIBUTION
OF A LIME-GENERATED SLUDGE
Equivalent spherical diameter, |jm
+18
-18
-12
- 9
- 8
- 6.4
- 5
- 4
- 3
- 1.5
Cumulative wt. percent
10
90
80
70
60
50
40
30
20
10
There are two types of colloidal sols, lyophobic and lyo-
c. Their properties are listed in Table 4.9-3. Colloidal
sols formed in FGD systems are usually lyophobic and in the
subsequent text only these sols will be discussed.
"Stability" refers to the inherent property of colloidal
particles to remain dispersed despite passage of time; whereas
"instability" describes the tendency of particles to coalesce
whenever particle-to-particle contact is made.
The stability of particles in the sol is due largely to the
phenomenon of the electrical double layer, consisting of a
charged-particle surface and a surrounding sheath of oppositely
charged ions. Several theories have been advanced to describe
quantitatively the concept of the electrical double layer.
Helmholtz proposed this first in 1879; it was subsequently
modified by a number of workers. in its simplest form, the
-theory states that particles in a colloidal sol have electrical
charges on their surfaces.
The charge on all particles in a given dispersion is the
saine (either positive or negative) but can vary considerably in
magnitude. It depends on the nature of the colloidal material.
The electrical double layers around particles inhibit their
close approach to each other. In this way the double layers
4.9-17
-------
Table 4.9-3. SUMMARY OF THE CHARACTERISTIC DIFFERENCES
BETWEEN THE TWO CLASSES OF COLLOIDS.6
vo
I
M
00
Lyophobic
(No affinity for solvent)
Low viscosity.
Normally irreversible.
Particles show Tyndall effect, i.e.,
easily detected by ultramicroscope.
Particles are all positively or all
negatively charged, i.e., they move
in one direction under influence of
applied electromotive force.
Particles coagulate and are precipi-
tated on addition of electrolyte.
Examples:
Metals, sulphur, metal sulphides,
hydroxides (Fe(OH)3,Al(OH)3>.
Lyophilic
(Considerable affinity for solvent)
High viscosity.
Reversible, i.e. will become colloidal
again, after coagulation, on addition
of solvent.
Particles not easily detected.
Particles move in both directions
or not at all.
No precipitation by electrolyte un-
less added in large quantities, when
"salting out"" may occur. Give pro-
tection to lyophobic colloids from
precipitating effect of electrolytes.
Examples:
Starch, gum, soaps, haemoglobin,
egg albumin, gelatin, agar agar; i.e.,
organic substances of high mole-
cular weight.
-------
confer stability to the sol. This property inhibits precipita-
tion of solid particles, and the settling velocity is often too
low to design an economical thickener with good overflow qual-
ity. It is necessary, therefore, to destabilize colloidal sols
by reducing the forces that keep the particles apart (coagula-
tion), and agglomerating the small particles into larger aggre-
gates (flocculation), which have the settling velocities re-
quired to be practical. This is done by adding chemical condi-
tioners to the colloidal sols.
Inorganic metal salts and polymers are among the chemicals
that can be used for conditioning. The Hardy-Schulze Law states
that the coagulating power of an electrolyte depends on its
valency. Thus, soluble ferric or aluminum compounds (3-valent)
are better coagulants than soluble ferrous compounds (2-valent)
and far much better than 1-valent compounds.
Both inorganic and polymer conditioners agglomerate fine
particles by neutralizing surface charges, thereby accelerating
settling, and clarifying thickener overflow.
Polymers possess less neutralizing power but a higher
capacity for "bridging" (simultaneous attachment) to two or more
solid particles than do inorganic coagulants. They also provide
the advantage of a very large increase in the size of the floe,
which greatly increases their settling rate. Dorr-Oliver
Laboratories (Stanford, Connecticut) performed a series of
settling tests to study the effect of polymer dosage.7 Figure
4.9-11 shows that the rate is increased 7 to 20 times, depending
on polymer dosage.
Flocculation, however, yields a more dilute underflow (25
to 35% solids from lime-generated slurry) because of the bound
water in the interstitial structure of the polymers. The cor-
rect polymer dose should be determined by trial-and-error
through electrical potential measurement (zeta potential) and
jar tests, since an incorrect dose may cause side effects such
as increased overflow turbidity, slower settling rate, or under-
flow plugging.
In general, a thickener using polymers need only be a third
or half as large as one without flocculation.
Most polymers are bought in powder form and may be stocked
£or use at strengths of 0.5 to 1 percent. Stock solutions may
be prepared as infrequently as twice a week, but they are usu-
ally made fresh every day and, in some cases, once per shift
The solution preparation system includes a manual or automatic
blending system in which polymer is dispensed by hand or bv a
feeder. . •*
4.9-19
-------
oo
CO
0.25 0.50 0.75 1.0
POLYMER (BETZ 1120) DOSAGE, ppm
1.25
Figure 4.9-11. Effect on polymer on settling rate of lime-
generated sludge.
Source: Dorr-Oliver
4.9-20
-------
A schematic of a typical automatic dry feed system is
presented in Figure 4.9-12. At Cane Run Power Station, a Penn-
walt dual tank polyelectrolyte feeding system adds Betz 1100
polyfloc to the thickener tank in a 0.5 percent solution.9
This system consists of a reciprocating screw feeder that
has an adjustable stroke and operates at constant speed It
volumetrically meters dry polymer into two stainless steel
wetting cones. These cones are provided with washdown water at
controlled pressure and each one is connected to a small hiah-
velocity stream eductor in which the particles are individually
wetted and dispersed. The combined discharge is passed to an
open cylindrical tank, equipped with level controls and a slow-
speed propeller mixer in which blending is completed. All water
for the preparation process passes through a Varea-Meter so that
an accurate water-to-polymer ratio may be set. The blended
polymer requires a 15- to 30-minute aging period, because poly-
mer chains in the dry product tighten like a wound spring and
need time to unwind in the solution. when aging is complete,
the slow-speed mixer is stopped and the batch awaits a demand
signal, which depends on the need of the metering tank.
The aged polymer solution is metered by a diaphragm pump.
As this pump lowers the level in the metering tank, a level
control switch is actuated. This initiates the operation of a
screw-type transfer pump, which withdraws solution from the
aging tank and refills the metering tank. A signal that the
level in the aging tank is low begins preparation for another
batch, and the cycle is repeated.
Because viscous polymer solutions require dilution before
application to processes, the system is provided with a second
Varea-Meter and a valve so that dilution water flow may be
observed and controlled. The metering pump discharge connects
into the dilution water line and mixing occurs en route to the
point of application.
Controls are mounted in a steel cabinet at one end of the
system platform. An interlock prevents operation of the trans-
fer pump until the batch has aged.
4.9.3.3 Sizing Criteria—
This section describes methods of determining the surface
area and the side water depth of a thickener.
Surface area—One method of determining the surface area of
a thickener is to use the result of a laboratory settling test
with a representative sample of slurry. The characteristics of
a slurry cannot be predicted, however, because quality varies
greatly from plant to plant, as do ash content and' chemical
4.9-21
-------
SOLENOID
VALVE
SCALE
HOT
WATER
DISPENSER
.SLOW-SPEED
MIXER
.LEVEL PROBE
BLENDER/ METER
NOTE: CONTROL AND INSTRUMENTATION
WIRING IS NOT SHOWN
I
to
to
SOLUTION FEEDERS
POINT OF APPLICATION
DISPENSER
SLOW-SPEED
MIXER
LEVEL
PROBE
MIXING-AGING
TANK
MIXING-AGING
TANK
-Cx»-
V TRANSFER PUMP
LEVEL
PROBE
I r
HOLDING TANK
Figure 4.9-12. Typical automatic dry polymer feed system.
-------
constituents. Because of this, an empirical method is fre-
quently used in designing a thickener for a new plant.
(A) Method I - If a representative slurry sample is available,
a settling test can be performed as follows: a column is
filled with a suspension of solids of uniform concentration
(C ) to height (H ). Figure 4.9-13 shows the position of
the interface as tne suspension settles. The rate at which
the interface is subsiding at any given time is then equal
to the slope of the curve at that time. When flocculation
is used, polymers can be tested to determine the type and
dosage.
The space required by the thickener can be determined
by several techniques using the settling curve. One tech-
nique is the direct calculation technique, i.e., direct
calculation of the unit area from a point on the curve in
the following equation:
A = C H (Eq. 4.9-2)
o o
where A = unit area, ft2/ton per day of dry solids
C0 = initial concentration, tons dry solids/ft3
HQ = initial height of solid suspension in the
column, ft
tu = time' days required for the interface to
reach the height (H ), under which the
average concentration is the same as the
desired underflow solids concentration (C ).
Since HQ CQ = Hu Cu (Eq. 4.9.3)
t is obtained from the settling curve at H = H . When C
is expressed in weight percent, Equation 4.9-31 becomes:
Co Ho PO = Cu Hu PU
where C^; C^ = weight percent of feed and underflow
solids concentrations
po; pu = densitY of feed and underflow,
tons/ft3, or g/cm3
Rearranging Equation 4.9-4, H becomes:
C1 H p
H - o o Ho
u ~ C1p(Eq. 4.9-5)
u Hu u -
4.9-23
-------
Figure 4.9-13 presents the result of a settling test made
by EIMCO on a slurry sample from Bruce Mansfield plant.
For this sample, H = 1.45 ft, p = 1.0386 g/cm3 , p =
1.206 g/cm3, C' = 6/7 wt.%, and C^ =30 wt. %.
Substituting tnese values in Equation 4.9-4:
6.7 wt. % x 1.45 ft x 1.0386
u 30 wt. % x 1.206
= 0.279 ft
From Figure 4.9-13, t is 104 min (or 0.0722 day). Since
the feed concentration^ C , is 0.002172 ton dry solids/ft3,
the thickener area from Equation 4.9-2 becomes:
A =
Co Ho
0.0722 day
0.002172 ton/ft3 x 1.45 ft
= 22.93 ft2/ton per day of dry solids
A variation of this technique has been modeled by
Kynch and modified by Talmadge and Fitch for suspensions
with hindered settling characteristics.10'11 The thickener
area is expressed as:
A' = --* (Eq. 4.9-6)
where A = unit area, ft2
Q = volumetric flow rate, ft3 /day
t = time in days required for the interface
X to reach the height (HU) under which the
average concentration is the same as
the desired underflow solids concen-
tration, (C ) .
Since the feed solids concentration is C ton per ft3 , the
solid loading Q1 is
Q1 (tons/day) = Q x CQ (Eq 4.9-7)
e^
Rearranging Equations 4.9-6 and 4.9-7 yields
A = — = c XR ft2 /ton per day (Eq. 4.9-8)
Q1 o o of dry solids
4.9-24
-------
1.5
H0=1.45 ft
DATA TAKEN BY EIMCO AT BRUCE MANSFIELD PLANT
FEED SOLIDS*6.7 wt. %
(NO POLYMERS)
UNDERFLOWS wt. %
1.0
I
0.5
Hu
Cu-30 wt. % SOLIDS
J L
J
0 20
40 60 tx 80 lOQtu
SETTLING TIME. mln.
12Q 140
Figure 4.9-13. Graphical analysis of interface
settling curve.
4.9-25
-------
The main difference between the two techniques is in
determining the settling time. The time, t , can be deter-
mined as follows:12 Determine the concentration C2 by
extending the tangents to the hindered settling and com-
pression settling regions of the subsidence curve to the
point of intersection, and then bisecting the angle thus
formed as shown in Figure 4.9-13. Construct a tangent to
the settling curve at the point indicated by C2 . The time
axis of the intersection of this tangent line with the
horizontal line passing through H = H is t . For this
example, tx is 71 min (or 0.0493 day) rrom Figure 4.9-13.
The thickener area can be calculated from Equation 4.9-8.
Usually a 1.25 to 1.33 scale-up factor is used to correct
the full-scale unit from the above test method. Therefore,
A = -£—|— x 1.33 (Eq. 4.9-9)
o o
0.0493 day
0.002172 ton/ft3 x 1.45 ft X
= 20.82 ft2/ton per day of dry solids.
(B) Method II - When representative sludge is not available,
the empirical method of determining the area of a thickener
is based on either the solids loading or the hydraulic
surface loading, whichever yields a greater size.13 Exper-
ience indicates, however, that solids loading generally
governs the design.14 Surface loading rates can be ob-
tained from the operating experience of similar processes.
These values, however, vary widely, depending on the degree
of sulfite oxidation and the use of flocculants and other
process design variables.
Recently, EIMCO conducted a series of sludge settling
tests using samples from operating plants. Thickener unit
areas were calculated with the results of these tests as
outlined in Method I. The areas, without polymer, ranged
from 12 to 30 ft2/ton per day for 25 to 35 wt. percent
solids in the underflow for one plant, and from 8 to 12
ft2/ton per day for 35 to 45 wt. % solids for another
plant. When polymers were used, the areas were reduced to
3.5 to 11 ft2/ton per day and 1.7 to 4.5 ft2/ton per day,
respectively.
Temperature also has an effect on thickener sizing.
One study showed that an increase in temperature from 60°
to 110°F resulted in a twofold reduction in thickener unit
area requirements.
4.9-26
-------
Thickener vendors usually have their own values for
the surface loadings from their design experience with
similar operations. The trend seems to be to design a
thickener without polymer and then to adjust the operation
(as discussed in a later section of this report).
Depth—After determination of the area of the thickener,
the vendor calculates the side water depth from design experi-
ence and the size requirements of the feedwell and the rake arm
mechanisms. The residence time for liquor in the thickener,
which is the main factor affecting overflow clarity, is calcu-
lated from the side water depth and thickener area. The side
water depth is often adjusted to give the desired residence
•time,15 the most common depths for thickener tanks of 50- to
100-ft diameter being from 8 to 14 ft.
4.9.4 Materials of Construction
In lime scrubbing facilities the thickener is usually a
lined or painted mild steel tank with a steel or concrete bot-
tom.
The feedwell and launder are also made of mild steel. The
gear mechanisms are of heat-treated steel and alloys such as
bronze or alloy steel. To facilitate lubrication, all gear
components are normally enclosed in a dust-tight, cast iron
housing.
Since the pH of the solution in the thickener tank varies
and chloride content may be high, most parts submerged below the
solution level are protected from corrosion (including chloride
stress corrosion) by an epoxy or rubber coating. The rubber
covering operation, however, limits the size of a thickener to a
maximum of about 130-ft diameter due to the size limitation of
commercial vulcanizers.
Some thickener vendors use special materials for certain
parts, because they show better resistance than stainless steel
to corrosion by chloride ions. For capital cost savings on tank
wall construction, mild steel with corrosion allowance is an
alternative material. The general section on corrosion (4.12)
provides further information.
4.9.5 Operating Procedure
The design of a thickener largely depends on experience and
empirical data, which are not always available. As a result,
the thickener is often incorrectly sized.
The retention time of an undersized thickener is shorter
than the time required by solid particles to settle down for a
4.9-27
-------
normal feed rate. Insufficient settling will result in a high-
solids content in the overflow and low underflow densities. If
this happens, a proper dosage of polymer can increase the set-
tling rate.
On the other hand, if the thickener is oversized, or if
low-solids loading is practiced because of plant load changes,
retention time increases. The settled solids may become com-
pacted and cause excessive torque. The proper action in this
case is to stimulate a higher feed rate by recycling part of the
underflow back to the thickener feedwell.
Another important aspect of the thickener operation is the
shutdown process. It is not normally advisable to shut down the
thickener for extended periods with appreciable amounts of
solids still in the tank. If the bulk of solids inventory in
the thickener is not removed before extended shutdowns, settled
solids may bed-in around the rakes, causing high torque and
extreme difficulty in restarting. Thickener underflow, there-
fore, should be pumped until it becomes thin. If the thickener
has a lifting device, the mechanism should be raised and, if
possible, rotated during the shutdown period. If it is not
possible to pump out settled solids before shutdown, an alterna-
tive procedure is to recirculate underflow back to the thickener
feedwell while the rakes are rotating in the lower position. It
may also be possible to reduce this recirculation rate below
that employed for pumping the thickener underflow during opera-
tion. The amount recirculated need only be sufficient to pre-
vent severe compression of the solids, which in turn might cause
excessive torque.
After an extended shutdown during which the rakes have been
raised, they should be lowered gradually to avoid excessive
torque loads. If appreciable solids have bedded-in under the
raised rakes, the mechanism should be lowered slowly enough to
avoid exceeding the recommended maximum operating torque. In
this manner the blades will slice into the bed as the rakes are
lowered, fluidizing the mass so that it will flow and be dis-
charged. The time necessary for lowering the rakes could range
from a few minutes to several hours, depending on the solids
level.
4.9.6 Existing Facilities
The use of the thickener in lime-scrubbing FGD systems is
popular; eight of nine plants surveyed in the United States use
thickeners. Design and operating data on these facilities are
presented in Table 4.9-4.
Performances are satisfactory and no major problems have
been reported. Minor problems include underflow plugging, the
result of such things as hard hats or welding rods being dropped
into the thickener.
4.9-23
-------
Table 4.9-4. EXISTING THICKENER FACILITIES FOR LIME SCRUBBING FGD SYSTEM
to
Name
and
Location
Paddy's Run
No. 6, Rubber town
Kentucky. LGiE
Cane Run
Louisville, Ky.
LGfcE
Conesville No. 5
Conesville, Ohio
Columbus fc
Southern Elec.
Phillips Power
Sta., Cresant
Township, Pa.,
Duquesne Light Co.
Elrama
Duquesne Light Co.
Bruce Mansfield
No. 1
Shippingport, Pa.
Penn. Power Co.
Mohave
Southern Cal.
Ed i ion
Four Corners - 5A
Arizona Public
Service
Thickener
Dimension, ft
SO D x 13 H
Dorr-Oliver
8S D x 14 H
EIMCO Type B
Swinglif t
14S D x 16 H
Dorr-Oliver
Cable Torq
2 units
75 D x 14 H
Dorr-Oliver
Cable Torq
2 units
. 120 D x 8.5 H
(side) Dorr-Oli-
ver Cable Torq.
1 unit
200 D x 13.5 H
Koppers Co.
60 D x 12 H
EIMCO Truss-type
100 D
EINCO
Material
of
Construction
Mild steel Tank
3/16-in. Natural
rubber-covered
carbon steel tank
316 SS fittings
Rubber-covered
carbon steel wall
concrete bottom
Rubber-covered
Monel cables
Rubber-covered
Monel cables
Rubber -covered
carbon steel
tank
Flocculant
Betz 1100 (anionic)
5-7 ppm
Betz 1100 (anionic)
5-7 ppm
Nalco 676 (anionic)
Betz 1120 (anionic)
1-2 ppm
Floe. -type unknown
None
Nalco
None
Operating
Conditions
10% in
19-25% out
pH - 7.8-9. 130T
10% in
17% out
pH 7.8-9
130?F
15% (max) in
30-40% out
pH - 6-7
125BF
5-10% in
35-40% out
pH - 6-8
110»F
5-15% in
35-40% out
pH » 6-7
160'F (max)
10% in
30% out
pH - 8-9
100T
7% in
pH » 7
130'F
1% in
pH - 7-9
-------
All the thickeners operate with flocculants, except at the
Bruce Mansfield plant. This facility has a Koppers Co. thick-
ener, 200 ft in diameter and 12 ft high. It is equipped with a
Pennwalt polyelectrolyte feed system. The underflow achieves 25
to 30 percent solids concentration without flocculants. Truss-
type arms are supported by a center column with an automatic
lifting device. A single pipe in a tunnel is arranged for
underflow discharge and a 120-psig water line provides back
flushing when needed. Improper lubrication caused high torque
in cold weather. This has been corrected. A flocculant testing
program is currently in progress.
Paddy's Run No. 6 unit is operating a Dorr-Oliver thicken-
er, 50 ft in diameter and 13 ft high, at 100° to 130°F with a
retention time of 4.3 hours. This unit has bridge-mounted,
truss-type rake arms with a hydraulic lifting device. Plow
blades are mounted on Thixo posts, and the underflow has a
multiple piping arrangement with a 40-psig back flushing water
connection. This thickener was originally too small to handle
untreated slurry at full load, since the slurry contained a
higher ratio of calcium sulfite to calcium sulfate than expect-
ed. The use of Betz 1100 Polyfloc improved the settling rate
sufficiently and a larger thickener was unnecessary. Flocculant
is injected into the thickener at a rate sufficient to maintain
a concentration of 5 to 7 ppm. Lime is also added to the thick-
ener tank to stabilize the sludge and is consumed at a rate of
about 100 Ib/ton of dry sludge generated. The feed rate is 200
gal/min at full load. The overflow is 120 to 170 gal/min with
less than 0.25 percent suspended solids.
The EIMCO Type B, Swinglift thickener (Figure 4.9-14) at
Cane Run is 85 ft in diameter, 14 ft high, and produces 176
gal/min of underflow at 17 percent solids concentration from a
feed of 300 gal/min (at full load) of 10 percent solids. The pH
ranges from 7.8 to 9 and all the submerged parts are rubber
covered. A 0.5 percent solution of Betz 1100 Polyfloc is pre-
pared in a feeding system (described in Section 4.9.3.2) and
added to the thickener to make a concentration of 5 to 7 ppm.
The underflow has multiple discharge piping in a tunnel with two
pumps and a 60- to 65-psig back-flushing water connection. The
overflow rate is 250 to 260 gal/min, with less than 0.25 percent
suspended solids.
Phillips Power Station installed two units of "Cable Torg"
thickeners (Dorr-Oliver), 75 ft in diameter and operating at a
pH of 6 to 8. The feed rate is 800 to 1300 gal/min with 4 to 10
percent solids. The underflow solids concentration is 35 to 40
percent. The slurry is flocculated with 1 to 2 ppm of Betz 1120
to aid settling. The underflow has multiple discharge piping in
a tunnel with a 90-psig backflush water connection. The over-
flow has less than 1 percent suspended solids. This station is
4.9-30
-------
VD
I
UJ
DRIVE CONTROL
H/LOAO INDICATOR
FEEDWELL SUPPORTS
VEEDPIPE AND SUPPORT!
-(NOT BY EWCO)=
LAUNDER AND WEIR
(NOT BY EIMCO)
ARMS AND BLADE
SUPPORT CABLES
ARMS AND BLADES RAKE TANK
BOTTOM TWICE PER REVOLUTION
DISCHARGE PIPE
CONCRETE AND STEEL '
CONE SCRAPER 316 SS TANK (NOT BY EIMCO)
Figure 4.9-14. EIMCO swinglift thickener at Cane Run Power Station.
-------
currently adding one more thickener by Denver Equipment Co.
This 75-ft-diameter thickener is a bridge-supported unit with an
automatic lifting device. The truss-type rake arms have blades
mounted on Thixo posts. This station tested the Lamella plate-
type thickener. The test showed a good overflow quality, but
there were some problems with sludge deposits near the bottom of
the inclined plates. The design was changed and this type of
thickener was tested at Shawnee demonstration plant. Some of
the findings were: it was efficient, required shorter residence
time than a conventional thickener, and generated good overflow
clarification. Insufficient data are available, however, to
evaluate a full-scale application in FGD systems.
The thickeners at Elrama Station are 120-ft diameter, Cable
Torq models by Dorr-Oliver. They are column-supported units.
Operating conditions are the same as at Phillips Station. This
station is currently adding two more thickeners, the same as
those described at Phillips Power Station.
Conesville No. 5 unit has one Dorr-Oliver, Cable Torq
thickener, 145 ft in diameter and 16 ft high (Figure 4.9-15).
The walls and launder are steel and the bottom is concrete. The
feed rate is approximately 1200 gal/min with a maximum 15 per-
cent solids. The thickener operates at 125°F and produces 25 to
40 percent underflow. The specific gravity of the sludge is
1.2. A 0.3 percent polymer solution is prepared with a BIF
polymer feeding system. Polymer consumption (Nalco 676) is 100
to 175 Ib/day. A single underflow discharge line has a tilted
design and a 70-psig backflush water connection. However,
solids settled in the discharge pipe and caused plugging.
Underflow pumping rate was therefore increased. The overflow
has about 1 percent suspended solids.
The bridge-mounted thickener at Mohave (EIMCO truss-type,
60 ft in diameter and 12 ft high) employed multiple underflow
discharge piping with both water (200 psig) and air (100 psig)
back-flushing provisions. The unit stood 10 ft above grade to
allow access to the underflow piping.
Four Corners Unit 5A used an EIMCO thickener (100 ft in
diameter) with an automatic lifting device.
It also had a tunnel for access and a single discharge pipe
with two underflow pumps and a water backflush line.
4.9-32
-------
CO
CJ
RAKE ARM WITH BLADES
(14 In. SCHEDULE 10 PIPE)
UPPER FEEDMELL
LOWER FEEDMELL
- UPPER FEEWELL L 8 ln- FE«> PII>E
-PVC DOHNCOMER PIPE
145 ft. OU.
LIFT INDICATOR11 ' ft-
11 ft 3 in FEEDUELL
31 ft. 3 in.
KLUDGE DISCHARGE PIPES
Figure 4.9-15. Dorr-Oliver cable torq thickener at Conesville Power Station.
-------
REFERENCES
1. Crowe, J.L., G.A. Hollinden, and T. Morasky. Status Report
of Shawnee Cocurrent and Dowa Scrubber Projects and Widows
Creek Forced Oxidation at EPA Industrial Briefing.
Raleigh, North Carolina, August 1978.
2. Dorr-Oliver Thickeners. Bulletin No. THIC-1, Dorr-Oliver,
Inc. 1969.
3. Enviro-Systems Division, Zurn Industries, Inc. Design
Information of a Thickener. Private communication, June
1977.
4. Cohen, J.M., and S.A. Hannah. Coagulation and Floccula-
tion. Chapter 3, Water Quality and Treatment, A Handbook
of Public Water Supplied, 3rd Ed. McGraw-Hill Book Co.,
New York. 1971.
5. Sawyer, C.N., and P.L. McCarty. Chemistry for Sanitary
Engineers. 2nd Ed., McGraw-Hill Book Co., New York, 1967.
6. Goddard, F.W., and E.J.F. James. Elements of Physical
Chemistry. 4th ed. Longmans 1967. p. 462.
7. Private communication with H. H. Oltmann, Dorr-Oliver, Inc.
October 1977.
8. Process Design Manual for Suspension Solids Removal. EPA
Technology Transfer, EPA 625/l-75-003a, January 1975.
9. Operating Manual for Cane Run Power Station.
10. Kynch, F.J. A Theory of Sedimentation. Trans. Faraday
Soc., 48:161, 1952.
11. Talmadge, W.P., and E.B. Fitch. Determining Thickener Unit
Areas. I&EC, 47(1): 38, 1955.
12. Metcalf & Eddy, Inc. Wastewater Engineering, Collection,
Treatment, Disposal. McGraw-Hill Book Co., New York, 1972.
4.9-34
-------
13. Process Design Manual for Upgrading Existing Wastewater
Treatment Plants. EPA Technology Transfer, EPA 625/1-71-
004a, October 1974.
14. Schroepfer, G.J., and N.R. Ziemke. Factors Affecting
Thickening in Liquid Solids Separation. National Institute
of Health, Sanitary Engineering Report No. 156s, March
1964.
15. Private communication with J. Wilhelm. EIMCO Process
Machinery Division of Envirotech, October 1977.
4.9-35
-------
BIBLIOGRAPHY
Cornell C.F. Liquid-Solids Separation in Air Pollution Removal
Systems'. ASCE Annual and National Environmental Engineering
Convention, Kansas City, Missouri, October 21-25, 1974.
4.9-36
-------
CONTENTS
4.10 MECHANICAL DEWATERING EQUIPMENT
4.10.1 Introduction
4.10.2 Centrifuge
4.10.2.1 Introduction 4.10-1
4.10.2.2 Service Description 4.10-1
4.10.2.3 Design Criteria 4.10-3
4.10.2.4 Available Equipment and
Operating Techniques 4.10-5
4.10.2.5 Existing Facilities 4.10-7
4.10.3 Continuous Vacuum Filters 4.10-7
4.10.3.1 Introduction 4.10-7
4.10.3.2 Service Description 4.10-8
4.10.3.3 Design Parameters 4.10-12
4.10.3.4 Available Equipment 4.10-23
4.10.3.5 Existing Facilities 4.10-23
•
References 4.10-28
4.10-i
-------
4.10 MECHANICAL DEWATERING EQUIPMENT
4.10.1 Introduction
When underflow from thickeners requires further solids-
liquid separation, continuous mechanical dewatering devices such
as centrifuges or continuous vacuum filters can be used. These
methods are used to remove sufficient water from liquid sludges
so that the sludge can be easily handled. Ideally, a dewatering
operation is designed to capture all the solids from a thickened
slurry at the lowest cost. The dewatering process produces a
solid cake having optimal physical handling characteristics and
moisture content for subsequent processing. Process reliabil-
ity, ease of operation, and compatibility with the plant en-
vironment also need to be optimized.
This section will acquaint the reader with the various
types of mechanical dewatering equipment that are currently used
or that have great potential for future application in lime
scrubbing FGD systems, and also with the parameters considered
to be important in the design and operation of the equipment.
4.10.2 Centrifuge
4.10.2.1 Introduction—
Centrifuges are widely used for separating solids from
liquids. They effectively create high centrifugal forces (about
4000 times the force of gravity). The equipment is normally
small and can separate bulk solids rapidly with short residence
time. The specially developed centrifuges are reliable and
efficient machines. Their products are consistent, uniform, and
easily handled; however, they are not effective in producing
clarified overflow and, because of high rates of wear, erosion
and corrosion require special materials of construction and
frequent maintenance.
4.10.2.2 Service Description—
Centrifugal separators are divided into two broad classes:
those that settle and those that filter. In the first class,
centrifugal force is utilized to increase the settling rate over
that obtainable by gravity settling; this is done by increasing
the apparent difference between densities of the phases. In a
filtering centrifuge, the pressure needed to force the liquid
through a septum is generated by centrifugal action. The main
interest in this section is the continuous settling centrifuge
for separating a slurry into a clear liquid and a very thick
sludge.
Figure 4.10-1 shows a continuous bowl centrifuge for solids
settling. The two principal elements of this centrifuge are the
rotating bowl (which is the settling vessel) and the rotating
screw conveyor (which discharges the settled solids). The bowl
4.10-1
-------
CONVEYOR
r-DRIVE
DRYING
LIQUID
ZONE
•INLET
CYCLOGEAR SOLIDS SCREW
DISCHARGE CONVEYOR
BOWL ADJUSTABLE IMPELLER
EFFLUENT
WEIR
Figure 4.10-1. Solid bowl centrifuge.
Source: Pfandler Co.
4.10-2
-------
has adjustable overflow weirs at its larger end for discharge of
clarified effluent and solids discharge ports on the opposite
end for discharge of dewatered sludge cakes. As the bowl ro-
tates, centrifugal force causes the slurry to form an annular
pool, the depth of which is determined by the adjustment of the
effluent weirs. A portion of the bowl is of reduced diameter to
prevent its being submerged in the pool; it thus forms a de-
watering zone for the solids as they are conveyed across it.
Feed enters through a stationary supply pipe and passes through
the revolving conveyor hub into the bowl itself. As the solids
settle out in the bowl, they are picked up by the conveyor screw
and transported to the solids discharge ports. A recent study1
indicated that, depending on feed rate and bowl rotation speed
(3300 to 5400 rpm), solids concentration ranged from 0.7 to 4
weight percent in the effluent and from 60 to 70 weight percent
in the cake when the feed was 16 to 20 weight percent.
4.10.2.3 Design Criteria—
It is extremely important to note that there are two opera-
ting zones in the horizontal bowl conveyor centrifuge: the
liquid zone and the drying zone. Early theoretical considera-
tion of centrifugal dewatering mechanisms focused primarily on
the relationship between the centrifuge and a hypothetical
sedimentation basin as they are affected by the employment of
very high gravity forces. The sigma formula is normally used to
describe the operation of a continuous, horizontal, helix-type
centrifuge. This formula shows that the centrifuge capacity
factor (which is proportional to the rate of liquid clarifi-
cation) varies with the surface area of the liquid and the
centrifugal force.2
where 2 = Sigma centrifuge capacity factor, ft2
b = Length of cylindrical bowl, ft
iu = Rate of rotation, rad/s
r2= Radius of inner bowl wall, ft
rx = Radius of retained liquid surface, ft
g = Gravitational constant, ft/s per second
Sigma and other theoretical relationships based on easily mea-
sured machine dimensions are useful tools when employed by the
centrifuge designer for estimating scale-up relationships in
geometrically similar machines. Unfortunately, the widespread
use of the sigma formula has led to some centrifuge specifi-
cations based only on sigma.3
4.10-3
-------
The capacity of a helical centrifuge is usually somewhat
lower than that predicted by theory. The action of the conveyor
tends to resuspend solids particles in the liquid. In addition,
at low feed rates, complex fluid dynamic effects have to be
taken into account.2
The important machine variables that affect centrifuge
performance are as follows:
0 Bowl design
Length/diameter ratio
Bowl angle
Flow pattern
0 Bowl speed
0 Pool volume and depth
0 Conveyor design
0 Relative conveyor speed
0 Sludge feed rate
Settling time and surface area can be increased for a given
diameter bowl by increasing the length/diameter ratio. Although
the detention time is increased by an increase in bowl diameter,
lower centrifugal forces result because of mechanical limita-
tions. Length/diameter ratios of 2.5 to 3.5 are customarily
employed. The designer can effectively increase the length of
the liquid zone of the bowl by making the discharge angle of the
screw conveyor steeper. Centrifugal forces can also be in-
creased by increasing the rotation speed.
In any centrifuge application, the centrifuge manufacturer
will determine the length/diameter ratio and the bowl angle;
however, wide variations in performance can be made by changing
other variables.
The primary operating variables are bowl speed and pool
volume. While increasing the bowl speed increases the centri-
fugal forces and favors increased clarification, it also makes
the settled solids become more difficult to discharge. Exces-
sive bowl speed tends to lock the bowl and conveyor together and
increases abrasion.
Pool depth affects both clarification and cake dryness.
Lowering the pool depth extends the drying zone, increases the
dewatering time, and produces a drier cake. Within limits,
increasing pool depth increases clarification by increasing
detention time. Just as in plain sedimentation, however, too
great a depth prevents a particle from reaching the sediment
zone prior to its being discharged in the effluent. At too
shallow a depth, the moving conveyor tends to redisperse settled
solids.
4.10-4
-------
Conveyor speeds are normally designed or adjusted to a
minimum turbulence inside the pool while still providing suffi-
cient conveying capacity. Low speeds also reduce the rate of
wear on the conveyor blades.
The sludge feed rate is clearly one of the more important
variables. It affects both clarity and sludge cake dryness.
The handling of a larger volume of sludge per unit of time in a
given bowl means less retention time and a decrease in solids
recovery. It also usually results in drier solids in the cake
because of the higher loss of fines with the centrate. Fines
have a tendency to retain more water.
Successful application of continuous bowl conveyor cen-
trifuges for removal of solids requires consideration of
numerous factors; proper scale-up is the major one. To obtain
predictable results, values must be available for the following
variables:
0 Wet cake discharge rate
0 Solids dewatering time under centrifugal force
0 Conveying torque for cake solids
0 Liquid clarifying ability
0 Resistance to abrasion from slurry solids
0 Stability of centrifuge feed
0 Physical nature of solids being handled.
The scale-up factors have provided accurate predictions of
full-scale performance.4
4.10.2.4 Available Equipment and Operating Techniques—
Bowl centrifuge —There are two types of bowl centrifuges
for solids removal, countercurrent and concurrent. The counter-
current centrifuge assembly consists of a rotating unit compris-
ing a bowl and conveyor joined through a planetary gear system
designed to rotate the bowl and the conveyor at slightly differ-
ent speeds in the same direction. The bowl, or shell, is sup-
ported between two sets of bearings and includes a conical
section at one end. This section forms the drying zone (on the
dewatering beach) over which the helical conveyor screw pushes
the sludge solids to outlet ports and then to a sludge cake
discharge hopper. The opposite end of the bowl is fitted with
an adjustable outlet weir plate to regulate the level of the
sludge pool in the bowl. This plate also discharges the cen-
trate through outlet ports, either by gravity or by a centrate
pump attached to the shaft at one end of the bowl. Sludge
slurry enters the rotating bowl through a stationary feed pipe
extending into the hollow shaft of the rotating bowl. The
sludge feed enters a baffled, abrasion-protected chamber for
acceleration before it is discharged through the feed ports of
the rotating conveyor hub into the sludge pool in the rotating
4.10-5
-------
bowl. The sludge pool takes the form of a concentric annular
ring of liquid sludge on the inner wall of the bowl. Separate
motor sheaves or a variable-speed drive can be used for adjust-
ing the bowl speed for optimum performance.
Usually, all parts of centrifuges that contact liquids are
made of ductile, generally corrosion-resistant, grade 316 stain-
less steel. The ductility of the stainless steel prevents
catastrophic brittle failure. Hard facing materials (such as
tungsten carbide) are applied to the leading edges and tips of
the conveyor blades, the discharge ports, and other wearing
surfaces, because of the abrasive nature of the lime-generated
sludges. Such wearing surfaces may be replaced, when required
by welding. M '
In a cocurrent centrifuge, incoming sludge is carried by
the feed pipe to the end of the bowl opposite the solid dis-
charge. As a result, settled solids are not disturbed by in-
coming feed. Solids and liquids pass through the bowl in a
smooth parallel-flow pattern. Turbulence is substantially
reduced. Solids are conveyed over the entire length of the bowl
before discharge to provide better compaction and a drier cake
and to reduce flocculant demands.
Addition of conditioners—Conditioners may be added to the
centrifuge feed to increase settling rates. Both inorganics and
polymers agglomerate fine particles by neutralizing surface
charges, thereby accelerating, settling, and clarifying thicken-
er overflow. Polymers possess less neutralizing power than
inorganics, but they have a higher capacity for "bridging"
(simultaneous attachment) to two or more solid particles than do
inorganic coagulants. They also provide the advantage of a very
large increase in the size of the floe, which greatly increases
their settling rate. (See Section 4.9.3.2.)
Lower-speed centrifuges—These centrifuges have been devel-
oped primarily in Europe to achieve high solids capture and
minimize the recirculation of solids without the use of high
polymer dosages. The sludge is introduced into the centrifuge
with the lowest possible acceleration and turbulence. The
machine is operated at about 1500 rpm, depending on the diameter
of the centrifuge. This low rpm gives a low noise level and a
minimum of wear and tear on the rotating parts. Low conveyor
differential speeds are also used. Among the reported advan-
tages of these machines are lower capital costs, lower power
requirements, lower noise level, and reduced maintenance when
compared with higher-speed centrifuges. The use of large pool
volumes, reduced internal turbulence, and low centrifugal forces
(500 to 800 g) combine to reduce shearing forces on the floe and
to improve conveying characteristics.
4.10-6
-------
4.10.2.5 Existing Facilities--
Bird 18-in. x 28-in. continuous bowl centrifuges are used
to dewater scrubber waste sludge and to recover dissolved Scrub-
bing additives at the test facility of the Tennessee v^mf,
Authority (TVA) coal-fired Shawnee Power Station near Paduc^
Kentucky Normal operating conditions usually consist of a feed
stream flow of 15 gal/min at 30 to 40 weight percent sol fdf *
centrate of 0.1 to 3.0 weight percent soUds, Pand a clke ol' 5f
to 65 weight percent solids. Approximately 30 percent of the
solids are fly ash; the remaining solids are prVdominantlv
calcium sulfate and sulfite. The centrifuge operates at 205?
rpni »
The material of construction is 316L stainless steel with
Stellite hard facing on the feed ports, conveyor tips and
ChaartS' The centrifu^ was inspected I in June' 1978
nv , e n une 978
after 6460 hours of operation since the previous factory servic-
ing. The machine was judged to be generally in flir condition
but .some components were badly worn and in need of factor^
repair. Serious wear was observed at the conveyor tips on the
discharge end and at the junction of the cylinder and the 10-deg
section of conveyor. Wear was also present at the casing head
plows and solids discharge head near the discharge ports The
bowl and effluent head were in good condition.5 P°rts • me
Recently, EIMCO (Division of Environtech Corp ) conducted a
test program for EPRI at Bruce Mansfield, Philffps and Cones
ville stations to determine design parameter^ and eva?ua?e Ihl
economics of centrifuges. Sharpless Models P-600 and I 660
Super-D-Canter (Pennwalt Co.) were used for these tests The
results indicated that over 90 percent solids could be recovered
with a bowl speed of 4000 rpm or higher. The discharged cake
7° W6iht 6 '
4.10.3 Continuous Vacuum Filters
4.10.3.1 Introduction —
Vacuum filters are normally the most economical mechanical
dewatering devices for continuous service. They are widely used
because they can be operated successfully at relatively high
turndown ratios over a broad range of solids concentrations in
the feed. A vacuum filter provides more operating flexibility
than any other type of dewatering device. J-J-exiomty
Five types of vacuum filters are applicable to lime-
generated sludge systems: drum, belt, disk, horizontal belt and
pan. Each has different characteristics and applicability
Since the rotary-drum vacuum filter is widely used for-
continuous service and is currently used in most scrubber sys-
4.10-7
-------
terns, this section will concentrate on it. A detailed dis-
cussion of other filter types will be presented under "Available
Equipment" in Section 4.10.3.4.
Since the vacuum filter will not provide an acceptable
filter cake if the solids content of the feed is too low, an
upstream thickener, centrifuge, or hydroclone is normally' re-
quired. •*
4.10.3.2 Service Description—
A rotary-drum vacuum filter (Figure 4.10-2) is widely used
for continuous service. The drum is divided into sections, each
connected through ports in the trunnion to the discharge head
The slurry is fed to a tank (or vat) in which the solids are
neld uniformly in suspension by an agitator. As the drum ro-
tates, the faces of the sections pass successively through the
slurry. The vacuum is applied in turn to each section (pickuo
or form zone in Figure 4.19-3) and the filtrate is drawn through
the filter medium, depositing the suspended solids on the filter
drum as cake. As the cake leaves the slurry, it becomes com-
pletely saturated with filtrate and undergoes dewatering by the
simultaneous flow of air and filtrate (cake drying zone). The
drying is negligible when air is used at room temperature.
Finally, the cake is removed in the discharge zone by a scraper*
which may be assisted by a slight air reversal through the
filter valve.
Continuous rotary-drum vacuum filters of this general type
provide high filtering rates and are available in a wide range
of sizes, from about 3 to 800 ft2 of filter area.
A typical filter system is presented in Figure 4.10-4. The
lower pipe connection at the filter valve accommodates the
liquid pulled through the sections in the pickup zone. The
upper filter valve connection carries the liquid and air pulled
through the cake in the dry zone. When a drum section reaches
the end of its cycle, the vacuum is released and a low-pressure
air supply discharges the cake through the filter tank chutes to
the conveyor below for final disposal.
Liquid and air enter the side connection of the filtrate
receiver, where the liquid drops down to the filtrate pump and
the air is pulled through the top connection of the receiver to
the moisture trap. Each receiver may be equipped with a
vacuum-limiting device to admit air if the design vacuum is
exceeded, a condition that would cause the pump to overload
The receiver also acts as a reservoir .for the filtrate pump
suction. The receiver is usually designed to give a maximum air
velocity of 2.5 to 5 ft/min and a minimum air detention time of
2 to 3 min to prevent carryover of the liquid. Check valves on
the discharge side of the pumps are usually provided to minimize
4.10-8
-------
o
I
vo
AIR FILTRATE LINE
CLOTH CAULKING
STRIPS
DRUM
AIR BLOW-BACK LINE
SLURRY FEED
Figure 4.10-2. Cutaway view of a rotary-drum vacuum filter.
-------
PICKUP OR FORM /^
Figure 4.10-3. Operating zones of vacuum filters.
4.10-10
-------
MOISTURE TRAP
O
I
FILTER CAKE
SLURRY FROM
A THICKENER
AGITATOR
SLURRY RECEIVING
TANK OR VAT
FILTRATE
FILTRATE RETURN
TO THICKENER
VACUUM OR
FILTRATE
RECEIVER
AIR TO ATMOSPHERE
SILENCER
FILTRATE 1=*^ VACUUM PUMP
PUMP BAROMETRIC
SEAL TANK
Figure 4.10-4. Flow sheet for continuous rotary-drum vacuum filtration.
-------
air leakage back to the vacuum pump through the filter pump and
receiver,
A two-receiver system is employed in those cases where
vacuum regulation is considered. This is especially advan-
tageous when the cake solids tend to crack and when regulation
is desired to limit air flow or control cracking in cake washing
systems.
The air pulled through the receiver enters a bottom tan-
gential connection at the moisture trap. The spiraling air,
upon entering the trap, expands, cools, and drops moisture in
the form of water vapor and entrained droplets from the receiver
through the bottom connection of the moisture trap. The mois-
ture trap has a barometric leg 34 ft long. Since a 34-ft water
differential pressure is approximately equal to atmospheric
pressure, any water that enters the trap will drop out the
barometric leg by gravity. The trap, therefore, offers protec-
tion to the vacuum pump in the event that the filtrate punro
should fail. v p
Air pulled out the upper tangential moisture trap connec-
tion is carried over to the intake of a vacuum pump and dis-
charged to the atmosphere. A silencer should be placed on the
pump discharge to reduce noise.
The lime-generated filter cake is compressible and may be
thixotropic, in that subsequent handling (via conveyor belt or
trucking, etc.) may liquify the sludge into a difficult-to-
handle putty.
A recent filtration test by EIMCO personnel for EPRI showed
that the lime-generated filter cakes cracked easily during the
early period of dry cycle.4 A short dry time (about 10 s or
less) is desirable to prevent this.
Plants that use vacuum filters dewater slurry from a thick-
ener containing 20 to 35 percent solids into filter cakes con-
taining 45 to 75 percent solids. The filtrate containing up to
1.5 percent solids returns to the thickener for reuse.
4.10.3.3 Design Parameters—
Mechanical design—The major components of a rotary-drum
vacuum filter consist of the drum, grids, internal piping,
receiving tank (or vat), and agitator. The mechanical design of
these components is a standard procedure provided by equipment
suppliers. Corrosion resistance is usually the controlling
factor in selection of the filter equipment. The sludge pH can
vary from 5 to 11. In addition to sulfurous acids, the slurries
contain chloride ions, which can be highly corrosive. Many
operators report evidence of chloride stress corrosion through-
4.10-12
-------
out the entire FGD system. K should therefore be used as
°nfo°o ^ ^¥n P*rameters'' threader is referred to Section
c^non
oo '
4.12.3 for additional information
All rotary-drum vacuum filters are available in a varietv
of materials, including carbon steel, stainless steel, special
alloys, rubber-covered steel, plastic-covered steel or all
plastic. Use of special materials requires higher capital
costs Nevertheless, when moving parts or wear surfaces are
exposed to corrosive/erosive environments, they should be prop-
erly protected with coatings of epoxy-based materials, FRP
rubber, or similar protecting substances. Specific areas of
concern are internal piping, wear plates, filter valves, filter
nal' i iGn m and grid' the filtrate pump, and ex-
(A) Major^sEonsnts - Figure 4.10-5 shows the drum without
covers and displays the internal piping, which may be made of
thin-wall stainless steel tubing, i.e., Schedule 10, 304 stain-
less steel. The drum heads are usually of mild steel; the drum
face and media are made of 304 stainless steel for long service
with minimum maintenance. Large manholes on the drum covers
provide easy access to all internal areas.
VACUUM LINES
FILTER SURFACE
-STRUCTURAL SUPPORT
MEMBERS
FILTRATE OUTLET
TO VACUUM SYSTEM
Figure 4.10-5. Drum and internal piping.
Source: Ameteck Process Equipment Division.
4.10-13
-------
The slurry receiving tank and the agitator can be made of
carbon steel, rubber-covered for corrosion protection.
Some suppliers offer plastic filter equipment made entirely
of high-strength FRP. The corrosion resistance of this equip-
ment is excellent. The available filtering area, however, is
limited to about 100 ft2, and the cost of the equipment is
greater than the cost of steel.
Grids that support the filter media are available in wire
screen, molded rubber, polypropylene, and other plastics with
open areas up to 55 percent. The snap-in grid design greatly
simplifies maintenance.
(B) Auxiliary equipment - The important auxiliary equipment
items are the vacuum receiver, the filtrate pump, and the vacuum
pump. Design considerations for these items will be discussed
in the following paragraphs.
To discharge 'the filtrate requires a pump to overcome the
suction head created by vacuum. A check valve is placed on the
filtrate pump discharge to ensure that no air is sucked back
into the system. Should this occur, the system would be inoper-
able. Thought must be given to the application of filtrate pump
discharge. The filtrate pump is rated for a given total dynamic
head (TDK) in gallons per minute and for net positive suction
head (NPSH).
Filtrate pumps should be specifically designed to operate
at very low net positive suction heads. The design inlet pres-
sure is at least 20 to 22 in. Hg vacuum. Centrifugal pumps are
common, but they should be protected against loss of prime in
the pump and have a balance or equalizing line connected from a
high point of the receiver to the eye of the pump impeller.
Nonclogging centrifugal pumps are used with coil filters or with
coarse metal filter media. They permit a somewhat higher solids
concentration in the filtrate. Self-priming centrifugal pumps
are used most frequently because they are relatively mainten-
ance-free. Self-priming, nonclogging centrifugal pumps are also
used.
The filtrate pumps must be sized to accommodate the entire
range of filtrate flow rates. In sizing a filtrate pump, the
designer must recognize that the rate of filtrate flow is a
function of the mode of chemical conditioning. Polymers allow
the sludge to drain much more rapidly than do inorganic condi-
tioners (see Section 4.10.2.4).
The piping from the filter valve on the filtrate pump
discharge must be in a horizontal plane or dropped vertically to
the receiver side connection. The moisture trap height above
4.10-14
-------
the receiver is inconsequential as long as the bottom connection
of the trap is above the top connection of the receiver, and at
least a 34-ft tail leg is on the trap. All pieces of equipment
should be placed as close as possible to each other, and un-
necessary pipe turns and bends should be avoided to reduce
friction head loss. Pipe connections and auxiliary equipment
sizes should be in accord with the filter manufacturer's recom-
mendations .
Filtrate may also be discharged from the receiver by means
of a barometric leg rather than a filtrate pump. This type of
system is beneficial from the standpoint that barometric legs
require minimal maintenance compared with filtrate pumps.
Discharging filtrate by a barometric leg is not always possible
however, because of plant elevation limitations. If a baro-
metric leg discharge is used, it should be immersed in the
cleanest circulating water in the plant to prevent its being
plugged with solids. it is recommended that an electrode be
placed in the line from the receiver to the trap so the elec-
trode, when contacted by water, will automatically shut off the
vacuum pump motor
Wet-type vacuum pumps are most popular because they are
easily maintained and provide sufficient vacuum. In such a
system, the vacuum pump uses water for its sealing medium, and
the moisture trap, the barometric leg, and the seal tank shown
in Figure 4.10-5 are eliminated. Because wet-type vacuum pumps
use seal water, the water must be of good quality; if it is hard
and unstable, a sequestering agent may be needed to prevent
carbonate buildup on the seals, but no moisture trap protection
is required.
Machine variables—A number of variables affect the opera-
tion of the filter system:
0 Feed solids concentration
0 Filter cycle time
0 Drum submergence
0 Agitation
0 Cake air requirements
0 Filter media.
The effect of each variable on the performance of the filter
system is discussed below:
Feed solids concentration - This variable is of utmost
importance in the filtration step, and for this reason a thick-
ening device precedes the filter to ensure a feed solids concen-
tration consistent with economic and efficient operation A
general plot of dry cake output vs. feed solids concentration is
shown in Figure 4.10-6.
4.10-15
-------
Each slurry has its own filtration characteristic curve,
but generally the slurry exhibits a sharp incremental rate above
"a." Controlling the solids concentration between "b" and "c"
will therefore require less filtration area, and filter operat-
ing costs will be reduced. Above point "c," the slurry becomes
relatively viscous and its transportation to the filter is
difficult. The curve becomes asymptotic, and further slurry
thickening is impractical and uneconomical in view of the slight
increment in cake rate.
(B) Filter cycle time - Cycle time of a continuous vacuum
filter is the time required for the filter to make one complete
revolution and is expressed in terms of minutes per revolution
(mpr). During the cycle, three phases of filter operation
occur: cake formation or pickup, cake dewatering or drying, and
cake discharge by air blowback or release. At the end of a
given cycle, the filter has discharged a given weight of cake
per given amount of filter area, and a dry cake rate in pounds/
hour per square foot of filtering area is obtained. The general
appearance of a log-log plot of dry cake rate vs. cycle time for
one filter feed solids concentration is shown in Figure 4.10-7.
The slope of the curve is negative and is theoretically equal to
-0.5.5 Empirical values are usually equal to the theoretical.
Stated in terms of increasing or decreasing filter cake output
as a function of changing cycle time, the resulting change in
cake rate is equal to the square root of [the original cycle
time divided by the new cycle time]. Expressed mathematically,
this relationship appears as follows and is based on the assump-
tion that solids concentration and cake compressibility remain
constant:
New filter caRe rate - old =a*e rate
It can be seen that cycle time is of great importance in
the filter operation. For this reason, the filter is equipped
with a variable-speed filter drive operating with 6:1 ratio
limits, usually of 1.5 to 9 mpr. Consequently, for a given
amount of filter area, cake output can be doubled, or possibly
tripled or halved, as the situation requires.
Cycle time is an important function of filter cake moisture
content and filter cake dischargeability. Where possible, it is
strongly recommended that the filter be sized at a cycle time at
least 3 mpr, and preferably as 4 mpr. Appreciable decreases in
cake moisture occur at cycle times slower than 3 mpr. In addi-
tion, thicker cakes resulting from slower cycle times give
complete and easy cake discharge. Easy cake discharge means
sizable reductions in filter maintenance costs.
(C) Drum submergence - Increasing the drum submergence
increases the form cycle time and usually results in an in-
creased yield of thicker but wetter cake. Submergence is usu-
ally kept between 15 and 25 percent to provide long drying time
and to keep the cake moisture content at a minimum.
4.10-16
-------
C\J
200
s-
-------
(D) Agitation - Proper agitation of the slurry requires
variable-speed mixing equipment for the vacuum filter vat. Only
enough agitation should be applied in the vat to prevent solids
classification and to keep them in suspension. Too much agita-
tion will loosen the filter cake from the filter medium. There-
fore, optimum control requires a variable-speed agitator.
(E) Cake air requirements - After the filter cake emerges
from the pickup zone, it is dewatered in the dry-zone part of
the cycle. To dewater the cake, it is necessary to provide
vacuum pump capacity to pull the required volume of air from the
atmosphere through the cake. Air flow through the cake creates
a resistance, which is recorded on the vacuum gauge as vacuum or
negative pressure differential. It is this pressure differ-
ential that effects cake formation in the pickup zone and cake
dewatering in the dry zone.
Each filter cake has its own air flow requirement, which is
approximately 3 ft3 of free air per ft2 of filter area. The air
flow rate, however, is mainly a function of cycle time and
solids particle size; in the case of compressible filter cakes
such as the lime-generated cake, an increase of air flow rates
would decrease cake moisture content. Unfortunately, lime-
generated cake cracks easily during the dry cycle and most of
the air passes through the cracks. A short dry time (about 10
s) and 2 to 3 ft3/min per ft2 of air therefore seem to be opti-
mum. This permits a vacuum differential of at least 22 in. Hg
across the cake, which is desirable to obtain minimum cake
moistures and maximum cake rates.
(F) Filter medium - The right selection of filter medium
is essential for the most effective operation of a continuous
filter. A great many types are available for drum filters.
Blinding characteristics and chemical conditioning play an
important role in medium selection. Filter leaf tests should be
conducted with the various media as an aid in selecting the
optimum one for a specific sludge. The ideal medium has the
following characteristics:
0 It is able to perform the desired liquid/solids separ-
ation and give a filtrate of acceptable clarity.
0 The filter cake discharges readily from it.
0 It is strong enough mechanically to give a long life.
0 It is chemically resistant to the materials being
handled.
0 Its resistance to flow is not too great.
0 It does not rapidly blind.
4.10-18
-------
Obviously, some reasonable compromise must be reached between
these objectives, since all of them cannot be optimized simul-
taneously Years ago, cotton duck was about the only filter
medium available to the vacuum filter operator. Today: a wide
range of choices exists. Filter media are available to cove r
any filtration situation, so blinding should not occur and a
maximum medium lifetime can be obtained. For a lime scrubbing
system, polypropylene appears to be most economical while a I so
providing adequate service and good chemical resistance.' oSe?
choices are polyethylene, nylon, and Dacron. utner
Operational features— Filters are normally installed -in
buildings with adequate weather protection. Besides Ihe machine
variables, the main considerations are proper feedina of the
slurry to the filter, the disposition OPf ?he dTs charged cake
S° ™tion Crated filtrate, and instrumentation of ?he
operation.
tank. Feeding should generally be accomplished at ?he side
feed; this is done in parallel flow with the direction of the
drU? r? wft^ JJ?f Si1UrrY ,WU1 haVG an Wortunity to filter upon
contact with the cleaned surface of the drum. In addition the
coarse particles will tend to collect first on the drum, thereby
providing a "precoat" to aid filtration in the remaining pickup
• Thn are installe<* at an elevated loca-
tion. This allows the cake solids discharging from the filter
to drop into a chute to a storage hopper for easy loading into a
truck. If an elevated position is undesirable, a belt conveyor
may be employed to collect the discharged solids from the filter
'"18
Finally, an electrical interlock system is worthy of men-
tion for the purpose of additional precautions. The simplest
and safest interlock in the filter system would be to interlock
the filter cake conveyor, filter drive, and vacuum pump such
that if the cake conveyor failed, first the filter drive and
then the vacuum pump would kick out. This would ensure that a
cake buildup would not occur in the filter discharge chute in
the event of cake conveyor failure. Moreover, should the filter"
drive fail, the vacuum pump would stop and slurry in the filter
tank would not have a chance to "dewater" itself to the deqree
that only solids remain. Degree
Sizing criteria— A standard rotary drum vacuum filter
be purchased from many suppliers. Correct sizing, i.eT,
can
the
4.10-19
-------
determination of the correct filter area, is important for
economical operation since size usually accounts for an appre-
ciable portion of the capital and operating costs. Sufficient
filter area must be provided for maintenance of the sludge
solids removal rate necessary to prevent excessive solids ac-
cumulation in the plant.
The size of a filter for a given application is inversely
proportional to the slurry feed concentration.7 Thus, if a
thickener is installed upstream, it is important to determine
the minimum underflow concentration encountered in average
operation of the unit. When the filter is sized at this minimum
solids concentration, it will have adequate capability to de-
water the solids output of the plant.
The filtration rate for sludges containing almost total
calcium sulfite from lime scrubbing appears to range from 50 to
60 Ib/h per ft2. On the other hand, filtration rates of sludges
in which calcium sulfate crystals dominate range between 150 and
250 Ib/h per ft2. 7
Because of the wide variations in slurry characteristics of
the lime-based FGD system, it is advisable to run laboratory
vacuum filtration tests on representative samples (if available)
of the sludges to be dewatered; this allows accurate sizing of
the filter equipment.2'3'8 The two test procedures used for
determining the filterability of sludges are the Buchner funnel
method3 and the filter leaf technique. The Buchner funnel
method enables a determination of the relative effects of vari-
ous chemical conditioners and the calculation of the specific
resistance of the sludge, but it is seldom used for the calcula-
tion of required filter area because it presents many diffi-
culties in providing data. The filter leaf test (Figure 4.10-8)
is used to determine the required filter area.2'3 It employs a
test leaf over which is fitted a filtering medium identical to
that which will be used on the full-scale filter. The procedure
for conducting filter leaf tests is as follows:3
1. Condition approximately 2 liters of sludge for filtra-
tion. The sludge should be thickened to a minimum
concentration of 2 percent or to that anticipated for
the full-scale application.
2. Apply desired vacuum to filter leaf and immerse in
sample 1-1/2 min (maintain sample mixed). The test
leaf normally is inserted upside down in a represen-
tative slurry to simulate the cake formation zone of
the drum filter. This portion of the cycle is cake
formation.
4.10-20
-------
FLEXIBLE VACUUM HOSE
TO VACUUM
(15-20 in. Hg)
FILTER LEAF
(0.10 ft2)
VACUUM HOSE
2-liter
VACUUM
FLASK
2-liter
SLUDGE SAMPLE
7 / / / / / / / /// /
7
Figure 4.10-8. Filter leaf test apparatus.'
4.10-21
-------
3. Bring leaf to vertical position and dry under vacuum
for 3 min (or other predetermined time). This is the
cake draining and drying part of the cycle.
4. Blow off cake for 1-1/2 min (this gives a total drum
cycle of 6 min). To discharge the cake, disconnect
the leaf and apply air (pressure not exceeding 2 psi).
5. Weigh cake, then dry and reweigh to determine percent-
age moisture. The filter rate (Y) in pounds/square
foot per hour is computed:
Y = dry weight sludge (q) x cycles/h
453.6 x test leaf area(ft2)
The test can easily be modified for other cycle times and dis-
charge mechanisms. Filter leaf is readily available from filter
manufacturers. It may be necessary to adjust the above result
by a factor to compensate for scale-up and partial medium blind-
ing over a long period of operation. The test results will
provide filtration parameters for the form, dry, and wash por-
tions (if necessary) of the filtration cycle. Although the
filter leaf test is a simple one, there are some precautions
that should be observed to ensure accurate results:
Representative sludge samples must be used.
0 Several (5 to 10) tests should be run to monitor
filter medium blinding.
The test sample must be agitated to ensure that it is
homogeneous.
0 The test filter vacuum must be regulated so that it
does not vary during the test and so that it is the
same as proposed for use in full-scale operation.
The filter leaf tests have been conducted for numerous
industrial and municipal waste treatment applications, and the
scale-up techniques are well established.6 The filter area
provided for in design should be for the peak sludge removal
rate required, plus a 5 to 15 percent area allowance.
If large variations in solids handling capability are
encountered, it is often more desirable to install two smaller
filters than one large filter. In this way, when the solids
amount decreases substantially over a long period, one of the
units may be shut down; this will allow the other to operate at
the proper submergence level and thereby to optimize perform-
ance. In addition, vendors recommend installation of a spare
for uninterrupted plant operation.
4.10-22
-------
4.10.3.4 Available Equipment—
Five types of vacuum filters are applicable for dewatering
sludges from lime-based FGD systems: drum, belt, disk, hori-
zontal belt, and pan.
The horizontal belt and pan vacuum filters are designed for
the dewatering of quick-draining, coarse solids that cannot be
retained on a vertical filter medium. They are also useful for
recovering valuable chemicals by washing.
The disk vacuum filter, which provides the highest fil-
tering surface area for the size of the equipment, is normally
used to handle large volumes of slurry, as in mineral processing
operations.
The rotary-drum vacuum filter is the most popular design.
It is usually the least expensive filter, in dollars/square foot
of filter area for a given application, that still permits cake
washing to be accomplished. The disadvantage of the unit is
that it is susceptible to medium blinding and wearing of the
medium; this is because the scraper at the discharge point
abrades the filter cloth. Replacement of the medium is time-
consuming because it must be caulked and possibly wire-wound
The rotary filter can therefore be costly from a maintenance
standpoint.
The belt filter (Figure 4.10-9) is an improved version of
the rotary-drum filter. The filter medium is lifted from the
drum after the dewatering portion of the cycle is completed and
is passed over a small-diameter roller to effect cake removal.
This rapid change in direction ensures a complete discharge of
cake without the need of a scraper. Thus, the filter cloth life
is comparatively long. After cake discharge, the medium is
washed on both sides. This arrangement provides a clean medium
for each filter cycle and prevents blinding, a particular advan-
tage in the filtering of solids, such as gypsum, that tend to
blind the medium. The installed cost of a belt filter, however,
is approximately 30 percent higher than that of the equivalent
size drum filter unit 7 A comparison of filter costs is pre-
sented in Figure 4.10-10. *
4 10.3.5 Existing Facilities—
The preferred methods of sludge disposal are ponding or the
use of a special stabilizing process such as Calcilox (by Dravo
Lime Corp.), Poz-0-Tec by International Utility Conversion
Systems, inc.), or Chemfix; however, one utility company has
installed filter equipment in a lime-based FGD system The
filter installation is at Paddy's Run No. 6 Unit (Louisville Gas
& Electric Co.). This plant has two rotary-drum vacuum filters
each with 150-ft2 filtering area and 10-tons/h sludge handling
capacity. Twenty to 24 percent solids feed from a thickener is
dewatered to 45 percent solids on nylon cloth medium, and the
filter cake is disposed of on an offsite landfill area.
4.10-23
-------
SPRAY PIPES
WASH TROUGH
TAKEUP ROLL
CLOTH BELT
DISCHARGE ROLL
DISCHARGE ZONE
WASH ROLL
Figure 4.10-9. Cross section of a belt filter.
4.10-24
-------
o
o
o
200
175
150
125
4/t
t? 100
o
S 75
_i
i* 50
z
>-H
25
0
100 200 300 400 500 600
FILTER AREA, ft2
Figure 4.10-10. Filter costs (7),
4.10-25
-------
Some utility plants (Conesville, Elrama, and Phillips Power
Stations) have contracts for waste disposal with IUCS, (Phila-
delphia) to produce the environmentally acceptable Poz-0-Tec.8
With this process (Figure 4.10-11), the partially dewatered
slurry from the FGD system thickener is pumped to the stabiliza-
tion system surge tank. The slurry is then pumped to a second-
ary thickener, if necessary, or directly to vacuum filters. The
capacity of vacuum filters is 150 lb/ft2 per h of 60 weight
percent solids from 36 to 40 weight percent feed.10 The filter
cake is then mixed with hydrated lime, Ca(OH)2, and silica,
SiO2, from the boiler fly ash; bottom ash is sometimes used as
well. The resulting product, Poz-O-Tec, is light-weight,
stronger than natural soils, and develops greater slope stabil-
ity on landfill. Its permeability is very low and the volume is
significantly less than the combined volume of untreated mate-
rials.11
4.10-26
-------
SLURRY
OVERFLOW TO FGD SYSTEM
ROTARY
VACUUM
FILTER
FILTRATE
BOTTOM ASH
FLY ASH
FILTER CAKE
SOLIDS
MIXER
CHEMICAL
ADDITIVE
POZ-0-TEC
TO LANDFILL DISPOSAL
Source:
Figure 4.10-11. Pox-0-Tec process.
IUCS, Philadelphia, Pa.
4.10-27
-------
REFERENCES
1. Personal communication with J. H. Wilhelm and R. W. Kobler,
EIMCO Process Machinery Division of Envirotech, October
1977.
2. Perry, R. H. , and C. H. Chilton, eds. Chemical Engineers'
Handbook. 5th ed. McGraw-Hill Book Co., New York, 1973.
pp. 19-93.
3. Process Design Manual for Sludge Treatment and Disposal.
U.S. EPA-625/1-74-006, October 1974.
4. Envirotech. Sludge Dewatering Methods for FGD Cleaning
Products. EPRI, 1978.
5. Rabb, David T. Selected Topics from Shawnee Test Facility
Operation. EPA Industry Briefing, Research Triangle Park,
North Carolina, August 29, 1978.
6. Tiller, F. M., et al. How to Select Solid Liquid Separa-
tion Equipment. Chemical Engineering, April 29, 1974, pp.
116-136.
7. Cornell, C. F. Liquid-solids Separation in Air Pollution
Removal System. Preprint 2363, ASCE Annual and National
Environmental Engineering Convention, Kansas City, Mis-
souri, October 21-25, 1974.
8. Heden, S. D., and J. H. Wilhelm. Dewatering of Power Plant
Waste Treatment Sludges. In: 36th Annual Meeting of
International Water Conference, Pittsburgh, Pennsylvania,
November 4-6, 1975.
9. Minnick, L. J. , W. C. Webster, and C. L. Smith. Lime-Fly
Ash-Sulfite Mixtures. U.S. Patent 3,785,840, January 15,
1974.
10. Boston, D. L., and J. E. Martin. Full-scale FGD Waste Dis-
posal at the Columbus and Southern Ohio Electrics Cones-
ville Station. In: FGD Symposium, Hollywood, Florida,
November 8-11, 1977.
4.10-28
-------
11. Mullen, H., L. Ruggiano, and S. Taub. The Physical anrt
Environmental Properties of Poz-O-Tec. in: Engineering
tion Solids, Hueston Woods, Cincinnati, October S19, "ig?!"
4.10-29
-------
CONTENTS
4.11 STACK GAS REHEATING
Page
4.11.1 Introduction 4 11-1
4.11.2 Reasons for Reheat 4 11-1
4.11.2.1 Prevention of Downstream Condensation and 4.11-1
Sebsequent Corrosion
4.11.2.2 Prevention of a Visible Plume 4 11-3
4.11.2.3 Plume Rise and Dispersion of Pollutants 4!11-3
4.11.3 Methods of Stack Gas Reheat 4.11-4
4.11.3.1 In-line Reheat 4 11-4
4.11.3.2 Direct-firing Reheat 4*11-6
4.11.3.3 Indirect Hot Air Reheat 4*11-8
4.11.3.4 Bypass Reheat 4!11-8
4.11.3.5 Exit Gas Recirculation Reheat 4!11-8
4.11.4 Reheat Requirement for Prevention of Downstream 4 11-11
Condensation
4.11.4.1 In-line Reheat 4.11-11
4.11.4.2 Indirect Hot Air Reheat 4!11-13
4.11.4.3 Direct-firing Reheat 4!11-15
4.11.5 Reheat Requirement for Normal Operation and 4 11-17
Prevention of Visible Plume
4.11.5.1 In-line Reheat 4 n_i7
4.11.5.2 Indirect Hot Air Reheat 4*11-18
4.11.5.3 Direct-firing Reheat 4!11-18
4.11.6 Reheat Requirement for Enhancement of Plume Rise 4 11-19
and Dispersion of Pollutants
4.11.7 Analysis of Bypass Reheat 4.11-21
4.11.8 No Reheat 4.11-22
4.11.9 Acid Condensation and Reheat 4.11-23
4.11.10 Selecting Optimum Sources of Energy for Gas 4119-5
Reheating -*.a.j. u
-------
CONTENTS (continued)
4.11.11 Existing Systems 4.11-24
4.11.11.1 Hawthorn (Kansas City Power and Light) 4.11-24
4.11.11.2 Four Corners (Arizona Public Service) 4.11-27
4.11.11.3 Colstrip (Montanta Power) 4.11-27
4.11.11.4 Phillips Stations (Duquesne Light & 4.11-27
Power)
4.11.11.5 Elrama Station (Duquesne Light & Power) 4.11-27
4.11.11.6 Bruce Mansfield (Pennsylvania Power Co.) 4.11-27
4.11.11.7 Paddy's Run 6 (Louisville Gas and 4.11-28
Electric Co.)
4.11.12 Recommendations 4.11-28
References 4.11-29
Bibliography 4.11-30
4.11-ii
-------
4.11 STACK GAS REHEATING
4.11.1 Introduction
One of the major drawbacks to wet scrubbing methods for
cleaning stack gas is unwanted cooling of the gas and its sat-
uration with water vapor as it exits the scrubber. The several
problems that result from this have led to reheating of the gas
in most of the operating or planned FGD scrubber installations.
The reheat system can be located directly above the mist elimi-
nator (Figure 4.11-1) or in the horizontal duct leading to the
stack. Hot air or bypassed gas may also be injected downstream
from the mist eliminator.
Before dealing with the subject of reheat, it is important
to know the following terminology and relationships:
Absolute humidity; The amount (in pounds) of water vapor
carried by one pound of dry air.
Percentage relative humidity; The partial pressure of
water vapor in air divided by the vapor pressure of water
at a given temperature.
Dew point or saturation temperature; The temperature at
which a given mixture of water vapor and air is saturated.
Wet-bulb temperature: The equilibrium temperature attained
by a water surface when the rate of heat transfer to the
surface equals the rate of heat transfer from the surface
because of the liquid evaporation.
4.11.2 Reasons for Reheat
A major question in designing wet scrubbing systems is
whether or not scrubbed gas should be reheated, and if so, how
to do it. The reasons usually advanced for reheating are as
follows:
(1) To prevent downstream condensation and subsequent
corrosion, and either
(2) To prevent emission of a visible plume (the stack gas
temperature required ranges from 180° to 220°F), or
(3) To enhance plume rise and dispersion of pollutants
(stack gas temperature required is above 220°F)
4.11.2.1 Prevention of Downstream Condensation and Subseauent
Corrosion—
At power plants, the flue gas exits the wet scrubber in a
.turated condition at about 125°F. The gas also contains mist
ea.
4.11-1
-------
FLUE GAS
ID FAN
STACK
REHEATER
MIST ELIMINATOR
SCRUBBER
Figure 4.11-1. Reheat system location.
4.11-2
-------
droplets, the amount depending on the efficienty of the mist
eliminator. If the gas is not reheated, an induced-draft (ID)
fan downstream from the scrubber will run wet, and entrained flv
ash and slurry will be able to deposit on the wet surface of the
fan blades. Removal of the resulting solid deposits from the
fan is troublesome and expensive. Although there has been
improvement in this area, corrosion problems in the ducts and
stack are still severe in most cases. In some cases, deposits
of solids on the ID fan, ductwork, and the inside walls may
absorb residual SOg from the flue gas. This could result in
highly acidic conditions capable of corroding fan blades and
ducting and damaging the stack coating.
4.11.2.2 Prevention of a Visible Plume (Normal Operation,
Reheat to 180° to 220°F Range)—
in the absence of stack gas reheat, "acid rain" can occur.
Acid rain is formed by the condensation of droplets and their
absorption of residual SO2 in the plume. Since the stack gas
contains more water vapor than that from normal plant operation,
formation of a steam plume can occur more easily. This is a
major consideration in some situations, especially if the plant
is in a densely populated area and the neighbors are sensitive
-to visible emissions. To avoid a visible plume, the gas is
reheated between 100° to 220°F.
4.11.2.3 Plume Rise and Dispersion of Pollutants (Reheat to
above 220°F)—
A plume may be several miles from the plant before it
finally reaches the ground. Ground-level concentration is of
concern to regulatory authorities who set maximum ambient
levels.
To achieve plume rise and dispersion of pollutants, the gas
can be reheated to 220° to 300°F, which in some cases is the
same as the scrubber inlet temperature. This practice is espe-
cially prevalent in Japan, where among the four major scrubber
installations operated by utilities, the reheat level at one is
250°F and at the other three, 290°F. This extra heating ac-
counts for as much as 5 percent of the total fuel costs, but is
considered justifiable because it promotes good community rela-
tions .
The Federal SO2 regulation for new coal-fired plants cur-
rently allows a maximum emission of 1.2 Ib S02/million Btu. The
regulation does not mention the degree of reheat or, for that
matter, the height of the stack. Thus, as far as the S02 emis-
sion regulation is concerned, the plant operator can save money
by not reheating the gas if he is willing to accept the other-
trade-offs.
4.11-3
-------
Ambient standards, however, apply to all sources; it is the
responsibility of the states to enforce compliance. If scrub-
bers are to be installed in a given plant, it becomes a matter,
(theoretically at least) of predicting whether or not the am-
bient air standard will be met at the proposed degree of reheat.
If the prediction (based on plume dispersion models) indicates
that the ambient standard will be exceeded, then the degree of
reheat or the stack height must be increased, or the SO2 emis-
sions decreased still further. The matter of reheat, however,
does not seem to have come up in permit hearings. Control of
the degree of reheat, as applied to the prediction of ambient
concentration, is a refinement that apparently has not yet come
into wide use in the United States.
4.11.3 Methods of stack Gas Reheat
The following methods can be used to increase the tempera-
ture of the gas from a wet scrubber.
(1) In-line reheat
(2) Direct-firing reheat
(3) Indirect hot air reheat
(4) Bypass reheat
(5) Exit gas recirculation reheat
Method (5) is not used commercially for lime scrubbing appli-
cations .
4.11.3.1 In-line Reheat—
The most popular system is a heat exchanger installed in
the flue gas duct following the scrubber mist eliminator. The
in-line reheater is simple in design and installation as shown
in Figure 4.11-2; however, it is difficult to maintain because
of corrosion and plugging in the tube bundles.
The tubes are usually arranged in banks. Since deposits on
the tubes reduce the heat transfer considerably and cause corro-
sion, soot blowers are normally installed between the banks.
Corrosion has been a problem, and even expensive alloys have
been unsatisfactory under some conditions. A large number of
materials have been tested in TVA's Colbert pilot plant, in-
cluding Cor-Ten A, Cor-Ten B, Incoloy 825, Inconel 625, various
300 series stainless steels, and Hastelloy C-276. This is
further discussed in Section 4.12.7. When used in lime scrub-
bing systems, 304 SS and 316 SS have sometimes failed within a
few months. In contrast, carbon steel tubes have given accepr
table service at some installations for up to 5 years.
No good explanation of this is available. Obviously some
differences in operating conditions are responsible, but it is
not clear what they are. Factors postulated to be significant
4.11-4
-------
REHEATER
i
01
BOILER
AIR
PREHEATER
ESP SCRUBBER
STFAM
/
s — •
>
/
, — .
s>
/
/^—
\
J ' t t
— ^»
STACK
Figure 4.11-2. In-line reheat system.
-------
include the amount of mist impinging on the tubes, distance
between the mist eliminator and reheater, temperature inside the
tubes, and adequacy of soot blowing. There is some indication
that soot blowing is the most important factor. Much of the
corrosion of tubes is of the pitting type and occurs under
deposits on the tubes. The corrosion of high-alloy materials is
generally attributed to stress corrosion caused by chloride.
Stress corrosion causes failure far more rapidly than does
general corrosion. Carbon steel is more susceptible to general
corrosion but more resistant to stress corrosion than high-alloy
materials, and vice versa, in the reheater environment. Carbon
steel reheaters thus may be more durable in operation than
stainless steel reheaters.
In-line reheat can be classified according to the heatina
medium (steam or hot water).
0 Steam in-line reheat-- Inlet steam temperatures range from
420 to 720°F, and steam pressures from 115 to 450 psig. The
.
,problems with this type of reheater have been corrosion of
the heating tubes and plugging in the tube banks. Superheated
steam requires a larger heat-transfer area than saturated steam
does, since it involves gas-to-gas heat transfer. Saturated
steam is preferred.
Hot water in-line reheat— The system configuration is
similar to that of the steam in-line reheaters, except that hot
water is the heating medium. Inlet temperature of the hot water
ranges from 250° to 350°F, and the temperature drop over the
heat exchanger is 70° to 80°F. The inlet pressure of the hot
water ranges from 15 to 120 psig. Finned tubes are usually
required for better heat transfer and soot blowers are needed to
clean the tubes.
Corrosion problems with these reheaters have been less
severe possibly because of the lower operating temperature.
Plugging, however, has been a major maintenance item at Lawrence
Station because of the finned tubes in the heat exchanger.
4.11.3.2 Direct-firing Reheat—
This type of system eliminates the use of heat exchangers.
As shown in Figures 4.11-3a and 4.11-3b, gas or oil is burned
and the combustion product gas (at 1200° to 3000°F) is mixed
with the flue gas to raise its temperature to between 150° and
300°F. in Japan, practically all scrubbers have direct- fired
oil reheat, and in most cases low-sulfur oil is used.
The main problem with direct firing is availability and
cost of gas and oil. in Japan, where most boilers burn oil, the
overall oil requirement is increased by only 2 to 5 percent, in
the United States, however, coal is the principal fuel; oil or
4.11-6
-------
REHEATER
BOILER AIR
PREHEATER
ESP SCRUBBER
IN-I
BUR
.INE
NER
~p
1
STACK
FUEL
AND
AIR
Figure 4.11-Sa. Direct-firing reheat system using in-line burner.
REHEATER
BOILER
AIR
PREHEATER
ESP
SCRUBBER
i
/
x^
EXTERNAL
COMBUSTION
CHAMBER
STACK
FUEL
A Ki n AID
ANU M1K
Figure 4.11-Sb. Direct-firing reheat system using external combustion chamber.
-------
gas would not only be difficult to obtain in some situations,
but would also be costly and usually require a new storage or
pipeline installation.
Direct firing requires some care in mixing the hot combus-
tion gas with the cool scrubbing gas. If mixing is not carried
out effectively, hot spots could develop downstream from the
heater and cause damage to the duct lining.
4.11.3.3 Indirect Hot Air Reheat—
As shown in Figure 4.11-4, ambient air is heated by an
external heat exchanger using steam to temperatures of 350° to
450°F. Finned tubes made of carbon steel are arranged in two to
three banks in the heat exchanger. Hot air and flue gas may be
mixed by a device such as a set of nozzles or a manifold.
The advantage of indirect hot air reheat over in-line
reheat is that the indirect system involves no corrosion or
plugging. Its disadvantages include the need for an additional
fan for hot air blowing, the relatively large amount of space
required to retrofit the reheat system to an existing boiler
system, and the increase in stack gas volume, which may be
undesirable because of the limited capacities of existing ID
fans and stacks. Another disadvantage is its higher energy
consumption; this extra energy is needed to heat air from the
ambient temperature level. An advantage that offsets the higher
energy requirement to some extent is that the dilution resulting
from the air addition reduces the incidence of steam plume
formation and also presumably gives better plume dispersion.
4.11.3.4 Bypass Reheat—
In the bypass reheat system, a portion of hot flue gas from
the boiler (approximately 300°F) is taken off ahead of the wet
scrubbing system, bypassed around it, and mixed with the flue
gas that has been processed through the wet scrubber (Figure
4.11-5). The limiting conditions for application of this system
are determined by (1) the properties of the boiler fuel, such as
heating value, sulfur content, and ash content; (2) particulate
control (ESP) preceding the scrubbing system; (3) the character-
istics of flue gas, such as temperature and flow rates; (4) the
efficiency of the scrubbing system; and (5) emission regula-
tions. Provision of enough reheat sometimes requires that as
much as 40 percent of the flue gas bypass the scrubber. This
requires very high SO2 removal efficiency and lenient perform-
ance standards. A performance standard requiring 90 percent SO2
removal efficiency would completely rule out the bypass reheat
option.
4.11.3.5 Exit Gas Recirculation Reheat—
As shown in Figure 4.11-6, a portion of heated stack gas in
the exit gas recirculation reheat system is diverted, heated
4.11-9
-------
REHEATER
BOILER
I
vo
AMBIENT
AIR
EXCHANGER
Figure 4.11-4. Indirect hot air reheat system*
-------
REHEATER
STACK
Figure 4.11-5. Bypass reheat system.
REHEATER
.1
H
O
BOILER
HEAT
EXCHANGER
Figure 4.11-6. Reheat by exit gas recirculation.
-------
further to approximately 400°F bv an ^xt-*T-n=,i h~ «.
and injected back into the flue cL from^ £eat exchan
-------
HEAT GAIN FROM
I.D. FAN.Q.-
HEAT LOSS FROM
DUCTQLD
FLUE GAS FROM
WET SCRUBBER, T,
1
HEAT LOSS FROM
STACK,QLS
1
NET HEAT INPUT = Q
STACK GAS TO
ATMOSPHERE, T,
DEWPOINT , V
Figure 4.11-7. Schematic of heat balance around
downstream system with in-line reheater.
4.11nl2
-------
C
pm °f flue gas at constant pressure,
IUO J. F
F = Flue gas flow rate, scfh
T = dew point of stack gas at top of stack, °F
(usually 125°F)
T! = temperature of flue gas exiting wet scrubber, °F
QLD = heat loss from ducts, Btu/h
QLS = heat loss from stack, Btu/h
L = heat required to evaporate mist carryover, Btu/h
Qp = heat gain from ID fan, Btu/h
For estimation purposes the overall heat-transfer coefficient
(the thermal conductivity ) can be assumed to be 10 Stu/?ft?«
(°U(h/ 4^°^^ (c°n4densing steam) heat exchange and to be
6 Btu/(ft)*('F)
-------
HEAT LOSS
FROM DUCTS
HEAT LOST
FROM STACK
A'LS
FLUE GAS FROM
WET SCRUBBER, T
1
STEAM IN
STEAM OUT
INPUT = Q ""*"
STACK GAS TO
•ATMOSPHERE, T,
DEWPOINT, T/
HOT AIR, T
ha
AMBIENT AIR
IN- Tca
Figure 4.11-8. Schematic of heat balance around
downstream system with indirect hot air reheater.
4.11-14
-------
Heat required Heat Heat
to raise gas + loss + loss
to its dew from from
P°int ducts stack
Heat
required
to evap-
orate mist
carryover
Heat
gain
from
heated
air
F C
t .
- Td)
(Eq. 4.11-2)
Heat required by
Net heat input = (F ) amount of ambient
air to reach temperature T
F C
a
379
ha " ca
,
ha
. 4.11-3)
where ,
Fa = ambient air flow rate, scfh
Cpa = sPecific
ha
of air, Btu/(lb-mol )°
= ternPerature of hot air, °F
Tca = temperature of ambient air, °F
and the other symbols are as previously defined
in this case, Equation 4.11-2 is used to determine the
temperature of hot air (Tha), then Equation 4.11-3 is used to
determine the net heat input.
4.11.4.3 Direct- firing Reheat —
Figure 4.11-9 is a schematic of the heat balance around the
stack gas downstream system; it includes a direct combustion
reheater. Natural gas or low-sulfur fuel oil may be burned" in
the combustion chamber. The minimum heat requirement, Q, can be
obtained from the following equations:
Heat required Heat
to raise gas
to its dew
point.
required
+ to evap-
orate
mist
carry-
over
Net latent heat
of combustion
product gas
between tern-
peratures T
and T , g
Heat Heat
loss gain
from + from
ducts fan
Heat
loss
from
stack
4.11-15
-------
FLUE GAS
FROM WET
SCRUBBER, T
1
HEAT GAIN
FROM I.D.
FAN (L
HEAT LOSS
FROM DUCTS
FUEL IN
HEAT LOSS
FROM STACK
STACK GAS TO
ATMOSPHERE, T,
DEW POINT, T/
d
COMBUSTION PRODUCT
GAS, T
COMBUSTION
AIR IN, T
ca
Figure 4.11-9. Schematic of heat balance around
downstream system with direct combustion reheater.
4.11-16
-------
(Eq. 4.11-4)
(Net heat input) = (Fuel consumption rate) x (Heating value of
fuel)
Q = q xv
(Eq. 4.11-5)
where,
Fg = combustion product gas flow rate, scfh
Cpg = fX?^10 heat of combustion product gas, Btu/(lb-
Tg = temperature of combustion product gas, °F
q = fuel consumption rate, Ib/h
v = heating value of fuel, Btu/lb
and the other symbols are as previously defined.
4.11-5 Reheat Requirement for Normal Operation and Prevention
of Visible Plume ~
The section explains the methods for determining the re-
quired heat that are currently applied in most reheat instal-
lations. For normal operation and prevention of visible plume,
the temperature desired at the top of the stack is selected by
the designer. It is usually between 125° and 220°F. The reheat
requirement for normal operation and prevention of visible plume
can be estimated by making heat balances as follows:
4.11.5.1 In-line Reheat—
Minimum Heat required Heat Heat Heat Heat required
heat - to raise gas + loss - gain + loss + to evaporate
required from tempera- from due from mist carryover
ture Tt to T2 ducts to stack ^yuver
fan
FC
A =
379 vx2 ~ M ) T
* i-1^3
(Eq. 4.11-6)
4.11-17
-------
where,
A = minimum heat required
T2 = stack gas temperature at the top of the stack.
All other symbols have been defined earlier.
4.11.5.2 Indirect Hot Air Reheat—
The computation here is more complicated than that for
in-line reheat, since the stack gas is diluted with hot air.
Heat
required
to raise gas
temperature
from T! to T2
Heat
required to
+ evaporate
mist
carryover
Heat
required by
F amount
or ambient
air to
reach tem-
perature T
from T
Heat
loss
from
ducts
Heat
gain
due
to
fan
Heat
loss
from
stack
ha
FC
379
(T2 - T!
F C
a pa
379
(T
ha
- QLD + QF ~ *LS
(Eq. 4.11-7)
Net heat input = Heat required by Fa amount of ambient air to reach
temperature T,
F C
Q =
y
P"1
379
(Tha - Tca>
(Eq. 4.11-8)
4.11.5.3 Direct-firing Reheat—
Heat
required
to raise
gas
tempera-
ture from
Tj to T2
Heat
required
+ to evaporate
mist
carryover
FC
m
379
(T2 -
+ L
Net latent
heat of
combustion
product gas
between
temperatures
Tg and T2
Heat Heat Heat
- loss + gain - loss
from due from
ducts to stack
fan
F C
g
379
- T2) - Q
LD
QF - -LS
(Eq. 4.11-9)
4.11-18
-------
innutf = (FUGl consumPtion (Heating value
input) rate x
rate) x Qf
v
(Eq. 4.11-10)
All symbols used in the above equations have been defined earlier.
Enhancement of Plume Rise and
without
-M-^""—— — ~~ ——•-'vtj.w j.j.1 IJUUX. D-LUMf* T-T c«i ^v-,,3 i •
characteristics, which in turn -could ,fai, P^°r- dlsPersi°n
ground-level concentrations of otherpollutants" su^17 ^
of nitrogen. The computation of reheat ™ ' Ch aS oxides
complicated because it requires a quant^ttgm VSrY
behavior in the atmosphere, which is a rive analysis of plume
ables as meteorological conditions, stack si^*^ J""* Va^i"
tics of the stack gas , at the emission nn-irfi- * charactens-
tails, refer to the report StaTk Gas R^V -r r further de~
Desulfurization Systems, prepared^by the^Bat^i P ^ Flue Gas
j_ ~.i ^<~ -F/-VV- c-i ^^4- • ^ v-^ttioj \jy tne oatelle ColiirnVMio r.aK^v_
atones for Electric Power Research Institute
To achieve dispersion of pollutants t
ture at the top of the stack should be ab'ovf
lation in the United States provides that
although some Japanese installations reheat t
In the Battelle study, a family Of c
using mathematical models that ored-i^t- m^ .urves ^as developed
trations under any given conditions Is^sho^^n C°ncen-
4.11-10, each curve represents a given Dercen? T ^ F^^re
from the gas; if nothing else were chanaed 1-hi ° ?E r.emoval
the percent reduction in ambient concentration ^ ° b!
concentration is directly proportional ^ ^' because Around
emitted from the stack. since thJ^ scrubbeJ^ a^^ ^ S°2
lower temperature (unless reheated to scrubbe 1V VS &
ture), the reduction in ambient • • -1 tempera
nigh '
cncentration. At 90 percent removal !lr ex^nlln 9roundT1evel
reheat the maximum ambient air c^ncentrai-^ ? ' ^Ven wlth°ut
^rcent; at about 40 percent removX " by "
4.11-19
-------
225
200
a:
LU
O-
f
ui
z
UJ
ex.
175
150
125
SCRUBBER
EFFICIENCY, %
-40 , -20 0 < 20 40 60 80 100
REDUCTION IN MAXIMUM GROUND-LEVEL CONCENTRATION, percent
Figure 4.11-10. Effect of reheat temperature
(for in-line reheat) on ground-level 862
concentration at various scrubber efficiencies,
Base: Maximum ground-level concentration at 300°F without scrubber.
4.11-20
-------
concentration from 83 percent to 88 percent. This is an impor-
tant consideration with respect to standards requiring 90 per-
cent S02 removal efficiency. At 70 percent removal, the effect
of the 50°F reheat is more pronounced; reduction in maximum
ground-level concentration changes from 48 percent to 65 per-
cent.
It should be noted that the higher water vapor content in
the gas offsets to some extent the adverse effects of gas cool-
ing. Since the water vapor has a lower density than other
constituents of the gas, it makes the plume more buoyant. The
effect is small, however, and has been omitted in developing the
curves.
It is concluded that for the high degrees of S02 removal
(85-95%), reheating is not likely to be economically justified
except in marginal situations where the inlet SO2 to the scrub-
ber is so high that, even with high SO2 removal, ambient concen-
tration is still close to exceeding the standard.
There are some considerations, however, that may make the
situation worse than it appears. If there is no reheat, then
the gas leaving the stack can already have a load of mist, in
which case evaporation of the droplets as the plume becomes
mixed with air can cool the plume and further reduce its buoy-
ancy. A high degree of mist elimination should be achieved if
no reheat is used. Moreover, very little NO (probably less
than 15 percent) is removed in SO2 scrubbing. ^Thus NO ambient
concentration will be greatly increased unless the ga^s is re-
heated.
4.11.7 Analysis of Bypass Reheat
This section analyzes bypass reheat to examine its applica-
bility from the standpoint of the emission limitations for
sulfur dioxide. Bypass reheat offers the advantages of low
capital investment and simple operation. The maximum degree of
reheat that can be obtained, however, is limited by the con-
straints of pollutant emission standards. As mentioned earl-
ier, a regulation requiring 90 percent SO2 removal efficiency
would completely rule out the bypass reheat option. The limita-
tion of sulfur emission to meet the emission standard for sulfur
dioxide of 1.2 Ib/million Btu can be written as:
v - i . 1 + 1-2
X ~ 1 E 2WSE (Eq. 4.11-11)
where,
W = amount of fuel required to generate one million Btu/lb
4.11-21
-------
S = weight fraction of sulfur in the fuel
X = fraction of bypass flue gas stream
E = fractional sulfur removal efficiency of the wet scrubbing
system
For details of the heat balance around the reheat system,
refer to the report Stack Gas Reheat for Wet Flue Gas Desul-
furization Systems, prepared by the Battelle Columbus Labora-
tories for EPRI.1
4.11.8 No Reheat
As mentioned previously, stack gas reheat is not required
by law. Some power plants have selected, at least temporarily,
a "no-reheat" design and accepted the possible consequences—
condensation in the ID fan and the stack.
Wash water can be sprayed periodically on the ID fan blades
to prevent solid deposits, and a wet stack can be installed to
protect the stack from acid attack.
Some advocate "no-reheat" by utilizing a "slow" stack (gas
velocity of 30 ft/s) versus a conventional stack (gas velocity
of 90 ft/s). The slow stack allows mist droplets (acid rain) to
settle out in the stack bottom. This requires special duct and
stack material and handling equipment. It also required larger
stacks, but these increase opacity problems.
Another alternative for prevention of ground concentration
of pollutants is to build a taller stack. A tall stack may be
more economical than reheating, even though it involves a high
capital cost. There is, by comparison, no energy cost. Under
certain circumstances, however, a stack of the required height
might not achieve the objective of dispersion for a particular
location. Meteorological modeling is a useful tool for deter-
mining the validity of such an alternative, but most dispersion
models have not been developed for wet plumes.
To limit corrosion in no-reheat operation, one may either
select materials that are inherently resistant to corrosion, or
use coatings to cover corrodible materials. Discussion of this
is included elsewhere in this book. If the purpose of reheat is
to protect a downstream fan, an obvious alternative is to place
a fan upstream from the scrubber. This is only feasible with an
upstream collector or ESP to remove abrasive particulate. Most
installations with wet stack operation have stack lining prob-
lems. The lining usually blisters and eventually comes off.
Once this happens, stack corrosion begins.
4.11-22
-------
As discussed in the chapter on mist elimination, an effi-
cient mist eliminator can reduce the reheat energy requirement
°£ fflevf n°~reh®at operation in some cases. It is also obvious
that the formation of large particles in the atmosphere by
agglomeration and growth is highly unlikely in the absence of
large drops Such formation should not contribute significantly
either to stack icing or to rain. stack icing is the ice that
forms on top of the stack under freezing conditions.
4.11.9 Acid Condensation and Reheat-.
one of the factors that limits boiler efficiency is the
effect of corrosion on heat recovery hardware as the temperature
°fnot?e ^V1"6 gaS aPProaches the critical region of 275° to
*° ;,F; a™.ZZ^.&^£*±£ • t»ical So^r.plS re!
materials, the cost of which cannot
While most of the sulfur is converted to sulfur dioxide
P°ti0n ^ reaCtS to form sulfur tri!
is*** /en \ M-i-t-v.-: vs j-i- , , — —wwo ww j_«jo.m t»uj.j.ui. tri-
oxide (S03). Within the scrubber, these acid gases react to
form sulfurous and sulfuric acids, some of which are removed in
the scrubber. Beyond the scrubber outlet, water vapor continues
to condense and reacts with residual SO2 to form sulfurous aSiS
Sulfur trioxide has a far more corrosive effect than sulfur
dioxide, although in most boilers the ratio of SO3 to SO2 is
between 0.01 and 0 03. Even this low concentration of SO3
raises the gas acid dewpoint temperature considerably. Conden-
sation occurs and corrosive sulfuric acid is formed. Only
partial removal of SO3 occurs in the scrubber. The degree of
removal is difficult to measure because of the limited accuracy
of S03 determination methods at these low levels.
At a typical S03 concentration of 6 to 8 ppm in the hot
- fiU6 9?aoir stream' theoretical acid dewpoints are between
~- ' i A\ typical FGD scrubber outlet level ranges
between 1 and 5 ppm. The water dewpoint of scrubbed flue gas is
typically about 128°F. The presence of as little as 1 ppm of
S03 m the flue gas creates an acid dewpoint of above 200°F.
in the final analysis, it is not economical to reheat above
the acid dewpoint, but only to reheat sufficiently to prevent
significant condensation of moisture in the gas. This means
drying the gas to an intermediate point between water dewpoint
and theoretical acid dewpoint. uewpomt
4.11-10 Selecting Optimum Sources of Energy for Gas Reheating
As fuel supplies become more scarce and prices continue to
rise, it becomes increasingly important to look for economies in
reheat designs. In general, four items must be consider"! in
arriving at an optimum design: «-««sj.aerea in
4.11-23
-------
(1) Initial costs of the auxiliary steam system components
(2) Operating costs to produce the auxiliary steam
(3) Pressure level of the source
(4) Boiler or turbine limiting factors
(5) Decreases in reheater area as steam pressure increases
It should be remembered that the penalty associated with
steam use from the boiler depends on the point in the cycle at
which the energy is extracted. If steam is extracted at the
highest temperature and pressure available, the cost of that
heat will be very high. On the other hand, if steam is extrac-
ted at some intermediate point, the cost is lower. In deter-
mining the cost of such heat, its value in terms of extracted
work as electrical energy must be considered. The boiler and
turbine for new installations must be designed with the reheat
steam penalty in mind; otherwise there may be a loss of net
power generation capability.
4.11.11 Existing Equipment
This section presents the reheat situation of lime scrub-
bing installations and reheater problems experienced by each.
Table 4.11-1 surveys lime scrubbing installations with in-line
reheaters. As the table shows, two of three in-line reheat
systems are designed for 50°F reheat and tubes of 1 in.
diameter; both have soot blowing arrangements.
Table 4.11-2 surveys direct-fired reheat installations.
Power plants that do not reheat the gas are Conesville (Columbus
and southern Ohio Electric Company), Green River (Kentucky
Utilities), and Cane Run (Louisville Gas & Electric). Problems
developing from not reheating the gas have been encountered in
equipment downstream from the mist eliminator. These problems
will be discussed in the chapter involving that particular
equipment.
4.11.11.1 Hawthorn (Kansas City Power and Light)—
The FGD system at Hawthorn was modified from a limestone
injection-based system to a lime slurry-based system. The unit
became available for service in the lime scrubbing mode on
January 1, 1977. No reheater corrosion problems have been
reported. Reheater plugging was a problem, particularly in the
B module of the scrubber system. This was solved by removing a
section of the reheater to facilitate cleaning and maintenance.
Currently, it is a normal practice to shut down the scrubber
every 3 days for cleaning the mist eliminator and the reheater.
Soot blowing is a heavy maintenance item at this installation!
The reheat hot water pump is normally started before placing the
scrubber in service.
4.11-24
-------
Table 4.11-1. SURVEY OF IN-LINE REHEAT SYSTEM
I
to
en
No.
1
2
3
Power plant
Colstrip
Montana Power
Hawthorne
Kansas City Power and
Light
Four Corners
Arizona Public Service
Heating
source
Steam ?
150 psig
and 360'F
Hot water
$ 325°F
Steam §
600 psig
and 650»F
Degree
of
reheat
•F
30
50
50
Tube
style
Plate
Finned
Finned
Heat
exchanger
diameter
inches
1.0
1.5
1.0
Number of
tube vanes
NA
NA
2
Soot
blowing
NA
Steam soot
blower
Steam blower
once a day
Material of
construction
Bottom section is
Inconel 625, top
section is Hastelloy G
Carbon steel
Carbon steel
NA - Not available.
-------
Table 4.11-2. SURVEY OF DIRECT-FIRED REHEAT SYSTEM
I
K)
Ho.
1
2
3
4
Power plant
Phillips
Duquesne Power
Elrama
Duquesne Power
Bruce Mansfield
Pennsylvania Power
Paddy ' s Run
Louisville Gas
and Electric
Deqree
of
reheat,
OF
30
30
40
40
Fuel
type
No. 2 fuel
oil
No. 2 fuel
oil
No. 2 fuel
oil
Natural gas
Fuel and Combustion
Sulfur
content,
%
0.3
0.3
0.2
0
Fuel
rate
440 gal/h
440 qal/h
NA
20,000
scfh
Excess
air,
%
NA
NA
NA
6-9
Gas
temperature,
Op
3000
3000
NA
NA
NA = Not available.
-------
4.11.11-2 Four Corners (Arizona Public Service)—
In this horizontal scrubber reheater, the carbon steel
tubes developed pinholes. Metallurgical examination revealed
that the pinholes, which developed from inside the tube were
caused by stagnant condensate and air that leaked into the
reheat system. It is not known when the pinholes developed;
however, the system had been idle for a year before testing, and
it is likely they occurred during this idle period. This prob-
lem was resolved by installation of a nitrogen (N2) blanket
system in the reheater loop to exclude air from inside the tube
The pinholes were brazed so that system testing could continue!
. This horizontal scrubber has now been shut down indefi-
nitely. There is no feedback regarding the nitrogen blanket
performance.
4.11.11-3 Colstrip (Montana Power)—
in this in-line reheater installation no corrosion problems
have been reported. Materials of construction are Inconel 625
and Hastelloy G. Although some loose scale formed on the re-
heater tubes, it did not cause any operating problems.
4.11.11.4 Phillips Station (Duquesne Light & Power)--
In this direct-fired reheat system, the design value for
reheat temperature was not high enough to protect the lined
stack. The blower failed because of mechanical problems.
Problems occurred with the oil pumps, burners, and temperature
control system, and corrosion has been reported in the combus-
tion chamber. No priority has been placed on operating the
reheater because of the oil shortages. A new acid-proof stack
liner has been installed. The scrubber system is currentlv
operated with a wet stack.
4.11.11.5 Elrama Station (Duquesne Light & Power)—
In ^is di"?*"*1*6'1 reheat system, problems similar to
those at the Phillips station have been encountered (see Section
4.11•11-4 ) •
4.11.H-6 Bruce Mansfield (Pennsylvania Power Co. ) —
In this direct-fired reheat system, combustion problems
have occurred because of the flue gas mixing with the combustion
air. This in turn has caused problems in maintaining a flame in
the combustion chamber.
Vibration in the reheat system is another problem All
these problems are to be resolved by the equipment supplier, but
plans for doing so were not available at the time of this writ-
ing.
4.11-27
-------
4.11.11.7 Paddy's Run 6 (Louisville Gas & Electric Co.)—
No problems have been reported in this direct-fired reheat
system.
4.11.12 Recommendations
There is a good deal of disagreement over the selection of
the appropriate type of reheat system and material of construc-
tion. Any one of the following four options is recommended, but
each has drawbacks:
(1) An in-line reheat system constructed of carbon steel,
which uses finned tubes and hot water.
(2) An in-line reheat system employing steam and a plate-
coil-type heat exchanger constructed of Hastelloy.
This type can be corrosion-resistant; however, use of
Hastelloy raises the capital cost.
(3) An indirect hot-air reheat system constructed of
carbon steel. The main drawback of this option is
that it involves heating a portion of ambient air and
hence has a higher operating cost.
(4) A direct-firing reheat system using oil. Its main
drawback is that oil is expensive and in short supply.
Other general recommendations are as follows:
0 A soot blower should be installed.
0 An efficient mist eliminator should be installed to
decrease the load on the reheat system.
0 Gas should be heated by 25° to 50°F to prevent down-
stream water condensation.
4.11-28
-------
REFERENCES
1. Choi, P. S. K., S. A. Bloom, H. S. Rosenberg, and S T.
DiNovo. Stack Gas Reheat for Wet Flue Gas Desulfurization
Systems. EPRI Research Project prepared by Battelle Colum-
bus Laboratories. November 1, 1976.
2. National Air Pollution Control Administration. Sulfur
Oxide Removal from Power Plant Stack Gas ~ Use of Lime-
stone in the Wet Scrubbing Process. Conceptual Design and
Cost Study Series, Study No. 2. Prepared for Tennessee
Valley Authority. 1969.
4.11-29
-------
BIBLIOGRAPHY
Hollinden, G. A., et al. Reheat Study and the Corrosion-Erosion
Tests at TVA's Colbert Pilot Plant. EPRI Report RP537-1.
September 1978.
PEDCo Environmental, Inc. EPA Utility FGD Survey. Prepared
bimonthly under Contract No. 68-01-4147.
4.11-30
-------
CONTENTS
4.12 CORROSION
Page
4.12.1 Introduction
4.12-1
4.12.2 Types of Corrosion
* • JL £ ™ J.
4.12.3 Corrosion Design and Material Selection 4.12-2
4.12.4 Coatings
4.12-4
4.12.5 Construction Techniques A i •> o
Tf • -L^HO
4.12.6 Process Operation /• i o o
**. J.^—o
4.12.7 Reheaters
T; . JL£~a
4.12.8 Stacks 4 12_15
4.12.9 Examples of Existing Plants 4.12-16
yf'^'o'J °reen *iver (Kentucky Utilities Company) 4.12-16
4.12.9.2 Cane Run Station 4 (Louisville Gas and 4 12-16
Electric Company)
4.12.9.3 Conesville No. 5 (Columbus and Southern 4.12-16
Ohio Electric Co.)
4.12.9.4 Phillips and Elrama, Boiler No. 2 and One 4 12-17
Other (Duquesne Light Company)
4.12.9.5 Hawthorn Nos. 3 and 4 (Kansas City Power 4 12-17
and Light Company) '
4.12.9.6 Bruce Mansfield Nos. 1 and 2 (Pennsylvania 4 12-17
Power Company)
References 4>12_i8
Bibliography 4 12_2Q
4.12-21
4.12-i
-------
4.12 CORROSION
4.12.1 Introduction
Consideration of the corrosion phenomenon is important in
the design and construction of any plant and is particularly so
in the case of flue gas desulfurization (FGD) systems. Many FGD
systems have experienced severe corrosion despite ongoing ef-
forts by the utilities to find corrosion-resistant materials.
The materials of construction for the scrubber were discussed in
Section 4.6.3. This section discusses some of the more impor-
tant factors concerning corrosion of stacks and reheaters as
well as the plant experience of operational lime FGD systems.
4.12.2 Types of Corrosion
There are various specific types of corrosion. General
corrosion, pitting, crevice corrosion, intergranular corrosion,
stress corrosion cracking, and erosion-corrosion are the types
of corrosion most commonly found in FGD systems and hence will
be given special consideration. other types of corrosion will
also be discussed.
General corrosion is the uniform dissolution of an entire
metallic surface.1 it is the best understood of all the corro-
sion processes. The required conditions are usually not very
specific, occurring over wide variations in solution composi-
tion, pH, etc.^ General corrosion is a controllable problem in
that the lifetime of the equipment can be accurately predicted
by laboratory tests or theoretical calculations.
Pitting is intense attack at certain locations on the
metallic surface because of local film breakdown.3'4 Pits or
holes form and usually result in rapid perforation of the mate-
rial.
Crevice corrosion is in many ways similar to pitting ero-
sion: intense attack occurs within preexisting crevices as a
result of the formation of concentration cells, etc.5
Intergranular corrosion is localized corrosion occuring at,
or immediately adjacent to, the grain boundary. Chemical heter-
ogeneities, such as a segregate or precipitate at the grain
boundary, cause a local galvanic cell to be established, and the
grain boundary dissolves.
Stress corrosion cracking encompasses a complex spectrum of
failure mechanisms.6'7 The area is still one of intense re-
search. Essentially, failure is caused by the combination of a
specific environment, a tensile stress of sufficient magnitude
4.12-1
-------
and, usually, a specific metallurgical requirement in terms of
the composition and structure of the alloy. Alloys subject to
stress corrosion cracking are not normally considered to be
markedly susceptible to general corrosion in the environment.8
Erosion-corrosion is the effect of the joint action of
mechanical forces and a corrosive environment. Debris or sus-
pended solids impinge upon a susceptible surface, destroy the
protective surface film, and thereby expose the alloy to the
corrosive agent. Cavitation and subsequent bubble collapse have
the same effect.9
Other corrosion processes of interest include corrosion
fatigue and galvanic corrosion. In corrosion fatigue, the
fatigue life of the structure is greatly reduced by the effect
of the corrosive environment.10 The process is not well under-
stood, but in several ways resembles both stress corrosion
cracking and erosion-corrosion. Galvanic corrosion occurs when
two dissimilar metals are joined in a conducting solution.
Severe corrosion of the less noble metal occurs at the metal-
metal junction. Many common forms of joining (welding, brazing,
soldering, bolting) provide junctions at which galvanic corro-
sion can develop; this should be considered in the design stage.
Galvanic corrosion on a microscopic scale can also occur between
the constituents of multiphase alloys.
4.12.3 Corrosion Design and Material Selection
A structural design that will minimize or allow for erosion
is of the highest importance. The science of corrosion design
is closely related to economics and process safety.
For a given piece of equipment and a given set of operating
conditions there is often a wide choice of possible materials.
Selection of a low-cost material when general corrosion prevails
in the scrubber environment can be the most economical option in
that corrosion will occur at a predictable rate and the equip-
ment can be, periodically replaced during planned maintenance
periods; however, the expense of frequent maintenance or exten-
sive downtimes of a particular section of a plant (and the
possible need for backup equipment) can be limiting factors.
Use of more expensive corrosion-resistant materials is another
alternative; however, these materials must be selected carefully
to ensure, first, that they will have a lifetime that will
justify the higher cost and, second, that they will not be prone
to the more severe types of corrosion such as pitting corrosion
and stress corrosion cracking. The corrosion rates for these
types of failure are difficult to predict, and the results can
be catastrophic if corrective action is not taken. Neverthe-
less, it is possible to derive theoretical relationships that
enable alloys susceptible to stress corrosion cracking in a
4.12^2
-------
given environment to be used with comparative safety. In recent
years it has been realized that real structures inevitably
contain ! flaws That is, actual cracks already exist and the
engineer, has therefore to develop criteria to ascertain whether
and at what applied load (a ), any of these cracks will
propagate to the point of failure.
• I
Fracture mechanics is the analysis of the stresses in the
neighborhood of preexisting cracks of specified geometry ll The
effective stress intensity, K, close to the root of a given
crack is directly proportional to a the crack length. Values
£orJ-^§,QQ can be obtained by experiment, and hence loading
conditim can be established for each section of the plant so
that the stress corrosion cracks do not propagate Current
'
Other alternatives in material selection are to isolate the
structural alloy from the corrosive environment by a barrier or
to design corrosion control systems. These include the addition
of inhibitors in circulating liquids or in paint coatings, or
the superimposition of anodic or cathodic protection.1^ The
concept of isolating the structural alloy from the environment
is simple, but it has proved difficult to achieve It is
discussed more fully in Section 4.12.4.
Correct materials selection and corrosion design must avoid
junctions between dissimilar metals that would promote galvanic
corrosion. One way of doing this is to use washers to insulate
bolts from a structure. Other possibilites include painting the
more noble constituent of a couple to give a large anode area
and a very small cathode area, which will ensure a very slow
corrosion rate. Some problems arise here, however. It is not
always possible to predict from the electrochemical table which
metals will be anodes and which will be cathodes when joined, or
what the cell voltage will be. For example, if a highly reac-
tive metal is covered by a protective oxide film, its ions
cannot easily enter into solution, even though the metal is
readily lonizable. Galvanic corrosion problems can be encoun-
tered in welded structures. This is discussed more fully in
Section 4.12.5 (Construction Practices) .
i
Correct material selection requires an understanding of
corrosioin, a careful evaluation of process conditions in the
proposed, section of plant, an extensive literature survey and
practical research. Much can be learned from past failures
Reference to standard works would also prove beneficial.14
= -K for ther most severe stress condition (plane strain)
in an environment promoting stress corrosion cracking.
4.12-3
-------
Proper structural design is equally important. Crevices,
high flow areas, and areas of evaporation are just a few exam-
ples of sites that promote corrosion. Debris collecting in a
crevice can give a low-oxygen concentration, and possibly set up
an anode-cathode system; debris can similarly collect at the
flashing around welds. High flow areas may cause the breakdown
of protective oxide fibers. Areas of evaporation can lead to
salt concentration and the possibility of increased corrosion.
External environmental conditions have their effect, too.
For example, marine environments have a high concentration of
chlorides. This can be of importance in areas where dew or
rainwater collects and possibly evaporates, which can lead to
chloride concentration and subsequent corrosion problems.
Incorporating additional components in an FGD system can be
helpful in preventing corrosion. For example, the inclusion of
reheaters immediately after the scrubbing module will reduce
corrosion problems in stacks and achieve plume enhancement and
pollutant dispersion. The reheaters heat the gas and reduce the
probability of condensation on the stack wall. It is necessary,
however, to optimize conditions to obtain a balance between the
occurrence of condensation and the maximum service temperature
of the stack material. Correct materials selection and design
are required to prevent severe corrosion of the reheaters. This
will be discussed further in Section 4.12.7.
4.12.4 Coatings
Coatings are of particular interest because they are used
extensively in scrubber systems. This section will discuss
paints, liners, bricks, and their application. The object of a
coating is to prevent the access of corrosive liquors to the
bare metal surface. In practice, however, all coatings are
somewhat permeable and a dual objective of some is to treat any
permeating liquor and thereby prevent corrosion.
Paints are suitable in some situations. Although they are
porous, they decrease the available surface area. Inhibitors
such as natural oils, lead, chromate, phosphates, and silicates
are often incorporated into the paint. Paints can be used
either as a final treatment or as an intermediate treatment to
prevent corrosion during construction when field application of
a liner or brick structure is intended.15,16
Glass linings have been used on heat exchangers in Russia
to protect steel in chloride environments.17 Although the heat
transfer coefficient of a glass lining is initially 10 percent
lower than that of carbon steel, it remains constant, whereas
the steel surface oxide film becomes thicker and its heat trans-
fer coefficient is progressively lowered.
4.12-4
-------
Linings of concrete, e.g., Precrete, have been employed in
FGD stacks with varying success (see Section 4.12.9).
Various brick linings are also used to protect stack struc-
tures in FGD units and also in venturi throats. They have
advantages in being able to withstand high temperatures and
corrosion, but those currently used have the disadvantage that
acid seeps through the brick mortar joints and leads to corro-
sion of the supporting structural steel.
Irrespective of the actual coating specified for the
stacks, an important consideration is its correct application to
the structural steel so as to prevent subsequent delamination
This can be accomplished by the employment of highly skilled
operators and the use of rigorous inspection techniques.
Correct preparation of the surface is also important, and the
following general guidelines are provided.
Prior to application of a coating:
(1) All surfaces should be cleaned in accordance with
Steel Structures Painting Council (SSPC) Spec. SP
6-63.18
(2) All air sources, including blasting, cleaning, and
spraying should be free of oil and water. Effective
traps and filters should be installed and frequently
inspected.
(3) Prior to any required abrasive cleaning, surfaces
should be precleaned according to SSPC Spec. SP 1-63
to remove grease, oil, and loosely adhering deposits.
(4) Cleaning should not be performed if any of the follow-
ing conditions are suspected or evident.
a. Moisture is present on the surface.
b. Moisture condensation is imminent.
c. The abrasive is wet.
d. The blasting operation interferes with painting.
e. The cleaning equipment is not in good operating
condition.
(5) Surfaces should be blasted in conformance with the
following requirements.
a. Blasting equipment should be in good condition to
prevent any moisture, oil, or other foreign
matter from depositing on the surface during
blasting in accordance with SSPC Spec. SP 6-63
Pressure should be 100 psi as measured at the
nozzle.
4.12-5
-------
b. A depth profile between 2 and 4 mils should be
attained.
(6) The surface should be cleaned by a vacuum cleaner and
dust blown off with compressed air (free from oil and
water) after sandblasting. Any oil, grease, or other
detrimental materials adhering to the surface after
blasting should be removed by solvent washing and
reblasting.
(7) The blast-cleaned surfaces should be coated within 4
to 6 hours of blasting and before any rusting occurs.
If rusting does occur, the surfaces should be re-
blasted to the degree specified.
(8) When a wet or water vapor sandblasting method of
surface preparation is required in hazardous areas
the procedure written by the SSPC should be followed!
(9) Machined parts, bearings, and motors must be adequate-
ly protected during sandblasting to prevent sand from
endangering their operation. If these pieces of
equipment cannot be protected, they should be removed.
Specification SP 6-23 of SSPC is a commercial blast clean-
ing process stating the maximum amount of residue permitted on a
steel surface before application of a coating. Near-white metal
blast cleaning (SSPC Spec. SP 10-63T) and white metal blast
cleaning (SSPC Spec. SP 5-63) result in cleaner surface finishes
than produced by commercial blast cleaning. The following are
definitions supplied by the SSPC for the three types of sand-
blasting:
0 Commercial Blast-Cleaned Surface Finish
All oil, grease, dirt, rust, scale, and foreign matter
have been completely removed from the surface and all
rust, mill scale, and old paint have been completely
removed except for slight shadows, streaks, or dis-
colorations caused by rust stain, mill scale, oxides,
or slight, light residues of paint or coating that may
remain; if the surface is pitted, light residues of
rust or paint may be found in the bottom of pits; at
least two-thirds of each square inch of surface area
should be free of all visible residues and the re-
mainder limited to the light discoloration, slight
staining, or light residues mentioned above.
0 Near-White Blast-Cleaned Surface Finish
All oil, grease, dirt, mill scale, rust, corrosion
products, oxides, paint, or other foreign matter have
been completely removed from the surface except for
very light shadows, very slight streaks, or slight
4.12-6
-------
discolorations caused by rust, stain, mill scale
oxides, or slight, light residues of paint or coating
that may remain. At least 95 percent of each square
inch of surface area should be free of all visible
residues, and the remainder limited to the light
discoloration mentioned above.
0 White Metal Blast-Cleaned Surface Finish
A surface with a gray-white, uniform, metallic color,
slightly roughened to form a suitable anchor pattern
for coatings. The surface, when viewed without mag-
nification, should be free of all oil, grease, dirt,
visible mill scale, rust, corrosion products, oxides
paint, or any other foreign matter. The color of the
clean surface may be affected by the particular abra-
sive medium used.
The following is a list of materials typically used when
pressure blasting for a specified anchor pattern. The profile
depth is an approximation and not a minimum or maximum depth
obtainable.
2-mil Profile
16/35-mesh silica sand
G-40 steel grit
S-230 steel shot
36-mesh garnet
36 Grit aluminum oxide
Clemtex No. 3
Black Beauty BB-50 or BB-2040
2.5-mil Profile
8/35-mesh silica sand
G-40 steel grit
S-280 steel shot
16-mesh garnet
24 Grit aluminum oxide
Clemtex No. 2
Black Beauty BB-400
3- to 4-mil Profile
8/20-mesh silica sand
G-25 steel grit
S-330 or 390 Steel shot
16-mesh garnet
16 Grit aluminum oxide
Clemtex No. 2
Black Beauty BB-40 or BB-25
4.12-7
-------
Shot-grit blasting has additional advantages in that it
gives a deformed surface layer that resists stress corrosion
cracking.19
4.12.5 Construction Techniques
From some points of view, the optimum design of any struc-
ture calls for its fabrication from a single sheet of metal,
which permits a smooth, continuous surface. Welds on overlapp-
ing pieces of metal create stagnant areas where corrosion attack
intensifies. Since it is usually impossible to design a struc-
ture with no welds, care should be taken in constructing the
welds. Crevices should be avoided and the possibility of gal-
vanic corrosion considered. Figures 4.12-1 through 4.12-5 show
some recommendations for welding techniques. Generally, the
filler metal should be more noble than the structural metal in
order to prevent galvanic corrosion.
4.12.6 Process Operation
Process operation is another important consideration in the
prevention of corrosion. Monitoring flow rates, temperatures,
pH, etc., will indicate when process conditions are deviating
from the design conditions. Chemical additions to restore pH or
bypasses to protect delicate parts of the process are suggested.
For example, at Green River, where the FGD system has been
retrofitted, the existing stack can be used for hot flue gases
when the FGD system is not operating. This protects the tem-
perature-sensitive walls of the FGD stack.
4.12.7 Reheaters
To date, only limited data have been published on reheaters
and corrosion. The work of Zotor et al. has already been dis-
cussed in Section 4.12.4.17 The Tennessee Valley Authority
(TVA) has also recently completed a study.20
In the TVA study, the main material problems arose from
mud-like carryover from the mist eliminator. The carryover was
deposited on the reheater tubes, the moisture evaporated, and
concentrations of chlorides and sulfur oxides caused pitting
corrosion. The use of air blowers should prevent this deposit
buildup and hence reduce corrosion. Problems were also experi-
enced because of shutdowns. When the temperature of the re-
heaters dropped below the acid dewpoint, acid condensed on the
tubes and caused severe corrosion. Condensation on the tubes
could be prevented by maintaining steam flow to the reheaters,
even when the scrubber is not operating.
During the TVA program, the materials listed in Table
4.12-1 were studied. The report notes that although these
results seem, on the whole, to be satisfactory, many can be
susceptible to stress corrosion cracking.
4.12-8
-------
POSSIBLE
CORROSION
AREA
Figure 4.12-1. The top drawing shows a continuous weld and the
lower one a series of spot welds. Spot welding is not recommended
since it has areas for potential corrosion around the joint.
4.12-9
-------
EB INSERT RING
CONSUMABLE FLAT RING
POSSIBLE CORROSION AREA
STANDARD BACKING RING
Figure 4.12- 2. The three possible ways to join two pieces
of pipe by using an EB insert ring, a consumable flat ring, or a
standard backing ring. The standard backing ring weld is not
recommended because of the possible corrosion area
between the ring and the pipes.
4.12-10
-------
DOUBLE BUTT WELD
LAP WELD, SINGLE FILLET
POSSIBLE CORROSION AREA
LAP WELD, DOUBLE FILLET
v///
\\X\\\\I
POSSIBLE CORROSION AREA
Figure 4.12-3. Examples of a double butt weld, single fillet
lap weld, and double fillet lap weld. The double butt weld is
recommended because the joint is filled.
4.12-11
-------
ROUGH WELD
SMOOTH WELD
SMOOTH THE WELDS BY GRINDING
WELD SPLATTER
REMOVE ALL WELD SPLATTER
ROUGH WELD
SHARP
CORNER
ROUND ALL SHARP CORNERS
ROUND
CORNER
Figure 4.12-4. Examples of good and bad finishing
techniques for welds.
4.12-,12
-------
OPEN SPACE
Figure 4.12-5. Always make a complete full weld.
4.12-13
-------
Table 4.12-1. CORROSION OF REHEATER TUBES
FROM COLBERT PILOT PLANT AFTER 3800 HOURS*9
Tube
No.a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Alloy
Inconel 625
Inconel 625
Inconel 625
Incoloy 825
Type 316L stainless steel
Cor-Ten A
Cor-Ten A
Hastelloy C-276
Cor-Ten A
Cor-Ten A
Inconel 625
Incoloy 625
Type 31bL stainless steel
Cor-Ten A
Cor-Ten A
Hastelloy C-276
Cor-Ten A
Cor-Ten A
Inconel 625
Inconel 625
Corrosion rate,
mils/yr
1
2
1
b
b
8
9
1
8
9
Neg.
b
b
13
12
b
8
8
1
Neg.
a Tubes are numbered from left to right, facing the incoming
gas.
b „.
Pits were visible, but they were too small to measure depth,
Neg. - Negligible
4.12-14
-------
Boiler stacks are massive, expensive structures that are
difficult to maintain or replace and are subject to corrosion
Before the introduction of FGD systems, and where such controls
are unnecessary, stainless and mild steels have proved to be
satisfactory stack construction materials. The use of Cor-Ten
has also been suggested. The satisfactory performance record
was due to the fact that the exhaust gases were dry and hot
(above the acid dewpoint) and did not condense on the stack
the" topaof thf stack ^ ^ experienced from condensation at
The use of wet FGD systems, which saturate the flue gas
with "loist necessitates the use of more sophisticated male-
rials for the stack and the reheaters. in a few cases, very
expensive corrosion-resistant alloys have been used, but more
commonly bricks are employed as stack liners. As previously
discussed in Section 4.12.4, seepage is a problem with brick
linings. Some utilities are attempting to solve this by using a
positive pressure between the steel structure and the brickwork.
Other coatings have been used in stacks to prevent corrosion
With or without reheat, most of the coatings have per?ormed
satlS^n° H-Y UndGJ n°rmal °Perating conditions; however, when
the FGD system is bypassed, the exhaust gases are considerably
hotter (in excess of 284°F) and may damage the lining. The
stack shell is exposed and is then susceptible to corrosion.
precrete has been used at some locations with good results.
It is not corrosion-resistant, but it fails at a predictable
rate. The utility companies have been able to apply a thick
layer of Precrete to stack shells at a reasonable cost. Pre-
crete may not be applicable for new large boilers, which have
limited downtime for relining.
4.12.8 Stacks
Boiler stacks are massive, expensive structures that are
difficult to maintain or replace and subject to corrosion.
Before the introduction of FGD systems, and where such controls
are unnecessary, stainless and mild steels have proved to be
satisfactory stack construction materials. The use of Cor-Ten
has also been suggested. The satisfactory performance record
was due to the fact that the exhaust gases were dry and hot
(above the acid dewpoint) and did not condense on the stack
walls, although problems can be experienced from condensation at
the top or the stack.
The use of wet FGD systems, which saturate the flue aas
with moisture, necessitates the use of more sophisticated mate-
rials of construction for the stack and/or the use of the re-
heaters. In a few cases, very expensive, corrosion-resistant
alloys have been used, but more commonly bricks are employed as
4.12-15
-------
stack liners. As previously discussed in Section 4.12.4, seep-
age is a problem with brick linings. Some utilities are
attempting to solve this problem by using a positive pressure
between the steel structure and the brickwork (see Section
4.12.9.4). Other coatings have been used in stacks to prevent
corrosion. With or without reheat, most of the coatings have
performed satisfactorily under normal operating conditions;
however, when the FGD system is bypassed, the exhaust gases are
considerably hotter (in excess of 284°F [140°C] ) and may damage
the lining. The stack shell is exposed and is then susceptible
to corrosion.
Precrete has been used at some locations with good results.
It is not corrosion-resistant but it fails at a predictable
rate. Utility companies have been able to apply a thick layer
of Precrete to stack shells at a reasonable cost. However,
Precrete may not be applicable for new large boilers, which have
limited downtime for relining.
4.12.9 Examples of Existing Plants
This section details examples of materials problems in
existing plants.
4.12.9.1 Green River (Kentucky Utilities Company)—
The stack lining was initially Carboline, which had to be
replaced because of excessive flaking. The stack was relined
with Precrete in 1977 to withstand design capabilities of a pH 5
to 9 and up to 300°F.
4.12.9.2 Cane Run Station 4 (Louisville Gas and Electric Com-
pany )~
The ductwork at the Cane Run plant was originally made of
mild steel and coated with Carboline. Bubbles formed in the
Carboline coating, but it did not fail. This coating was re-
moved in May 1977 and replaced with Plasite 4005. The per-
formance of the Plasite 4005 has not been reported.
The 250-ft stack for Boiler 4 is made of concrete and lined
with acid brick. The brick was coated with a Carboline layer.
The Carboline coating and the brick began to fail, and in May
1977 they were removed. A 2-in. layer of Precrete was installed
in their place. The stack was inspected after a year of opera-
tion; the Precrete was holding up very well.
4.12.9.3 Conesville No. 5 (Columbus and Southern Ohio Electric
Co.)--
The ductwork of the FGD system on Boiler 5 is made of
Cor-Ten steel and coated with Saurereisen 54 Gunite. No adverse
reports have been received regarding its performance. The outer
shell of the 800-ft stack is concrete. The shell was originally
4.12-rie
-------
lined with a Cor-Ten steel flue, which was coated with a Ceil
cote liner. This liner failed because of continual bypass of
the scrubbing system. When this occurred, flue gases enterino
the stack were too hot for the Ceilcote liner tohandll OncI
the liner failed, the Cor-Ten flue was exposed to the flue gases
and began to corrode. it was replaced in December 1977 with a
fireclay brick liner using Saurereisen 65 mortar.
4.12.9.4 Phillips and Elrama, Boiler No. 2 and One Other20
(Duquesne Light Company)—
The lime FGD systems at these plants are identical. The
original ductwork was made of 316L stainless steel, it corroded
and was replaced with a mild steel shell coated with Ceilcote
103. There have been no subsequent problems.
Each stack consists of a concrete shell with an acid brick
liner The mortar used is Saurereisen 65. Sulfuric acid seep-
ing through the brick caused problems. The brick liner was
repaired and a Positive pressure maintained in the space between
the concrete and brick. It is hoped this measure will prevent a
recurrence of sulfuric acid seepage.
4.12.9.5 Hawthorn Nos. 3 and 4 (Kansas City Power and Light
Company)--
The ductwork and the stack at the Hawthorn plant had no
corrosion problems. The ductwork was built of carbon steel and
lined with a 2-in. layer of Gunite. The stack was built of
steel and lined with Gunite. These materials were installed in
1972 when the FGD systems went into service.
4.12.9.6 Bruce Mansfield Nos. 1 and 2 (Pennsylvania Power
Company)—
Unit 1 has two carbon steel stacks, which were originally
lined with polyester flakeglass. In May 1977, the polyester
flakeglass lining failed and the flue gas caused extensive
damage to the carbon steel structure. After a 10-week main-
tenance outage, the stacks were repaired one at a time while the
unit was operating at approximately half load. The unit came
back to full load by January 1978. A very similar polyester
flakeglass lining has held up reasonably well in the scrubber
and absorber vessels and the ductwork. Plant personnel have
tested patches of various linings in one stack of Unit No 2
4.12-17
-------
REFERENCES
1. Shrier, L. L. Corrosion. 2nd ed. Newnes Butterworth,
London, Boston, Sydney, 1976.
2. Pourbaix, M. Atlas of Electrochemical Equilibria in Aque-
ous Solutions. Pergamon Press, London, 1966.
3. Hoar, T. P. et al. Corro. Sci., 5, 279, 1965.
4. Hoar, T. P. Corros. Sci., 7, 341, 1967.
5. Fontana, M. G., and N. D. Greene. Corrosion Engineering.
McGraw-Hill, 1967. pp. 28-115.
6. Parkins, R. N. Ref. 1, pp. 8.03-8.35.
7. Fundamental Aspects of Stress Corrosion Cracking.
N.A.C.E., Katy, Texas, 1969.
8. Terns, R. D. Thesis, University of Newcastle Upon Tyne,
1978.
9. West, J. M. Electrodeposition and Corrosion Processes, Van
Nostrand Reinhold, 1971.
10. Handbook of Fatigue Testing. ASTM, 1974.
11. Ref. 9, p. 133.
12. Chell, G. G. CEGB Research, Jan. 1978, pp. 36-44. (CEGB,
Leatherhead, Surrey, England).
13. Lui, A. W., and G. R. Hoey. Materials Performance 15,
13-16, 1976.
14. Schweitzer, P. A. Corrosion Resistance Tables. Marcel
Dekker, Inc., 270 Madison Ave. New York, New York 1976.
15. Brooks, M. E. Bull Inst. Cor. Sci. Tech., No. 4, 2-9,
1977.
4.12-18
-------
16. Lapasin, R. et al. Brit. Cor. J., 12, 92-102, 1977.
17. Zotov, A. G. et al. Teploenergetika, 58-60, May 1976.
18. Kaplan N and J. C. Herlihy. lERL-Research Triangle
Park, North Carolina. y
19. Brown, A., and R. Wilkins. Eurocor '77, pp. 563-570.
20. Hollinden, G. A. et al. Reheat Study and the Corrosion-
erosion Test at TVA's Colbert Pilot Plant. Published by
EPRI (EPRI RP 537-1). Palo Alto, California, September
4.12-19
-------
BIBLIOGRAPHY
NACE Surface Preparation Handbook. NACE, Katy, Texas.
staehle, R.w. et. al., eds. Stress Corrosion Cracking and
Hydrogar Embrittlement of Iron Base Alloys (Firming France
1973). NACE (No. 51102).
Resolving Corrosion Problems in Air Pollution Control Equipment
NACE 1976 (No. 52100).
Surface Preparation Specifications. Steel Structures Painting
Council, Pittsburgh, pp. 3, 31-38, 47-50, 1963.
Heil Process Equipment Company. Surface Preparation.
Dudick Corrosion Proof, Inc. Application Guide: Surface Prepa-
ration.
Corrosioneering, Inc. Resista-Flake 1200 Series. Thermosetting
Lining Systems Application Guide RFA 1200.
Wyatt, C.H. Coatings—A Chain Link Fence. In: Corrosion
Problems in Air Pollution Control Equipment Symposium, Atlanta,
Georgia, 1978.
Singleton, W.T., Jr. Protective Coatings Formulated from Vinyl
Ester Resins for the Air Pollution Industry. In: Corrosion
Problems in Air Pollution Control Equipment Symposium, Atlanta
Georgia, 1978.
Landrum, R.J. Designing for Corrosion Resistance of Air Pollu-
tion Control Equipment. In: Corrosion Problems in Air Pollu-
tion Control Equipment Symposium, Atlanta, Georgia, 1978.
Graver, D.L. Forms of Corrosion. In: Corrosion Problems in
Air Pollution Control Equipment Symposium, Atlanta, Georgia
1978. '
4.12-20
-------
GLOSSARY
Anchor pattern - The surface morphology that ensures adhesion of
a coating to the surface.
concentration cell - An area in which dissolved ion concentra-
tions differ from the bulk environment; it can include increased
concentrations of dissolved species or decreased concentrations
of a dissolved species; the pH may be affected; these differ-
ences can occur in crevices, cracks, pits, or restricted flow
areas; the term "occluded cell" is a concentrations cell in
which the oxygen concentration is reduced.
Film - A barrier between the metal and the solution formed by
the reaction between the same and usually consisting of a metal
oxide; this surface film may be protective and prevent further
general corrosion (e.g., the film on stainless steels) or it may
offer no protection at all (e.g., a rust film).
High flow areas - Here, an area in which high flow rates may
impinge on the surface film, damaging it, and exposing ?he
metal to further corrosive action. Apw*,j.«y
Plastic deformation - Damage to a material such that it cannot,
of its own accord, regain its original shape and character^
istics.
Precipitate - One or more of the constituents of an alloy that
settle out as a distinct phase in the matrix- this nhase can hf»
dispersed throughout the matrix (e.g.f g?apWte in grey Sas?
iron) or localized to particular areas (e./, chromium carbide
precipitated at the grain boundary when stainless steels suffer
from weld decay).
Segregate - Similar to precipitate, except that the constituents
do not settle out as a separate phase, but merely become con-
centrated within certain areas of the matrix.
Shot-grit blasting - Blasting using a mixture of 80 percent shot
and 20 percent grit; the grit gives a better anchor pattern
whereas the shot has other beneficial effects va^-exn
4.12-21
-------
CONTENTS
4.13 INSTRUMENTATION
Page
4.13.1 Introduction . , .
Q • JLo — J.
4.13.2 pH Control 4 13_1
4.13.2.1 Reference Electrodes A i •? i
4.13.2.2 Electrode Cleaning Devices I 1*9
4.13.2.3 Electrode Installation 41?"?
4.13.2.4 pH Controller A T^'A
4.13.2.5 Performance History 4 13-5
4.13.3 Solids Content A 13_5
4.13.3.1 Differential Pressure and Ultrasonic
Devices 4 13_5
4.13.3.2 Nuclear Absorption Meters 4*13 7
4.13.3.3 Existing Facilities 4] 13-7
4.13.4 S02 Measurement 4 13_7
4.13.4.1 Existing Facilities 4.13-9
4.13.5 Liquid Level 4.13-12
4.13.5.1 Some Design Considerations 4 13-12
4.13.5.2 Displacement Instruments 4*13-12
4.13.5.3 Differential Pressure Diaphragm 4*13-14
4.13.5.4 Capacitance Measurement 4*13-16
4.13.5.5 Ultrasonic Linear Height Gauges 4*~n ifi
4.13.5.6 Alarms y J'Tr t°
4.13.5.7 Existing Facilities 4! 13-16
4.13.6 Liquid Flow Meters 4 i3_18
4.13.6.1 Pressure Differential Instruments 4 iq IQ
4.13.6.2 Electronic Devices A i? 10
4.13.6.3 Open Channel Flow 4 1321
*
|lw Controlling and Recording Equipment 4*13-23
4.13.6.5 Existing Facilities 4 13.23
4.13.7 Control Panels and Panel Instruments 4.13-25
References
4.13-i 4.1.--26
-------
4.13 INSTRUMENTATION
4.13.1 Introduction
To date, lime scrubbers have not been highly instrumented
in comparison with many other process plants, and the instrument
reliability has been low. The previous poor performance of
instrumentation should not be used as a reason for not instal-
ling instrumentation on future systems. Instead, it should
provide impetus to install better designed process systems with
controls and instrumentation capable of functioning properly
with normal maintenance. The intent of this section is to
present the design engineer with sufficient information to be
able to purchase control equipment that is appropriate for the
job and as easy as possible to maintain.
The following paragraphs outline five common instrument
applications in a lime scrubber and suggest suitable hardware
for each. These applications include pH control, solids con-
tent, S02 measurement, liquid level, and liquid flow.
4.13.2 pH Control
Measurement of pH in the slurry of a lime scrubber is more
difficult than in many other industrial applications. Elec-
trodes are fragile devices, easily damaged by the action of an
abrasive slurry. Scrubber slurry can also form a deposit on
electrodes; this deposit acts as an electrical insulator giving
a false value of the electrode potential. Operators should
recognize that the pH electrodes of a lime scrubber will require
more maintenance and will have shorter lives than in cleaner
applications.
4.13.2.1 Reference Electrodes—
There are two types of pH sensor: a dip sensor and a
flow-through sensor.1 The dip sensor is merely inserted into a
slurry tank and can be removed for maintenance and calibration.
A flow-through sensor depends upon a continuous flow of slurry
in the sample line. Both have advantages and disadvantages.
The dip sensor is easy to maintain, but cannot be used effec-
tively inside scrubber vessels that must be gas tight. Flow-
through sensors depend upon sample lines that, if not well
designed, can block. The sensors are also prone to high rates
of erosion.
Both types of sensor have operated well in service condi-
tions, but in either case multiple sensors must be employed to
ensure accurate measurement.
4.13-1
-------
A major advance in recent years has been the development of
pressurizable "nonflowing" reference electrodes constructed of
nonbreakable plastics. Although the older "flowing-type" refer-
ence electrodes of glass construction are still sold, the newer
type is best suited to a slurry application.
Many companies now sell miniaturized electronics packages,
which can be serviced easily and quickly by replacement of
electronic modules.
No matter whose components are used, wiring between the
electrodes and the preamplifier should be as short as possible;
one vendor (Uniloc) mounts the preamplifier in the electrode
housing to eliminate the short circuits that occur readily in
this wiring. This arrangement, however, has the disadvantage of
placing the electronics in a wet atmosphere, which could lead to
failure of the preamplifier if the housing fails.
All vendors offer either voltage or current output signals,
most of which are field-adjustable for both range and span.
Almost all this equipment will provide adequate service; none
will be trouble-free.
4.13.2.2 Electrode Cleaning Devices—
The ultrasonic type of cleaner should be used with lime
scrubber electrodes, since this cleaner is specifically designed
for removal of the sort of brittle, insoluble, insulating coat-
ings that can occur in this application. Ultrasonic cleaners
work best when operated intermittently with a timed pulse de-
vice. This adaption can be purchased as a standard accessory.
Both types of sensor can be fitted with ultrasonic cleaning
devices, though it is reported that occasionally these have
caused the pH probes to break.1
4.13.2.3 Electrode Installation—
Installation techniques that have proved beneficial in
ensuring pH sensor reliability are presented in Table 4.13-1.
4.13-2
-------
Table 4.13-1. METHODS OF IMPROVING pH SENSOR RELIABILITY*
Dip-type sensor probes
Provide sufficient vessel
agitation.
Locate probe away from
quiescent zones but pro-
vide mechanical support.
provide external tank for
easy access.
provide redundant sensors
Conduct frequent cali-
bration.
Flow-through sensor elements
— •
Provide extremely short sensor
lines (1 to 2 ft), at least
1 in. in diameter.
Avoid installing sample taps
at the bottom of horizontal
slurry lines.
Provide piped-up backflushing
capability (also can be used
for calibrating).
Install upstream deflector bar
to prevent erosion of the pH
cell.
Provide redundant sensors.
Conduct frequent calibration.
is
,
Sh°Bn ^ installations that are
- e
the process line; no manufacturer recommends this type of in-
?^e electrodes' iT ^^^ *** ^^^ ^* ^nmersiin-
type electrodes. if this design is used, the unit must be
readily accessible in an open tank to ensure ease of mainte^
nance. Flow- through electrode holders, if used should !L
installed with valves to permit simplified service', stnce this
type also requires frequent service. Flow- through holders
should be supplied with slurry by a separate sample pump a
slipstream from the recirculation pump, or by means of a'
sure drop across a scrubbing nozzle. There must be enough
4.13-3
-------
pressure to produce a flow rate within the range recommended by
the electrode manufacturer. One option is to use small-diameter
(3/8 in. to 1/2 in.) tubing to maintain high velocity (~10
ft/s). Conversly, some engineers prefer to use large-diameter
tubing so that it can be reamed when it plugs. The piping
should be as short as possible and arranged so that it complete-
ly drains by gravity when shut down. At least two identical
electrode assemblies are desirable, with valves and switches
arranged for simple crossover to a set of standby electrodes
when a set requires service or a calibration check. All ampli-
fiers and calibration controls should be installed at the elec-
trode station to permit one man to perform the necessary adjust-
ments. This eliminates the need for communication between the
control room and the maintenance man during calibration.
4.13.2.4 pH Controller—
The signal from a pH measurement instrument is usually sent
to a main control panel, where a control instrument is used to
adjust the rate of lime feed. Although more sophisticated
control systems are built into some lime scrubbing FGD systems,
simple feedback control of pH has been used predominately to
date. The pH controller need not be purchased from the same
manufacturer that supplies the electrodes and amplifiers. Since
only the larger of the several specialty instrument companies
that make dependable pH measurement equipment produce control-
lers in sufficient quantity to maintain quality control, in some
instances more than one vendor should be used.
Electronic controllers have an advantage over pneumatic
instruments in that their signal conversion takes place at the
valve, which is a more favorable position in the loop, and their
operation is slightly faster. Either type should be purchased
as a three-mode instrument incorporating proportional, integral,
and derivative action, because even if three-mode control is not
required at the time, flexibility is maintained at a minimal
additional cost. Special electronic nonlinear controllers
specifically designed for difficult pH applications are also
available. These instruments were described in Section 3.0 of
this data book. Although it may be difficult to tune a non-
linear controller to match the process characteristics, the pH
control should improve.
Equipment for measuring pH is made by several U.S. com-
panies, including Beckman, Foxboro, Great Lakes, Leeds and
Northrup, and Universal Interloc. Several other companies make
part of the equipment and resell other components from both U.S.
and foreign manufacturers. Some of the foreign-made electrodes
give excellent service in difficult applications, especially
certain ones made in Japan and Switzerland and sold by labora-
tory supply jobbers such as Fisher Scientific Co. and A. H.
Thomas. All brands are electrically compatible with U.S. ampli-
fiers .
4.13-4
-------
4.13.2.5 Performance History —
D^ta,o°? °Perational PH control systems are presented in
43"-
n ,o
g?ven b4elo3w":2- **"* descriPtions of p system peranc are
Phillips and Elrama— These stations have no automatic DH
controls. The instruments only monitor the pH of the slurrv
Problems with pH at Phillips and Elrama are usually caused by
breakdowns in the lime slurry supply systems Neither station
reports pH measurement problems. y^ems. weitner station
Bruce Mansfield— At this station, the pH electrodes are
mounted in a 1-in slipstream from the recirculation line with
electrodes located on a platform with difficult access Because
cleaning and maintenance are difficult, problems Sith'dirtS and
^
Cation is
. Green River and Cane Run— si v sets of pH electrodes are
installed in the recirculation tank. Each set is checked
against the others daily. Recalibration and repairs are done
1 °f PH " 6XCellen a
Conesville No. 5— At this station, the PH electrodes are
submersed in a trough on the bleed-off line from the recircu^
lation loop. The pH is pneumatically controlled by changing ?he
W1U1 ^
Paddy ;B Run-Two sets of pH electrodes are installed in the
recirculation tank. The major operating problem has been the
manua^y
4.13.3 Solids Content
Scrubber installations should include instrumentation for
continuous control or recording of variables related to solids
content.
4.13.3.1 Differential Pressure and Ultrasonic Devices-
Slurry density can be directly measured with special dif-
ferential pressure instruments, but a 6-ft liquid depth is
needed to measure a 0.1 specific gravity span Ultrasonic
devices directly measure the percentage of suspended solids
Vibrating reed instruments measure the dampening effect of SU
slurry on vibrations from an electrically driven coil.
4.13-5
-------
Table 4.13-2. pH CONTROL INSTRUMENTATION.
OJ
I
91
Facilities
Conesville
No. 5
Columbus and
Southern
Ohio
Elrama
Duquesne Light
Phillips
Duquesne Light
Green River
Kentucky
Utilities
Cane Run
Louisville Gas
and Electric
Paddy ' s Run
Louisville Gas
and Electric
Bruce
Mansfield
Pennsylvania
Power
pH electrode assembly
Mfr.
Foxboro
Uniloc
Uniloc
Uniloc
Uniloc
Uniloc
Uniloc
Type
Immersion
Flow-
through
Flow-
through
Immersion
Immersion
Immersion
Flow-
throuqh
Model
NR
324
324
324
321
321
324
Location
Recirc. line,
bleed trough
Recirc. line
Recirc. line
Recirc. tank
Recirc. tank
Recirc. tank
Recirc. line
Single/multiple
Single
Single
Single
Multiple (1)
Multiple (1)
Multiple (2)
Single
Cleaning
type
NR
Ultra-
sonic
Ultra-
sonic
Manual
Manual
Manual
Manual
NR - Not recorded
-------
4.13.3.2 Nuclear Absorption Meters —
Nuclear absorption meters, which measure the deqree of
absorption of gamma rays from a radioactive source are ore!
f erred for this service. These instruments do not ^ physical! v
contact the slurry; they are strapped to a pipe through which
the slurry is flowing. They have the minor disadvantage of
producing a signal that is not linear with solids content iSilew
the unit purchased contains an electronic lineariler Thl
nuclear meter can be precalibrated by theoretical calculations
if an accurate chemical analysis of the slurry being iSered is
used, but vendor data for "average" slurry should I not b! used
since this may produce a calibration with a very large error
Each manufacturer specifies a source size range in mil! icSries
for each pipe size diameter, it is advisable to ourchasi on Jh^
high side of the range, since smaller sources produce erraUc or
sluggish output signals. yj.uuu^e erratic or
By government regulation, an NRC license certifying famil-
iarity. with radiation safety practices is required before Tany
'
w n '*
w ne aci, ,
plant service group is strongly recommended.
4.13.3.3 Existing Facilities—
Table 4.13-3 presents solids content design information
from several FGD installations. ebj.gn iniormation
-rv, on^v n?nMfm ^^ meter iB a low-maintenance instrument.
The only problem with these meters has been their inaccuracy and
inconsistency. At Green River and Cane Run facilities, the
density measurements are often verified by manual sampling and
•Kestincr. Bruce Mansfield, h™^^*. 2^- ^_. ^* ," . "u
4.13.4 S02 Measurement
Lime scrubbing systems are usually provided with instrumen-
tation to measure the SO2 content of gases entering and lea^ina
the scrubber. As with most instrumented analytical measure?
ments, the devices are costly and the operatingVrincites frZ
sophisticated. Available instruments operate on OMf Sf -
4.13-7
-------
Table 4.13-3. SOLIDS CONTENT INSTRUMENTATION
Facilities
Conesville No. 5
Columbus and
Southern Ohio
Elrama
Duquesne Light
Phillips
Duquesne Light
Green River
Kentucky Utilities
Cane Run
Louisville Gas
and Electric
Paddy ' s Run
Louisville Gas
and Electric
Bruce Mansfield
Pennsylvania Power
Densitv
Mfr.
Nuclear
None
None
Nuclear
Nuclear
None
Nuclear
meter
Type
K-Ray
Texas
Nuclear
NA
NA
Ohmart
Texas
Nuclear
NA
Texas
Nuclear
NA - Not applicable
NR - Not recorded
4.13-8
-------
(1) Coulometry - Gas is exposed to an electrolyte throuoh
electrSvtT^16 membrane' Chemical changes in the
* cl^ochemical oxidation
(2) Absorption spectrophotometry - Light is passed through
a gas, and the degree of absorption of certain infra
red or ultraviolet wavelengths is measured
"
The most consistent difficulty with the operation of an
tl&^t^^to^«* -^
intent of the sampling system is to remove solid particulars
and water droplets while avoiding condensation of water vapor
in practice, as the water collects it continues to absorb soj
and oxygen and creates a strong sulfuric acid solution Solids
preferentially collect on other precipitated solids and form
scale. Careful design of the sampling system is therefore
required. Electrostatic precipitators or filters heated linel
to prevent condensation, and a suitable back flush i to prevent
filter and sample pipe blockage are suggested. * This system is
shown schematically in Figure 4.13-1. t>ys>tera is
•iSiat5L+-°t^er ihan-d' .sPectr°Photometric instruments are
available that, by eliminating the sampling system, may provide
ettSer-e 01611*8 a rob* w P
1
-ah h-, a prob*' w*ic*' because of its
might be difficult to maintain. Another type uses a
The S02 concentration is generally recorded on a strip
chart potentiometric recorder. The operating record will be of
us! • ? ^ P u • °Perator in optimizing scrubber performance
and if the mechanical problems that have plagued these units can
of limrleed rate? ^doubtedly be used in closed-loop contrS
4.13.4.1 Existing Facilities —
Table 4.13-4 presents design information on S02 meters
installed in lime scrubber facilities. The only device current-
ly in service is the absorption spectrometer ^u-rrem:
4.13-9
-------
PARTICULATE REMOVAL
fl»«8flflgBtaBBBBflflB a p » B o ppt q t» »A°-R
5 5 5 a gyyg'd a e a m~n't s t vvvnt a
HEATERS
BACKFLUSH
VALVE
ANALYZER
PUMP HEATERS
HEATERS — »tBB
At t1119 8.BJUUUU
PROTECTIVE
LINING IN
PUMP
Figure 4.13-1. Ideal S02 sampling system.
-------
Table 4.13-4.
INSTRUMENTATION
Facilities
Conesville
No. 5
Columbus and
Southern
Ohio
Elrama
Duquesne Light
Phillips
Duquesne Light
Green River
Kentucky
Utilities
Cane Run
Louisville Gas
and Electric
Paddy ' s Run
Louisville Gas
and Electric
Bruce
Mansfield
(Pennsylvania
Power)
SO2 meter
Mfr.
DuPont
Lear-
Siegler
Environmental
Instruments
Environmental
Instruments
DuPont
DuPont
DuPont
DuPont
Model
460
NR
NR
NR
460
460 A
460 A
460 A
Sample
type
Wet
Dry
In situ
Dry
In situ
Dry
Wet
Dry
Dry
Dry
NR - Not recorded
4.13-11
-------
All the plants using Dupont S02 meters report very high
maintenance on the analyzer systems. Moisture condensation on
the probe, plugging of the sample lines, and frequent calibra-
tion requirements are some of the major operational problems.
Under an EPA contract, York Corporation is investigating some
modifications to blowback and the cleaning of sample lines at
the Bruce Mansfield facility. The modifications have been
successful in reducing maintenance somewhat.
4.13.5 Liquid Level
In a lime scrubber, level control is usually used to re-
lease excess slurry into a pond or thickener and may regulate
the quantity of water recycle. Local pneumatic control instru-
ments (instruments that have no external signal input) are often
used for simple level control applications (e.g., water tank
levels); when they are used, high- and low-level alarms are also
usually supplied to inform the control panel operator of mal-
functions. If electronic liquid level sensors are employed, as
on more complex applications such as the slurry tank, it is most
convenient to locate the controller on the control panel; there-
fore, supporting alarms are of less importance. Hardware for
control of liquid level should be dependable rather than abso-
lutely accurate.
4.13.5.1 Some Design Considerations—
Dependability is the most important criterion. Therefore,
the use of such systems as bubble tubes requires careful design
to prevent blockages, since they can easily become plugged in
lime slurry applications. Devices using mechanical floats are
not recommended for use with a slurry that may form deposits.
The displacement principle and the force-balanced, differen-
tial-pressure diaphragm measurement of liquid level are probably
the most satisfactory for this application. Capacitance instru-
ments are also proving dependable. Ultrasonic meters are also
available.
4.13.5.2 Displacement Instruments—
If a scrubber system contains a separate reaction tank with
an open top, an internal displacement transmitter is a good
choice for level measurement (Figure 4.13-2). A displacement
instrument is basically a simple scale, measuring the weight of
a stainless steel cylinder that is partly immersed in the
liquid. The cylinder does not float in the liquid, but as the
level rises or falls, its apparent weight decreases or increases
in proportion to the volume of liquid displaced by its submerged
portion. The instrument is mounted above the tank, and the
displacement cylinder is suspended in a pipe well or behind a
baffle to protect it from surface agitation. If liquids or
4.13,-12
-------
DISPLACEMENT
TRANSMITTER
I
(-•
U)
LEVEL
CONTROLLER
MAY BE
MOUNTED
ON CONTROL
VALVE YOKE
Figure 4.13-2. Displacement level instrument control system.
-------
solids do not impinge onto the parts of the instrument above the
surface of the liquid, and if any surface deposits are periodi-
cally removed, a displacement transmitter will be trouble-free.
4.13.5.3 Differential Pressure Diaphragm—
For measurement of liquid level in a closed vessel such as
the body of a scrubber, a flange-mounted, differential-pressure
transmitter (Figure 4.13-3) is suitable unless formation of
thick, hard deposits in the vessel is expected (in which case a
capacitance device is more suitable). This type of flange-
mounted instrument measures the force necessary to hold a flex-
ible metal diaphragm in a fixed position when one side of the
diaphragm is exposed to liquid pressure below the liquid sur-
face. There are two types of flanged differential-pressure
transmitters. The standard type, mounted on a 3-in. flange, has
the disadvantage of forming a pocket of stagnant slurry in the
vessel nozzle. Solids can collect and harden, causing the
instrument to operate improperly. The other type has an ex-
tended diaphragm that uses a 4-in. flange. The diaphragm is
placed on the end of a stainless steel cylinder that extends
through the vessel nozzle in such a way that the diaphragm is
flush with the inside wall of the vessel. The extended dia-
phragm type works best in slurry service.
Either type of flange-mounted, differential-pressure trans-
mitter is installed on the vessel without a shutoff valve;
therefore, the instrument cannot be removed for maintenance
without shutting down the scrubber. If the instrument is prop-
erly installed, this limitation is usually acceptable, since
maintenance of the diaphragm is seldom necessary. Proper in-
stallation requires that the vessel nozzle be located in a
turbulent zone of the tank, so that mild scouring of the dia-
phragm will occur and prevent scale deposits. A plastic-coated
diaphragm will often be used to minimize erosive damage.
Care should be taken in the installation of the pressure
balancing line. Use of a differential-pressure transmitter
requires that a small-diameter pipe be attached to the instru-
ment and extended to a vessel tap located well above the maximum
liquid level in the vessel. The line connects the static pres-
sure in the vessel to the back of the diaphragm, thereby balanc-
ing, or cancelling out, its effect. Many of the problems with
differential-pressure transmitters are related to the balancing
line. In a scrubber, the line is filled with water. Solids can
enter the line and plug it, or the line can lose water and
become partially filled with gas; either condition causes the
instrument to operate inaccurately or perhaps even fail. As
shown in Figure 4.13-3, the balancing line should be connected
to the vessel into a tap of at least 1-in. diameter. A small
rotameter should be installed to purge a continuous stream of
4.13-14
-------
U)
I
ROTAMETER
WATER
BALANCING LINE
LEVEL
CONTROLLER
MAY BE
MOUNTED
ON CONTROL
VALVE YOKE
I
FLANGE - MOUNTED
DIFFERENTIAL-
PRESSURE
TRANSMITTER
Figure 4.13-3. Flange-mounted differential-pressure level control system.
-------
water into the tap and thus prevent accumulation of solids. it
is also desirable to provide a tap into the vessel, located near
and at the same elevation as the transmitter. In addition a
valved tap should be supplied in the balancing line. These
connections can be used with a water-purged manometer to check
calibration of the instrument or to attach a substitute water-
purged transmitter in the event of instrument failure during
scrubber operation.
4.13.5.4 Capacitance Measurement—
For closed-tank applications, throttling control from
instrumentation based on electrical capacitance measurement
overcomes many of the problems of the differential-pressure
transmitter. Throttling-type capacitance instruments from
various suppliers vary widely, however, and the dependable ones
are quite expensive (approximately $3000 to $4000). These in-
struments, which use an electrical probe coated with TFE
(Teflon), measure the capacitance between the probe and the tank
wall with a relatively complex electronic circuit. Good quality
capacitance instruments are unaffected by deposits that accumu-
late on the probe and by the presence of spray or vessel acri-
tation. y
4.13.5.5 Ultrasonic Linear Height Gauges—
These are available from at least one manufacturer (Badger
Meter, Inc., of Tulsa, Oklahoma). The fact that Ultrasonic
Systems have no parts actually submerged in the liquid makes
them virtually free of maintenance problems. Ultrasonic energy
is transmitted by a transducer to the liquid surface. The
signal is reflected and received by another separate transducer
and the elapsed time is converted into fluid level. The meters
are constructed from stainless steel and are sealed with eooxv
resin.4 J
4.13.5.6 Alarms-
High- and low-level alarms inform the operator if a mal-
function occurs in the level control system. Completely separ-
ate instruments should be used for alarm sensing, especially if
differential-pressure transmitters are used for control sensing
Capacitance instruments are a better choice than mechanical
float or electrical conductance devices for alarm actuation.
Ultrasonic or vibrating reed probes may also be suitable in this
application.
4.13.5.7 Existing Facilities—
Data on level control equipment in existing lime scrubbers
are shown in Table 4.13-5.
The liquid level gauges have given reliable service at all
the facilities but Green River, where inaccuracy of the gauge
has sometimes necessitated manual control of pond water returns.
4.13-16
-------
Table 4.13-5. LIQUID LEVEL INSTRUMENTATION
Facilities
Conesville No. 5
Columbus and
Southern Ohio
Elrama
Duquesne Light
Phillips
Duquesne Light
Green River
Kentucky Utilities
Cane Run
Louisville Gas
and Electric
Paddy's Run
Louisville Gas
and Electric
Bruce Mansfield
Pennsylvania Power
NA - Not applicable
NR - Not recorded
Mfr.
Foxboro
Level gauge
Taylor
Taylor
B/W
NR
NR
Foxboro
Type
Transmitter
NR
NR
Float
NR
NR
Bubbler
4.13-17
-------
4.13.6 Liquid Flow Meters
Measurement of liquid or slurry flow rates is vital to the
optimization of a process plant. Although lime scrubber systems
have not used slurry flow meters to date, their adoption is
expected. The flow rate of fresh lime slurry is perhaps the
most important control application, but flow rate of recycle
slurry and of slurry drawoff to a thickener are also important
points of application.
Several physical principles are used to measure liquid or
slurry flow rate. Available instruments fall into three broad
categories: one based on mechanical measurement of pressure
differential, one encompassing various electronic measurements,
and the third designed for measurement in open channels. Each
category has its specific applications.
4.13.6.1 Pressure-Differential Instruments--
Pressure-differential instruments are best suited to clean
water flowing through piping under pressure. Examples of these
devices include orifice meters, flow nozzles, pitot tubes, Dall
tubes, venturi meters, target meters, and rotameters. Target
meters contain a metal plate in the flowing stream; this device
must not be used with any liquid that may contain abrasive
particles. Rotameters are intended primarily for local indica-
tion of small flow rates, such as water feed to a lime slaker.
The principle of measurement used in a rotameter creates a
mechanical force that is too weak for a dependable connection to
signal transmission accessories. All of the other types of
mechanical flow measurement instruments mentioned above require
the use of small ports connected into the process stream. If
the liquid contains suspended solids, the lead lines will become
plugged unless correctly designed. A continuous purge of fresh
water is needed if these instruments are to be used even with
thin slurries. The instruments also contain stagnant water,
which can freeze easily in winter weather; this necessitates the
use of heating jackets or insulation to prevent freezing.5
Pressure-differential instruments measure neither volumetric nor
mass flow rate, and their measurement is inaccurate unless the
density of the flowing stream remains constant. Measurements
are also inaccurate unless the instruments are installed with
the required lengths of straight piping both upstream and down-
stream from the meter location.
Despite their disadvantages, pressure-differential instru-
ments are in wide use. They are not only less expensive ini-
tially than electronic types, but they can also be calibrated
accurately using only standardized calculations and simple test
equipment. For these reasons, they are often used for slurries,
even though maintenance costs are high. Single-port cast ven-
turi meters are least affected by abrasive wear and suffer least
4.13-18
-------
from lead line plugging; they are about as expensive as the
electronic types. Some of the insert-type venturi meters and
flow nozzles operate almost as well and are significantly less
costly. Multiport Venturis, pitot tubes, and most types of
special flow tubes are more easily plugged by slurries. Sharp-
edged orifices are worn away quickly by abrasive slurry.
Quadrant-edged orifice plates are the least expensive practical
slurry measurement devices; they work best in vertical lines
(Figure 4.13-4).
Except for target meters and rotameters, pressure-differ-
ential instruments require the use of a differential-pressure
-transmitter, which need not be purchased from the same manufac-
turer as the meter itself. Force-balanced transmitters, such as
those made by Bailey, Fischer and Porter, and Foxboro, are most
often used, but a newer electronic-transmission principle unit
sold by Honeywell and Rosemount is gaining acceptance.
Equipment to provide freshwater purge of lead lines should
include two small purge rotameters with each transmitter. Purge
water must be filtered. If filtered water is distributed in
copper or stainless steel tubing, a single filter can be used.
Alternatively, individual filters may be supplied with each
transmitter.
4.13.6.2 Electronic Devices—
Electronic measurement of flow rate can be accomplished
with vortex-shedding instruments, ultrasonic transmission de-
vices, Doppler-effect ultrasonic meters, and electromagnetic
flow meters. The first two are unsuited to abrasive slurry.
Doppler-effect meters—The Doppler-effect ultrasonic meter
is a fairly new development that is intended for slurry appli-
cations. Its principal advantage is that the sensors are cemen-
ted to the outside of the pipe through which the slurry is
flowing; there is no penetration of the pipe. Badger Meter,
inc. (Tulsa, Okla.) supplies ultrasonic flow meters as a spool
section for attachment to metal, plastic, or asbestos cement
pipes, or for use in open channels with variable fluid height.
Accuracies within 2 percent are reported.4 The meters have a
linear output and a meter factor of l.OO.6
Hersey Products (Spartanburg, S.C.) also produces an ultra-
sonic flowmeter. It operates on a different principle and is
designed only for closed pipes. It requires at least 2 percent
solids or an injected gas bubble flow to operate. It is there-
fore less versatile than the Badger Meter product. It is suit-
able for use on most pipes and has an accuracy within 5 per-
cent .7
4.13-19
-------
U)
I
to
o
RELATIONSHIP OF
DIMENSIONS r, d,
AND D TO FLOW COEFFICIENT
ARE PUBLISHED. SEE SPINK,
"PRINCIPLES AND PRACTICE OF
FLOW METER ENGINEERING" FOXBORO
COMPANY
TII m»«vi n i»ii tti's it it »•>•»•>•»•»•» 5»tint
HEATER/INSULATION
FLOW
DIRECTION
HEATER/INSULATION
1111 t.t.t.u t at i n.i.t i»i i j 11 j«««t n 11 mu_
««t.tj.uu
D
DIFFERENTIAL PRESSURE
TRANSMITTER
FRESH
HATER
PURGE
Figure 4.13-4. Quadrant-edged orifice plate.
Note: Relationship of dimensions 4, d, and D to flow coefficient are published.
See Spink, "Principles and Practice of Flow Meter Engineering," Foxboro
Company.
-------
Tech/Sonics (Houston, Tex.) manufactures a meter similar to
that from Hersey Products. It requires gas bubbles or suspended
solids in the stream and has an accuracy within 2 percent.
Portable or dedicated versions are available.8
Electromagnetic devices—The electromagnetic flowmeter, or
"magnetic meter," is the best proven instrument available for
the measurement of pressurized water slurries (Figure 4.13-5).
It consists of a stainless steel pipe section lined with an
electrically insulating material. Two metal electrodes protrude
through the lining, and a coil is arranged to supply a magnetic
flux perpendicular to the slurry flow direction. The slurry
itself acts as a conductor that cuts across the flux, thereby
inducing an electrical potential between the electrodes. When
amplified, the signal is adjusted to indicate the true volu-
metric flow rate of slurry. The signal is linear with flow rate
and can be recorded on a uniformly graduated chart. The mag-
netic meter does not require installation in a straight piping
run and introduces no pressure drop into the flowing stream.
The lining can be made of an abrasion-resistant material; poly-
urethane resin is recommended by most manufacturers, but Neo-
prene synthetic rubber is probably better in meters smaller than
4 in. Teflon is available, but is not as resistant to abrasion.
Electrodes can be of any metal; hardened Type 316 stainless
steel is the usual manufacturer's standard and is suitable for
most scrubber services. Magnetic meters should be recalibrated
at least annually. Magnetic meters, which are expensive instru-
ments, are made by several companies, but Brooks, Fischer and
porter, Foxboro, and Taylor market them most actively.
Figure 4.13-5. Typical magnetic flow meter.
4.13-21
-------
Less expensive devices are also made that use the electro-
magnetic principle. One, which is usually called the insert-
type magnetic meter, consists of a coil in a thin probe which
sets up a magnetic flux inducing a voltage in two electrically
insulated sections of the probe casing. This instrument mea-
sures the velocity of the slurry in a small region near the tip
of the probe. The other magnetic-type instrument consists of a
small, conventional magnetic meter totally immersed in a larger
pipeline. It measures the velocity of the portion of the flow
that passes through the small meter. The only advantage of
these two instruments is their lower cost. Their disadvantage
is the uncertainty of their calibration accuracy.
4.13.6.3 Open Channel Flow—
Open channel flow measurements are suited to streams that
can be made to flow by gravity, such as the feed slurry to a
thickener. These devices are most often used in plants with a
civil engineering design basis, such as waste treatment plants,
but they are equally suitable for use in chemical processes such
as lime scrubbing. A relatively inexpensive calibrated flume is
installed in a freely flowing, unpressurized pipeline or chan-
nel. Measurement of the level of the flowing liquid in the
throat section of the flume is directly related to fluid flow
rate. The best known device for this application is the Par-
shall flume, although others are made to fit into either rec-
tangular channels or partly filled, circular pipes. Most are
sold as preassembled fiberglass and plastic constructions. They
may also be constructed of poured concrete using forms sold for
this purpose. A variety of mechanical and pneumatic instruments
are available that fit onto the flumes to transmit a signal
related to flow rate. Ultrasonic devices are also available for
open channel flow measurement, as discussed in Section 4.13.6.2.
4.13.6.4 Flow Controlling and Recording Equipment—
Signals from flow measurement instruments are usually
brought to a centralized control panel to be recorded or used
directly in the control of a scrubber. Most flow measurement
signals change rapidly and erratically over a rather wide band;
therefore, a chart record can usually be read with greater
accuracy than can an indicating pointer or a digital instrument.
Signals from pressure-differential flow instruments are
nonlinear, since they are proportional to the square root of
flow rate. These signals may be either passed through square
root extraction instruments to linearize the signal or recorded
on chart paper with square root graduations. It is customary in
many electric power industries to use square root extractors
with all these measurements, since these signals are most often
used in a boiler plant for ratio computation or cascade control,
where linearization is necessary. In a lime scrubber, however,
where some flow signals are not associated with complex control
4.13-22
-------
loops, the chart record may be read more accurately without
square root extraction. If adjusted improperly, extractors can
introduce substantial error at the low end of the scale.
Signals from at least one of the brands of ultrasonic
equipment are linear.4'6 This is another advantage of the use
of ultrasonic systems.
4.13.6.5 Existing Facilities—
Flow metering equipment used at existing lime scrubbers is
shown in Table 4.13-6.
The most widely used liquid flow meter is the magnetic
type. Cane Run and Paddy's Run facilities have had no opera-
tional problems, whereas at Green River there was initially a
minor pluggage problem. At Bruce Mansfield, it was found that
the magnetic flowmeter generates heat when shut off and causes
lining material failures. Foxboro has alleviated this problem
by changing the lining material on the flowmeters.
4.13.7 Control Panels and Panel Instruments9
Control panels and panel instruments do not need detailed
discussion, since the utility industry already has detailed
specifications.
It is important to emphasize, however, that uniformity in
design and spares in a new system will simplify operation and
maintenance. Similarly, uniformity between a retrofit system
and the existing boiler will also prove beneficial unless the
existing panel system has caused too many problems.
Systems should be designed with consideration to opera-
bility, efficiency, ease of maintenance, and safety.
4.13-23
-------
Table 4.13-6. FLOW INSTRUMENTATION
Facilities
Conesville No. 5
Columbus and
Southern Ohio
Elrama
Duquesne Light
Phillips
Duquesne Light
Green River
Kentucky Utilities
Cane Run
Louisville Gas
and Electric
Paddy's Run
Louisville Gas
and Electric
Bruce Mansfield
Pennsylvania Power
Flow meter
Mfr.
Foxboro
Foxboro
Brooks
Brooks
Fischer
Porter
Foxboro
Brooks
Foxboro
Foxboro
Brooks
Type
Magnetic
Magnetic
Magnetic
AP
Magnetic
Magnetic
Magnetic
Magnetic
4.13-24
-------
REFERENCES
1. Jones D. G. , A. V. Slack, and K. S. Campbell. Lime/
and c°ntro1
2. Private communication with P. S. Lowell, P. S. Lowell and
Associates, September 1978.
3. Private communication with T. Moraski, EPRI , August 1978.
4. Badger Meter. Product literature, Tulsa, Oklahoma.
5. Private communication with P. S. Lowell, P. s. Lowell and
Associates, September 1978.
Spartanburg, South
8. Private communication with Tech/Sonics, Houston, Texas,
September 1978 .
9. Private communication with P. s. Lowell, P. S. Lowell and
Associates, September 1978.
4.13-25
-------
SECTION 5
BID REQUEST/EVALUATION
5.1 INTRODUCTION
This section presents information to assist a utility
engineer in the preparation of bid requests for an FGD system
and subsequently to evaluate bids received. This information
supplements a utility's normal process for bid requests and
evaluations on proposed capital expenditures.
5.1-1
-------
CONTENTS
5.2 DESIGN BASIS
Page
5.2.1 Coal and Ash Analyses 5.2-1
5.2.2 Boiler Conditions 5.2-2
5.2.3 Gas Flow Rate and Temperature 5.2-3
5.2.4 Particulate Control Strategy 5.2-4
5.2.5 S02 Loading and Emission Regulations 5.2-4
5.2.6 Lime Properties 5.2-5
5.2.7 Makeup Water Composition 5.2-5
5.2.8 Waste Disposal Requirements 5.2-5
5.2.9 Miscellaneous Information 5.2-5
5.2-i
-------
5.2 DESIGN BASIS
To submat bids that are cost-effective and responsive to a
utility's needs, prospective FGD system suppliers require speci-
fic information in many areas; thus, utility companies that
provide sufficient information are in a better position to
obtain an optimum emission control system. Bid requests should
be specific so that bids received from the various vendors are
Sont8a?n °C°** ^ "" °f '"^ison and should
of
0 The required sparing capacity (redundancy)
Guarantees to meet applicable emission regulations at
all operating conditions
0 Well-defined maintenance requirements
A successful design depends on a few key parameters that
are essential to FGD system design:
0 Coal and ash analyses
0 Boiler conditions
0 Gas flow rate and temperature
0 Particulate control technology
0 S02 loading and emissions regulations
0 Lime properties
0 Makeup water composition
Sludge disposal requirements (if sludge disposal is
within the scope of the bid request)
5.2.1 Coal and Ash Analyses
An extensive sampling and analysis program is recommended,
particularly when coal from more than one coal seam is used
The coal analysis, used in conjunction with boiler firing pracl
tices, can be used to define accurately the composition of the
flue gas to be treated. Prospective bidders should be supplied
wlth a proximate and an ultimate coal analysis form with mean
values and ranges noted. The following example is a sample coal
analysis form. ^ <-«cu.
5.2-1
-------
Proximate analysis,
as received, percent.
by weight
Moisture
Ash
Volatile matter
Total
Sulfur
Heating value,
Btu/lb
Ultimate analysis,
as received, percent
by weight
Moisture
Ash
Sulfur
Nitrogen
Carbon
Hydrogen
Oxygen
Chlorine
Total
Coal Analysis
Mean
value
Range of values
Mean
value
Range of values
It is also recommended that a fly ash alkalinity evaluation
be performed and the results provided to prospective bidders.
The method of measurement used in determining this alkalinity
should be identified. The fly ash might have sufficient alka-
linity to enhance S02 removal; if so, a lime/fly ash system
(several are in operation) that saves considerably on the cost
of reagent (when compared with a system using lime alone) can be
designed.
5.2.2 Boiler Conditions
As previously mentioned, the boiler conditions, in conjunc-
tion with the nature of the coal fired, determine the compo-
sition of the flue gas being treated. Items that should be
provided are boiler type, coal size when fired, coal firing
rate, Btu input rate, excess air in the boiler, and any air
leakage expected throughout the plant life. Boiler material
balance calculations that identify concentrations in the gas
stream should be performed on at least the following:
5.2-2
-------
0 SCh
0 so:
O f~\
o
CO 2
2
° H2O
0 Chloride
Fly ash
The reader is referred to Section 2.3 as an aid to performing
these material balance calculations.
Because FGD systems operate best when running in a steady
state mode and swing conditions greatly increase the probabil-
ity of costly problems, it is important to provide prospective
bidders with the expected variability in any of the above condi-
tions. Boiler operating conditions are no exception.
5.2.3 Gas Flow Rate and Temperature
The unit size of an FGD system is usually classified by
power generating capacity, i.e., megawatts (MW). In a bid
specification, however, the only meaningful method of deter-
mining size is the specification of flue gas flow rate and
temperature. Flow rate and temperature should be provided from
several points along the gas path as it exits the boiler, such
as the boiler outlet, the inlet and outlet of the economizer,
air heater, particulate collection device, and any other equip-
ment preceding the SO2 absorption train.
The most important location for specifying flow rate and
temperature is the inlet to the proposed FGD system, as this is
the key parameter in determining sizing of gas handling equip-
ment (ducts, quencher, absorber, fan, reheater, and stack).
This parameter has often been incorrectly specified, leading to
underdesigned FGD systems. Again it is essential to delineate
any variability in gas flow and temperature.
The primary purpose in purchasing an FGD system is to meet
the S0? emission regulation for the life of the boiler plant
One major factor that is not considered frequently, however is
future air leakage. The effects of the increased air flow'due
to air leakage are as follows: increased requirement for fan
capacity, increased pump capacity to maintain design level for
L/G, increased module cross-sectional area to maintain desiqn
level for gas velocity, increased piping capacity, increased
tank capacity to handle the increased liquor flows and levels
increased demister loading (more liquor being entrained and the
gas traversing the demister at higher velocities), and increased
reheater requirements.
5.2-3
-------
FGD system suppliers do not guarantee FGD system operations
throughout the life of the boiler plant because there are too
many variables beyond their control. Usually the vendor guaran-
tee applies only through the test run. Because the utility will
be required at all times to comply with the emission regulation,
the following options should be evaluated:
0 Request the bidders to design a system that will
remove sufficient SO2 at the increased air flows to be
expected as air leakage worsens throughout the plant
life, based on expected air flows (provided by the
utility) throughout the plant life. The bidder should
not be requested to make a guarantee, but simply to
consider the expected flow rates. This procedure
should increase the chances for adequate S02 removal
over the life of the plant.
0 Realize that the increased air flow problem will
probably occur, and plan to operate the boiler at a
reduced load when it does occur. The primary deter-
minant of boiler load would be air flow to the FGD
system. This scheme of reducing boiler load as a
function of air flow to the air pollution control
equipment would follow normal utility planning method-
ology wherein ever-decreasing amounts of power genera-
tion are required from any particular boiler.
5.2.4 Particulate Control Strategy
The particulate controls that precede the proposed absorp-
tion train in the gas flow loop, inlet and outlet particulate
loadings, and the regulation for final particulate emission and
opacity should be specified. Expected variations at the scrub-
ber inlet should also be discussed.
5.2.5 SO? Loading and Emissions Regulations
The expected SO2 loading at the FGD inlet and the required
SO2 loading, with the specified averaging time, at the stack
exit should also be included. To ensure compliance, the utility
may desire to specify an outlet loading that is lower than the
applicable regulation.
For example, if the regulation is 1.2 Ib SO2/million Btu input,
the utility may specify 1 Ib SO2/million Btu input to allow a
safety factor because of time averaging requirements. To sum-
marize, the following stack conditions should be specified:
0 S02 flow, Ib/million Btu input
0 Particulate flow, Ib/million Btu input
0 Mist loading, gr/scf
5.2-4
-------
0 Plume opacity, percent (asually controlled by deter-
mining removal efficiency required to meet regulation)
0 Stack exit temperature, °F (if desired)
5.2.6 Lime Properties
If a lime supply has been obtained, the bid specifications
should give information about the reagent such as magnesium
content, size, composition, reactivity, and slaking rate.
5.2.7 Makeup Water Composition
A complete analysis of the makeup water supply should be
provided. The water source should be named, e.g., service
water, river water, cooling tower blowdown. Sodium, magnesium,
and chloride ion concentration, pH, sulfite/sulfate content, and
solids level are especially important. In addition, the amount
of water discharge (if any) allowable under local regulations
should be specified. This amount, evaporation losses, and
interstitial water exiting with the sludge permit calculation of
makeup water requirements.
5.2.8 Waste Disposal Requirements
Regarding waste disposal requirements, the utility should
specify the proposed disposal site and the desired quality of
the final product with respect to solids content, pH, leaching
characteristics, and impact strength (minimal if sludge is being
landfilled, high if it is to be usod for building foundations).
EPRI's Sludge Manual provides greater detail concerning waste
disposal requirements.
5.2.9 Miscellaneous Information
Other items that should be included in bid specifications
are:
Annual weather and temperature conditions. Inadequate
cold weather protection has caused extensive downtime
in existing systems.
Retrofit restrictions (if applicable). Space limita-
tions, current fan placement and materials of con-
struction, and current duct and stack placement,
sizing, and materials of construction should all be
specified.
Startup date required. If a rapid job is planned, it
could significantly affect the cost of the system.
5.2-5
-------
CONTENTS
5.3 GUARANTEE REQUESTS
Page
5.3.1 General 5 3_^
5.3.2 SO2 Removal 5 3_2
5.3.3 Particulate Removal Efficiency 5 3_2
5.3.4 Mist in the Outlet 5 3_3
5.3.5 Power Consumption 5 3_3
5.3.6 Reheat Energy Consumption 5 3_4
5.3.7 Lime Consumption 5 3_4
5.3.8 Water Consumption c q A
D • -3 ~rr
5.3.9 Waste Streams 5 3_4
5.3.10 Turndown Ratio 5 3_5
5.3.11 Availability 5 3_5
5.3.12 General _-
5.3-i
-------
5.3 GUARANTEE REQUESTS
5.3.1 General
in the past utilities sometimes were given vague process
guarantees with their purchase of an FGD system. in some in-
stances, these guarantees proved to be less than binding because
they were not specific in covering the possible range of operat-
ing conditions. Currently, however, major system suppliers are
willing to supply detailed guarantees in several important
areas. ^
A utility requesting bids for an FGD system should simul-
taneously request the accompanying guarantees. These guarantees
should not only ensure satisfactory process and equipment per-
formance but also set limits on certain operating parameters
For example, clearly stipulated guarantees for SO, and partic-
ulate removal relate to satisfactory process and equipment
performance. In addition, guarantees on items such as power
consumption, reheat energy consumption, lime consumption, and
waste stream quality and quantity establish a basis for pro-
jecting accurate operating costs.
This process of requesting guarantees provides two advan-
tages for the utility. It allows an in-depth comparison of the
strength and scope of the guarantees among the various bids and
permits operating costs to be predicted accurately.
following °n lnClUdeS S discussi- «>* guarantees for the
O
o
O
o
o
o
o
o
o
o
S02 removal efficiency
Particulate removal efficiency
Mist in the outlet gas stream
Waste stream quality/quantity
Power consumption
Reheat energy consumption
Lime consumption
Water consumption
Turndown ratio
System availability
in requesting guarantees, it is essential to specify measurement
prSCr^ndinans tTail^it° av°id later' Potentially Mostly mis-
understandings. in flue gas or waste stream sampling* items
such as test port location and accessibility, sampling proce-
dure, analysis procedure, data reporting procedure, and assign-
reSp°nsibilit* of conducting sampling"shouTd
5.3-1
-------
Another important concept is to require as many guarantees
as possible to be consistent with the applicable environmental
regulations, especially for SO2 removal, particulate removal,
mist in the outlet, and waste stream quality.
5.3.2 SO? Removal
SO2 removal guarantees are requested most often. In the
past, guarantees were written specifying that the FGD system
would effect a certain percentage of S02 removal, usually for a
specific coal sulfur content. The percent removal was such
that, for the specified coal, sufficient SO2 would be removed
from the flue gas to meet the applicable regulation; however,
problems can occur because there can be significant deviations
in the coal sulfur content and flue gas flow rate. Both of
these factors affect required SO2 removal and the operation of
the FGD system. Bidders should therefore be requested to guar-
antee meeting the SO2 emission regulation (usually expressed in
allowable Ib S02/million Btu input) over the entire range of
operating conditions set forth in the design basis.
Regulations being considered by the U.S. Environmental
Protection Agency (EPA) for new utility boilers include an
averaging time over which the emissions must not exceed a cer-
tain value. If a regulation of this type becomes effective, the
utility should ascertain that the system is guaranteed to meet
it.
As mentioned earlier, it will be important to specify in
the guarantee request the sampling procedure (probably EPA
Method No. 6) that will be used and the exact location of test
ports (according to the EPA test procedure). In addition, all
conditions that require testing for compliance should be speci-
fied in the request. For instance, if the regulation requires
the unit to be tested under varying boiler load conditions
(e.g., 50%, 100%), this should be included in the guarantee
request.
5.3.3 Particulate Removal Efficiency
Even though particulate removal is effected in an ESP,
baghouse, or scrubber upstream of the FGD system, particulate
discharge can present a problem when particulate matter generat-
ed by slurry carryover is not removed in the mist eliminator.
To avoid particulate emission problems, one of three types
of guarantee is suggested. The most straightforward guarantee
applies the appropriate particulate emission regulation to the
FGD system discharge; however, system vendors are reluctant to
guarantee the absorber system outlet particulate concentration
if they are not supplying the primary upstream particulate
control system.
5.3-2,
-------
The second method, and one in use in several current
cations, guarantees that the
S
in tp
trol system does not bring the unit in?o compliance b^ftse!?:
5.3.4 Mist in the Outlet
measure, however, because the I accurarv «? djfflcult quantity to
is not known Anv a,,^^ accuracy of the measuring methods
project o^let6 m'st^Ladlng ncc"! ^^ ^ "*
usual cause of excessive mist carryover* 7t is ^^ 1S ^
ating problem. carryover, it xs more of an oper-
5.3.5 Power Consumption
.nentaSon",
operaions, is
Another form, and perhaps the most i ™-; 0=1
maximum power consumption ( e . g., the plant J-m ' guarantees
'-- '^'^S
*"* 1*1 i-I. J» >Ji L \J 1,ill S iJTrC 1 T *1 P*C! A ivk A ^*^ vm.-._«u.
tr ^^ "^ * •* ** *3 Q •••^UmUlllllll T^f^ T"/*^f^T^ ^ a 4T^^ ^N -f* 4_ T_
scrubber plant will consume n^more tha/^ SYpSetrcen
-------
5.3.6 Reheat Energy Consumption
Once again the simplest and most requested guarantee for
reheat energy consumption specifies a guaranteed maximum fuel
consumption in Btu per hour. Two refinements of this simple
approach warrant consideration. The first relates fuel consump-
tion to scrubber inlet gas flow; thus, the guarantee would be
for a maximum Btu consumption per volume of gas (Btu/scfm)
Another refinement stipulates minimum acceptable downstream
temperature at this heat flow. Whatever is guaranteed the
location and methods of energy and gas flow measurement, as' well
as the test interval, should be specified.
5.3.7
Lime Consumption
The guarantee for lime consumption is usually a maximum
usage rate (lb/h), which can be related to a process variable
(the rate of S02 removal). The guarantee should be written in
pounds of lime consumed per pound of S02 removed.
The lime usage rate is difficult to measure unless the
plant is equipped with a gravimetric-type lime feeder in advance
or the slaker. It is therefore important to specify the test
method for lime feed rate measurement. One method measures rate
in gallons per minute with a magnetic flowmeter and the density
with a nuclear density meter. Both of these pieces of equipment
are expensive and are seldom included in the bid package A
second method operates the lime slurry system as a batch process
during the test run. (Each batch must be measured and sampled
for solids content.)
5.3.8 Water Consumption
The amount of water consumed by the process may be guaran-
teed, but it is better not to specify this parameter. The
critical item in the plant water balance is the need, in manv
areas, to run closed loop. if this is the case, it is advan-
tageous to specify closed-loop operation and let the vendor use
as much water as needed. Closed-loop operation should be clear-
ly defined at all boiler loads and SO2 concentrations for which
the scrubber would be operated. "-«-^«
5.3.9 Waste Streams
The waste streams that are acceptable should be specified
in the request because the plant material balance is highlv
influenced by these streams. Thus, it should be specified
whether the plant will produce dry landfill, use a settling
pond, or operate open or closed loop. if regulations permit the
plant to discharge, then the limits of wastewater quality should
appear in the guarantee. Again, the utility should specify the
sampling and analysis method as well as at what level of boiler
operation these waste streams should be measured.
5.3-4
-------
5.3.10 Turndown Ratio
efficiently produced. The
could operate at the lowest :
expect satisfactory FGD operation This
and therefore no number is offi^
down ratio. The tuTndown ratio "sh^n rf
as well as by vessel h°Uld
x Can be
°be assured that it
be slte-sPecific,
^ maximum turn-
ldentified by system
Bidders should be requested to rlaT-i^, ~
when operations are proceeding ma»^ J S lmP°rtant P°int:
the FGD system response f to" a Led 1- turndown, what will be
manner or in a continuous maSJL Thus^ * r" * steP-function
ratio of 4 to 1 is guaranteed" £ x'-i & maximum turndown
whether the boiler load cSulS f a?l ,n x bldders should indicate
percent to 100 percent or^hL^T^^ the range from 25
operate at certain di.cr.te^ ^^2^5 5
ij v/2 — ~ * — — ^ ** *P»^ %*4p* ^\i. ^
relate to the utilities cum j
well as closed-loop guarantees
5.3.11 Availability
and wil1 also
consumption guarantees, as
ed or not) divided by the
in
is the
(Whether °Pe"t-
^^
Connecticut, November 1977.
, Stanford,
5.3-5
-------
It is therefore recommended that the utility request an
availability guarantee from the prospective bidders. When the
bid evaluation procedure is initiated, this can be an essential
area of comparison.
5.3.12 General
In evaluating the proffered guarantees it is important to
evaluate the financial liability being assumed by the various
bidders to achieve the guarantee levels. As an example, con-
sider the situation wherein two vendors have each guaranteed an
availability (as defined in the bid specification) of at least
90 percent for one year. One vendor, however, sets a limit on
his expenditures to meet the guaranteed level of availability
(e.g., the supplier shall expend no more than $1,000,000 to
achieve the guarantee level), and the other supplier is willing
to make expenditures up to an amount equal to the total system
cost.
5.3-6
-------
CONTENTS
5.4 EQUIPMENT AND INSTRUMENTATION
Page
5.4.1 Introduction
5.4-1
5.4.2 Venturi Scrubber
5.4-1
5.4.2.1 Corrosion
5.4.2.2 Erosion 5.4-1
5.4.2.3 Plugging 5-4-3
5.4-3
5.4.3 Venturi Recirculation Fumes r „ ^
~ o . 4—3
5.4.3.1 Corrosion .. . „
5.4.3.2 Erosion 5.4-3
5.4.3.3 Pump Seals !'4~3
5.4.3.4 Suction Head 5* 4"5
5.4-5
5.4.4 Venturi Recirculation Tank
c * r-
o .4-5
5.4.4.1 Corrosion c . _
5.4.4.2 Erosion l'*t
5.4.4.3 Tank Size 547
5.4.5 Presaturator
5.4-7
5.4.5.1 Corrosion/Erosion c-
5.4.5.2 Nozzle Construction c "?~ -7
*5 * 4t~-/
5.4.6 Absorber
5.4-7
5.4.6.1 Pressure Drop
5.4.6.2 Scaling and Plugging l'l,n
5.4.6.3 Erosion 5.4-10
5.4.6.4 Corrosion 5.4-11
5.4-11
5.4.7 Absorber Recirculation Pumps c
b « 4—12
5.4.7.1 Corrosion
5.4.7.2 Erosion 5.4-12
5.4.7.3 Pump Seals 5.4-12
5.4.7.4 Suction Head 5.4-12
5.4.7.5 Maintenance Simplicity 5 'Jill
5.4-i
-------
CONTENTS (continued)
Page
5.4.8 Absorber Recirculation Tank 5.4-14
5.4.8.1 Corrosion/Erosion 5.4-14
5.4.9 Mist Eliminator 5 4_14
5.4.10 Soot Blowers 5 4_18
5.4.11 Stack Gas Reheater 5.4-20
5.4.11.1 In-line Reheat 5 4_21
5.4.11.2 Direct-fired Reheat 5"4-21
5.4.11.3 Indirect Reheat 5*4-21
5.4.11.4 Bypass Reheat 5.'4-21
5.4.12 Dampers 5 4_23
5.4.13 Duct Work 5.4-24
5.4.14 Booster Fan 5.4-27
5.4.14.1 Corrosion 5.4-27
5.4.14.2 Cleaning and Inspection 5.4_27
5.4.14.3 Temperature Rise 5*4-27
5.4.15 Lime Silos 5.4_29
5.4.15.1 Materials of Construction 5.4-29
5.4.15.2 Conveying Systems 5.4-29
5.4.15.3 Closed Construction 5.4_29
5.4.16 Lime Feeder 5.4-29
5.4.16.1 Conveying Mechanism 5.4-29
5.4.16.2 System Shutdown 5."4-30
5.4.17 Lime Slaker 5.4-30
5.4.17.1 Lime Type 5 4_3Q
5.4.17.2 Slaker Type 5".4-30
5.4.17.3 System Reliability 5*.4-32
5.4.17.4 Miscellaneous 5.4-32
5.4.18 Lime Stabilization/Storage Tank 5.4-32
5.4-ii
-------
CONTENTS (continued)
Page
5.4.19 Lime Slurry Feed Pump
O * 4t ™ J £t
5.4.19.1 Abrasion
5.4.19.2 Pump Seals ^.4-35
5.4.19.3 Lime Slurry Feed Control 5.4-11
5.4.20 Freshwater Pump
K 5.4-35
5.4.21 Thickener
5.4-38
5.4.21.1 Corrosion
5.4.21.2 Solids Concentration I'Tla
5.4.21.3 Clarification R'AQQ
5.4.21.4 Recirculation S'A"^
5.4.21.5 Solids Removal s A~^
5.4.22 Flocculant Proportioning Pump 5.4-39
5.4.23 Thickener Underflow Pump 5 4_3g
5.4.24 Thickener Overflow Pump 5 4_42
5.4.24.1 Corrosion .
5.4.24.2 Pump Seals 5 4 44
5.4.25 Thickener Overflow Tank 5 4_44
5.4.25.1 Corrosion c A
5.4.25.2 Erosion K*A~
5.4.25.3 Tank Size 54!
5.4.26 Centrifuges
5.4.26.1 Materials of Construction 5 4.47
5.4.26.2 Rotational Speed c 'I JL
5.4.26.3 Conveyor Speed lijlj?
5.4.27 Vacuum Filters
5.4-49
5.4.27.1 Materials of Construction c A
5.4.27.2 Drying Time 0.4-49
5.4.27.3 Barometric Legs c't"49
5.4.27.4 Filter Medium \'\
5.4-iii
-------
CONTENTS (continued)
Page
5.4.28 Filtrate or Centrate Pump 5.4-49
5.4.28.1 Erosion 5.4-51
5.4.28.2 Corrosion 5*4_51
5.4.28.3 Pump Seals 5*4-51
5.4.29 Fixation Additive Silos 5.4-51
5.4.29.1 Materials of Construction 5.4-51
5.4.29.2 Conveying Systems 5.*4-53
5.4.30 Fixation Tank or Pug Mill 5.4-53
5.4.30.1 Corrosion 5.4-53
5.4.30.2 Erosion 5*4-53
5.4.30.3 Tank Size 5*4-53
5.4.31 Sludge Disposal Pump 5.4-55
5.4.31.1 Corrosion 5.4-55
5.4.31.2 Erosion 5*4-55
5.4.31.3 Pump Seals 5*4-55
5.4.31.4 Pumping Distance and Head 5*4-55
5.4.32 Sludge Conveying System 5.4-57
5.4.32.1 Belt Conveyor 5.4.57
5.4.32.2 Screw Conveyor 5.4-57
5.4.33 Pond Water Return Pump 5.4-59
5.4.33.1 Erosion 5.4-59
5.4.33.2 Corrosion 5*4-61
5.4.33.3 Pump Seals 5*4-61
5.4.33.4 Pump Placement 5.'4-61
5.4.34 pH Sensors and Controller 5.4-61
5.4.34.1 Probe Location 5.4-63
5.4.34.2 Probe Type 5.4-63
5.4.34.3 Nonlinear Control 5.4-64
5.4.35 Level Controllers 5.4-64
5.4.35.1 Displacement Controllers 5.4-64
5.4.35.2 Differential Controllers 5*4-64
5.4.35.3 Capacitance Controllers 5.4-66
5.4-iv
-------
CONTENTS (continued)
Page
4.3.36 Flowmeters 5 4-66
5.4.36.1 Magnetic Flowmeters 5.4-68
5.4.36.2 Nuclear Density Sensors 5*4-68
5.4.37 S02 Analyzers 5.4-68
5.4.37.1 Extractive Analyzers 5 4-70
5.4.37.2 Probe Locations 5'.4-70
5.4.37.3 Sample Line Maintenance 5*4-70
5.4.37.4 In Situ Analyzers 5 4-70
5.4.38 Pressure Sensors and Controller 5.4-71
5.4.39 Temperature Sensors and Controller 5.4-71
5.4.40 Control Valves 5 4_75
5.4-v
-------
5.4 EQUIPMENT AND INSTRUMENTATION
5.4.1 Introduction
This section presents a description of
equipment that comprise a lime FGD system a cw^r-iTM-i,* * ^
service these components will experience' factors t->^
sidered when specifying the coirmonen^ A fact,ors to be con-
sheets for them. Tables 5 4-la and % and sanple specification
and instrumentation addressed in this section ^ equipment
5.4.2 Venturi Scrubber
A venturi scrubber is a gas-atomized spray device- i e it
uses a moving gas stream to atomize 1 imiiSi«+•«.? if" ^
then to accelerate the drooletJ H?«K liquid nlntP droplets and
400 ft/s impart a high relativevelociL^ °^leS °f 2°° tO
liquid drops and promote particle collection^? iner??*!*?™^6
tion. In a venturi scrubber limii^ ? • ^. ^ tial lmPac-
trance to the throat though'sever a ]. introduced at the en-
radially inwards. tnroug^ several nozzles that are directed
^ peases with throat
increases as the squarl of ' the velocity ,t f, P,re^sure dr°P
neering practice to use a high L/G ratin ^A !u betteur fngi"
velocity to obtain a given overall ^^ h&n a hlgh gas
ranging from 5 to 20 gai/5>00 ft- "hull fl^ienCV' , L/G ratios
requestor should specif/^^muf 1^!^ La?S^il en
droplet ^ concentration sufficient to sweep the gas streak
The following factors should be considered in the desion of
the venturi scrubber: -mcieu j.n une aesign 01
1. Corrosion
2. Erosion
3. Plugging
5.4.2.1 Corrosion —
Corrosion from the acid contaminants and solids builduc
-1-" w"— —r.~, .*'""'=: »-W"<~LIIUOUS tiusning of all inle
averts particulate buildup, whereas the uniform wettina
reduces damage from localized acid concentrations i?9^^ •
resistant materials are not used it i? mi ? • If4corrosion
continuous control of the PH of the reelro«i *+^mP° i &^ that"
maintained during operation recirculating slurries be
5.4-1
-------
Table 5.4-la. EPRI LIME FGD SYSTEMS DATA BOOK EQUIPMENT LIST
en
•
•b.
I
ro
Particulate control subsystem
Venturi scrubber
Venturi recirculation pumps
Venturi recirculation tanks
Sulfur dioxide absorption subsystem
Absorber
Absorber recirculation tank
Absorber recirculation pumps
Mist eliminator
Soot blowers
Reheaters
Dampers
Duct work
Fan
Sludge disposal subsystem
Reaction tank pump
Thickener
Flocculant proportioning pump.
Thickener underflow pump
Thickener overflow pump
Thickener overflow tank
Vacuum filter or centrifuge
Filtrate or centrate return pump
Pug mill or fixation tank
Sludge disposal pump
Fixation additive silo
Fixation additive feeder
Sludge conveying system (belt and screw conveyor)
Front end loader or bulldozer
Pond water return pump
Lime preparation subsystem
Storage silos
Feeders
Slakers
Stabilization/storage tank
Lime slurry feed pumps
Fresh water pump
Table 5.4-lb. EPRI LIME FGD SYSTEMS DATA BOOK INSTRUMENTATION LIST
Instrumentation List
pH sensors and controllers
Level controls
Flowmeters
SO2 analyzers
Pressure sensors and controllers
Temperature sensors and controllers
Control valves
-------
5.4.2.2 Erosion--
The inlet section of a venturi scrubber is subjected to
high-velocity erosion. It is often made of a much higher grade
31 T ^?Vrrn,COIVel 6,25' than is the main ^rubber body
may be 316 ELC (extra low carbon).
5.4.2.3 Plugging —
The nozzles
The nozzles may plug frequently as a result of grit mate-
rial. Plugging can be avoided by installing strainers in the
recirculation line upstream from the nozzles srrainers in the
5.4.3 Venturi Recirculation Pumps
vessel or from a recxrculation tank located under the
in
1. Corrosion
2. Erosion
3. Pump seals
4. Suction head
5.4.3.1 Corrosion—
.rticSla^e1 mltter^r ^^ ^ Venturi ^ designed to remove
particulate matter or not, the venturi recycle pumps operate in
a corrosive atmosphere. m a particulate-only system without DH
control the pH 18 normally about 1. Rubber-liners or high-allov
;^i«1X£™?«en ^~5?f? f^uire,d ^r corrosion resis^ance^
5.4.3.2 Erosion—
The fly ash removed in the venturi is a highly erosive
material containing small particles of silica and «?«»t
derived from fly ash. Rubber-lined pumps and pumps cons?ruct2d
of an erosion resistant alloy (such as Ni-Hard) are suitable f«?
this erosive service; however, if the PH is not maintained above
4, Ni-Hard pumps should not be used. i*±ni-ainea asove
5.4-3
-------
Table 5.4-2. EPRI LIME FGD SYSTEM DATA BOOK
VENTURI SCRUBBER SPECIFICATIONS
CHECKED BY DATE
COMPUTED BY DATE
COMPANY
SPEC. NO.
PROJ. NO.
LOCATION
EQUIPMENT NO. FOR USE ON TOTAL NO. REO'D
SUPPLIER
1
c
3
P.O. NO. PRICE EACH $
GENERAL DESCRIPTION
4
5
b
7
bftUHUUIONS: FABRICATION
WORKING PRESSURE:
WIND LOAD
E5T. WEIGHT:
8
y
10
_
\i
UlAMtltK
LENGTH:
' MATERIALS: SHELL
CORROSION ALLOWANCE
TYPE OF FLANGED JOINTS
13
14
1 5
16
1 j
UIAMLTER:
' LENGTHS
MATERIALS: SHELL
CORROSION ALLOWANCE
TYPE OF LIQUID INLET:
18 ' DIAMETER
19
-------
5.4.3.3 Pump Seals —
Since the solids in the slurry contain highly erosive
particles, sealing with water to prevent erosion of the pump
Controlled seal water soured
5.4.3.4 Suction Head —
„ TxiL?Ump Sh°Uld be desi
-------
Table 5.4-3. FPRI LIME FGD SYSTEM DATA BOOK
VEMTURI RECIRCULATION PUMP SPECIFICATIONS
COMPANY
LOCATION
EQUIPMENT N0._
SUPPLIER
FOR USE ON
TOTAL NO. REQUIRED
P.O. NO.
PRICE EACH $
GENERAL INFORMATION
TYPE
DUTY: CONTINUOUS INTERMITTENT
SERVICE
PROCESS INFORMATION
LIQUID:
DESIGN FLOW: NORMAL MAX GPM
PUMPING TEMPERATURE °F
SP. GR. 9 PUMPING TEMPERATURE
VISCOSITY 9 PUMPING TEMPERATURE
VAPOR PRESS. 0 PUMP TEMP. (FT. LIO.)
PH VALUE
CORROSIVE MATERIAL
SOLIDS (MAX. DIA.)
HYDRAULIC INFORMATION FT. LIO.
SUCTION PRESS. ABOVE LIO. (ARS 1 (+)
STATIC SUCTION LIFT (-): HEAD (+)
SUCTION FRICTION HEAD (-)
TOTAL SUCTION HEAD (17+18+19)
STATIC DISCHARGE HEAD
DISCHARGE FRICTION HEAD
DISCHARGE PRESS. ABOVE LIO. (ABS.)
TOTAL DISCHARGE HEAD (21+22+23)
TOTAL DYNAMIC HEAD (24-20)
NPSH AVAILABLE (20-13)
NPSH REQUIRED
PUMP
MANUFACTURER
RPM
PFRFDRMANfF fllRVF
SERIAL NO.
BPH @ SERVICE CONDITIONS
a MAX Finu FOR IMPFIIFR
POTATinw a rmjvF SHAFT FND
COPIES REQUIRED OF:
PERFORMANCE CURVES . - - -
DIMENSION DRWGS.
CRATING Awn MAINTENANCE INSTRUCTIONS
MATERIALS
MATERIAL CODE - EXTERNAL CASING INTERNAL PARTS
I - CAST IRON INTERNALS CODE I B S C
B - BRONZE IMPELLER I B S C
S - STEEL INNER CASE PARTS I B S C
C - 11-13% CHROME SLEEVE (PACKED) Ch Ch Af Af
A - ALLOY SLEEVE (SEAL) C C C C
h - HARDENED WEAR RINGS B Ch Ch
f - FACED SHAFT S S S S
LANTERN RING
PACKING GLAND
SUCTION CONN: SIZE POSITION
DISCHARGE CONN: SIZE POSITION ..
CONN. RATING TYPE
PACKING TYPE
IANTFRN RINGS MATFRIAi
COOLING
BEARINGS: TYPE GREASE ntl_
IMPELLER: TYPE SIZE FUR. MAX
VENT CONN: DRAIN CONN.
!
FIUSHING CONNECTION:
DRIVER
FURNISHED WITH PUMP BY OTHERS
TYPE:
FRAME :
MANUFACTURER:
FNtt nSMRF
VOLTS PHASE CYf -
HP RPM
RFARINGS IMBRICATION
COUPLING GUARD
NOTES:
5.4-6
-------
for the tank vendor.
5.4.5 Presaturator
c wh
This increases S02 removal ef ficiencv *nH scrubb^g slurry.
tial for corrosion and scalinS ft * tne d »xnim«es the P°ten-
areas. taxing at the slurry/gas interface
-"sidered in the design of
1. Corrosion/erosion
2. Nozzle construction
A possible specification sheet is shown in Table 5.4-5.
5.4.5.1 Corrosion/Erosion —
the
possible chloride attack. iTno. D^V^^V and c°rrosfon from
acidic pH may accentuate corrosion scrubber is used,
5.4.5.2 Nozzle Construction
The potential of erosive att-ar-v i« *-u
often an area of concern, "he use of refract"7 n°2ZleS, is
such as those constructed of silicon ca?b±SS P PraY no;2les'
erosion becomes a problem. Slilcon carbide, is recommended if
5.4.6 Absorber
sss-ss
concept at the Shawnee test facility of TVA anrf hxl ^ • t ?
investigations at TVA's Colbert ^ °ne pllot
however, *v2 absorption efficiencv w*0 Tthr
-------
Table 5.4-4. EPRI LIME FGD SYSTEM DATA BOOK
VENTURI RECIRCULATION TANK SPECIFICATIONS
Sheet
of
CUSTOMER
PLANT LOCATION
SERVICE
JOB NO.
EQUIPT. NO.
FILE NO. _
P.O. NO.
Type of Tank
Size:
Diam.
Height
Capacity_
GENERAL NOTES
1) For required capacity as shown, Mfg. to advise
diameter and height of tank for the most
economical utilization of plate.
2) Nozzle orientation to be furnished later.
3) Nozzle location and design tube furnished
later with mechanical design.
4) Ladder Clips & Ladder:
Inside Outside
5) Design P.
Design T
6) Paint
7) Lining - Fiberglass or rubber Note 1
NOZZLES
Inlet
Outlet
Drawoff Elbow
P & V Vent
Level Gage
Thermowell
Roof Manhole
Shell Manhole
MARK
NO.
SIZE
RATING
Tank Materia
DESIGN DATA
Min. Plate Thick
Corrosion Allowance: Shell
SP GR
in.; Bottom
1n.L
APPURTENANCES (BY
Roof
in.
Level Gage or Gate Column:
Make
Yes
No
LL9L.
Pressure & Vacuum Vent Valve:
Make Fig
or Equal
Float
Yes
No
Pressure
oz.
or Equal
Vacuum
oz.
Gage Hatch:
Yes
No. Hake
or Equal
Thermometer Well:
Model
Yes
No
Length
Material
In
Make
or Equal
Thermometer:
Yes
No
Stem Length
or Equal Range
Make
Model
REMARKS:
5.4-8
-------
Table 5.4-5. EPRI LIME FGD SYSTEM DATA BOOK
PRESATURATOR SPECIFICATIONS
CHECKED BY DATE
COMPUTED BY DATE
COMPANY
SUPPLIER P.O. NO.
1 GENERAL DESCRIPTION
2 GAS FLOW: lb/h
~ 1 MiTFDTAI
LOCATION
PRICE EACH <;
acfm AT °F
ESTIMATED WEIGHT
4 INLET DIMENSIONS
5 OUTLET DIMEJSIONS
6 L/G RATIO
"7 COMPOSITION OF CQpIlM_LIQU6R~
8 SPRAY CONFIGURATION
9 SPRAY HASH, TYPE (STEAM, WATER, AIR)
10 GAS COMPOSITION: C02 ' 02 H20 ~SOX N0~~
(?: by wejjghtj x
Tl "PARTICULAT'E LOADING: gr/scf~
"TT~WORKI_NG PRESSURE: psig AT
J3_MANHOLE GASKET MAT'L
14 PRESSURE DROP: IN. WU. ACROSS PRESATURATOR
5.4-9
-------
In recent years, there has been a shift toward the use of
mobile-bed absorbers (a modification of packed towers) and spray
tower absorbers. (See Section 4.6 for a discussion of mobile-
bed absorbers.) The following discussion focuses on these two
configurations.
Several factors should be considered in the design of the
absorber:
1. Pressure drop
2. Scaling and plugging
3. Corrosion/erosion
5.4.6.1 Pressure Drop—
A typical pressure drop in a spray tower is 2 in. H20
whereas in a mobile-bed absorber a pressure drop of 6 to 8 in.
H20 is typical. In a comparison based only on pressure drop
the spray tower is desirable. However, many other factors enter
into the selection process.
A problem that may occur in mobile -bed absorbers is a
phenomenon known as flooding. As the liquid rate increases, the
pressure drop increases until the liquid actually begins to form
a layer above the packing. This is known as a flood point. An
important measure in avoiding flooding is to monitor inlet and
outlet pressures continuously to identify a pressure drop higher
than design with respect to gas flow rates.
5.4.6.2 Scaling and Plugging—
Spray towers have few internal components in the gas/liquid
contact zone. They offer the potential for higher availabili-
ties because of the lack of sites for deposition of solids. The
accumulated solids provide sites for precipitation of dissolved
mineral matter present in the coal and water sources. Under
these deposits the chloride concentration grows, increasing the
potential of stress-corrosion cracking of the metal surfaces.
Mobile-bed absorbers have the advantage of reducing scaling
and plugging as a result of the motion of the mobile-bed mate-
rial, usually polypropylene or rubber balls. However, solids
may be deposited on the trays supporting the bed.
The use of a presaturator to cool flue gas to its adiabatic
saturation temperature reduces the potential for scaling and
corrosion at the slurry/gas (or wet/dry) interface areas of the
absorber.
5.4.6.3 Erosion—
Absorber internals are subject to abrasion from the fly ash
and recirculating solids inherent in a lime slurry system. To
5.4-10
-------
One additional area of potential erosion in absorbers is
the spray nozzles. Refractory- type spray nozzles, such as toose
made of siHcon carbxde, are recommended to prevent erosion
5.4.6.4 Corrosion —
The areas of most serious corrosion attack in «-*= „ ,, • ,
bed absorbers are the inlet to the absorber anrt iJL *? x*e~
eli.in.tor. At the entrance to %e ^s£*Lg^gtT ™
may impinge on partially wetted surfaces Jnd create
rfaeS ^Tmist eli^at^
-
in a spray tower are subject to major corrosion attack Shen
'10 9 SlUrrY breakS d° resistan?
In absorbers havina
Walls abrasion attacks I
ing and exposes the inner resin T a^*- +- t . -Lcli'er 01 rne coat-
USe - con-
If shredding occurs in a spray tower, the spray nozzles mav
plug; in systems containing pump screens, plugging of the
screens may bring serious damage through pump cavitation
5.4.7 Absorber Recirculation Pumps
Some absorber systems are designed to recycle
directly from the bottom of an absorber vessel. However mo
lime scrubbing systems incorporate a recirculatio« Tor recycle
5.4-11
-------
Table 5.4-6. EPRI LIME FGD SYSTEM DATA BOOK
ABSORBER SPECIFICATIONS
CHECKED BY
COMPUTED BY _
COMPANY
EQUIPMENT NO.
SUPPLIER
DATE
DATE
SPEC. NO.
PROJ. NO.
LOCATION
FOR USE ON
TOTAL NO. REO'D.
P.O. NO.
PRICE EACH $
1
2
3
4
~r
8
y
IU
Ti
12
13
14
Ib
Ib
i/
IU
GENERAL DESCRIPTION
FLOW DIAGRAM
GENERAL INFORMATION
TYPE ABSORBER
DIAMETER HEIGHT
SPECIFICATIONS: FABRICATE
WORKING PRESSURE: PSIG AT
TEST PRESSURE: PSIG
WIND LOAD
EST. WEIGHT LBS.
HEIGHT: RECYCLE TANK
MAT'L. AND THICKNESS: BASE
CORROSION ALLOWANCE
TYPE OF FLANGE JOINT
DEMISTER: TYPE
MATERIAL IN TOWER
TEST
"f. DESIGN PRESSURE: PSIG AT «F.
rfATER PSIG AIR
SEISMIC LOAD
EMPTY LBS. FILLED WITH WATER. CAPACITY GAL.
SHELL SECTIONS
DEMISTER ABSORPTION SECTION
TOP INTERNED.
SEAMS: WELDED BRAZED
MATERIAL
HEADS
TYPE BOLTEC
ON WELDED ON BRAZED ON
MAT'L. AND THICKNESS: BOTTOM TOP
CORROSION ALLOWANCE
19 I TYPE
HEIGHT
MAT'L. AND THICKNESS
NOZZLES
20 I TYPE
T)ST
21 ! GAS INLET
NUMBER
"SIZF
LOCATION
22 GAS OUTLET
23
24
25
2;
28
29
301
31
32
33
34
35
3b
37
38
39
40
41
42
LIQUID OUTLET ~ •
RECYCLE
LIQUID INLET
LEVEL CONTROL
GAUGE GLASS
MANOMETER
MANHOLE
GASKET MATERIAL
NO, OF TRAYS
BUBBLE CAPS/SIEVES: NO.
RISERS: SIZE
DOWNCOMERS: NO. AND SIZE
MATERIALS AND THICKNESS:
PRESSURE DROP
DIA.
HOLES ON ROWS 'APART
TYPE OF PAINT
TRAYS
SPACING
PER TRAY
I RISER AREA
TRAYS
IN. WG.
DIAMETER
SIZE
MAT'L.
TYPE MAT'L.
BUBBLE CAPS
PER TRAY
PACKING
TYPE AND SIZE:
WEIGHT OF PACKED SECTION
PRESSURE DROP
PACKING SUPPORT
IN. WG.
MATERIAL
HEIGHT FACTOR
PER PACKED SECTION
MATERIAL
SPRAY NOZZLES
43
44
46
4!
48
49
iO
51
52
NO. AND TYPE MATERIAL
SIZE
ORIENTATION
PRESSURE DROP IN. WG. ACROSS
THE ABSORBER
PROCESS INFORMATION
GAS COMPOSITION (I): C02 02 H?0
GAS FLOW: Ib/hr,
SO?: Ib/hr, SO;?:
SO* REMOVAL EFFICIENCY: REQ'D.
LIQUID: Ib/hr, SOLIDS:
LIQUID FLOW: RECIRCULATION 6PM. BLEED
SO, NOx
acfm AT °F
Ib/hr
DESIGN
X, COMPOSITION^):
6PM
LIQUID pH: RECIRCULATION BLEED
REMARKS AND SPECIAL DETAILS:
5.4-12-
-------
tank that receives the effluent by gravity from the absorber
The slurry is then pumped from these tanks. Whichever system is
used, the design and specification of the pumps are identical
The major factors to be considered in recirculation pump desiqA
are as follows: ^
1. Corrosion
2. Erosion
3. Pump seals
4. Suction head
5. Maintenance simplicity
5.4.7.1 Corrosion—
Although the PH of the slurry in the reaction tank pump
should be greater than 7, it may on occasion dip to 3 or 4 In
those cases, erosion-resistant alloy pumps would be attacked.
High grade alloy (e.g., Hastalloy) or rubber-lined pumps are
therefore required for completely reliable corrosion resistance.
5.4.7.2 Erosion—
The abrasive nature of the slurry requires rubber-lined
pumps or erosion-resistant alloy pumps; however, the corrosion
potential eliminates erosion resistant alloy from consideration
if, as stated, completely reliable service is the goal.
5.4.7.3 Pump Seals—
The pump must be packed with seal water to prevent erosion
°f n-,^PUm^ ? fr°m erosive Particulate. Individually con-
trolled seal water supplies should be used.
5.4.7.4 Suction Head—
The pump should have an NSPH greater than 10 ft to prevent
outgassing. In addition to reducing the flow rate, the cavita-
tion caused by outgassing can destroy the rubber lining in the
pump.
Table 5.4-7a is a typical specification for a reaction tank
pump as it would be completed by the utility, architect/engi-
neer, or system supplier for the pump vendor.
5.4.7.5 Maintenance Simplicity—
Most manufacturers have designed pumps that are easy to
maintain. Pumps should be placed so that these maintenance
features can be best exploited. Table 5.4-7b is a typical
specification for an absorber recirculation pump.
5.4.8 Absorber Recirculation Tank
The absorber recirculation tank holds absorber underflow
and recirculates it to the absorber. since fly ash may carry
over from the fabric filter, precipitator, or venturi, the
5.4-13
-------
Table 5.4-7a. EPRI LIME FGD SYSTEM DATA BOOK
REACTION TANK PUMP SPECIFICATIONS
COMPANY LOCATION
EQUIPMENT NO. FOR USE ON TOTAL NO. REQUIRED
SUPPLIER P.O. NO. PRICE EACH $
GENERAL INFORMATION
TYPE
DUTY: CONTINUOUS INTERMITTENT
SERVICE
PROCESS INFORMATION
LIQUID:
DESIGN FLOW: NORMAL MAX GPM
PUMPING TEMPERATURE °F
SP. GR. 9 PUMPING TEMPERATURE
VISCOSITY » PUMPING TEMPERATURE
VAPOR PRFS1;. fl PUMP TEMP. (FT.LIO.)
PH VALUE
CORROSIVE MATERIAL
SOI IDS (MAX. DIA.)
HYDRAULIC INFORMATION FT. LIQ.
SUCTION PRESS ABOVE LIO. (ABS ) ( + )
STATIC SUCTION | IFT (.)• HFAD ( + )
SUCTION FRICTION HEAD (-)
TOTAI SUfTinN HfAp (17+1R+19)
STATIC DISCHARGE HEAD
DISCHARGE FRICTION HEAD
PISCWABGF PBF^ Apnvf \ ]t) (ARS )
TOTAL DISCHARGE HEAD (21+22+23)
rnTAi DYNAMO HFAD (24-20)
NP«,H AVAII ARIF (20-13)
NPSH REQUIRED
PUMP
MANUFACTURER
RPM
PFRFQRMANCE CURVF
SERIAL NO.
BPH 0 SERVICE CONDITIONS
0 MAX. FLOW FOR IMPELLER
ROTATION » DRIVE SHAFT END
COPIES REQUIRED OF:
PERFORMANCE CURVES
DIMENSION DRWGS.
OPERATING AND MAINTENANCE INSTRUCTIONS
MATERIALS
MATERIAL CODE - EXTERNAL CASING INTERNAL PARTS
I - CAST IRON INTERNALS CODE I B S C
B - BRONZE IMPELLER I B S C
S - STEEL INNER CASE PARTS I B S C
C - 11-m CHROME SLEEVE (PACKED) Ch Ch Af Af
A - ALLOY SLEEVE (SEAL) C C C C
h - HARDENED WEAR RINGS B Ch Ch
f - FACED SHAFT S S S S
LANTERN RING
PACKING GLAND
SUCTION CONN: SIZE POSITION
DISCHARGE CONN: SIZE POSITION
CONN. RATING TYPE
PACKING TYPE
1 ANTFRN RINGS MATFRIAI
COOLING
1 BEARINGS: TYPE GREASE niL
IMPELLER: TYPE SHE FUR. MAX
VENT CONN: DRAIN CONN.
FLUSHING CONNECTION:
DRIVER
FURNISHED WITH PUMP BY OTHERS
TYPE:
FRAME:
MANUFACTURER:
FNrinsiiRF _
VOLTS PHASE CYCI r
HP RPM
pFABlNKS H1BRICATION.
i (-nilPI ING GUARD
5.4-14
-------
Table 5.4-7b. EPRI LIME FGD SYSTEM DATA BOOK
ABSORBER RECIRCULATION PUMP SPECIFICATIONS
v-1-"" ' — LULATION
FOUTPMENT NO. .. ,._. . , FOR USE ON Tmfl Brnmi,rn ~ ~ ~
«-jjppi IER p n N^
TYPE _ — .
pllTY: CONTINUOUS ... . ... INTERMITTENT
S
~~~ PROCESS INFORMATION
==r^nT" ~
DC5ir,N FLOW: NORMAL MAX .. r,PM
,rjNr, TFMPERATURE -t
r.B 0 PUMPING TEMPERATURE
-,rn (1RIVF SHAFT END
HOT*' '" .
W rc BFOUIREDOF:
^Snn-r n,PVF<
|irf|C'nH DRUGS.
RflTTNG AND MAINTENANCE INSTRUCTIONS
OP^i=^. • — J
-SOTEST
MATERIALS
MATERIAL CODE - EXTERNAL CASING INTEPNflL PfPT'
j I - CAST IRON INTERNALS CODE I B S C
j B - BRONZE IMPELLER I B S C
S • STEEL INNER CASE PARTS I B S C
C - 11-131 CHROME SLEEVE (PACKED) Ch Ch «f Af
* ' *LLOY SLEEVE (SEAL) C C C C
h - HARDENED WEAR RINGS B Ch Cn
f ' PACED SHAFT S S S S
LANTERN RING
PACKING GLAND
SUCTION CONN: SIZE P"S"ION
DISCHARGE CONN: SIZE ""MTION
CONN. RATING TVPE
PACKING TYPE
LANTERN RINGS MATERIAL
COOLING
BEARINGS: TYPE r.orASE nj|
IMPELLER: TYPE ^I7r FUR
VENT CONN: PPM N CONN
FLUSHING CONNECTION:
DRIVER
FURNISHED WITH PUMP p. nTH[RS
TYPE: "
MANUFACTURER: ~
ENCLOSURE _.
VOLTS PHAS£ nf...
BEARINGS LUBRICATES
COUPLING GUARD
. —
5.4-15
-------
absorber tank may contain fly ash in addition to lime, calcium
sulfite, and calcium sulfate. The slurry should therefore be
considered mildly abrasive. The pH in a properly controlled
lime slurry tank should be 5.5 to 6.5; however, the pH can have
excursions as low as 3. The primary factor to consider is
corrosion/erosion.
5.4.8.1 Corrosion/Erosion--
The pH of the slurry in the recirculation tank should be at
least 5.5. In a well-controlled system, coated carbon steel is
probably adequate; pH excursions can occur, however, making
stainless steel, rubber-lined carbon steel, and fiberglass
coated steel the better alternatives.
Table 5.4-8 illustrates a typical specification to be
completed by the utility, architect/engineer, or system supplier
for the tank vendor.
5.4.9 Mist Eliminator
A mist eliminator is a device used to collect and return to
the scrubbing liquor the slurry droplets entrained with the gas
exiting the scrubber. A well-designed mist eliminator is needed
to prevent corrosion and scaling of downstream equipment. It
can also substantially reduce the reheat energy requirement,
because there would be less water to vaporize.
Several types of mist eliminators are available. For lime
scrubbing operations, however, the zigzag baffle configuration
(chevron) is almost universally used. The design of a complete
mist eliminator system is complex, because several conflicting
objectives must be considered. The desire for high collection
efficiency and for methods to reduce reentrainment must be
weighed against washability and susceptibility to plugging.
Associated factors to consider are the scrubber system design
and operating conditions, system construction, scrubbing medium,
solids content of the slurry, and sulfur content of the coal.
Bulk separators and knockout devices are used to remove
most large liquid droplets from the gas before the stream passes
through the mist eliminator. Special drainage features, such as
hooks and pockets, have been applied to lime scrubbing systems.
For lime scrubbing systems, mist eliminator design speci-
fies three- or four-pass, 90-degree bend, chevron mist elimi-
nators with vanes made of reinforced plastic and spaced 1 to 3
in. apart. Mist eliminators are usually housed atop the absor-
ber and placed 4 to 20 ft above the last absorbing stage.
Superficial gas velocities range from 7.5 to 21.5 ft/s.
5.4-16
-------
Table 5.4-8. EPRI LIME FGD SYSTEM DATA BOOK
ABSORBER RECIRCULATION TANK SPECIFICATIONS
Sheet
of
CUSTOMER
PLANT LOCATION
SERVICE
JOB NO.
EQUIPT. NO.
FILE NO. _
P.O. NO.
Type of Tank
Size:
Diam.
Height
Capacity_
GENERAL NOTES
1) For required capacity as shown, Mfg. to advise
diameter and height of tank for the most
economical utilization of plate.
2) Nozzle orientation to be furnished later.
3) Nozzle location and design tube furnished
later with mechanical design.
4) Ladder Clips 4 Ladder:
Inside Outside
5) Design P.
6) Paint
Design T
7) Lining - Fiberglass or rubber Note 1
NOZZLES
TnT¥T_
"Outlet
Drawoff Elbow
TTV Vent
Level Gage
Thermowel'
~Roof Manhole
Shell Manhole
MARK
NO.
SIZE
RATING
Tank Material
Corrosion Allowance: j
level Gage or Gate Coli
Ttaki
^Pressure * Vacuum Vent
TEki
HSage Hatch: ?eT
Thermometer Well:
Ttaki
Thermometer:
Ttaki
vk^iun unln
Min Plate TKiVL- CD r*r> ~~ —— —•"• — — —
'hen in.; Bottom In.; Roof
APPURTENANCES (BY
«">:- Yes No Tvoe
-r-r^- or Eaual TH5
v«lve: Yes No Pressure —
Lli or Equal Van
NO^ Make Fiq.
•H-T-. 5!§1 _Ho Lent
Ho3ei or Eoual ffiteTl
t. . t Tes No Stem Length
Model : or Eoual RanaP
in.
_t
urn 02
or Equal
|th In
al
In
REHARKS:
5.4-17
-------
The following recommendations should be considered in mist
eliminator design for lime scrubbing systems:
0 The continuous chevron is better than the noncon-
tinuous because of its greater strength and relatively
lower cost.
0 Blade spacings between 1.5 and 3.0 in. are desirable.
0 On vertical units, special features such as hooks and
pockets can be used to decrease reentrainment.
0 Bulk separation and knockout devices are desirable,
because they yield increased removal efficiency and
greater design flexibility.
0 Wash systems using blended water, consisting of pond
return water or thickener overflow and freshwater, are
recommended. Intermittent high-pressure, high-veloc-
ity wash systems are preferred to continuous wash
systems because of impact on water usage and closed-
loop operation.
Table 5.4-9 illustrates a few typical specifications for a
mist eliminator system.
5.4.10 Soot Blowers
The soot blowers described in this section have the func-
tion of removing deposits occurring in duct work downstream of
the absorber. These deposits consist of fine fly ash particles
and condensed acids. Choice of the soot blower cleaning medium
(compressed air or steam) is determined by economic considera-
tions, since compressed air and steam offer comparable service.
If the proposed plant is located in an arid region or if service
water is expensive, compressed air would be more economical. In
general, however, maintenance costs are higher for steam blow-
ing.
The compressed air requirements for soot blowing are cyclic
in nature. Thus, air compressors must be designed to respond
efficiently to varying airflow demands. The extent of these
variations would determine the best compressor control from the
three types most commonly used: constant pressure, pressure
differential, and automatic dual. Systems with properly sized
compressors can keep pace with soot blower air requirements and
maintain a stable header pressure. An operating condition that
can cause a momentary pressure drop in the header is the fre-
quent loading and unloading of the compressor and simultaneous
releasing of blower valves. This condition can be taken into
account in the design of the air piping system (increase the
pipe diameter) or the control circuit (include a time delay to
override the low-pressure switch in the header).
5.4-18
-------
Table 5.4-9. EPRI LIME FGD SYSTEMS DATA BOOK
MIST ELIMINATOR
1. Type: continuous chevron
2. Shape: Z-shaped
3. Number of passes: four
4. Vane spacing: 1 to 3 in.
5. Number of stages: one
6. Freeboard distance: 4 to 6 ft
7. Superficial gas velocity: 9 to 14 ft/s
8. Material of construction: FRP
9. Wash duration: intermittent high velocity
10. Wash water pressure: 30 to 50 psig
11. Features recommended; a) hooks and pockets and b) bulk
separation and knockout devices
5.4-19
-------
Mechanical design integrity is an important consideration
in compressor selection. For long-term durability of the gear
train, the manufacturer must exceed the requirements of the
American Gear Manufacturers Association (AGMA standard 921.06)
in terms of the actual gearing service factor applied to each
stage of compression. Internal air passages of the compressor
must be lined with corrosion resistant material; critical compo-
nents, such as impellers, should be made of stainless steel.
The interstage air coolers must be made of nonferrous materials
if the compressor is subjected to a corrosive atmosphere. The
compressor should be designed so that internal components are
accessible for maintenance.
Table 5.4-10 is a typical specification to be completed by
the utility, architect/engineer, or equipment supplier for the
centrifugal compressor vendor.
5.4.11 Stack Gas Reheater
Stack gas reheat may be necessary for the following rea-
sons :
1. Prevention of downstream condensation and subsequent
corrosion.
2. Reduction or elimination of plume visibility.
3. Enhancement of plume rise and pollutant dispersion.
Several methods are available for reheating the saturated flue
gases exiting from the absorber.
5.4.11.1 In-line Reheat—
This system consists of tube banks arranged perpendicularly
to the gas flow. Steam or hot water flows through the tubes,
imparting the required amount of temperature increase.
5.4.11.2 Direct-fired Reheat—
An in-line burner is installed in which a clean fuel (No. 2
oil or gas) is fired. The flue gas is thereby heated to the
required temperature. In a direct-fired reheat design, combus-
tion should occur in a separate chamber and not in the duct
work. This assures the integrity of the flame. A number of
systems had trouble keeping a flame when the burners were lo-
cated in the duct work.
5.4.11.3 Indirect Reheat—
Ambient air is heated either in a combustion chamber or
through a heat exchanger and then mixed with the flue gas. The
system thus avoids the plugging and corrosion problems that come
from particulate and acid mist in the flue gas.
5.4-20
-------
Table 5.4-10. EPRI LIME FGD SYSTEM DATA BOOK
CENTRIFUGAL COMPRESSOR SPECIFICATIONS
CMECICD IT BATE
CONFUTED IT DATE
COMPANY
EQUIPMENT NO.
SUPPLIER
B»EC.
PROJ.
LOCATION
FOR USE ON TOTAL iO. «EO'
P.O. NO. MICE EACH
•0.
NO.
D.
t
GENERAL INFORMATION
i
2
3
4
5
(
7
TYPE
DUTY: CONTINUOUS [3 INTERMITTENT Q
SERVICE
PROCESS INFORMATION
t
9
10
11
12
13
11
15
It
17
It
19
20
21
2?
23
24
25
2t
*7
21
29
JO
11
12
CAS, NAME t COMPOSITION.
CONDITION OF 6AS: MET. DRY, SOLIDS, CORHOS.
KOI. HEIGHT $p. CR. ( 70-F.
5P. (R. » SUCTION FLOW TEMP.
SP. HT., CP. CP/CV
CRITICAL TEMP. -F. CRITICAL PRESS. PSIA
SUCT. TEMP. «F. SUCTION PRESS. fSIA
VOL. Ft 0« t SUCT. NORMAL, CFM
VOL. FLOU » SUCT. MAI. REQ'D.. CFM
VOL. FLOW ( SUCT. DESIGN. CFM
HEIGHT FLOK. DESIGN. LBS./MIN.
OISCH. PRESS, PSIA. NORMAL. MAX.
DISCH. TEMP. «F.
HT. CONTENT OF CAS , ITU/CU. FT.
COOLING MATER: PRESS. PSIG TEMP. »F.
THEORETICAL HP NAS. IHP
VOL. IFF. BECK. EFF .
ATM. PRESS, PSIA
PERFORMANCE CURVES VES Q HO D
CONSTRUCTION
33
)l
35
36
37
38
39
40
41
42
43
44
4S
46
47
IMPELLER: SIZE TTPE
NO. STAGES «PM
SUCT. CONN. TYPE
OISCM. CONN. TYPE
CONN. BATING
SHAFT SEAL
ROTATION
8EAR1NGS
LUBRICATION COOLING
MATERIALS: CASE
IMPELLER
SHAFT
SEALS
MOUNTING BASE
ACCESSORIES
48
49
SO
SI
S2
S3
S4
SS
56
INTERCOOLER
AFTERCOOLER
RECEIVER
REGULATION: CONST. PRES. D CONST. VOL . D
_ STABLE __
ADJ. INLET VALVES D LIMIT CONTROL D
INLET FILTER
DRIVER
57
58
59
(0
61
(2
63
FURNISHED MITH COMPRESSOR Q BY OTHERS Q
TYPE t HAKE
VOLTS PHASE CYClt HP MM
COUPLING TYPE
REMARKS -
9/78
FGD Systems Data Book
5.4-21 Equipment and Instrumentation
-------
5.4.11.4 Bypass Reheat—
A portion of flue gas, having been treated for fly ash
removal, is bypassed and mixed with the gas at the scrubber
exit. This method may not apply to systems where regulations
require 90+ percent SO2 removal efficiency. (A minimum of 10
percent of the flue gas is needed to provide adequate reheat.
Therefore, the FGD system would have to remove virtually all of
the inlet S02 to achieve overall removal of 90 percent.) in
addition, Federal New Source Performance Standards (NSPS) cur-
rently being considered may eliminate this option.
The following are general recommendations for reheat sys-
tems, regardless of the chosen strategy:
0 Soot blowers should be installed on in-line reheaters.
0 An efficient mist eliminator should be installed to
decrease the load on the reheat system.
0 Gas should be heated by 25° to 50°F to prevent down-
stream water condensation.
The following are general specifications for the stack gas
reheating system.
1. Design temperature increment: 50°F.
2. Preferred type: in-line reheater.
3. Heating medium: steam.
4. Material of construction: 316L SS.
5. Recommended feature: installation of soot blowers.
The trend in reheat strategies, as evidenced by FGD systems
scheduled for immediate and future operation, is away from
in-line and direct combustion methods and toward indirect hot
air reheat. The rationale for this trend is the problems en-
countered in the former and the need for oil or natural gas in
the latter.
In-line reheat systems have been subject to corrosion and
plugging in the tubes. The corrosion in many cases has been so
severe that even the heartier alloy listed above has been un-
satisfactory under some operating conditions. A number of the
major system suppliers still recommend in-line reheaters, espe-
cially when parasitic energy demand must be minimized. The
corrosion of high-alloy materials is attributed to stress corro-
sion caused by chloride, whereas carbon steel is more suscep-
tible to acid corrosion caused by high sulfur dioxide concen-
trations. If low sulfur/low chloride, low sulfur/high chloride,
or high sulfur/low chloride environments can be identified,
in-line reheaters may be successfully used. Many of these
problems have been attributed to inefficiency of the upstream
mist eliminator and inadequate self-cleaning techniques (soot
blowers).
5.4-22
-------
5.4.12 Dampers
isolation dampers are used in a coal-fired power plant to
prevent the flow of the gas into the FGD system. The purpose^
of isolation are to continue boiler operations while the alsor-
ber modules are under maintenance and to take a scrubber module
tween them. This pressurized seal air system increases the
parasitic energy demand by the FGD system^ but sTgnTfTcantly
improves damper operation. s>j.yuj.j.xi.
-------
Table 5.4-11. EPRI LIME FGD SYSTEM DATA BOOK
DAMPER SPECIFICATIONS
CHECKED BY _
COMPUTED BY
COMPANY
DATE
DATE
SPEC. NO.
PROJ. NO.
LOCATION
EQUIPMENT NO.
SUPPLIER
FOR USE ON
P.O. NO.
TOTAL NO. REQ'D
PRICE EACH $
1 GENERAL DESCRIPTION
3 GAS FLOW:
DESIGN CONDITIONS
4 GAS COMPOSITION^,'
ACFM, MAX
AT
°F. MAX
5 MAX. PRESSURE:
IN. WG., MAX. LOSS
MAX
6 MAX. LEAKAGE: TO SYSTEM
ACFM. TO AMOUNT
IN. WG. ACROSS FULL OPEN BLADE
ACFM
TEST INFORMATION
1 PRESSURE TEST:
PSIG WATER
PSIG AIR
8 WIND LOAD
SEISMIC LOAD
9 LEAKAGE TEST
10 DEFLECTION TEST
MAX. DEFLECTION
OF BLADE SPAN
WELDING
11 SECTION
OF BOILER CODE
12 DYE PENETRANT TEST
BLADES
THICKNESS
13 MATERIAL
ACTIVATOR
TYPE
14 TYPE
DRIVE
BEARINGS
MATERIAL
15 TYPE
SIZE
DAMPER
16 TYPE
LEAKAGE
ACFM
PRESSURE LOSS
IN. WG.
SPECIAL INSTRUCTIONS:
5.4-24
-------
intervals. The ducting downstream of the absorber and to the
reheater should be lined with or constructed of corrosion resis-
tant material to combat the wet, acidic environment.
Expansion joints are an essential part of the duct work
because of their ability to absorb thermal movements, vibra-
tions, and limiting forces on equipment. Expansion joints
designed to compensate for axial movements are suitable in
utility applications. The joints must be resistant to erosion
and corrosion. They normally operate with a residual stress
pattern that amplifies stress corrosion problems. Because of
this situation, condensation in expansion joints during shut-
downs may attack the metal. Expansion joints cannot be designed
with a corrosion allowance because even a small allowance will
materially limit the movement capability of a joint. Type 321
stainless steel is resistant to stress corrosion over a wide
range of temperatures. Any application of this type of expan-
sion joint above 800°F may subject it to carbide precipitation
and subsequent intergranular attacks. Therefore, care should be
taken to prevent major temperature excursions in the duct work
after a reheater. Expansion joints in the utility industry
should be equipped with a replaceable liner to reduce erosion by
particulate matter.
Table 5.4-12 is a preliminary data sheet for duct work and
expansion joints.
5.4.14 Booster Fan
To overcome the pressure drop in a scrubbing system, fans
are used to push or pull the gas through the system. This FGD
system pressure drop may be overcome by the main boiler fan or
by a control system booster fan.
Although wet booster fans have a size advantage as a result
of the wet gas being cooler and having less volume, the trend
has been to dry (high volume), forced draft (FD) fans. A dry
fan is defined as one that does not see a saturated gas and is
not sprayed with wash water. Normally a dry fan is located
upstream of the absorber or downstream of the reheater. Because
of abrasion effects of particulate matter in a gas stream, a dry
fan should only be placed before a scrubber absorber if there is
a particulate removal device upstream from it. Factors that
should be considered in the specification of a dry fan are as
follows:
1. Corrosion
2. Cleaning and inspection
3. Temperature rise
5.4-25
-------
Table 5.4-12. EPRI LIME FGD SYSTEM DATA BOOK
DUCT WORK SPECIFICATIONS
CHECKED BY DATE
COMPUTED BY DATE
COMPANY
EQUIPMENT NO.
SUPPLIER
SPEC. NO.
PROJ. NO.
LOCATION
FOR USE ON TOTAL NO. REQ'D
P.O. NO. PRICE EACH $
1 GENERAL DESCRIPTION
'i
3
4 GAS FLOW: ACFM
5 STATIC PRESSURE
6 GAS COMPOSITION
OPERATING CONDITIONS
AT °F, MAX
IN. WG.
UPSTREAM DUCTWORK
7 GAS VELOCITY
8 MATERIAL ' "
9 WELD: TYPE
10 SUPPORTS: TYPE
11 GAS VELOCITY
12 MATERIAL
13 WELD: TYPE
14 SUPPORTS: TYPE
15 LINING: MATERIAL
16 TYPE
17 MATERIAL
18 SIZE
19 MOVEMENTS: in. AXIAL
FPS AT °F, LENGTH FT, WIDTH FT
THICKNESS
TEST
SPACING NOS.
DOWNSTREAM DUCTWORK
FPS AT °F, LENGTH FT, WIDTH FT
, THICKNESS in.
, TEST
SPACING NB
THICKNESS in.
EXPANSION JOINTS
NOS. LOCATIONS
LINER
in. LATERAL
SPECIAL INSTRUCTIONS:
5.4-26
-------
5.4.14.1 Corrosion--
Although fly ash and corrosive gases are handled by the dry
fans, carbon steel is an adequate material of construction for
forced draft fans. with proper mist elimination and reheat,
scrubber solids that can cause deposition, corrosion, and im-
balance should not appear on induced draft fans.
5.4.14.2 Cleaning and Inspection—
All fans should have adequate cleanout doors. This is
especially important on induced draft dry fans. Inspection
ports are also useful to determine if deposits are accumulating.
5.4.14.3 Temperature Rise—
The flue gas temperature rises slightly (10°F) as it ab-
sorbs the compressive energy of the fan. Reheat temperatures
and duct work velocities should be designed for this.
In view of the poor performance record of wet fans, their
specification should be accompanied with a rationale for circum-
venting known problems of erosion, corrosion, and solids deposi-
tion.
Table 5,4-13 is a typical specification to be completed by
the utility, architect/engineer, or equipment supplier for the
fan vendor.
5.4.15 Lime Silos
The following factors are important in specifying a lime
storage silo:
1. Materials of construction
2. Conveying systems
3. Closed construction
5.4.15.1 Materials of Construction--
Cylindrical, unlined, carbon steel tanks with cone bottoms,
or concrete silos can be used for lime storage. Concrete silos
must be cured for several months before use.
5.4.15.2 Conveying Systems—
Either pneumatic or mechanical conveying systems are used
to move lime. Pneumatic systems are more common and offer
easier maintenance, but they allow water vapor into the system
if not operated closed loop. Larger fabric filters are required
for pneumatically conveyed silos than for mechanically conveyed
silos.
5.4.15.3 Closed Construction—
Any lime silo or transfer hopper must be closed to prevent
the entry of water and to minimize absorption of water vapor and
dioxide. r
5.4-27
-------
Table 5.4-13. EPRI LIME FGD SYSTEM DATA BOOK
BOOSTER FAN SPECIFICATIONS
IT
e.
COMPUTED
COMPANY _
»ATE
LKATIM
EQUIPMENT 10..
ret we M
».0. HO.
TOTAL *0. tEO'D.
MICE CACH t
GENERAL INFORMATION
,
r
i
4
t
f
7
TYPE:
MFR.
MODEL:
DUTY: CMTKUOUS Q UTUMITTENT Q
SERVICE:
PROCESS INFORMATION
1
,
in
11
17
11
H
It
17
11
All D FUMES D SAS D VAPOR Q
COMPOSITION:
EXPLOSIVE L~] MON-EXPLOSIVE
CORROSIVE D NOH-CORROSIVE
OPERATING TEMP.
U
D
•F
DISCHARGE CFM
STATIC PRESSURE INCHES N{0
OUTLET VELOCITY MM
ORIENTATION
CONSTRUCTION
It
20
21
I?
21
M
M
7(
,7
ri
M
10
11
12
11
14
IS
If
17
NON-SPARKING LJ
KHEU DIAMETER: INCHES
MHEEL MATERIAL:
COATINt OR LIRINC:
SINGLE INLET D MUILE INLET Q
VARIABLE DAMPER D ""-tT D CMTLET Q
SCREENED D >"1-IT D »«HET Q
NOUS1NC MAT'L.
COATING OR LINING:
HOUSING CLIANOUT [3 KW» Q
SHAFT MAT'L.
SHAFT CIAMETER
•EARING MAKE 1 TYPE:
EITERIOR FINISH:
DRIVE
it
n
40
41
42
41
44
4!
46
47
41
41
FURN. WITH ILOHER D OTNEtQ
IELT D »« D "•" D
DIRECT D OTMERD
ILOMER RPM:
MOTOR MAKE :
MOTOR TYPE:
MOTOR FRAME:
MOTOR HP i»M
VOLTS CYCLES PNASE
IASE:
NOTES -
5.4-28
-------
5.4.16 Lime Feeder
C°ntr°ls the amount °f lime going into the
o
adeauKe If lx"tion tank - used, then volumetric feeders are
accurately rL^r^/^7 °f lime USed in the system must be
SilS be needed in aCCOUntin^ Purposes, gravimetric meters
little additfon^i In,many cases' the gravimetric feeder has
little additional cost. m addition to feeder types the fol-
lowing factors are important: j-«euer types, tne 101
1. Conveying mechanism
2 . System shutdown
5.4.16.1 Conveying Mechanism —
vibra?ina hZ^0^' SCreW convey°rs' oscillating hoppers, and
listems 9 Th?P™St- are Convevin9 methods used in lime feeding
systems. The most common systems are the belt conveyor and the
, operate continuously, since intermittent
T?1Ugging When risin9 slak^r moisture reacts
eeder 6 are n° maj°r ^advantages of the belt
5.4.16.2 System Shutdown —
taken with all types of feeders to minimize this effect through
1011 *
?eeders ^onT1011 ?* ^ sl^*™- ^ severe cases even belt
feeders could be plugged. Table 5.4-14 illustrates a tvoical
conveyor specification as it would be prepared I S a utili?^
architect/engineer, or system supplier for ?he conveyor vendor.
5.4.17 Lime Slaker
^ ^ Slf^er ^s needed to Produce a consistent absorbent
to remove the sulfur dioxide in the absorber, since the proper-
feth°e avatem IT ^^ **" ^^^ efficiency and economics
of the system, the design of the slaker is important to the
5UC^"fS ?SeHatl°n -°f a Drubbing system. The following
factors should be considered in the design of. a slaker:
1 . Lime type
2. Slaker type
3. System reliability
5.4-29
-------
Table 5.4-14. EPRI LIME FGD SYSTEM DATA BOOK
BELT CONVEYOR SPECIFICATIONS
CMCttl «V WTl
COMPUTED IT tATt
COMPANY LOCATION
IWIPMENT no. rot kst on
luMiitu P.O. NO.
ItlltT 1 Of .
SPEC. «0
PROJ, NO.
TOTAL 10. tEO'O.
PRICE (ACH 1
DUTY
\
T
J
4
S
(
7
1
CONVEYOR HTM. LIFT
TPK: Art. SURGE OESI6N
NOW LOADED:
ANGLE TO DIRECTION OF IELT TRAVEL
HOURS/DAY
HO. Of LOADIH6 »TS.
INSTALLATION: OUTOOO«I~~| 1«DOO«[~~|
O^EIATION: CONTINUOUSQ INTE«K'T.Q
FEED DESCRIPTION
9
10
u
12
13
14
IS
16
17
ia
19
40
41
4?
43
44
45
46
47
41
41
50
51
sz
i)
14
II
NAME
OENSITt 11. /CU. fT.
M01STUDE
TEHP. *f. IUXP5 *f. fI«Ei *f.
SIZE:
MAI. LUMP 1 Of FEED
CHAtACTERISTICS:
STICK*
COIIOS1VE
AHASIVE
ANGLE Of •fOSE
ACCESSORIES
?o
21
22
2i
24
?S
n
27
26
it
10
)l
32
33
3'
35
36
37
38
39
SlIITIOARDS: LCTH.
UALtUlV: WIDTH TYPE
OECKING MANORAILS
COVER (TYPE)
GUARD
HOLDIACt: SAND, ItTCNET
NOLDIACt LOCATION
STRINGERS: MIRE NOPE, CHANNEL
TRUSS. OECKPLATE
CELT HIPER: (RUSH, MIRE, SPIAY
•LADE (SINGLE. OUPLEI. OUAD.)
TACE-UP: TRAVEL
SCNEV. ClAVITf. SPRING
VERTICAL. HORIZONTAL
DRIVE: V-IELT, CHAIN, SHAfT -MOUNT ED
MOTOR TTPE:
NP NPM V. PH. CY.
DESCRIPTION
IELT: NIDTH L6TH. {PEED IN f.f.K. m j COVEItS
MAT'L. CKADE FINISH 02.
»«EAItE« STRIP: Us[~l NO (~) TT»E Of SPLICE
IDLER: tm
DIA.
SPACING
IEARINC
LUIE.
PULLEY: TYPE
HAT'L.
OIA.
L6TH.
CIOMN
LAtClNC
HUIS
MtC.
TIOUCMERS
MEAD
tETURNS
TAIL
IMPACTS
SNUB
TROUGH
TRAINERS
IENC
RETURN
TRAINERS
TAIC.UP
5.4-30
-------
5.4.17.1 Lime Type--
which ssYak^^M10*11^ Wlth reSPect to the ease with
which it is slaked. Magnesium li require
bcarefuny vaatid6 pr°perties- «» *»• of lime
5.4.17.2 Slaker Type-
Two types of slakers are available for lime slurrv
"
size, more economical,
5.4.17.3 System Reliability—
Paste slakers, which are limited to 4 tons per hour of
slake?sPUare needed **""**** to ^tem reliability^ Since more
siverelv affer? ?h 1?-lar«e s^ems, slaker malfunctions do not
improved. ^ SUPPly' °Vera11 system reliability is
5.4.17.4 Miscellaneous —
The slaking system should be located as close to the scrub-
ber system as possible. A maximum 200-ft pumping distance ^ is
recommended for lime slurry feed. In order to maintain water
but'raJher bToY 'T8 •^°Uld nOt be Washed out when shut down!
but rather blown out with air.
Table 5.4-15 illustrates a typical specification as it
iebr ^S4!^ ^.^iiSi,"
5.4.18 Lime Stabilization/Storage Tank
si
^sif
v HI ' ^k 1S desi9ned to hold a 6- to 12-hour supply of
slaked lime The partition between the two chambers is desiLed
to prevent short-circuiting of the stabilizing lime desi
-------
Table 5.4-15. EPRI LIME FGD SYSTEMS DATA BOOK
SLAKER SPECIFICATIONS
CHECKED BY DATE SPEC. NO.
COMPUTED BY DATE PROJ. NO.
COMPANY
EQUIPMENT NO.
SUPPLIER
1
i
J
4
b
b
/
8
y
10
i)
LOCATION
FOR USE ON TOTAL NO. REQ'D.
P.O. NO. PRICE EACH $
GENERAL DESCRIPTION
LIME COMPOS ITION(i): CaO
LIME REACTIVITY: RESIDENCE
HEAT EVOLVED:
FRESH WATER: pH;
RECLAIMED WATER: pH;
LIME FEEDER: TYPE
SLURRY REQ'D.:
PROCESS INFORMATION
, MgO , CaC03
TIME REQ'D. MIN
BTU/LB LIME
GPM; °F
GPM; °F
, FEEDRATE LB/HR AT °F '
LB/HR, pH, °F, % SOLIDS
SLAKER
SIZE:
12 MATL.
13 1 LINING MATL.
4
5
16
17
18
19
i!U
THK.
THK.
BAFFLES
TYPE:
MATL.
NO.
THK.
AGITATOR
TYPE:
MATL:
SIZE:
THK.
MOTOR
PHASE,
RPM VOLTS
DRIVE
TYPE:
MATL.
FABRICATION
WELD: TYPE
, TEST
SPECIAL INSTRUCTIONS:
5.4-32
-------
The tank should be covered to reduce the absorption of C02
from the air, and it should be vented to allow water vapor to
leave. The height of the tank should be designed to provide
sufficient NPSH for the lime slurry feed pumps.
Table 5.4-16 is a typical specification as it would be
completed by the utility, architect/engineer, or system supplier
for the tank vendor.
5.4.19 Lime Slurry Feed Pump
The lime slurry pumps supply slaked and diluted lime slurry
to the absorber or absorption recirculation pumps. Factors that
should be considered in the lime slurry pumps are as follows:
1. Abrasion
2 . Pump seals
3 . Lime slurry feed control
5.4.19.1 Abrasion —
Lime slurry would not be erosive if it did not contain grit
and unreacted limestone cores; however, since these materials
are present, erosion resistance must be provided.
Cast iron, erosion resistant alloy, and rubber-lined pumps
are common in lime slurry supply systems. Some designers prefer
to use rubber-lined pumps for single manufacturer consistency
through the plant. Rubber-lined pumps are not required for
corrosion prevention because low pH in the lime slurry feed
should not occur. J
5.4.19.2 Pump Seals--
The abrasive nature of the slurry requires pumps with seal
water. Each seal water supply should be independently regulated
to ensure uniform flow through the packing gland.
5.4.19.3 Lime Slurry Feed Control —
The lime slurry supply rate can be regulated by: varying
the speed on the pump, letting the slurry slip in the pump II
the control flow changes, or using a fresh slurry pump system
that recycles the lime slurry in a loop and returns the excess
to the lime slurry tank. The last technique is the preferred
method since it prevents excessive wear on the lime slu?ry pumps
and brings an adequate supply to the absorption train. Ca?ef Si
selection of the control valve is critical to minimize erosion/
JTJros i on •
the lime slurry pump vendor.
5.4-33
-------
Table 5.4-16. EPRI LIME FGD SYSTEM DATA BOOK
LIME STABILIZATION/STORAGE TANK SPECIFICATIONS
Sheet of
CUSTO
PLANT
SERVI
Type
Size:
MER JOB NO.
LOCATION EQUIPT. NO.
CE FILE NO.
P.O. NO.
of Tank
Diam. Height Capacity
R
E
V
GENERAL NOTES
1) For required capacity as shown, Mfg. to advise
diameter and height of tank for the most
economical utilization of plate.
2) Nozzle orientation to be furnished later.
3) Nozzle location and design tube furnished
later with mechanical design.
4) Ladder Clips & Ladder:
Inside Outside
5) Design P.
6) Paint
Design T_
7) Lining - Fiberglass or rubber Note 1
NOZZLES
Outlet
Drawoff Elbow
P & V Vent
Level Gage
Thermowell
Roof Manhole
Shell Manhole
MARK
NO.
SIZE
RATING
DESIGN DATA
Min. Plate Thick
Tank Material
Corrosion Allowance: Shell'
in.; Bottom
In.; Roof
SP GR
APPURTENANCES (BY
J.cvel Gage or Gate ColumnT
Yes
Make
No
or Equal
Jy.PJL_
Float
Pressure 8 Vacuum Vent Valve:
Yes
No
Pressure
oz.
Make
or Equal
Vacuum
oz.
or Equal
Gage Hatch
Yes
No. Make
Thermometer Hell:
Yes
No
or Equal
Length
Material
In
Make
Model
Thermometer:
Yes
No
Stem Length
Make
Model
or Equal
Range
REMARKS:
5.4-34
-------
Table 5.4-17. EPRI LIME FGD SYSTEM DATA BOOK
LIME SLURRY PUMP SPECIFICATIONS
LOCATION
EOUJPHFNT NO. FOR USE ON TOTAL NO. REQUIRED
r,irPLlirR p-°- N0- PRICE EACH $
' GENERAL INFORMATION
=^===
i VPC
TV. CONTINUOUS INTERMITTENT
DUTY •
btKVJC
PROCESS INFORMATION
. t ni I T D :
L I QU » u —
GN F|_0|J- NORMAL MAX GPM
plNG TEMPERATURE °F
GR (a PUMPING TEMPERATURE
5P' CQ'ITY « PUMPING TEMPERATURE
VI>l"a pRESS B PUMP TEMP. (FT. LID.)
-nriur MATERIAL
COKKO-IVt
,„, ins (MAX. DIA.)
Z-— ' HYDRAULIC INFORMATION FT. LIO.
JC SUCTION 1 I FT (-): HEAD ( + )
rl0N FPIfTION HEAD (-)
AUCTION HEAD (17+18+19)
TOTAL ->u
- nlSfHARGE HEAD
rT^TIt "*-"•
AKCE PRITTION HEAD
riARCE PRE^^ ABOVE LIQ. (ARS.)
0 nIcrHiRC,F HEAD (?l+22+23)
T0'" Yfl.y,r HFAH (24-20)
JQ1MU "
A\/A 1 LABLf ( ?0* 1 3 )
NPSH prouIRED
Ji£i! PUMP
MANU*-'aC
RPH '^cTci'""
S£R>MI" RVICE CONDITIONS
BPH ^ I4AX FtO" FOR IMPELLER
0N 0 DRIVE ^"AFT END
C CIIRUFS
cjON OR1''*^-
0lHEN*lNG AND MAINTENANCE INSTRUCTIONS
MATERIALS
MATERIAL CODE - EXTERNAL CASING INTFRNA1 DAPTS
I - CAST IRON INTERNALS CODE I B S C
B - BRONZE IMPELLER I B S C
S - STEEL INNER CASE PARTS I B S C
C - 11-13J CHROME SLEEVE (PACKED) Ch Ch Af Af .
A - ALLOY SLEEVE (SEAL) C C C C
h - HARDENED WEAR RINGS B Ch Ch
f - FACED SHAFT S S S S
LANTERN RING
PACKING GLAND
SUCTION CONN: SIZE POSITION
DISCHARGE CONN: SIZE POSITION
CONN. RATING TYPF
PACKING TYPE
1 LANTERN RINGS MATFDiAi
COOLING
' BEARINGS: TYPE GREASF OIL
IMPELLER: TYPE SIZE FtIR MAX.
VENT CONN: DRAIN r.ONN.
1 '
FLUSHING rnNNFrTTDN:
DRIVER
FURNISHED WITH PUMP BY OTHERS
TYPE:
|
FRAME:
MANUFACTURER :
FNd nsuRF
i
VOLTS PHASE CYCL-:
HP RPM
PFARINGS IMBRICATION
COUPLING GUARD _
\
•sons:
5.4-35
-------
5.4.20 Freshwater Pump
In a lime FGD system, the most likely points at which
freshwater would enter are the slakers, pump seals, and mist
eliminators. The pump, if one is needed for this service, can
be a standard centrifugal pump.
The important information that must be specified includes
physico-chemical properties of the service water, available
NPSH, materials of construction, type of drive, and motor.
Table 5.4-18 is an engineering data sheet for a typical
freshwater pump vendor.
5.4.21 Thickener
To reduce the volume of wastes, most systems use a thicken-
er to concentrate solids. Since the thickener has an important
role in the water balance, the sludge characteristics, and in
some cases the chemical reactions, care should be taken in its
design. The following factors must be considered:
1. Corrosion
2. Solids concentration
3. Clarification
4. Recirculation
5. Solids removal
5.4.21.1 Corrosion--
Thickeners in lime scrubber systems are subject to wide
variances in pH. Underflow slurry has been known to be as low
as 4 when operating in very extreme upset conditions, and over-
flow pH can be as high as 11. Under normal operations, the pH
of both the overflow and underflow will fluctuate around the
operating pH of the scrubber. The pH gradient, from highly
alkaline to medium-high acidity, brings a need for corrosion
protection throughout the thickener. To prevent acid attack,
the walls must be coated with epoxy or rubber. The floor of the
thickener should be lined or made of concrete. The rake and the
lifting mechanism should be rubber-coated stainless steel or
rubber-coated alloy. Gear boxes should be sealed, and all
walkways and support steel should be galvanized and properly
coated. For additional detailed information, refer to the EPRI
report entitled, "Sludge Dewatering Methods for Flue Gas
Cleaning Products."
5.4.21.2 Solids Concentration—
Solids concentration in thickeners varies significantly
from site to site. The calcium sulfite-to-sulfate ratio has an
important impact on settling rates. The use of magnesium to
enhance the efficiency of S02 removal causes lower settling
5.4-36
-------
Table 5.4-18. EPRI LIME FGD SYSTEM DATA BOOK
FRESHWATER PUMP SPECIFICATIONS
COMPANY. LOCATION
r0,,TPMENT NO. FOR USE ON TnTAi -0. REQUIRED
SUPPLIER_ _ _ ____ P.O. NO. PB,rE E4rH $
• GENERAL INFORMATION
TYPF . • —
v CONTINUOUS TNTFRMITTENT
PROCESS INFORMATION
, tnilTP" —
rf, FL0|J- NORMAL MAX r.PM
QtiJG" r
DINC TEMPERATURE °F
PUMP I" •
GR p PUMPING TEMPERATURE
SP' g,JTY A PUMPING TEMPERATURE
V1> D PRESS « PUMP TEMP. (FT.LIQ.)
vAPOR PKt->-) ' '
... ufli Uf — • •
-05I.F MATERIAL
^LIP«; (HAX. DIA.) .
i- ' HYDRAULIC INFORMATION FT. I IQ
====^==FRFST. ABOVE L1Q. (ABS.) ( + )
5UAiic v"""™ LIFT (-): HEAD (+)
QN FBirT10N HEAD (-)
5UU"W'UCTIPN HEAD '17+18+19)
T°Inc n^H«RGE H"°
,.L(lAf7r,F FRICTION HEAD
l)ISl-H
uAROE PD^V ABOVE LI°- (AES-> --
-jr^H^Bfif HE.AD (2i+tt*Zj)
TOTAL DyrjAMI'- Hran '21-201
£VAlLflRI F '20-13)
NP*" RfOUlRED. . -. _.
fTi— ' PUMP
-^===lfTUR^="
SERVICE CONDITIONS
8PH LlAX Fl nu FflR IMPELL"
0 DRIVE SH*FT END ,
n£OUTRED Of:
_gplE$ •*twu —
PE JOH pBUGS. . .
OIMt ir,r. AND MAINTENANCE INSTRUCTIONS
MATERIALS
I - CAST IRON INTERNALS CODE I B S C
B - BRONZE IMPELLER I B S C
5 - STEEL INNER CASE PARTS I 8 S C
C - 11-131 CHROME SLEEVE (PACKED) Ch Ch Af Af
A - ALLOY SLEEVE (SEAL) C C C C
h - HARDENED WEAR RINGS B Ch Ch
f - FACED SHAFT S S S S
LANTERN RING
PACKING GLAND
SUCTION CONN: SIZE POSITION
DISCHARGE CONN: SIZE POSITION
CONN. RATING yypf
PACKING TYPE
i LANTERN RINGS MATpRIAl
! COOLING .
BEARINGS: TYPE. GRFASE OIL
: IMPELLER: TYPE SI7F FUR MAX
VENT CONN: DRAIN CONN
FLUSHING CONNECTION:
DRIVER
FURNISHED WITH PUMP RV OTHERS
TYPE:
FRAME:
MANUFACTURER:
i " : — • -—-————__
• ENCLOSURE
,
VOLT^ Duftcr f -
' HP BDM
BEARINGS 1 IRRIGATION
COUPLING GUARD
•
5.4-37
-------
rates. For proper design, settling tests should be performed on
the sludge actually produced by the system. However, since this
is normally impossible (except for duplicate systems), absorber
effluent should be tested from a plant using a coal and scrub-
bing system as similar as possible.
5.4.21.3 Clarification—
The primary purpose of thickeners is to regulate the solids
content of the underflow. Clarification is also important since
the liquid is recycled to the system. Clarification of the
liquid should be sufficient to prevent damage to the recycle
pumps (if the liquid is used as pump seal water). The maximum
solids content should be the design criterion, with the minimum
solids level variable as a function of system load.
5.4.21.4 Recirculation—
Recirculation of underflow liquid has two functions. One
is to prevent plugging of the underflow pipeline. The second is
to guarantee a uniform flow to the thickener. The design of the
thickener should consider an underflow recirculation option, to
account for changes in solids loading when the scrubber is
operating at turndown or with a lower sulfur coal.
5.4.21.5 Solids Removal—
Thickeners should be designed with rake mechanisms that can
handle varying solids levels in the thickener. Because of the
varying amounts of sludge that may be produced, the thickener
must have a limitorque system to prevent the rake drive from
breaking due to excessive torque. For more details, refer to
the EPRI document entitled, "Sludge Dewatering Methods for Flue
Gas Cleaning Products."
Table 5.4-19 is a typical specification for a thickener to
be completed by the utility, architect/engineer, or system
supplier for the thickener vendor.
5.4.22 Flocculant Proportioning Pump
Flocculant is usually added to the thickener by a recipro-
cating pump, which can be of a piston, plunger, or diaphragm
type. The pump is invariably associated with a check valve in
the discharge piping and is characterized by its pulsating flow.
Diaphragm pumps have no packings and seals exposed to the
liquid pumped. It is possible to mount the pumping head of a
low-capacity diaphragm pump in a location entirely separate from
the drive. A major consideration in choosing a diaphragm pump,
however, is the frequency of diaphragm failure. The flexible
diaphragm, fabricated of metal, rubber, or plastic, has a short-
er life than a piston or a plunger; and routine maintenance
procedures should be established accordingly.
5.4-38
-------
Table 5.4-19. EPRI LIME FGD SYSTEM DATA BOOK
THICKENER SPECIFICATIONS
CHECKED BY
COMPUTED BY
COMPANY
EQUIPMENT 'NO
SUPPLIER
1
2
3
4
b
b
7
8
9
10
11
12
13
14
Ib
Ib
17
18
19
20
21
22
DATE sPFf..
DATE . ppfu
LOCATION
FOR USE ON TOTAL Nf>
P.O. NO. PRICE
NO.
NO.
. REQ'D.
EACH $
GENERAL INHUKMAIIUN
ABSORBER
BLEED:
f SOL IDS COMPOS IT
' SOLIDS REQ'D.:
MATERIAL
SIZE
TORQUE:
OPERAT
GEAR: TYPE
hREDUCER:
1 BEARINGS
SUPPORT:
HP
OVERLOAD
TYPE
TYPE
TYPE
DEVICE
CARE MATERIAL:
NO.:
MATL:
MATL:
NOTCH: 1
fYPE
PROCESS INFORMATION
FLOW GPM. pH TEMP.
•F
ION, i (DRY BASIS : CaS03.l/2 H?0 CaSOa-2 H?0 Ash
UNDERFLOW MIN, OVERLFOW MAX.
CENTRAL COLUMN
THK.
DRIVE UNIT
ING FT LBS, CUT-OUT FT LBS
MATERIAL SIZE
MATERIAL SIZE
MATERIAL SIZE
MATERIAL SIZE
MOTOR
VOLTS PHASE
: TYPE
TORQUE CAGE AND INFLUENT BAFFLE
BAFFLE: )IA HT
TRUSS AIMS
PIVOT AXIS INCLINATION ' MATERIAL
SCRAPER BLADES
THK. 1n., DEEP in.
EFFLUENT WEIR
THK. in., DEEP in.
NO.
BRIDGE WALKWAY
HERTZ
THK.
bIZL: WlUt IHK., MAX. Lb/f-K
HANDRAILS: TYPE SIZE
SPECIAL INSTRUCTIONS:
b.4-39
-------
Table 5.4-20 is the engineering data sheet for a typical
reciprocating pump.
5.4.23 Thickener Underflow Pump
The thickener underflow pump performs several duties in
lime slurry systems. The primary function is to remove solids
from the bottom of the thickener, then transfer the sludge to
one of three destinations: a waste pond, a thickener underflow
transfer tank, or a vacuum filter. Depending on the destina-
tion, the pumping head may vary significantly for a uniform
solids content slurry. The underflow pump must be designed to
handle high solids concentration with abrasive components and,
during upset conditions, corrosive conditions. Although the
underflow from a thickener should have a pH between 7 and 8,
early systems had insufficient pH control and suffered severe
corrosion at the pump. Even though rubber-lined pumps should
not be required, in many cases they are used as extra protec-
tion.
The pumps should have water seals to prevent erosion of the
pump shafts. Each seal water supply should be individually
controlled. The type of downstream equipment or the method of
solids control affects the size of the thickener underflow pump.
If solids are recirculated to maintain solids concentration in
the thickener, then the pump must be oversized. If possible,
the control system should be designed so that the thickener
underflow pump operates at a continuous uniform rate. However,
intermittent or variable operation also brings the need for an
oversized pump.
Table 5.4-21 is a typical specification to be completed by
the utility, architect/engineer, or system supplier for the pump
vendor.
5.4.24 Thickener Overflow Pump
The thickeners are normally so arranged that overflow goes
by gravity to a collection tank. From this tank, one or more
pumps return the clarified liquor to the system. Thickener
overflow has a variety of uses: as dilution water for stabi-
lized lime slurry, prequencher water, wash water for mist elimi-
nators, recycle water to the recirculation tank of the absorber,
and makeup water for an ash disposal system.
Because of this wide variety of uses, the designer should
determine the end use before the pump specification is prepared.
The pump head will vary considerably between uses. In addition,
the following must be considered when specifying the thickener
overflow pump:
5.4-40
-------
Table 5.4-20. EPRI LIME FGD SYSTEM DATA BOOK
FLOCCULANT PROPORTIONING PUMP SPECIFICATIONS
CUSTOMER
PI ANT
PUMPMFR:
SIZE » TYPE
.ITEM NO..
. P. O NO..
OPERATING CONDITIONS
SP CH AT PT • tO F_
VAP PRESS. AT PT, PSIA___
VIS AT PT. CP • 60°F_
CORR/EROS CAUSED ay
II CAPACITY RCQ'O: (QPH) (CC/MR| AT PT. MIN
OISCH PRESS., PSIO, DES
SUCT PRESS.. PSIG, Gf*. ___
DIFF PRESS. PSI. DES
SUCT. LINE: LENGTH. FT.. VEI
ACCEL. HO , FT. NPSH AVAIL.
*** MoanAi
rfAX. (R. V. SET)
*AX. nc<
MANUFACTURER'S DATA
TYPE: SIMPLEX.
MAX. CAPACITY
MIN. CAPACITY
.DUPLEX.
NPSH REO'O
FT
PLUNQ. DIA. (INCHES).
STROKE LENGTH (INCHES).
PLUNGER SPEED (STROKES/WIN)
LENGTH or STROKE ADJUSTMENT
STROKE ADJUSTMENT i j
TVC ( )
. ( )
BRAKE H. P.
CYI_ DESIGN PRESS. HYDRO. TEST PRESS..
Z) RELIEF VALVE. INTERNAI EXTERN Al
SIZE SETTINO
WHILE OPERATING
WHILE SHLTT DOWN
AUTO. WHILE OPER.
WEIGHTS AND DRAWINGS
OUTLINE DWG. NO._
SECTION DWO. NO._
RATIO
. H. P. RATING.
DRIVER
-FRAME NO
VOLTS_
ENCLOSURE.
MATERIALS AND CONSTRUCTION
LIQUID END IBABOFI 1
W-UNOER.
EXTERNAL CASING:.
.INTERNAL PARTS:.
GLANO/FOLLOWE °
VALVe* BALLS/CONES.
VALVC SEAT*
VALVE annv
VACVt SLEEVE
VALVE CAP
VALVE GASKETS
PACKINO
BASEPLATE (COMMON FOR PUMP AND DRIVES
COUPLING MFO. AND TYPE IW/OUARQ)
CONNECTIONS
DISCHAROE (SIZE. ASA RATING. FACINQI
LUBRICATION RECOMMENDED
RACK AMD PINION
P Arm >in
CONNECTION ROD BHQ5
HYDHAULIC Pr<
UIU-
NOTES:
1) CATIONIC AND/OR ANIONIC POLYELECTROLYTE.
2)
SHALL 6E FURNISHED BY PURCHASER AND SHALL BE MOUNTED ON THE
SHALL BE FURNISHED BY VENDOR A
VENDOR TO SPECIFY.
5.4-41
-------
Table 5.4-21. EPRI LIME FGD SYSTEM DATA BOOK
THICKENER UNDERFLOW PUMP SPECIFICATIONS
COMPANY
EOUIWENT NO. FOR USE ON
SUPPLIER p-°-
GENERAL INFORMATION
TYPE
DUTY: CONTINUOUS INTERMITTENT
SERVICE
PROCESS INFORMATION
LIQUID:
DESIGN FLOW: NORMAL MAX GPM
PUMPING TEMPERATURE °F
SP GR. 0 PUMPING TEMPERATURE
VISCOSITY e PUMPING TEMPERATURE
VAPOR PRESS » PUMP TEMP. (FT. LIO.)
PH VALUE
CORROSIVE MATERIAL
VII mi (MAX. DIA.5
HYDRAULIC INFORMATION FT. LIO.
SUCTION PRESS ABOVE LIO. (ABS ) ( + )
STATIC SUCTION | JPT (-)• HEAD ( + )
SUCTION FRICTION HEAD (-)
TOTAL SurTiriN HFAn (17+1R+19)
STATIC DISCHARGE HEAD
p[srHARr,r FRICTION HEAD
DISCHARGE PRESS ABOVE IIQ (ARS )
T(1TA| DISCHARGE HEAD (21+22+23)
TOTAL DYNAMIC HEAP (?4-?0)
NPSH AVAII API F (20-13)
NPSH RFQIIIRFn
PUMP
MANUFACTURER
RPM
PfRFfiRMANCF CURVE
SERIAL NO
p MAX FIQU FOR IMPfl 1 FR
ROTATION P PBIVE SHAFT fNp
COPIES REQUIRED OF:
PERFORMANCE CURVES
DIMENSION DRUGS.
OPERATING AND MAINTENANCE INSTRUCTIONS
LIA.AI iun
TOTAL NO. REQUIRED
NO. PRICE EACH J
MATERIALS
MATFRI4I CnnF . FXTFRNAI CASING tNTfONAL PARTS
I - CAST IRON INTERNALS CODE I B S C
B - BRONZE IMPELLER I B S C
S - STEEL INNER CASE PARTS I B S C
C - 11-131 CHROME SLEEVE (PACKED) Ch Ch Af Af
A - ALLOY SLEEVE (SEAL) C C C C
h - HARDENED HEAR RINGS B Ch Ch
f - FACED SHAFT S S S S
LANTERN RING
PACKING GLAND
SUCTION CONN: SHE POSITION
DISCHARGE CONN: SIZE POSITION
fONN RATING TYPE
PACKING TYPE
| 1NTFRN BTNC,<; MATFRIAl
mm ING
RFAR1NGS: TYPE GREASE OR
IMPFIIFR: TYPE SIZE FUR. MAX.
VFNT CONN: DRAIN CONN.
FIIISHING CONNECTION:
DRIVER
FURNISHED WITH PUMP BY OTHERS
TYPE:
FRAME:
MANUFACTURER:
FNCI OSIIBF
VOLTS PHASE CYCLr
HP RPM
BEARINGS IMBRICATION
rmiPI ING fiUARD , ,
NOTES:
5.4-42
-------
1. Corrosion
2. Pump seals
5.4.24.1 Corrosion--
The dissolved salts content of the overflow liquor is hdgh
Therefore, to avoid corrosion, the pump must be made of high-
alloy steel or be rubber lined.
5.4.24.2 Pump Seals—
Since the pump may on occasion have to handle high solids
levels, water seals should be used instead of mechanical seals.
Individually controlled seal water systems should be used to
protect the shaft from solids erosion.
Table 5.4-22 is a typical specification to be completed by
the utility, architect/engineer, or equipment supplier for the
pump vendor.
5.4.25 Thickener Overflow Tank
The thickener overflow tank acts as a surge tank and stores
supernatant liquid from the thickener to be pumped to the vari-
ous locations in the scrubbing system to maintain a water bal-
ance. Factors to be considered in the design of the thickener
overflow tank include:
1. Corrosion
2. Erosion
3. Tank size
5.4.25.1 Corrosion—
The pH of the thickener overflow should be between 6 and 7;
however, pH excursions may occur. The tank, therefore, should
be made of carbon steel clad with stainless steel, or carbon
steel lined with rubber or fiberglass.
5.4.25.2 Erosion—
Unless there are upsets in the thickener, there should be
no large quantities of slurry solids in the overflow tank, and
serious erosion problems should not occur. Tanks may be lined
with stainless steel, rubber, or fiberglass.
5.4.25.3 Tank Size—
in many cases, the thickener overflow is reused to achieve
a water balance in the scrubbing system. The tank should be
sized to allow for some system swings. Adequate NPSH for the
overflow pump must also be considered.
Table 5.4-23 illustrates a typical specification as it
would be completed by the utility, architect/engineer, or system
supplier for the tank vendor. =>*si-em
5.4-43
-------
Table 5.4-22. EPRI LIME FGD SYSTEM DATA BOOK
THICKENER OVERFLOW PUMP SPECIFICATIONS
COMPANY
EQUIPMENT NO. TO* USE ON
SUPPLIER p-°-
GENERAL INFORMATION
TYPE
nilTY- CONTINUOUS INTERMITTENT
SERVICE
PROCESS INFORMATION
LIQUID:
DESIGN FLOW: NORMAL MAX GPM
PUMPING TEMPERATURE °F
SP GR 9 PUMPING TEMPERATURE
VISCOSITY B PUMPING TEMPERATURE
VAPOR PRESS B PUMP TEMP (FT.LIO.)
PH VA1UE
CORROSIVE MATERIAL
SOLIDS (MAX, D1A.) , „ _„
HYDRAULIC INFORMATION FT. L10.
SUCTION PRESS ABOVE LIO (ABS ) (+)
STATIC SUrTION | I FT (-}• HEAD ( + )
SIICTIDN FRICTION HEAD (-)
TOTAL WTION HfAn (17+18+19)
STATK n'SCHARGE HEAD
pJSCHARCiF FRICTION HEAD
PI<;CWARC,F PRfSS ARDV.E LIQ- (AflS )
TOTAI DISCHARGE HEAD (21+22+23)
TOTAL DYNAMIC HEAD (24-20)
NPSH AVAIIABIF (?n-!3)
NPSH RFQIIIRFD
PUMP
MANUFACTURER
RPM
PFRFORMANCF CIIRVF
SERIAL NO
BPH (B SERVICE CONDITIONS
PMA Y Fl OU FOB THPFI t FR
ROTATION (9 PRIVF SHAFT END
COPIES REQUIRED OF:
PERFORMANCE CURVES
DIMENSION DRUGS.
OPERATING AND MAINTENANCE INSTRUCTIONS , .
LIX.AI iun
TOTAL NO. REQUIRED
NO. PRICE EACH $
MATERIALS
MATEP1AL ™pf - fKTFRNA! CASING TNTF_PHA|_ PARTS
I - CAST IRON INTERNALS CODE I B S C
B - BRONZE IMPELLER I B S C
S - STEEL INNER CASE PARTS I B S C
C - 11-13* CHROME SLEEVE (PACKED) Ch Ch Af Af
A - ALLOY SLEEVE (SEAL) C C C C
h - HARDENED MEAR RINGS B Ch Ch
f - FACED SHAFT S S S S
LANTERN RING
PACKING GLAND
snnins COMN- sm POSITION
pfSCHADGF C.ONN- SI7F POSITION
rnNN BATING TYPE
DACItlNC,. TYPF
I^TFPW B1MC,S MATFRIAI
rnni TNG
BEARINGS: TYPE GREASE OR
THPFI 1 FR: TYPE SIZE FUR. MAX.
VFNT CONN: DRAIN CONN.
FMISHINC mNNFCTTON:
DRIVER
FURNISHED WITH PUMP BY OTHERS
TYPE:
FRAME:
MANUFACTURER:
Fnrm
-------
Table 5.4-23. EPRI LIME FGD SYSTEM DATA BOOK
THICKENER OVERFLOW TANK SPECIFICATIONS
Sheet
of
CUSTOMER
PLANT LOCATION
SERVICE
JOB NO.
EQUIPT. NO.
FILE NO.
P.O. NO.
Type of Tank
Size:
Diam.
Height_
Capacity
GENERAL NOTES
1) For required capacity as shown, Mfg. to advise
diameter and height of tank for the most
economical utilization of plate.
2) Nozzle orientation to be furnished later.
3) Nozzle location and design tube fu
later with mechanical design.
4) Ladder Clips & Ladder:
Inside Out*
5) Design P.
6) Paint
Desion T
7) Lining - Fiberglass or
NOZZLES
Inlet
Outlet
Drawoff Elbow
P & V Vent
Level Gage
Thermowell
Roof Manhole
Shell Manhole
Tank Materia1
Level Gage 01
Gage Hatch:
Thermometer:
Make
MARK
rnished
ide
rubber Note 1
NO.
SIZE
RATING
DESIGN DATA
Min. Plate Thick
owance: Shell in. ; Bottom T
APPURTENANCES (
" Gate Column: Yes No
Fig. or FQU;,
icuum vent Va
Mot
le
?J Yes No
19. or
No. Make Fia
Yes
1 ~ ' oFT
Yes NO
•qua
Ro~
qua
s
5 or Eaual
SP GR
n.; Roof in
fil . L_
, I*PJ
1 Float
Pressure
1 vacuum
. or E
Length
Material
tern Length
Range "
qual
In
In
• -in
REMARKS:
5.4-45
-------
5.4.26 Centrifuges
To date, centrifuges have only been used experimentally for
dewatering scrubber sludges. They do offer a consistent product
that is uniform and easily handled. If a consistent sludge
product is required as part of a regulation program, then cen-
trifuges may be used extensively. The lack of clarified cen-
trate is not a problem in lime slurry systems, since the cen-
trate can be recycled to the scrubbing system. The following
factors should be considered in specifying a centrifuge for lime
slurry applications:
1. Materials of construction
2. Rotational speed
3. Conveyor
5.4.26.1 Materials of Construction—
The erosive and sometimes corrosive nature of scrubber
sludges requires that all liquid contact materials in the cen-
trifuge be made of 316L stainless steel. The tips of the con-
veyor should be made of tungsten carbide to reduce abrasive
wear.
5.4.26.2 Rotational Speed—
Rotational speed should be midrange, 3000 rpm or less, to
gain some of the benefits of high-speed clarification while
preventing excessive abrasions and difficult solids discharge.
If centrifuge speeds are too high, the conveyor and the bowl
will lock.
5.4.26.3 Conveyor Speed—
The screw conveyor within the bowl should turn at the
minimum speed required to remove solids without making excessive
turbulence. Since the scrubber cake rates may vary, a variable-
speed conveyor should be specified.
Table 5.4-24 is an illustration of a specification as it
might be prepared by a utility, architect/engineer, or systems
supplier for the centrifuge vendor. For more details, refer to
EPRI report, "Sludge Dewatering Methods for Flue Gas Cleaning
Products."
5.4.27 Vacuum Filters
Vacuum filters are used at several locations to dewater
lime slurry sludges. Although they are bulky and use a signifi-
cant amount of energy, they respond well to varying sludge
properties and quantities. The following factors should be
considered in specifying a vacuum filter:
5.4-46
-------
Table 5.4-24. EPRI LIME FGD SYSTEM DATA BOOK
CENTRIFUGAL SEPARATOR SPECIFICATIONS
CHECKED BY _
COMPUTED BY
COMPANY
DATE
DATE
SPEC. NO.
PROJ. NO.
LOCATION
EQUIPMENT NO.
SUPPLIER
FOR USE ON
P.O. NO.
TIME NO. REQ'D.
PRICE EACH $
DUTY
1ISLUDGE, MAX. LB/HR"
2
J1
4
j
b
7
B
LIQUID IIMHLKAIURL: °f
SOLIDS: LB/HR
PRESSURE, IN:
PRESSURE, OUT:
AP ACROSS SEPARATOR:
INSTALLATION: INDOOR " OUTDOOR "~"
OPERATION: CONTINUOUS™ INTERMITTENT a
9 CUNblANI: PRLSSURL FLOW
10
1
T2"
U
14
Ib
WINU VELOCITY: M. >.H.
ISEISMIC LOAD: ' G'S "
SPEED RPM
MOTOR: TYPE , VOLTS HERTZ
DRIVE: TYPE , MATERIAL
SOLIDS STORAGE CU.FT
SOLIDS DESCRIPTION
16
\l
18
19
20
21
i7
i!J
24
2b
COMPOSITION
LB/CU.FT. S.G.
TEMPERATURE: *F.
CORROSIVED HYGROSCOPICL" ABRASIVE LT
PERMEABILITY:
COMPRESSIBILITY:
CAKE THICKNESS: MAX.
PHYSICAL DESCRIPTION
?fi
?l
?H
M
311
31
M
3,H
34
3S
.%
3/
38
39
40
VESSEL MAT'L.
THICKNESS:
LEG SUPPORTEDD B
VESSEL SIZE:
PIPE LINE SLUDG
SIZE
WALL THICKNESS
MATERIAL
JACKET SUPPORTED L
E LIQUID OUT
INTERNALS
SCREEN: SIZE
OPENINGS: SIZE
CLOTH: MATERIAL
, MATERIAL
.PITCH
. THK.
BREAKING TENACITY:
SPECIFIC RESISTANCE:
SPECIAL INSTRUCTIONS
5.4-47
-------
1. Materials of construction
2. Drying time
3. Barometric legs
4. Filter medium
5.4.27.1 Materials of Construction—
To prevent corrosion, the piping and support members in a
vacuum filter in lime slurry applications should be made of
stainless steel. Drum heads in the filter can be made from
carbon steel if sufficient corrosion allowances are provided.
5.4.27.2 Drying Time—
The drying zone in a vacuum filter should be designed for a
10-second drying time. If longer drying times are used, the
sludge will crack. The drying air then short-circuits through
the cracks and does not continue to dry the cake. This condi-
tion would require the purchase of an excessively large vacuum
pump to pull in the excess air. The 10-second drying time
maximizes the cake solids content of the filter cake without
requiring an excessively large vacuum pump.
5.4.27.3 Barometric Legs—
To keep filtrate from going through the vacuum pump, a
barometric leg should be installed to protect the pump by trap-
ping liquid before the suction of the vacuum pump.
5.4.27.4 Filter Medium—
The filter medium should be cheap, durable, and noncorro-
sive, and it should allow easy cake discharge. In lime slurry
applications, polypropylene appears to be the best material.
Table 5.4-25 is a typical specification for a vacuum filter
to be prepared by a utility, architect/engineer, or system
supplier for a vacuum filter vendor.
For more information, refer to EPRI report, "Sludge De-
watering Methods for Flue Gas Cleaning Products."
5.4.28 Filtrate or Centrate Pump
This pump returns the clarified filtrate from the centri-
fuge or vacuum filter to the scrubbing system. The liquid is
normally returned to the thickener overflow tank or to the
thickener itself. In either case, the solids content of the
filtrate or centrate should be low and the pH should be 7 or
greater. This pump should see the mildest duty of any process
pump within the scrubbing system. The following factors should
be considered when specifying the filtrate or centrate pump:
5.4-48
-------
Table 5.4-25. EPRI LIME FGD SYSTEM DATA BOOK
VACUUM FILTER SPECIFICATIONS
CHECKED BY DATE <;pFr Nn
COMPUUD BY DATE ppm *n
COMPANY LOCATION
EQUIPMENT NO. FflR I.SE ON TIME NO. REO'D
SUPPLIER _ p.o. Nn pplr[ EflfH s
t.
3
5
6
9
T
T
5
6
^
8
9
U
1
't
3
r
^J
DUTY
ISLUDGE, MAX. ~ LB/HR
LIQUID TEMPERATURE: ~ Sp
SOLIDS: LB/HR
"PRESSURE, OUT:
;^P ACROSS SEPARATOR: ' '
INSTALLATION: INDOOR H OUTDOOR D
'OPERATION: CONTINUOUS M INTERMITTENT |
CONSTANT: PRESSURE * FLOWfi
4^
"MOTOR : t Y P E , VOLTS HERTZ
DRIVE: TYPE 7 MATERIAL
SOLIDS STORA~GE ' CU FT "
SOLIDS DESCRIPTION
COMPOSITION
LB/CU.FT. S.G.
TEMPERATURE: *>F "
CORROSIVED HYGROSCOPICD ABRASIVE"
PERMEABILITY:
COMPRESSIBILITY:
CAKE THICKNESS: MAX
PHYSICAL DESCRIPTION
. 26 IVESSEL MAT'L.
27 THICKNESS:
Ljg_LEG SUPPORTEPn BRACKET SUPPORTED H
i.29 VESSEL SIZE:
i ^OlPlFE LINE SLUDGE LIQUID OUT
1 31 (SIZE |
i 32 IWALL THICKNESS! I
JJ riM i tKIAL
.. J INTERNALS
I 34ISCREEN: SIZE , MATERIA' "
[35 (OPENINGS: SIZE .PITCH
36 CLOTH": MATERIAL THK^
37 IBREAKING TENACITY •
[38 (SPECIFIC RESISTANCE:
19] ~ ~
40 — —
SPECIAL INSTRUCTIONS
5.4-49
-------
1. Erosion
2. Corrosion
3. Pump seals
5.4.28.1 Erosion—
Operating upsets in the centrifuge or the vacuum filter may
allow solids to enter the filtrate or centrate. Although these
occurrences should be rare, the designer may want to protect
against them. In this case, an erosion resistant alloy or
rubber-lined pump should be used.
5.4.28.2 Corrosion—
In most thickeners on lime slurry systems, the pH of the
underflow will be between 7 and 9 because of the unreacted lime
in the slurry. Therefore, the feed to the vacuum filter or
centrifuge is in the same range. In cases where the pH is so
low that an erosion resistant alloy is not suitable, the thick-
ener will not operate properly. Also, the vacuum filter could
be stopped to protect a pump made of an erosion resistant alloy.
From a corrosion and erosion standpoint, both erosion resistant
alloy and rubber-lined pumps are suitable.
5.4.28.3 Pump Seals--
Since the pump on occasion handles solids, water seals
should be used instead of mechanical seals. Individually con-
trolled seal water systems should be used to protect the shaft
from solids erosion.
Table 5.4-26 is a typical specification as it would be
completed by the utility, architect/engineer, or equipment
supplier for the pump vendor.
5.4.29 Fixation Additive Silos
The following factors are important in specifying a fixa-
tion additive storage silo:
1. Materials of construction
2. Conveying systems
5.4.29.1 Materials of Construction—
Cylindrical unlined carbon steel tanks with cone bottoms or
concrete silos can be used for storing fixation additive.
5.4.29.2 Conveying Systems—
Either pneumatic or mechanical conveying systems are used
to move additives. Pneumatic systems are more common and easier
to maintain, but allow water vapor into the system if not oper-
ated in a closed flue manner. Larger fabric filters are re-
quired for silos handling pneumatically conveyed lime than for
silos handling mechanically conveyed lime.
5.4-50
-------
Table 5.4-26. EPRI LIME FGD SYSTEM DATA BOOK
FILTRATE OR CENTRATE PUMP SPECIFICATIONS
LOCATION
Cum-""1 — •-• " •
irt1FMT NO. FOR USE ON TOTAL NO. REQUIRED
rlirp| TPH P.O. NO. PRICE EACH $
"====::=:^ GENERAL INFORMATION
TYPE
TV CONTINUOUS INTERMITTENT
DUi ' •
SER* l
~~ PROCESS INFORMATION
t m 1 TO '-
L j QU i "•"
rN pLnw NORMAL MAX GPM
-ipJNG TEMPEHATURF °F
r0 a PUMPING TEMPERATURE
Vn'ITY » PUMPING TEMPERATURE
^Q_,lll V
,^D PRESS C PIIMP TEMP- (FT.LIQ.)
p^ VALUE • ••••
,«, IDS («**• DIA->
COL * v * —
:-- — HYDRAULIC INFORMATION FT. L1Q.
^^^Q^RFSS. ABOVE LIQ. (ABS.) ( + )
,ir SUCTION IIFT (-): HEAD ( + )
rTlON FRICTION HEAD (-) ,
, --UCTIOM wFAn (17+18+191
TOTAL ->"'-"'-'
,. pjscHAR^F HEAD
STA • *
uARGE FPITT10N HEAD
nlS^"
.nnrr PRES1^ . APOVF LIQ. (ARS ) , .
nlSCH>»KUt
01 niSCH"" HFAD (21+22+23)
TOTAL ul;"-r
j HYNAMIC "F&n (24-20)
TOTAL D'""nlL
AVAILABLF (?0-13)
2!-2 — —^ PUMP
MANyFA
flpH - —
ORMANCE CU°VF
5E^l^SERVIC£ CONDITIONS
B MAX FLO" FOR IMPELLER
IQN t DRIVE NG AND MAINTENANCE INSTRUCTIONS
MATERIALS
MATERIAL CODF . FXTFRNAI CASING INTFRNAI P4RT^
I - CAST IRON INTERNALS CODE I B S C
B - BRONZE IMPELLER I B S C
S - STEEL INNER CASE PARTS I B S C
C - 11-m CHROME SLEEVE (PACKED) Ch Ch Af Af .
A - ALLOY SLEEVE (SEAL) C C C C
h - HARDENED WEAR RINGS B Ch Ch
f - FACED SHAFT S S S S
LANTERN RING
PACKING GLAND
SUCTION CONN: SIZE POSITION
DISCHARGE CONN: SI7E ....POSITION .
CONN RATING TY"F
PACKING TYPF
1 ANTFRN RINM MATCRIAI
COOLING
BEARINGS: TYPF GREASF OIL
1MPFHFR: TYPF SIZE FUR .MAX.
VENT CONN: DRAIN CONN.
FIIKHTNG CONNFrnnN:
DRIVER
FURNISHED WITH PUMP BY OTHERS
TYPE:
FRAME:
MANUFACTURER:
FNCI fKMRF
VOLTS PHASE CYCLE
HP RPM
RFARINGS IIIBRICATION
COUPLING GUARD
5.4-51
-------
5.4.30 Fixation Tank or Pug Mill
The fixation tank is used to mix a fixation agent with the
spent slurry (sludge). In the Dravo system, the clarifier
underflow is used as a feed, the fixation agent is added, and
the mixed product flows or is pumped to a sludge pond. In the
IUCS system, lime, fly ash, and sludge from a vacuum filter are
mixed in a pug mill. The product from the pug mill is a moist
fixation t^nk ^Wlng fact°rs should be considered in designing a
rixacion tank or a pug mill:
1. Corrosion
2. Erosion
3. Tank size
5.4.30.1 Corrosion--
sion TQGnSp-ent Slurry has a pH of 8 or more; therefore, corro-
orotert JV Wa,m^°r problem- The equipment could be stopped to
protect it from low pH swings.
5.4.30.2 Erosion--
ie major design consideration. To achieve
mixing of the slurry, high torque agitators are needed.
1-ant ,'i? m°vem,^ «, ~i
be considered SH "^««nt8 of he pump must
Table 5.4-27 is a typical specification as it would be
5.4.31 Sludge Disposal Pump
Vacuum filtatio or
pump Include :t0 ^ C°nsidered when specifying a sludge disposal
5.4-52
-------
Table 5.4-27. EPRI LIME FGD SYSTEM DATA BOOK
FIXATION TANK SPECIFICATIONS
Sheet
of
CUSTOMER
PLANT LOCATION
SERVICE
JOB NO.
EQUIPT. NO.
FILE NO.
P.O. NO.
Type of Tank
Size:
Diam.
Height
Capacity
GENERAL NOTES
1) For required capacity as shown, Mfg. to advise
diameter and height of tank for the most
economical utilization of plate.
2) Nozzle orientation to be furnished later.
3) Nozzle location and design tube furnished
later with mechanical design.
4) Ladder Clips & Ladder:
Inside Outside
5) Design P.
Design T
6} Paint
7) Lining - Fiberglass or rubber Note 1
NOZZLES
Inlet
Outlet
Drawoff Elbow
P & V Vent
Level Gage
Thermowell
Roof Manhole
Shell Manhole
MARK
NO.
SIZE
RATING
•
DESIGN DATA
Tank Material
Corrosion Allowance:She!'
Min. Plate Thick
SP GR
1n.; Bottom
in.; Roof
Level Gage
Make
Pressure &
Make
Gage Hatch:
Thermometer
Make
Thermometer
Make
or Gate
Vacuum
Well:
I
A
Column:
F1q.
Vent Valve:
Fig.
Yes No.
Model
Yes
Model
P P U R T E N
Yes
Yes
Make
Yes
No~
A N C E S BY ]
No Type
or Equal Float
No Pressure
oz .
or Equal Vacuum n?
Fig.
or Eoual
No Lenqth in
or Equal Material
Stem Length in
or Equal Ranqe
REMARKS:
5.4-53
-------
1. Corrosion
2. Erosion
3. Pump seals
4. Pumping distance and head
5.4.31.1 Corrosion—
Excess lime should cause the sludge to have a pH between 7
and 9. Low pH should only occur during system upsets; there-
fore, erosion resistant alloy or rubber-lined pumps would be
acceptable from a corrosion standpoint. During the rare pH
excursions, the sludge disposal system could and should be
stopped to protect the erosion resistant alloy pump.
5.4.31.2 Erosion—
Sludge having a high solids content and containing abrasive
fly ash requires rubber-lined or erosion resistant alloy pumps
to prevent excessive pump wear.
5.4.31.3 Pump Seals—
Since the pump handles abrasive solids, seal water should
be used instead of mechanical seals. Individually controlled
seal water systems with alarms should be used to protect the
shaft from erosion by the solids.
5.4.31.4 Pumping Distance and Head—
The ultimate disposal site affects the type of pump select-
ed. Rubber-lined pumps are limited to 120 ft of head, erosion
resistant alloy is limited by excess erosion to 400 ft of head,
and reciprocating pumps are limited only by economic design.
Table 5.4-28 is a typical specification as it would be
completed by the utility, architect/engineer, or equipment
supplier for the pump vendor.
5.4.32 Sludge Conveying System
5.4.32.1 Belt Conveyor—
If the sludge is solid and nonthixotropic after the addi-
tion of the fixation agent and vacuum filtration, it can be
transported to the disposal pond by the sludge conveying system.
The most commonly used system is a belt conveyor that, given
good routine maintenance, can outlast any other type of con-
veyor .
Belt conveyor design begins with the material to be
handled. Since weight per cubic foot is an important factor, it
should be determined accurately in an "as-handled" condition,
rather than taken from published information. Lump size is
another important factor. If the feed chute and the belt slope
are properly designed, there is little problem with lumps fall-
ing off. Belt conveyor slopes are limited to a maximum of 30
degrees; those in the range of 18 to 20 degrees are more common.
5.4-54
-------
Table 5.4-28. EPRI LIME FGD SYSTEM DATA BOOK
SLUDGE DISPOSAL PUMP SPECIFICATIONS
LOCATION
CUW1"— — • • ..... .
...TDMFNT NO FOR USE ON TOTAL NO. REQUIRED
TFR P.O. NO. PRICE EACH $
===::::::=^ GENERAL INFORMATION
TYPE — —
TV. cnNTlNlinu<; INTERMITTENT
SERV
"" ~~ PROCESS INFORMATION
======TP==~
cTTN FLnu' NORMAL MAX GPM
rn ffl PUMPING TEMPERATURE
" r,<-TTY a PUMPING TEMPERATURE
-, DDC*;^ 13 PUMP TEMP (FT. LIO.)
PH VAUUt
_«ctut MATERIAL
t — HYDRAULIC INFORMATION FT. LIO.
==H^PBF«. ABOVE LIO. (ABS.) ( + )
1C SUCTION I I FT (-)• HEAD ( + )
rTTON FRICTION HEAD (-)
i AUCTION "FAD (17+18+19)
TOTAL -ui-"u
rf DISCHARGE HEAD
ST*1 *
-inBGE FRITTION HEAD
fljSl-^'
ARGE PPES^ AROVF LIQ. (ARS )
DlSLfl"nlSCufiBr,F HFAD (21+22+23)
TOTAL J'iLnr • -
nYNAMIC HFAn (24-20)
MANU^*0
«**—• 7c r,,DUF
DfRfO""
nl NO
SERTsERVlCE CONDITIONS
MAX FLOW FOB IMPFI1FR
mw a DRIVE SH1(rT END
j c REOUIRFP OF:
C0plt5 Ittwu""1
•j^lON DRW^S .
D'MtnT'NG AND MAINTENANCE INSTRUCTIONS
MATERIALS
MATFRIAI rnnp . FXTFPNAI TAKING INTERNAL PARTS
I - CAST IRON INTERNALS CODE I B S C
B - BRONZE IMPELLER I B S C
S - STEEL INNER CASE PARTS I B S C
C - 11-13X CHROME SLEEVE (PACKED) Ch Ch Af Af
A - ALLOY SLEEVE (SEAL) C C C C
h - HARDENED HEAR RINGS B Ch Ch
f - FACED SHAFT S S S S
LANTERN RING
PACKING GLAND
Sljr.TIdN CONN: SI7E POSITION .
PHI-HARRF mNN: <;I7F POSITION . —
CONN PATlNr TYPF_
PACKJUr, TYPF —
1 ANTFPN R1NC,S MATERIAL
COOLING ______ _
! RFAHlNrA- TYPF GREASE ..... OIL
IMPELLER: TYPE SIZE FUR. ,,. MAX
UFNT r.ntiti: DRAIN CONN.
FLUSHING CONNECTION:
DRIVER
FURNISHED WITH PUMP BY OTHERS
TYPE:
FRAME:
MANUFACTURER:
FNCKKMRF
VOLTS PHASE CYCLE
HP RPM
PFAHINK<; IIIBRICATION
COUPLING GUARD
TO STATE WITH QUOTE
5.4-55
-------
The temperature and the chemical activity of the sludge
play an important role in belt selection. Belts can be quickly
damaged by high temperatures; a high-priced belt may prove most
economical in the long run. The elastomers available for belt
construction include Neoprene, Teflon, Buna-N rubber, and vinyl
rubbers.
The belt conveyor tonnage requirement should be specified
at the peak rather than the average load. It is advisable to
work with a manufacturer to calculate the horsepower needed to
drive a belt conveyor. Belt tension can be calculated from
drive shaft horsepower. Since various combinations of width and
ply thickness will develop the required strength, final selec-
tion is influenced by lump size, ability of the belt to form a
trough, and ability of the belt to support the load between the
idlers. Once final belt selection is made, idlers and return
rolls can also be selected.
Table 5.4-29 is a typical engineering data sheet for a belt
conveyor.
Care should be taken not to use conveyor belts and chutes
for the transfer of unfixed thixotropic sludge. The vibrations
in such a system would cause undesired liquefaction. If the
sludge to be conveyed is gypsum, there is little chance that it
will be thixotropic.
5.4.32.2 Screw Conveyor—
Another common sludge conveying system is the screw con-
veyor, which consists of a helical flight mounted on a shaft and
turning in a trough. Power to convey must be transmitted
through the shaft and is limited by its allowable size. Screw
conveyor capacities are generally restricted to about 10,000
ft3/h.
In addition to a wide variety of designs for the compo-
nents, screw conveyors may be fabricated of several materials,
ranging from cast iron to stainless steel. Since sections are
coupled together, special attention should be given to bending
stresses in the couplings. Screw conveyors operate at low
rotating speeds, and the outer edge of the flight may be moving
at a relatively high linear speed. This can create a wear
problem, which can be reduced by the use of hard-surfaced edges,
rubber covering, or high-carbon steels.
Horsepower calculations for screw conveyors are well stan-
dardized; however, each manufacturer has grouped numerical
contents in a different fashion on the basis of individual
design variation. Thus, in comparing horsepower requirements,
it is advisable to use a specific formula for specific equip-
ment. The typical feed arrangements for heavy or lumpy material
5.4-56
-------
Table 5.4-29. EPRI LIME FGD SYSTEM DATA BOOK
BELT CONVEYOR SPECIFICATIONS
C4KCHI IT MTE
COMPUTED IT UTE
CBMPANf
MtMMENT ID. roi til OR
SUPPLIER , o
»«IT 1 »r
1PEC. to. .
»ROJ. NO. .
LOCATION
TBTii mn ftn-f.
NO.
DITTY
\
t
I
4
1
i
7
1
CON»EYO« IITH. LIFT
TPNi »»E. SURCt BESI6H
MO* LOADED:
AX6LE TC D1RECTIOH Of KlT TRAVEL
MOORS/DAT
HE. OF LOADING PTS.
INSTALLAT10I1: OUTDOOR IIDOOR
OPCMTION: COHT1NUOUS IITEIK' T . f~l
FEED DESCRIPTION
t
10
11
"
13
11
IS
li
17
11
It
40
41
42
43
44
41
4t
47
41
41
10
1)
4
II
NAME
OfSITY LI./CU. FT.
MOISTURE
TEMP. *f. LUMPS »f. FINES »r
S12E:
«»l. LUMP t or FEED
CHARACTERISTICS:
STICKY
CORROSIVE
A»RASI»E
ANCLE OF REPOSE
viler IACM t
ACCESSORIES
K
n
u
13
}t
„
It
??
n
?»
3tl
31
3?
33
31
n
36
37
38
3»
S(1»T|04EOS : L(T«.
MAllHIT: UIO'M TY»[
DECHNC MANDRAlLS
COVE! (TYPE)
CUAtC
xoiDiAct: BAN;. IATCNET
"OLDIACI LOCATIOk
STI|«6!«S: HIRE lOPt . CNAIRti
TRUSS, BEOPLATE
ICLT WIPER: BRUSH, WIRE, SPNtr
ILAOE (SIRtLE. BUPlEl. OuAC. )
TAtE-uP: TIA»EL
SCREW CRAVITT. JPIU;
VERTICAL, MORUORTAL
DRUE: »-IElT, CNAIR, SHAFT. MOUNTED
MOTOR TTPE:
"' «M ». »N. CT.
DESCRIPTION
IELT: KIDTM LtlM. tPECD |« f
•WT'L. CRADE FINISH
IREAtER STRIP: Y£sl] NO T
'*'"' »l» I CO»tlS
02.
TPE Or SPLICE
lOLfl: TTPE TIOUtMERS tETURNS IMPACTS
BIA.
fPACINC
(EA«IR(
IUIE.
PULLEY: TYPE HUD TAIL >NU|
«UT'l.
• IA.
UTN.
CIOMN
lAStl.C
MUIS
MCS.
TROUti- RETURN
TtAINERS TRAINERS
•END TAtE-UP
1 , . .
5.4-57
-------
include rack-and-pinion, bin, and side inlet gates. Typical
discharge arrangements include open-end trough, discharge trough
end, and rack-and-pinion side gates.
Table 5.4-30 is a typical engineering data sheet for a
screw conveyor.
5.4.33 Pond Water Return Pump
In systems that do not incorporate a thickener, or that do
not produce a 50 to 60 percent solids sludge that can be ponded
or landfilled, excess water from the sludge pond must be recy-
cled to the scrubbing system. Water recycling is required to
make the system a closed loop. Large sludge ponds separate out
suspended solids and allow reactions to go to completion. The
pH in the sludge pond rarely changes. The following factors
should be considered when specifying the pond water return pump:
1. Erosion
2. Corrosion
3. Pump seals
4. Pump placement
5.4.33.1 Erosion—
With good pond design there should be few solids in the
pond return water; however, reliability of return water pumps
can be improved by accomodating some suspended solids.
5.4.33.2 Corrosion—
The pump should be designed to take a pH range of 6.5 to 9.
Since chloride levels are usually higher in closed-loop systems,
the pump should be chloride resistant. This will be true for all
pumps, if closed-loop operations are specified and if it can be
determined that chloride levels will be high. This corrosion is
perhaps best deterred by using rubber (natural or synthetic)
liners.
5.4.33.3 Pump Seals—
Since the pump will occasionally handle solids, water seals
should be used instead of mechanical seals. Seal water should
be controlled with an independent regulator and alarm for each
pump.
5.4.33.4 Pump PIacement—
The pond water return pump should be placed so that the
water has traveled as far as possible from its point of entry to
the pond. As a consequence, the desired precipitation reactions
will occur and as many solids as possible will settle out.
Table 5.4-31 is a typical specification to be completed by
the utility, architect/engineer, or equipment supplier for the
pump vendor.
5.4-58
-------
Table 5.4-30. EPRI LIME FGD SYSTEM DATA BOOK
SCREW CONVEYOR SPECIFICATIONS
CHECKED »Y BATE
COMPUTED It, BATE
rqillPHIIT Hfl t$m utt B.
IU"LIER_ , B ,n
MEET i
IPEC. NO.
PROJ. *0..
purr r.ru f
.tF
I
t
I
4
1
(
7
1
1
10
11
1?
11
14
IS
11
17
18
19
20
21
22
23
24
IS
H
r
re
M
10
11
12
DUTY
COWTINS LENGTH un
T.P.H. AVERAGE SURGE
HOURS PER DAY
DO. LOADING POINTS
*0. DISCHARGE POUTS
1»ST*LL»TIOX: INDOOR Q OUTDOOR ~!
OPERATION: COITTIMUOUSQ INTERMITTENT Q
MOM LOADED:
FEED DESCRIPTION
IAMC OF MAT 'I.
OEIISITT: 11. PER CU. FT
MOISTURE (SURFACE) ,
TEMPERATURE .,
SIZE: I
• t FEED
ANGLE OF REPOSE:
STICKY Q CORAOSUt Q r«IABLE D
AmSIYE D COHTAHINAILE ~1
H»tROJCOPIC PI
OPERATING CHARACTERISTICS
SIZE SCREW DIA
SPEED: »,n
TROUGH LOADING: 1SI Q 101 C «** CI
LEFT Q IUMTQ MAUD CONVEYOR
DUST TUHT D HATE* TItNT C OPEN ~
El£CTRICAL DATA
"««: ». PH. CT.
«OTO«: NP OI1P-PROOF D
IIPL. «oor D unv G Ttft D
CONSTRUCTION
)}
34
3S
36
37
38
3«
40
41
4?
43
14
4S
46
THOUGH: FLAN6ED D BUSTSlAl f~|
0«OP lOTTOMn rLARED R TUIULAR R
JAC«TEO D ClD CS D OTHEUD
COVER: FLANGED D SENI-FLANCED O
OUST SEAL D SNAP-ON D IOLTED D
SCREW: THICKNESS
HELICOIOQ SECTIONAL n *H»ON D
CUT FLISHTS n „,,,« MDDLES f]
«D "D OTMERQ
MTCN: STO.D LONG D *M°«T D
DOUBLE FLIGHT P] TAPERED fl
BEARINGS: JQLLtR D JALL D SLIEVE D
rWTERlALS OF CONSTRUCTION
47
4R
49
SO
51
s;
HANGER IRG.
THRUST IRG. END 1RG.
COUPLINGS IOLTS
END PLATES
WIVE PREFERRED
51
*,t
55
(
7
t
SCKEK CONV. IEOUCER D
FLOOR MOUNTED HEOUCER D
SHAFT MOUNTEO REDUCER Q
»-»ELTQ CHAIN D VARIABLE G
«UARO: TESQ 10 Q
NOTES:
5.4-59
-------
Table 5.4-31. EPRI LIME FGD SYSTEM DATA BOOK
POND WATER RETURN PUMP SPECIFICATIONS
COMPANY
LOCATION
EQUIPMENT N0._
SUPPLIER
FOR USE ON
TOTAL NO. REQUIRED
P.O. NO.
PRICE EACH $
GENERAL INFORMATION
TYPE
DUTY: CONTINUOUS INTERMITTENT
SERVICE
PROCESS INFORMATION
LIQUID:
DESIGN FLOW: NORMAL MAX GPM
PUMPING TEMPERATURE °F
SP. GR. P PUMPING TEMPERATURE
VISCOSITY 0 PUMPING TEMPERATURE
VAPOR PRESS. P PUMP TEMP. (FT. LID.)
PH VALUE
CORROSIVE MATERIAL
SOLIDS (MAX. DIA.)
HYDRAULIC INFORMATION FT. LIQ.
SUCTION PRESS. ABOVE LIQ. (ABS ) ( + )
STATIC SUCTION LIFT (-): HEAD ( + )
SUCTION FRICTION HEAD (-)
TOTAL SUCTION HEAD (17+18+19)
STATIC DISCHARGE HEAD
DISCHARGE FRICTION HEAD
DISCHARGE PRESS. ABOVE LIO. (ABS )
TOTAL DISCHARGE HEAD (21+22+23)
TOTAL DYNAMIC HEAD (24-20)
NPSH AVAILABLE (20-13)
NPSH REQUIRED
PUMP
MANUFACTURER
RPM
PERFORMANCE CURVE
SERIAL NO.
BPH P SERVICE CONDITIONS
0 MAX. FLOW FOR IMPELLER .
ROTATION @ DR1VF SHAFT END
COPIES REQUIRED OF:
PERFORMANCE CURVES
DIMENSION DRWGS.
OPFRATINf, AND MATNTFNANfF INSTRUCTIONS
MATERIALS
MATERIAL CODE - EXTERNAL CASING INTFBNAL PARTS
I - CAST IRON INTERNALS CODE I B S C
B - BRONZE IMPELLER I B S C
S - STEEL INNER CASE PARTS I B S C
C - ll-13t CHROME SLEEVE (PACKED) Ch Ch Af Af
A - ALLOY SLEEVE (SEAL) C C C C
h - HARDENED WEAR RINGS B Ch Ch
f - FACED SHAFT S S S S
LANTERN RING
PACKING GLAND
SUCTION CONN: SI2E POSITION
DISCHARGE CONN: SI7F POSITION
CONN. RATING TYPE
PACKING TYPF
1 ANTFRN RTNGS MATFRIAI
COOLING .
BFARINGS: TYPF GREASE OR
IMPFLIFR: TYPE SIZE FUR. MAX
VFNT CONN: DRAIN CONN.
F1USHING CONNECTION:
DRIVER
FURNISHED WITH PUMP BY OTHERS
TYPE:
FRAME:
MANUFACTURER:
run nsHRF
VOLTS PHASE CYCLC
HP RPM
pFMHNfiS IMBRICATION
COUPLING GUARD ._
NOTES:
5.4-60
-------
5.4.34 pH Sensors and Controller
P5 «alue °f * samPle solution is proportional to the
*r difference between a sensing electrode and a reference
electrode that contains a small chemical battery with a liquid
salt bridge conductor. when the pH probe is immersed in the
sample solution, the electric impulse from the potential differ-
ence is amplified by a high impedance circuit. This signal is
used as the input to a controller that changes the lime feed
ITci L.G •
controllers are either pneumatic or electronic; the latter
has less lag time, hence the operation is slightly faster. The
controller should be a three-mode instrument incorporating
proportional integral, and derivative modes. A standard con-
troller is linear and the neutralization of an alkali is a
process' To obtain a dependable pH reading, multiple
1 Ofi n°1llnear controllers should be used. A non-
™-i T* recently been developed specifically for
™« & T Permits small variations in pH value with little
JnSL^ ^lmS /need rate and P^vides proportionately larger
changes in lime flow when PH exceeds present limits.
*nH J^ /Oll°wing /actors should be considered in the design
and specification of a pH sensor and controller:
1. Probe location and maintenance
2 . Probe type
3 . Nonlinear control
5.4.34.1 Probe Location —
In many existing systems, improper design limits access to
the pH sensors, which causes inadequate service. The pH sensors
in lime scrubber service need frequent cleaning, because depos-
* °n e electrode- This deposit insulates the
^A Preve»ts development of an accurate electrode
-Q A, P^Perly designed electrode station should have
r- -,/°>f PH SenS°r mai^enance. If possible, each PH
bench a cabinet^ * "a^e™»c* stati°n equipped with a work-
™' * " 16t t0 hold
H • t0 hold sPare Parts' small tools, and stan-
dardizing solutions.
5.4.34.2 Probe Type —
There are two types of pH sensors currently in use in lime
scrubbing applications: dip-type and flow through
the
each S1de for easy maintenance. Care
5.4-61
must be taken to ensure
-------
adequate flow through the slipstream so that accurate readings
can be obtained. At each probe location, pH sensors should be
installed parallel for good maintenance and consistent readings.
All amplifiers and calibration controls should be installed at
the electrode station to permit one man to perform maintenance
and adjustments.
The new pressurized "nonflowing" reference electrode, made
of unbreakable plastic, is better suited for lime scrubber
application than the older "flowing type" electrode of fragile
glass construction. Miniaturized electronic packages that allow
easy service and replacement are now available. In every probe,
the wiring between the electrodes and the preamplifier should be
as short as possible to prevent short-circuiting. In some
models, the preamplifier is mounted in the electrode housing.
Most of these instruments have either voltage or current output
signals that are adjustable for both range and span.
To remove scale formation, electrodes can be equipped with
ultrasonic cleaning devices. These devices are effective in
removing mild scale deposits.
5.4.34.3 Nonlinear Control—
With the advent of nonlinear controllers for pH sensors, pH
control has been less of a "hit or miss" situation. If pos-
sible, nonlinear controllers should be used. Table 5.4-32 is a
specification for a pH control system to be completed by the
utility, architect/engineer, or equipment supplier for the pH
sensor vendor.
5.4.35 Level Controllers
Controls in a lime scrubber serve two functions: to feed
makeup water into the system, and to control the levels in the
various hold tanks. Dependability, not accuracy, is the goal
for level control in the lime scrubbing system. The system
should be supported by high- and low-level signals for notifica-
tion of malfunction. Several controllers are available. Con-
trollers using sensors that operate on the pneumatic bubble
tube, mechanical flow, and ultrasonic principles are not suit-
able for lime scrubber applications because of the abrasive and
chemical nature of the slurry. Displacement controllers and
flange-mounted differential controllers are best suited for lime
scrubber applications. Capacity level controllers are also
effective.
5.4.35.1 Displacement Controllers—
This instrument is based on the principle that the weight
of air in a cylinder containing liquid is proportional to the
volume of liquid displaced by the submerged portion of the
cylinder as the liquid falls and rises. An internal displace-
ment transmitter is suitable when operating in an open-top tank,
5.4-62
-------
Table 5.4-32. EPRI LIME FGD SYSTEM DATA BOOK
pH INSTRUMENTS SPECIFICATIONS
CHECKED BY DATE
COMPUTED BY DATE
COMPANY LOf.ATTnil
EQUIPMENT NO. FOR ,,SE QN
SUPPLIER P.O. Nn
SPEC. NO.
PROJ. NO.
TOTAL NO. REQ'D.
PRICE EACH $
2
2
t
8
.0
3
1 RECORDER D
CONTROLLER "
CASE: CIRCULAF
CASE COLOR:
GENER
INDICAT
rn
OTHER
BLACK
CHART SIZE: 12" CIRC.
SCALE RANGE:
CHART DRIVE: SPRING n
"Fun — nnr- — P""l n
EXP PRF. D V
CHART SPEED
MODEL NO.
[TYPE": PNEL
OTHE
MODE PROP.D
ON-OFFH
OUTPUT: 3-15 P
ON MEASUREMENT
OUTPUT INCREASE
ELEC. SW. TYPE
CONTACTS OPENC
CONTACT RATING:
c
AL
OR
n —
TRA
REC
SLRFACE"
IND.
OTHE
NUMBER
TYPE
ELECTRIC
BLIND 11
NSMITTERI
TANGULARC
YOKE
Rin
U PNEUM.T
AIR PR.
WIND
CONTROL
MATICg
RFSFin
sirr
INCREA5E:
5
OTHER
OTHER
ELECTRICT
Rflipr 'i
DECREASES
- ON MEASUREMENT
AMPS
AUTO-MANUAL
5
6
NU. HUblllUNi
it!
7
If
"0
1
3
MANUAL :
AUTO-SET:
BANK:
CLASS:
IMPEDANCE
3|pH-KANUt
OTES:
CLOS
INCREASE:
• '
VOLTS
"SWITCH
INTERNAL
INTEGRAL
-3OINT ADJUSTMEia
INTERNALL
PNEUMATIC
•IXED
EXTERNAL'
S
EXTERNAL
ELECTRIC
ADJUSTABLE
n
H4
HS
~4T
43
44
47
sn
bl
IPLAIN E
LENGTH:
INSERT1
DIA. (I
MATERI/
FLANGE
BUSHING
ISANITAF
LENGTH:
pH
LECTRODE [
ON LGTH.
N.):
ELE
: 3
Y:
C
CIROD
IHD.l
/4"
3A b
AP1LL
TYPE: ARMOREDT
MATERI*
METER
J REFERENCE
(IN.):
E CONNECTIONS
J CLAMP [I
OTHER
ARY TUBING
PI
_E_LECTR_ODED__
OTHER
ATNP
L: CAPILLARY
ARMOR
[CONN. AT CASE: "
INSTR
MATERIAL: 304 S
CONSTRUCTIO
WELL: 3/4"
FLANGE:
^:
BACKD
JMENT CASE
.S.D 316 S.S.
U EXTERNAL D
BOTTOM D
_ OTHER
JLNTERN_A_LD_
MODEL NO.
ACCESSORIES
S?
S'l
S4
55
Sfi'
sr
SR
s
-------
and it would be useful in such instruments as the thickener if
the reaction tank were separate from the lime slurry storage
tank. The instrument should be mounted from the top of the
tank, and the cylinder suspended in a well or behind a baffle so
that agitation in the tank does not affect the accuracy of the
level controller. It gives dependable, trouble-free service if
the slurry does not impinge on the instrument parts above the
liquid level. Deposits in the tube or the air cylinder should
be removed periodically.
5.4.35.2 Differential Controllers—
For closed-vessel applications, e.g., in a scrubber vessel,
a differential pressure (DP) transmitter is best suited. This
instrument measures the force necessary to hold a flexible metal
diaphragm in a fixed position when one side is exposed to the
liquid pressure below the liquid surface. Flange-mounted DP
transmitters are suitable for scrubber application. The dia-
phragm is mounted in the end of a 4-in. stainless steel cylinder
that extends through a nozzle into the chamber where the level
is. A plastic-coated diaphragm should be used in an abrasive
slurry. In such installations, the scrubber has to be down in
order to carry out maintenance on the DP transmitters, because
they have no shutoff valves. However, a properly designed DP
transmitter requires little maintenance. A good installation
should have the following features: a nozzle located away from
agitation to minimize scouring of the diaphragm; and a small-
diameter pipe connected to the instrument via a vessel tap that
is located well above the minimum liquid level, in order to
balance the static pressure of the DP cell. The most common
problem of DP transmitters is that the balancing line is either
plugged with solids, or the line is filled with gas as a result
of evaporation. For proper operation, the balancing line
should be installed with a rotometer to purge the line with
water, and a valve tap should be installed in the tank near and
at the same elevation as the transmitter.
5.4.35.3 Capacitance Controllers--
For closed-tank application, capacitance controllers are
effective though expensive. These instruments use an insulated
electrical probe that measures the capacitance between the
electrode and the grounded vessel. They are simply constructed,
have no moving parts, and are accurate. The accuracy of the
readings is unaffected by slurry deposits on an electrode or by
the presence of scrubber agitations. The electrodes should not
be placed at the axis of the tank.
Table 5.4-33 illustrates a typical specification for a
level sensor as it would be prepared by a utility engineer for
equipment vendors.
5.4-64
-------
Table 5.4-33. EPRI LIME FGD SYSTEM DATA BOOK
LIQUID LEVEL INDICATOR SPECIFICATIONS
tBCCltl IT MT[
COMPUTED BY Pirn
CWPAHY .nriiin.
fOUIPMENl »0. tn. ,,SE ON
SUPPLIER , n „
PROJ KO
TOTAL §0. IEO'0.
»lltr EACH 1
1
?
3
4
5
«
7
8
9
10
n
1Z
13
14
IS
16
17
18
19
ZO
n
zz
23
24
n
It
27
28
71
10
11
12
iOT
GENERAL
LIQUID LEVEL RANCt :
TIH: ISOLATING DIAPHRAGM Q DIP TUBE
OTHER D
NFR. 1 MODEL:
TANK
LOCATION: INDOORS C OUTDOOR C
CONNECTIONS: TOP SIDE £
MELDED C BOLTED C
INTERNAL PRESS. VENTED TO ATM. C PSI
MATERIAL MET IT PROCESS:
SERVICE CONDITIONS
FLUID:
NORK. PRESS. (M1N/NORM/MAI):
MORI;. TEMP. (MIN/NORM/MAX):
S.G. » IORM. TEMP.
INDICATOR
TTPE: «LL HAND. Q U TUBE D SAGE C
OTHER C
SIZE:
INDICATING FLUID - TTPE t S.t.
CONNECTIONS: VENTED TANK d»ON. VENTED [~
LOCATION: LOCAl _ SIDE Q TOP C
• EMOTED SURFACE D FLUSH Q
INDOORS OUTDOORS
ANIIENT TEMP. RARGE:
DISTANCE FROM TANK:
SCALE UNITS:
SCALE/PRESSURE RANGE:
SMALLEST SCALE DIVISION-
EMPHASIS/NUMERALS EACH: SCALE DIV
CASE TTPt:
CASE MATERIAL:
ES;
11
14
n
if
17
38
39
40
41
4?
43
44
4S
4*
47
48
49
sn
51
1?
S3
54
s1;
56
S7
SB
59
to
n
(2
(3
(4
ACTUATION
SOURCE: IKST. »uD PLART A1RQ HARD PUMP Q
OTHER U
PRESSURE REQ'D.
FLOM MODE: CONTINUOUS H DURING READING H
FLOW RATE:
FLOK CO»TROL:VALVE[nDIF. RELAY (""iNONE RtQ'D. H
DIP TUBE
OVERALL LENGTH:
DISTANCE-LOWEST OPENING TO TAN I 80T .
INACTIVE LENGTH AT LOVER END:
SUE:
MATERIAL:
ISOLATING D1APHRAG",
DO. PER INDICATING POINT:
MTG. LOCATION: TOP D IOTTOMQ SIDE [I
TANK CONNECTION-TYPE I SUE:
CONNECTION MATH.
DIAPHRAGM MAT'L.
ACCESSORIES
FLOM INDICATOR:
HAND PUMP:
CHECK VALVE:
INDICATOR MTG. BRACKETS-
CONNECTION TUBING:
TANK FITTING:
PRESS. FITTING:
FILTER REGULATOR:
DIFFERENTIAL RELAY:
PACKING NUT 1 GLAND:
"
5.4-65
-------
5.4.36 Flowmeters
In a lime scrubber, flow rates are usually measured on
several streams. Lime feed flow rates are measured to allow pH
control. Flow rate of slurry to the thickener is measured to
improve the operation of the thickener. Several types of flow-
meters can be used to measure flow. These include target
meters, rotometers, or purge differential-pressure transmitters
with venturi tubes or flow nozzles. These mechanical flowmeters
are not suitable for the abrasive slurry encountered in lime
scrubbers. The most acceptable devices to measure flow rates in
lime scrubbers are electromagnetic flowmeters and nuclear den-
sity gauges. They have no operating parts in contact with the
fluid, produce very little pressure drop, and are fairly accur-
ate .
5.4.36.1 Magnetic Flowmeters—
These devices use electromagnetic induction to produce an
AC voltage signal that is directly proportional to the flow rate
of the liquid through the piping. They can normally be cali-
brated from zero to full range. This output signal is converted
by signal converter for hookup to various readout and indicating
instruments. A common arrangement is a panel-mounted indicating
transmitter coupled with a strip chart recorder. An electronic
indicating controller should be used in cases where the fluid
must be controlled. To avoid constrictive flow and short meter
life, flowmeters should be the same size as the pipe on which
they are measuring the flow. Many materials are available for
liners and electrodes. Polyurethane, neoprene, and synthetic
rubber liners are best suited for scrubber applications. Stain-
less steel is the best material for electrodes. Some vendors
offer field-replaceable electrodes only in polyurethane, neo-
prene, rubber, and lined meters. These electrodes can be re-
placed without the meter being taken out of the process stream.
Among the accessories are ultrasonic electrode cleaners. A
metering installation should have a vertically mounted flowmeter
with upward fluid flow. The electrodes should be horizontal.
The pipe should be full when the slurry is measured. To obtain
accurate measurements, there should, if possible, be no air in
the fluid stream. Good quality magnetic flowmeters give
trouble-free performance; however, when selecting the vendor,
the availability of startup personnel to inspect and calibrate
the meter should be considered. Magnetic flowmeters come in
sizes from 1/10 in. to 6 ft in diameter.
5.4.36.2 Nuclear Density Sensors—
Nuclear density meters use a nuclear source to emit gamma
rays through a flowing stream. The density is proportional to
the number and intensity of the gamma rays that pass through the
fluid and the surrounding pipe. The main advantage of the
nuclear density gauge is its ease of application. It can be
5.4-66
-------
applied directly to existing process piping without inserting
anything into the pipe line. There is no wear on a liner, as
happens in a nuclear density cell. Table 5.4-34 illustrates a
typical flowmeter specification as it would be prepared by a
utility engineer for the instrument vendor.
5.4.37 S0? Analyzers
Continuous monitoring of gaseous emissions can be performed
by extractive or in situ analyzers. This section discusses both
types, their advantages and disadvantages. A continuous moni-
toring system has three subsystems: sampling, analyzer, and
data logging devices.
5.4.37.1 Extractive Analyzers—
These instruments have undergone greater development than
the in situ class, since they apply a standard laboratory method
for the analysis. The most common extractive analyzers use
nondispersible infrared and ultraviolet absorption, chemilumi-
nescence, fluorescence, and electrochemical techniques.
Several contaminants in the flue gas will interfere with
the analytical method. Hence, the sample has to be processed by
removing particulate matter, water vapor, and other contaminant
gases. Filters, refrigerators, and pumps are used to condition
the sample before the analysis. The following factors should be
considered when specifying an extractive SO2 analyzer.
1. Probe location and type
2. Sample line maintenance
5.4.37.2 Probe Locations—
The simplest method of extracting a sample with a repre-
sentative concentration is to insert a probe in the source at a
point where SO2 is present. Compositional stratification,
however, can lead to single-point measurement errors up to 15
percent. Therefore, the adequacy of single-point measurement
should be verified. A multiple sample probe works by drawing
equal volumes of samples through each port, mixing them within
the probe body, and delivering an integrated gas sample through
a single port.
5.4.37.3 Sample Line Maintenance—
The primary consideration in constructing a probe is par-
ticulate control. A filter at the probe inlet (perhaps 30 mesh)
is typically used to remove the larger particulate matter A
moisture control system is in-line immediately preceding the
sample chamber. Blowback air is used to clean the lines of any
particulate matter that is deposited. Pumps, valves, and sam-
pling lines should be made of 316 stainless steel or Teflon
Pumps are often the weak link of the sampling chain, developing
5.4-67
-------
Table 5.4-34. EPRI LIME FGD SYSTEM DATA BOOK
FLOWMETER SPECIFICATIONS
CHECKED BY _
COMPUTED BY
COMPANY
DATE
DATE
SPEC. NO.
PROJ. NO.
LOCATION
EQUIPMENT NO.
SUPPLIER
FOR USE ON
P.O. NO.
TOTAL NO. REQ'D.
PRICE EACH $
GENERAL
1
2
3
4
ITEM NUMBER
MODEL NUMBER
.OCATION
TYPE INSTRUMENT
SERVICE CONDITIONS
~T
6
7
8
9
In
n
12
FLUID
CONDITION
MAX. FLOW/NORMAL FLOW
OPERATING PRESS. PSIG
OPERATING TEMP. °F.
SP. GR. AT 60°F_LyAT FLOW COND.
VISCOSITY AT FLOW COND.
CLARITY OF FLUID
METER
TT
14
15
16
17
1R
11
?n
?i
??
n
?4
^
?6
?7
?R
?q
in
?i
3?
-------
leaks or becoming plugged. Diaphragm and metal bellows pumps
have been used successfully upstream from the analyzers, whereas
water or air aspirators have been used on the downstream side
5.4.37.4 In Situ Analyzers--
In situ monitors have been specifically designed to over-
come many of the problems encountered in extractive analyzers
These instruments use electro-optical techniques based on infra-
red or ultraviolet absorption. The monitors are placed across a
^f*; °£ have a.P5obe placed in a stack, and perform the analy-
S L iT gaS "t^ any samPle modification. The instruments
on nnJ inHC°nS1 °J( ^^ * long-slotted probe with a mirror
on one end, or a reflector and analyzer placed on opposite sides
SLJ?6 /tack. Air curtains are used to prevent particulate
matter from covering the instrument mirrors or windows located
in the stack.
in situ analyzers have the advantage of continually scan-
?™?v « " Cr°SS section of ^ck gases and thus reading a
£™?r* if**96 Concentration. m addition, in situ monitors
*?S£n % Sam£^e extraction °r conditioning system and thus
eliminate possible sample interactions.
m°nitors' ho"ever, have some basic design disadvan-
1. They are limited to the monitoring of only one stack
at a time whereas extractive systems may draw samples
from a number of sites.
2. In situ monitors are limited in their location and are
otten exposed to extremes in weather conditions and to
harsh environments. They may be difficult to reach
for repairs.
•„ c-Jab^ 5'4"35 is a tvPical data sheet for an extractive or
in situ SO2 analyzer.
5.4.38 Pressure Sensors and Controller
Three types of devices may be used to measure process
pressure: 1) manometers, 2) electrical elements, and 3) elec-
trical sensing devices. m a lime FGD system, pressure gauges
are commonly used to measure pressure drop across the scrubber
absorber and mist eliminator. t-r°ss i:ne scrubber
5'4"36 ** * *™ical data shee* for a pressure control
5.4-69
-------
Table 5.4-35. EPRI LIME FGD SYSTEM DATA BOOK
S02 ANALYZER SPECIFICATIONS
CHECKED BY
COMPUTED BY
COMPANY
EQUIPMENT NO.
SUPPLIER
DATE
DATE
LOCATION
FOR USE ON
P.O. NO.
SPEC. NO.
PROJ. NO.
TOTAL NO. REQ'D.
PRICE EACH $
'1
£
J
Ti
5
6
7
8
9
10
11
TT
13
4
TF"
15
16
IV
18
•19
2IT
GENERAL INFORMATION —
LOCATION
GAb
TYPICAL GAS COMPOSITION^):
EXTRACTIVE ANAL
PROBE: TYPE
MAIL. :
HLTER: TYPE
MAIL. :
CONDENSER: TYPE
MAIL. :
FINE FILTER: TYPE
MAIL. :
PUMP: TYPE
MATL. :
WIPLE LINE: SIZE
AUTO BLOWBACK: YES
ANALYSIS PRINCIPLES:
PROCESS INFORMATION
STATIC PRESSURE
ACFM, MAX. AT
502, 0?. CO*.
YZER
SIZE
SIZE
SIZE
SIZE
SIZE
MATL.
NO
•
IN. WG.
*F
NO*.
IN SITU ANALYZER
PROBE : TYPE
MATL.:
MIRROR: TYPE
MATL.:
BLOWER: TYPE
MATL.:
FILTER: TYPE
SIZE
SIZE
RATING
SIZE
MATL.:
GRATING: TYPE
IMATL.:
SAMPLE LINE: SIZE
DETECTOR: YES 3
ANALYSIS PRINC PLES:
SIZE
MATL.
NOC
5.4-70
-------
Table 5.4-36. EPRI LIME FGD SYSTEM DATA BOOK
PRESSURE SENSORS SPECIFICATIONS
me w.
IOCMION
TOTAL NO. tEQ'O.
MICE (ACM t
1
t
1
4
i
(
7
1
1
10
11
It
11
14
11
11
17
II
II
10
tl
It
It
14
GENERAL INFORMATION
BESCRIPTIOH RECORDER Q IRDICATDRQ ILINDQ
CONTROLLER QTHANSKITTER Q
CASE RECTANtULARQ CIRCULAR Q
OTHER
CASE COLON (LACK Q OTHER
NOUNTINS FLUSH Q SURFACE Q YOKE R
10. PTS. (ECOKDINC IIDICAT1MC
CHART Yr 11* CUC.Q OTHER
CHART RANGE NUMIER
SCALE RANCE TTPE
CHAUT DRIVE »MIK6 Q ELECTRIC Q "EU.n
CHART SPEED MIND (DATS)
»OLTS/CTCLCS £». MF. O
AIR HESS.
TRANSMITTER
tin MEUMATIC Q ELECTRIC Q
OUTPUT 1-15 HI Q OTHERS
RECEIVERS ON SHEET DO.
CONTROL
tin PREUNAT1C Q ELECTRIC Q
OTHER
HOP i AUTO-RESET QRATE ACTIORQ OR-OFF n
OTHER
OUTPU I- 11 *SI Q OTHER
OR MEASUREMENT INCREASE:
OUTPUT INCREASES Q DECREASES fj
/WTO MANUAL SWITCH
10. MIITIOHS EITERNALQlXTERNALn
INTERNAL Q
SETPOIHT ADJUSTMENTS
MANUAL INTERNAL Q EITERNALQ
AUTO-IET fNEUMATICQ ELECTRICfJ
•AM: FiuoQ AOJUSTAILEQ
OTHERS
PRESSURE ELEWEKT
5
6
7
1
9
30
11
It
1)
14
It
1C
17
11
SPIRAL Q lELLOdSQ iOURDON Q
OIAPHIASMQ HELICAL D
OTHER
MATERIAL
IRONZE D STA1NLESSD STEEL D
OTHER
AISOLUTE PRESS. COMPENSATION
STATIC HEAD COMPENSATION
MEAD
RANSE
PSICQ IN.Mt.YAC. D PSIAO
OTHER
CONNECTION-NPT 1/4' [^ '/I'D
lACRQ tOTTON D OTHER
ACCESSORIES
FILTER 1 IESULATOR
AIR SUPPLY IAUCE
LOCAL INDICATOR
CHARTS 1 INKSET
MOUNTINC VOICE
PULSATION DAMPENER
SYPHON
ALARM SVITCH
HERMETICALLY SEALED D E.P D I.P.D
OPERATING CONDITIONS
PRESSURE. NORMAL MX.
TENPERATUIE NORMAL IUI.
FLUID
SEAL FLUID t.l. • »0«F.
HOTES
5.4-71
-------
5.4.39 Temperature Sensors and Controller
The most commonly used temperature measuring devices are
thermocouples, resistance thermometers, liquid-in-glass thermom-
eters, and pyrometers. In a lime FGD system, temperature sen-
sors are installed at several points, including the scrubber
inlet, absorber outlet, slakers, and recirculation tank.
Table 5.4-37 is a typical data sheet for an electrical
temperature control system.
5.4.40 Control Valves
Control valves have been described in Section 3.4.3 (Con-
trol Modes). Butterfly or ball valves should not be used to
control the flow of lime feed slurry, since they are susceptible
to frequent plugging and jamming. The wedge should not be
rubber coated because it can erode rapidly. Pinch valves are
generally suitable for slurry flow control. This is a major
problem area where considerable research is needed.
Table 5.4-38 is a typical data sheet for a control valve.
5.4-72
-------
Table 5.4-37. EPRI LIME FGD SYSTEM. DATA BOOK.
TEMPERATURE SENSORS AND CONTROLLERS SPECIFICATIONS
ewtifc ir MU
COMPUTED BY BATE
COMPANY LOCATION
EQUIPMENT «0. rot usi Oil
SUPPLIER r o, NO
tree.
f«OJ.
TOTAL NO.
MICE
•0
10.
•EQ'O.
EACH $
1
?
3
4
5
6
7
e
9
10
11
12
13
14
15
16
17
18
19
10
21
22
23
24
25
26
V
it
29
30
31
6ENFRAL
OtSCRIPTjOH: RECORDER D INDHA nl. t~~]
CON1ROLLERD TRANSl-imuD CTMSP £1
MFR. 1 MODEL:
CASE. RECTANGULARQ CIRCULAR d
COLOR: CASE DIAL
KOUNT1HO: FLUSHQ- SoRIACE D
KO. Of POINTS;
INDICATOR: AHALOcD DIOlTAi
MORIZ. vtp? D
CHAR' TYPE: IK. STRIPD IN. CIHCLf D
CHART RANGE: dUHBER-
SCALE RANGE: TYPE-
CHART SPEED: REV. /DAY IN /HR
fE« SPEED: SECONDS fULl SCALF THAVtt
PRINT SPEED: SECONDS PER POlin
BALANCING: P1ANUALQ AUTOMATIC D
STAHDARDIZATIOK: HANUAL D AUTOMATIC
CHART DRIVE: »OLTS CYCLES
INPUT IMPEDANCE: OHKS KIN. OHMS AT IAL
INSTRUMENT ERROR UNIT :
LOCATION: INSIDE „ OUTSIOE HAZARDOUS C
AMBIENT TEMP. RANGE:
SENSING ELET1ENT
FORM: THERMOCOUPLE D RESISTANCE
OTHERd
MATERIAL OR CALIBRATION:
REF. JUNCTION COMPENSATION' TES NO
STD. LIMIT OF ERROR:
SENSOR IS: ISOLATCOQ (ROUNDED D
CONTROL
3?
11
34
1[
IS
37
3B
39
TYPE: PNEUMATIcD ELECTRIC D
OUTPUT:
MODE: PROPORTIONAL D RESETD tATE D
ON-OFrD
ON MEASUREMENT INCREASE -
OUTPUTMNCREASESG BECtfASES (H
CONTACTS: OPENQ CLOSE D
CONTACT RATING: »MPS.» »OLTS
1 SET POINT ADJUSTJOT
40
41
42
43
MANUAL: 1NTERNALD PlTfl.li Q_
AUTO-SET:PNEUMATIcQ ELECTRICAL D
B»ND: FIUOD *OJUST«Ba D
ACCESSORIES
44
45
46
47
48
49
$0
SI
s?
S3
S4
ss
56
AUTO MANUAL SWITCH: POSITIONS
1IITERNALD tITtRNALD l»TrC»»l D
ALARM CONTACTS:
CHARTS 1 IIUSET-
FILTER t RESULATOR:
AIR SUPPLY CAtE:
LOCAL INDICATOR:
NOTES:
IEPIARKS: ~~~ ~ ~
' "
5.4-73
-------
Table 5.4-38. EPRI LIME FGD SYSTEM DATA BOOK
CONTROL VALVE SPECIFICATIONS
OKCUt »Y »«Tt
CONPUTED l» BAT!
•rtc. M.
»«OJ. 10.
LOCATID*
ro, ust 0. W»l iO. IEO'0.
r.o. 10. »«icc i»c« t
GENERAL
I
?
}
4
t
7
ft
9
10
11
1?
13
14
IS
16
17
1"
19
?0
?1
22
DESCRIPTION MESS.DtiW- U HOK C
ItVcL D OTHtll I]
TYPE 6LO«t Q «UTTt*rU I] I»U D
OTHCIl D
MFK. 1 MODEL:
VALVE BODY
FORM. STRAIGHT O »«6L£ I]
OTHER D
sin .
TYPE: C*SlQ •»«STOC»D
OTHER LJ
EKDS: SCHEMED DrL»«6toD f L*««IESS Q
TT'E t SHE
PRESS/TEMP RATING: 'S" • *f
BONNET STD.D OTHERD
INNER VALVE
CHARACTERISTIC: EO. xQ L IKEARDoUICi: 0»E«D
OTHER D
PLUG FORM:
DO. OF POUTS 1 SIZE:
TRIh: fULL SIU D MSTKICHD Q
PLl>6 TRAVEL:
GUIDING: TOP t IOT.D TO*D WIT D
OTHER n
MATERIALS
23
?fi
27
2B
?»
»AL»[ »oor: »RO«:ED C*RB. su^Q ss I]
OTHER O
TRIH- PLUG 3U SS C] OTHER D
SEAT: )1( SS D OTMIR Q
SPRING: 31t SS D OWR [^1
PACHHG: TEFLON Q OTHtR Q
BIAPHBACH BUKA-liriOK [""I OTHER Q
WLVt AQUATOR
10
11
12
13
14
IS
36
17
11
M
40
41
42
41
44
4S
TtPE: MEUHATIC "^ IltCTIICn
OTHER O
ClAPHRAtN ^3 flSTOR CD
OTHtR 3
SIHAL LE»EL:
AIR TO: ClOSlD Of£»Q
FAILURE POS. «PEnQuHCHA«SEDO ClOSEoO
SERVICE CONDITIONS
FLUID
TEMP[*ATU*E: IOR" *F MAI. *F
IILET PRESSURE: IORM. »Slt. MI. Kit
HAI. PRESS. DROP THRU VALVE:
S.t.t » «'F 1 T.T.
VISCOSITY * F.T.
Hid. FLO* &P: C»:
H»>. FIOH: AP Cv:
HORN. FLOW: ^P: t»:
ACCESSORIES
4i
47
48
4*
10
SI
12
S)
POS1TIOIER
SOLEKOIO »AL»t
FILTER REGULATOR
LIMIT SWITCHES
LUK1CATOR G ISOLATOR VALVE D
•OTES-
REMARKS •
5.4-74
-------
5.5 BID EVALUATION
Economics is a key element in the evaluation of bids. The
utility or architect-engineering (A-E) firm that is deciding
among proposed systems will have to assess various capital and
rlwiart!!9 C°StTS; and ^ Wil1 fre<3^tly find it hard to make
comparisons. it is easy to imagine a situation where System A
needs a lower capital investment than System B, but System B
needs lower annual operating and maintenance (O&M) expenditures
than System A. In these cases, it is necessary to investigate
the overall economic impact of the system.
The utility will not actually make an initial expenditure
in the amount of the overall capital cost. Circumstances will
vary among utilities, but in general, funds in the amount of the
overall capital cost will be borrowed and paid back within the
course of 15 to 20 years.
The best economic analysis is probably one that compares
"present-worth" or "present-value." This methodology reduces
all future economic differences among the various systems to a
single equivalent present amount. To work properly, the method
relt ?5°n accurate forecasts of operating costs. Every
°
e1val,ua^ion team will have to factor the costs for plant
+ ?r throu9h the Projected years of operation.
nrc* ageS' 10ad factor wil1 decrease and operating costs
riJXno ^ CUrrent Collars) will, therefore, be reduced. For
details on the "present-worth" analysis, the reader may consult
an economic decision-making text, such as Process Plant Esti-
mating, Evaluation, and Control by K.M. Guthrie.
Tables 5.5-1 through 5.5-8 present "spread sheets" useful
itL^i^n119 ibldS f°r the following eight major equipment
items used in a lime slurry FGD system:
Slaker
Lime slurry pump (with minor modifications, this sheet
could be used for any pump in the system. The thick-
ener underflow pump and flocculant proportioning pump
sheets will need the most modifications )
S02 absorber
Mist eliminator
5.5-1
-------
Table 5.5-1. BID EVALUATION FOR SLAKERS
Characteristics
Ul
•
en
to
'Design Performance
Lime feed, Ib/h
Total water required,
gal/min
Residence time, min
Slaker size, ft
Lime slurry
Flow, gal/min
Solids, wt.%
PH
Temp., °F
N'-'Tiber of slakers
Agitator
Type/rpm
Size
Drive: type
Hotor
hp/rpm
Capital cost, $
Slakers
Agitator
Motor
Subtotal
Vendor 1
Vendor 2
(others as needed)
-------
Table 5.5-2. LIME SLURRY PUMP
Characteristics
Oi
•
tn
I
u>
Pump performance
Pump selection
Pump size
Efficiency, %
Pump speed, rpm
BHP 0 design
BMP installed
Material of construction
Lining
Casing
Impeller
Shaft
Hear rings
Cost, $
Pump
Coupling
Motor
Subtotal
Vendor 1
Vendor 2
(others as needed)
-------
Table 5.5-
3. BID EVALUATION FOR S02 ABSORBER
Characteristics
in
•
in
Design performance
Internal configuration
Size
No. of stages of absorption
Gas flow, °F
Inlet
Outlet
SO- removal efficiency, %
Pressure drop, in. WG
Turndown ratio
Max L/G, gal/1000 acf
Operating L/G
Materials of construction
Shell
Linear
Abrasion zones
Stagnant zones
Internals
Packing, if any
Nozzles
Capital cost, $
Absorber
Vendor 1
Vendor 2
(others as needed)
-------
tn
•
en
ui
Table 5.5-4. BID EVALUATION FOR MIST ELIMINATOR
Characteristics
Dasign performance
Gat flow, acfm 9 °F
Mitt
Inlet, gr/scf
Outlet, gr/scf
Gas velocity
Mlit eliminator
Type/shape
No. of passes/stages
Vane spacing, in.
Freeboard distance, ft
Material of construction
Mist eliminator
Hash water headers
Hash water spray nozzles
Wash water collectors
Wash water system pump
Hater flow, gal/min
Pressure, psig
Pump speed, rpm/bph
Design installed
Motor, hp/rpm
Capital cost, $
Mist eliminator
Wash water headers/
sprays/collectors
Wash water pump
Subtotal
Vendor 1
Vendor 2
(others as needed)
-------
Table 5.5-5. BID EVALUATION FOR REHEATER
Ul
i
Ul
(Ti
Characteristics
Design Performance
Reheat
«F
Btu/h
Direct/indirect
Gas flow, acfm 9 °F
SOj/moisture, ppm/vol.%
Fuel/heating medium flow, lb/h
•Overall heat
Transfer coefficient,
Bf-.u/h-ft2 "F
T'-.i.; bundle
No. of tubes
No. of runs
Support
Material of construction
Tubes
Baffles
Supports
Fuel heating medium pump
Speed, rpm/bhp, design
Drive: type/rpm installed
Motor
hp/rpm
Capital cost, $
Tube bundle
Pump
Subtotal
Vendor 1
Vendor 2
(others as needed)
-------
Table 5.5-6. BID EVALUATION FOR BOOSTER FAN
en
•
(SI
I
-vl
Characteristics
Design performance
Fan selection
Pan size
Fan hp, design
Fan hp, installed
Air flow through fan, acfm
Temperature, °F
Fan speed, rpm
Pressure increment, in. WG
Materials of construction
Housing
Blades
Bearing
Shaft
Capital Cost, $
Fan
Motor
Subtotal
Vendor 1
Vendor 2
(others as needed)
-------
Table 5.5-7. BID EVALUATION FOR RECIRCULATION TANK
ui
00
Characteristics
Design performance
Tank size, ft/shell thk, in.
Tank capacity, gal
Slurry height, ft
Lime slurry.
Residence time, min
Temperature, °F
PH
Operating pressure, psig
Wt. of absorber, tons
No. of recirculation tanks
Material of Construction
Shell/lining
Supports
Agitator
Baffler
Agitator
Type/rpm
Size/nos.
Drive types
Motor, hp/rpm
Capital cost, S
Recirculation tank
Supports
Agitator
Absorber
Subtotal
Vendor 1
Vendor 2
(others as needed)
-------
Table 5.5-8. BID EVALUATION FOR THICKENER
ui
•
en
Characteristics
Design performance
Settling, ton/ft2/day
Desired
Maximum
Size
Capacity, gal
Underflow, gal/min
% solids
Overflow, gal/min
Lifting mechanism
Rake type
Torque limit
Materials of construction
Tank shell
Rake coating
Rake metal
Support stanchion
Motor
Type
hp
Cost, $
Tank
Rake
Subtotal
Vendor 1
Vendor 2
(others as needed)
-------
0 Reheater
0 Booster fan
0 Absorber recirculation tank
0 Thickener
These sheets list major design information, materials of con-
struction, and costs submitted by the various bidders for each
item. This comparative strategy enables the utility to select
the best system.
To obtain useful estimates of operating costs, the utility
or A-E firm will have to provide the costs of items such as
electricity, water, fuel for reheat, and steam. The equipment
suppliers should also be provided with normal design operating
levels so that the operating costs will all refer to similar
load conditions.
It is also reasonable to ask for guidelines for maintenance
costs. These figures will probably be expressed in the form of
a percentage of the overall capital cost. The vendors should be
asked to document those projected maintenance charges.
5.5-10
-------
GLOSSARY
Following is a guide to the terminology, abbreviations and
tions used "ievj.rtuj.ons, ana
assumptions used in this text.
wlTH re*£^\ CUb±C leet Per minute'- a 9^8 flow rate, expressed
with respect to operating conditions (temperature and pressure).
~ the Deterioration of a material surface by mechanical
Process ^ which 5^5 molecules are transferred
aH^ v
a liquid phase by scrubbing.
a aa*™ *?* process bv which 9^3 molecules are removed from
a gas stream by means of adhesion to the surface of a solid.
°f hours an FGD svstem "as "acces-
sen
of actu? ™Pera tlor\to the h°urs in the time period (regardless
of actual operation time), expressed as a percentage.
temper^tn^^1 "n1t ~ the amount of heat required to raise the
temperature of one pound of water one degree Fahrenheit.
seEaration - the separation of phases in a composite
forcn« I aPPlying a circular motion to the stream and
forcing the higher density component to the outside wall of the
device, where it is collected.
c,ontinuous settling basin used in wastewater
' Producin9 a dear overflow and a concen-
tra n '
trated sludge from the bottom.
rmc»hK " ?irculation in the same direction of two
streams through a piece of equipment.
-T PKeCe °f e(3uiPment used to convert a vapor state to
- K
state by compression or extraction of heat.
G-l
-------
Counter-current flow - circulation in opposite directions of two
streams through a piece of equipment.
Cyclone - a piece of air pollution hardware used for particulate
removal by centrifugal separation.
ESP - electrostatic precipitator; an air pollution device used
to remove particulates from an exhaust stream by initially
charging them with an electrode and then collecting them on an
oppositely charged plate.
Efficiency - ratio of the amount of a pollutant removed to the
total amount introduced to the removal operation.
Entrainment - the suspension of liquid droplets or mist in a gas
stream.
FGD - flue gas desulfurization; the process by which sulfur is
removed from the combustion exhaust gas.
FD - forced draft; a fan, or blower, used to produce motion in
an enclosed stream of gases by creating a positive pressure in
the stream and pushing it.
Flooding - the situation in countercurrent gas-liquid operations
when gas velocity is high enough to impede the flow of liquid
through the tower; excessive entrainment.
Free space - voids within the packing material in a packed
L-OWGjT •
ID - induced draft; a fan used to move an enclosed stream of
gases by creating a negative pressure in the stream and pulling
J_ L- »
MW - megawatts; unit used to describe gross or net power genera-
tion of a particular facility. One watt equals one joule per
second. One megawatt equals 106 watts.
Mist - dispersion of relatively large liquid particles in a gas
stream; carryover from a gas-liquid contact operation.
Mist eliminator - a piece or section of pollution hardware used
to remove a dispersion of liquid particles from a gas stream.
Nm3/h - normal cubic meters per hour; unit of gas flow rate
under the standard condition of O°C and 760 mm Hg.
NSPS - New Source Performance Standards; environmental regula-
tions that apply to a new installation.
G-2
-------
Operability factor - ratio of hours an FGD system operated to
hours of boiler operation during a particular time period
expressed as a percentage.
parts per million; units of concentration; in wastewater
applications equal to milligrams per liter; in air pollution
applications equal to moles of pollutant to million moles di-
luent.
Packed-bed scrubber - a piece of pollution equipment using small
pieces' "ith high surface area-to^volume
transfer
v (usually cylindrical) used for pollutant
PaC*ed scrubber. Untreated gas enters the bottom
ward
n
UP ' and li<3Uid GnterS the tQP and ^ravels down-
SC,rubber - Pollution control equipment that
n ,
hich laniH ^ gaS throu9h h°les in a series of plates on
hv SMv££ ^ ' ,causin9 an intimate contact between phases
by breaking the gas flow into bubbles.
" ^he.fiffe^nce in force per unit area between
& stream, due to resistance to the flow of
n " a uuVi°e US6d to spread out th« flow of liquid
material ^rubber to insure uniform wetting of packing
rati° °f hours an FGD system operated to
that stream.11"
hnnr
hours
was called on to operate, expressed as a percentage.
°f time a unit volume of ^as or
H- ,
spends in a pollution control device.
f^oo"FS^nd,ardH°UbiC feet Per minute' ™its of a gas flow rate
at 60 °F and 1 atmosphere pressure.
Saturated - the situation when a gas or liquid is filled to
capacity with a certain substance. No additional amount of the
same substance can be added under the given set of conditions
Scrubber - a device used in the removal of pollutant gases from
exhaust streams of combustion or industrial processes.
scrubber train - a series of physical and/or chemical unit
operations that remove pollutants from exhaust streams, carried
out in a series of modules. <"-ij.eu
G-3
-------
Slagging - scaling of precipitate or buildup of particulate
material on equipment surfaces.
Spalling - the deterioration of stone, concrete, or ceramic
materials because of chemical or physical action.
Stabilization - the addition of a flocculating agent to a waste-
water to enhance the settling of solid materials.
Stabilization pond - a large excavation, usually manmade, for
the storage and settling of stabilized sludge.
Stacked packings - ceramic, plastic, or wood materials placed in
layers in a packed tower.
n*r?fand c™**-£™* ' a set of physical constants for the com-
parison of different gas volume flow rates. English = 60°F, 1
atmosphere pressure.
"*™ " m°lar ratio of "actants in a chemical
SXtent lime is added to the reaction
in exr, x
in excess of the theoretical amount required.
t™ t0KS Per year; units used to express amounts of
material,, H ** US?d °r that are generated (usually solid
materials such as coal or sludge).
ron^n^rv" a c°ntinuous settling basin used to increase solids
concentration from influent to underflow.
Underflow -concentrated solids flow from the bottom of a clari-
fier, scrubber, or thickener.
Utilization factor - ratio of hours an FGD system operated to
hours in the given period, expressed as a percentage.
Venturi scrubber - an air pollution device used to accelerate
concurrent liquid and gas flows for more turbulent and intimate
contact.
WC - water column; units of pressure expressed as inches of
water.
Weeping - the situation in an absorption tower when gas velocity
is too low and does not provide for intimate contact between
phases.
G-4
-------
TECHNICAL REPORT DATA
(f'lease read /nOfuctions on the reverse before completing)
EPA-600/8-79-009
4. TITLE AND SUBTITLE
Lime FGD Systems Data Book
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
April 1979
6. PERFORMING ORGANIZATION CODE
T.C.Ponder Jr. , J.S.Hartman, H.M.Drake.
> A-N- Patkar> R- D-Tems,
8. PERFORMING ORGANIZATION REPORT NO.
P/N 3283
PEDCo Environmental, Inc.
11499 Chester Road
Cincinnati, Ohio 45246
10. PROGRAM ELEMENT NO.
EHE624A
11. CONTRACT/GRANT NO. ~
68-02-2603, Tasks 5 and 35
1E AND ADORES
EPA, Office of Research and Development*
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; 12/76 - 12/78
14. SPONSORING AGENCY CODE
EPA/600/13
vi^w A T>TTEAil* /??s
------- |