FINAL TECHNOLOGICAL REPORT
                   on
   A SYSTEMS ANALYSIS  STUDY  OF
     THE INTEGRATED IRON AND
            STEEL  INDUSTRY
         (Contract No. PH 22-68-65)
                    to
DIVISION OF PROCESS CONTROL ENGINEERING
    NATIONAL AIR POLLUTION CONTROL
             ADMINISTRATION
    DEPARTMENT OF HEALTH, EDUCATION,
              AND WELFARE

    (Complementary Final Economic Report on a
    Cost Analyses of Air-Pollution  Controls in the
       Integrated Iron and Steel Industry is
           also dated May 15, 1969.)
               May 15, 1969
        BATTELLE MEMORIAL INSTITUTE
           Columbus Laboratories
             505 King Avenue
           Columbus, Ohio 43201

-------
            FINAL TECHNOLOGICAL REPORT
                            on
            A SYSTEM ANALYSIS STUDY OF
     THE INTEGRATED IRON AND STEEL INDUSTRY
               (Contract No.  PH 22-68-65)
                            to
     DIVISION OF PROCESS CONTROL ENGINEERING
 NATIONAL AIR POLLUTION CONTROL ADMINISTRATION
DEPARTMENT OF HEALTH,  EDUCATION, AND WELFARE
                      May 15,  1969
                            by
          J. Varga,  Jr. , Principal Investigator
          H. W. Lownie, Jr. , Project Director
           BATTELLE MEMORIAL INSTITUTE
                 Columbus Laboratories
                     505 King Avenue
                 Columbus, Ohio  43201

-------
Battelle Memorial Institute. COLUMBUS LABORATORIES
505 KING AVENUE COLUMBUS, OHIO 43201. AREA CODE 614, TELEPHONE 299-3151' CABLE ADDRESS: BATMIN
Ma y 2 9, 1 96 9
Mr. No rman Plaks
Division of Process Control Engineering
National Air Pollution Control
Administration
5710 Wooster Pike
Cincinnati, Ohio 45527
Dr. Paul Kenline
Division of Economic Effects Research
National Air Pollution Control
Administration
1055 Laidlaw Avenue
Cincinnati, Ohio 45237
Mr. N. G. Edmisten
Division of Economic Effects Research
National Air Pollution Control
Administration
411 W. Chapel Hill Street
Durham, North Carolina 27701
Gentlemen:
Final Technological Report on
A Systems Analysis Study of
The Integrated Iron and Steel Industry
(Contract No. PH 22-68-65)
Two copies of the subject report are being sent to both Mr. Edmisten and
""'r. Kenline, and 96 copies to Mr. Plaks.
For the companion Final Economic Report on Cost Analyses, two copies are being
sent to both Mr. Edmisten and Mr. Plaks, and 96 copies to Dr. Kenline.
Very truly yours,
~~\.
H. W. Lownie, Jr.
Project Director
HWL:jls

-------
T ABLE OF CONTENTS
SECTION I
INTRODUCTION AND SCOPE.
PURPOSES OF THIS STUDY.
SCOPE OF THIS STUDY
SECTION II
SUMMAR Y AND RECOMMENDATIONS
.tHE NATURE OF THE INTEGRATED IRON AND STEEL INDUSTRY
Industry Efforts for the Immediate Futur e
FUTURE PLANS OF THE INDUSTR Y .
AIR-POLLUTION-CONTROL RESEARCH AND DEVELOPMENT.
General Recommendations for Phase II Res earch on
Technical Aspects of Control of Air-Pollution Emissions
From the Integrated Iron and Steel Industry
Improved Control of Emissions From the
Manufacture of Metallurgical Coke
Instrumentation for In- Plant Emis sions Measur ement
and Control
Improved Characterization of Emissions From
Process Segments.
Other Subjects of Research Interest
Specific Recommendations for Phase II Research on
Technical Aspects of Control of Air -Pollution Emissions
From the Integrated Iron and Steel Industry
Project 1: Processes for Making Metallurgical Coke.
Project 2: Sulfur-Bearing Emissions from
Blast-Furnace Slag.
Project 3: Characterization of Emissions from Sintering Plants.
Project 4: Evaluation of Instrumentation for In-Plant
Emissions Measurement and ControL
Project 5: Development of Improved Flow Sheets and
Material Balances for American Integrated Steelworks
Project 6: Control of Wide-Area Emissions from
Storage Piles
Project 7: Design and Costing of Systems for Collection
and Control of Wide-Area Emis sions Inside a Steelworks.
BATTELLE MEMORIAL INSTITUTE ';"'COLUMBUS LABORATORIES
Page
I-I
1-1
1-2
II-I
II-I
II-2
II-4
II-5
II-5
II-6
II-7
II-8
II-9
II-II
II-II
II-l7
II-I8
II- 19
II-20
II-2I
II-2I

-------
TABLE OF CONTENTS
( Continued)
Page
SECTION III
THE INTEGRATED IRON AND STEEL INDUSTRY.
III- I
CHARACTERISTICS ON A NATIONAL SCALE
III - I
GEOGRAPHIC DISTRIBUTION OF THE INDUSTR Y
III- 3 2
Present Facilities
Current Modernization and Expansion.
Future Plans
III-32
III - 3 2
III -40
MAKING OF IRON AND STEEL.
III-40
Raw Materials.
Making Ir on
Making Steel
Manufacture of Semifinished Products.
Manufacture of Finished Products
1II-4l
III-42
III- 43
III - 44
III-44
REFERENCES FOR SECTION III
III-45
SECT ION IV
PROJECTIONS REGARDING STEEL-INDUSTRY GROWTH
AND CHANGES IN PROCESS TECHNOLOGY
IV-I
Anticipated Growth in Production and Consumption of Raw Steel.
Anticipated Growth in Production and Consumption of Pig Iron.
Anticipated Growth in Consumption of Iron Ore.
Anticipated Changes in Iron and Steel Proces s Technology.
Processing of Iron Ore at or Near the Source of the Ore.
Coke Ovens and Coke Production
The Ironmaking Blast Furnace
Electric-Furnace Pig Iron.
Direct-Reduction Processes.
Basic-Oxygen-Furnace (BOF) Steelmaking
Electric-Furnace Steelmaking
Mini Steel Plants.
Sponge Iron for Steelmaking
Open-Hearth Steelmaking.
Continuous Casting of Steel
Pressure Pouring. .
Vacuum Degassing
Rolling
Finishing
Heating Furnaces and Controls
IV-I
IV-5
IV-6
IV-8
IV-8
IV-9
IV-II
IV - 13
IV - 14
IV-IS
IV - 16
IV - 17
IV - 18
IV - 18
IV-19
IV-19
IV -20
IV -20
IV - 2 2
IV - 2 3
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
TABLE OF CONTENTS
. (Continued).
Explosive Bonding
Electrical-Energy Consumption and Generation.
SECTION V
SOURCES OF AIR POLLUTION INTH~ INTEGRATED
IRON AND STEEL INDUSTR Y .
Process Segments as Sources of Air Pollution.
Receipt, Storage, and Handling of Raw Materials.
Coking Process.
Coke-Oven Charging.
Coke -Oven Pushing.
Coke Quenching
Coke Handling.
By-Product Processing.
Preprocessing of Raw Materials.
Sintering
Pelletizing.
Ironmaking .
Charging.
Smelting.
Casting and Flushing
Pigging of Iron
Other Ironmaking Processes.
Steelmaking
Open Hearth
Basic Oxygen Furnace.
Electric Furnace.
Vacuum Degassing of Molten Steel.
Continuous Casting
Pressure Casting.
Steel Shaping
Soaking Pits and Primary Breakdown.
Conditioning, Reheating, and Hot Rolling.
Acid Pickling.
Cold Rolling and Cold Forming
Annealing
Steel Finishing
FUEL AND ENERGY UTILIZATION
10- Year Summary of Fuel Usage
Analysis by Applications
Firing of Coke Ovens
Sintering of Dusts and Ores .
Smelting of Iron Ores and Agglomerates.
BATTELLE MEMORIAL .INSTITUTE - COLUMBUS LABORATORIES
Page
. IV - 24
. IV - 2 5
V-I
V-I
V-I
V-2
V-2
V-3
V-3
V-4
V-4
V-5
V-5
V-6
V-6
V-7
V-7
V-8
V-9
V-9
V -10
V-I0
V-12
V-13
V-15
V-17
V-17
V-18
V -18
V -18
V -19
V -19
V -19
V -19
V-21
V-22
V-30
V -31
V-33
V-35

-------
TABLE OF CONTENTS
(Continued)
Steelmaking
Steam Raising.
Heating of Steel in Proces s
Principal Findings
Some Facets of the Relationship Between Clean Air
and the Fuel,and Energy Aspects of the Steel Industry.
SOURCES AND AMOUNTS OF NOXIOUS EMISSIONS
IN THE IRON AND STEEL INDUSTR Y
Sulfur in Iron and Steel Plant Processes.
Making Coke
Sintering Machine
Blast Furnaces
Open Hearth Steelmaking
Basic Oxygen Furnace Steelmaking.
Electric Furnace Steelmaking
Summary Balance of Sulfur for Iron and Steel Industry
CO Balances for the Iron and Steel Industry
Coke-Oven Gas
Blast Furnace Top Gas.
BOF Process.
Fluoride Emissions.
Nitrogen Oxide Emissions.
References.
SECTION VI
ANAL YSIS OF APPLIED CONTROL SYSTEMS.
Electrostatic Precipitators
Sinter - Plant Applications of Electro static Pr ecipitator s .
Blast-Furnace Applications of Eleqtrostatic Precipitators.
Open-Hearth Applications of Electrostatic Precipitator-s .
Basic-Oxygen-Furnace Applications of Electrostatic Precipitators
Electric-Furnace Applications of Electrostatic Precipitators.
Wet Scrubbers.
Application of Wet Scrubbers in Sinter Plants
Application of Wet Scrubbers to Blast Furnaces
Application of Wet Scrubbers to Open-Hearth Furnaces.
Application of Wet Scrubbers to Basic Oxygen Furnaces.
Application of Wet Scrubbers to Electric Furnaces
Fabric Filters
Application of Fabr ic Filter s to Sinter Plants
Application of Fabric Filter s to Open-Hearth Furnaces.
Application of Fabric Filters to Electric Furnaces
Cyclone Dust Collectors
BATTELLE MEMORIAL INSTITUTE -COLUMBUS LABORATORIES
Page
V-37
V-40
V-41
V-43
V-44
V-47
V-47
V-50
V-52
V-53
V-54
V-56
V-60
V-63
V-63
V-65
V-65
V-66
V-66
V-67
V-67
VI-l
VI-3
VI-7
VI-7
VI-S
VI-9
. VI- 12
. VI-12
. VI-15
. VI-16
. VI-IS
. VI-IS
. VI-20
. VI- 20
. VI-21
. VI-21
. VI-21
. VI-23

-------
TABLE OF CONTENTS
( Continued)
Cost of Applied Control Equipment.
Cost-Effectiveness of Applied Systems
Effect of Efficiency Specifications Greater Than Current
Legal Requirements on the Cost of Air-Pollution Control Equipment
Performance Equations.
Theoretical Factors Affecting Performance.
Control System Cost Changes.
Conclusions
Technological Factors Affecting Gas-Cleaner Performance.
Effect of Gas Volume Changes
Effect of Pressure Drop
Effect of Dust Loading.
Particulate Characteristics From Different Proces s Segments
Relationship of Particulate as Generated by Different
Processes to Collecting Efficiency
Effect of Temperature.
Effect of Humidity
Effect of Electrical Resistivity of Dust
Effect of Particle Size on Precipitator Efficiency.
Adaptability of Particulate Removal System
to Removal of Gas eous Pollutants.
REFERENCES FOR SECTION VI
SECTION VII
PROBLEMS AND ASSOCIATED OPPORTUNITIES FOR RESEARCH.
First Priority for Research and Development
Second Priority for Research and Development.
Third Priority for Research and Development
Evaluation of Research and Development Priorities
Coke Making
Making Sinter.
Raw Material Storage and Handling.
Ir on Making
APPENDIX A
PROCESSES IN THE INTEGRATED IRON AND
STEEL INDUSTRY.
APPENDIX B
GENERAL DESCRIPTION OF AIR-POLLUTION
CONTROL EQUIPMENT.
BATTELLE MEMORIAL INSTITUTE -.COLUMBUS LABORATORIES
Page
. VI-23
. VI-25
. VI-36
. VI-37
. VI-39
. VI-42
. VI-56
. VI-58
. VI-69
. VI-70
. VI-70
. VI-70
. VI-7l
. VI- 74
. VI-77
. VI- 77
. VI-78
. VI-79
. VI-8l
VII-l
. VII-IO
. VII-IO
. VII-II
. VII-12
VII-12
. VII-16
. VII-16
. VII-19
A-I
B-1

-------
TABLE OF CONTENTS
( Continued)
Page
APPENDIX C
CHARACTERISTICS OF EMISSIONS BY THE
INTEGRATED IRON AND STEEL INDUSTRY.
C-l
APPENDIX D
COSTS AND PERFORMANCE OF CONTROL SYSTEMS
AND CONTROL EQUIPMENT
D-l
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
i' Igure III - 1.
Figure III - 2.
Figure III-3.
Figure 1II-4.
Figure 1II-5.
Figure 1II-6.
i' igure 1II-7.
i'igure 1II-8.
Figure 1II-9.
Figure III -1 O.
Figure III - 11 .
Figure III-12.
Figure III -13.
Figure 1II-14.
Figure 1II-15.
Figure V-I.
i'lgure V-2.
Figure V -3.
i'igure V -4.
Figure V-5.
Figure V -6.
LIST OF FIGURES
Value Added by Manufacture by State - 1963 .
Steelmaking Districts in the United States.
. ~
Raw-Steel Production in the Various Geographical Steelmaking
Districts.
Northeast Coast District
Pittsburgh District
. .
Detroit District.
Buffalo District.
Cleveland and Youngstown Districts.
Cincinnati District
Southern District
Chicago and St. Louis Districts
Western Districts.
Size Distribution 'of Open Hearth Furnace s in the United State s .
Trends in Installed Capacity and Raw S,teel Production in
BOF Furnaces.
Size Distribution of Electric Steelmaking Furnaces
Total Energy Consumed Per. Ton of Iron and Steel, According
to Physical Form of the Fuels Used.
Total Energy Consumed Per Ton of Iron and Steel, According
to Ultimate Source of the Energy.
Trend in Use of Sinter in Blast Furnaces in the United States
Source s of Smelting Ene rgy
Consumption of Hydrocarbon Fuels in the Making of Iron and
Steel in the United States
Consumption of Hydrocarbon Fuels Per Net Ton of Raw Steel
in the United State s
BATTELLE' MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
Page
III - 2
III - 3 2
III - 3 3
III - 34
III - 34
III - 3 5
III - 3 5
III-35
III - 36
III - 36
III - 3 7
III - 3 7
III - 3 8
III - 3 9
III-39
V-28
V-29
V-34
V-36
V-48
V-49

-------
Figure V-7.
Figure V -8.
Figure V-9.
Figure V-10.
Figure V-II.
Figure V-12.
Figure V -13.
Figure V -14.
Figure V-15.
Figure V -16.
Figure V-17.
Figure V -18.
Figure VI - 1 .
Figure VI-2.
Figure VI-3.
Figure VI-4.
Figure VI - 5.
LIST OF FIGURES
( Continued)
Sulfur Content of Coke as Related to the Sulfur Content of the
Corresponding Coal
Pounds of Sulfur in Coke Re sulting From the Coking of One
Net Ton of Coal.
Sulfur in Coke-Oven Gas Resulting From the Coking of One
Net Ton of Coal.
Sulfur Content in Steel as Affected by Increased Sulfur in
Fuel Oil
Sulfur Content of BOF Steel as Affected by Basicity of the
Tap Slag.
Relationship Between Sulfur Content in Hot Metal and Sulfur
Content in BOF Steel as Affected by Slag Basicity
Effect of Slag Basicity on the Percent of Sulfur Removed From
the Hot Me tal
Effect of Sulfur Content in the Lime and Slag Basicity on the
Sulfur Content of BOF Steel
Relationship Between Carbon Content in the Steel and Sulfur
Content in the Steel .
Relationship Between Slag Basicity, Iron Oxide Content of the
Slag, and the Partition of Sulfur Between the Metal and the
Slag
Effect of Time Under a Reducing Slag on the Sulfur Content of
AISI 52100 Grade Steel
Sulfur Reduction in Electric Furnace Steelmaking
Ringelmann Smoke Chart
. .
Apparent Resistivities of Metallurgical Dusts.
Apparent Resistivities as Affected by Moisture
Precipitator Efficiency as Related to Collecting-Surface Area,
Gas-Flow Rate, and Drift Velocity
Effect of Gas Volume on Precipitator Efficiency.
BATTELLE MEMORIAL. INSTITUTE - COLUMBUS LABORATORIES
Page
V-50
~
V-51
V-51
V-55
V-58
V-58
V-59
V-59
V-60
V-61
V-61
V-62
VI-1
VI-4
VI-5
VI-6
VI-6

-------
Figure VI-6.
Figure VI-7.
Figure VI-8.
Figure VI-9.
..l'lgure VI-10.
Figure VI-II.
Figure VI-12.
Figure VI-13.
Figure VI-14.
Figure VI-IS.
~ igure VI-16.
~ 19ure VI - 17.
Figure VI-18.
~ 19ure VI-19.
LIST OF FIGURES
( Continued)
Theoretical Total Off-Gas Volume From BOF Furnaces as
Influenced by Heat Size, Percent Hot Metal, and Excess
Combustion Air for a 4.0 Percent Carbon Hot Metal and
20-Minute Blowing Time.
Theoretical and Actual Gas Rates During Blowing
Operating Conditions for a Venturi Scrubber.
Effect of Water Rate on Output Dust Loading for a Venturi
Scrubber.
Performance of a Venturi Scrubber on a Metallurgical Fume
Performance of a Venturi Scrubber on Open-Hearth Fume
Effectiveness of Gas Cleaning by a Fixed-Orifice Scrubber
and a Variable -Orifice Scrubber When Gas -Flow Rate is
Varied
Operating Characteristics of a Blast-Furnace Venturi
Scrubber.
Calibration Curve for a Blast-Furnace Venturi Scrubber.
Relationship Between Clean-Gas Dust Loading and Pressure
Drop for a Wet Scrubber on an Open-Hearth Furnace (Oxygen
Lancing Used During the Refining Period) .
Installation of Electrostatic Precipitators and High-Energy
Scrubbers for Air-Pollution Control at BOF Steelmaking
Plants .
Estimated Installed Capital Costs (1968) of Air-Pollution-
Control Equipment as Related to Different Steel-Making
Processes, on the Basis of Designed Actual Cubic Feet Per
Minute (at Temperature) of Gas Flow Rate.
Estimated Annual Operating Costs (1968) for Air-Pollution-
Control Equipment for Steelmaking Processes (Depreciation
and Capital Charges Are Not Included) .
Estimated Annual Operating Costs (1968) Plus Capital Charges
and Depreciation for Air-Pollution-Control Equipment for
Steelmaking Processes
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
Page
VI - 11
VI - 11
VI-13
VI - 13
VI - 1 5
VI - 15
VI - 1 7
VI - 1 7
VI - 18
VI - 1 9
VI-19
VI-26
VI - 2 7
VI-28

-------
Figure VI-20.
Figure VI-2l.
Figure VI-22.
Figure VI-23.
Figure VI-24.
Figure VI-25.
Figure VI-26.
Figure VI-27.
Figure VI-28.
Figure VI-29.
Figure VI-3D.
Figure VI-3l.
Figure VI-32.
Figure VI-33.
LIST OF FIGURES
( Continued)
Estimated Installed Capital Costs of Air-Pollution-Control
Equipment Installed on Electric-Arc Steelmaking Furnaces.
Control Equipment Designed'to Handle Emissions From Any
One Furnace at One Time
Estimated Installed Capital Costs of Air-Pollution-Control
Equipment Installed on Open-Hearth Furnaces. Control
Equipment Designed to Handle Emissions From Furnaces
Ope ra ting at the Same Time.
Estimated Installed Capital Costs of Air-Pollution-Control
Equipment Used in Sinter and Pellet Plants
Estimated Annual Operating Costs for Air-Pollution-Control
Equipment Used in Sinter and Pellet Plants (Depreciation and
Capital Charges Are Not Included)
Estimated Capital and Annual Operating Costs for Air-Pollution-
Control Equipment Used on Scarfing Machines (Depreciation
and Capital Charges Are Not Included in the Operating Costs)
Range of Estimated Operating Costs for Air-Pollution-Control
Equipment Per Net Ton of Raw Steel - Open-Hearth Furnaces,
BOFs, and Electric Furnaces (Two-Furnace Operations) .
Electrostatic -Precipitator Costs as Affected by Collection
Efficiency
Installed Cost for Steel Plant Electrostatic Precipitators as
Affected by Collection Efficiency.
Relationship Between Target Efficiency
Fabric Filters.
D'
D
Dg
and Vf for
Relationship of Electrostatic Precipitator Collecting Surface
to Collection Efficiency for Open-Hearth Emissions
Relationship of Output Dust Loading to Pressure Drop and Fan
Operating Cost for a Venturi Scrubber Operating on an Open-
Hearth Furnace. .
Nomogram for Estimating the Filter Ratio of a Reverse -Jet
Filter.
Nomogram for Estimating the Size of a Reverse-Jet Filter
Relationship of Particle Size to Collection Efficiency for a
Fabric Filter
BATTELL.E'MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
PalYe
VI - 29
VI - 3 0
VI - 3 1
VI - 3 2
VI-33
VI-34
VI - 3 4
VI - 3 5
VI - 3 8
VI-43
VI-48
VI-53
VI - 54
VI - 5 7

-------
Figure VI-34.
Figure VI-35.
Figure VI-36.
Figure VI-37.
Figure VI-38.
l! igure VI-39.
Figure A-I.
Figure A-2.
Figure A-3.
Figure A-4.
"Figure A-5.
Figure A-6.
Figure A-7.
l! 19ure A-8.
Figure A-9.
l! igure A-I O.
Figure A-II.
l! 19ure A-12.
Figure A-13.
"LIST .OF FIGURES
( Continued)
Page
Grade-Efficiency Curves for Different Types of Air-Pollution
Control Equipment ' .
VI-73
Collection Efficiency as Affected by Particle Size in a
Fabric Filter. .
VI-74
Variation of Volume With Temperature When Gases are Cooled
by the Evaporation of Water
VI-76
Waste Gas Volume Variation for Different Gas Conditioned
Temperatures by Air Infiltration and Water Spraying
VI-76
Variation of Precipitation Efficiency With Particle Size
VI - 80
Electrical Resistivity of Red Oxide Fume From Various Oxygen-
Blown Steelmaking Proces ses .
VI - 8 0
Change in Burden Characteristics for United States Blast
Furnaces. '
A-2
Sinter Plant.
A-3
Grate -Kiln Pelletizing Plant at the Empire Mine
A-4
Typical Blast Furnace Stove
A-7
Effect of Fuel-Oil Injection and Blast Temperature on
Coke Rate
A-8
Effect of Coal Injected at the Tuyeres on Coke Rate.
A-8
Typical Blast Furnace
A-IO
Dwight-Lloyd-McWane Direct Reduction Process
A-II
Production of Carbon Raw Steel in the United States by
Various Processes
A-13
Cross Section of a Basic Open-Hearth Furnace
A-14
Basic Oxygen Furnace
A-17
Stora-Kaldo Rotary Oxygen Converter
A-18
Direct-Arc Electric Furnace
A-19
-BATTELLE ;MEMORIAL'INSTITUTE ;... COLUMBUS LABORATORIES

-------
Figure A-14.
Figure A-IS.
Figure A-16.
Figure A-17.
Figure A-18.
Figure A-19.
Figure A-20.
Figure A-21.
Figure A-22.
Figure A-23.
Figure A - 24.
Figure A-2S.
Figure B-1.
Figure B-2.
Figure B-3.
Figure B-4.
Figure B-S.
Figure B-6.
Figure B-7.
Figure B-8.
Figure B-9.
Figure B-IO.
LIST OF FIGURES
(Continued)
Page
Relationship Between Electric -Arc Furnace Capacity and
Transformer Rating
A-19
Number of Vacuum Degassing Installations in the
United States .
A-21
Ingot Stream Degassing.
A-22
Ladle-Stream Degassing.
A-22
Circulation-Degassing Processes.
A-23
Ladle-Degassing Processes
A-23
Conventional Ingot-Casting Practice.
A-24
Typical Vertical Continuous-Casting Machine for Steel
A-26
Later Designs of Continuous-Casting Machines for Steel.
A-26
Installation for Pressure Casting a Slab
A-27
Nature of Horizontal Processing Line
A-28
Nature of Vertical Processing Line.
A-28
Example of Comparative Energy Levels to Maintain Typical
Exit-Dust Levels
B-1
Typical Cyclone Configurations
B-3
Nested Tubular Cyclones
B-3
Cellular Cyclone
B-3
Wire -in-Tube Electrostatic Precipitator
B-4
Wire-and-Plate Electrostatic Precipitator.
B-4
Internal Construction of Wire-and-Plate Electrostatic
Precipitator.
B-S
Diagram of a Typical Filter Fabric.
B-6
Low- Velocity Bag Filter.
B-7
Shaker-Type Bag Filter.
B-8
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
.t' igure B-ll.
.t'lgure B-12.
Figure B-13.
Figure B-14.
Figure B-1S.
Figure B-16.
Figure B-17.
Figure B-18.
.t' Igure C - 1 .
Figure C-2;
Figure C-3.
.t'igure C-4.
.t'lgure C-S.
Figure c-6.
Figure C-7.
Figure C-8.
.t' igure C-9.
.t'lgure C-IO.
.t' Igure C-ll.
Figure C-12.
Figure C-13.
LIST OF FIGURES
(Continued)
Page
Reverse-Jet Bag Filter.
B-8
Irrigated-Target Scrubber.
B-9
Disintegrator
B-IO
Spray Tower.
B-ll
Venturi Scrubber
B-ll
Variable -Throat Venturi Scrubber
B-12
Venturi Scrubber Using Slurries.
B-12
Flooded-Disk Scrubber
B-13
Consumption of Sinter and Pellets in Blast Furnaces in the
United States
C-3
Typical Flow Sheet for. a Sintering Plant
C-4
Size Distributions of Various Sinter-Plant Dusts.
C-6
Dust-Flow Distribution Along a 92-Foot Sintering Machine
C-7
Consumptions of Limestone and Lime by the Integrated Iron
and Steel Industry in the United State s
C-12
Trend in Limestone Consumption in the United States per
Net Ton of Pig Iron
C-12
Trends in the Consumption of Coke and in Coke Rate in the
Production of Hot Metal in American Blast Furnace s
C-1S
Effect of Amount of Sinter and Pellets in Blast Furnace
Burden on Coke Rate.
C-15
Typical Flow Sheet for a By-Product Coke Plant.
C-17
Particulate Emission During Coke Quenching.
C-20
By-Product Plant Flow Sheet
C-22
Relation of Reflectance to Ultimate Carbon
C-24
Relation of Reflectance to Volatile Matter.
C-24
BATTELLE MEMORIAL INSTITUTE - COLUMBUS "LABORATORIES

-------
Figure C-14.
Figure C-15.
Figure C-16.
Figure C-17.
Figure C-18.
Figure C - 19.
Figure C-20.
Figure C-21.
Figure C-22.
Figure C-23.
Figure C-24.
Figure C-25.
Figure C-26.
Figure C-2 7.
Figure C-28.
Figure C-29.
Figure C -30.
LIST OF FIGURES
(C ontinued)
Relation of Coal Reflectance to Coke - Wall Reflectance.
Production of Pig Iron in the United State s .
Relationship of Hot-Metal Production and Number of Blast
Furnaces in the United States, 1958-1967 .
Trend in the Average Production per Blast Furnace in the
United States
Typical Blast-Furnace Operation on a Burden of
100 Percent Unscreened Ore
Typical Blast-Furnace Operation on a Burden of
100 Percent Screened Ore
Typical Blast-Furl1ace Operation on a Burden of Lump
Iron Ore and Sinter
Typical Blast-Furnace Operation on a Burden Consisting
Mainly of Sinter
Typical Blast-Furnace Operation on a Burden of
100 Percent Pellets
Typical Blast-Furnace Operation on a Burden of 100 Percent
Pellets, and W'ith Injection of Natural Gas.
Typical Blast-Furnace Operation With a Burden Consisting
Mainly of Sinter, and With Injection of Natural Gas.
Effect of Burden Improvement on Dust Rate From a Blast
Furnace to Its Dust-Collecting System.
Effect of Coke Rate on Volume of Blast-Furnace Gas Produced.
Effect of Coke Rate on the Calorific Value of Blast-
Furnace Top Gas
Relationship of Moisture in Blast Air to Hydrogen in
Blast-Furnace Top Gas.
Flow Diagrams of Two Early Blast-Furnace Gas-
Cleaning Systems.
Flow Diagrams of Blast-Furnace Gas-Cleaning Systems.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
Page
C-24
C-28
C-29
C-29
C-31
C-32
C-33
C-34
C-35
C-36
C-37
C-39
C-42
C-42
C-46
C-48
C-49

-------
Figure C-31.
Figure C-32.
.1' Igure C-33.
.1' igure C-34.
Figure C -35.
Figure C-36.
Figure C-37.
Figure C-38.
Figure C-39.
Figure C-40.
Figure C-41.
Figure C-42.
LIST OF FIGURES
(Continued)
Page
Flow Diagrams of Three Recent Blast-Furnace Gas-
Cleaning Systems. .
C-49
Size Distribution of Open-Hearth Furnaces in the
United State s in 1968 and 1973
C-51
Size Distribution of Open-Hearth Furnaces in Several
Geographical Districts for 1968 and Subsequent Years
C-52
Size Distribution of Open-Hearth Furnaces in Several
Geographical Districts for 1968 and Subsequent Years
C-53
Open-Hearth Furnace Operating with a Cold-Metal Charge
Consisting of 30 Percent Pig Iron and 70 Percent Steel
Scrap (Ore Practice) . .
C-55
Open-Hearth Furnace Operating with a Cold-Metal Charge
Consisting of 40 Percent Pig Iron and 60 Percent Steel
Scrap (Oxygen Practice)
C-56
Open-Hearth Furnace Operating with Hot-Metal Practice
Consisting of 50 Percent Hot Metal and 50 Percent
Steel Scrap (Ore Practice) .
C-57
Open-Hearth Furnace Operating with Hot-Metal Practice
Consisting of 60 Percent Hot Metal and 40 Percent Steel
Scrap (Ore Practice) .
C-58
Open-Hearth Furnace Operating with Hot-Metal Practice
Consisting of 70 Percent Hot Metal and 30 Percent
Steel Scrap (Ore Practice) .
C-59
Open-Hearth Furnace Operating with Hot-Metal Practice
Consisting of 50 Percent Hot-Metal and 50 Percent Steel
Scrap (Oxygen Practice)
C-60
Open-Hearth Furance Operating with Hot-Metal Practice
Consisting of 60 Percent Hot Metal and 40 Percent Steel
Scrap (Oxygen Practice) .
C-61
Effect of Different Fuels on Dust Loadings from an 80-Ton
Open-Hearth Furnace Operating Without Oxygen Injection
(German ~ractice)
c-63
BATTELLE.
MEMORIAL INSTI'TUTE ~ COLUMBUS LABORATORI'ES

-------
Figure C-43.
Figure C-44.
Figure C-45.
Figure C-46.
Figure C-47.
Figure C-48.
Figure C -49.
Figure C- 50.
Figure C-5l.
Figure C-52.
Figure C-53.
Figure C-54.
Figure C -55.
Figure C-56.
Figure C-57.
Figure C-58.
LIST OF FIGURES
( Continued)
Page
Sulfur Dioxide .and Carbon Dioxide Contents of an
Oxygen-Lanced Open Hearth Fired With a Tar-Oil
Fuel (German Practice) .
C-63
Dust Loading During Oxygen-Lanced Open-Hearth
Practice With a Tar -Oil Fuel Mixture (German Practice) .
C-64
Size Distribution of Open-Hearth Particulate Emissions
for Operation With Oxygen Lancing (U. S. Practice)
C-65
Flow Diagrams of Open-Hearth Dust-Collecting
Systems Using Electrostatic Precipitators.
C-67
Flow Diagrams of Open-Hearth Dust-Collecting
Systems Using Scrubbers and Bag Houses.
C-68
Trends in Installed Capacity and Raw-Steel
Production in BOF Furnaces
C-69
Relationship Between the Volume of Oxygen Blown
and Volume of Exhaust Gases
C-70
Effect of Number of Holes in Oxygen Lance on
Emis sions During Oxygen Blowing
C-70
Effect of Velocity on Emission During Oxygen
Blowing of BOF Furnace (Laboratory Results)
C-7l
Effect of Carbon Content on Emissions From
BOF Furnace (Laboratory Results)
C-7l
Effect of Metal Temperature on Generation of Emissions
From BOF Furnace (Laboratory Results)
C-72
Basic Oxygen Furnace Operating With 80 Percent
Hot Metal and 20 Percent Steel Scrap
C-74
Basic Oxygen Furnace Operating With 70 Percent
Hot Metal and 30 Percent Steel Scrap
C-75
Rotary Oxygen Furnace Operating With 55 Percent
Hot Metal and 45 Percent Steel Scrap
C-76
Off-Gas Analysis From a 60-Ton Basic Oxygen
Converter (Before Combustion)
C-79
Examples of Wet-Cleaning Systems for BOF

Steelmaking Furnaces.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
C-80

-------
Figure C-59.
.I:' Igure C-60.
Figure C-61.
.I:' 1 gur e C - 62 .
Figure C-63.
.I:' Igure C-64.
Figure C-65.
Figure C-66.
Figure C-67.
Figure C-68.
Figure C-69.
Figure C-70..
.l:'lgure C-71.
Figure C-72.
Figure C-73.
Figure C-74.
LIST OF FldURES
( Continued)
Examples of Electrostatic-Precipitation Gas-Cleaning
Systems for BOF Steelmaking Furnaces.
Annual Production of Raw Steel in Electric
Furnaces in the United States
Size Distribution of Electric Steelmaking Furnaces.
Dust Emissions During Electric-Furnace Melting of Steel.
Dust Loadings During Oxygen Blowing of a 40-Ton
Heat in an Electric-Arc Steelmaking Furnace.
Example of Electric-Furnace Steelmaking Using
a Charge of Cold Steel Scrap (Ore Practice)
Example of Electric-Furnace Steelmaking Using
a Charge of Cold Steel Scrap (Oxygen Practice)
Example of Electric-Furnace Steelmaking Using a
Charge of Cold Steel Scrap (Oxy-Fuel Burners for
Meltdown; Oxygen Practice)
Dust Loadings During Preheating of Scrap in Norway
Size Distribution of Dust Particulates During
Scrap Preheati~g in Norway.
Gases Generated During the Production of a
Ball-Bearing Steel in a 22 -Ton Electric Furnace
Carbon Monoxide Evolution During Oxygen Lancing
in a 16. 5-Ton Electric Steelmaking Furnace
Relationship Between the Evolution of Dust Particulates
and Carbon Monoxide. in a 16. 5-Ton Electric Furnace.
Manganese Recovery in Steel as ~ffected by the Method
of Extrac"tion of Electric-Furnace Emissions.
Examples of Direct-Extraction Emission Control
Systems With Bag Houses
Examples of Electric-Furnace Dust-Collecting
Systems Using Wet Scrubbers.
BATTELLE. MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
Page
C-81
C-82
C-82
C-83
C-83
C-85
C-86
C-87
C-89
C-89
C-90
C-90
C-91
C-92
C-94
C-95

-------
Figure C-75.
Figur e C-76.
Figure C-77.
Figur e C-78.
Figure C-79.
Figur e C-80.
Figur e C-81.
Figur e C-82.
Figur e C-83.
Figure C-84.
Figur e C-85.
Figure C-86.
Figure C-87.
Figur e C-88.
Figure C-89.
Figure C-90.
Figur e C - 9 1.
Figure C-92.
LIST OF FIGURES
(C ontinued)
Examples of Electric-Furnace Shop-Roof
Emis sion-Control Systems,. '.
Vacuum-Stream, Degassing:~f Molten Steel.
Vacuum-Chamber Degassing of Molten Steel
Variatio'n in Gas Contents During Vacuum-Stream Degassing.
Typical Steam-Ejector System for Vacuum Degassing.
Cross Secti~n of a Steam Ejector.
Annual Production of Semifinished Steel Products
Ingot Structures for Different Types of Steel
Casting of Steel Into Ingots; and Rolling to
Slabs, Blooms, and Billets.
Continuous -Cast Production as Compared With
Total Pr.oduction of Billets, Blooms, and Slab s
Continuous Casting of Steel.
Pressure Casting of Steel Slabs
Typical Hot-Rolling of Bar and Merchant Mill
Products From Billets and Blooms
Typical Hot-Rolling of Sheet and Strip From Slabs
Trend in Ingot Tonnage of Steel That is Machine Scarfed
Cold Rolling of Strip Steel
Production of Sheet Steel Coated Products.
Typical Galvanizing Processes.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
Page
C-96
C-98
C-99
C-IOO
C-I02
C-I03
C -1 04
C-I05
C-I07
C-I09
C-lll
C-1l3
C-115
C-l15
C-1l6
C-118
C-119
C -121

-------
Table III-I.
Table III-2.
Table 1II-3.
Table 1II-4.
Table 1II-5.
.1 able 1II-6.
.1able III-7.
Table 1II-8.
Table III-9.
Tab Ie III - 10.
Table III-ll.
Table III-12.
Table III-13.
Table III-14.
Table III-15.
Table III-16.
LIST OF TAB LES
Production, Shipments, and Foreign Trade of the United
States Iron and Steel Industry - 1967
Financial Balance Sheet: U. S. Steel Industry - 1967.
Income, Dividends, Employment, and Employment Costs:
United States Steel Industry - 1967 .
Employment, Stockholders, and Capital Expenditures:
United States Steel Industry - 1967 .
Total Net Shipments of Steel Products by United States
Steel Industry - 1967
Total Production of Hot-Rolled Products (by Type of
Product) by United States Steel Industry - 1967 .
Total Production of Hot-:Rolled Steel Products (by
State) by United States Steel Industry - 1967 .
Production of Plates and Skelp (by State) by United States
Steel Industry - 1967
Production of Flat Hot-Rolled Products (Except Plates)
by United States Steel Industry - 1967 (by State)
Production of Merchant Bars and Light Shapes (by State)
by United States Steel Industry - 1967 .
Production of Concrete Reinforcing Bars (by State) by
United States Steel Industry - 1967 .
Production of Heavy Structural Shapes and Steel Piling
(by State) by United States Steel Industry - 1967
Production of Wire Rods (by State) by United States
Steel Industry - 1967
Production of Blanks, Tube Rounds or Pierced Billets
for Seamless Tubing (by State) by U. S. Steel
Industry - 1967
Production of Semi£inished Steel for Forgings and
Export (by State) by U. S. Steel Industry - 1967
Production of Cold-Rolled Sheets and Strip by United
States Steel Industry - 1967
BATTELLE MEMORIAL INSTITUTE... COLUMBUS LABORATORIES
Page
III-3
III-4
III - 5
III - 6
1II-7
III - 9
. III - 1 0
. Ill-ll
. III-II
. III-12
. III-12
. III-13
. III-13
. III-14
. III-14
. III-15

-------
Table III-17.
Table III-18.
Table III-19.
Table III-20.
Table III-21.
Table llI-22.
Table III-23.
Table 1II-24.
Table III-25.
Table 1II-26.
Table 1II-27.
Table 1II-28.
Table III-29.
Table 1II-30.
Table 1II-3l.
Table 1II-32.
LIST OF TAB LES
(Continued)
Production of Coated Sheets and Strip by United States
Steel Industry - 1967
Production of Pipe and Tubing (by Type) by United States
Steel Industry - 1967 , . '
Steel Production (by Grades) by United States Steel
Industry - 1967 . .
Raw Steel Production (by State and by Type of Furnace)
by United States Steel Industry - 1967 . .
Production of Pig Iron and Ferroalloys by United
States Steel Industry - 1967
Production of Pig Iron and Ferroalloys (by Grades and
Kinds) by United States Steel Industry - 1967
Number of Blast Furnaces Producing Pig Iron and
Ferroalloyson January 1, 1968 (by State)
Productionof Pig Iron and Ferroalloys (by State) by
United'States Steel Industry - 1967 .
Materials Used by United States Blast Furnaces in
Manufacture of Iron and Ferroalloys - 1967 .
Materials Used by United States Blast Furnaces in the
'Manufacture of Pig Iron (by States) - 1967
Consumption of Materials per Net Ton of Pig Iron
Produced in the United States - 1967
Scrap - Stocks, Production, Receipts, and
Consumption by Grade - 1967
Consumption of Scrap and Pig Iron, and Production
of Steel by Types of Furnaces - 1967 .
Consumption of Fuels by United States Steel
Industry - 1967
Consumption of Electric Power by United States
Steel Industry - 1967
Consumption of Fluxes by United States Steel
Industry - 1967
BATTELLE'MEMORIAL INSTITUTE- COLUMBUS LABORATORIES
Page
. III-15
. III-16
. III-16
. III-17
. III - 18
. III - 1 9
. III-20
. 111-21
. lII-22
. lII-23
. 111-23
. lII-24
. 111-24
. 111-25
. lII-25
. 111-26

-------
Table 1II-33.
Table 1II-34.
.table III-35.
Table III-36.
Table 1II-37.
Table III-38.
Table 1II-39.
Table 1II-40.
.t able 1II-41.
Table III-42.
Table IV -1.
.table IV-2.
Table IV -3.
Table IV -4.
Table IV -5.
Table IV -6.
LIST OF TABLES
(Continued)
. Consumption of Oxygen by United States Steel
Industry - 1967
Production and Use of Agglomerated Products by
United States Steel Industry - 1967 .
Consumption of Nonferrous Metals for Coating Purposes
by United States Steel Industry - 1967 .
United States Steel Industry Iron Ore Inventories,
Re ce ipts, and Consumption - 1967 .
Shipments of Iron Ore (by State) - 1966
Production, Receipts, Consumption and Shipments of
Coke by United States Iron and Steel Industry - 1967 .
Coke Production by United States Iron and Steel
Industry (by State) - 1967 .
Coke Consumption by United States Iron and Steel
Industry (by State) - 1967 .
Coke Consumption by Uses in the United States Iron
and Steel Industry - 1967 .
Coal Consumption by Uses in the United States Iron
and Steel Industry - 1967 .
Production (1000 Net Tons) of Raw Steel in the United
States, by Type of Furnace, and Projections of
Production to 1980
U. S. Production of Raw Steel, as Forecast by
Various Sources
Production of Raw Steel, by Area, in 1960 and 1967
and Forecasts to 1980
Apparent Per -Capita Consumption of Raw Steel in the
United States for 1960 and 1967 and Projections to 1980 .
Pig-Iron Requirements for the United States in 1960
and 1967 and Projections to 1980
Consumption of Iron Ore and Derivatives in the Iron and
Steel Industry in the United States and Projections to 1980 .
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
Page
. III-26
. III - 2 7
. III - 2 7
. llI-28
. III- 28
. III-29
. III-30
. III-30
. III - 3 1
. III-31
IV -2
IV -3
IV-4
IV -5
IV -6
IV -7

-------
Table IV -7.
Table IV -8.
Table V-I.
Table V -2.
Table V -3.
Table V -4.
Table V-5.
Table V -6.
Table V-7.
Table V -8.
Table V -9.
Table V-10.
Table V -11.
Table V-12.
Table V -13 .
TableV-14.
LIST OF TABLES
(Continued)
Physical Form of Iron Ore Consumed in the United
States and Estimates to 1980 .
Average Production of Pig Iron per U. S. Blast Furnace
per Day for 1960 and 1967 and Projections to 1980 .
Total Fuels, Steelmaking Oxygen, and Purchased
Electricity Consumed by Steel Plants in the United
States, 1958-1967
Fuels, Steelmaking Oxygen, and Purchased Electricity
Consumed per Ton of Steel Produced in the United
States, 1958-1967
Conversion Factors for Fuels, Steelmaking Oxygen,
and Electricity as Applied in Steelmaking.
Distribution of Amounts of Energy Used per Ton of Iron and
Steel Produced in United States Steel Plants, 1958-1967 .
Estimated Use of Various Fuels for Underfiring Coke
Ovens (1966-1967 Averages) .
The Three Major Steelmaking Processes:
Attributes, and Bases for Competition.
Nature,
Estimation of Sources of Energy for Steelmaking
Processes (Other Than Sensible Heat in Charges) .
Use of Fuels to Fire Heating and Annealing Furnaces
in the United States Steel Industry, 1958-1967
Sulfur Balance for Coke Oven Operation
Sulfur Balance for Sintering Machine Operation.
Sulfur Balance for Blast Furnace From Figure C-24
in Appendix C .
Sources of Sulfur in Open Hearth Steelmaking.
Sulfur Balance for Open Hearth Furnace With 60 Percent
Hot Metal and 40 Percent Steel Scrap Oxygen Practice
(From Figure C-4l in Appendix C) .
Sulfur Balance for BOF Steelmaking Based on Material
Balance Given in Figure C-55 With 70 Percent Hot
Metal and 30 Percent Steel Scrap

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
Page
IV -7
. IV - 12
V-23
V-24
V-25
V-27
V-32
V-38
V-39
V-42
V-52
V-53
V-54
V-55
V-56
V-57

-------
Table V-15.
Table V-16.
Table VI-l.
Table VI-2.
Table VI-3.
.lable VI-4.
Table VI-5.
Table VI-6.
Table VI-7.
Table VI-8.
.lable VI-9.
.1 able VI-10.
Table VI-ll.
Table VI-12.
LIST OF TABLES
( Continued)
Sulfur Balance for Electric Furnace Steelmaking Using a
Charge of Cold Steel Scrap and Oxygen Practice (From
Figure C-65 in Appendix C)
Estimated Summary of Sulfur Balance for the U. S. Iron
and Steel Industry (1967)
Representative Emission-Control Applications in the
Integrated Iron and Steel Industry
Basic Oxygen Furnace Installations and Associated
Air -Pollution Control Equipment
Design and Operating Data for Sinter-Plant Fabric
Filters on Sinter Strand Discharge.
Estimated Cost Differences for a Sinter Plant Electro-
static Precipitator as Affected by the Output Dust Loading
From the Wind box
Estimated Cost Differences for a Sinter Plant Electro-
static Precipitator as Affected by the Output Dust Loading
From Material Handling
Estimated Cost Differences for a BOF Electrostatic
Precipitator as Affected by Output Dust Loading
Estimated Cost Differences for an Open Hearth Electro-
static Precipitator as Affected by Output Dust Loading
Estimated Cost Differences for an Electric Furnace
Electrostatic Precipitator as Affected by Output
Dust Loading
Estimated Cost Differences for a Scarfing Machine
Electrostatic Precipitator as Affected by Output
Dust Loading
Estimated Cost Differences for a .BOF Wet Scrubber
as Affected by Output Dust Loading.
Estimated Cost Differences for an Open-Hearth Wet
Scrubber as Affected by Output Dust Loading
Estimated Cost Differences for a Scarfing Machine Wet
Scrubber as Affected by Output Dust Loading
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
Page
V -62
V-64
VI-3
. VI - 1 0
. VI - 2 2
. VI-45
. VI-45
. VI-46
. VI-46
. VI-47
. VI-47
. VI- 51
. VI-51
. VI- 52

-------
j-
Table VI-13.
Table VI-l4.
Table VII-I.
Table VII-2.
Table VII-3.
Table VII-4.
Table VII-5.
Table VII- 6.
Table VII-7.
Table VII-S.
Table VII-9.
Table VII-IO.
Table VII-II.
Table VII-12.
Table B-1.
Table C-l.
Table C-2.
LIST OF TABLES
(Continued)
Grading of W. C. 3 Test Dust.
Calculated Relative Efficiency of Collecting Equipment
for Various Process Dusts and Collectors
Identification Key for Problem Identification.
Priority for Research and Development Efforts.
Research and Development Evaluation for Making Pellets
and for Raw-Material Handling
Research and Development Evaluation for Coke Making.
Research and Development Evaluation for Making
Sinter and Making Iron.
Research and Development Evaluation for Making
Steel and Pouring Ingots
Research and Development Evaluation for Primary
Rolling, Continuous Casting, Pressure Casting, and
Conditioning
Research and Development Evaluation for Hot Rolling,
Cold Rolling, Coating of Finished Products, Waste
Incineration, and Power Generation
Analysis and Possible Solutions to Coke -Plant
Air Pollution Problems
Analysis and Possible Solutions to Sinter-Plant
Air Pollution Problems.
Analysis and Possible Solutions to Raw Material
Storage and Handling Air-Pollution Problems
Analysis and Possible Solutions to Iron Making
Air-Pollution Problems
Operating Conditions for Typical Baghouse Fabrics
Sintering-Machine Stack-Emission Test Data on Exhaust
Gas Dust Loading at Ignition End
Screen Analysis of Particulate Emission From a
Sintering Machine
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
Page
. VI - 71
. VI-75
VII-l
VII-3
VII-4
VII-5
VII-6
VII-7
VII - S
VII-9
. VII-13
. VII-I7
. VII-IS
VII-20
B-6
C-6
C-6

-------
Table C-3.
.Lable C-4.
Table C-5.
Table C-6.
.Lable C-7.
Table C-8.
Table C-9.
Table C-10.
Table C-ll.
Table C-12.
Table C-13.
.Lable C-14.
.Lable C-15.
.L able C -16.
Table C -1 7 .
Table C-18.
Table C-19.
Table C-20.
LIST OF TABLES 
(Continued)
Characteristics of Some Sinter-Plant Emission-
Control Systems.
Iron Ore Pellet Plants in the United States
Iron-Ore Pellet Plants in Canada
Estimated Lime Production Captive to the Integrated
Iron and Steel Industry in the United States
Screen Analyses of Coke-Plant Particulates,
Weight Percent
Screen Analysis of Quench-Tower Particulates.
Chemical Analysis of Dry, Blast-Furnace Flue Dust
Size Analysis of Blast Furnace Flue Dust From U. S.
Blast Furnaces
Dust Loadings for Great Lakes Blast Furnaces.
Top-Gas Analyses for Different Blast-Furnace Burdens.
Dust Generation in Open-Hearth Steelmaking
Chemical Compositions of Open-Hearth Particulate
Emissions, Oxygen Lancing, Weight Percent
Size Distribution of Silica Fume.
Chemical Compositions of Basic Oxygen Furnace Steel-
making Dust, Weight Percent.
Composition of Off Gases From Oxygen Steelmaking
Process, Volume Percent.
Chemical Compositions of Electric-Furnace Dusts,
Weight Percent
Changes in Composition of Electric -Furnace Dust
During a Single Heat.
Size Distribution of Particulate Emissions From
Electric Steelmaking Furnaces, Percent.
BATTELLE MEMORIAL INSTll'UTE - COLUMBUS LABORATORIES
Page
C-8
C-9
C-10
C-13
C-26
C-26
C-45
C-45
C-45
C-46
C-64
C-65
C-73
C-78
C-78
C-84
C-88
C-88

-------
Table C-2l.
Table C-22.
Table C-23.
Table C-24.
Table C-25.
Table C-26.
Table C-27.
Table C-28.
Table C-29.
Table C-30.
LIST OF TABLES
(Continued)
Composition of Off Gases During Vacuum-Stream
Degassing of Steel
Off-Gas Composition During Vacuum-Stream Degassing
of Ball-Bearing Steels.
Dust and Metal Analyses for Vacuum-Treated Steels.
Effect of Treating Rate and Amount of Carbon Removed
as Carbon Monoxide on the Gas Load During Vacuum
Treatment With Steam Ejectors.
Production of Pressure Cast Slabs.
Surface -Treatment Facilities for Sheet Products in the
Integrated Iron and Steel Industry
Chemical Composition of Galvanizing Emissions
Foundry Facilities Located At Steel Plants in the
United States
Size Characteristics of Particulates Emitted From
Gray Iron Cupolas
Chemical Analysis of Particulates From a Gray Iron
Cupola
BATTELLE MEMORIAL INSTITUTE ~ COLUMBUS LABORATORIES
PaP'1'!
. C -100
. C-IOl
. C-IOl
. C -1 03
. C-1l2
. C-120
. C-IZl
. C-123
. C-124
. C-124

-------
I - 1
SECTION I
INTRODUCTION AND SCOPE
The National Air Pollution Control Administration (NAPCA) in fulfilling part of
its responsibilities under the Air Quality Act of 1967 is sponsoring systems-analysis
studies of selected industries. These studies will describe (1) the present status of
control technology, (2) the economic burden of controlling air-pollutant emissions, and
(3) the research and development needs requiring solution to provide for both better and
more economical means for controlling air pollution. It is recognized that if solutions
to air-pollution control problems are to be responsive to the needs and objectives of
both industry and government, they require joint participation and involvement. There-
fore, industry has been requested to participate in these studies by providing technical
and cost information which will keep these studies responsive to industry's needs.
The Division of Economic Effects Research and the Division of Process Control
Engineering of the National Air Pollution Control Administration jointly have. funded
this study of the integrated iron and steel industry. The Division of Economic Effects
Research is responsible for determining the economic impact of air-quality standards
on the Nation's industries, communities, and other contributing sources of pollution.
The Division of Process Control Engineering is responsible for federally sponsored
research and development of air-pollution control devices and processes for stationary
emission sources.
The integrated iron and steel industry is among those chosen by NAPCA for study
because the processes used to manufacture coke, pig iron, and steel make a contribu-
tion to air pollution. The pollution potential and the degree of control practiced vary
widely depending upon the particular processes involved and the geographical location
of the steel works. In addition to the technical problems associated with controlling
these emissions, the industry trend toward use of newer and higher production manu-
facturing processes puts an increased decision-making burden upon steel-industry
management when deciding to what extent emissions from existing equipment shall be
controlled.
This report is concerned mainly with the technological aspects of ironmaking and
steelmaking and an analysis of air-pollution technology as it applies to the integrated
iron and steel industry. A companion report is concerned primarily with costs of
air-pollution control. However, for orientation and fuller understanding of the techno-
logical problems, some cost information is included in this present report. For more
complete information on costs, reference should be made to the companion report.
PURPOSES OF THIS STUDY
This iron and steel study has the purposes of (1) analyzing emission-control tech-
nology as it applies to the various manufacturing processes in the integrated iron and
steel industry, (2) analyzing emis sian-control alternatives, (3) identifying the practical
problems associated with controlling emissions from the iron and steel industry, and

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
1-2
(4) defining the problem areas and needs, thereby providing one basis for establishing
priorities for federally sponsored research and development projects to be undertaken
independently or on a joint- sponsorship basis with industry. In furtherance of a pur-
pose of the Air Quality Act of 1967, the research programs are to be directed toward
new, improved, and lower cost control techniques having industry-wide application.
SCOPE OF THIS STUDY
To accomplish the overall goals of this study, the following tasks were under-
taken:
( 1 )
A general description of the various processes used in each
manufacturing segment, including proces s block-flow diagrams
and schematic drawings for each proces s segment and for
principal variations within each process segment
(2 )
Identification of the particulate and gaseous emis sions from each
manufacturing proce s s at the point of emis sian
(3 )
A material balance to identify particulate and gaseous emissions,
especially sulfur oxides, and to trace the flow of sulfur through
representative models of the process segments
(4)
A comprehensive and quantitative review of fuel and energy
utilization to identify the total industry consumption, to identify
the types of fuels used, the quantities of each, identification of
the determinants of fuels choice, and interpretation of the influ-
ence on pollutant emissions from the standpoints of fuel chemistry
and fuel usage
(5 )
A projection of industry growth and proces s technology changes
for the periods 1970-1975 and 1975-1980, emphasizing processes
that will come into future use and their associated pollution poten-
tial, as well as processes which will be phased out
(6)
Description of the principal types of control equipment for the
various processes, and information on the performance charac-
teristics of the various designs in use
(7)
An in-depth narrative analysis of the relative merits and limita-
tions of applied control equipment
(8 )
Identification of pollution-control alternatives for the various
process segments in the industry, and comparison of the rela-
tive cost and effectivenes s of each
(9)
Identification of gaps in air-quality control technology, and recom-
mendations for research to fill such gaps.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
1-3 and 1-4
The scope of the study included the following process segments of the integrated
iron and steel industry with respect to the generation of data on particulate and gaseous
emissions:
( 10)
( 11)
(12)
(13)
(1) Ore preparation (sintering and pelletizing plants)
(2 )
Coke -manufacturing plants and by-product facilitie s
(3 )
Blast furnace s
(4) Open-hearth furnaces (with and without oxygen lancing)
(5 )
Basic oxygen furnaces (upright and rotary)
(6 )
Electric furnaces (arc and induction, including the preheating
of scrap)
(7)
Vacuum degassing
(8)
Casting operations (ingot practice, continuous casting, and
pressure casting)
(9) Roughing-mill operations (soaking-pit practice and rolling to
slabs, blooms, and billets)
Hot rolling (slab, bloom, and billet conditioning; reheating and
rolling to plate, sheet, strip, structural and merchant products)
Cold rolling of strip and bar products
Surface cleaning and coating (galvanizing, tin plating, and plastic
coating)
In-plant power generation.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-I
SECTION II
SUMMARY AND RECOMMENDATIONS
The overall objectives of this study of the steel industry are to determine
(1)
The present status of air-pollution-control technology
(2)
The cost of applying present technology to the lessening of air-
pollution by steelworks
(3)
The subjects on which research and development are needed to
improve the technology and economics of air-pollution control.
The study was conducted in 1968 and early 1969 by Battelle Memorial Institute,
Columbus Laboratories, with subcontracted assistance from Swindell-Dressler Com-
pany of Pittsburgh and voluntary assistance from a number of steel companies that
contributed information and guidance throughout the project,
This Final Technological Report is complementary to a companion report titled
"A Cost Analysis of Air Pollution Controls in the Integrated Iron and Steel Industry",
dated May 15, 1969.
The partition of the final Battelle reports reflects the design of this study to serve
both the Division of Economic Effects Res"earch and the Division of Process Control
Engineering of the National Air Pollution Control Administration. "
This Final Technological Report is the result of investigative, data-gathering
operations and data analyses with respect to ironmaking and steelmaking technology
and air-pollution-control technology in the integrated iron and steel industry. The
status of the respective technologies is discussed in relation to present and future air-
quality requirements. Technology gaps are identified so that research and development
may serve to minimize the generation of emissions in the various processes, and efforts
may be directed toward the development of improved pollution-control procedures.
THE NATURE OF THE INTEGRATED IRON AND STEEL INDUSTRY
The integrated iron and steel industry, as considered in this study, consists of
those companies with resources of iron ore and with facilities for ore preparation,
ironmaking, and steelmaking. Included as part of steelmaking are the production of
semUinished products such as ingots, billets, blooms, and slabs; and the production of
finished products such as sheet thal is galvanized, tinplated, painted, or plastic coated,
Many smaller companies utilizing scrap as their source of raw materials and electric
furnaces as their melting facility have come into operation in recent years, Such small
secondary (nonintegrated) plants are established to satisfy local markets or markets
for a limited range of specific steel products.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-2
Industry Efforts for the Immediate Future
The integrated iron and steel industry is faced with a dual situation, as are many
industries in the United States today. First, they are in a highly competitive market
which requires that they maintain an advancing metallurgical technology to meet the
increased demands of an increasing population and maintain a position in the world
market. Second, the industry must, as a part of the overall scheme to improve man's
environment, reduce the amount of its process emissions reaching the atmosphere,
The last 10 years have seen significant advances in the technology of making iron
and steel. The use of ores has changed from run-of-the-mine varieties to closely
sized, select ores, and to sinter and pellets made from less-suitable ores. Use of the
improved burden materials has incr eased ironmaking production and has resulted in a
lessening of emissions from the blast furnace. However, this improvement in blast-
. furnace technology has resulted in an aggravated emission problem at the sinter plant,
particularly when self-fluxing sinters are made.
Coal preparation and coke-making technology have been advanced to make a bet-
ter, lower sulfur fuel for the blast furnace, but the coke-making process itself has not
been changed basically and the coke -plant emis sions are a source of continuing concern
to the industry. Auxiliary fuels such as oil, natural gas, and coal have been combined
with air-blast temperatures as high as 2000 F and with other technological advances to
lower the amount of coke required to make a net ton of iron from 1600 pounds in 1958
to 1250 pounds in 1968, The total overall tonnage of coke, however, continues to in-
crease because of requirements for more hot metal to make more steel, The improve-
ments in ironmaking technology have permitted an increase in physical size of the blast
furnace and a reduction in the total number of blast furnaces to satisfy ironmaking re-
quirements,
Steelmaking technology has changed from one dominated by open-hearth technol-
ogy at the start of the present decade to one shared 50 percent by open hearths, 37 per-
cent by basic oxygen furnaces, and 13 percent by electric furnaces in 1968. Basic
oxygen furnaces with capabilities of producing from 150 to 300 net tons of steel per hour
have played a major role in changing steelmaking technology, The use of gaseous oxy-
gen in the refining of steel has been the prime mover for achieving these higher pro-
duction rates but, at the same time, this use of gaseous oxygen has become a major
problem in control of air pollution because of the severe generation of red iron oxide
fumes during its use. This pollution problem is rapidly being brought under control by
the steel industry.
Conversion of steel from its molten state to a solid, more-usable state is pres-
ently undergoing technological change, The older technology which had been used in
the industry practically from its inception consisted of pouring the molten steel into
ingot molds, permitting the steel to solidify, removing the hot ingot from the mold,
. transferring the ingot to a soaking pit, and then rolling the heated ingot into billets,
blooms, or slabs. The newer technologies of continuous casting and pressure casting
bypass much of the intermediate processing. In continuous casting, the molten steel is
poured into a special mold and the solidified steel is extracted continuously from the
bottom of the mold in the form of billets, blooms, or slabs. The continuously cast
semifinished products are cut to length, and then are ready for further processing
into finished products. Pressure casting uses air pressure on the surface of molten
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-3
steel contained in a ladle which is positioned in a special tank to force the molten steel
into a graphite mold via a ceramic tube. After solidification, the slabs are ready for
rolling into finished products. Although the particulates and gases generated in the cast-
ing of ingots, and rolling into semifinished products have not been a large source of air
pollution, these operations are still of concern from air-quality aspects in the iron and
steel industry. The new technologies of continuous and pres sure casting have lowered
emis sions from this stage of steel processing.
The processes used in the manufacture of finished products have not been sources
of large amounts of emis sions to the atmosphere, but steps have been and are being
implemented to les sen the small amounts of emis sions generated during the rolling
operations. Iron oxide fume is generated in the high-speed, high-powered, hot-strip
mills, in the last few stands. This problem does not exist in the older, lower speed
mills. The proces ses for making coated products have of necessity been designed to be
made in equipment that is almost completely enclosed, either because of the safety
aspects (such as the high speeds with which these processing lines operate) or because
of quality requirements which dictate the use of closely controlled atmospheres in the
processing lines. Overall, the processes used in the making of finished steel products
are minor contributors to air pollution.
Particulate emissions have been a point of major concern in the iron and steel
industry and have been the focus for concentrated efforts in the control of air pollution.
However, gaseous emissions, some of which are noxious in character, should also be
considered. Among these are (1) sulfur dioxide, (2) carbon monoxide, (3) fluorides,
and (4) oxides of nitrogen.
Sulfur dioxide is released to the atmosphere as a result of the combustion of sulfur
originating in raw materials and fuels. Although the steel industry attempts to minimize
its intake of sulfur-bearing raw materials for metallurgical reasons, it is estimated that
about 955,000 net tons of sulfur enter the various process segments annually in the mak-
ing of iron and steel in the United States. Of this amount, about 651,000 net tons(68 per-
cent) originates in the coal used to make coke. In accounting for the sulfur in products,
by-products, and gases, about 464,000 net tons (49 percent) of sulfur is contained in
blast furnace and steelmaking slags, and 39,000 net tons (4 percent) in raw steel and
scrap. The burning of fuels (coke-oven gas, fuel oil, coke breeze, tar and pitch)
involves an estimated 374,000 net tons of sulfur annually emitted to .the atmosphere as
sulfur dioxide (or removed in cleaning systems). In addition, there. are about 78,000 net
tons of sulfur unaccounted for.
Carbon monoxide emis sion to the atmosphere has not yet been treated as a problem
of concern in the making of iron and steel, even though significant amounts of carbon
monoxide are generated in various process segments. For example, about 550 cubic feet
(43 pounds) of carbon monoxide are generated in carbonizing 1 ton of coal to make coke;
and depending on specific blast furnace practice, about 16,000 to 25,000 cubic feet (about
1300 to 2000 pounds) of carbon monoxide are generated in the production of 1 net ton of pig
iron. However, the large amounts of carbon monoxide generated in these processes do
not reach the atmosphere because the carbon monoxide is used as a fuel in the heating of
coke ovens, blast stoves, soaking pits, and other in-plant applications. This in-plant
use of carbon monoxide as a fuel is one of the major economies realized by the iron and
steel industry in the manufacture of steel products. Carbon monoxide is also generated
in considerable amounts during the refining of steel in the basic oxygen furnace (BOF).
The BOF off gases may contain as much as 90 percent carbon monoxide. However, here
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-4
again, little carbon monoxide reaches the atmosphere. The carbon monoxide is burned
to prevent pos sible explosions in pollution-control systems, or (as in the case of some
wet-collection systems which pass carbon monoxide through the system) the carbon
monoxide is burned and the products are flared to the atmosphere. The total amount of
carbon monoxide reaching the atmosphere may be as little (in BOF steelmaking) as
33 cubic feet (2.6 pounds) per net ton of steel.
Fluoride emissions have long been suspected as contributors to air pollution be-
cause of the fluoride contained in some local iron ores, and the use of fluorspar (a
calcium fluoride mineral) in the making of steel. Information concerning the amounts
of fluoride emissions reaching the atmosphere have not been reported in the open
literature. A recent report from Germany indicates that present air-pollution control
equipment may be capable of controlling the emission of fluorides to the atmosphere.
Nitrogen oxide emis sions by the iron and steel industry are an unknown quantity.
The use of natural gas by the iron and steel industry has increased from about 5. 7 mil-
lion tons annually in 1958 to about 12.5 million tons in 1967. Combustion of natural gas
is known to be a contributor to nitrogen oxide emissions. The lack of information on the
amount of nitrogen oxide emissions combined with the increased use of natural gas indi-
cates a need for investigative work in this area of air-pollution control.
FUTURE PLANS OF THE INDUSTRY
Changing market demands have resulted in the integrated iron and steel industry
making plans for increasing and establishing new production facilities in the Midwest,
Southwest, and Far West. Some of this has already been implemented with Bethlehem
Steel Company's new plant at Burns Harbor, Indiana, and Jones & Laughlin's new
plant at Hennepin, Illinois. United States Steel Corporation is placing a new electric-
furnace steelmaking facility into operation in the Houston, Texas, area. National Steel
has purchased property in the Corpus Christi, Texas, area, as well as in the San
Francisco Bay area.
Technologically, the steelmaker has long had dreams of developing a process
that would convert iron ore directly into semifinished products without the necessity of
the blast furnace, steelmaking furnace, and casting operation. However, this still ap-
pears to be in the distant future. Work is being carried out at laboratory and pilot-
plant levels but, as yet, none of the processes are in a position to produce the desired
products at a cost competitive with present methods.
Direct-reduction processes that eliminate the need for blast furnaces to reduce
iron ore are only beginning to make their entry': into U. S. commercial steelmaking
practice. Two plants are scheduled to start operations in 1969: one in Mobile,
Alabama, and the other in Portland, Oregon. The steel industry as a whole will un-
doubtedly be watching the companies involved to determine how their two processes
may influence decisions in the integrated iron and steel industry with respect to future
expansion plans.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-5
AIR-POLLUTION-CONTROL RESEARCH AND DEVELOPMENT
Research and development efforts are required to produce better emission-
control equipment and systems, as well as to improve processes themselves in the iron
and steel industry. Major needs include the development of sampling and analytical
equipment that will operate on a continuous basis as a means of providing better design
data to equipment manufacturers, better control data for steel plants and improved
methods to monitor and measure emis sions.
Temperature-sensing instrumentation is too slow in basic oxygen steelmaking
practice, with the result that particulate emissions occur at the start and end of the
oxygen-blowing periods. Instrumentation with faster response is necessary to correct
thi s deficiency.
Additional basic information is needed on the electrical properties of particulate
emissions and how they are affected by chemical composition; and studies are needed
on more effective methods of conditioning metallurgical dusts for collection by electro-
static precipitators. For wet scrubbers, informatio~ is needed and work needs to be
done to develop the necessary criteria to apply the wet system to the removal of ob-
noxious gases from the various processes in which wet scrubbers are used.
In the integrated iron and steel industry, research and development efforts are
needed on a first-priority basis for the reduction of particulates and gaseous emissions
at coke ovens and sinter plants. Another high-priority problem is the evolution of hy-
drogen sulfide from blast-furnace slags. An economical method for removing hydrogen
sulfide from coke-oven gas is still another subject needing development. Also in the
high-priority category is a need for research to lessen air-polluting emissions of par-
ticulates generated in the loading, transfer, and storage of the vast amounts of bulk
raw materials. Longer range studies are required that will be directed toward the
possible development of continuous-coking processes as a means for eliminating pres-
ent coke-plant emission problems. Other subjects in need of research are the control
of emissions from casting and flushing operations at the blast furnace, preparation of
ingot molds, and methods of collecting emissions generated during the casting of ingots.
General Recommendations for Phase II Research on
Technical Aspects of Control of Air-Pollution Emis sions
From the Integrated Iron and Steel Industry
The present report deals with Phase I, which since its inception has been expected
to be followed by a Phase II research effort to fill gaps in the information developed
during Phase I. Analysis and judgment by the Battelle staff of all the information accum-
ulated during Phase I leads to the recommendation that Phase II research be conducted
in the following three subject areas, each of which will be discussed in more detail:
(1)
Improved control of emissions from the manufacture of
metallurgical coke
(2)
Instrumentation for in-plant emissions measurement and control
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-6
(3)
Improved characterization of emissions from certain process
segments.
These subjects have been selected because of (a) their importance to future control
of air pollution, and (b) their apparent amenability to research action.
Following this general presentation, including discussion of reasons why some
other subjects have been assigned lower priority for research attention, specific infor-
mation is given on seven new projects that could be undertaken to implement these
recommendations.
Improved Control of Emis sions
From the Manufacture of
Metallurgical Coke
In Section VII of this report, two of the seven proces s segments identified as hav-
ing first-priority research needs involve coke plants - oven charging, and pushing of
coke. The conventional by-product coke-oven battery as it exists today as the traditional
method for making metallurgical coke is viewed by Battelle and by many other observers
of the situation as inherently unsuitable for close control of air-polluting emissions. In
addition to charging and pushing, problems of air pollution of quantitatively undetermined
magnitude are represented by gaseous emissions from lids and door seals, emissions of
particulates during quenching of the coke, handling of the coke, and the coke-oven-gas
system; all with the usually complex by-product system superimposed as an air-quality
problem. A number of approaches mentioned throughout this report attempt to deal with
pollution from one or more of these sources in the coke plant. However, until it can be
demonstrated that these coking proces ses would have an impact on reducing air pollution
from cokemaking, Battelle views many of these approaches as expedients aimed at mak-
ing the coke plant tolerable until really satisfactory solutions can be found. The air-
pollution problem associated with coke plants is of sufficient magnitude and the controls
so inadequate that it is rare that anyone knowledgeable on the subject will deny that the
coke plant is a major polluter of air. The severity of emissions is easy to see qualita-
tively because a number of coke-oven plants physically are sufficiently remote from
other operations in a steelworks that there is little question as to which process segment
in the plant is producing contamination.
Two approaches seem to warrant attention in Phase II; the first deals with improved
methods of controlling emis sions from present-type coke ovens, and the second involves
entirely new methods of making metallurgical coke. Although a third remedy is con-
ceivable and is receiving some attention, there appears to be little that could be done in
the way of air-pollution research to promote this third alternative. The third alternative
is direct reduction of iron ore - the reduction of iron ore by methods that bypass the
blast furnace, and thus do not require metallurgical coke.
It is recommended that Phase II attention be given to (I) improved systems for
controlling emis s ions from conventional slot-type coke ovens, and (2) new and improved
methods for making metallurgical coke by methods other than conventional slot-type
ovens, over which better control emissions could be exercised. More specific reasons
for such a study, background, and the recommended scope are given in more detail under
under the heading IIProject 1: Processes for Making Metallurgical Cokell.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-7
Throughout consideration of new methods for making metallurgical coke, one
encounters the problem that the best way to verify the suitability of a coke is actually
to use it on a commercial scale in a blast furnace. This is an expensive procedure for
two reasons: (1) some thousands of tons of coke are needed for a full-scale test, and
few pilot plants for coke have the capability to produce such quantities, and (2) the blast-
furnace tests are expensive to perform, difficult to evaluate, and involve the risk of
damaging the multimillion dollar blast furnace. A reasonable step between small-s cale
research and full commercial operation involves the use of small "experimental" blast
furnaces. There are two such furnaces in the United States. One is owned by U.S.
Steel Corporation, and the other by the United States Bureau of Mines. Both are in the
.t""lttsburgh area. U. S. Steel operates theirs when they have an appropriate problem.
It is conceivable that it could be used under contract for investigations for other parties.
The Bureau of Mines furnace has not been operated for several years and is being con-
sidered for disposal.
As suming that no small experimental blast furnace is available for evaluating new
types of coke, there are some possible alternatives. The chemical composition of the
coke gives some indication of its suitability for blast-furnace use. Beyond this, there
is a need for information on its combustibility and reactivity. Laboratory tests can be
used and devised to give good information on this aspect, and some useful information
can be developed from small-scale shaft-furnace tests that involve perhaps several
hundred pounds of coke per test. Finally there is the question of the mechanical
strength of the new coke at the temperatures involved in the blast furnace. A procedure
was developed for determining such strength, and was applied to research for three
steel companies. ':' The method is still applicable. These alternatives taken together
can form a sound basis for planning commercial-scale runs on a new coke with high
confidence.
Instrumentation for In-Plant
Emissions Measurement and Control
Instrumentation is considered from the viewpoint that the general air-pollution
problem involves needs for two types of instrumentation. The first is equipment that
generally is most useful externally to the plant to determine levels of pollution existing
in a general area, and to gather general information on air quality. That category is
considered to be outside the scope of the present study. The second type, which is
within the present scope, is anticipatory equipment which is useful within the steelworks
in connection with process control and prompt measurement and detection of undesirable
emissions, either particulate or gaseous. Preferably such instrumentation should be
highly reliable for operation continuously in ducts under what often are trying conditions
of temperature and 'corrosion. Continuous operation should be such as to give fast read-
outs of the level and nature of emissions, and offer predictive aspects. The sensors
might be interlocked with the proces s so as to accomplish some type of corrective
action rapidly before the air is polluted. The instrumentation should contribute to the
avoidance of air pollution; not only, such as is sometimes now installed, to "protect the
equipment'!. Although the emission-control equipment usually requires some type of
. John Varga, Jr. and H. W. Lownie, Jr., "Influence of Temperature on Mechanical Strength of Coke" Transactions of American
Foundrymen's Society, 1956.
M. W. Lightner (Assistant Vice President, United States Steel Corp.), "Burdening the Blast Furnace"; Steel, February 7, 14,
21, 28, 1955.
John Varga, Jr.. "High-Temperature Evaluation of Blast Furnace Coke". Proceedings of AIME Blast Furnace, Coke Oven, and
Raw Materials Conference, 1960.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-8
protection, this frequently takes the form of a bypassing
at a frequency high enough to interfere with air quality.
needed, therefore, would perform three functions:
system that can cause "puffingl'
Instrumentation of the type
(1)
Measure the nature and amount of emis sions between the proces s
segment and the collection system; do this in such a way as to
provide signals that some undesirable change has occurred in the
process; and give this readout fast enough (or act through inter-
locks) so that corrective action can be taken in connection with
operation of the process.
(2)
Provide measurements of conditions that if ignored would destroy
or lower the efficiency of the emission-control equipment, and
provide signals (or act through interlocks) in such a way as to
protect the equipment without simple bypas sing of the effluent into
the atmosphere.
(3)
Measure the nature and amounts of emissions from the emission-
control equipment. If the emission-control system is properly
designed and adequately maintained, and if the other instrumenta-
tion does its job properly, then the sensors at this stage merely
confirm the proper operation of the overall system. If any up-
stream portion of the total system is inadequate or fails, then
this location will signal the need for immediate attention to up-
stream process, instrumentation, and collection equipment.
During the course of this study, the investigators have been impressed by the
apparent lack of availability of such instrumentation that meets the standards of the
steel companies. The nature of equipment that is available and that is under develop-
ment both in the United States and in Europe should be investigated in more detail in
Phase II. Also the experiences of steel companies (and other heavy industry) both in
the United States and overseas with the present designs of instrumentation should be
investigated and analyzed to determine both the good features and the deficiencies of
present equipment, and to prepare more specific recommendations for research to
develop new acceptable instrumentation. "Project 4: Evaluation of Instrumentation for
In-Plant Emissions Measurement and Controlll deals with this subject.
Improved Characterization of
Emissions from Proces s Segments
One of the objectives of this study was to obtain and interpret data and information
on the amount, size, composition, electrical properties, and corrosive properties of
particulates; composition and amounts of gaseous effluents; and to develop material
balances that would permit a definition of the nature and amounts of emissions generated
by ironmaking and steelmaking process segments in various parts of the United States.
Attainment of the objective was handicapped by the sparsity of reliable information that
could be obtained. In particular, data were generally inadequate for the preparation of
complete material balances directed at the identification of emissions problems.
Although the literature is fairly extensive, much of it is not adequately definitive of the
conditions under which samples were taken, and some of the literature is frankly
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-9
skeptical about the accuracy of the data. The European literature generally is more
complete and more extensive than American literature on this subject, but there is a
hesitancy to apply foreign results to the American iron and steel industry because of
differences in operating practice. The shortage of reliable information on American
practice stems from two factors: (1) it is Battelle's opinion that reliable data actually
are rarer than many persons care to admit, and (2) some potential sources of informa-
tion consider the data proprietary and confidential.
Battelle recommends that research be conducted to fill the gaps in the information
on the amount, size, composition, and other properties of particulate and gaseous
emissions from steelworks process segments. Such an activity will for the most part
require active cooperating participation by at least one steel company. Given access
to the processing equipment itself, investigators would be able to obtain the necessary
samples and make the necessary analyses.
The process segment that appears to Battelle to warrant first attention in this
regard is the sintering plant. Although there is some information presented in this
report on particulate emis sions from sintering plants, their gaseous emis sions appear
to warrant particular attention. "Project 3: Characterization of Emissions from
Sintering Plants " is an example of a high-priority project of this type.
Characterization of the mechanism and nature of emis sions of sulfur-bearing
gases from blast-furnace slag during tapping from the blast furnace, during handling
in the cast house, and during weathering in outside dumps has been recognized as a
problem by the steel industry. Since 1966, the American Iron and Steel Institute has
been sponsoring research on this subject. The research has the two objectives of
identifying the mechanism of formation of the undesirable gases, and developing means
for their suppression. Progress has been made, and results have been published. ~,
More research is needed. With additional funding, the research could be conducted
at an accelerated rate. Then, possibly, the results could be placed into practice some
years earlier than would otherwise be the case. The problem is particularly worthy of
attention because the evolutions occur over wide areas in diffuse amounts. When the
slag is disposed of in dumps, the area can involve many acres; a condition that makes
collection of the emis sions by conventional means extremely difficult, perhap's impos-
sible. "Project 2: Sulfur-Bearing Emissions from Blast-Furnace Slag" deals with
this problem.
Other Subjects of Research Interest
In addition to the three foregoing subject areas that deserve (in Battelle's judg-
ment) primary consideration, three additional subjects that warrant attention are
(1) wide-area emissions, (2) sulfur content of fuels and of products of combustion, and
(3) incineration of wastes. The reasons for assigning these subjects to a lower priority
when viewed in the context of the integrated iron and steel industry are given below.
Wide-Area Emissions. These are emissions that occur generally over a wide
area, often intermittently. Examples include the smoke and fume generated during
the pouring of ingots and the kishing of hot metal as it is tapped from the blast furnace
"F. H. Woehlbier and G. W. P. Rengstorff; "Preliminary Study of Gas Formation During Blast-Furnace Slag Granulation with
Water"; Annual Meeting of Air Pollution Control Association; June, 1968.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-IO
and as it is transferred from vessel to vessel in the plant. These emissions occur over
a wide area, often under physical conditions that make it difficult to install adequate
hooding and ventilation. Often, overhead structures, and especially access to the area
by overhead cranes, interfere with hooding. The technology of the process segments is
such that it is unrealistic to expect that major improvement in emission control can be
accomplished by a change in the process itself. Therefore, hooding or general evacu-
ation of the building using large amounts of air exhausted through a suitable emissions-
collection device appears to be the only possible solution. Although such a system
would intrude upon the usual layout in a steel plant, the fact of the matter is that suitable
equipment can be engineered for this purpose. Therefore, Battelle takes the view that
this subject is a technical problem that can be solved by the application of existing
engineering know-how. For this reason, Battelle does not consider this subject a prime
candidate for research. The problem more properly belongs within the scope of engi-
neering design and construction of steel plants.
In this general subject area, however, there are two facets deserving of research
attention. These are:
(I)
Experimental investigations of improved means for holding in
place against weather and wind the fines stored in outside piles,
especially coal piles. "Project 6: Control of Wide-Area
Emissions from Storage Piles" deals with this subject.
(2 )
Pre-engineering and cost estimates for systems to collect and
control wide-area emis sions. "Project 7: Design and Costing
of Systems for Collection and Control of Wide-Area Emissions
Inside a Steelworks" deals with this subject.
Sulfur Content of Fuels and of Products of Combustion. Sulfur is an undesirable
element in iron and steel. Therefore, steel companies have strong technological
reasons to maintain the sulfur content of their raw materials at the lowest possible
level consistent with considerations of cost. However, the problem of air pollution
from the combustion of sulfur-bearing fuels exists in the steel industry as it exists in
many other industries that burn such fuels.
Steel plants generate products of combustion at numerous locations scattered
often over hundreds of acres of a steelworks. NAPCA is well aware of the problems of
dealing with the combustion of sulfur-bearing fuels, the premium that must be charged
for low-sulfur fuels, and the preliminary stage of development of methods developed to
date to control sulfur-bearing gaseous emissions from the combustion of sulfur-bearing
fuels. The treatment of coals, coke -oven gas, and petroleum products to lower their
sulfur content has been studied on many fronts for many years, so that technically
feasible methods have been worked out. Insofar as is known to the present investigators,
no desulfurization process exists that meets today's criteria for economic feasibility.
However, the present criteria may have to be changed if the nation is to have cleaner
air. The general subject of desulfurization of steelworks fuels and the collection and
control of sulfur-bearing gaseous emissions from combustion is not included in the list
of recommendations for Phase II research because the general problem goes far be-
yond the steel industry. Because of the nationwide severity of this problem involving
many industries, efforts should be continued to attack this problem vigorously. 1£ an
economically practical solution can be developed for the general case, its application
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-II
to integrated steelworks would substantially lower the overall contribution of steelworks
to air pollution by gaseous effluents.
Incineration of Wastes. Steelworks generate much combustible waste that often is
burned on the premises. Because suitable incinerators and practices for their operation
are considered to be available to solve this problem, this subject is not included in the
present recommendations for research. However, as in the case of sulfur-bearing fuels
and their products of combustion, the general problem of incineration should continue to
be investigated on a base broader than that of the steel industry alone.
Specific Recommendations for Phase II Research on
Technical Aspects of Control of Air-Pollution Emissions
From the Integrated Iron and Steel Industry
Seven specific subjects that in Battelle's opinion warrant further research are now
described. The subjects are not necessarily in any formal order of priority, but there
has been an attempt to place the subjects in approximate order of urgency as viewed by
the present investigators. Other investigators will have other judgments on ranking.
Particular attention will be given to Project 1 - "Processes for Making Metallurgical
Coke" - because it is in Battelle I s judgment most urgent.
Project 1: Processes for Making
Metallurgical Coke
Metallurgical coke is high-strength coke used in blast furnaces to smelt iron ores
to produce molten iron that is the main ingredient for steelmaking. In the United States
in 1967, about 56 million tons of metallurgical coke were used in about 170 blast fur-
naces to produce about 87 million tons of pig iron. Virtually all of this coke was made
by the American iron and steel industry in coking plants in or associated with integrated
steelworks. The production of this amount of coke required about 83 million tons of coal
as the raw material. This was about 93 percent of all the coal used by the American
integrated iron and steel industry in 1967 (which was a typical year). These statistics
show that the American steel industry annually converts very large amounts of coal to
metallurgical coke.
The manufacture of coke by the American integrated steel industry is widespread
geographically. Of the total production in 1967; about 38 percent was made in
Massachusetts, New York, plus Pennsylvania; about 13 percent in Indiana; about 12 per-
cent in Maryland plus West Virginia; about 12 percent in Ohio; about 10 percent in
Tennessee, Alabama, Texas, plus Oklahoma; about 6 percent in Michigan plus Minnesota;
about 5 percent in Colorado, Utah, plus California; and about 4 percent in Illinois.
Therefore, problems involving air pollution during the manufacture of metallurgical coke
are of national interest.
The blast-furnace process is used in the United States as the method to produce
over 95 percent of the pig iron on which the steel industry is based. Advancements in
blast-furnace technology (such as use of higher grade ores and agglomerates, high blast
temperatures, and injection of hydrocarbons) are lowering, year by year, the amount of
coke needed to make 1 ton of pig iron, but the rate of production of pig iron is increasing
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-12
annually so that the total tonnage of metallurgical coke required per year has been
remaining about constant. The present state of the technology of the blast-furnace pro-
ces s does not include any factor that is expected to eliminate the high degree of depen-
dence of the blast furnace on metallurgical coke. The main technological factor affect-
ing the future need for metal from the blast furnace is the growth of the basic oxygen
furnace (BOF) as the predominant method for making steel. Of the three major steel-
making processes, the BOF process conventionally requires about 70 percent of its
charge in the form of hot metal from blast furnaces; the open-hearth process conven-
tionally uses about 40 to 60 percent blast-furnace hot metal; and the electric-furnace
process usually uses no hot metal. The BOF process (which uses high amounts of blast-
furnace hot metal) has been growing rapidly in the United States, and elsewhere through-
out the world. Invented in Europe in 1952, the BOF process accounted in 1968 for about
37 percent of the production of raw steel in the United States, and is expected to reach
about 60 percent by 1980. This growth of the basic -oxygen process will put continuing
pressure on demands for increasing amounts of hot metal from blast furnaces, which
in turn will have to be supplied with increasing amounts of metallurgical coke.
The chances are low for development in the next decade or two of a commercially
important method for reducing iron ore by a process other than the blast furnace in the
quantities needed by the steel industry. Processes that eliminate the blast furnace are
generically and loosely called "direct reduction" processes. They make use of rotary
kilns, traveling grates, batch retorts, and electric smelting furnaces. Instead of
metallurgical coke, they make use of coal, reformed natural gas, and/or electricity.
A few percent of the world's production of iron from ore is made by such processes, and
increasing attention is being given to them as possibilities for the future. Several new
small plants utilizing such processes are now under construction near Mobile, Alabama;
Portland, Oregon; Auckland, New Zealand; and in Korea. These plants deserve careful
evaluation for their future potential in affecting the importance of the blast furnace, but
it is unlikely that such processes will replace as much as 20 percent of blast-furnace
pig iron in the United States within the next 15 years. Therefore, although such pro-
cesses have potential commercial significance during the next decade for special situ-
ations, the blast-furnace proces s will remain for many years as the backbone of the
American iron and steel industry.
The manufacture of metallurgical coke is conducted in batteries of slot-type by
product ovens. The nature of these ovens is such that they give rise to air pollution.
The research done since June, 1968, on Contract PH 22-68-65 has identified coke plants
as probably the largest source of uncontrolled emissions to the air by the American
steel industry. Because of the shortage of accurate and reliable data and measurements
on emissions from coke plants, it is not possible at this time to prove the foregoing
statement quantitatively, but knowledgeable persons in the steel industry and affiliated
with it almost completely agree that coke plants represent the largest source of uncon-
trolled emissions by the steel industry. When the Industry Liaison Committee asso-
ciated with Contract PH 22 -6 8-6 5 was asked to identify air-pollution problems deserving
of further study by NAPCA, every respondent mentioned some aspect of coke-plant
operation. The most frequently mentioned problem was that of controlling the gaseous
and particulate matter when the blend of fine coals (about 1/8 inch and down) is charged
through the top into the incandescent coke oven.
In Section VII of this technological report on Contract PH 22-68-65, an evaluation
is made of the process segments in the steel industry on which research and development
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-13
may best be undertaken to achieve improved control of emissions to the air from the
integrated iron and steel industry. Process segments are evaluated from the stand-
points of (1) severity of emissions, (2) degree of present control of emissions, and
(3) availability of methods for control. Of 109 sources of emissions evaluated, 7 are
rated to deserve first-priority attention, 7 are given second priority, and 8 are given
third priority. Of these 22 high-priority sources, 6 are associated with coke plants
(2 in each of the three priority classifications). One of the first-priority subjects is
control of gaseous and particulate emis sions during charging of coal to the coke ovens.
It_is suggested that further research attention be given to improved systems for
charging coke ovens of the present design. An adequate system should be capable of
injecting coal into the incandescent ovens under conditions that avoid the emission of
particulates or gases during charging. Some research and development being conducted
by others along these lines are, for purposes of discussion, considered to fall into three
subdivisions.
The first subdivision is systems under development in Europe for this purpose.
Although appare'ntly tried to some degree in France and Canada, Germany seems to be
the source of these systems. The European systems have been examined by American
steelmakers, who so far have not seen fit to adopt them generally. American steelmen
have stated several objections to the application of these systems in the United States.
One of the objections is low reliability. One member of the steel industry said that he
saw seven installations in Europe, of which one was operational and six were shut down
undergoing repairs or modification. No further information was available from this
.source. Other American steelmen have observed that the height of some of the equip-
ment (called a larry car and running on tracks on top of the ovens) is too great to fit
under existing super-structure on American ovens. Other observers have questioned
whether present American ovens can support the additional weight of the European
systems. There also is an unconfirmed report that two installations of European
systems on Canadian ovens resulted in damaging explosions. On Contract PH 22-68-65,
it has not been possible because of contract limitations on time and money to investigate
these European systems in sufficient depth to arrive at independent evaluation of their
potential for the American situation. Suitability of any system that is installed on
American ovens is important because installation of inadequate systems would not only
be costly but could delay by several years the installation of really effective systems.
The second subdivision of systems that might be used to charge existing designs
of coke ovens is identified here, for want of a better name, as lithe AISI system". The
American Iron and Steel Institute has had the subject under study by at least one of its
committees. As a result of that study, a member of the Industry Liaison Committee
has told NAPCA representatives that the AISI intends to approach NAPCA soon with a
proposal for financial support for development of the "AISI system". Based on the
information presently available to Battelle, the AISI system probably is a modification
of one or more European systems. However, this is speculation. If the AISI (or any
other organization) does ask NAPCA for support for a new system, the cost of a demon-
stration unit would be large. Because of this, any new system proposed for development
should be compared with other alternatives that are possible. There would appear to be
some merit, however, ih supporting development of a system proposed by the American
steel companies, because this would make the steel companies a direct part to finding a
satisfactory solution to this challenging problem.

The third subdivision of this category suggested for research attention is an oven-
charging system under development by the Allied Chemical Company, Wilputte Division.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-14
This system attempts to accomplish the same thing as the foreign systems (cleaner
charging ef the coke oven), but also includes preheating of the coal - a technique that
should decrease the coking time and increase the productivity of any oven on which it is
installed. Allied has not released detailed information on this system (except that it
involves preheating and pipeline charging), but is installing the system on a battery of
ovens of its Semet-Solvay Division at Ironton, Ohio. It is suggested that NAPCA obtain
more information on this system, and evaluate it in connection with improved informa-
tion on the other two subdivisions in this category (the European equipment and the AISI
system).
Turning now from the foregoing innovations aimed at better control of emissions
from conventional coke ovens, it is further suggested that research attention be devoted
to what Battelle considers the real long-term solution - the development of cokemaking
methods that inherently lend themselves to a high degree of control over emissions.
Battelle views such a system as being probably continuous so as to avoid the batch-
charging and batch-ejection operations that contribute so much to air pollution from con-
ventional ovens. Also, a new process should be designed from the ground up as a closed
chemical system with specific attention to the avoidance of air pollution. At least one
coal company in the United States has a process that should be investigated in more
detail. This is the Chemcoke process of the Peabody Coal Company. Further investi-
gation may show that other coal companies have something similar, or that the process
is unsuitable for the manufacture of metallurgical coke. A demonstration plant for the
Chemcoke proces s was built in 1961 at Columbia, Tennes see, and continues to operate
for process development and to produce coke for chemical plants in the vicinity. The
plant is rated at 100 tons of coke per day, which is quite large for a pilot plant. The
process is thought to involve preliminary treatment of coal on a traveling grate in a
dutch oven, followed by finishing in a shaft furnace. Most of the product has not been
coke of metallurgical quality, but sources within Peabody Coal have said that they have
produced coke that appears to meet the requirements of the blast furnace, but that this
coke has never been tested in blast-furnace use. From these remarks, it is concluded
that the process deserves more investigation and analysis. An official of Peabody has
agreed orally to cooperate in further study of this subject. Based on the results in the
demonstration plant, Monsanto Chemical Company in 1966 built a similar plant (rated
at 500 tons of coke per day) at Decatur, Alabama, which reportedly is producing 550 tons
of coke per day as reducing agent for phosphorus furnaces, an application considerably
less demanding as regards mechanical strength than blast-furnace coke. The Chemcoke
process at both plants is alleged to be continuous, requiring a throughput time of be-
tween 1 and 2 hours, and "completely clean" as to emissions. These allegations need
confirmation, as do the prospects for the ability of this process to make metallurgical
coke.
At the same time that the Peabody proces s is studied, an up-to-date evaluation
should be made of the present status of the FMC process that produces a briquetted
coke called FMC coke. This process has been under development by the FMC
Corporation for more than a decade, and resulted in the construction of a demonstration
plant at Kemmerer, Wyoming, to produce coke mainly for electric-furnace use. A few
years ago, FMC and United States Steel cooperated in a venture to produce blast-furnace
coke and evaluate this proces s. Such coke was made and evaluated in a small blast
furnace, but the results were not sufficiently good that U.S. Steel was influenced to com-
mit further large investment in research on this process. About 2 years ago, a larger
scale evaluation was conducted in an Armco blast furnace at Hamilton, Ohio. Dusting
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-IS
of the FMC Coke during transportaion and handling was a problem, but FMC officials
believe that the trouble was in the use of marginal briquetting equipment, and would not
recur if briquetting rolls of proper size were used. The Kemmerer plant is now
operating at a rate of 150 to 200 tons of coke pe r day for electric phosphorus furnaces
of FMC Corporation at Pocatello, Idaho. In principle, the FMC process treats the coal
in a fluidized bed and then agglomerates the product into briquettes. The method is one
that seems to lend itself to close control of emissions.
It is recommended that new research be undertaken to make a nonexperimental
technological evaluation in more depth of (1) improved systems for charging coal, pushing
coke, and quenching coke made in conventional ovens so as to achieve improved control
of gaseous and particulate emissions, and (2) new processes that have potential as means
for manufacturing metallurgical coke under conditions that lend themsleves better than
conventional coke ovens to the overall control of gaseous and particulate emissions.
Based on the new information and data that would be accumulated during such research,
it would then be possible to make an independent authoritative and meaningful evaluation
of the future potential for such systems, and to prepare recommendations for additional
hardware research that might be needed, leading perhaps to recommendations for hard-
ware demonstration units or pilot plants that may be appropriate.
Drawing upon the foregoing background, it is suggested that the scope of new
studies in Project 1 be as follows:
I.
Improved Systems for Charging Coal, Pushing Coke, and Quenching
Coke Made in Conventional Coke Ovens
A.
Acquire new information by direct visits to offices,
laboratories, and/or installations of U. S. builders
of coke ovens.
(1)
Allied Chemical Company
(2)
Koppers Company
(3)
AnDid Corporation
B.
Acquire new info rmation by dire ct vis its to office s,
laboratories, and/or installations of European and
Canadian companies.
(1 )
Builders of coke ovens and coke-oven

equipment
(2 )
Steel companies with pertinent experience
C.
Evaluate the suitability of the technology for application
to the present domestic oven design and coal blends.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-16
II.
New Processes With Potential for Manufacturing Metallurgical Coke
A.
Acquire new information by direct visits to offices, laboratories,
and/or installations of American companies and organizations
with pertinent experience. New processes probably will be
continuous, but potential proces ses will not be excluded solely
because they are of a batch type.
(1)
Peabody Coal Company
(2 )
Mons anto Chemical Company
(3)
FMC Corporation
(4)
United States Bureau of Mines
(5)
Others as may be determined during the study
B.
During visits in Europe under 1, search out status of European

technology on such processes
C.
Evaluate the suitability of the technology for application to the
smelting of iron ore in blast furnaces in the United States.
III. Overall Evaluation and Recommendations
A.
Evaluation
B.
Recommendations
(1)
New nonexperimental studies
(2 )
New research on hardware
(3)
For demonstration units or pilot plants
The foregoing scope has been prepared with the knowledge that Japanese steel-
makers also face severe problems of air pollution from cokemaking operations. During
the last year, Japanese visitors to the United States have made a point to observe U.S.
practice along these lines, but have divulged little about the nature of solutions that they
are considering or developing. During the course of the study outlined above, further
information might be forthcoming about the situation in Japan. If that occurs, there may
be improved reason for a later effort to acquire more information by direct visits to
Japan, but such visits are not included in the present scope.
The conduct of Project 1 can be expected to require 5 or 6 months of elapsed time,
and to require about 10 man-months of professional time (including supervision), plus
costs for travel and support services.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-17
Project 2: Sulfur-Bearing Emissions
from Blast- Furnace Slag
The American integrated iron and steel industry annually receives in its raw
materials about 1 million tons of contained sulfur. About 85 percent of this incoming
sulfur content is contained in coal; mostly in the coal that is converted to coke. All of
the contained sulfur that enters the steelworks leaves the works in some form. Only
relatively minor amounts of sulfur are contained in the iron, steel, and steelmaking
slags. The major identifiable fraction is the amount of sulfur contained in the blast-
furnace slags. About 30 million tons of solid blast-furnace slag containing an average
of about 1.5 percent sulfur is generated annually. This is almost half of the total
amount of contained sulfur entering the American integrated steelworks.
Although a statistical analysis of th~ sulfur content of American blast-furnace
slags is lacking, the range of sulfur contents appears to be from substantially less than
1 percent to mo re than 2 percent, with an average of about 1.5 percent.
What happens to the sulfur in molten slag that is tapped from a blast furnace?
Some burns to S02 as the slag flows down the runners into granulators or slag thimbles;
some converts to S02 and H2S during granulation in water or in other proces ses where
the molten slag contacts water; some is dissolved in the water and mayor may not pre-
cipitate out; and some remains in the solid slag. It is known that during subsequent
contact with water (as during weathering on a slag dump), additional sulfur is evolved
into the air. The amount that is evolved is believed to be small, but factual data are
not yet available. The fact that the evolutions do occur at these points has been
established, but the relative amounts have not.
The amount of sulfur lost at various stages in the handling of blast-furnace slags
is influenced by the nature of the handling process, the composition of the slag, the
temperature of the slag,. and presumably by other factors. It is known that some gran-
ulation processes yield a solid product with about half the sulfur content of the molten
slag exiting from the blast furnace. On the other hand, some of the "hard slag'l pro-
cesses might yield a product with a sulfur content reduced only from about 1. 5 percent
in the molten condition to about 1. 3 percent in the solid state, and most of this loss
might report to the water used to cool the slag.
Recognizing the dearth of information on the mechanism of evolution of sulfur from
blast-furnace slags, the American Iron and Steel Institute has been conducting experi-
mental research on this subject since 1966, and the research is continuing. The experi-
mental work is proceeding, but does not include identification of amounts of sulfur being
evolved. The emphasis is on the fundamentals of the mechanism of the evolution.
Although there remains much to be learned, it appears likely that the research will be
extended to include some measurements on a full-size blast-furnace operation.
Battelle estimates that commercially useful results from studies on the mechanism
of sulfur evolution from blast-furnace slags could be several years in the future at the
present rate of effort on this subject. An expanded rate of effort would be expected to
lead to earlier solution of the problem. The program also could be usefully expanded to
determination of the amounts of sulfur emitted under various conditions and at the sev-
eral points of handling or treatment of the slag.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-18
The conduct of Project 2 can be expected to require at least I more year of
experimental work at an annual rate of about 20 man months of professional time (includ-
ing supervision), plus costs for support services (including considerable chemical
analyses) .
Project 3: Characterization of Emissions
from Sintering Plants
About 50 million tons of sinter is made annually in the United States for charging
to blast furnaces. Most of this sinter is made within the perimeter of'integrated steel-
works. The emis sions include particulates and gases. The particulates are generally
amenable to control by scrubbers, precipitators, and baghouses (but at substantial cost
for equipment and operation). One of the major gaseous evolutions is S02. Sulfur enters
the system in the ore, in the coke used in the sinter mix, and in fuel oil (when fuel oil is
used). On the average, the amount of sulfur entering the system amounts to about
2 pounds per net ton of sinter produced. This is about 50,000 tons of contained sulfur
per year. About 50 to 60 percent of this sulfur reports to the sinter, and about 30 to
40 percent probably appears in the off gases. The off gases may also contain fluorides
under some conditions, but the amounts and conditions under which significant amounts
of fluorides are evolved have not been identified adequately.
The present research has identified some of the characteristics and amounts of
pollutants evolved from sintering plants, but the fund of reliable information requires
reinforcement if the nature of sintering machines as polluters of air is to be understood
thoroughly. Therefore, more research of an experimental nature is recommended.
Measurements should be made in at least two sintering plants, preferably more.
The plants should be representative, but could involve at least two installations of gen-
erally similar conditions (so as to serve as a check on each other), and/or two or more
different practices (to define a range). The different practices might well involve
unfluxed and fluxed sinters of different basicities.
,
For each plant included in the project, the following is the minimum that should
be dete rmined:
(1)
Prior to experimentation, define the plant and the system under
study, including the nature and performance of air-pollution con-
trol equipment that is installed. Prepare flow sheets and material
balances, including air. Identify unknowns in the system.
(2)
Measure unknowns. Sample and characterize the amounts and
compositions (including size distribution) of particulates and
gases at various locations ahead of and behind any air-pollution
control equipment.
(3)
Develop improved and more definitive flow sheets and material
balances than was possible in (1). Prepare an improved state-
ment of the nature of the problem of control of emis sions from
s inte ring machine s .
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-l9
(4)
Prepare recommendations for further research, if needed; and/or
for construction of hardware or demonstration installations.
Project 3 requires that the investigators have virtually free access to the sinter-
ing plants being studied. Problems of sampling are expected to be considerable, and
costs for particle-size and chemical analyses probably will be high. If two plants are
involved in the study, Battelle estimates that about 9 months of elapsed time will be
required, and that the investigation will require about 24 man-months of professional
time (including supervision), plus costs for support services (including laboratory
technicians and chemical and mechanical analyses of particulates and gases).
.t"'roject 4: Evaluation of Instrumentation for
In-Plant Emissions Measurement and Control
Investigation to date shows that in-plant control of emissions from steelworks is
hampered by a lack of suitable sensors, instrumentation, and control, and by lack of
correlation of such equipment with that used external to the plant by abatement personnel.
.lne nature of the problem is discussed earlier in Section II. New research on this
problem is recommended.
New research in Project 4 should encompass the following scope:
(l)
Develop from existing information a set of written hypotheses
covering the present state of the art, including capabilities
and limitations of present equipment.
(2)
Modify or confirm each hypothesis by visits within the United
States. The hypotheses will serve as a definite focal point to
guide information seeking. The visits will develop new infor-
mation on what is being used, what is available, and what the
needs are. Visits should include steel plants, instrument
companies, and abatement agencies, plus other sources of
information as they develop during the "investigation.
(3)
Prepare a draft report on the state of the art in the United
States and review this report with interested and affected
parties in industry.
(4)
Visits as in (2) above to Europe and to Japan.
(S)
Prepare final report, including analysis of the present situation,
recommendations for new research, and recommendations for
hardware development or demonstration units as may be appropriate.
Project 4 is expected to require about 1 year of elapsed time. This includes pro-
vision for overseas travel and for interaction with industry on preparation of reports.
The study is expected to require about 28 man-months of professional time (including
supervision), plus costs for travel and support services.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-20
Project 5: Development of Improved
Flow Sheets and Material Balances
for American Integrated Steelworks
In the conduct of Contract PH 22-68-65, flow sheets and material balances on
proces s segments in the steel industry were developed from the open literature, from
information accumulated during visits, and from Battelle's knowledge of the steel indus-
try. Flow sheets and material balances prepared in this manner, generally descriptive
of the situation that exists on the average for the steel industry viewed as a single entity
on a national scale, have deficiencies and will introduce inaccuracies if applied to smal-
ler segments of the industry, as for example the steelworks in a particular geographic
area.
In April, 1969, the Board of Directors of the American Iron and Steel Institute
offered to provide certain AISI quarterly operating reports for a one-time nonrecurring
use for input to NAPCA studies of the iron and steel industry. Specified pages or sec-
tions of the reports would be provided with all individual company references deleted
by mechanisms still to be worked out. None of the operating data would be deleted.
The AISI quarterly reports are prepared by the AISI from operating data submit-
ted by process segment and by individual furnaces by companies that comprise almost
the entire American iron and steel industry. A few companies do not report such infor-
mation even to the AISI. These quarterly operating reports are closely held within the
AISI. Approval in this case was secured by the efforts of the Industry Liaison Committee
associated with Contract PH 22-68-65. Unfortunately, insufficient time was available
after their release to make use of these reports on Contract PH 22-68-65.
Battelle recommends that Project 5 involve activity as follows:
( l)
NAPCA would provide the investigators with definition of a number
of geographic areas. These areas should not be such that analyses
during the investigation will disclose the identity of particular steel
plants unless the steel companies involved agree to such disclosure.
As an example, geographic definitions could coincide with AISI steel-
making districts, or with state boundaries.
(2 )
For each area, and for the United States as a whole, flow sheets
and/or material balances would be prepared so as to integrate all
steelworks in the area. Breakdowns would be by major processes
(e. g., blast furnaces, sintering machines, coke plants, BOF
furnaces, etc.) in each area, not by company or plant within the
area. It also would be possible not only to derive by direct statis ':"
tical analysis the average condition within each area, but also to
derive the frequency distribution of conditions deviating from the
average.
(3)
In the statistical analysis, emphasis could be focused on pollutants
of special interest, e. g., sulfur in gases.
Project 5 would be essentially an exercise in statistical analysis of technological
operating data. The AISI quarterly reports do not include information on capital or
operating costs. The time required for analysis and reporting would be expected to be
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-21
about 3 to 6 months, depending on the degree of detail desired, and the study is expected
to require about 5 to 10 man-months of professional staff time (including supervision),
plus supporting services (including the use of a digital computer).
Project 6: Control of Wide-Area
Emissions from Storage Piles
Many steelworks raw materials are stored outside in bulk and in large quantities
under conditions that allow fines to be airborne. The problem of air pollution is par-
ticularly severe when the material is being placed into or removed from storage. Coal
is thought to be the worst offender, but fine iron ore also can be seen by simple obser-
vation to be a substantial contributor to air pollution.
It is recommended that new research be conducted to prepare a definitive and
anal ytical report on the state of the art. In addition to the open literature (which con-
tains little information on this subject), visits should be made and conferences held
with steel companies, coal companies, ore companies, and other organizations that
are potential sources of information. The report should define the nature of the prob-
lem and the state of the art in sufficient detail to support recommendations for experi-
mental work and/or demonstration units. Among future experimental activities that
should be evaluated in the report are the following:
( 1 )
In-plant experiments on actual storage piles, including dust-fall

measurements before and after changes are made.
(2)
Model studies on a reduced-size scale, and including factors of
pile arrangement, wind breaks, and other factors that might
minimize air pollution by particulates.
Project 6 (the preparation of a definitive report to serve as the basis for decisions
on experimental work) is expected to require about 5 calendar months and to involve
about 10 man-months of professional time (including supervision), plus supporting ser-
vices and travel costs.
Project 7: Design and Costing of Systems
for Collection and Control of Wide-Area
T"-nissions Inside a Steelworks
This final technological report on Contract PH 22-68-65 identifies wide-area
emissions as those that occur over an expanse or space such that concerted attention
has not yet been devoted to their collection and control. Usually this condition exists
when the process segment is such that contained streams of air or other gases are not
involved as part of the process. Examples (given a ranking of first-order priority in
Section VII of the report) include the casting of blast-furnace hot metal, the flushing of
blast-furnace slag, the disposal of blast-furnace slag, and the pigging of molten iron.
Other examples include the preparation of ingot molds and the teeming of ingots
(especially when exothermic hot tops are used).
In many existing steelworks, emissions of this general type are largely uncon-
trolled for two reasons:
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
II-22
( 1)
Collection hoods or other devices would be expected to interfere
with other structures, especially with the operation of overhead
cranes that service the areas.
(2 )
Amounts of air that might have to be handled to collect the
emis sions can be expected to be large, and from this capital
and operating costs can be expected to be substantial.
It is recommended that activity on Project 7 involve the design and preparation of
engineering drawings for suitable representative hooding systems, and the estimation of
costs for the installation and operation of the systems. The report to accompany such
designs and drawings should be suitable for serving as the basis for decisions regarding
the construction of in-plant demonstration units. It is recommended that two types of
designs be included in the project:
(1)
Equipment for a typical difficult situation in an existing steelworks
(2)
Equipment suitable for use in a new steelworks where the equip-
ment can be installed as part of new construction.
Project 7 is estimated to require about 6 months of elapsed time at a cost equiva-
lent to that of about 10 man-months of professional staff.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III-l
SECTION III
THE INTEGRATED IRON AND STEEL INDUSTRY
This Section describes some of the characteristics of the integrated iron and
steel industry in the United States, and describes briefly some of the types of process-
ing that are involved. An integrated steel plant is one that performs all of the functions
necessary to convert iron ore to finished steel. Thus, an integrated steel plant usually
will be large and include coke ovens, at least one blast furnace, one or more steel-
making furnaces, casting facilities, and rolling operations that yield finished steel
characteristically in the form of plates, shapes, sheets, and/or strip. If a steel plant
does not start with the reduction of iron ore, but for example purchases iron or steel
scrap that is then remelted for processing into steel, then such a plant is not considered
to be "integrated". Likewise, a steel plant that does no rolling or .other processing be-
yond semifinished products such as billets or slabs is usually considered "nonintegrated".
In the broadest sense, the integrated iron and steel industry, therefore, includes the
operations of raw-material receipt and stocking, coking, reduction of iron ore, manu-
facture of molten steel, casting of the steel, and rolling of the steel into useful shapes.
In the context in which the term is usually used, the integrated steel industry does not
include small-scale operations (such as mini steel plants) that perform only a limited
number of the full range of functions.
Most companies in the integrated steel industry are large, and most of the largest
have several plants. About 25 steel companies in the United States have annual sales in
excess of $100 million, and of these about 6 have annual sales in excess of $1, 000
million.
Several distinctions should be kept in mind when considering the information and
interpretations in this Section. For example, it sometimes is easy to overlook the
distinction between capacity to make steel and production that deals with the amount
actually made. Ratings of capacity are so difficult to make thtit the ste.el industry no
longer publishes official capacities. Elimination of official but meaningless ratings for
capacitycame about a few years ago when production figures routinely ran over 100 per-
cent of "capacity".
Statistics on steelmaking quantities are frequently given in terms of two main
bases, and the distinction between them must be recognized. One basis is raw steel,
which is steel in the form of ingots or continuously cast billets, blooms, or slabs.
Until the advent of continuous casting, this basis was called ingot tonnage. The other
basis is the tonnage of finished steel, which characteristically amounts on a national
annual gros s scale to about 70 pe rcent of raw steel.
CHARACTERISTICS ON A NATIONAL SCALE
A statistical picture of the American integrated iron and steel industry is given
in Tables III-l through III-42. The tables are self explanatory, and cover aspects such
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
-I
III - 2
as production, financial status, number of persons employed, employment and other
costs, processes used, products manufactured, and raw materials consumed. Some of
the data are given by states where such information is available.
In this group of tables, the industry is pictured at a point in time - the year 1967*.
In addition to being the latest year on which statistics are available, the production of
127 million net tons of raw steel in 1967 was reasonably close to the long-term trend line
that would indicate a "normal" production of about 118 million tons in 1967. Thus, 1967
was neither exceptionally "good" nor exceptionally "bad". Whereas this Section deals
with the state of the industry in a single year, Section IV includes considerations of his-
torical changes with time, and statistical projection of performance into the future.
Except where otherwis e noted, information for Tables Ill-l through 1II-42 was
or adapted from the 1967 Annual Statistical Report of the American Iron and Steel
Institute. ':0:<
taken
In addition to some statistical breakdowns by states in various tables from
Table 1II-8 through III-40, Figure 1II-1 maps the location of the American iron and steel
industry on the basis of value added by manufacture by states. (1 )':<':<* The areas of the
circles centered over each steel-producing state are proportional to that state's contri-
bution to value added. The concentration of the United States steel industry in the region
of the Great Lakes is strikingly apparent. The three states Pennsylvania, Ohio, and
Indiana account for well over half of the steel industry's value added.
In Million. of Dollon
500. 999--
200.499-,
100.199-.......
50. 99-.........
20. 49--
10. 19--

}. 9--.
-
FIGURE III-I.
VALUE ADDED BY MANUFACTURE BY STATE - 1963
(Blast Furnaces, Steel Works, and Rolling and Finishing
Mills: SIC 331).
"Table III-37 is an exception because shipments of iron ore are given for 1966, the latest year reported.
""The AISI Annual Statistical Report covering 1968 is expected to be released in June, 1969.
''''References for Section III are given at the end of the Section.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 3
TABLE III-I.
PRODUCTION, SHIPMENTS, AND FOREIGN
TRADE OF THE UNITED STATES IRON AND
STEEL INDUSTRY - 1967
Thousands of Net Tons(a)
Raw Steel Production by Type of Furnace
Open Hearth
Basic Oxygen Process
Electric
Total
70,690
41,434
15,089
127,213
Net Shipments of Steel Products by Type of Steel
Carbon Steel
Alloy Steel
Stainle s s Ste e 1
Total
76,042
7,018
837
83,897
Foreign Trade in Mill Products
Exports
Imports
1,685
11,455
(a) All tons are net tons of 2000 pounds.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III-4
TABLE III- 2.
FINANCIAL BALANCE SHEET: U. S. STEEL INDUSTRY - 1967
Assets
Millions of Dollars
Current As sets
Cash
United States
Receivables,
Inventories
Government securities
less allowances for doubtful accounts
694.6
1,119.5
1,742.1
3,835.9
Total current assets
7,392.1
Miscellaneous investments and other assets
Fixed Assets
Property, plant and equipment (les s depletion)
Less - Depreciation and amortization
1,792.3
26,484.4
14,951. 7
Net fixed assets
11,532.7
Intangibles
Deferred charges
2.9
173.4
Total assets
20,893.4
Liabilities and Stockholders' Equity
Current Liabilities
Accrued taxes, including Federal income taxes
Long-term debt maturing within one year
All other current liabilitie s
855.7
96. 0
2,344.5
Total current liabilities
3,296.2
Long-term debt less amount maturing within one year
Reserves for future Federal income taxes, insurance,
contingencies, deferred credits, etc.
Minority interest in companies not wholly owned
Stockholders' Equity
Preferred stock
Common stock
Capital in excess of par or stated value of capital stock
Income reinvested in business
4,154.8
1,013.0
9.4
128.7
3,243.7
1,149.3
7,898.3
Total stockholders I equity
12,420.0
Total liabilities and stockholders' equity
20,893.4
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
:-
III - 5
TABLE lII-3. INCOME, DIVIDENDS, EMPLOYMENT, AND EMPLOYMENT COSTS:
UNITED STATES STEEL INDUSTRY - 1967
Reporting Companies Represent:
Blast furnace production, million net tons
Percent of industry
Raw steel production, million net tons
Percent of industry
Shipments of finished steel products, million net tons

Statement of Income, millions of dollars
Revenues:
Net billing value of products shipped and other services
Interest, dividends and other income
84.0
95.8
119.3
93.8
80.3
Total revenue
$17,045.7
195.6

17,241. 3
Costs Applied to Billings and Other Services millions of dollars
Employment costs:
Wages and salaries
-------
III - 6
TABLE 1II-4.
EMPLOYMENT, STOCKHOLDERS, AND CAPITAL
EXPENDITURES: UNITED STATES STEEL
INDUSTRY - 1967
Labor: (All employees)
Number of Employees:
Average monthly number of employees
wages and salaries during the year:
receiving
Wage employees
Salaried employees

Total Employees
538,522
171,483

710,005
Man Hours Actually Worked During the
in millions
Wage employees
Salaried employees

Total Hours
Year,
1, 0 09 . 3
349.5

1,358.8
Total Payroll, millions
Wage employees
Salaried employees

Total Payroll
of dollars(a)
$ 4,028.6
1,738.4

5,767.0
Less portion of payroll included above but charged

to own construction or other nonoperating
accounts
Balance
104.4
$ 5,662.6
Stockholders:
Number at end of year:
Preferred
Common
Total(b)
22,972
1,149,123
1,166,777
Expenditures for additions, improvements and acquisitions,
in millions of dollars
$ 2,172.8
Estimated cost to complete construction in progress,
in millions of dollar s
$ 3, 125. 9
(a) Wage and salaried payroll includes payments applicable to the prior year.
(b) Excludes duplication of holders of more than one class of stock of a single company but does not
exclude a holder of stock of more than one company.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
1II-7
TABLE III-5.
TOTAL NET SHIPMENTS OF STEEL PRODUCTS BY
UNITED STATES ST'EEL INDUSTRY - 1967 .
All Grades (Including Carbon, Alloy, and Stainless Steels)
Product
Net Tons
Ingots and steel castings
Blooms, slabs, billets, sheet bars
Tube rounds
Skelp
Wire rods
Total semi-finished
289,261
2,478,621
25,524
2,008
1,266,516
4, 061 , 93 0

5,591,735
541,186
7,948,349
14,081,270
Structural shapes (heavy)
Steel piling
Plates
Total shapes and plates

Rail s
Standard (over 60 pounds)
All other
Joint bars
Tie plates
T rack spikes
Wheels (rolled and forged)
Axle s
Total rails and accessories
684,523
45,955
30, 934
144,791
61,998
285,502
180,669
1,434,372
Bars
Hot rolled (incl. light shapes)
Reinforcing
Cold finished
Tool steel
Total bars and tool steel
7,961,028
3,249,479
1,732,575
109,929
13,053,011
Pipe and tubing
Standard
Oil country goods
Line
Mechanical
Pressure
Structural
Stainless
Total pipe and tubing
2,712,499
1,339,772
3,094,920
1, 051, 711
260,291
473,360
36,481
8,969,034
Wire
Drawn
Nails and staples
Barbed and twisted
Woven wire fence
Bale ties and baling wire
Total wire and wire products
2, 5 07 , 25 1
334,106
102, 175
116,050
73,757
3,133,339
BATTELLE MEMORIAL INSTITUTE - COLUMBUS' LABORATORIES

-------
III - 8
TABLE 1II-5.
(Continued)
Product
Net Tons
Black plate
Tin and terne'plate - hot dipped
Tin plate - electrolytic
Total tin mill products

Sheets
Hot rolled
Cold rolled
Sheets and strip
Galvanized - hot dipped
Electrolytic
All other metallic coated
Electrical
Strip
Hot rolled
Cold rolled
Total sheets and strip
6 04, 487
38, 734
5,947,638
6,590,859
9,311,592
14,709,006
4,218,700
. 326, 726
503,334
721,031
Total shipments
1,409,339
1,373,797
32,573,525

83, 897,340
'. BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 9
TABLE III- 6.
TOTAL PRODUCTION OF HOT-ROLLED PRODUCTS (BY
TYPE OF PRODUCT) BY UNITED STATES STEEL
INDUSTRY - 1967
Product
Tho.usands of Net Tons
Plates
Sheets
Strip(a)
Coils for cold reduced black plate and tin plate
Bars - Merchant
- Concrete reinforcing
- Light shape s
Structural shapes - heavy
Steel piling(b)
Rails
Joint bars and tie plate bars
Ske lp
Blanks or pierced billets for seamless tubing
Wire rods
Rolled forging billets
Blooms, billets, etc., for export
Car wheels (rolled steel)
All other
10,120
37,007
1,417
8,548
8,921
4,089
1,305
5,681
519
763
201
3,484
3,949
4,787
1,695
120
289
189
Total
93,084
(a) Includes cotton ties, balling bands and hoops.
(b) Includes H -bearing piles.
BATTELLE MEMORIAL INSTITUTE .:.. COLUMBUS. LABORATORIES

-------
III - 10
TABLE III-7.
TOTAL PRODUCTION OF HOT-
ROLLED STEEL PRODUCTS (BY
STATE) BY UNITED STATES
STEEL INDUSTRY - 1967
State  Thousands of Net Tons
Massachusetts } 
Rhode Is land 187
Connecticut 
. New YOrk } 4,049
New Jersey
Pennsylvania  21,060
Ma ry land } 4,786
Delaware
Virginia  
West Virginia  3,559
North Carolina 
Georgia  
Flo rida  
Ohio  15,789
Illinois  7,942
Indiana  14,665
Michigan } 
Minnesota 8,068
Missouri 
Alabama  3,181
Kentucky  1, 285
Mis sis sippi } 
Oklahoma 2,362(a)
Texas 
Colorado  
Arizona  
Utah  2,819
Washington  
Oregon  
CalifOrnia } 3,332
Hawaii
Total  93,084
(a) Includes Tennessee and Arkansas. 
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
TABLE III-8.
III - 11
PRODUCTION OF PLATES AND
SKELP (BY STATE) BY UNITED
STATES STEEL INDUSTRY - 1967
State
Thousands of Net Tons
New York
Pennsylvania
Delaware
Maryland
West Virginia
Ohio
Indiana
Illinois
Michigan
Kentuc ky
Alabama
Texas
Colorado
Utah
Washington
California
Total
}
}
3,647
1,215
}
}
}
1,330
2,294

1,138
2,504
1,476
13,604
TABLE III- 9.
PRODUCTION OF FLAT HOT-ROLLED
PRODUCTS (EXCEPT PLATES)(a) BY
UNITED STATES STEEL INDUSTRY -
1967 (BY STATE)
State

Rhode Island
New York
New Jersey
Pennsylvania
Maryland
West Virginia
Georgia
Kentucky
Alabama
Texas
Ohio
Indiana
Illinois
Michigan
Minnesota
Colorado
Utah
Washington
California
Total
Thousands of Net Tons
}
9,653
8,295
 9,535
 9,254
 1,425
} 6,858
} 1,952
 46, 972
(a) Includes sheets, strip, coils for cold -reduced black plate and tin plate,

hoops and cotton ties.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
[ll-12
TABLE III-10. PRODUCTION OF MERCHANT BARS AND LIGHT
SHAPES (BY STATE) BY UNITED STATES STEEL
. INDUSTRY - 1967 .
State
Massachusetts
Connecticut
New York
Pennsylvania
Maryland
Virginia
West Virginia
Norrh Carolina
Georgia
Florida
Ohio
!llinois
Ind iana
Michigan
Minnesota
Missouri
Alabama
Kentucky
Oklahoma
Texas
Colorado
Utah
Washingron
Oregon
California
Total
Thousands of Net Tons
}
989
2,437
}
205
1,982
1,617
1,595
}
}
}
542
161
205( a)
159
(a) Includes Tennessee,
334
10, 226
TABLE III-ll. PRODUCTION OF CONCRETE REINFORCING BARS
(BY STATE) BY UNITED STATES STEEL INDUSTRY
- 1967 .
Srate 
Massachusetts }
New York
pennsyl vania 
Maryland }
North Carolina
Georgia
Florida
Ohio 
!llinois 
Indiana }
Minnesota
Michigan
Missouri
Alabama }
Mississippi
Oklahoma }
Texas
Colorado }
Arizona
Wash ingron
Oregon
Hawaii
California 
Total 
Thousands of Net Tons
72
872
444(a)
192
506
392
302
464(b)
293
552
4.089
(a) Includes West Virginia and Virginia.
(b) Includes Tennessee and Arkansas.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 13
TABLE III-12. PRODUCTION OF HEAVY STRUC-
  TURAL SHAPES AND STEEL
  PILING (BY STATE) BY UNITED
  STATES STEEL INDUSTRY - 1967
 State  Thousands of Net Tons
New YOrk } 3,402
Pennsylvania
West Virginia } 
Georgia 351(a)
Ala bama
Texas 
Ohio  1 
Indiana 2,023
Illinois
Mis souri J 
Colorado } 
Utah  26,2
Washington
Oregon 
California  162
 Total  6,200
(a) Includes Kentucky.  
T'ABLE III- 13. PRODUCTION OF WIRE RODS (BY
  STATE) BY UNITED STATES
  STEEL INDUSTRY - 1967
 State  Thousands of Net Tons
Massachusetts } 
Rhode Island 273
New Jersey 
New York  46
Pennsylvania  732
Maryland } 355
Georgia
Ohio   699
Illinois  1,392
Ind iana  271
Michigan } 
Minnesota 370
Mis souri 
Alabama } 177
Texas 
ColOradO } 472
California
 Total  4, 787
BATTELLE MEMORIAL INSTITUTE';" COLUMBUS LABORATORIES

-------
1--
TABLE 1II-14.
III- 14
PRODUCTION OF BLANKS, TUBE
ROUNDS OR PIERCED BILLETS
FOR SEAMLESS TUBING (BY
STATE) BY U. S. STEEL INDUS-
TRY - 1967
State
Thousands of Net Tons
New York
Pennsylvania
Ohio
Indiana
Illinois
Colorado
Total
23
1,671
1,824
}
431
3,949
TABLE III-IS.
PRODUCTION OF SEMIFINISHED
STEEL FOR FORGINGS AND EX-
PORT (BY STATE) BY U. S. STEEL
INDUSTRY - 1967
State
Thousands of Net Tons
Connecticut
New York
Pennsylvania
Maryland
Kentuc ky
Alabama
Texas
Ohio
Ind iana
Illinois
Michigan
Minnesota
Colorado
Washington
California
Total
}
}
}
}
910
266(a)
. 622
17
1,815
(a) Includes Georgia and West Virginia.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 1 5
TABLE III-16.
PRODUCTION OF COLD-ROLLED SHEETS AND
STRIP. BY UNITED STATES STEEL INDUSTRY -
1967
Product
Thousands of Ne.t Tons
Finished on continuous or tandem mills
Finished on single- stand mills
19,351
2,494
Total
21,845
TABLE III-17.
PRODUCTION OF COATED SHEETS AND
"STRIP'BY UNITED STATES STEEL INDUS-
TRY - 1967
Product
Thousands of Net Tons
Galvanized sheets and strip
Long terne sheets
Electrolytic tin plate
Hot-dipped tin and terne plate
4,582
258
5,776
31
Total
10,647
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - J 6
TABLE III-18.
PRODUCTION OF PIPE AND TUBING (BY
TYPE) BY UNITED STATES STEEL
INDUSTRY :- 1967
Product
Thousands of Net Tons
Seamless
Buttweld
Electricweld
Gasweld and spiral weld
3,445
1,886
4,146
13
Total
9,609
TABLE III-19.
STEEL PRODUCTION (BY GRADES) BY
UNITED STATES STEEL INDUSTRY -
1967
Thousands of Net Tons
Raw Steel plus Castings
Carbon steel
Alloy steel
Stainles s steel
113,190
12,572
1, 451
Total production
127,213
Steel for Castings (included in total production)
Carbon steel
Alloy steel
Stainless steel
134
59
1
Total for castings
194
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 1 7
TABLE 1II-20.
RAW STEEL PRODUCTION (BY STATE AND BY TYPE OF FURNACE)
BY UNITED STATES STEEL INDUSTRY - 1967
State s
New York
Pennsylvania
Rhode Island, Connecticut, New Jersey,
Delaware, Maryland
Virginia, West Virginia, Georgia, Florida,
North Carolina, South Carolina
Kentuc ky
Alabama, Tennessee, Mississippi
Ohio
Indiana
Illinois
Michigan
Minnesota, Missouri, Oklahoma, Texas
Arizona, Colorado, Utah, Washington,
Oregon, Hawaii
California
Total
(a) Deleted because disclosure would reveal proprietary information.
(b) Includes Bessemer.
Total
7,298
29,881
8, 132
4,268
2,410
4,444
20,378
17,610
10,649
9,245
4,644
4,154
4,100
127,213
Thousands
Open
Hearth
(a)
20, 854(b)
5,225
(a)
(a)
(a)
10,763
11,157
5,387
(a)
2,310
( a)
1, 984
70,690
of Net Tons
Basic
Oxygen
Process
(a)
5,854
(a)
( a)
(a)
(a)
7,452
(a)
2,316
7,250
(a)
(a)
41,434
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
Electric
199
3, 173
(a)
958
575
344
2, 163
(a)
2,946
( a)
2,334
777
(a)
15,089

-------
III - 1 8
TABLE III-21.
PRODUCTION OF PIG IRON AND FERROALLOYS BY
UNITED STATES STEEL INDUSTRY - 1967
, Product
Thousands of Net Tons
Blast furnace production
Pig iron (including silvery pig iron)
Ferromanganese and spiegel
Total blast furnace production
86,984
663
87,647
Ferroalloys made in electric furnaces
Ferromanganese
Ferrosilicon
Ferrochrome
Ferrochrome silicon
Ferrophosphorus
Silico- manganese
All other
Total fer roalloys made in electric furnace s
280
528
263
154
III
230
259
1,825
Total pig iron and ferroalloys
89,472
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 1 9
TABLE III- 22.
PRODUCTION OF PIG IRON AND FERRO ALLOYS (BY GRADES AND
KINDS) BY UNITED STATES STEEL INDUSTRY - 1967
Product
Thousands of Net Tons
Pig iron
Basic
Bessemer
Low phosphorus
Foundry
Malleable
All othe r
81,344
1, 722
166
1,550
1,83 °
372
Total pig iron
86,984
l' erroalloys (made in blast furnaces
Ferromanganese and spiegeleisen
Ferrosillcon
All other ferroalloys
Total ferroalloys
and electric furnaces)
943
528
1,017
2,488
Total pig iron and ferroalloys
89,472
BATTELLE MEMORIAL INSTiTUTE"-- COLUMBUS LABORATORIES

-------
III - 20
TABLE III- 23.
NUMBER OF BLAST FURNACES
PRODUCING PIG IRON AND
FERROALLOYS ON JANUARY 1,
1968 (BY STATE)
State In Blast Total
 Pig Iron 
Alabama 9 17
California 4 4
Colorado 4 4
Illinois 14 18
Ind iana 22 24
Kentucky 2 3
Maryland 10 10
Michigan 9 9
Minnesota 1 2
New York 12 15
Ohio 33 47
Pennsylvania 39 58
Tennessee 0 3
Texas 2 2
Utah 3 3
West Virginia 4 4
Total' 168 223
 Ferroalloys 
Blast furnaces 5 7
GRAND TOTAL 173 230(a)
(a) Of the 57 furnaces which were idle on January I, 1968, 5 were relining.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 2 1
/
TABLE III- 24.
PRODUCTION OF PIG IRON AND FERROALLOYS (BY
STATE) BY UNITED STATES STEEL INDUSTRY - 1967
State
Thousands of Net Tons
Pig Iron(a)
To ta 1
6,172
20,542
10,824
4,290
14,485
12,167
.' 6,309
7,439
4, 756

86,984
New York
Pennsylvania
Maryland, West Virginia, Kentucky, Tennessee, Texas
Alabama
Ohio
Indiana
Illinois
Michigan, Minnesota
Colorado', Utah, California.
Ferroalloys
New York
Pennsylvania
Virginia, West Virginia,
Ohio
Othe r State s
South Carolina, Tennessee
109
468
559
734
618
Total
2,488
GRAND TOTAL
89,472
(a) Includes silvery pig iron.
BATTELLE MEMORIAL. INSTITUTE.- COLUMBUS LABORATORIES

-------
III - 22
TABLE III- 25.
MATERIALS USED BY UNITED STATES BLAST FURNACES IN
MANUF ACTURE OF IRON AND FERROALLOYS - 1967
Thousands of Net Tons of
Material Used
Products
Iron 
(Pig and 
Molten) Ferroalloys
42,390 12
133 1,257
96,875 11
139,398 1,280
3,749 58
135,649 1,222
4,218 9
809 7
3,409 2
5,476 2
14,422 341
56, 197 1,088
1,280 21
54,917 1,067
are
Iron ore (including manganiferous and block)
Manganese ore (including ferruginous manganese)
Agglomerated products (sinter, pellets, etc.)

Total ores and agglomerated products
consumed
Less flue dust and sludge produced

Net ores and agglomerated products
consumed
Scrap
Total scrap consumed
Less produced at blast furnaces and
auxiliary units
Net Scrap consumed
Mill cinder, roll scale, etc.
Limestone, dolomite, other flux materials
Coke
Total coke consumed in blast furnaces
Less coke breeze (dust) recovered

Net Coke consumed
Total
42,402
1,390
96,886
140,678
3,807
136,871
4,227
816
3,411
5,478
14, 763
57,285
1,301

55,984
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 23
TABLE III- 26.
MATERIALS USED BY UNITED STATES BLAST FURNACES IN THE
MANUFACTURE OF PIG IRON (BY STATES) - 1967
. . .
. .
State
Thousands of Net Tons 
   Mill 
Iron Ore and   Cinder, 
Agglomerate s( a) Scrap(a) Scale, etc. Total
9,392 218 487 10,097
33,155 820 1,574 35,549
16,612 252 875 17,739
7,215 121 19 7,355
20,453 1, 114 1,453 23,020
10,108 367 297 10,772
19,299 189 460 19,948
11, 545 205 126 11,876
7,870 123 185 8, 178
135,649 3,409 5,476 144,534
New York
Penns y1 vania
Maryland, West Virginia, Kentucky,
Texas
Alabama
Ohio
Illinois
Indiana
Michigan, Minnesota
Colorado, Utah, California
Total
(a) Total material charged less products recovered.
TABLE 1II-27.
CONSUMPTION OF MATERIALS PER
NET TON OF PIG IRON PRODUCED
IN THE UNITED STATES - 1967
Material
Net Tons
Iron ore and agg1omerates(a)
Scrap(a) . .
Mill cinder, scale, etc.
Total
1.560
0.039
0.063

1.662
Limestone and dolomite(b)
Coke(a)
0.257
0.631
(a) Based on total material charged less products recovered.
(b) Based on total limestone and dolomite charged directly into blast furnaces,
plus tonnage consumed in the production of agglomerates.
BATTELLE MEMORJ'AL rNSTITUTE - COLUMBUS LABORATORIES

-------
III - 24
TABLE III-28. SCRAP - STOCKS, PRODUCTION, RECEIPTS, AND CONSUMPTION BY GRADE - 1967
    Thousands of Net Tons  
  Stocks    Shipped, 
  On Hand    Transferred Stocks
  Beginning    or Otherwise On Hand
Grade  of Year Produced Received Consumed Disposed End of Year
Steel scrap       
Carbon  5,221 36,752 23, 158 56,533 3,499 5,099
Alloy (excl. stainless) 226 2,294 403 2,598 90 235
Stainless  90 477 297 752 26 86
Cast iron, incl. borings 1,022 4,331 2,041 5,539 984 871
Total  6,559 43,854(a) 25,899 65,422 4,599 6,291
(a) Production by source:
Source
Recirculating (including home, plant or
recycled iron and steel scrap) .
Obsolete (including molds, stools,
machinery, buildings, except rerolling rails)
Other (including slag,) etc.
Total
Produced
38, 118
3,389
2.347
43,854
TABLE III-29. CONSUMPTION OF SCRAP AND PIG IRON, AND PRODUCTION OF STEEL BY
TYPES OF FURNACES - 1967
  Thousands of Net Tons
  Consumed 
Type of Furnace Scrap Pig Iron(a) Total
Open hearth and Bessemer 32,298 42,266 78,564
Basic oxygen process 13,955 33,607 47,562
Electric 13,351 279 13,630
Cupola 1,048 216 1,264
Air 46 15 61
Blast 4,280  4,280
Other 444 2,467 2,911
Total 65.422 82,850 148.272
(a) Including molten metal.   
(b) Total raw steel production: 127,213.  
Raw Steel production(b)
70,560
41,434
13,489
125, 483
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III -2 5
TABLE III-30.
CONSUMPTION OF FUE'LS BY UNITED STATES STEEL
INDUSTRY - 1967
Blast Furnace Area Ste e 1 Heating and Heating  
Blast  Me lting Annealing Ovens for  
.t' urnaces Other Uses Furnace s Furnaces. . Wire Rods Othe r Total
  Fuel Oil~ thousands of gallons  
67,225 14,011 567,663 403,426 (a) 201,663 1,253,988
  Tar and Pitch,. thousands of gal ions   
20,936 (a) 248,222 (a) (a) 33,741 302,899
Liquid Petroleum Gas, thousands of gallons
(a)
( a)
11,112
3,853
13, 965
Natural Gas, millions of cubic feet - based on
1,000 Btuper cubic foot
44,255
5,766(b)
85,932
277,322
5,682
115,504
534,461
Coke Oven Gas, millions of cubic feet -
based on 500 Btu per cubic foot
13,433
276 923( c) .
,
32,762
352,236
9,681
201,193
886,228
Blast Furnace Gas, millions of cubic feet -
based on 95 Btu per cubic foot
1,400,147
1,291,217(d)
(a)
162,898
1,276,741
4,131,003
. (a) Included with "Other".
(b) Includes coke-oven underfiring.
(c) Includes 237,541 coke-oven underfiring.
(d) Includes 334,068 coke-oven underfiring.
TABLE III-31.
CONSUMPTION OF ELECTRIC
POWER BY UNITED STATES
STEEL INDUSTRY - 1967
Power Source
Millions of kwhr
Generated
Purchased,
Total
11.,954
30,557
42, 511
. " .
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 26
TABLE 1II-32.
CONSUMPTION OF FLUXES BY UNITED STATES STEEL
INDUSTRY - 1967
   Net Tons  
    Other 
Use' Fluorspar Limestone Lime Fluxes Total
In agglomerated products  7,929, 735   7,929,735
In blast furnaces  14,763,656   14,763,656
In steelmaking furnaces     
Open hearth 124,689 2,921,853 652,316 458,837 4,157,695
Basic oxygen process 254,630 363,211 2,997,223 238,689 3,853,753
Electric 108,843 118,880 405,336 70,738 703,797
Total 488, 162 26,097,335 4,054,875 768,264 31,408,636
TABLE 1II-33.
CONSUMPTION OF OXYGEN BY UNITED STATES STEEL
INDUSTRY - 1967
Millions of cubic feet in gaseous form(d)
High Purity( c) Low Purity Total
Purchased(a)
Produced(b)
Total
155,120
15,456
170,576
5,415
(e)
5,415
160,535
15,456
175,991
Consumption by l!s.es:
Conditioning
Scrap preparation
Other burning and welding
Blast furnaces
Ste e Ima king:
Open hearth
Basic oxygen process
Electric and Ees semer
18,355
1,998
2,617
8,662
Total Steelmaking
58,407
78, 707
3,352

140,466
Maintenance and construction
All othe r
779
3,114
To ta 1
175,991
(a) Purchased from vendors with facilities in or adjacent to plant or facilities located away from plant.
(b) Produced on companies own facilities or leased facilities. .
(c) The term high purity oxygen means a gaseous mixture containing 99. fP/o by volume, plus or minus 0.10/0 of pure oxygen.
(d) Oxygen consumed in liquid form was reduced to its gaseous equivalent in these statistics.
(e) Included in high purity.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 27
TABLE III- 34.
PRODUCTION AND USE OF AGGLOMERATED PRODUCTS BY
UNITED STATES STEEL INDUSTRy(a) - 1967
   Net Tons 
   Briquettes, 
Agglomerates Sinter Pellets nodules and other Total
Produced(b) 51,068, 111 38,050,646 1,695,004 90,813,761
Imported (c) 9,293,103 (c) 9,516,559
Cons umed in    
blast furnaces 51,638,941 43,453,448 1,793,657 96,886,046
Consumed in    
steelmaking furnaces (c) 603,974 (c) 988,474
(a) Includes ore agglomerates whether produced at consuming plants, at or near the mine site or several nonconsuming plants in
the United States. The latter two includes plants owned and operated by Bethlehem Steel Corp., Buffalo Sintering Corp. ,
Cleveland -Cliffs Iron Co., Erie Mining Co., Jones and Laughlin Steel Corp., Republic Steel Corp., Reserve Mining Co. ,
and United States Steel Corp. .
(b) Agglomerates produced from 96,124,000 Tlet tons of metallic-bearing materials, 3,228,000 net tons of coal and coke,
8,259,000 net tons of fluxing materials, and 1,976,000 net tons of miscellaneous materials.
(c) Deleted because disclosure would reveal proprietary information.
TABLE III-35.
CONSUMPTION OF NONFERROUS
METALS FOR COATING PURPOSES
BY UNITED STATES STEEL INDUS-
TRY-1967
Metal
Gras s Weight, net tons
Aluminum
Copper
Lead
Nickel
Tin
Zinc
7,186
8,751
4,506
454
34,250
334, 776
BATTEL.L.E MEMORIAL. INSTITUTE - COL.UMBUSL.ABORATORIES

-------
III - 28
TABLE III-36. UNITED STATES STEEL INDUSTRY IRON ORE INVENTORIES, RECEIPTS, AND CONSUMPTION - 1967
Thousands of Net Tons
    Original Source of Iron Ore   
   United States   Canada All Other 
  Great     Great  Foreign Total
  L
-------
III - 29
TABLE 1II-38.
PRODUCTION, RECEIPTS, CONSUMPTION
AND SHIPMENTS OF COKE BY UNITED
STATES IRON AND STEEL INDUSTRY -
1967
Coke Movement
Net Tons
Production
Metallurgical
Other coke
57,064,984
400,113
Total Production
57,465,097
Receipts from
Other steel companies
Merchant producers
Other plants of parent company
604,354
3,376,116
11,238,015
Total Receipts
Total Production and Receipts
15,218,485
72,683,582
Consumption
Blast furnaces
Foundries
Other uses
Total Consumption
56,204,659
14,345
216,408
56,435,412
Shipments to other
Steel companies
Outside companies
Plants of parent company
694,077
1, 134,800
11,864,037
Total Shipments
Total Consumption and Shipments
13,692,914
70,128,326
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 3 °
TABLE III-39.
COKE PRODUCTION BY UNITED STATES
IRON AND STEEL INDUSTRY (BY STATE)
- 1967
State
Net Tons
Massachusetts, New York }
Pennsylvania
Maryland, West Virginia
Tennessee, Alabama, Texas,
Ohio
Ind iana
Illinois
Michigan, Minnesota
Colorado, Utah, California
Total
Oklahoma
Total (Bureau of Mines)(a)
21,603,852

6,898,895
5,425,566
7,180,977
7,730,385
2,202,976
3,318,378
3,104,068
59,465,097
64,570,000
(a) Includes the merchant producers in addition to coke plants associated with iron and
steel plants.
TABLE III-40.
COKE CONSUMPTION BY UNITED STATES
IRON AND STEEL INDUSTRY (BY STATE) -
1967
State
Net Tons
Rhode Island, Massachusetts
New York
Pennsylvania
Ohio
Ind iana
Illinois
Michigan
Minnesota, Missouri
Delaware, Maryland, Virginia, West Virginia,
Georgia, Florida, North Carolina
Alabama
Tennessee, Kentucky
Oklahoma, Texas, Colorado, Utah
Washington, California
Total
n.a.
3,765,987
13,730,906
9,651,766
7,295,580
4,283,999
4,122,323
n.a.
5,111,664
3,556,575
975, 720
2,446,281
1,152,375
56,435,412
na - not available.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 31
TABLE III;'41.
COkE CONSUMPTION BY USES
IN THE UNITED STATES IRON
AND STEEL INDUSTRY - 1967
Use
Net Tons
Blast furnaces
Foundries
Other Uses
Total
56,204,659
14,345
216,408
56,435,412
TABLE 1II-42.
COAL CONSUMPTION BY USES
IN THE UNITED STATES IRON
AND STEEL INDUSTRY - 1967
Use
Net Tons
Production of coke
Production of steam( a)
Othe r purpo s e s
Total
82,698,190
6,097,949
780,505
89,576,644
(a) Includes coal consumed in generating electric power.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 32
GEOGRAPHIC DISTRIBUTION OF THE INDUSTRY
The integrated iron and steel industry in the United States is often divided into
areas or districts to facilitate statistical comparison and grouping of industrial areas.
The division into eleven districts, as used by the American Iron and Steel Institute, is
illustrated geographically in Figure III-2, and the tonnage of production of raw steel in
each of these districts annually since 1958 is given in Figure llI-3. The Chicago and
Pittsburgh districts produce the most steel - each about 24 to 27 million net tons per
year. These rivals are followed in descending order by the Northeast Coast District,
the Youngstown District, and the Detroit District. The district with the lowest produc-
tion rate (St. Louis) produces about 3 million tons of raw steel per year. The fastest
growing district in terms of percentage growth is Detroit, where production has about
doubled in the last decade.
,"
!
".
)
.....~......
""'"''''
'"\
I
<,
'.i'-
\'" W"t'r~
\
--~~
;"'-"",
'''''-'''''M'"
;-J
~
, . :\
.----...
,"
'"
"
n'
"
If
FIGURE 1II-2.
STEELMAKING DISTRICTS IN THE UNITED STATES
Present Facilities
Maps of each district, with the names of major steelmaking facilities in each
district, are given in Figures III-4 through llI-12.
Current Modernization and Expansion
Estimates of the total steelmaking capacity (as differentiated from production) of
the U. S. integrated iron and steel industry for 1968 varies from 165 million to as high
as 192 million net tons of raw steel per year. (2) These estimates undoubtedly include
many of the idle facilities, particularly open-hearth furnaces that have been on a standby
basis. A more practical estimate places the current steelmaking capacity at 155 million
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 3 3
 30
 28
 26
 24
 22
 20
 18
(/) 
c: 16
o
+- 
+- 
Q) 
c: 14
'0
(/) 
c: 
0 12
"E 
c: II
o 
+= 
0 
:J 10
"0
o 
~ 
Cl.. 
Q) 9
Q) 
+- 
(j) 
~ 8
o
a:: I
 7
 6
 5
/,
// Pittsburgh
_/
---
--
~.---.~. .
/.... "-.--

/ Northeast
/ Coast
/'-'-'-'/'
./'
(Change in scale)
St. LDuis
3
FIGURE III-3. RAW-STEEL PRODUCTION IN THE VARIOUS GEOGRAPHICAL
STEELMAKING DISTRICTS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 3 4
WATERVlIET
ALLEGHENY
LUDLUM
\ --
,-
Alban1

I
L--
\
New York
------"""


\
/?-'--....:.

) New Jersey
Pennsylvonla
MILTON
- CECO CORP.
BURNHAM - BETHLEHEM
- BALDWIN-LlMA- BET,HLEHEM STEE\
HAMILTON READING
- CARPENTER )
STEEL TON - STEEL FAIRLESS HILL
BETHLEHEM CONSHOHOOKEN U.S. STEEL.,'
ALAN WOOD- .PHILADELPHIA ~
COATSVILL~ ,/ MIDVALE -HEPPENSTALL
- -- - LUKEN.:..{) ~~6E~,~T

,./)\ ~"
SPARROWS POINT o~
BETHLEHEM %
\ '\
\
ARMCO
BAL TIMORE
EASTERN
STAINLESS
Atlantic
Ocean
Maryland
FIGURE III-4.
NOR THEAST COAST DISTRICT
I
I
I

OhiO Ri el
MIDLAND
I CRUCI BLE
STEEL
NATIONAL
STEEL
WEIRTON JONES a LAUGHLN
. I
STEUBENVILLE
WHEELlNG- PI~ HOUSTON
/ STEEL WASHINGTON
I STEEL
WASHINGTON.
JESSOP STEEL

I ~
West I i/
~~~~ 0
...
I ''t
"":::SYIVOnia - %.
Ohio
BEAVER FALLS
. BABCOCK a WILCOX
ittsburgh
MUNHALL
BRADDOCK
DUQUESNE
McKEESPORT
CLAIRTON
} U. S. STEEL
JOHNSTOWN.
BETHLEHEM
U.S. STEEL
------
FIGURE III-5.
PITTSBURGH DISTRICT
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 3 5
~
Canada
Buffalo
JONES a LAUGHLIN
Michigan . WARREN
_ALLEGHENY LUDLUM
FERNOAlE
Windsor
Lake Erie
'REPUBLIC STEEL



7 "'''U,," "'''



DUNKIRK. ROBLIN STEEL
ALLEGHENY LUOLUM
Canada
I New York
---------
McLOUTH
TRENTON.
Pennsylvania
IRVINE
-NATIONAL FORGE
FIGURE 1II-6.
DETROIT DISTRICT
FIGURE 1II-7.
BUFFALO DISTRICT
Lake Erie
~ MANSFIELD
~MPIRE-REEVES
REPUBLIC
STEEL~
MASS I LON
~ANTON
TIMKEN ROLLER
BEARING CO
REPUBLIC STEEL
I
I
(
\

Ohio I Penn



COPPERWELD STEEL I
, /177}.. ,,,,,,FARREL
REPUBLIC STEE~WARREN ~
SHARON STEEL

YOUNGSTOWN YOUNGSTOWN
~I
SHEET 8 TUBE CAMPBELL
REPUBLIC STEEL -.yOUNGSTOWN
U.S. STEEL SHEET. 8 TUB~

I
I
I
~. STEEL
~ \,. LORAIN

',,- ,

".............. Cleveland District ,,/
------'....-'
I
CLEVELAN,~ PU LlC STEEL
JidN{'S 8 LAUGHLIN
/
Youngstown District
FIGURE III-S.
CLEVELAND AND YOUNGSTOWN DISTRICTS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 36
.MIDDLETOWN
ARMCO STEEL
Ohio
0\1.0
",E.~se S"tE.E.1..
~O~\I.\'IE.\I.
GflE.E.
Kentucky
FIGURE III-9.
CINCINNATI DISTRICT
I\..
r/ "'-t

( Vlrginlo

/",1" f~(ffA"Wcf~ E
~ . ELECTRIC-
___-C--r-STEEL
- -, ~ARRIMAN KNOXVILLE N.Carahna
SANO SPRiNGS---z. r TENNESSEE ..~ti8~VILLE CROFT
. ARMCO STEEL f FORGING.r ../"-~FLORIDA
I .> Tennessee L-- . STEEL
Okl ho I J__GADSDEN--," S.Carahna "-
a ma A k I REPUBLIC ATLANTA. f
r ansas BIRMINGHAM. \ . ATLANTIC CAYCE
~ (CONNORS. STEEL \ OWEN J
L,,~ I ~ ANNISTON \. ELECTRIC
. . . FAIRFIELD KILBY STEEL STEEL
- -- US. STEEL \ \r
LONE STAR. I (MIssIssippi , ~ f
LONE STAR . ). JACKSON Alabama ( Georgia
STEEL l loUISiana ~ MISSISSIPPI I
\. L( STEEL, L--\---~\
Texas ----"'\
? I,.,.J-!\... ~ Flanda
ARMCO STEEL I ~ ~"" \
~~U~TPE~~)1P-.J-II~~VS,- '\ \.
f V I TAMPA'\
\-? ~~ ~W~~A\
\If. )
\.."&1
,.,..~
FIGURE III-l O.
SOUTHERN DISTRICT
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
.

-------
III -37
St. Louis District
( Includes A,mco 51001 . Konsas City. Mo. i


@ Springfield ;'
~ake Michigan r"
STERLI NG
~ORTHWESTERN Chicago J
STEEL a WIRE LEMONT. ~ - - -
LEMONT MFG.. .../"""
CHICAGO HTS. GARY
"" BORG-WARNER
" . JONES a LAUGHLIN I
Illinois -....... HENNEPIN KANKAKEE
-......., .KANKAKEE
PEORIA'" ELECTRIC
.KEYSTONE STEEL
STEEL a WIRE I
\ Chicago District
') (Includes U.S. Stool,Duluth,Minn.
and North Slor St.ol,St. Paul,
Minn.) I

I
I
I
I
Indiana
FIGURE III-II. CHICAGO AND
ST. LOUIS DISTRICTS
KOKOMO
.
CONTINENTAL
STEEL
~."'''



LAItEDE STEEL
GRANITE CITY
SI. Louis GRANITE'
CITY STEE~ /

/'
---
Indianapolis
@
Pacific Ocean
Missouri
SOU H CHICAGO



Q}S' STEEL
~ Lake MlchlQCIn

W~Ne .INTERLAKE STEEL ( Iron Plont)
. REPU8UC S1£EL
INTERLAKE STEEL (Coke Plant)
"" Indiana Harbor
~EElIfrT~'~ ~l~~D
.lNlERLAKE STm~ \
- ( Steel Plant)
/ /
/ \
RTLAND---I '7
:.::~' I ~


-- \....,,""\r--_-
-- Idaho I

California ---/t-----l--,... I -
/ ---I Wyoming

/ Nevada I L
JUOSON STEEL ~?t~e. ---I -
EfinERYV1LLE /. GENEVA --
~~;nclsca ~~~~,g'~~~1'ES U.S. STE~L
T I
Utah I Colorado

/ PUEBLO.
---- C.F. 81.
- STEEL
I --

I
I
I
Metropolitan ChicoQo Area
Montano
FIGURE III-I2.
WESTERN
DISTRICTS
Denver.
Arizona
New Mexico
. Albuquerque
Metropolitan Los Angeles Area
,Phoenix
TEMPE
ALLISON
STEEL
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III-38
tons per year(2), taking into consideration only those facilities that can be placed into
operation on short notice. The changes in production capacity are illustrated in
Figures III-l3, III-l4 and III-15.
Figure IIl-13 shows the size distribution of open-hearth furnaces in 1968, and the
furnaces of the respective sizes that are expected at this time to be removed from
operation by 1973. The open hearths removed from operation are primarily those re-
moved to make room for additional BOF capacity.
100
90
80
70
60
....
~ 50
E
:J
Z
40
30
20
10
FIGURE III-13.
o
Total of 467 open hearth furnaces in 1968.
1973 = 257 furnaces (forecast)
Shaded portion represents furnaces forecast to
be removed from operation by 1973.
125 175 225 275 325 375 425 475 525 575
100 150 200 250 300 350 400 450 500 550 600

Nominal Open Hearth Furnace Capacity, net tons
SIZE DISTRIBUTION OF OPEN HEAR TH FURNACES
IN THE UNITED STATES
The trend in installation of BO F capacity is illustrated in Figure III-14. Installa-
tions planned for operational status in 1969 include BOF's by Armco Steel at Middletown,
Ohio; Bethlehem Steel at Bethlehem, Pennsylvania; U.S. Steel at South Chicago, Illinois;
and Youngstown Sheet and Tube at Indiana Harbor, Indiana; for a total capacity increase
of 9. 9 million net tons in 1969. During 1970, an additional 7 million nettons of BOF
capacity is expected to be 'placed into operation by Bethlehem at Burns Harbor, Indiana;
National Steel-Great Lakes at Ecourse, Michigan; Republic Steel at Buffalo, New York;
and U.S. Steel at Lorain, Ohio. During 1972, U.S. Steel is expected to place an addi-
tional 2.25 million capacity plant in operation at the Edgar Thomson Works, followed in
1973 by Inland Steel's 2-million-net-ton plant in East Chicago, Indiana. On the basis of
these expectations, by 1973 there will be a total BO F steelmaking capacity of about
76 million net tons pe r year.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IIl- 3 9
 70   - -
    /-
  Future /
 60 '
II) planned I
c  ,
o  i nsta Ilations---l
+- 
+- 
Q) 50  '
c  /
'+-   /
o  
II) 40  /
c Rated
o  /
  installed ./
E 30 capacity .
/ 
Q)   
Q)  / 
+- 20 
(f) / 
~  ,/'  
0   
0:. 10 /'  
  /'  
 """'-  
o
1957 1959 1961 1963 1965 1967 1969 1971 1973
1958 1960 1962 1964 1966 1968 1970 1972

Year
FIGURE IlI-14.
TRENDS IN INSTALLED CAPACITY AND RAW
STEEL PRODUCTION IN BOF FURNACES
Electric-furnace steelmaking capacity is also on the increase. The size distribu-
tion of electric-arc furnaces is shown in Figure III-IS. The trend toward the installa-
tion of the larger furnaces is evident by the furnaces planned for installation by 1971.
The most recent announcement involves Republic Steel's plans for 250-ton furnaces at
their Chicago plant. In 1967, electric steelmaking furnaces accounted for 15 million
tons of steel (11.9 percent of the total). This was increased in 1968 to 16 million net
tons (12.3 percent of the total). (3)
28
26
24
1/1 22
~ 20
g 18
:; 16
It.. 14
~ 12
~ 10
.0 8
~ 6
Z 4
2
o
10 20 30 40 50 60 70 80 90 Iro 200 250
15 25 35 45 55 65 75 85 100 175 225

Furnace Capaci ty, net tons
FIGURE III-IS. SIZE DISTRIBUTION OF ELECTRIC
STEELMAKING FURNACES
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III-40
The shift in the market for steel products into the Midwest has been the motivation
in heavy expansion of steelmaking facilities in the Chicago area, in addition to com-
pletely new plants at Burns Harbor, Indiana, and at Hennepin, Illinois. Considerable
expansion in facilities has taken place in the Houston, Texas, area at Armco Steel's
existing electric furnace shop and at U. S. Steel's electric furnace plant which is ex-
pected to be completed late in 1969. (4)
Future Plans
Plans for expansion of steel companies are usually preceded by purchases of real
estate for the plants. For example, National Steel recently purchased a large tract of
property at Corpus Christi, Texas. (5,6) The site has deep-water channel access to the
Gulf of Mexico, with resulting ocean transport possible for high-quality ores from
South America. National Steel Corporation also appears to have long-range plans for
the possible construction of steelmaking facilities in California, as indicated by its
contractin& to purchase approximately 3300 acres on the north shore of the Sacramento
River. (7, ) This site also has deep-water acces s to San Francisco Bay and the Pacific
Ocean, so as to improve the feasibility of using high-grade foreign ores in the produc-
tion of iron and steel at the new site.
Another future development (this time in technology) is heralded by the construction
of a plant in Venezuela by the U.S. Steel Corporation to make one million net tons per
year of pre reduced agglomerate in the form of "high iron-ore briquettes ". (9) The oper-
ation will make use of natural gas to perform the reduction operation, after which the
briquettes will have been upgraded to an iron content of 86 percent. (l0) The "high iron-
ore briquettes II are expected to lower transportation cost, reduce coke consumption in
the blast furnace, and result in increased production of hot metal from blast furnaces.
This technology has not yet gained general acceptance in the United States. Its advent
in Venezula is prompted by proximity to high-grade ore, and availability of natural gas
at a price about one-quarter of that in the United States.
MAKING OF IRON AND STEEL
The major process segments concerned with the making of iron and steel are
described briefly in this Section. More -detailed descriptions of the various steps in
the making of iron and steel will be found in Appendix A, IIProcesses in the Integrated
Iron and Steel Industry".
In 1967 the integrated iron and steel industry in the United States consumed
135,649,000 net tons of iron ore, 8,885,000 net tons of scrap and mill scale,
14,422,000 net tons of flux, and 54,917,000 net tons of coke to make 82, 850, 000 net tons
of pig iron. The pig iron combined with 65,213,000 net tons of scrap was used in mak-
ing 127,213,000 net tons of raw steel which in turn was converted into 83,897,000 net
tons of finished steel products. The magnitude of these quantities shows that the
integrated iron and steel industry is involved in a massive, continuous, material-
handling problem. Much of this handling which is of an intermittent nature, is a major
contributor to the problem of air pollution by the industry.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 4 1
Raw Materials
The major raw materials required to make pig iron (or "hot metal" as it is called
in its molten state) are iron ore, coke, limestone, and air. In recent years, auxiliary
fuels (such as oil, natural gas, coal, and tar) have been used to lower the coke require-
ment for making iron. Raw materials are received at a steel plant by boat, barge, or
railroad. Some lump iron ore is used for making hot metal, but about 70 percent of the
metallics charged to American blast furnaces are agglomerated products such as pellets
and sinter.
Pellets are made at the mine site by grinding the ore toa very small size (usually
less than 325 mesh), adding a binder, and then forming the ore into small balls or pel-
lets, which are subsequently hardened. Prior to pelletizing, if the ore is low-grade
(i. e., low in iron content), it can be beneficiated or up-graded by various methods that
remove unwanted gangue materials and result in a higher iron content in the feed to the
pellet plant. After hardening, the pellets are shipped to the steel plants for use directly
in the blast furnace.
Sinter plants, on the other hand, are located at the steel plant for two reasons:
(1) sinter is friable and does not withstand shipping without degradation, and (2) sintering
was initially developed and installed to recover and convert to a useful form the ore
fines, blast-furnace flue dust, mill scale, and other iron-bearing materials that could
not be us'ed directly as charge to a blast furnace. Advancing technology showed that a
properly prepared sinter would result in still further improvements in productivity if
even good ore was crushed and made into sinter. This was the motivation for the con-
struction of many sinter plants.
Although coke is the major fuel and reducing agent used to make hot metal in blast
furnaces, coke is made from still another raw material (coal). Coal received from the
mines (again by boat, barge, or railroad) is usually less than 1/2-inch in maximum
size. It is crushed and ground further, blended with other coals, and then charged into
coke ovens for conversion to coke. Conversion takes place by subjecting the coal to in-
direct heating in long, thin ovens, for periods of 16 to 20 hours. During the coking
operation, coke-oven gas, tar, and other by-products are collected. The coke-oven
gas is used for heating the coke ovens, and for other in-plant use. The tar and other
materials are recovered and processed in the by-product plant. On completion of the
coking cycle, the coke is pushed from the ovens into a special car, which carries it to
a quenching tower where the incadescent coke is quenched by a deluge of water. The
coke is then crushed and screened prior to its use in the blast furnace.
Limestone is crushed and screened to size at the quarry site, and is received at
the steel plant in the proper size for use in the blast furnace. Similarly, lime required
in steelmaking is prepared in lime kilns at the quarry site and shipped to the steel
plant ready for use. However, there is a growing tendency to calcine limestone to lime
at the steel plant.
The large amounts of raw materials needed to make hot metal require extensive
storage facilities at the steel plant to ensure a continuity of operation.
BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

-------
III-42
Making Iron
The prepared materials (coke, iron ore, pellets, sinter, and flux) are placed in
a transient storage in the blast-furnace stockhouse, from which they are withdrawn in
weighed fractions and charged into the top o,f the blast furnace via a skip hoist. Col-
lectively these solids are called the "burden". The newest blast furnaces constructed
in the United States consume such a vast quanity of materials that continuous conveyor
belts are used for charging rather than the intermittent-operating skip hoist, but this is
not yet a common practice.
The heat energy necessary to affect the required chemical reactions between the
iron ore, pellets, sinter, and flux is supplied by blowing air (preheated to temperatures
of 1600 to 2000 F) into the bottom of the blast furnace via blowpipes known as "tuyeres".
The heat energy also serves to maintain the pig iron in a molten state and at a temper-
ature such that the iron can be removed from the furnace by "casting". The coke in the
combustion zone at the tuyeres does not burn to carbon dioxide. Because of the high
temperature (about 3500 F to 3600 F) and the presence of the large amount of carbon in
the form of coke, carbon monoxide is formed and passes on up through the solid burden
where it takes part in some of the chemical reactions necessary to produce metallic
iron. The excess of carbon monoxide (diluted with carbon dioxide, nitrogen, and mois-
ture) passes off the top of the blast furnace, and is collected for use as a fuel to heat the
air blown into the blast furnace and for other in-plant heating purposes.
The hot metal produced in the blast furnace and the liquid slag (which is a fused
mixture of the flux and impurities removed from the ore, pellets, sinter, and coke) are
removed periodically from the blast furnace. The hot metal is "cast", while the slag is
said to be "flushed". Hot metal is cast into special containers called submarine or
torpede ladles (because of their elongated shape) for transfer to the steelmaking plant.
Slag is disposed of at a dump, or it may be granulated to produce an aggregate that is
sold.
Although almost all pig iron in the United States is made in blast furnaces, another
type of processing of iron ore to metal is just starting to make its commercial appear-
ance. This is "direct reduction", a name applied loosely to any method that bypasses
the blast furnace. Direct-reduction processes characteristically are solid-state pro-
cesses. Reduction of the iron ore to metal is conducted entirely in the solid state,
although the metallic iron might subsequently be melted. Direct-reduction processes
appear in many variations, but all use some form of carbon and/or some form of hydro-
gen as fuel and reducing agent. Carbon can be supplied as coal, coke, or other hydro-
carbons. Carbon monoxide is one of the useful forms of carbon for this purpose.
Hydrogen is supplied sometimes from reformed natural gas (when it often appears in a
mixture with carbon monoxide) or from coke-oven ga~. Regardless of the process, the
product usually is a spongy or powdered form of metallic iron (still mixed with gangue
from the iron ore) which then is intended for charging directly into steelmaking furnaces,
much as steel scrap is used for steelmaking.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 4 3
Making Steel
Steelmaking in the United States is done in three types of furnaces: (1) the open-
hearth furnace, (2) the basic oxygen furnace, and .(3) the electric-arc furnace. The
open hearth is a rectangular furnace with a comparatively shallow hearth for containing
and processing the steel. Scrap, flux, and hot metal are charged into the furnace via
several doors located in the front of the furnace. Heating of the charge materials is
done with various fuels such as oil, tar, natural gas, and combinations of these. Steel
is produced by lowering the carbon, manganese, and silicon contents by oxidation to
levels desired for the type of steel required. Impurities such as sulfur and phosphorus
are also lowered by using slags of appropriate compositions. Lowering of the carbon,
silicon, and manganese contents is accomplished by the .addition of oxygen to the charge
to oxidize these elements so that the carbon passes off as carbon monoxide, and the
silica and manganese oxides dissolve in the slag. The older steelmaking technology
used iron ore and air to supply the necessary oxygen, while the newer technology makes
use of gaseous oxygen introduced into the molten bath through a water-cooled lance.
Open-hearth steelmaking requires comparatively lengthy time periods, varying from
8 to 10 hours for practices not using oxygen lancing to time periods of 4 to 5 hours for
practices using oxygen lancing.
The basic oxygen steelmaking (BOF) process makes use of a pear-shaped vessel
to contain the charge materials required to make steel. As in the open-hearth process,
the major materials are hot metal and scrap. In the basic oxygen process, no external
heat is supplied to achieve the required lowering of carbon, silicon, and manganese
contents. Gaseous oxygen is blown onto the surface of the molten charge at a very high
rate, and the heat generated by the oxidation of the elements is sufficient to carry the
process to completion and produce steel. Flux (in the form of lime) is charged into the
vessel to form a slag by combining with the oxides of iron, silicon, and manganese, and
to lower the contents of sulfur and phosphorus in the, steel to acceptable limits.
The heat necessary to produce steel in an electric-arc furnace is supplied as
electrical energy to the charge materials which in the majority of cases is solid steel
scrap. Only two steelmaking plants in the United States are known to use hot metal as
part of the charge in large electric furnaces. The scrap used in conventional practice
is of a high quality, with compositions approximating those required in the finished steel.
Oxygen (in the form of ore or gaseous oxygen) is used to achieve the desired chemical
reaction necessary to produce the steel. The use of gaseous oxygen is by far the more
prevalent method. While most of the stainless and alloy steels are made in electric
furnaces, in recent years the increase in size of the furnaces, combined with higher
powered transformers, and a greater availability of scrap, has made electric-furnace
steelmaking of the plain-carbon, high-tonnage steels competitive in cost with the open
hearth and with BOF steelmaking.
In all steelmaking practices, the refined steel is tapped from the furnace into
ladles, after which it is transported to an adjacent area of the steel plant where it is
cast into ingots, cast directly into continuous casting machines, or cast into pressure-
casting molds for conversion to semifinished products.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
III - 44
Manufacture of Semifinished Products
Molten steel must be converted to some type of semifinished product before it can
be processed to the finished products which are either sold as such or are sold to other
intermediate proces sors for manufacture into finished products. Three methods of
converting the molten steel into semifinished products are (1) casting of ingots and rol-
ling, (2) continuous casting, and (3) pressure casting.
The practice used to process most steel into semifinished products involves teem-
ing of the molten steel into cast iron ingot molds, permitting the steel to solidify, re-
moving the ingots from the mold, and transporting the ingots to soaking pits for heating
and temperature equalization before primary rolling. This technology is time consuming
and requires large amounts of plant space and numerous personnel. After the ingots
have been properly heated in the soaking pits, the hot ingots are transported to the
roughing mill where they are rolled into billets, blooms, or slabs for subsequent pro-
cessing into finished products.
Continuous casting of steel is a relatively new process that can eliminate much of
the processing required by ingot casting and rolling. The molten steel is transported
in a ladle to the continuous-casting machine, where it is poured into a water-cooled
mold in which solidification starts. Solidified steel (often with a molten core) is ex-
tracted from the bottom of the mold continuously in one or several strands. The solid-
ified steel is cut into desired lengths for further processing into finished products.
Continuous -casting machines are designed to produce billets, blooms, or slabs, depend-
ing on the products to be manufactured in later stages in the plant. The trend in the
United States has been toward combining electric steelmaking furnaces with billet and
bloom continuous -casting operations; and the basic oxygen furnace with large slab-
casting machines. One operation has succeeded in continuous-continuous casting of 1500
tons of steel in a slab-casting machine. (This was achieved by casting five, 300 ton
heats consecutively, without any stoppage of the continuous casting machine.) Such
technology is conducive to greater economics in the making of steel products.
Pressure casting is another new process that bypasses the ingot stage in the
manufacture of steel products. Molten steel is forced into a large slab-shaped cavity
via a ceramic tube inserted in the steel. The force is developed by exerting air pres-
. sure on the surface of the molten steel. The steel solidifies in the reusable graphite
mold, after which the mold is opened and the slab removed. The process has been
used to date mostly for casting stainless steels. However, a plant scheduled to come
on-stream in 1969 will produce plain-carbon-steel slabs by the pressure-casting
process.
Manufacture of Finished Products
Billets, blooms, and slabs (the semifinished products) are processed into rods,
bars, wire (wire in this sense may be as large as I-inch in diameter), angles, channels,
structurals, plate, sheet, and strip depending on the particular capabilities of a given
plant. Rolling is the method used to convert the semiiinished products into finished
products. Prior to the rolling operation, the billets, blooms, and slabs are inspected
so that defects potentially detrimental to the finished products I}lay be removed by chip-
ping, grinding, or scarfing. After this "conditioning" operation, the billets, blooms, or
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IIl-45 and III-46
slabs are heated in special reheating furnaces. On reaching the required temperature
and after an appropriate soaking time, the semi-finished products are removed from the
furnaces and rolled in highly specialized mills.
Further processing of the products may take place in the steel plant before ship-
ment is made. This may include annealing, cold-rolling, galvanizing, tin electroplating,
painting, or plastic coating. The coatings are applied primarily to the sheet and strip
products.
( 10)
REFERENCES FOR SECTION III
(1)
Census of Manufacturers, United States Department of Commerce, 1963.
(2)
165 Million Tons", American Metal Market, February 14, 1969,
"Steel Capacity:
p. 2.
(3)
Roche, J. P., "Year End Statement; Steel Market Registers Record Growth in
1968", Blast Furnace and Steel Plant, 57 (1), p. 51 (January, 1969).
(4)
"U.S. Steel's Works to Take Shape Soonl', Blast Furnace and Steel Plant, ~ (11),
p. 1049 (November, 1967).
(5)
"National Steel to Purchase Large Texas Tract", Blast Furnace and Steel Plant,
56 (12), p. 1063 (December, 1968).
(6 )
"Signs Contract to Buy Corpus Christi Site", Iron and Steel Engineer, 45 (12),
p. 159 (December, 1968).
(7)
"National Steel Plans West Coast Output", Metal Working News, November 13,
1 967, p. 12 .
(8)
"National Steel to Buy California Property'l, Blast Furnace and Steel Plant, ~ (12),
p. 1123 (December, 1967).
( 9)
"U.S. Steel to Make Iron Ore Briquettes in Venezuela", The Iron Age, 201 (21),
p. 26 ( May 23, 1968).
"Venezuelan Ore Beneficiation Plant Under Constructionl', Blast Furnace and
Steel Plant, ~ (12), p. 1108 (December, 1968).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV -1
SEC TION IV
PROJECTIONS REGARDING STEEL-INDUSTRY GROWTH
AND CHANGES IN PROCESS TECHNOLOG Y
The purpose of this section is to analyze and project the nature and extent of
changes expected to take place in the integrated iron and steel industry between now and
about the year 1980. Emphasis is focused on changes expected to be greatest in their
effects upon air-polluting emissions. The background for these projections includes a
wide variety of information sources, including the published literature, plans of various
industrial companies, statistical reports, and forecasts and analyses prepared by
Battelle during many other projects recently conducted or currently in process.
Although some of the projections rest upon a background of hard statistics and solid
technological information, the project~ons in most cases are set by judgmental evalu-
ations after thorough analysis of all the available and pertinent inputs.
Anticipated Growth in Production and
Consumption of Raw Steel
The steelmaking capacity of the United States is about 155 to 160 million net tons*
of raw steel per year. There is no "official" statement of capacity, because capacity
of any particular installation depends heavily on how it is operated. The capacity con-
stantly changes; new facilities are always under construction, obsolete facilities are
being dismantled, while existing equipment is being modified.
Steelmaking capacity is so hard to define that the AISI has abandoned its old
practice of annually issuing a new rating. Part of the difficulty is that the steelmaking
equipment can be operated for considerable periods at above its "nominal capacity".
A few months during the first half of 1968 exemplified such a situation ~ production
was high in anticipation of a possible strike. Many furnace installations operated con-
siderably above their nominal capacity. How long this condition could have continued
without serious equipment or labor troubles is speculative, but it is likely that the rate
that was probably in excess of "nominal capacity" could not have been continued
indefinitely. Thus, the "annual capacities" presented here are in a sense integrated
over a period of a year, using judgment and some speculation as to how long and how
great temporary displacements may be in future years.
In 1967, capital expenditures for expansion and modernization of the domestic
steel industry were about $2.3 billion. A large part of the outlay went for basic-
oxygen steelmaking furnaces, electric steelmaking furnaces, and continuous-casting
facilities. The new steelmaking furnaces will replace existing open-hearth furnaces
to a large degree.
"Throughout this section, "tons" are net tons of 2000 pounds.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV-2
In recent years, there have. been important shiits in the geographical concentra-
tion of steelmaking in the United States. Most significant is the buildup of new capacity
in the Midwest, especially near the Chicago market area.
The growing West Coast market has often been investigated by interested steel
companies, but competition from Japanese steel producers has been a big factor tending
to discourage U. S. companies from establishing additional West Coast steel-producing
facilities. Despite this competition from overseas producers, it is likely that new
integrated steelmaking installations will be built by U. S. companies on the West Coast
in the course of the next decade.
The annual production of raw steel in the United States since World War II has
fluctuated from a low of just over 66 million tons in 1946 to a high of 134 million tons
in 1966. In 1967, the production was 127 million tons. "Raw steel" is in the form of
ingots or continuously cast billets or slabs and is converted to a lower tonnage of
finished steel.
The production of raw steel for 1960 and 1967 and projections to 1980 are given
in Table IV-I. Annual production of raw steel in the United States is projected to reach
about 157 million tons in 1975 and about 180 million tons in 1980. These forecasts of
total production are based primarily on judgments concerning 1975 and 1980 technology
and on U. S. input-output tables prepared by Battelle's Socio-Economic Research Group
as part of its Aids :::. Corporate Thinking program.
 TABLE IV-I. PRODUCTION (1000 NET TONS) OF RAW STEEL 
  IN THE UNITED STATES, BY TYPE OF FURNACE, 
  AND PROJECTIONS OF PRODUCTION TO 1980 
  Source: Annual Statistical Report, American Iron 
   and Steel Institute, 1967. Forecasts by 
   Battelle.   
 Open  Basic-Oxygen Electric 
Year Hearth Bessemer Furnace  Furnace Total
1960 86,368 1189 3,346  8,379 99,282
1967 70,690 (a) 41,434  15,089 127,213
1975 44,000  80,000  33,000 157,000
1980 36,000  99,000  45,000 180,000
(a) Included with open-hearth production; 278,000 tons of Bessemer steel were reported in 1966. 
Production forecasts of raw steel in the United States by several other organiza-
tions or individuals are summarized in Table IV-2, which shows that the estimated
annual production of raw steel for 1975 varies from 145 to 180 million net tons. For
1980, the estimates vary from 176 to 184 million net tons per year.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
lV-3
TABLE IV-2.
U. S. PRODUCTION OF RAW ST~EL,
AS FORECAST BY VARIOUS SOURCES
Source
Million Net Tons
1975 1980
UN/ECE Steel Committee
(current study)(a)
163
184
H. S. Harrison, Cleveland-Cliffs
Iron Co. (October, 1965)
148
F. K. Maxcy, Jr., Pittsburgh
Steel Co. (October, 1966)
-145
F. Jaicks, Inland Steel Co.
(September, 1966)
165
Chemical Week (September 24, 1966)
(September 24, 1966)
171
Iron Age (January 6, 1966)
180
Dr. Pierre Rinfret, Lionel D. Edie &
Co. (September, 1962)
170
Resources for the Future (Medium
Forecast) (September, 1962)
176
Arthur D. Little, Inc. (December, 1962)
-160
-180
(a) Forecast in a draft report made public at a meeting in Geneva, October 16-20, 1967. Report
was to be published before June, 1968.
Production of steel in the United States by geographical area for 1960 and 1967 and
forecasts to 1980 are given in Table IV-3. The forecasts by area are based on the most
recent 5-year average production indexes published by the AISI, as adjusted to meet.
tonnage forecasts for 1975 and 1980. The Chicago area is expected to continue to be the
largest producer of steel in the United States, and Pittsburgh is expected to remain in
second place. On a percentage basis, the Detroit area is expected to show the fastest
future growth. Production of raw steel in the Pittsburgh, Youngstown, and St. Louis
areas is expected to grow at a slower rate than for other areas in the United States.
The apparent per-capita consumption of raw steel in 1960 and 1967 and forecasts
to 1980 are given in Table IV -4. Apparent per-capita consumption is calculated by adding
the raw-steel equivalent of imported steel to U. S. production and subtracting the raw-
steel equivalent of exports, then dividing by the population. The projected per-capita
consumption for 1975 and 1980 was based on the assumption that imports of steel would
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV-4
TABLE IV-3. PRODUCTION OF RAW STEEL, BY AREA, IN 1960 AND
 1967 AND FORECASTS TO 1980  
  Millions of Net Tons 
Area 1960(a) 1967(b) 1975 1980
Chicago 20.80 26.45 33. 10 37.80
Pittsburgh 20. 10 24.45 27.30 31.10
Northeast Coast 14.35 17.65 20.80 24.50
Youngstown 8.35 9.90 10.90 12.50
Detroit 6.50 9. 15 14.80 16.90
Western 6. 15 8.25 10.20 11. 70
Southern 5.65 8.25 10.30 11. 70
Cleveland 5.55 6.80 9.00 10.30
Buffalo 5.20 7.25 9.00 10.30
Cincinnati 4.00 5. 75 7.80 8.90
St. Louis 2.65 3.30 3.80 4.30
Total 99.30 127.20 157.00 180.00
(a) Estimated from steel-production indexes published by AISI.
(b) Preliminary breakdown by area gave production at 126.9 million net tons. This was adjusted up to 127.2 million net tons
to be consistent with AISI Statistical Yearbook.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV-5
amount to 10 percent of the U. S. production of raw steel and that exports would amount
to 2 percent. The apparent per-capita annual consumption of steel is expected to rise
from 1408 pounds in 1967 to 1598 pounds by 1980.
TABLE IV-4.
APPARENT PER-CAPITA CONSUMPTION OF RAW STEEL
IN THE UNITED STATES FOR 1960 AND 1967 AND
PROJECTIONS TO 1980
  Apparent Consumption of Per-Capita Raw
 Population (a), Equivalent Raw Steel, (b) Steel Consumption,
Year millions million net tons pounds
1960 180.7 99.5 1101
1967 199.2 140.2 1408
1975 223.8 169.6 1516
1980 243.3 194.4 1598
(a) 1980 population estimate from Series B projection by U. S. Bureau of Census.
(b) Raw-steel production. plus imports, minus exports of mill products; both of the latter converted to raw steel
equivalent based on 75 percent yield.
The foregoing estimates of consumption are based on a net importation of steel
products in an amount equivalent to 8 percent of the annual U. "'8:" production. The
experience of 1967 and 1968 suggests that this net difference could reach or even exceed
20 percent by 1975 or 1980. It is possible to speculate on the future net difference, but
it is impossible to prove any assumption that might be made. The net difference will be
affected not only by relative costs of producing steel in various countries but also by
Governmental policies and by voluntary quota arrangements. In the case of the United
States, the uncertainty about the net difference between imports and exports is caused
mainly by the uncertainty about the amount of imports; the amount of exports is expected
to fluctuate relatively little from year to year. The net difference of 8 percent used in
the present estimates is reasonably defensible on a judgmental basis. To the extent that
history proves it to be in error, an increase in net imports will detract from projected
U. S. production of steel rather than increase projected per-capita consumption.
Anticipated Growth in Production and
Consumption of Pig Iron
The production and apparent consumption of pig iron for 1960 and 1967 and fore-
casts to 1980 are given in Table IV-5. During the 18-year period from 1950 to 1967, the
ratio of apparent total pig-iron consumption to raw-steel production has ranged between
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV-6
a low of O. 651 and a high of O. 691. The trend up to 1966 was for a gradual increase in
the ratio of pig iron to raw steel. This was the result of an increase in tonnage of BOF
steelmaking at the expense of the open-hearth process. In 1967, the ratio dropped
slightly as the production of electric-furnace steel increased percentagewise. In the
future, the ratio of pig iron to raw steel is expected to remain at O. 68 as BOF and
electric-furnace steelmaking continue to increase at the expense of the open-hearth
process.
TABLE IV-5.
PIG-IRON REQUIREMENTS FOR THE UNITED STATES
IN 1960 AND 1967 AND PROJECTIONS TO 1980
  Apparent Consumption 
 Raw Steel Produced, of Pig Iron, Ratio of Pig Iron/
Year millions of net tons millions of net tons Raw Steel
1960 99.3 66.5 0.670
1967 127.2 87.0 0.684
1975 157.0 106.8 0.680
1980 180. 0 122.4 0.680
Anticipated Growth in Consumption of Iron Ore
The depletion of domestic direct-shipping iron-ore deposits during the past 20 years
has caused the United States to rely increasingly on outside sources. In an effort to
forestall continuation of the trend, domestic producers and consumers of iron ore turned
to the upgrading of relatively low-iron domestic ores, such as taconites. The success
of the taconite beneficiation program has in turn aifected the physical and chemical speci-
fications of imported ores.
The consumption of iron ore (and its derivatives) by the iron and steel industry in
the United States in 1960 and 1967 and projections to 1980 are given in Table N -6. The
consumption in 1967 was about 132 million net tons. Consumption in 1975 is projected to
be about 158 million net tons, with about 179 million net tons expected in 1980. The ratio
of the consumption of iron ore (and derivatives) to production of pig iron was 1. 47 in 1967
and is expected to decrease to 1. 44 by 1980, because the average iron content of the ore
has gone up with the increased use of pellets in the blast furnace. This trend is expected
to continue.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV-7
TABLE Iv-6.
CONSUMPTION OF IRON ORE AND DERIVATIVES IN
THE IRON AND STEEL INDUSTR Y IN THE UNITED
STATES AND PROJECTIONS TO 1980
 Iron Ore and Agglomerates Used, (a)  
 1000 net tons Total Consumed, Tons of Ore
Year In Blast Furnaces In Steel Furnaces 1000 net tons Per Ton .of Iron
1960 106,260 8,250 114,510 1. 60
1967 128,045 4, 266 132,311 1. 47
1975 155,000 3,500 158,500 1. 45
1980 176,200 2,800 179,000 1. 44
(a) Includes lump ore and ore fines used for sinter, pellets, briquettes, and nodules.
Changes in percentages of the different forms of ore consumed in the United States
and forecasts for future consumption are given in Table IV -7. Burden preparation and
upgrading has increased rapidly during the past 12 years, and the trend is expected to
continue. Sinter and pellets represented about 65 percent of the burden in 1967; by 1980,
sinter and pellets are expected to be over 80 percent of the blast-furnace burden. This
will tend to reduce the amount of dust produced per ton of pig iron in the blast furnace.
TABLE IV-7.
PHYSICAL FORM OF IRON ORE CONSUMED IN THE
UNITED STATES AND ESTIMATES TO 1980
 Lump Ore Sinter Fines (a) Pellets Total
     Millions 
 Millions  Millions  Millions  
Year of Tons Percent of Tons Per cent of Tons Percent of Tons Percent
1960 62.0 54. 1 41. 5 36.3 11. 0 9.6 114.5 100.0
1967 46.7 35.3 42.2 31. 9 43.4 32.8 132.3 100.0
1975 31. 7 20. 0 46.0 29.0 80.8 51. 0 158.5 100.0
1980 34.9 19.5 48.3 27.0 95.8 53.5 179.0 100.0
(a) Includes only iron-ore fines used for making sinter. For example, sinter production in 1967 was 51. 6 million net tons but
required on! y 42. 2 million tons of iron -ore fines. The remainder was supplied as mill scale, dust, fluxes, and the like.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
N-8
As steel companies pay increasing attention to air quality, they direct their atten-
tion in two directions - (1) production of lower amounts of pollutants by the processes
and (2) collection of pollutants to avoid their emission into the atmosphere. The second
of these courses results in the accumulation of large amounts of fine particles, many of
which are contaminated in some way that makes them unsuitable for recycling into the
processes as those processes are now being run. Many of these "dusts" introduce a
cost for disposal, and this cost is in addition to the cost of collection. Few of these
dusts have actual by-product value. Research is needed to develop better means for the
disposal of such materials. The most logical concept would involve recycling within the
steel plant, but attainment of this will require further development of recycling schemes
and modification of the steelmaking processes to accept such recycled materials without
intolerable technological or economic penalty.
Anticipated Changes in Iron and Steel Process Technology
The continued growth of the iron and steel industry in the United States has
resulted in gradual changes and improvements in processing technology. Most of these
have been evolutionary rather than drastic or rapid. Many changes have taken place
within the past 10 years, and gradual evolutional changes are expected to continue.
Modernization and expansion programs have been conducted by most of the steel com-
panies during the 1960s aimed at improving the competitive position of the American
producer. Marginal and poorly productive facilities are gradually being shut down or
phased out, but the industry probably will continue for decades to contain a substantial
amount of marginal equipment.
With respect to the manufacture of iron from ore, the blast furnace will remain'
by far the major production unit. More pig iron will be produced each year in fewer
blast furnaces that will, on the average, become larger and have higher productivity
per furnace. Electric smelting and direct reduction of iron ore are expected to be
relatively minor in overall importance as iromnaking processes in the United States at
least until 1975, but these practices are expected to find increasing application outside
the United States.
With respect to steelmaking in the United States, the BOF process and the electric-
furnace process will continue to grow at the expense of the open-hearth process. Vacuum
degas sing of steel and continuous casting will assume increasing importance over the
next two decades. Large steel plants will tend to become larger and the number of small
steel plants ("mini steel plants") will increase, but the latter will account for relatively
little tonnage of steel. From an overall air-quality viewpoint, the large integrated
plants will continue to pose the major problems.
Proces sing of Iron Ore at or
Near the Source of the Ore
Processing of iron ore includes all steps in beneficiation and agglomeration as well
as reduction (smelting) of the ore to metal:
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV-9
(1) There will be a continuing strong trend toward closer sizing of lump ore.
There is no uniformity from company to company as yet' on specifications
for size of lump ore, but the trend is for the upper dimensional limit to
become smaller. The trend in size is toward something like plus 1/4 inch
by minus 1 inch. The allowable percent of minus 1/4-inch material will
also trend downward. Fines generated from the crushing and grinding of
lump ore during sizing operations will be used as sinter fines or will be
ground further for pelletizing.
(2) There is expected to be a continuation of a present strong trend toward
further upgrading of the chemical quality of ore as mined (increased
iron content and decreased gangue content). This will be accomplished
through the following:
(a) Probable increase in the use of autogenous grinding
when the characteristics of the ore permi,t
(b) Use of grinding equipment of larger size
(c) Wider application of more sophisticated beneficiation
practices
(d) Possible application of reduction roasting, followed
by magnetic beneficiation of hematitic taconites and
other ores amenable to such a process
(e) Construction of additional new large oxide pellet plants
at or near iron-ore mines to permit broader use of fines
(f)
Probable establishment during the 1970s of direct-
reduction plants at or near ore mines to produce
prereduced materials. The use of this material will
be discussed later in this section.
Coke Ovens and Coke Production
Coke production reached its peak in 1953 with an annual output of 78 million tons
and bottomed out in 1962 at 51 million tons. In 1966 (a record-tonnage steel year), coke
production was 67 million tons.
Improvements in blast-furnace technology during the past 15 years have lowered
by 30 percent the amount of coke necessary to produce 1 ton of hot metal. Little capital
has been put into new coke ovens because of the great number of coke ovens installed
from 1950 to 1960 (5,617 ovens accounting for 52 percent of the total number of ovens
operating in 1969), and commitments to other major projects from 1960 to date. Of
the total number of coke ovens in operation in 1969 at iron and steel plants, 32 percent
are over 20 years old, 52 percent are between 11 and 20 years old, and 16 percent are
less than 10 years old. From this information, it is likely that within the next 10 years,
much of the 32 percent increment will be replaced with new ovens, and after that there
may be a heavy period of coke-oven construction to replace much of the 52 percent in-
crement. If research and development is to be done to replace the slot-oven method of
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV -10
making blast-furnace coke, the work should be completed by 1980 before the period for
replacement of the 52 percent inc rement of coke ovens arrives.
During the past 8 years, conventional by-product slot-type ovens have produced
98 percent of the coke consumed in North America. Beehive coke ovens produced the
remaining 2 percent.
Beehive ovens are generally located in remote areas, partially because of their
air-polluting characteristics, and partially because they tend to be located near the coal
mines. Beehive ovens produce a low-grade metallurgical coke that can be used only for
a limited number of applications in the steel plant. Although the cost of producing bee-
hive coke may be from $3 to $4 per ton cheaper than producing conventional coke, there
is little interest in beehive coke. The conventional beehive ovens now in existence are
mostly in western Pennsylvania. They may continue to remain in existence as a hedge
to provide reserve cokemaking capacity in case of an emergency, but, from technologi-
cal, economic, and air-quality aspects strong arguments could be mustered to encour-
age their abandonment. At some future date, nonrecovery type coke ovens may have
greater application due to the deteriorating coal chemical market.
With respect to conventional slot-type ovens, there has not been much activity or
success in the steel industry in improving coking practice, but people directly engaged
in coking and coke-oven research and development probably will deny this generality,
because there has been a gradual evolutionary change in the direction of better coke for
blast-furnace use. Nevertheless, present cokemaking equipment and practices persist
not because they are really good and appropriate but because there is no proven better
method at this time. Starting and stopping of a conventional by-product coke-oven
battery is one of the most complex procedures in the steel-plant operation. Replacement
of existing coking facilities is not expected to occur until the requirements of blast fur-
naces for coke exceed the capacities of existing ovens, which have high operating costs
because of their antiquity (but low allocable fixed charges for the same reason).
Of the problems remaining to be resolved in the control of air pollution in the iron
and steel industry, coke -plant operations are among the worst offenders as far as air
pollution is concerned. Coke-oven operations cause air pollution mainly through three
aspects of the process: (1) coal preparation, (2) coal charging, and (3) coke pushing and
quenching.
In spite of the low amount of new coking-facility construction in recent years and
the advanced age of existing coke plants, several developments in processing are taking
place which lead to the following expectations about the future:
(1) An improvement in coke-oven refractories by changes in shape (design),
composition, density, and the processes by which the refractories are
produced. "
(2) A shortening of coking time by using higher flue temperatures and by
developing methods for reversal of flue-heating gases. The improved
refractories are essential for this practice.
(3) More research in coal petrography to develop a wider range of blends
of coal for coking. The Bureau of Mines has determined the coking
properties of more than 1200 individual coals in the United States.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV-ll
(4) The use of larger coke ovens capable of producing 80 percent more
coke per oven than the largest ovens now in use in the United States.
New ovens are 20 feet high, 50 feet long, and 18 inches wide; older
ovens are 10 to 14 feet high, 30 to 45 feet long, and 17 to 18 inches
wide.
(5) The development of new coking processes. Several new methods for
carbonization of coal are under study and development. Some of these
are continuous. To date, no continuous coking process has been
judged suitable by the U. S. steel industry.
(6) Development of equipment to control, especially, particulates emitted
into the air. New method.s currently in use in Europe may fall short
of really effective control, if applied to operations on U. S. coke
ovens.
(7) Nominal continued improvement in the strength and composition of the
coke (but no drastic improvement is foreseen).
All of the foregoing relates to coke made from coal. Petroleum-based coke seems
to warrant more attention than it has received in the past. It has the advantage of higher
carbon content and less ash, but these advantages have been more than offset to date by
the higher sulfur content of petroleum coke. Currently known processes for producing
petroleum coke low in sulfur have been too expensive for high-tonnage use,. and no break-
through on this problem appears to be in sight.
The Ironmaking Blast Furnace
The blast furnace will continue to be the major producer of pig iron in the United
States, as it is throughout the world. Blast-furnace technology has improved a great
deal in the past few years and has led to increased productivity of existing and new
furnaces. New furnaces are larger than those they replace, thus causing the statistical
average size of furnaces to increase from year to year.
The most important single factor that will influence the future of the blast furnace
is the continuing growth in capacity of BOF steelmaking capacity in the United States.
The BOF process uses a high percentage of hot metal in its charge (say 70 percent), and
the blast furnace is the main source of this hot metal.
Size and Number of Blast Furnaces. A trend in the United States and elsewhere has
been to enlarge the inner volume of existing blast furnaces during relining and rebuilding.
Because of the large number of small blast furnaces that have been maintained in opera-
tion, the average size of the blast furnaces has not increased much during recent years.
The mean diameter has increased from about 23 feet in 1960 to about 24 feet in 1967. The
largest blast furnace in the United States has a hearth diameter of 32 feet.
The total number of blast furnaces in the United States has decreased from 250 in
1960 to 225 in 1967. The total number of furnaces on blast (excluding those on standby
and undergoing relining) in 1960 was 218 and on January 1, 1967, the number on blast
was 158. Production of pig iron in 1960 was 66.5 million tons and in 1967 the production
was 87 million tons.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV-12
The average production of pig iron per blast furnace per day in 1960 and 1967 and
projections to 1980 are given in Table IV -8. The average rose from about 1200 net tons
per furnace day in 1960 to almost 1600 in 1967. By 1980, average production of pig iron
per blast furnace per day is expected to be about 2200 net tons.
TABLE IV-8.
AVERAGE PRODUCTION OF PIG IRON PER U. S.
BLAST FURNACE PER DAY FOR 1960 AND 1967
AND PROJECTIONS TO 1980
Year
Total Days in Blast
Average Per Blast Furnace,
net tons/ day
1960
1967
1975 (a)
1980(a)
52,266
55,414
55,625
55,636
1181
1570
1920
2200
(a) Projections by Battelle.
The average production per furnace day will continue to increase because of the
retirement of some of the older and smaller furnaces, increases in the sizes of some
existing furnaces during relining, and the very large size of new furnaces that will be
built. Improvements in controls and in processing techniques will also augment the
productivity per furnace, especially for newer and larger furnaces. Probably little
money (on a relative basis) will be devoted to improvement of older and smaller furnaces.
Because of these factors, the total number of blast furnaces in operation in the United
States is expected to decrease, thus decreasing the number of potential sources of air
pollution.
Changes in Blast-Furnace Technology. The productivity of blast furnaces as a
class has increased because they have tended to become larger on the average, the raw
materials have been upgraded, and improved operating techniques have been developed.
The major factors that have affected the recent past and that are expected to affect the
future of bIas t fur nace s are as follows:
(1) The increased use of pellets and sinter in the burden, which provides
a richer ore and less gangue and permits a lower coke rate.
(2) Emphasis on burden preparation to upgrade the chemical composition
and the physical form of the burden. Sizing of the ore, sinter, coke,
and limestone provides a more uniform material in the stack of the
furnace.
(3) There has been some improvement in the quality of coke used in the
blast furnace. Coke has tended to become somewhat harder and denser
and to contain a lower ash content, so less coke is needed per ton of
iron produced.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV -13
(4) The use of higher wind rates to increase the driving rate of blast furnaces
generally and, in turn, to raise productivity.
(5) Increased top pressure to aid in controlling the furnace operation when
higher wind rates are used. However, progress along these lines has
been slow and is not expected to be really substantial in the next decade.
(6) The use of increased blast temperature to lower coke consumption
per ton of iron and thus to increase' productivity per furnace. The upper
blast temperature, however, is limited by problems with regard to
refractories, valving, and ducting.
(7) The injection of cheaper hydrocarbons to replace some of the coke and
permit use of higher blast temperatures.
(8) Increased use of instrumentation to aid in controlling the more modern
high-production units.
The productivity of U. S. blast furnaces is expected to continue to increase in the
future. The changes in technology just tabulated are expected to increase in degree;
for example, even higher wind rates and blast temperatures will be used in the future
as well as higher percentages of pellets and sinter in the burden.
The development of new equipment is expected to extend the practical limits of the
processing techniques now being employed.
In addition to the established methods of increasing the productivity of the blast
furnace, the use of metallized or pre reduced material may also become an accepted
practice in achieving additional improvements in blast-furnace productivity. Limited
investigations of use of such material indicate an increase in productivity of the furnace
and a decrease in coke consumption,
The improved prevailing technology and expected future changes in technology of
U. S. blast-furnace operations will decrease air-pollution problems for the following

reasons:
(1) The number of operating units is expected to decrease because of the
larger average size and increased productivity per unit.
(2) A decrease is expected in the amount of dust produced per ton of hot
metal because of constantly improving burdens and lower coke rates.
,
(3) Improved control, including improved collection and contr~l of
emissions and reduction in the number of "slips" that emit dust
,
into the atmosphere is expected.
Electric-Furnace Pig Iron
Electric smelting furnaces are used in several parts of the world to pro.duc~ p~g
iron. Their pig-iron capacity is about 4. 4,million tons per year. The electnc pig-uon
furnace is well suited to locations having ready access to cheap electric power and/ or

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV - 14
low-strength raw materials that are not suitable for conventional blast furnaces. A
clear distinction must be recognized by the reader between electric smelting furnaces
(now under discussion) and electric melting or steelmaking furnaces. At the present
time, electric smelting furnaces are not used for the production of pig iron in the United
States nor are they expected to be a major factor in the production of pig iron in the
United States in the future.
An electric smelting furnace is being installed near Mobile, Alabama, at the pres-
ent time as part of an operation known as the D-LM process. Pellets containing iron
ore and coal are partially reduced on a traveling grate and then hot charged into a
smelting furnace to complete the reduction of the ore to produce molten pig iron. The
initial production capacity is announced at 200,000 net tons per year. Some of the claims
of the D-LM process with respect to air quality and water quality control are as follows:
(1) Minimum combustion products are generated because coking, agglom-
erating, and pre reduction are all carried out coincidentially.
(2) Entrainment of solids during pelletizing is minimized because the
agglomerates are maintained in a quiescent bed (nonagitating) and
the draft is exhausted through a filter bed of wet green pellets.
(3) Direct charge of preheated burden avoids air contamination from water
and draft quenching of hot intermediate products.
(4) Electric smelting provides endothermal energy without direct combus-
tion products.
(5) Wet systems for grinding, gas cleaning, and returns transport avoid
secondary dust entrainment.
(6) Full water recycle for mill and cooling systems involves extensive
evaporation which does not require continual discharge of liquid effluent.
The next two decades may see several more small electric-furnace pig-iron plants
installed in the United States, but there are several reasons (in addition to rarity) to
expect that, in the aggregate, they will not be important contributors to air pollution.
The components of the facility can be compactly arranged, and the economy of the method
generally may be expected to depend upon full use of all by-products which must be col-
lected before utilization.
Direct-Reduction Processes
Several direct-reduction processes have been developed to produce sponge iron or
molten pig iron without the use of the blast furnace.
The D-LM installation mentioned in the discussion on electric smelting is the first
commercial venture of this nature in the United States. A direct-reduction plant for pro-
ducing sponge iron is under construction near Portland, Oregon. Its rated sponge-iron
capacity is 450,000 net tons a year. The sponge iron will be converted to steel in elec-
tric steelmaking furnaces nearby. Other small plants of this type are likely to be con-
structed in the United States in the future if the first two are successful, but proliferation
,
of such plants in the United States during the next 20 years is unlikely.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV - 1 5
During the past 10 years there has been considerable discussion about the eco-
nomic potential for large sponge-iron plants in the United States. "Large" in this sense
means an annual capacity of 1 million tons or more of metallized product. U. S. Steel
Corporation is building such a plant in Venezuela. Standard Oil of New Jersey has a
process of this nature for which the imminent construction of a plant in the United States
frequently has been rumored. For the most part, however, there are no confirmed
plans for such construction in the United States; and this means that it is extremely un-
likely that there will be any large U. S. plants of this type operational before 1975.
Conditions for economical operation of such plants currently favor locations outside the
United States because of proximity to suitable ore deposits, lower labor rates, lower
costs for natural gas and electricity, and other factors including political aspects. In
the event that such plants are built in the United States, they probably would be located
near the sources of ore, which means generally remote from present steelmaking
centers and urban areas. Furthermore, the sophisticated technology involved in such
plants, plus the fact that they involve new concepts, tends to maximize opportunities
to build in truly modern and effective air-quality control.
Basic-Oxygen-Furnace (BOF) Steelmaking
The annual production of steel in the United States by the BOF process has
increased from about 300,000 tons in 1955 to 41.4 million tons in 1967. It is estimated
that, by 1980, the BOF process will produce 99 million tons of steel annually in the
United States. Similar growth in this process also is taking place in most other countries
that are major producers of steel.
Of significance in the phenomenal growth of BOF steelmaking is that it is displacing
some relatively modern open-hearth shops. One example of the new trend is Granite
City Steel Company's new BOF shop that has replaced production from seven large open
hearths constructed during and after World War II. To date, about twelve open-hearth
shops in the United States have been placed on standby or have been demolished to make
way for new BOF capacity.
Some steel companies are now entering their second phase of BOF construction
and are replacing small BOF vessels with larger vessels. For example, Jones and
Laughlin Steel Corporation at Aliquippa, Pennsylvania, has constructed a new three-
vessel (160/ 190 tons each) shop and retired a two-vessel (80 tons each) plant built in
1957; and McLouth Steel Corporation is replacing three 60-ton vessels with two 110-ton
units, with corre sponding enlargement of the supporting meltshop facilitie s. Crucible
Steel Company at Midland, Pennsylvania, demolished three open-hearth furnaces to pro-
vide space for a new BOF shop. Youngstown Sheet and Tube Company at Indiana Harbor
has under construction two 265-ton BOF vessels in a shop scheduled for completion in
1970 to replace their No.1 open-hearth shop consisting of nine furnaces, which will be
dismantled. The No.2 open-hearth shop (already equipped with air-pollution control
equipment) will remain in use to make open-hearth grades of steel and to give the com-
pany continued flexibility to melt scrap in quantities larger than can be handled conven-
tionally in the BOF vessels.
The first installation in the United States of a new Japanese gas-collection process
(OG Process) will be located in Armco's new BOF plant at Middletown, Ohio. This
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV - 16
process, which incorporates a noncombustion technique for collecting CO gas and fumes
from BOF furnaces and cleans the gas by high-energy scrubbing, is now operating over-
seas on twelve BOF furnaces, with several more installations under construction. Beth-
lehem Steel is equipping two 250-ton BOF vessels currently under construction with
tubular-membrane hoods for steam production.
Future trends for BOF steelmaking in the United States are expected to include the
following:
(1) Increased total capacity, estimated at 99 million tons annually by 1980
(2) Replacement of small BOF vessels rated at 30 to 80 tons per heat by
larger vessels with capacities of 100 to 300 tons per heat
(3) Increased use of waste heat from fume-collection systems as a source
of in-plant energy.
These trends will reduce the number of open-hearth furnaces in the United States
by about 50 percent by 1980. This change-over will tend to alleviate the air-pollution
problem, because the open hearth has been a major source of air pollution in the past,
especially when oxygen lancing has been applied.
Electric-Furnace Steelmaking
The annual production of steel in the electric-arc furnace has increased from about
8 million tons in 1955 to over 15 million tons in 1967. The electric furnace's future is
so promising that one forecaster has stated that, within 25 years, U. S. steelmaking
capacity will be divided between electric furnaces (45 percent) and BOF furnaces
(55 percent) - no surviving open-hearth furnaces are foreseen by that forecaster.
There are presently about 200 electric-arc steelmaking furnaces (over 12-foot
shell diameter) operational in the United States, with about 14 in Canada and another
17 million tons of capacity under construction.
As examples of the trend, Armco Steel Corporation is installing three 22-foot fur-
naces rated at about 150 tons per heat at its Butler Works. Bethlehem Steel Corporation
has announced the construction of an electric-furnace meltshop to replace ten open
hearths at Steelton, Pennsylvania. This Bethlehem complex of three 150-ton furnaces
will be equipped with "the most modern facilities for dust collection and fume abate-
ment". Completion of the Steelton shop is scheduled for late 1969. Republic Steel Cor-
poration's plant at Canton, Ohio, has constructed and placed into operation a new melt-
shop housing four ZOO-ton electric steelmaking furnaces, each about 26 feet in diameter.
Several other steel companies have announced plans to install electric-arc furnaces.
Each of these undertakings is by a major steel company that is adding new capacity to
its plants or replacing existing open-hearth capacity by electric -arc capacity.
From the standpoint of the large integrated steel companies, BOF furnaces and
electric steelmaking furnaces are natural complements to each other. The fading open- '
hearth process had considerable flexibility to accept various ratios of hot metal and
solid-steel scrap. Considered independently of each other, the BOF and the electric
furnace each have less flexibility along these lines than the open hearth. But the BOF
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV - 1 7
process is essentially a hot-metal process, and the electric furnace is essentially a
scrap-based process. Therefore, combinations of BOF furnaces and electric furnaces
cannot only restore the flexibility of the former open-hearth situation but even can extend
that flexibility. This is a strong reason for the general expectation that the continued
growth of the BOF process will be a strong factor encouraging parallel growth of
electric -furnace steelmaking in large integrated plants that produce millions of tons of
steel per year.
A notable trend in electric steelmaking involves the use of "high-powered" and
"ultra-high-powered'l furnaces. New steelmaking furnaces now are being designed and
built to operate from transformers of much greater power input than was common only
a few years ago. For example, 80,000 kva on a 22-foot furnace (say, 150 tons per heat)
is a high-powered installation, and there is some discussion that such a furnace may be
capable of handling power input at a level of even 150,000 kva. Such installations permit
much faster meltdown of the scrap charged to the furnace, shorten heat times, increase
the productivity of the furnace, and, hopefully, will lower steelmaking costs (but such
units do involve increased capital investments and some experimentation that truly is
"large scalell).
To the extent that steel is made in electric furnaces instead of in BOF or open
hearths, the potential for air pollution is lowered. The BOF and the open-hearth pro-
cesses both have large volumes of effluent gases, because both involve large-scale com-
bustion and oxidation. However, even though electric furnace steelmaking produces
lower gas and particulate emis sions the problem of pollution control is a difficult one.
Top-charge furnace emissions combined with those from electrode parts, slagging
spouts, and tapping spouts, make the entrapment and containment of emissions difficult.
Mini Steel Plants
Although there were a few such plants prior to the 1940s, about 20 new small steel
plants have been built in the United States since World War II. These are described in
the industry's vernacular as "mini steel plants II. They are nonintegrated plants in that
they do not smelt ore, but generally melt iron and steel scrap in electric furnaces to
produce their steel. Steelmaking capacity per plant is low when compared with large
integrated plants and averages about 100,000 tons per year. Although a few of the mini
steel plants are in areas that also contain large integrated plants, most are in areas
remote from high-tonnage steelmaking and have been located primarily so as to be close
to specific local markets. Because they depend on steel scrap melted in electric fur-
naces, the availability and price of scrap and of electric energy become important deter-
minants in the location of such plants. While their products are sold on the open market
in competition with integrated steel plants, their product line is limited and usually con-
sists of reinforcing bars, merchant rounds, and light shapes.
The total annual production of mini steel plants at the present time is variously
estimated at about 2 or 3 million tons. By 1980, it is expected that the annual production
of steel by mini steel plants will rise to about 4 to 5 million tons. The number of such
plants is expected to increase to about 35 by 1980.
This particular type of steelmaking may pre sent a problem with air pollution. The
plants are usually located in remote areas and to date generally have not paid much atten-
tion to pollution of the air. However, the compact nature of the plants and the nature of
the processing that they use lend themselves technologically to virtually complete control
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV - 18
of emissions. The main deterrent is that these plants exist and profit mainly because
of low capital investments per ton of steelmaking capacity, and substantial investments
for air-control equipment could change the entire picture of their e,conomic viability.
The situation is not unlike that faced by the ferrous foundry industry in controlling emis-
sions from cupolas. Although requirements for air quality are expected to be a factor
tending to slow the future growth of mini steel plants, it is expected that this dete rrent
will be offset by other factors and that such plants will increase substantially in number
during the next two decades.
Sponge Iron for Steelmaking
"Sponge iron" is a product of the solid-state reduction of iron ore. The reductant
is usually carbon, carbon monoxide, hydrogen, or some combination of these. Because
melting is not involved, the product contains the gangue (such as silica) that was origi-
nally present in the ore. The terminology regarding this type of product is inconsistent.
For the present purposes, the term "sponge iron" is taken to include "pre reduced ore",
"pre reduced pellets", "metallized pellets", and the like.
The use of sponge iron for producing carbon steel in electric furnaces has been
practiced in Mexico for more than 10 years. Swedish steelmakers have used sponge iron
in electric furnaces and in open hearths for producing alloy steel since 1941. The Steel
Company of Canada has conducted many commercial-scale experiments for producing
carbon steel from sponge iron in the electric steelmaking furnace, with successful
re suits.
When sponge iron or similar materials are used in steelmaking, they generally will
be used in much the same way that steel scrap presently is used. An exception is that
some forms of sponge iron (particularly metallized pellets) lend themselves to continuous
charging to the melting furnace, in contrast to the conventional batch-charging methods
now used for scrap. From the standpoint of air pollution, replacement of some or all of
the scrap in a steelmaking operation with sponge iron is unlikely to increase problems
of control of emissions and offers an opportunity (for example, through continuous charg-
ing) for lowering the amount of such emis sions.
The Oregon Steel Division of the Gilmore Steel Corporation is building an electric-
furnace shop near Portland, Oregon, that will use sponge iron for producing carbon
steel. The associated sponge-iron facility will have a rated annual capacity of about
450,000 net tons and therefore is a larger operation than the so-called mini steel plants.
The success of the Oregon Steel Company operation and other similar facilities now under
construction around the world will determine the amount of sponge iron used in the future
for electric-furnace steelmaking. Some large integrated steel companies (such as
Armco) also are considering the use of sponge iron in their electric-furnace meltshops.
It is reasonable to expect that about 2 million tons of sponge iron may be consumed in
steelmaking by 1975 in North America.
Open-Hearth Steelmaking
Steelmaking by the open-hearth process has declined rapidly in the United States
from its peak of 103 million tons in 1956 to 70.7 million tons in 1967. This trend is
expected to continue, as has been discussed previously in this section.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV - 1 9
Continuous Casting of Steel
In the traditional method of manufacture, molten steel is teemed (poured) into
large cast-iron ingot molds, in which it is allowed to solidify. The solidified ingot of
steel then is stripped out of the ingot mold and, after suitable conditioning (surface prep-
aration and reheating to the desired uniform temperature), rolled down to smaller sizes.
Because the ingots often are very large, the rolling mill is large, massive, and, hence,
expensive. This traditional procedure has given rise to the practice of rating steel pro-
duction and capacity in terms of "ingot tons ".
Most steel still is cast by the conventional method just described, but during the
past decade a different method called "continuous casting" has emerged commercially.
This eliminates the ingot mold and the ingot. Molten steel is poured into water-cooled
copper molds, from which the resulting quickly cooled shape is withdrawn continuously
and cut to desired lengths. The advent of this new method has brought a change to the
tonnage rating system from "ingot tonsil to "raw steel", in which the latter term includes
ingots cast in the conventional manner plus billets and slabs cast continuously.
Most of the continuous casting units in production before 1968 in the United States
have been in plants with less than 200,000 tons of annual capacity. The total installed
U. S. annual capacity for continuous casting, as reported in May, 1968, was about
1.3 million tons. Continuous casting has been limited to date in the United States pri-
marily to the production of billets (about 4 by 4 inches) for subsequent rolling into rebar
and wire rod. By the end of 1968, the installed U. S. continuous-casting annual capacity
was about 3.5 million tons, and by the end of 1969 it probably will be about 13 million
tons. The rapid increase is a result of large installations for the continuous casting of
slabs and blooms in large integrated steel plants. An additional 30 million tons of annual
capacity is expected to be installed by the end of 1972.
Even the most optimistic proponents of continuous casting believe that only about
50 percent of U. S. steel production can eventually be cast continuously. However, any
process that may be a factor in the production of up to half of all U. S. steel introduces
a drastic change in process technology.
From the standpoint of air pollution, continuous casting offers considerable poten-
tial for mitigation of air-quality problems. The casting operation is conducted at a fixed
location rather than being spread out over the larger area required by conventional ingot
molds. Thus, collection systems for fumes can be more localized and, hopefully, can
be made more effective. Furthermore, the coatings used to protect molds during con-
tinuous casting appear to offer fewer opportunities for undesirable emissions than do the
coatings used on the conventional ingot molds.
Pre s sure Pouring
This process (originally developed to produce cast-steel railway car wheels in
graphite molds) is now being applied to slabs, billets, and pressure-cast hollow tubes.
The process is expected to be used only for special alloys and for special products and
not to be a major proce s s for tonnage steel production. Air pollution is not likely to be
a problem with pre s sure pouring.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV -20
. Vacuum Degassing
Originally developed for application to high-cost alloy and special forging steels,
vacuum degassing is rapidly being extended to tonnage grades of carbon steel. Vacuum
processing of steel lowers the incidence of refractory nonmetallic inclusions in the steel
and produces high-quality, clean steel with high consistency. Vacuum degassing may
provide an economical method for producing continuously cast rimming-type steel, but
because of other applications it will probably be used at an increasing rate that will be
more rapid than the growth of continuous casting of steel. Vacuum degassing is not
expected to have any major effect on air-quality problems, pro or con. Most of the par-
ticulate matter that is generated in the proces s is trapped by the steam ejectors used to
produce the vacuum. The ejectors act similarly to high-energy scrubbers in this
respect. The gas from a vacuum-degassing unit will contain carbon monoxide, but if
this gas was not emitted at this fixed location, it would be emitted somewhere else where
collection would be more difficult. Depending on the system, the carbon monoxide may
be burned or emitted from stacks. The total amount of carbon monoxide emitted during
degassing can be as high as 90 cubic feet per ton (7 pounds), or 675 cubic feet per
minute, for an average degassing time of 20 minutes.
Rolling
Hot Strip Mills. Installation of new hot strip mills in the United States presently
is at low ebb, because a considerable number of new units of this type have been erected
during the past 10 years. In that period, new wide, powerful, sophisticated units were
installed, and many of these are now in production, startup, or final stages of construc-
tion. As one example, the Granite City Steel Company rolled the first coil on its new
80-inch hot strip mill in August, 1967. This facility is powered by a total of 105,000 hp
and is computer controlled. .
The installation and modernization of hot strip mills invariably includes provision
for control of emissions. As with all steel-rolling mills, water pollution is the main
pollution problem. Emissions to the air consist largely of steam, and are confined to
the immediate vicinity of the mill.
Future trends in hot strip processing are expected to include the following:
(1) Increased consumption of electrical energy and faster processing of
hot strip
(2) Closer control of processing with the aid of computers
(3) Provisions for better control of water pollution.
Plate Mills. The tonnages of plate shipped by the steel industry have been on the
increase during recent years, and this trend is expected to continue for the next decade
or two. As with hot strip installations, new plate mills will be more powerful, will pro-
vide for control of the mill by computer, and will include increased attention to water-
pollution problems. The requirements for wider, stronger, and better shaped plate have
been met by several recent installations of new facilities. For example, Republic Steel
at Gadsden, Alabama, installed a new mill in 1967 that will produce plate up to
120 inches wide with good shape control. U. S. Steel Corporation has begun construction
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV -21
of a l60-inch plate mill at its Texas Works, and Armco will improve its plate mill at
Houston. When present plans are completed, plate capacity in the United States will be
more than 13 million tons annually.
Plate mills involve practically no problem in the control of air quality.
Pipe and Tube Mills. The recent trends involving longer lengths, higher speeds,
and faster roll changes are expected to continue. For example, the increasing market
for longer lengths of heat-exchanger and condenser tubing has influenced Crucible Steel
Company to build a new tube mill at its Trent Tube Mill at Midland, Pennsylvania.
Initially, conventional-length tubing will be produced, but as anticipated requirements
for specialized tubing in power plants and desalinization plants become a reality, the
mill will be in a position to supply a wider range of market demands. Tube diameter
will range from 1/2 to 3 inches in various materials including stainless steel, titanium,
and other metals. Roll changes necessary to alter the diameter of the tubing, which
formerly required 4 to 8 hours, are expected to require less than 1 hour on the new unit.
Designed speed for production on this mill is up to 300 fpm.
Pipe and tubing mills present no problem as far as air-quality control is
concerned.
/
Merchant, Rod, and Bar Mills. Bar-mill facilities are now undergoing a cycle of
modernization and expansion. Much of this will take place over the next 2 or 3 years.
Atlantic Steel Company's new rod mill is using a novel approach. From reheating to
finishing, the mill is designed for continuous operation. Billets proceed from a reheat
furnace directl~ to a swing-forging stand for reduction before rolling. The advantage of
the swing-forging machine is that it permits a 7 by 7-inch by 30-foot bloom to be pro-
cessed into one SOOO-pound coil of finished rod up to 1/2 inch in diameter. The swing?
forging machine reduces the bloom in a single pass. Four dies work the bloom, two
vertically and two horizontally. Other features include a high-speed roll-changing unit,
a designed finishing speed of 10,000 fpm, and a high utilization rate.
U. S. Steel, Republic Steel, Inland Steel, Allegheny Ludlum Steel, Latrobe Steel,
and others are among those planning modernization of merchant, rod, and bar mills.
Future trends involving this type of mill are expected to include the following:
(1) Higher speeds of operation
(2) Faster roll changes, with spare mill stands completely assembled and
ready for installation as needed
(3) Improvements in surface quality and in dimensional tolerances in the
heavier rod sizes
(4) Improved control by means of computers.
Mills of this type are not important contributors to air pollution.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV - 2 2
Cold-Rolling Mills. Higher speed, computer control, higher power, and thinner
gage s are the future trends expected in cold mills.
As an example of current developments, Armco broke ground in 1968 at Middle-
town, Ohio, for a new cold-rolling facility. The 86 -inch (wide) mill will be one of the
widest in the world. It will produce strip and sheet stock 24 to 80 inches wide and
0.070 to 0.200 inch thick. The five-stand mill will be powered by 41,536 hp and will
deliver strip at a speed of 5200 fpm. To facilitate production of 200,000 tons per month,
the mill will be equipped with automatic coil-handling equipment, gage-control equip-
ment, and it will use computers to log operating data. Completion is expected in 1970.
As another example, Inland Steel Company at Indiana Harbor will add a five-stand tan-
dem cold mill and a single-stand temper mill to increase capacity by 1 million tons per
year. The unit will be capable of producing 100, OOO-pound coils in widths up to
76 inches. Completion is expected in late 1970. The new mill will be operated by a
process-control computer and equipped with automatic gage control.
Modern, well-ventilated cold-rolling mills are relatively minor contributors to
air pollution.
Finishing
Coating Processes. Various forms of surface protection, such as galvanizing, tin
plating, cladding, or painting, have been applied to steel sheet for many years, but
more time and attention recently have been focused on improving old methods and devel-
oping new ones. Precoated steel strip is a good example. A number of steel producers
have facilities for applying alkyds, acrylics, vinyls, vinylalkyds, silicones, and organ-
sols. Coil coating (painting or laminating) has been growing at a rate of about 25 percent
annually over the past 5 years, and fast growth is expected to continue at least for
another decade.
Electrocoating (electrophoretic deposition) is a method whereby an electric charge
is applied to a solution of paint in a water medium, causing the resins to migrate to
any conductive surface immersed in the solution. Paint particles adhere so tightly that
the water is squeezed out. Parts are then washed and the paint is baked on the surface.
The process has in less than 2 years changed from pilot-plant investigation to full pro-
duction status and appears to have technical and economic aspects that make greater use
of the method quite likely during the next decade. As of January, 1968, there were about
36 electrocoating lines in operation in the United States, and the total is expected to
double by 1970. One forecast estimates that 50 percent of all paint applied to metal
eventually will involve the electropainting process.
Coating with lead phosphate (a new type of conversion coating) has advantages over
conventional zinc phosphate coating. The major advantage is that it does not impair the
weldability of the steel. The lead phosphate coating can be applied to the steel at the
mill, serve as a protective coating during storage and shipment, and facilitate subse-
quent forming, drawing, and coating operations.
The use of galvanized steel has increased because of its ability to resist corro-
sion and abrasion. Increased use of galvanized materials by the automobile industry
stimulated production of galvanized strip steel. In early 1967, there were 53 continuous
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV -23
hot-dip galvanizing lines and nine continuous electrogalvanizing lines in the United
States (all handling steel strip over 24 inches wide).
One major producer of cans has stated that, within a few years, tin-free steel will
account for more than 50 percent of its can production. The problem of closing the side
seam (soldering) has limited the acceptance of tin-free steel in the can industry, but
two recent developments (bonding and welding) make it practical to use this metal exten-
sively. Improved versions of tin-free steel (such as chromium-plated sheet) hold much
promise for the container industry because of their better corrosion resistance and
higher luster finish.
A chromium-coating process developed in England looks promising. Steel strip is
coated with chromium chemicals and then annealed in a hydrogen atmosphere for 5 hours.
.lne process produces a layer of chromium steel 0.002 inch thick with 38 percent chro-
mium on the outer side. The process is reported (perhaps too optimistically) to produce
a strip with the corrosion resistance of stainless steel at the cost of mild-carbon steel.
All signs point to greatly increased production of coated steels of many types by
the integrated steel companies. Because most of the methods of coating involve the use
of chemicals and solvents, there appears to be a substantial opportunity for air pollution,
unless each installation is engineered to control emissions. The emissions of concern
are most likely to be mists and organics. Solid particulates probably will not be evolved
in sufficient amount to constitute a problem. However, general statements of this type
may not apply to every type of coating process. The variety of processes is wide, some
are very new, and each installation should be considered as a separate problem until
there emerges more of a pattern of performance than exists today.
Pickling. The use of hydrochloric acid for pickling of steel has increased rapidly
in the past few years. It is estimated that about 18 million tons of steel per year now
are pickled with hydrochloric acid, which is rapidly replacing sulfuric acid as the pick-
ling agent. A recent prediction is that half the lines in the U. S. will be of the
hydrochloric-acid type by 1970. However, not all steel men are thoroughly convinced
that this is the correct approach. The relative advantages and disadvantages of hydro-
chloric acid and sulfuric acid are still being evaluated. The uncertainties involve mat-
ters of original cost, maintenance problems, economics of operation, customer prefer-
ences, water pollution, production capability, and equipment arrangement. As an
example of current thinking, the U. S. Steel Corporation's expansion at its Irvin Works
includes an 84-inch continuous pickling line that is designed to operate on either sulfuric
or hydrochloric acid.
Heating Furnaces and Controls
The trend among U. S. steelmakers to better control of quality has increased the
industry's demands for more complex heating and reheating equipment. Automation of
furnace operations, specially designed material-handling systems, and scale-free heat-
ing is now being applied more frequently and more extensively. The use of electric heat-
ing has apparently broken through an earlier economic barrier and applications are
increasing. For example, McLouth Steel will use very large electric induction coils for
heating large slabs to rolling temperature. The facility will heat 20 slabs per hour at a
rate up to 680 tons per hour. Copperweld Steel has installed an electric induction heater
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV -24
to equalize billet temperature just prior to entering the first roughing stand of an 18-
inch continuous bar mill. Ford Motor Company is installing 24 recuperative-type soak-
ing pits at its mill in Dearborn, Michigan. The new pits will incorporate a high-velocity
combustion system and a newly developed air seal for scale-free operation. Jones and
Laughlin Steel Corporation this year completed installation of a heat-treating facility at
the rod and wire mill at Aliquippa, Pennsylvania. The new furnaces are instrumented to
insure positive control of temperature and furnace atmosphere in each zone through
which the coil travels during heat treatment, thereby providing ultimate decarburiza-
tion control and uniformity of microstructure. A continuous annealing facility for cold-
finished carburized bar was installed at Republic's Union Drawn Division at Gary,
Indiana. The new furnace is a continuous type that can carburize up to 300 tons or an-
neal 1500 tons of carbon or alloy bars per month.
The future trends in heating furnaces in the integrated steel industry are expected
to include the following:
(1) An increase in the number of electric heating units
(2) More emphasis on scale -free reheating furnaces to increase yield and
to improve surface quality
(3) Improved instrumentation to control the temperature and the atmosphere
in the furnace s
(4) Faster heating cycle s.
Because of their nature (tight enclosure and use of clean fuels), heating and reheat-
ing furnaces do not constitute a major threat to clean air. Many furnaces of this type
already are only minor contributors to air pollution (if at all), and the trend in the
design and use of the new furnaces is toward further cleanliness. In particular, auto-
matic combustion controls are being applied more generally and perform well in pro-
moting cleanliness of the waste gases from combustion systems. Especially when com-
bined with the use of natural gas as the fuel, combustion-control systems have the poten-
tial for almost eliminating even fuel-fired furnaces as a source of major pollutants from
reheating furnaces in the steel industry.
Explosive Bonding
Explosive bonding offers a relatively inexpensive method of forming high-strength
metallurgical bonds between dissimilar and incompatible metals. Several of the avail-
able conventional explosives are not sensitive to shock or heat and can be handled with-
out usual precaution. Suitable chambers can be erected to control the noise and blast
and to help overcome the reluctance of some potential users to work with explosives.
This new process is expected to receive widespread application in the specialty field,
such as in cladding of carbon-steel plates with stainless steel, tantalum, or titanium.
The process is not a contributor to air pollution and its use may displace some of the
conventional bonding processes that do tend to pollute the air.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
IV -25 and IV -26
Electrical-Energy Consumption
and Generation
The consumption of electrical energy has been increasing in the steel industry as
the processing becomes more sophisticated. The consumption of electrical energy in
the steel industry has increased from an average of 305 kwhr per ton of raw steel in
1965 to an average of 334 kwhr in 1967. In 1967, the steel industry consumed about
42.5 billion kilowatt hours of electrical energy. About 28 percent of this was generated
within steel plants. This was down from 33 percent generated in-plant in 1964, and the
trend is continuing.
The future trend in the steel industry will be toward an increase in the consumption
of electrical energy (both total and per ton of steel) and a decrease in the percentage of
energy generated at the plant site. The increased consumption of electrical energy will
be caused by the following:
(1) An increase in production of steel in electric-arc furnaces
(2) Increased use of oxygen in ironmaking and steelmaking and the use
of large blocks of electricity to produce this oxygen
(3) An increase in the total horsepower of mill equipment
(4) An increase in automation and labor-saving devices that are consumers
of electrical energy.
In general, the expected increased use of purchased electrical energy in the steel
industry will help significantly to reduce this industry's contribution to air pollution.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V -1
SECTION V
SOURCES OF AIR POLLUTION IN THE INTEGRA TED
IRON AND STEEL INDUSTRY
The American iron and steel industry must handle a va,st tonnage of solid mate-
rials (many of which are 'of very fine size) and must carry out much of its processing at
high temperatures (in some instances exceeding 3500 F), often with the use of high vol-
umes of gas. This section presents an overview of some of the processing methods and
the air-pollution problems that result.
This section is divided into three parts. The first describes processes and emis-
sions in the steel industry. The second part concentrates on the use of fuels and energy
in the steel industry. The third is concerned with noxious emissions - sulfur dioxide,
carbon monoxide, fluorides, and nitrogen oxides.
Process Segments as Sources of Air Pollution
The following discussion is oriented on the basis of major processing segments
used in the manufacture of iron and steel. More detailed information concerning the
types of steelmaking equipment, identification of emissions, amounts of emissions, and
methods used for their control are given in Appendixes A, B, and C of this report.
Receipt, Storage, and Handling of Raw Materials
Raw materials arrive at steel plants by boat, barge, or railroad, with some minor
amounts of materials arriving via truck. The raw materials are unloaded to open
stockpiles, with the exception of a few materials such as lime which is shipped in closed
hopper s or container cars and then transferred to enclosed hoppers without exposure to
the atmosphere. Most of the materials received are in a presized condition; most of
the crushing and sizing having been done at the mine, quarry, or pelletizing plant.
Transfer of the materials into and out of storage and to the processing centers
creates a somewhat persistent dust problem because of the fine material always associ-
ated with bulk handling of are, coal, and limestone. Dusts from iron ores are oxides
of iron, silicon, calcium, and magnesium. Dusts from limestone and dolomite are
carbonates of calcium and magnesium, magnesium oxide, and silica. Dust from fluor-
spar (which is used as a fluxing material) consists of calcium fluoride, calcium car-
bonate, and oxides of iron, aluminum, and silicon.
One of the more persistent sources of dusting is coal, which, because of its small
size (one-half inch or less as received from the mine), offers ample opportunity for
dusting during handling and storage. Open stockpiles of coal may reach a height of
100 feet and cover up to 10 acres.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-2
Because stockpiles of raw materials are exposed to the weather, and can be af-
fected seriously by dry conditions, various techniques have been tried to suppress dust-
ing. These methods have included simple wetting with water and spraying with special
plastic materials, but none have been truly successful. The large scale and available
methods of stockpiling and reclaim.ing make it economically impractical to house or
shroud the stockpile s .
Transfers from stockpiles usually are by means of overhead clam-bucket gantries
to bottom-dump cars, or by endless conveyor belts for upward movement, and gravity
chutes for downward movement. Dust is created at each transfer point. Outdoor belts
conventionally ar e cover ed but not enclosed, and dusting can occur during windy weather.
Emission of particulates to the atmosphere from materials once received in a particular
building usually can be controlled. Emission control at indoor transfer points often is
controlled by cyclone dust collector s. At transfer points inside a coal crushing and
grinding plant, the dusting can be controlled by the use of wetting agents combined with
water sprays.
Although dusting from raw-material stockpiles is a problem of concern, there is a
shortage of reliable data on the amounts of dusts that become airborne.
Coking Process
Coke is the major source of fuel in the blast furnace. Over 98 percent of the total
production of metallurgical coke is made by the by-product oven process. Unlike the
older beehive process, the valuable ingredients in the volatile matter from heating coal
in the absence of air are recovered in the by-product oven rather than exhausted to con-
taminate the atmosphere. After the extraction process, the gas is used for underfiring
the coke ovens, as well as for heating in other proces ses in the steel plant. This gas is
free of particulates, but usually contains some hydrogen sulfide.
A modern by-product oven may receive a charge of up to 40 tons of coal through
four or five ports aligned at the top. Each oven may be up to 60 feet long, 18 feet high,
and 20 inches wide. As many as 100 ovens may be set together in a battery. Flues
between adjoining ovens are heated by hot combustion gases. Regenerative checkerwork
located under the battery preheats the combustion air. Charging vehicles called larry
cars have separate hoppers for bottom discharging into each port of the oven. Charging
is done with the oven walls at an incandescent temperature. The coal has a top size of
about 1/8 inch. After a coking period of 16 to 20 hour s in the sealed ovens, door s at
both ends of the oven are opened and a ram pushes the entire load of coke onto special
railway car s. At a quenching station the load is deluged with water to cool it, after
which the coke is crushed and screened to size.
Coke-Oven Charging
Steam, flame, smoke, and fine particles of coal rush out of the oven ports during
the charging cycle. During the coking period, leakage of vapors may occur at the lids
sealing the charging ports and at door seals. Characteristics of the emissions from
these specific sources are not available. The origin of particulates for all operations
in a coke plant appear to be about 40 percent from coke, 30 percent from coal, and
30 percent from other mineral dusts normally found in a steel plant.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-3
Leakage of toxic, obnoxious vapors at door and port seals during the early coking
stages (when the oven is under some positive pressure) can be minimized by adequate
attention to maintenance of the seals. A satisfactory solution to the problem of emis-
sions during charging has not been found. Some improvement results from the use of
steam jets in the raw-gas ascensio'n pipes to decrease the pressure in the open oven
during charging. A method receiving attention in Europe involves the use of a larry
car that drops the coal through a sleeve that is lowered into the oven, and which collects
the emissions in a disintegrator after exercising controlled combustion. Without con-
trolled combustion ahead of the collector, an explosion could occur. Unfortunately,
space and structural limitations of present American ovens may not permit direct adop-
tion of this system. A pneumatic charging method is being worked on in the United
States, but details on its development have not yet been released.
The emission during charging and during the early stages of coking (when the
ovens are under positive pressure) is considered by Battelle and by steel companies to
be a serious problem that may require a long time to reach a satisfactory solution.
Coke-Oven Pushing
When the incandescent coke is pushed out of the oven onto an open hopper car, a
strong draft is induced in the immediate surroundings. Fine particulates and smoke
are blown high into the atmosphere. Some of the particulates come from coal dust that
had settled in the area because of prior coal handling. Most of the particulates come
from abrasion of the coke during pushing and from incompletely coked coal. If the
amount of uncoked coal is great, then smoke may also be emitted for a short time dur-
ing combustion of the green coke. Good coking practice results in reduced emission of
particulates and smoke, but details on the sizes and concentrations of the emissions
are not available.
In Europe, a hood at one plant was used over the hopper car holding the coke. This
experiment has been abandoned. In a high wind, most of the smoke does not reach the
hood. The near-future prospects for development of a pushing system having good con-
trol of emissions is considered to be unencouraging. The need for control is great.
Coke Quenching
The open hopper car contains the incandescent coke on a sloping bottom with side
gates made of grating. The car is moved to the quench tower which is a large brick
chimney that will accept the car at its lower end. Water sprays in the chimney deluge
and cool the hot coke. The cloud of steam rising in the chimney lifts particulates of
coke into the atmospher e, but the particulates tend to fall out in the vicinity of the
tower. About 87 percent of the particulate emission occurs in the first minute of a
2-minute quenching time.
Baffles installed in a trial quench tower reduced the emission of particulates into
the atmosphere from 6 pounds to about 1-1/2 pounds per load of coke. This can amount
to a capture of 900 pounds of particulates per day for one tower. A still more efficient
control system is desirable. However, the localized fall-out of particulates results in
coke quenching being rated as a relatively low contributor to air pollution.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-4
Coke Handling
The quenched coke is dumped onto a sloping brick wharf from which conveyors
transport it to sizing and screening operations. Steel and rubber shrouding at the trans-
fer points, at the crusher, and at the screens minimize dust loss. Indoor transfer
points may use a cyclone to collect the dust. Fines are removed before the sized coke
is transferred to stock or to the blast furnace. The fines are usually dumped via
chutes into open hopper cars for transport to the sinter plant.
Although data are not available on the amounts of emission, dusting during coke
handling is considered not to be a serious problem, except in windy weather.
By-Product Processing
The gases and volatile products that come off the coal during coking are drawn by
downstream exhaust fans into ascension pipes at both ends of each oven and into collector
mains. The gases are carbon monoxide, hydrogen, methane, hydrogen sulfide, ammo-
nia, and nitrogen. The volatile products are anthracene and other tarry compounds,
benzene, toluene, xylene, naphthalene, phenols, and pitch. Recovery of these by-
products by well-established chemical techniques starts as the cooled raw gases leave
the mains.
Usually only minor emissions can occur at the primary end of the system, be-
cause it is under negative pressure. Some odor of free vapor is present at the tar col-
lectors and decanters and at locations where the liquor runs in lines that are not fully
closed. Ammonia and organic fumes are particularly strong at sumps where decanted
liquor and other flush liquor is recycled to the sprays for cooling the collector mains.
Emission of ammonia is minor because leakage is detected promptly and the leaks are
repaired. In older plants, the addition of strong sulfuric acid to the ammonium sulfate
precipitator tank can cause considerable fuming of the acid. In large plants, phenol and
pyridine bases dissolved in the ammonium sulfate precipitator are recovered in sys-
tems that are closed except for tank vents which do not create an emission problem.
In small plants, the weak liquor containing ammonia and phenols is added to the water
that is used to quench coke.
Tar s are heavily loaded with polycyclic aromatic hydrocarbons that are consid-
ered to be hazardous. Emissions from vents on processing and storage tanks are led
through scrubbers to absorb or destroy the fumes. However, the tars tend to condense
and foul the scrubbers, whereupon the scrubbers discharge the raw vapors.
Vapors of the light oils are toxic and flammable. Condensers and some of the
process tanks are vented. The sweet odor of the aromatic vapors often pervade a large
area. Some leakage and vapor loss is inevitable in the by-product system. Forced
ventilation in the pumphouse is a necessity. Major leakage or fires are rare because
the high hazard level prompts strong pr eventative measur es.
The stripped gas from the by-product plant enters the plant fuel system. It has a
heat content of 500 to 550 Btu per cubic foot. Tar also finds some use as a fuel for
open-hearth furnaces. The tar used for fuel is not cleaned. It contains about 0.6 per-
cent sulfur, which becomes sulfur oxide upon combustion.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-5
High-volume petrochemical processing has drastically lowered the market prices
of many organic chemicals. By-product recovery from coke-oven gas is no longer as
economically attractive as in the past. However, most coke plants have by-product
systems, and these systems do provide credits to the coking process.
By-product processing of coke-oven gas results in the escape of a relatively
small amount of obnoxious vapors, chiefly from vents, sumps, and seals. Data on the
concentration of these vapors are not available. Because the vapors are obnoxious,
they are considered to be a serious contributor to air pollution and are, therefore,
collected for disposal or flared to the atmosphere.
Preprocessing of Raw Materials
Sintering
Sintering plants convert iron-ore fines and metallurgical dust, including mill
scale, into an agglomerated product that is strong enough and large enough for charging
into the blast furnace. The iron-bearing materials are mixed with coke dust to form
a bed on a slow-moving grate in the form of an endles s belt. The bed is raised to its
kindling temperature in a gas -fired or oil-fired ignition furnace; then a down draft of
air keeps the bed burning. A screen catches the sintered material when it is dumped
from the traveling grate. Undersized material is recycled. Cooling air is blown over
the sinter as it passes through an enclosed cooler on a moving apron or grate. The
cooled sinter is dumped directly into cars or conveyors for.transport to the blast fur-
nace. Plant capacities range from 2000 to 6000 tons of sinter per day.
Sintering machines process a wide variety of feed materials and produce a con-
siderable amount of emissions. The quantity and nature of the emissions are variable
from plant to plant.
Minor amounts of dust are created in the handling and grinding of raw materials.
'"'cnissions include dust sucked through the grate bars into the windbox, combustion
gases from ignition and firing, and dust generated in the screening and cooling opera-
tions. Complete combustion exists during sintering and makes it unlikely that the
exhaust gas contains unburned hydrocarbons. However, the coke-oven gas used for
ignition and the sulfur in the sinter-mix coke contribute to the presence of sulfur dioxide
in the combustion produCts. Sinter dust may contain particles of iron oxides, calcite,
iron-calcium silicates, and quartz. Dusts going into the atmosphere from the dust
collector are probably similar in composition but finer in size;
The concentration of particulates in the windbox under the grate during ignition
appears to be about 0.5 to 3 grains/sc!':'. According to one report, about 20 pounds of
dust per ton of sinter is produced in 225,000 sC£ of gas per ton of sinter; this gives a
concentration of 0.62 grains/sC£ from the windbox and traveling grate. The size dis-
tribution was 3.7 percent over 420 microns, 22.6 percent at 178 to 420 microns,
36.8 percent at 76 to 178 microns, and 36.9 percent under 76 microns. At the dis-
charge from the grate, the size distribution was about 50 percent over 100 microns and
.scf = standard eu bie foot.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-6
10 percent under 10 microns. At the discharge from the enclosed cooler (at an exhaust
volume of 2000 sdm':' per ton of sinter), the particulate concentration is 5.8 to 6.6
grains/ sd.
Multicyclones, electrostatic precipitators, venturi scrubbers, mechanical col-
lectors, and baghouses have been used in various combinations at the various points pf
emission. Wet scrubbers have high maintenance cost from lime buildup and add to
water-treatment problems. Baghouses suffer from the abrasiveness of the dust. The
abrasive dust can also create a major problem in maintaining operation of the fans in
the sintering machines. With sinter that contains flux, the efficiency of electrostatic
precipitators decreases with increase in the basicity of the sinter. In subsequent
handling of sinter and sinter fines, spraying with water containing wetting agents has
been used effectively in some cases to reduce dusting at transfer points.
The state of the art for effective, efficient control of emissions is considered to
be at a low level. Operation and maintenance costs are high. Control of particulate
emissions to a level under 0.05 grain/sd is possible, but can be so expensive that
pellets might become a more economical burden material than sinter.
Pelletiz ing
Pellets are made by rolling fine concentrates of iron ores mixed with a clay
binder (usually bentonite) to form damp balls about 1/2 inch in diameter. The balls are
dried and fired (usually in a shaft furnace) to harden them. Pellets are made at plants
located at the mine sites. In crushing the ores, fines that are unsuitable for pelletizing,
because of mineral characteristics, are shipped to steel plants for processing into sin-
ter. Although pellets are strong and attrition is low during loading for shipment to and
receiving at the steel plant, the large volume of material handled in a shipment can
result in some airborne dust.
Concentrates received at the pelletizing plant are usually in a moist condition, so
that dust generation during receiving frequently is not a problem. Bentonite is received
in covered hopper cars and is unloaded into special bins that meter the bentonite into the
pelletizing operation. Particulate emissions are magnetite, hematite, and bentonite.
The minor amounts of dust generated in the plant are handled usually by simple cyclones
or baghouses. The induration operation (heat hardening) is conducted at relatively low
air -flow rates so that formation of particulate emissions usually is not substantial.
The concentration of airborne dust generated by pellets at shipment-transfer
points is not available, but the accumulative amount is considered to be so great that
the fall-out is a serious problem.
Ironmaking
The blast furnace is a refractory-lined structure about 100 feet high and up to
30 feet or more in diameter at the hearth. Raw materials are charged at the top
through a series of seals called bells. Air, preheated in regenerative stoves, is forced
through ports called tuyeres arranged around and just above the hearth. The air
"scfm = standard cubic feet per minute.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-7
(usually augmented with oil, gas, oxygen, or steam) rises through the charge and reacts
with the hot coke to generate hot reducing gases which, in turn, react with the oxygen
in the ore to start the formation of metallic iron. As the burden (the charge) moves
downward into the fusion zone, the iron becomes molten and collects in the hearth. The
limestone in the burden reacts y.rith impurities in the ore and coke and forms a molten
layer of slag on the pool of iron. Periodically, the hot metal (molten iron) is cast
(drawn off) into special enclosed railway tank cars that deliver it to the steelmaking
furnace. Slag is cast into steel pots and transported to a dump area, or is granulated
with water. The gas ascending in the blast furnace is removed at the top, stripped of
dust, and then used to fire the regenerative air-blast stoves of the blast furnace. The
blast-furnace gas also is used as fuel in the powerhouse and as fuel for other opera-
tions in the plant.
Char ging
The charging practice for a blast furnace can vary widely. Modern burdens con-
sist of screened ore, sinter, pellets, or a mixture of these. Some blast furnaces op-
erate on 100 percent pellet burdens or 100 percent sinter burdens.
Raw materials are moved from the stockpiles to surge hoppers (called pockets)
at the blast furnace. Coke is usually transferred to the pockets by conveyor belt; ore
and flux by bottom-dump cars. Materials drawn from the pockets are weighed and
. transferr ed to the top of the blast furnace by a skip hoist, or, in some newer installa-
tions, by a conveyor belt. Usually a scale-mounted car is used for the transfer of ore
and flux from the pockets, and a conveyor belt for the transfer of co~e.
The raw materials are dumped into a receiving hopper at the top of the furnace.
In step-wise fashion, the materials are dropped through a series of two or more sealed
hoppers. This system minimizes the escape of furnace gas as charges are added and
pas s succes sively from one sealed hopper to the next sealed hopper.
The transfer at the top of the furnace is highly exposed, but partial shrouding is
possible. Leakage of gases, which also contain particulates, can develop in the bell
system as a result of abrasion wear, creep, and distortion. In high-pressure blast
furnaces, three-bell hoppers are used, and a steam system maintains back-pressure
between two closed bells during each transfer. Leakage from instrumentation, such as
from ports for rods used to determine the height of the burden in the furnace, is con-
sidered to be minor.
The particulate emissions attributable to blast-furnace charging operations con-
sist of ore dust (FeZ03 and Fe304), coke dust (chiefly carbon), and limestone dust
(CaC03 and MgC03). . Shrouding at transfer points aids in keeping air pollution to a
moderate level.
Smelting
The cleaned blast-furnace gas is used for firing the regenerative stoves that pre-
heat the air blast. If the gas is too lean, it is enriched with coke-oven gas or natural
gas. Because the gas is burned with excess air, the chimney effluent is relatively free
of objectionable particulate or gaseous components, except for the presence of sulfur
oxides if sulfur-bearing coke-oven gas is used to raise the caloric value.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-8
Auxiliar y fue 1 s added to the air b la s t no rmall y do not alte r the type s of emi s s ions.
At start-up, however, inj ection of large amounts of auxiliary fuels may cause incom-
plete combustion and create soot. The soot usually will blow through the furnace and
foul the dust-collecting system. Operating experience is expected to resolve such
occurr ences.
Slips are sudden movements of the burden in the furnace. Slips generate high
pressures that are relieved by bleeders (safety valves) in the uptakes (gas-collection
mains) at the top of the furnace. Relief of the pressure is accompanied by the emission
of dust and gas into the atmosphere. Improved raw materials lower the dust loading in
the furnace and minimize abnormal operating conditions that cause slips. Control of
blast furnaces has advanced to the. stage where bleeders on some furnaces rarely open,
or open infrequently and only for short time intervals. Some blast furnace operators
deliberately use slips (i. e., controlled slips) to facilitate operating practice. However,
these controlled slips rarely cause the release of emissions to the atmosphere.
The blast-furnace gas consists chiefly of steam, nitrogen, carbon monoxide, and
carbon dioxide. The carbon monoxide content is about 25 to 30 perce.nt. The hydrogen
content is higher with higher moisture content in the blast, and ranges from 1 to 4 per-
cent. It is likely that at least some of the gas discharges continuously through leaks at
the bells and in areas where the furnace shell is pierced for instruments and coolers.
Dust entrained in the blast-furnace gas results from abrasion of the burden during
charging and during the early stages of its passage down the blast furnace. Improve-
ment in the burden by increasing the amounts of sinter and pellets decreases the amount
of dust generated. For example, the addition of 3000 pounds of sinter plus pellets per
net ton of hot metal can decrease the dust from 300 to 30 pounds per net ton of hot metal.
The amount of dust i.n the blast-furnace gas generally varies from 30 to 90 pounds
per net ton of hot metal. The dust typically contains 15 percent metallic iron, 40 per-
cent red iron oxide (hematite) that is usually finer than 2 microns, 40 percent magnetic
iron oxide (magnetite) that is partially or completely covered by red iron oxide, and
5 percent limestone that is also usually covered with red iron oxide.
The gas is cleaned to a dust concentration of less than 0.01 grain/ scf to assure
that clogging and slagging reactions do not occur when the gas is used as a fuel to fire
the regenerative blast stoves. About 30 percent of the gas is used to heat the stoves;
the rest is used as fuel for other in-plant heating purposes.
Systems for emission control depend upon the operating practice and the available
space around the blast furnace. Commonly a system will include a dust catcher and a
primary washer. Additional equipment is used as dictated by the in-plant use of the gas.
A system consisting of a dust catcher, two automatically adjustable venturi scrubbers,
and a gas -cooling tower has been claimed to clean the gas to a dust concentration of
0.005 grain/ scf. It is concluded that control of emissions during smelting in the blast
furnace is quite good.
Casting and Flushing
The molten iron in the blast furnace is saturated with carbon. As soon as the hot
metal emerges from the furnace, graphite flakes (called kish) are rej ected and rise to
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-9
the surface where currents of heated air sweep the kish into the atmosphere. Manga-
nese vapor is also given off and oxidizes to form a fine dust, but the amount is quite
small. The kish is a nuisance because it is readily carried long distances by a light
breeze, and its oily tenacity makes it difficult to remove after it has settled. The emis-
sions of kish consist of about 90 percent graphite, 5 percent Fe304, 5 percent FeZ03,
and trace s of quartz and calcite.
Kish control consists of running the iron short distances to minimize the amount
that can form before the iron flows into closed, preheated II submarine II ladles. Kish
rising to the opening of the submarine ladle can be raked out and disposed of before it
becomes airborne. Older plants that use long runners to open-top ladles release con-
siderable amounts of kish to the air. A need exists for control of kish emissions during
casting. Transfer of hot metal to a mixer ladle or from the submarine ladle to pouring
ladles that are used to charge steelmaking furnaces releases a moderate amount of
kish. Some steel plants are starting to give attention to control of kish emis sions at
this latter transfer point.
The slag volume from the blast furnace, somewhat less than the volume of hot
metal, is flushed out of the furnace usually twice for each cast of hot metal unless fast
operations require frequent casts. The sulfur load on the slag is 6.5 to 9.5 pounds per
ton of hot metal. The content of sulfur in the slag range s from 1. 0 to 1.8 percent.
During flushing, the sulfur in the slag reacts with oxygen in the air to form sulfur
dioxide near the slag runners. In damp weather, hydrogen sulfide may also be formed.
Long runners at older furnaces give greater surface exposure and increased fouling of
the air. Most slags are transported in open ladles to dump areas where hydrogen sul-
fide is evolved as the slag cools and weathers. In newer practices, the slag runs a
short distance and is granulated with high-pressure water. However, the formation of
hydrogen sulfide may continue in the granulation pit at low temperature.
Data on the amounts of hydrogen sulfide formed upon granulation of the slag or at
the slag dump are not available, but even low concentrations have an unpleasant odor.
This emission is considered to be a serious problem and research to control it is being
conducted under the sponsorship of the American Iron and Steel Institute.
Pigging of Iron
Only a small portion of the molten iron from the blast furnace is solidified into
"pigs" for distribution in the solid state. The kish problem discussed earlier occurs
during pigging. Hooding has resulted in modest effectiveness in the control of kish
emissions during transfer from the open ladles to the pigging machines. The problem
in this area probably will be solved when an effective control is developed to control
kish emissions in the transfer of hot metal to steelmaking furnaces.
Other Ironmaking Processes
Two plants are under construction in the United States for production of iron by
different direct-reduction processes that bypass the blast furnace. It is estimated that
their combined output will be about 0.3 percent of the total metallic iron produced in
the United States. In the Dwight-Lloyd McWane (D-LM) process pellets of powdered
ore and powdered coal are partially reduced in a sintering machine. Final reduction
and melting are then completed in an electric smelting furnace to produce pig iron.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-10
The raw materials are ground wet. Particulates in the products of combustion during
reduction are small. Data on the amount and characteristics of particulate and gaseous
emissions are not available. The overall process probably contributes little to air
pollution.
.......
The other method is known as the Midland-Ross "Midrex" process. .It uses re-
formed natural gas as the main source of heat and reductant in shaft furnaces to
produce metallized pellets that are processed in a manner similar to scrap in steel-
making furnace s .
Because direct reduction of iron ore on a high-tonnage basis is still to be made
operational in the United States, the problems of emissions from such plants are still
to be determined. The future likely will see a modest growth of such plants in the
United States, but for the next decade their number and output likely will be so small
as to involve attention only on the local level. Because such plants are being born in a
time that involves much attention to control of the quality of air and water, one hopes
that the innovators at the se new plants will incorporate the latest in control systems.
Battelle's expectations are that plants of this type will be considerably cleaner than
conventional plants based on blast furnaces.
Steelmaking
Open Hearth
With the increase in new BOF (basic oxygen furnace) steelmaking capacity, the
percentage of American steel made in open-hearth furnaces has decreased, and was at
55 percent in 1967. Some forecasters have predicted the virtual extinction of the open
hearth furnace by 1990. This is fortunate from the standpoint of air quality, because
the physical nature of the open-hearth shop and the large volumes of gas involved in the
process render control of air quality difficult and expensive. Although the advent of the
BOF process has exerted strong pressure favoring retirement of even existing open
hearth furnaces, the cost of controlling emissions adequately probably has applied the
coup de grace as far as open hearth furnaces are concerned.
Open-hearth furnaces differ greatly in capacity, with the median being between
100 and 200 tons. Time to produce a heat ranges from 8 to 12 hours, which can be
shortened by lancing the bath with oxygen.
Special charging machines dump the solid raw materials through side-placed
doors onto the hearth of the furnace. The original charge usually includes solid pig
iron, iron ore, limestone, scrap iron, and scrap steel. The flame from combustion of
oil, tar, coke-oven gas, natural gas, or producer gas travels the length of the furnace
and heats and melts the charge. The hot gases are led through regenerative chambers
called checkers. Upon reversal of the flow of the flame (once every 15 to 20 minutes),
the heated set of checkers preheats the combustion air. When the solid material has
melted, either (1) two batches of additional solid pig iron and/or steel scrap are added
in a cold-metal furnace, or (2) molten hot metal from a blast furnace is added in a
hot-metal furnace. The hot metal is poured from a large ladle into a spout set
temporarily in the furnace door. After these additions, release of carbon monoxide
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-ll
by oxidation of carbon in the molten bath creates a gentle boiling action called the ore
boil. This is followed by a more violent turbulence, when carbon dioxide is released
by calcination of the limestone, called the lime boil. After the heat is refined to the
desired composition with the slag conditioned for partitioning of impurities into it, the
molten steel is ready for deoxidation and tapping. The tap hole is at the base of the
back wall and is located above a pit that contains a ladle. Ingot molds are filled from
this ladle.
To shorten the time of a heat, oxygen is injected into the bath by water-cooled
lances extending through the roof of the furnace. High-carbon heats are usually not
lanced. When lancing is used, it may occur during the refining period, or it may start
at the time of hot-metal addition. In the latter case, oxygen consumption may range
from 600 to 1000 cubic feet per ton (900 to 1667 sdm during injection), and production
rates of over 90 tons per hour are possible in a 300-ton furnace.
Minor emissions of particulates occur during the charging of the furnace and dur-
ing tapping of the heat. The dust consists of two kinds of particles. One is magnetic
iron oxide having jagged edges and generally, coated with red iron oxide. The other is
red iron oxide and is rounded and usually smaller than 2 microns. They occur as free
particles or as simple agglomerates.
Three components make up most of the dust during the period of hot-metal addi-
tion to lime-up: (1) loose agglomerates of transparent grains of hydrated iron oxide,
usually smaller than 1 micron, (2) rounded, transparent grains of red iron oxide,
usually less than 1 micron, and (3) rounded, opaque spheres of magnetic iron oxide
which may be coated with hydrated iron oxide or red iron oxide. Lime dust and sulfur
in the form of sulfates are also present in small quantities.
Dust from the period of tap to charge (during which repairs are made to the
hearth) is similar to the dust during charging of the furnace.
Particulates in the combustion product usually consist of about 85 percent rounded,
transparent red iron oxide smaller than 1 micron, and 15 percent opaque spheres of
magnetic iron oxide ranging in size from 3 to 5 microns. The smaller grains appear
to be orange in top light and tend to form loose agglomerates.
Composition of the products of combustion was not available for open hearths in
the United State s. It is known that the gas usually contains sulfur compounds from the
use of fuels such as oil, tar, and coke-oven gas. Firing with oil is reported to create
a lower average dust loading than firing with tar.
The amount of dust generated varies at different stages of the steelmaking process
and with the particular practice. One source gives a dust loadi:t~g of 0.78 grain/sd at
60,000 sdm gas flow at meltdown, 1.9 grain/ sd at 64,000 sdm gas flow at hot-metal
addition, 2.70 grain/sd at 66,000 sdm gas flow at the lime boil, up to 5 grain/sd dur-
ing lancing, and 0.21 grain/sd at 64,000 sdm gas flow during refining. Typical dust
loadings for oxygen-lanced furnaces are estimated to be 20 to 22 pounds per net ton of
raw steel. Without oxygen lancing, a value of about 8 pounds of dust per net ton of
steel was reported.
The only data available on the size distribution of particulate emissions compares
a composite sample with a sample taken during the lime boil. About 46 percent of the
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-12
particles in the composite sample were smaller than 5 microns, while 77 percent from
the lime boil were below this size.
The slag pockets, checker chambers, and flues to waste -heat boilers provide
opportunities for settling out of dust from the waste gases, and, thus, served as fairly
efficient dust collectors before the advent of oxygen lancing. The use of oxygen lancing
increases the dust loading and generates large volumes of fume. Effective control of
the emissions is obtained with electrostatic precipitators. Venturi scrubbers and bag
houses are also being used.
Control technology is adequate for handling the particulate emissions which other-
wise constitute a serious air-pollution problem.
In the teeming area, ingot molds on a string of cars are filled from the ladle con-
taining the entire batch of steel from the furnace. In 1967, about 94 percent of the raw
steel produced in the United States was poured into ingot molds. The use of tar and
bitumens as mold coatings has been curtailed in the past decade to give a decrease in
the amount of visible emission during teeming. However, with some present teeming
practices, visibility can become highly restricted at the teeming station. No informa-
tion is available on the characteristics of the emissions at the teeming station or on
methods of controlling the emissions. Generally, the degree of pollution contributed
by emissions during teeming may be classified as moderate, but deserving of more
attention than it receives.
Basic Oxygen Furnace
About half of the steel in the United States will be made by the basic oxygen
furnace (BOF) proce s s by the end of 1969. The furnace is a pear - shaped steel shell
lined with refractory brick, charged through the top and tilted for tapping. The charge
usually consists of hot metal and scrap in the ratio of about 70 to 30, plus burnt lime.
A water-cooled lance impinges oxygen at high velocity on the surface of the charge to
cause violent agitation and intimate mixing of the oxygen with the molten iron. Rapid
oxidation of carbon, silicon, manganese, and some of the iron occurs. These exo-
thermic reactions supply heat to reach the tapping temperature. lmpuritie s in the
charge enter the slag. A typical ISO-ton furnace can produce a batch of steel in about
35 minutes.
Kish is released when the hot metal is charged and cooled by contact with the
cold-scrap charge. Only part of the kish is contained in the furnace. Dark brown
smoke evolves at the start of the blow from oxidation of the iron. It persists until the
silicon, manganese, and phosphorus begin to oxidize; these oxides enter the slag. Some
silica, lime fume, and small amounts of manganese enter the off gas. Then carbon is
oxidized and evolved, chiefly as carbon monoxide. An excess of air is usually mixed
with the off gases to burn the carbon monoxide as the off gases are collected. This
eliminates the possibility of an explosion from ignition within the exhaust system. Many
furnaces are equipped with waste-heat boilers for thermal recovery from burning of the
carbon monoxide. Some newer practices, not yet in use in the United States, deliberately
minimize the amount of air aspirated into the off gases to reduce the problems from
overheating of the hood and duct work.and to reduce the volume of gas that must be
cleaned. The cleaned gas is then flared ,safely into the atmosphere, or is collected and
stored for subsequent use as fuel.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-13
The kish formed during the charging of hot metal consists of angular flakes of
graphite having smooth surfaces. The emission of kish may include coarse fragments
of opaque magnetic iron oxide particles, rounded particles of red iron oxide, and
trace s of quartz and calcite.
The silica fume that occurs in the early part of the blow is collected as a gray to
off-white material. It often contains small amounts of iron, manganese, and carbon.
During the BOF blow with oxygen the predominant particulate emission is brown
iron oxide. The particulates are rounded, transparent, smaller than 1 micron, and
tend to agglomerate. Some fine, black spheres of magnetite are present and are
covered with red iron oxide. 1£ galvanized scrap is part of the charge, zinc oxide in
the cpllected dust makes the dust unsuitable for sintering for feed to the blast furnace.
The fineness of the BOF particulates makes them difficult to measure. As a
result, size analyses have varied widely. Reports show (1) 95 percent are smaller than
1 micron, with a median of 0.45 micron, and (2) 99 percent smaller than 0.2 micron,
with a median of 0.065 micron. A more recent report gives a median diameter of
0.012 micron. Another report states that 20 percent is smaller than 0.5 micron,
65 percent is between 0.5 and 1.0 micron, and 15 percent is between 1.0 and
1. 5 microns.
Dust concentration (as reported by one source) varied from 2.02 to 4.96 grains/
scf. Another source states it is up to 20 grains/scf. The volume of exhaust gas (at
temperature of combustion) is about 25 times greater than the volume of oxygen blown.
The amount of dust per net ton of raw steel was reported in 1959 to range from 14.5
to 27.4 pounds. In 1968, an average of 40 pounds per ton was reported by one plant.
Others state it is 1 to 2 percent of the weight of the metallic charge.
In the Stora-Kaldo Oxygen Process, a cylindrical vessel is rotated in almost a
horizontal position. Operating principles are similar to those for the upright BOF
vessels. More scrap can be charged than in the BOF, but the heat time is longer. The
Sharon Steel Corporation has the only two Kaldo furnaces in the United States, rated
at 150 tons per heat. The size of particulate emissions is reported to be larger than
particulate emissions from the BOF; only 6 percent is smaller than 1 micron, probably
as a result of agglomeration. About 10 pounds of dust are said to be generated per net
ton of raw steel.
Teeming of BOF steel is conducted in the same manner as was described for open-
hearth steel.
Electric Furnace
In 1967, electric furnaces produced about 11 percent of the total raw carbon steel
made in the United States and 36 percent of the alloy and stainless steels. About 59 per-
cent of the electric-furnace heats were carbon steels. These furnaces are refractory-
lined cylindrical basins having a capacity of up to 250 tons. In 1968, 40 percent were
under 50 tons, 36 percent were between 50 and 90 tons, and 24 percent were over 90 tons
in capacity.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
'-
V-14
High-current electrical energy is passed through large graphite electrodes extend-
ing down through the roof of the furnace. The charge consists of select steel scrap,
cast-iron scrap, pig iron, alloying elements to achieve the desired composition, and
flux. Melting is accomplished by the heat of the arc between the electrodes and the
charge.
Preheating of the steel scrap is not a common practice for shortening the heat
time. If preheating is practiced, it is done most commonly with air-fuel burners.
Oxygen-fuel burners are used to a limited extent. The scrap is rarely heated to above
1800 F. Thus, preheating creates no significant emission problems, unless com-
bustible materials such as oil or rubber are in the scrap.
Electric induction furnaces melt special alloys on a small scale. If the charge is
free of tramp combustibles, emission from these furnaces is minor and is readily
collected in simple equipment.
Emissions in the electric-arc furnace can originate from light scrap that oxidizes
readily, from dirty scrap (a major source), and from oxygen lancing. The emissions
are fume from scrap preheating, dust from the melting operations, and furnace off gas.
Information is not available on the nature of the dust from preheating scrap. The com-
position of the dust will be influenced by the freedom of the scrap from volatile or com-
bustible matter. The dust from operation of the furnace consists of opaque, rounded
grains that are peach to reddish in top light. Its composition is dependent upon the
composition of the steel being melted and stage s of making a heat.
The amount of dust per net ton of steel is dependent upon the condition of the scrap
and whether or not oxygen lancing is used. Particularly dirty scrap can raise the dust
emissions to as high as 30 pounds per net ton of steel. It has been estimated that
oxygen lancing produces 20 percent of the total emis sions in making a heat.
The composition of the off gas from the electric furnace varies with the slag
practice, the stage of the heat, and whether or not oxygen lancing is used. The chief
constituents are carbon monoxide, carbon dioxide, nitrogen, and oxygen. Change s in
the composition of the off gas during the course of a heat can vary considerably.
Typically the carbon monoxide content rises sharply during the boil and again during
oxygen lancing, while the carbon dioxide content usually stays below 15 percent at all
time s .
Emissions leave the furnace through the electrode ports in the furnace roof, out
of the tapping spout and slagging door, and (in the case of top-charged furnaces) through
the open furnace top during charging. Three main types of systems are used to collect
the emissions: (1) hoods over and around the furnace at points of emission, (2) direct
extraction from the interior of the furnace, and (3) shop-roof systems.
In the first system, hoods are fitted at the points of emission, and ducts pass the
emissions to the dust collector. Hoods must be movable when used with top-charging
furnaces. They also add to operational hazards because they tend to obscure the
visibility from the crane -operator's cab. Warpage of the hoods is a problem. They
seldom last 1 year.
Direct extraction by means of a duct entering the furnace roof causes air to flow
into the furnace, and thus minimizes the dis charge of emis sions through the doors and
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-15
electrode ports. This system increases roof life by 16 percent and decreases elec-
trode consumption by 8 percent. It causes lower recovery of some alloying elements
in the steel bath, and creates some difficulties with carbide slags that are used in re-
fining of special steels. Heat exchangers must be used, or a minimum duct length
(600 feet is required in some cases), to cool the gases by radiation to a safe tempera-
ture before they enter a bag house.
In the shop-roof extraction system, the shop building serves as the collector hood.
Ducts in the roof of the building exhaust the emis sions to the dust-collecting system.
Teeming practice is the same as for open-hearth and BOF steels. Some electric-
furnace shops use roof-extraction exhaust systems in the teeming building to collect
the fumes and exhaust them to the main bag house.
Heats of leaded and resul£urized grades of steel can be made in the open hearth,
BOF, or electric furnace. For leaded steels, a hood exhaust system to a bag house
collects the toxic fume during teeming. Shrouding of the stream with argon (as is
practiced to prevent reoxidation during continuous casting of vacuum-degassed heats)
is probably impractical for emission control when teeming into a series of ingot molds.
High-energy scrubbers installed on one oxygen-lanced electric furnace producing
a dust concentration of 3.2 to 6.4 grains/sdhave reduced the dust output to the range
of 0.026 to 0.0512 grains/sd. Bag houses have reduced it to the range of 0.004 to
0.0064 grains/ sd. Electrostatic precipitators, not performing as well, reduced the
dust loadings to a range of only 0.256 to 0.512 grains/sd.
Electric furnaces do not evolve the high gas volumes generated by either open
hearth or BOF furnaces. This is not to imply that the control of emissions from electric
furnaces is easier. The majority of electric furnaces are top-charging furnaces, which
means that during the charging operation the furnace is open to the atmosphere and con-
siderable amounts of emissions can be released. Other design characteristics of elec-
tric furnaces contribute to the problem of effectively controlling electric-furnace emis-
sions. These problems have resulted, in some cases, of using complete building
evacuation as a means of pollution control. In spite of the fact that many new electric
furnaces are now being installed and have been installed recently, most of these with
emission-control systems, accurate information on the amount, size, and composition
of particulate and gaseous effluents were remarkably sparse and difficult to obtain during
the present study. This dearth of reliable and complete information should be remedied
because the next decade will see continued growth of the electric-steelmaking process.
Vacuum Degassing of Molten Steel
For certain critical applications, the steel must be cleaner with respect to in-
clusions and lower in hydrogen content than is commonly achieved by conventional steel-
making practices. These objectives are attained by vacuum treatment of the molten
steel. Three methods are used: (1) stream degassing, (2) circulation degassing, and
(3) ladle degas sing.
In stream degassing, a bottom-pour ladle is sealed to a vacuum chamber contain-
ing an ingot mold. The stream of steel from the ladle is degassed as it falls into the
mold. If the stream is collected in a second ladle in the vacuum chamber, the process
is called ladle-stream degassing. The second ladle is removed from the chamber for
pouring the steel into ingot molds or into a continuous -casting machine.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-16
To achieve the highest degree of degassing, either the circulation-degassing or
ladle-degassing method is used. In circulation degassing, part of the molten steel in
the open ladle is lifted via a snorkel into a vacuum chamber. Only one snorkel is used
in the D-H process in which the steel is sucked up and returned repeatedly until the
hydrogen and oxygen contents in the steel drop to the desired levels. In the Thermo-
Flow process and the R-H process, two snorkels are used. Both processes inject an
inert gas in one snorkel to cause the steel to flow up that snorkel, into the vacuum
chamber, and down the other snorkel back to the ladle. In the Thermo-Flow process,
circulation in the snorkels is enhanced by the dire ctional stirring force from ele ctrical
induction coils surrounding the snorkels.
In ladle degassing, the filled ladle is set in a vacuum tank. Degassing time is de-
creased by agitating the molten metal by bubbling argon through it or by stirring it by
means of an induction coil.
The source of emissions is the molten steel in the vacuum chamber. The three
principal gases emitted are carbon dioxide and carbon monoxide from the reaction of
carbon and oxygen in the steel and the hydrogen from solution in the steel. Violent
agitation of the molten steel and vaporization of the metallics (because of the low pres-
sure) both generate dust, much of which deposits on the walls in the vacuum chamber.
The composition of the off gases varies with the grade of steel and the extent of
the vacuum treatment. In the circulation-degassing D-H process, the average composi-
tion is up to 80 percent carbon monoxide, up to 20 percent carbon dioxide, and up to
15 percent hydrogen. The composition of the off gases from vacuum-stream degassing
can vary from 18 to 50 percent carbon monoxide, 20 to 70 percent hydrogen, 15 to
75 percent nitrogen, and 1 to 6 percent carbon dioxide. Nitrogen and argon are
present in the off gases from air in the system or from aspirated air. Water is always
present from the steam ejector.
The dust that is deposited on the walls of the vacuum chamber is finely divided -
often so fine that it is pyrophoric and ignite s when air is admitted into the chamber. The
composition of the dust varies according to the composition of the steel being treated.
Data on the amount and size of the dust carried from the vacuum system is sparse and
difficult to obtain.
The steam ejectors that are used to create the vacuum also serve as scrubbers.
Thus, little dust is released to the atmosphere. One source of information states that
about 10 pounds of dust are collected in degassing 100-ton heats.
Depending on the amount of steel to be treated and on the level of vacuum required,
a degassing installation may have from 4 to 6 steam ejectors with intercondensers.
Usually the cleaned gases are emitted to the atmosphere. If the gases are passed
through a water hot well (to avoid a hazard to personnel), the gases are ignited and
flared to the atmosphere. The amount of gases depends on the carbon content of the
steel and the amount of hydrogen that is to be removed.
Control of particulate emissions is good and is accomplished economically by
entrapment in the hot wells of the steam ejector system. The cleaned off gas contains
considerable amounts of carbon monoxide and hydrogen which sometimes is flared for
dispersal in the atmosphere.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-17
Although vacuum degassing of steel is increasing in commercial importance and
use, its contribution both to particulate and gaseous emissions from a steel plant is
minuscule, and is expected to remain so.
Continuous Casting
In continuous casting, molten steel, poured into a water -cooled copper mold,
quickly cools to the shape of the mold and is withdrawn continuously. The estimated
capacity in the United States in 1968 is about 7 million net tons per year, and is ex-
pected to be doubled in 1969. The ultimate capacity is expected not to exceed 50 per-
cent of the total steel production.
Emissions during continuous casting are markedly lower than for teeming into
ingot molds, because the rape seed mold coating creates only a small amount of fume.
Continuous casting is conducted in one location rather than over a large area such as
used for conventional teeming. This confinement permits the use of a more localized
fume-collecting system to enhance the efficiency of collection. The tundish is some-
times blanketed with a reducing gas to minimize oxidation of the steel. To prevent
reoxidation of vacuum-degassed heats, the stream of steel from the tundish to the mold
sometimes is shrouded with argon. This also serves to inhibit escape of emissions
from this area. Steam generated by the secondary cooling sprays is collected in an
exhaust system.
Continuous casting contribute s very little to air pollution. As in the case of in-
creasing replacement of open-hearth furnaces by BOF and electric furnaces, it is
fortunate that the newer and growing technology (continuous casting) presents much less
of an air-pollution problem than the process that is tending to be replaced (conventional
teeming of ingot molds).
Pre s sure Casting
In pressure casting, the ladle of molten steel is placed in a pressure tank. When
air pressure is increased in the tank, the molten metal flows up through a ceramic tube
leading from the bottom of the ladle, through the top of the tank, and into a graphite
mold where solidification takes place. Total U. S. production is estimated to be only
about 1/2 million net tons in 1967, with an increase to about 1 million in 1969. Although
this process has been used mainly for casting stainless steel slabs and some billets and
tubes, in 1969 a plant will be placed in operation to cast carbon steels.
Emissions contributing to air pollution are considered to be negligible in the
pressure-casting process. Minor amounts of iron oxide fumes are generated by the
exothermic mixture in the hot top and from cutting off the riser and gate from the cast-
ing with a torch.
Pressure casting represents such a small percentage of steel cast that this
method deserves consideration from an air-quality standpoint only on a local basis
where the process is employed.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-18
Steel Shaping
Soaking Pits and Primary Breakdown
Soaking pits are furnaces in which ingots are brought to the desired equalized
temperature for rolling. They are fired with blast-furnace gas, coke -oven gas, or a
mixture. Data are not available on the emissions that occur during charging and re-
moving of ingots, or during firing, but the amounts of particulate emissions are con-
sidered to be negligible. Gaseous emissions are mixtures of carbon dioxide and carbon
monoxide. Sulfur dioxide is also present if the coke -oven gas has not been freed of
hydrogen sulfide. No economical commercial method is available for removal of the
sulfur dioxide in the exhaust gases.
During breakdown of the ingots by rolling into billets, blooms, or slabs, the major
emission is steam which is confined in the building. Emission of dust is minor.
Conditioning, Reheating, and Hot Rolling
After ingots are rolled to billets, blooms, or slabs, they are cooled and inspected,
and surface defects are removed by grinding, chipping, peeling, or scarfing with hand
torches. Then the steel is reheated in a once-through furnace fired with coke-oven
gas, blast-furnace gas, or natural gas. Some plants reheat slabs and billets with
electrical induction coils. Reheated slabs are usually scarfed with automatic machines
before they enter the hot-strip mill.
The fine particulates generated during grinding are airborne only a short distance,
and probably only rarely escape the building. Sometimes they are collected at the grind-
ing station. Chipping and peeling create insignificant emissions. Billet scarfing
causes a metallos s ranging from 3 to 6 percent; but most of the loss is spatter, not
fume. Usually no emission control is used in hand scarfing, except that hoods are used
sometimes if the shop practice calls for extensive hand scarfing. The loss in scarfing
of slabs is up to 2.5 percent, and up to 7 percent for blooms. One reference gives a
concentration of 0.2 to 4.4 grains/sd for scarfing slabs. Another places it at 2 to
3 grains/sd when the gas flow rate is 75,000 to 135,000 dm during the short period of
operation of a scarfing machine. The emissions are chiefly iron oxides and are col-
lected with electrostatic precipitators or high-energy scrubbers. Machine scarfing
would result in serious air pollution if collection-control systems were not used.
Emissions during reheating of steel billets and slabs are mixtures of carbon mon-
oxide, carbon dioxide, nitrogen, and moisture from the combustion of natural gas.
Emissions at the rolling mills themselves consist mostly of steam which is confined to
the area of the mills. The emission of submicron iron oxide at strip-finishing stands
is considered to be significant enough that some mills collect these emissions with
high-energy scrubbers.
Except for scarfing operations, no serious emission problem exists in the opera-
tions that involve conditioning, reheating, and hot rolling. This conclusion applies also
to hot-forging and hot-forming operations. Control systems are available to contain
the emissions from scarfing operations.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-19
Acid Pickling
Acid treatment (called pickling) is used to clean the oxidized surface of hot-rolled
steel in preparation for cold rolling. Although some pickling lines are being converted
from sulfuric to hydrochloric acid, the overall economics and advantages of the use of
one over the other have not yet been resolved. Acid fumes are the chief emissions
during pickling. Hoods exhaust these fumes usually to a wet scrubber and packed
tower. One source states the collection rate is about 100 grains of hydrochloric acid
per ton of steel. For pickling low tonnages of steel, the cleaning system may be con-
sidered to be too costly and the fumes maybe ejected to the atmosphere through a roof
exhaust.
Pickling can contribute moderately to air pollution if systems are not used to
collect and capture the acid fumes.
Cold Rolling and Cold Forming
In high-speed cold rolling, a water-oil mist is generated from the emulsion
lubricant applied to the rolls. Collection of this emission is by mechanical mist
eliminators or wet scrubbers. Neither cold-rolling nor cold-forming operations con-
tribute to air pollution.
Annealing
Annealing of steel sheet or strip is done in a batch or continuous manner. In the
batch process, a removable shell covers the coiled steel. Combustion gases heat the
shell, and a prepared. gas or protective atmosphere inside the shell prevents oxidation
of the coil. In the continuous process, the coil is annealed by passing it as a single
strand through a furnace having a controlled atmosphere in the heating, holding, and
cooling zones of the furnace. Products other than sheet and strip are usually heated
directly by combustion gases.
Air pollution from annealing operations is insignificant, except when sulfur-
bearing coke -oven gas is used as the fuel for firing the furnaces.
Steel Finishing
Continuous -strip lines lend themselve s to ready control of emissions during coat-
ing operations. Coatings are zinc, tin, terne (lead alloy), aluminum, chromium,
nickel, copper, phosphate, and a broad spectrum of paints and other organic materials.
Prior to coating, the strip is heated to carbonize grease or oil films, and the lightly
oxidized surface is cleaned in an acid tank or by the action of a reducing atmosphere in
a furnace. A hot alkaline solution may be used instead of heat to remove the grease
film before pickling.
In dip coating (zinc, aluminum, and terne coat), the preheated strip is treated with
a flux or passed through a flux layer on top of the metal bath. In electrolytic processes
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-2Q
(zinc, chromium, nickel, and copper), the cleaned strip enters directly into the plating
bath. Recent processes for coating with chromium and nickel involve coating with
chemicals that are reduced in a hydrogen atmosphere to produce the coating metal.
Phosphate coatings are produced by dipping steel in a dilute acid phosphate solution
saturated with a metal such as zinc, cadmium, aluminum, or lead. The metal surface
is converted to an insoluble crystalline phosphate coating. Paints are applied to
cleaned strip by an electrophoretic process by passing cleaned., warm strip through a
paint tank, or by roller coating. In the electrophoretic process, an electrical charge
applied to a solution of paint in water causes the resins to move to the oppositely
charged steel strip immersed in the solution. Upon contacting the steel strip, the
paint particles squeeze out the water. After washing, the paint is baked on the surface.
Painting operations emit solvent vapors which are collected by an exhaust system that
disperses the vapors to the atmosphere outside the building. During baking, the evolved
vapors are combusted in direct-fired ovens or are exhausted to the air in indirectly
heated ovens.
In the se finishing operations, emis sion of particulate s is negligible. Heating of
the steel results in stack gases having the normal compositions from combustion of
natural gas. Acid mists from pickling and vapors from electroplating baths are
readily collected by hoods and removed in wet scrubbers and packed towers. Some
particulates and/or vapors are emitted from dip-coating baths, but data are meager:
Information from two job-shop dip-galvanizing plants indicate that the size of the particu-
lates is about 2 microns. No information was available on the emis sion-control systems.
In general, the amounts of particulate emissions from steel-finishing operations
are relatively minor and appear to be susceptible to good control. Gaseous emissions
appear to be a problem only (1) when fuels contain sulfur, and (2) when the process emits
vapors from organic materials used in coatings. The former always is a problem;
the latter is amenable to good control by application of existing washing equipment.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-2l
FUEL AND ENERGY UTILIZATION
During the past 10 years, the patterns of use of fuel and energy in iron and steel
plants have been changing significantly. The most important changes include (1) the
advent of oxygen steelmaking, (2) major improvements in blast-furnace technology,
(3) increase in average iron content of materials in the blast-furnace charge, and (4) a
16 percent decline overall in the amount of fuel energy used per ton of iron and steel
produced.
From the standpoint of the steelmaker, the sulfur content of his fuels would be an
item of high concern, even if air-polluting influences of high-sulfur fuels could be
ignored. Fuels are the main source of inputs of sulfur into the steelmaking process,
and sulfur is an undesirable element in most iron and steel. Because most steelmaking
energy is derived ultimately from coal (in part by coking it and then burning the coke,
and in part by burning coal to produce electricity), steelmakers must be vitally con-
cerned with a variety of coal-based problems - (1) particulate matter from the com-
bustion of coal or coke, (2) sulfur that enters the iron or steel because of the use of
coal and coke in the process, and (3) gaseous emissions from combustion of the coal
or coke.
This discussion was prepared to analyze some of the principal trends and counter-
trends of fuel and energy use in steel plants. It will be useful as background for
evaluating the possible effects of changes in air-quality standards upon future patterns
of use of steelmaking energy. There are three parts to the analysis:
The first part presents a ,statistical summary of fuels used over the
10 years from 1958 to 1967 by type of fuel. The summary includes infor-
mation on purchased electricity and steelmaking oxygen. Although elec-
tricity and oxygen are not fuels, they have had a strong influence on overall
energy patterns in the steel industry, and their manufacture requires heavy
use of fuels and/or energy. Electricity can be used directly for melting
metal, and this use is growing. Oxygen used during steelmaking causes
some components in the metal to serve as fuels, and thus indirectly sup-
plants conventional fuels.
In the second part, the application of energy sources to key operations
of ironmaking, steelmaking, and steel processing is examined. The known
determinants of fuel choice are stated, with attention to fuel economics,
process technology, and other influences. From this part of the analysis,
it is pos sible to draw some preliminary conclusions about future evolution
in the absence of new constraints such as tighter emission standards.
The third part draws upon the first two and illustrates how considera-
tions of pollution control and process economics may be expected to inter-
act. In those situations in which present trends are in the direction of
cleaner fuels, the rate of change determines the effect on air quality. When
the present trend of emissions is unfavorable, the magnitude of economic
pres sure behind the change is important. It may well be that anticipated
changes in air-quality standards will tend to accelerate some of the present
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-22
trends and to reverse others. Attention is called, finally, to the long-term
future situation regarding U. S. fuel supplies. It is possible that short-term
decisions to use "clean" fuels, made now on the basis of pollution criteria,
may not be sustainable in the long term if natural gas and oil become scarce.
10-Year Summary of Fuel Usage
The American Iron and Steel Institute publishes annual data on the use of fuels in
iron and steel plants. These data are partly subdivided according to broad categories
of use (e. g., steelmaking furnaces). Corresponding data for use of oxygen and elec-
tricity in steel plants are also given by the AISI. Typically, these statistics cover
about 95 percent of the productive capacity of the U. S. steel industry.
Table V-I summarizes the total use of steelmaking oxygen, purchased electricity,
and each principal fuel for the 10 years from 1958 to 1967. Although the weight of coal
used to make coke is reported by the AISI, coking coal has been deleted from this sum-
mary because the fuel use of coking coal occurs in the forms of coke, coke-oven gas,
and tar or pitch. Table V -1 shows consumption of blast-furnace gas for 1966 and 1967,
but data are missing for prior years. All blast-furnace gas is ultimately derived as a
recuperation from use of other fuels in the blast furnace.
The few trends illustrated in Table V -1 are uncomplicated. There are no definite
trends in total consumption industry-wide of coal, tar and pitch, or coke -oven gas, but
there may be a downward trend (since 1965) in total use of fuel oil. The overall con-
sumptions of coke, natural gas, and LP gas'~ have clearly risen with increased steel
and iron production. Total consumption of steelmaking oxygen and purchased electricity
(listed for purposes of comparison) have shown strong and sustained increases over the
past decade.
The information of Table V-I is presented in a different context in Table V -2, in
which the data are adjusted to consumption per ton of crude steel plus merchant pig
iron. In Table V -2, clear downtrends appear in the use of coal, coke, oil, and tar per
ton of steel. No clear trends may be cited for consumptions of LP gas, natural gas,
and coke -oven gas; but (just as for total consumption in Table V -1) both oxygen and
purchased electricity show strong growth per ton of steel. The trends toward use of
more electricity and more oxygen per ton of steel and of much less liquid and solid fuel
are principal findings of this part of the study. However, it is to be noted that not all
of the changes are continuing: unit consumption of coal and tar per ton of steel seems
to have stabilized as of 1964, and coke consumption per ton of steel has been steady
since 1963.
Table V -3 presents a suggested basis for direct energy comparisons among fuels.
The stated energy contents (expressed in British thermal units) are typical for the
fuels studied, but a given lot of any fuel could differ substantially from the quoted
energy rating. The terms "high" or "gross" applied to energy ratings of fuels merely
signify that the value includes heat from condensation of the water vapor produced during
combustion. The assignment of a specific energy value to steelmaking oxygen is
*Liquefied petroleum gas or bottled gas, usually propane.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
m                   
)0                   
-i  TABLE V-I. TOTAL FUELS, STEELMAKING OXYGEN, AND PURCHASED ELECTRICITY CONSUMED BY STEEL PLANTS IN THE UNITED STATES, 1958-1967 
-i  
III                   
r    Source: American Iron and Steel Institute, Annual Statistical Report, 1962 and 1967 editions.     
r        
III                   
~         Total Consumption by the Steel Industry     
III             
~  Item Units 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 
0  
;!! Fuels                  
)0                   
r Coal(a) 103 NT 7,158 6,819 7,337 7,235 6,994 6, 882 7,045 7,372 7,392 6,878 
z Coke  103 NT 47,818 49,487 52,464 47,241 46,725 49,222 57,558 59,519 59,926 56,435 
en  
-i                   
-i Fuel oil 106 gal 1,607 1,735 1,699 1,543 1,418 1,575 I, 678 1,611 1,451 1,254 
C                  <:
-i Tar and pitch 106 gal 309 322 365 274 284 292 295 334 315 303 r
III N
I                   V>
LP gas 106 gal 8.4 8.6 8.8 7.5 7.8 10.9 10.5 18.0 13.9 14.0 
n                
0 Natural gas 109 ft3 279 317 361 399 434 464 513 547 517 534 
r 
C   109 ft3                
~ Coke-oven gas 675 750 900 708 690 610 716 788 915 886 
m                 
C Blast-furnace gas(b) 109 ft3 ------ ------ ------ - - - Unreported - .- - - - ----- ------- 4,076 4, 131 
en 
   106 ft3 13, 049 18,307 29,213 44,170 54,675 69, 761 90,626 104,396 128,681 140,466 
   106 kwhr 15,934 17,050 18,810 . 19,429 21,229 23, 652 26, 049 28, 006 29,891 30, 557 
~ Oxygen
m
o Electricity
:u
)0
-i (a) To avoid duplication, coal used in the manufacture of coke is excluded.
~ (b) Ultimately derived from coke and tuyere injectants, hence duplicative in the sense of energy sources.
III
en

-------
m
)10
~
~
111
r
r
111
TABLE V-2. FUELS, STEELMAKING OXYGEN, AND PURCHASED ELECTRICITY CONSUMED PER TON OF STEEL PRODUCED IN THE UNITED STATES, 1958-1967
Source: American Iron and Steel Institute, Annual Statistical Report, 1962 and 1967 editions. Data of Table V-I were divided by the sum of crude steel production
plus pig-iron shipments to outsiders. It is assumed that electric-furnace ferroalloys are reflected in steel production.
~               
111      Approximate Consumption per Net Ton of Steel    
~         
0 Item Unit 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 
::0               
- Fuels              
)10              
r Coal(a)              
 Lb 160 138 142 142 138 122 108 109 107 105 
Z               
en Coke Lb 1,071 1, 013 1,016 927 920 876 880 879 869 866 
~ 
~ Fuel oil Gal 17.99 17.55 16.45   15.14 13.97 14.01 12.83 ll.9G 10.53 9.62 <:
c  
~ Tar and pitch  3.46 3.26    2.69 2.80 2.60 2.26    '
111 Gal 3.53   2.47 2.29 2.32 N
I               ~
o LP gas Gal 0.09 0.09 0.08   0.07 0.08 0.10 0.08 0.13 0.10 O.ll 
0 Natural gas Ft3 3,123 3,207 3,495 3,916 4,274 4, 127 3,923 4,041 3,751 4,096 
r 
C               
~ Coke-oven gas Ft3 7,556 7,588 8,713 6,948 6, 796 5,425 5,476 5, 822 6, 638 6,796 
m 
c Blast-furnace gag{b) Ft3      Unreported - -    29,600 31,700 
en ----- - - - - ------ -  ------  
r Oxygen Ft3 146 185 283 433 538 620 693 771 934 1,077 
)10 
m               
0 Electricity kwhr 178 172 182 191 209 210 199 207 217 234 
::0              
)10               
~               
0 (a) To avoid duplication, coal used in the manufacture of coke is excluded.          
::0 (b) Ultimately derived from coke and tuyere injectants, hence duplicative in the sense of energy sources.      
111      
en               

-------
V-25
TABLE V-3.
CONVERSION FACTORS F9R FUELS, STEELMAKING OXYGEN,
AND ELECTRICITY AS APPLIED IN STEELMAKING
Item
Common Physical Unit
Energy Factor,
approximate gross or high
Btu released per unit used
~ uels
Coal
Coke
Lb (bituminous, 7 percent ash)
Lb (by-product, 7 percent ash)
Gal (residual)
14,030
13, 220
148,600
145,200
91,500
1, 000
Fuel oil
Tar and pitch
LP gas
Gal (crude) .
Natural gas
Coke -oven gas
Gal (commercial propane)

Ft3 (standardized)

Ft3 (standardized)

Ft3 (standardized)

Ft3 (based on refining of hot
metal to steel)
500
95(a)

435 (b)
Blast-furnace gas
Oxygen
""lectricity
kwhr
3,413
(a) Memorandum only - duplicates heat assigned to coke, fuel oil, and gas charged to blast furnace.
(b) ConjectUral; for use of 153.5 pounds of oxygen at O. 084 pound per cubic foot to refine 18 lb silicon, 84 lb carbon, and
14 lb manganese from hot metal, incidentally consuming 60 lb iron. Thermodynamic data taken from Tables 8D-1
and 8D-2 of Elliott, Gleiser, and Ramakrishna, Thermochemistry for Steelmaking, Vol. II, Reading, Mass., 1963,
Addison-Wesley. (This information is invalid for oxygen used in the electric furnace or for enriching combustion in
end wall burners of open-hearth furnaces.)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-26
conjectural, approximate, and invalid for some instances, but this does not bias the
energy analysis significantly, because the greatest part of steelmaking oxygen reacts
as illustrated in the explanatory footnote.
Table V -4 summarizes the 10 years of fuel, oxygen, and electricity data in energy
equivalents per ton of iron and steel produced. This final adjustment of the industry-
wide consumptions of Table V -1 shows the relative amount of energy supplied per ton
of iron and steel produced and allows direct comparisons among the fuels. It is' note-
worthy that roughly half of all steelworks' energy is supplied as coke and that coke -oven
gas and natural gas dominate the remainder. Electricity and oxygen, despite their
crucial importance to modern steelmaking, are seen to be of minor significance as
contributors to total energy - oxygen mainly has contributed savings in time, not in
energy, to the business of converting iron to steel.
Figures V-I and V-2 portray the 10-year energy patterns graphically. In
Figure V-I, the decline inuse of solid and liquid fuels per ton of steel is shown clearly,
as is the irregular pattern in use of gases. Despite the steady growth pattern for the
nonfuels (oxygen and electricity), energetically these sources have not even offset the
decline in use of oil, tar, and pitch as fuels. Also, in Figure V-I, the sharp breaks
in use of solid fuels from 1958 to 1959 and from 1960 to 1961 are shown to be irregu-
larities in an otherwise regular downtrending pattern.
Figure V -2 divides the graphic portrayal of energy use to distinguish among non-
fuels, fossil fuels used directly in their normal form, and coal-conversion fuels (coke,
tar, pitch, and coke -oven gas) made from coal at the coke ovens. This shows that
ordinary fossil fuels were replacing coal-conversion fuels from 1958 to 1962. There-
after, both declined for 2 years. Since 1964, use of coal-conversion fuels per unit of
iron and steel has been steady or increasing, and direct use of fossil fuels has continued
to drop. The 1966-67 average consumption of coal, oil, LP gas, and natural gas was
almost 6 percent below the 1964-65 average.
It is appropriate here to restate the principal trends, fuel by fuel, as a prelude
to the next discussion, which analyzes fuel applications according to categories of use:
(l) The use of coal (as coal) has been steady in total amount used by the
industry but has declined more than 34 percent per ton of iron and
stee.1 produced.
(2) The industry's consumption of coke has risen with steel production
but has declined about 19 percent per ton of iron and steel produced.
(3) The total consumption of fuel oil was steady, except in the past 2
years when it appeared to decline. Per ton of steel produced, con-
sumption of fuel oil has declined steadily to a level 44 percent below
that of 1958.
(4) The use of tar and pitch has been stable in total amount but has
dropped by 33 percent in terms of use per ton of iron and steel
product.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
 TABLE V -4. D:S' ~ UBUTIO \I OF A1'vOUN' ~S OF E' ~RGY USED PE ~ '~ON 0 ~ RON AN) STE ~L PRODUCED IN 
  UNITED STATES STEEL PLANTS, 1958-1967      
     Source: Derived from Tables V-I, V-2, and V-3.    
     Energy Consumption, millions of Btu (high or gross ), Per Ton of Iron. and Steel  
 Item   1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 
 Fuels              
m               
~ Coal(a)   2.245 1.936 1.992 1.992 1.936 1.712 1.515 1.529 1.501 1.473 
-i   
-i               
IT! Coke   14. 158 13 . 392 13.432 12.255 12.162 11.581 11.634 11.620 11. 488 11.449 
r   
r               
IT! Fuel oil   2.673 2.608 2.444 2.250 2.076 2.082 1.907 1.768 1.565 1.430 
~               
IT! Tar and pitch 0.502 0.473 0.513 0.391 0.407 0.378 0.328 0.359 0.333 0.337 
~ 
0               
;Q LP gas   0.008 0.008 0.007 0.006 0.007 0.009 0.007 0.012 0.009 0.010 
~               
r Natural gas 3.123 3.207 3.495 3.916 4.274 4.127 3.923 4.041 3.751 4.096 
-               
z Coke-oven gas 3.778 3.794 4.357 3.474 3.398 2.713 2.738 2.911 3.319 3.398 
I/) 
-i 
-i               
C Summary, Fuels:            <:
-i              r
IT!               N
I Solid fuels 16.403 15.328 15.424 14.247 14.098 13.293 13. 149 13.149 12.989 12.922 -J
o Liquids(b) 3.183 3.089 2.964 2.647 2.490 2.469 2.242 2.139 1.907 1.777 
0 
r Gases   6.901 7.001 7.852 7.390 7.672 6.840 6.661 6.952 7.070 7.494 
c    
~ Total   26.487 25.418 26.240 24.284 24.260 22.602 22.052 22.240 21.966 22.193 
m   
c               
I/)               
r Oxygen(c)   0.064 0.080 O. 123 O. 188 0.234 0.270 0.301 0.335 0.406 0.468 
~              
m               
0 Electricity   0.608 0.587 0.621 0.652 0.713 0.,717 0.679 0.706 0.741 0.799 
;Q   
~               
-i               
0 Summary, Energy:            
;Q            
IT! Fuels   26.487 25.418 26.240 24.284 24.260 22.602 22.052 22.240 21.966 22.193 
I/)   
 Nonfuels   0.672 0.667 0.744 0.840 0.947 0.987 0.980 1.041 1.147 1.267 
 Total   27.159 26.085 26.984 25.124 25.207 23.589 23.032 23.281 23. 113 23.460 
 (a) Excludes coking coal.            
 (b) Includes LP gas.            
 (c) Conjectural, but overall effect on energy balance is not greatly distorted. See Note (b), Table V-3.     

-------
 30 
 28 
 26 
\D 24 
0 
)(  
2 22 
CD  
-c  
<1> 20 
() 
:::J  
-c  
0  
....  
a.. 18 
<1>  
<1>  
+-  
(/) 16 
-c  
c::  
0  
c:: 14 
0  
....  
......  
- 12 
0 
c::  
~ 10 
....  
<1>  
a.  
 8 
>.  
01  
....  
<1>  
c:: 6 
w 
 4 
 2 
 0 
 1958 1960
- V-28
1962
1964
1966
1968
A-57769
FIGURE V-I. TOTAL ENERGY CONSUMED PER TON OF IRON AND STEEL,
ACCORDING TO PHYSICAL FORM OF THE FUELS USED
Year
BATTEL.L.E. MEMORIAL. INSTITUTE - COL.UMBUS L.ABORATORIES

-------
 30
 28
 26
CD 24
a
)( 
::J 22
+-
m
"0 
Q) 20
u
::J 
"0 
e 
a.. 
 18
Q) 
Q) 
+- 
(/) 
 16
"0 
c 
0 
c 14
o 
~ 
.- 
- 12
o
c 
~ 
 10
~ 
Q) 
a. 
>. 8
01 
~ 
Q) 
c 
w 6
 4
 2
 o
 1958
V-29
1960
1962
1964
1966
1968
A-57770
FIGURE v-2. TOTAL ENERGY CONSUMED PER TON OF IRON AND STEEL,
ACCORDING TO ULTIMATE SOURCE OF THE ENERGY
Year
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-30
(5) LP gas contributes only trivially to steelmaking-energy consump-
tion' There has been a rise in total consumption but no clear trend
in consumption per ton of steel.
(6) The total use of natural gas has risen sharply, but irregularly, by
about 91 percent over 1958 levels. Per ton of iron and steel,
almost all of the 31 percent rise occurred between 1958 and 1961,
(7) The total use of coke-oven gas has been irregular, with sharp peaks
in 1960 and 1966. Consumption per ton of iron and steel for the
years 1965 -67 averaged 15 percent below unit consumption per ton
in 1958 - 59.
(8) Incomplete data indicate that about 13 percent of the energy
originally paid for as coke is ultimately used in the form of blast-
furnace gas (a by-product of ironmaking),
(9) The total use of steelmaking oxygen has increased almost eleven-
fold in 10 years, The use of oxygen per ton of iron and steel has
increased more than seven-fold in the same period.
(10) The amount of electricity purchased from outside suppliers almost
doubled in the 10 years from 1958 to 1967. Consumption per ton
of iron and steel rose by about 56 kilowatt-hours, or 31 percent,
(11) Solid and liquid fuels lost ground during the past decade to gases
and nonfuels,
(12) Although both direct-used fossil fuels and fuels converted from
coking coal lost ground in the 10-year period, the conversion
fuels have fared better since 1963,
Analysis by Applications
There are seven principal uses of fuels, steelmaking oxygen, and electrical
energy within the iron and steel industry:
(l) Firing of coke ovens for conversion of coal to coke and by-products
(2) Sintering of dusts and fine ores to make blast-furnace feed
(3) Smelting of iron ores and agglomerates in the blast furnace
(4) Steelmaking, including remelting of scrap and conversion of hot metal
to steel
(5) Steam raising for compressing blast and for generating electricity
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-31
(6) Reheating, annealing, and heat treating of steel during processing
(7) Driving of mills, forges, and process lines to shape, coat, and
finish steel products.
Minor uses (as for comfort heating, curing of refractories, welding and cutting, and
operation of peripheral equipment) are numerous.
Energy sources compete with each other technically and economically in all but
the last of the listed uses. Electricity long has had a technical near -monopoly upon the
driving of mill mechanical equipment used in the rolling of steels. When steam is still
used, as in some forges, it is commonly raised in the same boilers that power the
plant's electrical generators. Accordingly, Use (7) is not discussed as an application
of energy to major processes.
.t' uing of Coke Ovens
The slot-like ovens that convert coal to coke are arranged in long batteries, like
books on a shelf. Between each pair of ovens there are flues heated by hot combustion
gases. The system of flues, ovens, and burners rests upon regenerative checkerwork
preheaters for the combustion air, and firing is reversed on a preset schedule so that
waste heat from the flues is recuperated.
The principal criteria for fuels used in underfiring coke ovens are cleanliness,
uniformity of heat release, and proper (nonexplosive) behavior during reversal of the
regenerative cycle. Risks of explosive conditions, deposits in the flue system, and
uneven heating must be avoided, because a coking battery is expected to operate con-
tinuously for 30 years or more without major repairs.
Coke -oven gas and blast-furnace gas furnish most of the energy requirements for
underfiring American coke ovens. Natural gas is used only to the extent that it is mixed
with by-product gases to enrich them. Although coke -oven gas commonly is used to
enrich blast-furnace gas in steel plants, special mixing of these gases for coke -oven
underfiring is not in ordinary practice. AISI statistics indicate that some fuel oil was
used for underfiring during 1966, but the data (which are not explicit) show that the
amount used was very small.
Table V -5 presents the estimated breakdown of fuels used in underfiring coke
ovens averaged for the years 1966 and 1967. The estimated use of natural gas is ob-
tained differentially from the estimate that, typically, 1000 Btu are required per pound
of coal coked. (It was inferred from oil-consumption figures for 1965 and 1967 that the
use of oil for underfiring in 1966 involved about 8 million gallons. )
The main determinant of fuel choice for underfiring coke ovens is the plant gas
balance. Use of coke-oven gas and blast-furnace gas made in the plant is preferred
economically to any purchase of fuels from outside. Blast-furnace gas is applicable
only to controlled-combustion systems with large ducts; thus, it is used, first, in the
blast-furnace stoves, second, in steam raising for blast compression, and, third, to
underfire coke ovens or to heat process furnaces (mainly soaking pits for reheating steel
ingots), If there is a net shortage of energy in the gas for these uses, it may be en-
riched or partly diverted away from steam raising to bring about balance. If the
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-32
TABLE V -5.
ESTIMA TED USE OF VARIOUS FUELS FOR UNDERFIRING COKE OVENS
0966-1967 AVERAGES)
Average coke produced, net tons /year
Average coal consumed in coking, net tons /year
Average coal per ton of coke, tons
Estimated underfiring heat, per ton of coal, Btu
58,180,591
84, 108, 905
1,446
per ton of coke, Btu
Average annual underfiring heat, million Btu
2,000,000
2, 892,000
Percentage of total industry energy used
168,000,000
5.39
Average Annual Coke-Oven Gas Used In Underfiring:
Millions of cubic feet
Cubic feet per ton of coke
Millions of Btu
231,330
3,976
116,000,000
68.74
Percent of total underfiring energy
Average Annual Blast-Furnace Gas Used in Underfiring:
Millions of cubic feet
346, 178
5,950
33,000,000
Cubic feet per ton of coke
Millions of Btu
Percent of total under firing energy
Estimated Average Annual Fuel Oil Used In Underfiring: (a)
Thousands of gallons
19,54
Gallons pe r ton of coke
-3, 900
....0,07
Millions of Btu
-600,000
-0.35
Percent of total underfiring energy
Estimated Average Annual Natural Gas Used In Underfiring:(b)
Millions of cubic feet
Cubic feet per ton of coke
Millions of Btu
-19,000
-330
Percent 0;£ total underfiring energy
-19,000,000
-11.00
(a) Estimated by comparing 1966 rise in "other uses, blast-furnace area" over average for 1965 and 1967.
(b) By difference - contains cumulative error of all estimates.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-33
shortage is great, blast-furnace gas is not used to underfire the coke ovens, in which
case any resulting surplus may be assigned to steam raising or may even be flared.
A second determinant is physical plant layout. Blast-furnace gas is under low
pressure after cleaning, and its low energy content does not justify use of booster
pumps to raise the transmission pressure. Accordingly, coke plants that are any sub-
stantial dis tance from the blast furnaces are underfired with their own coke -oven gas.
.tnis practice consumes roughly half of the coke-oven gas produced; the balance is
compressed and sent to the plant for other uses.
Merchant-iron plants that have coke ovens usually have a great surplus of both
coke -oven gas and blast-furnace gas owing to the absence of heating furnaces and
soaking pits. For these plants, it is often advantageous to underfire the ovens with
blast-furnace gas and to sell most of the coke -oven gas produced to an electric utility,
another steelworks, or to some other single user.
Sintering of Dusts and Ores
Dusts collected during the cleaning of blast-furnace gas, along with mill scale
and fine ore, are commonly mixed with coke breeze and fluxes, then sintered in an
8 to 12 -inch layer on a traveling grate. Coal may also be used as a part of the mix.
The sintering mix is ignited as the grate passes under an intense gas flame. The
sintering reactions are sustained during and after ignition by heat from the combustion
of coke. This heat is forced by suction of air down through the ignited mix.
The American Iron and Steel Institute's statistics on raw materials show that an
average of about 53 million tons of sinter were charged into blast furnaces annually
between 1963 and 1967. This consumption was fairly stable, as shown in Figure V -3.
Over the 10 years from 1958 to 1967, average sinter consumption per ton of hot metal
increased from 989 to 1178 pounds, or 19 percent. All of the increase occurred prior
to 1965, with total annual consumption of sinter nearly doubling from 1958 through 1964.
The average use of coke breeze (coke too small for most other uses) in manufac-
ture of agglomerates (including sinter, pellets, and briquets) was 2.48 million tons
annually from 1963 to 1967, or 4.69 percent by weight of sinter consumed. It may be
safely estimated that little or no coke breeze was used in making pellets or other
agglomerates.
When supplies of coke breeze are insufficient, anthracite may be used in making
sinter. The 1963-67 average use of coal (for all agglomerates) was 0.89 million tons
per year, or 1.68 percent of sinter consumed. As with coke breeze, it is estimated that
nearly all coal used in agglomeration is mixed with sinter feed. . The total of coal and
coke is about 6.37 percent of sinter consumed, or 1.8 million Btu per ton of sinter.
This is a maximum, allowing for no use of coke b~eeze or coal for other agglomeration.
At present, the determinant of solid fuel for the sintering mix is the availability
of coke breeze. When sintering was first used, it served mainly as a means for re-
cycling dust and other wastes. As the excellent smelting properties of sinter became
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
~  
c  
(1)  
>.  
~ 60 
(1) 
a.  
VI  
c  
0  
+-  
+- .
(1)  
c  
 50 
-  
0  
VI .
C  
0  
E 40 
.:  
(1)  
+-  
C  
Cf)  
(1)  
<.) 30 
c 
c  
~  
l.C  
I  
+-  
VI  
C  
CD  
- 20 
0  
c  
0  
+-  
a.  
E  
:::I  
VI 10 
c 
0  
()  
C  
+-  
~  
 0 
  1968
  A-57773
70
V-34
FIGURE V-3.
TREND IN USE OF SINTER IN
BLAST FURNACES IN THE
UNITED STATES
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-35
known, sinter plants became larger and more numerous. Whereas once there was
plenty of coke breeze for sintering and for boiler use, more and more breeze was
diverted to the sinter plants, until a breeze shortage occurred. Just as coal replaced
coke breeze in the boilers, it has been used to supplement breeze in sintering. Coal
is inferior to coke breeze for this purpose, but anthracite works passably.
Unfortunately, the AISI offers no statistics on the use of ignition fuels in sintering.
Coke -oven gas usually is preferred, because it is cheaper than natural gas, develops
a similar flame temperature, and is readily used in downward-pointing burners.
Blast-furnace gas does not develop adequate temperatures, and fuel oil would tend to
coke the fuel nozzles as it is burned in the radiant ignition hood.
With the growing shortage of coke breeze, some plants make extra coke to be
ground into sinter fuel. An alternative is to extend the length of the ignition hood and
to use extra ignition energy to offset a part of the solid-fuel requirements.
Smelting of Iron Ores and Agglomerates
Because the principal fuel used in smelting of iron ore is coke, an important con-
cern of ironmakers is to minimize the amount of coke used. If fixed charges and over-
heads are included, the cost of energy from coke is over 60 cents per million Btu, well
above the cost of energy from other fuels. For example, interruptible':' natural gas in
the Chicago area sells for less than 35 cents per million Btu in large amounts.
The decade from 1958 to 1967 was one of great progress in ironmaking, including
a 21 percent decline in pounds of coke (or energy from coke) applied to smelting. Fig-
ure V -4 illustrates this change. Some coke has been eliminated by use of hotter blast,
some by better sizing and beneficiation of ores and other burden materials, and some
by injection of gas, oil, or tar into the blast tuyeres. In 1967, auxiliary fuels supplied
about 4. 1 percent of all smelting energy.
The use of coke as the primary smelting fuel is dictated by chemical and physical
cons iderations. Chemically, coke has proper reactivity: it burns efficiently at the
tuyeres, yet it reacts but slowly with carbon dioxide and other oxygen-bearing gases in
the blast-furnace stack. Physically, coke retains its strength at high temperatures and
so helps to maintain good overall permeability of the burden in the shaft of the blast
furnace. Without coke, iron ores would tend to pack tightly and to block the flow of
reducing gas. If coal were used, it would decrepitate (by giving off mixed hydrocarbons
and by disintegrating physically) long before it reached the zone of combustion.
The injection of fluid fuels (or powdered coal) to replace some blast-furnace coke
depends upon technical conditions. The hydrocarbons in oil, coal, and natural gas must
be dissociated endothermically in the tuyere flame, and even coke-oven gas contains
appreciable methane, the most endothermic of hydrocarbons. These endothermic ten-
dencies and the lack of sensible heat in the injectants depress the flame temperature at
the tuyeres. The depression in flame temperature lowers the availability of high-
temperature heat, which is vitally needed to offset heat losses from the hearth, to fuse
metal and slag and to perform critical chemical tasks such as the reduction of silica to
silicon. The amount of injected fuel, therefore, must not be so great as to interfere
"Gas contract provides that the gas company can interrupt service during periods of peak demand. ActUal interruption is
usuall y confined to evenings in January and February.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-36
26
c:
o
~
24
<1>
g 22
c:
~
::J
......
I
+- 20
VI
o
.Q
...... 18     
0      
c:      
0      
+- 16     
~      
<1>      
a.      
::J 14     
+-      
en      
c:      
0 12     
E      
 10     
"'0      
<1>      
VI      
:J      
 8     
VI      
Q)      
~      
 6     
......      
0      
+-      
c:      
<1> 4     
+-     
c:      
0      
U      
+- 2     
0      
<1>      
I      
 0     
 1958 1960 1962 1964 1966 1968
 Year   A-57772
FIGURE V-4.
SOURCES OF SMELTING ENERGY
(a) Includes enrichment of blast-

furnace gas for stoves.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-37
with the heat balance in the hearth. Commonly, the installation of fuel-injection
apparatus is accompanied by efforts to decrease the blast moisture and/ or increase the
blast temperature. To date, the maximum efficient use for fuels injected at the tuyeres
per ton of pig iron produced has been:
(I) About 12 gallons (100 pounds) of oil or tar, or
(2) About 2000 cubic feet of natural gas or coke -oven gas, or
(3) About 120 pounds of bituminous coal.
The definition of "efficient use" depends upon local fuel economics.
There is a significant trend toward using tar as an injected fuel. This trend is
promoted by the decline of its use in open-hearth steel furnaces. Tar need not be sold
as fuel; it also is useful for highway construction and maintenance. But the low price
of asphalt in recent years has tended to keep coal tar in the steel plants as a fuel, and
the blast-furnace is one logical outlet. Almost 7 percent of all fuel tar was injected
into blast furnaces during 1967. In prior years, no separate listing was made.
Steelmaking
Steelmaking includes the remelting of scrap steels, the oxidation of carbon and
other unwanted elements from blast-furnace iron, and the careful adjustment of final
chemistry and temperature to specification. The energy that is supplied to a steel-
making furnace must offset heat losses, fuse the solids that are charged, and raise the
temperature of all materials to their hottest level (usually above 2800 F).
Three steelmaking processes compete vigorously - basic -oxygen, open-hearth,
and electric -arc furnace. They are strikingly different from each other, in that each
uses a different source of energy, a different mix of charges, and a different physical
configuration. Table V -6 outlines some of these major differences, indicates the
bases for competition, and shows the proportion of steel made in each type in 1958 and
In 1967.
AISI statistical summaries distinguish among the steelmaking processes in their
use of oxygen. Although there is a separate listing of fuels applied to steelmaking
furnaces taken as a group, no breakdown by process type is given. The main application
of fuels in steelmaking is to the endwall burners of the presently dominant open-hearth
process. In spite of the use of gases (and perhaps oil) for preheating ladles and new
linings for furnaces, and in spite of various ancillary uses for fuels in any melt shop,
the analysis must assume that fuels assigned to steel-melting furnaces by the AISI were
used in the open hearth.
Similarly, no breakdown of applications for electrical energy is given in the
statistical summaries. The requirements for energy in electric -arc melting are fairly
well known, however. A good working average is 600 kwhr per ton of carbon steels
and 800 kwhr per ton of alloy and stainless grades. Subject to these necessary assump-
tions, the energy sources for steelmaking are shown for the 6 years from 1962 to 1967
in Table V -7. Prior to 1962, applications of steelmaking oxygen were not separated by
process.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
TABLE v -6.
THE THREE MAJOR STEELMAKING PROCESSES: NATURE, A TTRIB UTES, AND
BASES FOR COMPETITION
Basic -Oxygen Steelmaking
Open -Hearth Steelmaking
Electric -Arc Steelmaking
1958:
1967:
1,323,000 net tons
41,434,000 net tons
1958:
1967:
75,880,000 net tons
70, 690,000 net tons
1958:
1967:
6,656,000 net tons
15,089,000 net tons
m
~ Furnace shaped somewhat like a
-i pear or a teacup, with upper
~ walls slanted conically inward.
~ Discharges gases to a hood.
~ Fired by a vertical water -cooled
~ oxygen lance, inserted verti-
g cally to within a foot or so of the:
» bath.
r
z Energy obtained primarily from
~ sensible heat in hot metal,
-i secondarily from reactions
c
-i between oxygen and unwanted
111
I elements dissolved in the hot
o metal (Si, Mn, P, C).
o
r
c Charges 22 to 31 percent scrap
~
m steel, 69 to 78 percent molten
~ hot metal from blast furnace.
Furnace shaped like a wide, flat
rectangular bathtub to maximize
slag-metal interface and heat-
receiving surface. Fired by
burners at each end, used
alternately. Combustion air
preheated by regenerative
checkerwork in flue systems.
Furnace shaped like a circular
bowl of variable depth. Roof
may lift and swing to admit
charges. Fired by 3 -phase
moderate -voltage power applied
to equispaced carbon electrodes.
Electrode heights adjust auto-
matically to maintain arcs.
Energy obtained primarily from
endwall burners, supplemented
by sensible heat in hot metal
and by oxidation of impuritie s
from hot metal. Oxygen lances
now in general use for refining.
Energy obtained almost entirely
from electrical discharge in
the arcs. No fuels used in
operation. Use of blown oxygen
confined to relatively minor
adjustments in carbon analysis.
<:
r
W
ex>
Charges any combination of hot
metal and scrap, most plants
close to a 50/50 mixture.
Charges mainly steel scrap, but
hot metal has been used to a
limited degree.
r
» Best mix of speed and versatility
~ for new or modernized plants,
~ at low capital cost, provided hot
6 metal is available in large
~ quantities. To reach 99 million
111
(/I tons in 1980.
Versatile but slow. Used heavily,
and competitive solely on the
basis of sunk capital. No new
installations anywhere since
1959. Oxygen roof lance s are
a competitive modification.
Extremely versatile and faster
than open-hearth. Suited to
small plants with no hot -metal
supply. Strong in stainless
and alloy steels.
Oxygen per
1958
1962
1967
ton crude steel:
1925 cu ft
1 917 cu ft
1900 cu ft
Oxygen per
1958
1962
1967
ton crude steel:
120 cu ft (estimated)
508 cu ft
826 cu"ft
Oxygen per
1958
1962
1967
ton crude steel: (a)
175 cu ft (estimated)
188 cu ft
222 cu ft
(a) Includes oxygen used in the Bessemer process, .... lich was fading fast in 1962 and all but gone in 1967.

-------
TABLE V -7.
ESTIMATION OF SOURCES OF ENERGY FOR STEELMAKING PROCESSES (OTHER THAN
SENSIBLE HEAT IN CHARGES)
  Note: This tabulation necessarily includes estimation as explained in the text and in footnotes. Statis tic s  
m   given herein should not be applied indiscriminately to uses other than for those intended here.  
»    
-f    
-f            
111    Energy   Consumption Per Ton of Steel    
r         
r Process and Energy Source Unit 1962 1963 1964 1965 1966 1967 
111 
3: Open-hearth steelmaking(a)         
111 Fuel oil Gal 10.66 10.37 10.28 9.44 8.49 8.03 
3: 
0 Tar and pitch Gal 3.19 3.09 2.86 3.38 3.52 3.51 
;u 
» Natural gas Ft3 1355 1073 1106 1121 1129  1216 
r Coke -oven gas Ft3 817 845 778 763 647  463 
z Steelmaking oxygen Ft3 508 578 592 625 737  826 
en  
-f  as energy(b) 10 6 Btu        
-f Total, 4.03 3.74 3.70 3.67 3.55 3.51 <:
c            r
-f            (,J.)
111 Basic-oxygen steelmaking         ...0
I         
0 Steelmaking oxygen Ft3 1918 1909 1944 1864 1853  1900 
0 Total, as energy(b) 106Btu        
r 0.83 0.83 0.85 0.81 0.81 0.83 
c 
3:            
m Electric -arc steelmaking( c)         
c         
en         
r Electricity (total) kwhr 643 678 677 678 677  666 
» Steelmaking oxygen(d) Ft3 188 178 190 199 208  222 
m  
0 Total, as energy(b) 10 6 B tu        
;u 2.28 2.39 2.39 2.40 2.40 2.37 
»  
-f            
0            
;u            
iii (a)           
en (b)           
Assumed to use all fuels ascribed to steel-melting furnaces in AISI statistical reports.
Oxygen valued at 435 BtU/cubic foot energy equivalent. This is correct for roof lances of open -hearth and for the basic-oxygen process, but only approximate for use of
oxygen in open -hearth endwall burners or in refining of electric -furnace heats ma de from melting of scrap steel.
(c) No figures on energy consumption are available. Estimate based on assumed 600 kwhr per ton of carbon steel and 800 kwhr per ton of stainless or alloy steel.
(d) Statistic includes the (fading) Bessemer process, both in oxygen used and in steel produced.

-------
V-40
The statistics of fuel consumption for open-hearth steelmaking (as approximated
in Table V -7) illustrate a decline in energy requirements per ton of open-hearth steel.
A more exact assignment of fuels among the three processes would tend to increase
this trend, inasmuch as use of both basic -oxygen and electric -arc steelmaking is
growing. For the open-hearth, the lower consumption of energy per ton of steel re-
flects replacement of the more obsolete shops (by basic -oxygen furnaces), offsetting
of fuel requirements by time -saving injection of oxygen, and use of oxygen in endwall
burners to augment flame temperature and save additional time.
The use of fuel oil and of coke-oven gas in open-hearth furnaces has declined
significantly, but the use of natural gas and tar has been steady. The reasons for these
trends are not obvious and deserve more investigation. Possible causes include
(l) Growing difficulties in the sale of tar
(2) Hazards in the use of fuel oil and tar with oxygen-augmented endwall
burners
(3) Problems with the high flame velocity of coke-oven gas (a long and
luminous flame is desired)
(4) Natural displacement of oil- and coke -oven gas -fired open hearths
by basic-oxygen steelmaking
(5) Concern over sulfur contamination from oil, tar, and coke-oven gas.
Proponents of basic -oxygen steelmaking seem certain that the open -hearth must
go the way of the Bessemer converter, and proponents of electric steelmaking are
equally sure that huge electric furnaces will take a constantly increasing share of total
steelmaking production. These claims may be overstated, at least for the next decade.
The existence of large investments in the larger, newer open hearths will slow the
demise of the open-hearth process as an important producer of steel.
Steam Raising
An integrated steelworks usually has substantial boiler capacity, because steam
power is very useful and because it represents convenient recovery of the energy in
certain plant by-products, especially waste heat and blast-furnace gas. The uses of
steam in the steel plant include
(l) Compression of air for blast furnaces
(2) Generation of electric energy
(3) Distillation of coking by-products
(4) Pumping of coke-oven gas
(5) Powering of forges and presses
(6) Warming of residual oil and tar lines
(7) Comfort heating and housekeeping.
(via turbine)
Most integrated plants use steam as the common denominator of power. The
boiler station often is equipped to burn almost any available fuel. When the plant is in
full operation, a minimum number of turbogenerators will be operating at full electrical
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-4l
output, and purchased fuel (usually coal) will be brought in to complete the balance. In
some instances, the generating plant is competitive in cost with outside power sources
and will be operated at rated capacity most of the time. This is especially true when
the power station supplements inadequate outside sources of electricity.
The management of steaming and generating capacity depends on company philos-
0phy. If the cost of the boilers is charged mainly to the blast-furnace blowing engines,
there will be a tendency to generate considerable power internally. Overall, the steel
industry buys about 70 percent of the electricity it uses. If the in-plant generators are
envisioned as noncompetitive, their use will be minimized. At any rate, the first
priority rests with recovering the fuel value of blast-furnace gas and the second, with
balancing out coke -oven gas and coke breeze on a plant basis and utilizing the heat
recovered in waste -heat boilers. .
The principal fuel used in the boiler plant (after blast-furnace gas) is coal. Use
of other fuels (at least in the East) is incidental. In the 10 -year period from 1958 to
1967, steel plants used from 6.1 to 6.7 million tons of coal per year (the average was
6.44 million tons), and there was no consistent trend from year to year. Roughly
90 percent of all noncoking coal used by steelmakers was applied to steam raising.
Heating of Steel in Process
When new molten steel first solidifies into an ingot, it has a temperature some-
what above 2800 F. From this time until it has been almost completely formed to
saleable shape, the steel must be maintained at (or reheated to) red heat. Maintenance
of temperature for steel in process is a major application of energy.
The kinds of furnaces used for in-process heating include soaking pits (in which
newly stripped ingots are allowed to equalize in temperature for a few hours before
rolling begins), Soaking pits (which may be fired with either blast-furnace gas, coke-
oven gas, or a mixture) are intended to make the temperature of the steel uniform, not
to add large amounts of heat to the steel.
An ingot is conventionally removed from its soaking pit and rolled to a bloom or
slab, then allowed to cool for inspection. The cool steel is carefully examined, and
defects are removed with an oxygen torch in a process called scarfing. The prepared
steel is then reheated in a once -through furnace, with careful attention to both accuracy
and uniformity of the final temperature. The reheated slab or bloom is then finish
rolled to plate, strip, merchant shapes, or wire rod. Thereafter, any reheating is
usually intended to produce some metallurgical change (such as an improvement in
strength or ductility), Furnaces used to alter steel metallurgically (called heat-
treating furnaces) often are operated with special atmospheres to promote a clean
appearance on the product. Table V-8 presents a 10-year analysis of fuels used in
heating and annealing furnaces. The tabulated data illustrate no particular trend in the
use of fuel energy per ton of steel in heating and annealing furnaces. However, prior to
.1962, both oil and tar declined in importance, as did coke -oven gas. The use of natural
gas rose during the same period, indicating a general preference for lower sulfur fuels.
The rise in use of LP gas may be a correspondent change for areas in which natural gas
is not available or in temporary relighting of furnaces built for fluid fuels.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
 TABLE V -8. USE OF FUELS TO FIRE HEATING AND ANNEALING FURNACES IN THE UNITED STATES STEEL INDUSTRY, 1958-1967  
     Source: Derived from AISI statistics.        
m               
~               
-i      Consumption Per Ton of Finished Steel Products      
-i           
IT! Fuel Unit 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 
r               
r Fuel oil 103 gal 372, 652 410,201 367, 053 333,956 335,283 391, 046 392,300 448,649    
IT! 461,939 403,426 
3: per ton product gal 6.32 6.17 5.37 5.25 4.94 5.39 4.80 5.09 5.37 5.02 
IT! energy per ton 106 BtU 0.94 0.92 0.80 0.78 0.73 0.80 0.71 0.76 0.80 0.75 
3:       
0               
;0 Tar and pitch 103 gal 5,296 4,519 5, 163 4,282 4,146 - - - - - - - - - - - Not separately listed - - - - - - - - - - - 
~ per ton product gal 0.09 0.07 0.08 0.07 0.06        
r        
 energy per ton 106 Btu 0.01 0.01 0.01 0.01 0.01 - - - - - - - - - - Assumed 0.01 million Btu - - - - - - - - - 
Z               
In               
-i LP gas 103 gal 4,861 5,176 6,278 4,941 6,395 9,317 7,828 13,952 9,901 11,112 
-i per ton product gal 0.08 0.08 0.09 0.08 0.09 0.13 0.10 0.16 0.11 0.14 
c: -<
-i energy per ton 106 Btu 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 ,
IT!              ~
I               N
o Natural gas 106 ft3 149,388 177,907 199,352 206,732 222, 739 236,515 256,256 277,317 264,228 283,004 
0 per ton product ft3 2,532 2,675 2,914 3,250 3,280 3,258 3, 133 3,148 3,069 3,524 
r  106 Btu 2.53 2.68           
c: energy per ton 2.91 3.25 3.28 3.26 3.13 3.15 3.07 3.52 
3:              
m               
c: Coke-oven gas 106 ft3 318,835 353, 063 477,481 324,736 280,789 255,673 339,567 354,500 365,327 361,917 
In 
 per ton product ft3 5,404 5,309 6,981 5,106 4, 135 3,522 4,151 4,024 4,243 4,507 
r  106 BtU 2.70            
~ energy per ton 2.65 3.49 2.55 2.07 1. 76 2.08 2.01 2.12 2.25 
m               
0               
;0 Blast-furnace gas 106 ft3 - - - - - - - - - - - - - - - - - - - - - - - Data not given - - - - - - - - - - - - - - - - - - - - - - - - 156,010 162,898 
~ 
-i per ton product ft3          1,812 2,028 
0 energy per ton 106 BtU - - - - - - - - - - - - - - - - - - - - - Assumed 0.17 million Btu - - - - - - - - - - - - - - - - - - - - 0.17 0.19 
;0 
IT!               
In               
 Total Energy 106 Btu 6.36 6.44 7.39 6.77 6.27 6.01 6.11 6.11 6.06 6.73 

-------
V-43
Principal Findings
The following statements summarize the existing applications of fuels to steel-
making activities and refer to determinants that are expected to dominate future choices:
Coking - Blast-furnace gas and coke-oven gas are now the preferred
fuels for underfiring coke ovens; this preference will not change. Between
them, the main factors of choice include physical layout and the overall
plant gas balance. All factors tending to improve the value of coke -oven
gas as a plant fuel similarly tend to increase the likelihood of underfiring
with blast-furnace gas. Natural gas participates in underfiring only
insofar as it is used to enrich the other gases,
Sintering - Two kinds of fuels are required. Among solid fuels, coke
breeze is preferred, and the sinter plant preempts breeze requirements
in other parts of the steelworks. Coal, especially anthracite, is usable
as a poor substitute for coke breeze. Longer ignition cycles can lessen
solid-fuel requirements for sintering. Most ignition is done with coke-
oven gas or mixed coke -oven and blast-furnace gas. These fuels usually
are preferred over natural gas for economic reasons. Fuel oil is not
used, because the configuration of the ignition system is technically
unfavorable.
Smelting - Full development of hot-blast operation with prepared burdens
and auxiliary injected fuels could cut coke rates to well under 1000 pounds
per ton of hot metal (from present levels averaging over 1250 pounds per
ton), However, the suggested improvements usually must wait upon a
strain on ironmaking capacity - at present, America has blast furnaces
to spare. The vast improvements in ironmaking fuel economy which began
in 1958 have nearly stopped under present conditions of steel demand.
Among the auxiliary fuels, regional economics will probably govern, Coal
injection, now in development, should dominate the inland East, oil prob-
ably will be preferred near the coasts, and natural gas probably will be the
preferred injectant 'west of Cleveland.
Steelmaking - Two of the three steelmaking processes (the basic-oxygen
process and electric steelmaking) do not use fuels in the usual sense, and
there is no reason to suggest that they may become fuel consumers in the
future. The third (open-hearth steelmaking) is much maligned for its
obsolescence but will be with us for at least two decades. There is a
clear trend away from both oil and coke -oven gas as open-hearth fuels;
yet, tar, which is also sulfurous, continues as an important source of
melting energy. The changes are probably compound and include the con-
sideration of an oversold market for road and waterproofing tars. More
investigation of open-hearth fuel determinants is advisable. The increased
use of oxygen in open hearths has cut the overall energy requirement, but
at a high cost in dust-control problems. Much has been written about pre-
heating of scrap to increase its use in basic-oxygen steelmaking, but this
remains only marginally economic, except as an expedient in special
situations.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-44
Steam Raising - The management of steam raising is founded on the
principle that the boiler plant is to balance plant-energy sources and
needs for optimal net-energy cost. Surpluses and shortages of gases
are balanced at the boilers on a day-to-day basis, and variable amounts
of electric power are generated. This compound economic system is a
study in itself and worthy of much more attention. At least in the East
and Midwest, purchased coal is the most probable fuel for adjustment of
boiler -energy requirements.
Steel Heating - Various furnaces (usually at least two) are used to equalize
and control steel temperature during processing. Others may be used near
the end of primary manufacture to soften, harden, or resurface the product.
These furnaces run mainly on coke-oven gas and natural gas, with natural
gas preferred when sulfur in the furnace atmosphere could cause metal-
lurgical trouble.
No aspect of the analysis by uses disclosed strong trends toward the "clean"
fuels. The primary economic determinant of fuel choice in any steelworks is and will
be to use the homemade fuels completely before buying energy in the open market.
However, the sulfur content of fuels will continue to receive examination, because
fuels are the major source of undesirable amounts of sulfur in iron and steel.
Some Facets of the Relationship Between Clean Air
and the Fuel and Energy Aspects of the Steel Industry
This present study is concerned, of course, with the broad relationship between
the integrated steel industry and national concern with problems of polluted air. One
portion of that relationship can be examined by considering the nature of some of the
avenues open to the steel industry to raise air quality by making adjustments in the
industry's usage of fuels and energy. It will be assumed for the balance of this dis-
cussion that increasingly stringent standards for air quality are established in steel-
producing areas. Another assumption is that decreased emission of sulfur -bearing
gases will be a specific objective and that meeting of this objective will require treat-
ment of combustion gases resulting from'the burning of coal, tar, and coke-oven gas.
At present, some steel plants remove hydrogen sulfide from coke-oven gas, but the
hydrogen sulfide that is recovered is burned in power boilers.
One general response by the steel industry (already foreseen in existing tE:~chnol-
ogy) would be the cleaning of coke-oven gas to remove hydrogen sulfide. It is anticipated
that the relative cost of natural gas will rise generally in the future and that this eco-
nomic readjustment of the relative costs of coke -oven gas and natural gas might justify
the cost of cleaning the coke-oven gas. However, present operating practices make the
recovery of sulfur dioxide, in most steel plants, an uneconomical process.
Another general response would be a sharp decline in use of fuel oil, even as an
injectant to blast furnaces. As a fuel in open-hearth furnaces and steel-heating fur-
naces, fuel oil would be seriously disadvantaged by sulfur-control regulations. It is to
be expected that a new wave of conversions from open-hearth steelmaking to the basic-
oxygen or electric-arc process would occur and that surviving open hearths would be
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-45

fueled with tar or coke -oven gas. The use of tar is being forced by a deteriorating mar-
ket for coal chemicals and by-products. (These conclusions assume that fuel oil cannot
be economically de sulfurized. )
It should be noted that an increase in the relative price of natural gas could make
fuel oil relatively more competitive as a blast-furnace injectant than it is now. Fuel
oil causes no problems in the blast furnace because its sulfur content goes mainly into
the slag (not directly into the air). However, tar and pitch would be displaced from
melting furnaces equally with oil, and then probably would preempt oil in blast-furnace
injection.
Another general response would be an alteration in the economic rules for opera-
tion of in-plant boilers. Blast-furnace gas is a clean fuel, and its use would continue.
But the use of coal as a balancing supplement would be much altered by the obligation to
process the stack gases from solid-fuel boilers. Indeed, barring a large -scale short-
age of electric energy, it is possible that strong efforts would be made to altogether
eliminate steelworks I use of coal for this purpose. It would probably be more econom-
ical to trim steam raising to a minimum and to use purchased natural gas for minor
energy balancing.
Overall, it is anticipated that in the event of stringent stack-emission standards,
the steel industry would suffer rather less per unit of energy consumed than some
other industries, such as the electric-power industry. The blast furnace, with its
ability to use dirty fuels,. acts as a partial shield. The technology for desulfurizing
coke -oven gas exists, although there is a handicap economically. Steelmaking prac-
tices may be converted to fuel-less techniques (like the basic-oxygen process and the
electric-furnace process).
It is a general conclusion of this portion of the study (concerned with fuels and
energy) that requirements for cleaner air would not unreasonably interfere with the
technology of steelmaking and would not require unreasonable responses by the steel
industry, providing that sufficient time is allowed for the response. Nonetheless, the
steel industry uses well over 23 million Btu of total energy per ton of product. If the
average cost of that energy is 75 cents per million Btu, the associated cost to raw
steel is at least $17 per ton. This works out to an annual cost of something over
$2,000 million per year for steelworks I energy. Upward changes in the costs that are
this large already will of course be reflected in higher costs of manufacturing steel
and higher costs to consumers for steel.
In October, 1968, the Energy Division of the Chase Manhattan Bank (New York)
published the pamphlet IIOutlook for Energy in the United States ". This review and
forecast highlighted some elementary but important considerations in planning long-
range energy sources and showed the following:
(l) The demand for high-grade energy (gasoline, distillate oil, jet fuel,
and natural gas) is skyrocketing, with no end in sight. Both the
number of users and the demand per capita are increasing sharply.
(2) In those geographic areas that contain most U. S. steel plants, there
is already a serious net energy deficit and substantial traffic in
importation of energy from the Gulf Coast and abroad.
(3) The anticipated expanded use of nuclear energy will not meet antici-,
pated requirements for additional energy. '
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V -4:6
(4) The expanded use of coal relative to other fuels is inevitable in
the far future - it is our only certain resource and the only one
potentially convertible to fluid fuels replacing those now derived
from gas and oil.
These conclusions serve as a reminder to be cautious about taking action that
lowers our national capability to use coal. Compared with ordinary fuel-burning uses,
the steel industry seems to offer an excellent opportunity to use coal under controlled
conditions that minimize air pollution. The use of coal in coking and thence to the
blast furnace seems to involve an excellent opportunity for concentration of the
technologies that suppress emission of sulfur compounds.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-47
SOURCES AND AMOUNTS OF NOXIOUS EMISSIONS
IN THE IRON AND STEEL INDUSTRY
Raw materials and fuels consumed in the iron and steel industry contain sulfur in
varied amounts. Coal, iron ore, steel scrap, fluxes, and ferroalloys are the major
bulk raw materials used by the steel industry. Large quantities of liquid and gaseous
fuels and electrical energy are also consumed in the many processing steps requiring
heat.
The presence of sulfur in the raw materials and fuels gives the ironmakers and
steelmakers a problem of controlling the amount of sulfur that ends up in the final fin-
ished product. Sulfur is a deleterious element in most irons and steels. Much of the
refining of steel is conducted to lower its sulfur content. Steelmaking technology per-
mits low sulfur levels to be attained in the metal, accompanied by correspondingly high
sulfur levels in the slags.
During the combustion of fuels in iron and steel processing, some of the reactions
are not complete, with the result that CO is present in some of the combustion gases.
Coke-oven gas, blast-furnace gas, and BOF gas normally contain CO due to the pro-
cessing techniques. Gas from coke ovens and blast furnaces is collected and used as a
fuel. Processes have been developed to collect and use BOF gas, but more commonly
it is burned within the collecting system and then exhausted to the atmosphere. Other
combustion processes within the steel plant generally go to completion unless there is a
malfunction of equipment. However, small amounts of CO (less than 0.5 percent) may
be found in some exit gases.
Fluoride emis sions can be a problem in the making of iron and steel if they are
uncontrolled. Some iron ores in certain local areas contain fluorides of sufficient con-
tent that vegetation has been affected by emissions from steelworks. Calcium fluoride
(in the form of fluorspar) is used in steelmaking to make slags more fluid, and emis-
sions from this use can be ejected into the atmosphere if uncontrolled.
Seven oxides of nitrogen theoretically can exist in the atmosphere. However, for
practical purposes, those of concern appear to be nitrous oxide (NZO), nitric oxide (NO),
nitrogen dioxide (NOZ), and possibly nitrogen pentoxide (NZ05)( 1)*. Little is known of
these in relation to steelmaking activities.
Sulfur in Iron and Steel Plant Processes
The major sources of sulfur in the making of iron and steel exist in the fuels
required to carry: out the various processes. By far most of this sulfur originates in
the coal used to make the coke that is a vital requirement for making the blast-furnace
hot metal. During the coking operation, various by-products are recovered and in turn
some of these are used as fuels. These are coke-oven gas, tar, and pitch. Fuel oil
also contains sulfur which is a contributor to sulfur-oxide emissions. The total con-
sumption of the various hydrocarbon fuels in the manufacture of iron and steel is shown
for the period from 1958 to 1967 in Figure V-5. The rapid increase in consumption of
natural gas is quite evident. Consumption of hydrocarbon fuels per net ton of raw steel
"Numbers in parentheses refer to reference list at end of this section.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
v- 48
65
60
 55
en 
c 
0 Coke
-
-
(I) 
c 
- 
0 50
en
c 
.Q 
.e 

-------
~ 950
c:
:J
o
a.
....: 900
Q)
Q)
-
(/)
~
!£ 85
....
o 30
c:
o
I-
-
Q)
z 25
~
Q)
a.
c:
o
~ 20
E
:J
If)
c:
o
u 15
Q)
:J
LL
V-49
1100
 \            
 \            
   \          
.   \         
    \         
      ~IC oke    
       ~      
        -     
0             
0 ~            ~
  " ~\     n  
  "      
 --- ~/  \     
0    \  VCOke oven gas  
    --....::     
      ,     ;1' 10--- 
0      ,    ,/  
    ~ -'=""'~ --' I'  
   ~   ,- -  ~ / 
   'Natu ra I gas   
0 " -----       
~ -            
  -...-- ~-- --- ...........~ .......Coal    
  -- ~--     
   -- '...     
     , -- --~---   
0       ~- - ~-- 
     "Fuel oi I - ~--  
        r-........ 
       . Ih    
0    Tar and Pltc~    
1050
1000
10
5
o
1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968
Year
FIGURE v-6. CONSUMPTION OF HYDROCARBON FUELS PER
NET TON OF RAW STEEL IN THE UNITED STATES
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-50
produced is shown in Figure V -6 for the same period of time. The trends illustrated in
Figure V -6 are somewhat biased in that the total consumption of each hydrocarbon has
been divided by the total production of raw steel. During the period considered, electric
furnace steelmaking accounted for about 3 percent of all steel in 1958, which increased
to 12.5 percent in 1967. The decreasing use of fuel oil shown in both Figures V-5 and
V -6 is a reflection in the decline of open-hearth production of steel, which was a large
consumer of fuel oil. Increased use of natural gas occurred mainly in reheating fur-
naces, as well as a slight increase in firing power plant and steam boilers. The latter
applications have become more common as a means for lowering the generation of sulfur
emissions by replacing sulfur-containing coals with natural gas that is practically
sulfur-free. The rather constant consumption of tar and pitch is also quite interesting.
Early use of tar and pitch was in the firing of open hearths, but as this use declined
because of the decrease in the number of open hearths, tar and pitch began to be used as
auxiliary fuels in the blast furnace, where the sulfur contained in the tars and pitch was
controlled by the slag action of the blast furnace, rather than entering the atmosphere
as a sulfur-bearing gas, as was the case when used as an open-hearth fuel.
Making Coke
The major source of sulfur in the making of iron and steel is the sulfur content of
the coal used for making coke and the coal used for the firing of power-plant boilers to
generate electricity. A relationship between the percent sulfur content of the coal and
the percent sulfur content of the resulting coke is shown in Figure V -7 for experimental
cokes and for steel-plant blast-furnace cokes(2, 3,4,5). A similar relationship is shown
in Figure V -8 between the percent sulfur content of the coal and pounds of sulfur con-
tained in the coke made from 1 net ton of coal.
 1.5
~ 
~ 
Q) 
0. 
- 
.s::. 
C\ 
'Q) 1.0
3
Q) 
x 
0 
u 
'0 
-
~ 0.5
-
c
o
U
~
:J
.....
"3
en
o Experimental cokes
. Blast furnace cokes
/
~oo
o 0
/
/'
,/
"
0.00.0
0.5 1.0
Sulfur Content of Coal, weight percent
1.5
FIGURE V-7.
SULFUR CONTENT OF COKE AS RELATED TO THE SULFUR
CONTENT OF THE CORRESPONDING COAL
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
  V-51  
 18   
 16   
II)    
"'C    
c: 14   
~   
0    
a.    
Q.) 12   
~    
0    
U    
,5 10   
....    
~    
.....    
- 8   
~   
(/)    
 6   
 4   
 0.0 0.5 1.0 1.5
  Sulfur Content of Coo I, percent 
FIGURE V-8.
POUNDS OF SULFUR IN COKE RESULTING FROM
THE COKING OF ONE NE T TON OF COAL
The sulfur content of the resulting coke-oven gas, also produced from the same coals, is
shown by the relationship in Figure V-9.
8
~ 7
c
~
o
a. 6
II)
o
CJ
5
c
Q.)
>
o 4
I
Q.)
~
8 3
c:
:; 2
.....
~
(/)
FIGURE V-9.
/
8
//
0.5 1.0

Sulfur Content of Coal, percent
1.5
SULFUR IN COKE-OVEN GAS RESULTING FROM
THE COKING OF ONE NET TON OF COAL
The sulfur content of bituminous coals shipped to American coke plants varies
between 0.5 and 2. 1 percent(6). The sulfur content of blast-furnace cokes in the United
BATTELLE MEMORIAL ,INSTITUTE - COLUMBUS LABORATORIES

-------
V-52
States varies between 0.5 and 1. 10 percent, with the average at 0.80 percent. A sulfur
balance for a coke-oven operation in the United States for a coking coal containing
0.8 percent sulfur is given in Table V-9. About 63 percent of the sulfur remains with
the coke and 37 percent goes to the by-products and coke-oven gas. Table V-9 shows
that the production of 1 ton of coke produce s about 15,000 cubic feet of coke -oven gas
with a sulfur content of 1.45 percent.
TABLE V-9.
SULFUR BALANCE FOR COKE OVEN OPERATION
(Based on Production of One Net Ton of Coke)
Item
Amount,  Sulfur, Amount of
pounds  percent Sulfur, pounds
 Input  
2950  0.8 23.60
 Output  
2121  0.700 14.85
290  0.344 1.00
534  1. 45 7.75
  Total 23.60
Coal
Coke and breeze
By-products
Coke-oven gas(a)
(a) Coke-oven gas produced is 15,000 ft3.
About 40 percent of the coke -oven gas is burnt in the oven to supply heat for the coking
proce s s. The combus t,ion of 1 standard cubic foot of coke - oven ga s produce s 5. 78 stan-
dard cubic feet of combustion gases.
The coke-oven gas not consumed in the coke oven may be used as a fuel elsewhere
in the plant, or exhausted to the atmosphere. Sulfur is generally not removed from the
coke -oven gas before use or before venting to the atmosphere.
Sintering Machine
Table V-I0 gives a sulfur balance for a sintering-machine operation. Sources of
sulfur are the iron-bearing material, coke breeze, fuel oil, and limestone. Sulfur is
carried out of the system with the product sinter and as S02 in the combustion gases. It
was estimated that about 36 percent of the sulfur entering the system leaves in the com-
bustion gases.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-53
TABLE V-I0.
SULFUR BALANCE :FOR SINTERlNG MACHINE OPERATION
(Based on the Production of One Net Ton of Sinter)
Item
Amount,
pounds
-Sulfur
Content, percent
Amoun t of
Sulfur, pounds
Input
Iron-bearing material 2200 0.041 0.90
Coke 100 O. 70 0.70
Oil 50 0.55 0.27
Limestone 200 0.049 0.10
  Total 1.97
Output
Sinter
2000
0.055
1. 10
Sinter fines
289
0.055
0.16
Sulfur in combustion gases
0.71
Total
1.97
Blast Furnace s
Sulfur enters the blast furnace from practically all of th~ raw materials used for
making iron; the major source being the sulfur in the coke. Because-of the reducing
conditions in the blast furnace, it is not possible for any of the sulfur present in the
charge materials to be oxidized and leave the blast furnace as sulfur dioxide. Sulfur
balances, based on chemical analyses of the iron and slag, have shown this to be the
case. However, chemical analyses of slags are subject to variations, and such balances
do hold a possibility for error. Analysis of blast-furnace gases during full-scale tests
of oil-injection in the Bureau of Mines experimental blast furnace verified that sulfur
gases are absent from the flue gas. Quoting from the reports, "It was suggested that
some of the sulfur from the oil might leave the furnace in the top gas. However, careful
analyses of the top gas throughout the test program showed no traces of sulfur
compounds. ,,( 7)
A sulfur balance for a blast furnace operating with sinter and screened ore (based
on the material balance for a blast furnace shown in Figure C-24 in Appendix C) is given
in Table V-II.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
1--'-"--
V-54
TABLE V-I!.
SULFUR BALANCE FOR BLAST FURNACE FROM
FIGURE C-24 IN APPENDIX C
(Based on Production of One Net Ton of Pig Iron)
 Amount,  Sulfur  Amount of
Item pounds Content, percent  Sulfur, pounds
  Input   
Sinter 2300  0.055  1. 26
Screened ore 790  0.040  0.32
Coke 1025  0.700  7.18
Limestone 125  0.049  0.06
Natural gas 45  Nil(a)  
    Total 8.82
  Output   
Pig iron 2000  0.030  0.60
Slag 460  1. 786  8.22
    Total 8.82
(a) Columbia Gas stated that the sulfur content of natural gas is less than 0.001 percent.
When sulfur-bearing slag is flushed from the.blast furnace, the sulfur reacts with
oxygen to form sulfur dioxide near the runner of the blast furnace, In wet weather, the
sulfur in the solid slag may react with water to form hydrogen sulfide which is released
to the atmosphere. The latter reaction also occurs during the granulation of hot slag
with water.
Open Hearth Steelmaking
Sulfur enters the open hearth steelmaking system from the raw materials and from
the fuels used for combustion. A comprehensive survey conducted on sulfur balances
in open hearths showed average source values listed in Table V-12(8), Other investiga-
tors have shown that the type of scrap can influence sulfur pickup by the scrap from the
fuel, with subsequent partition of the sulfur to the slag or metal. The results showed
that heavy scrap resulted in a sulfur pickup of 0.015 percent, while a light scrap melted
with the same fuel resulted in a sulfur pickup of 0.044 percent(9). Natural-gas firing in
open hearths is also reported to remove sulfur from the charge materials up to as much
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-55
as 0.020 percent(9). This means that even though a sulfur-free fuel is used, sulfur
emissions to the atmosphere may occur because of pickup of'sulfur from the open-
hearth charge materials. It has been reported that 25 percent of the sulfur in fuel oil
used to fire open hearths is retained in the metal and slag, and that this entire amount
is absorbed by the scrap during the meltdown period' of the heat( 10). A relationship
TABLE V-12.
SOURCES OF SULFUR IN OPEN HEARTH STEELMAKING
Source
Sulfur Concentration, percent
Hot metal
Fuel
Scrap
Limestone
Ore
Other s
38
22
19
14
5
2
between the sulfur content of the fuel oil and the effect of increasing sulfur content on
the final sulfur in the steel is shown in Figure V-10.
0.050
.-
c
Q)
u
~ 0.045
0.
-
Q)
Q)
en 0.040
-
o
.-
c
~ 0.035
c
o
U
....
~ 0.030
::J
(/)
0.025
0.0
1.0 2.0 3.0
Sulfur Content of Oi I t percent

SULFUR CONTENT IN STEEL AS AFFECTED
BY INCREASED SULFUR IN FUEL OIL
4.0
FIGURE V-IO.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-56
About 47 percent of the sulfur entering the open-hearth system leaves in the metal
and slag. The remaining sulfur exits with the products of combustion in the form of
sulfur dioxide. A sulfur balance for an open-hearth furnace using 60 percent hot metal
and 40 percent steel scrap (as shown in the flow sheet and material balance in Fig-
ure C-41 in Appendix C) is given in Table V-l3.
TABLE V-13.
SULFUR BALANCE FOR OPEN HEARTH FURNACE WITH
60 PERCENT HOT METAL AND 40 PERCENT STEEL
SCRAP OXYGEN PRACTICE (FROM FIGURE C-41 IN
APPENDIX C)
(Based on Production of One Net Ton of Raw Steel)
Slag
200
 Sulfu r  Amount of
Content, percent  Sulfur, pound s
Input   
 0.030  0.41
 0.020  O. 18
 0.040  0.03
 0.049  0.07
 0.55  0.61
 0.07  0.01
  Total 1. 31
Output   
 0.02  0.41
 O. 10  0.20
   0.70
  Total 1. 31
Item
Amount,
pounds
Hot metal 1361
Steel scrap 907
Iron ore 70
Flux 150
Fuel oil 111
Ferroalloys 14
Steel and scrap
2060
Combustion gases
Basic Oxygen Furnace Steelmaking
The major source of sulfur in the basic oxygen steelmaking process is the sulfur
in the hot metal from the blast furnace. External fuels are not required to supply heat
to bring about the desired chemical reactions for refining. High-purity gaseous oxygen
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-57
impinged on the surface of the molten iron causes rapid oxidation of silicon carbon, and
iron; which in turn supplies the heat necessary for completion of refining,
Slag chemistry controls the removal of sulfur from the hot metal during the
refining operation, in a manner similar to that for other steelmaking processes. The
major difference in the BOF process is that burnt lime (CaO) is used to provide the
desired slag characteristics, rather than limestone (CaC03), as in open-hearth steel-
making. The rapidity of the refining process in BOF practice (again as compared to
open hearth practice) dictates this requirement. There is not sufficient time to calcine

limestone. The effects of slag basicity (in this case the ratio of ~i6~) on the sulfur

content of the refined steel is shown in Figure V -11 (11). A relationship bet~een the
sulfur content in the hot metal and sulfur content in the steel as affected by slag basicity
is shown in Figure V-12. The percentage of sulfur removed from the hot metal as
affected by slag basicity is shown in Figure V -13( 11). Sulfur content of the burnt lime
has little effect on the final sulfur content of the steel, as shown in Figure V -14(11).
The relationships shown in Figures V -11 through V -14 can be considered as representa-
tive of sulfur removal in basic oxygen steelmaking practice. Specific results may vary
from plant to plant depending on the silicon content of the hot metal, sulfur content of
the scrap, and specific slag practices.
A sulfur balance for a BOF practice using 70 percent hot metal and 30 percent
steel scrap, and based on the material balance shown in Figure C-55 in Appendix C, is
given in Table V -14. Almost all of the sulfur leaves the system in the metal and slag,
with only about 4 percent of the sulfur leaving as sulfur dioxide in the off-gas. Detailed
information on the sulfur content of BOF off-gases is not available in the published
literature.
TABLE V-14. SULFUR BALANCE FOR BOF STEELMAKING BASED ON
MATERIAL BALANCE GIVEN IN FIGURE C-55 WITH
70 PERCENT HOT METAL AND 30 PERCENT
STEEL SCRAP
(Based on Production of One Net Ton of Raw Steel. )
 Amount,
Item pounds
Hot metal 1581
Steel scrap 678
Burnt lime 14Z
Ferroalloys 14
Steel and scrap
Z077
Slag
Z63
Sulfur ;n off-gas
 Sulfur Amount of
Content, percent Sulfur, pounds
Input  
 0.030 0.47
 o. no o. 14
 0.060 0.08
 0.070 0,01
 Total 0.70 .
Output  
 O.OZO 0.41
 O. 100 0.Z6
  0.03
 Total 0.70
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
0.026
+-
C
Q)
o
~ 0.024
a.
~
~ 0.022
+-
CJ)
. '00.020
+-
C
Q)
~ 0.018
o
u

~ 0.016
-
:J
CJ) 0.014
V-58
2.5
o
o
3.0
3.5
o
4.0
4.5
Tap Slag Basicity I (CaO/Si02)
- .
FIGURE V -11. SULFUR CONTENT OF BOF STEEL AS AFFECTED
BY BASICITY OF THE TAP SLAG
0.026
+-
C
~ 0.024
...
Q)
a.
-: 0.022
Q)
Q)
+-
CJ) 0.020
-
o
c: 0.0 18
Q)
+-
C
o
U 0.016
...
:J .
-
:J 0.014
CJ)
510 Basiei ty (Co 0/ 5i 02)1


'2. .5
\.)~de{
'2. 5 - 3.0 .
~ \ -3.6
3'7- 4.'2.
O~e{ 4.'2.
0.030 0.040 0.050 .
Sulfur Content of Hot Metal, percent

FIGURE \7-12. RELATIONSHIP BETWEEN SULFUR CONTENT IN
HOT METAL AND SULFUR CONTENT IN BOF
STEEL AS AFFECTED BY SLAG BASICITY
0.020
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
- 60
c:
(1)
u
~ 50
a.
A
~
~ 40
o
E
(1)
a:: 30
~
~
-
"3 20
(f)
V-59
70
10

0.020 0.025 0.030 0.035 0.040 0.045 0.050 0.055
Sulfur Content of Hot Metal, percent
FIGURE V -13. EFFECT OF SLAG BASICITY ON THE PERCENT OF
SULFUR REMOVED FROM THE HOT METAL
0.023
C 0.022
~
...
(1)
a. 0.021
-
(1)
(1)
 0.020
-
o
~ 0.019
-
c:
o
u 0.018
~
~
-
~ 0.017
0.016
0.025
SIJ bOSiCify(coLSi/
/
Under 2.7
I
3.1 3.6
I
Over 3.7
0.050 0.075 0.100
Sulfur Content of Burnt Lime, percent
0.125
FIGURE V -14. EFFECT OF SULFUR CONTENT IN THE LIME AND SLAG
BASICITY ON THE SULFUR CONTENT OF BOF ST~EL
BATTELLE MEMORIAL INSTITUTE - COLUMBUS. LABORATORIES

-------
V-60
Electric Furnace Steelmaking
Like the BOF process, the electric-furnace steelmaking process does not depend
on a sulfur-bearing fuel as the major source of energy. Electrical energy provides the
heat for melting and refining steel scrap to produce molten metal. Sulfur enters the
system with the steelmaking raw materials. Small amounts of coke breeze are added to
. provide carbon for the carbon boil during the refining period. As in the other steel-
making processes, slag control plays a major role in removal of sulfur from the steel.
Because many high-carbon steels are made in electric furnaces (as compared to the
lower-carbon steels made in open hearths and BOFs), the carbon content of the steel
produced must also be taken into consideration in the desulfurization slag practice. A
relationship between the carbon content of the steel and the sulfur content for several
different kinds of steel made in two different plants is shown in Figure V -15(12, 13).
The relationship between slag basicity and the partition of sulfur between the metal and
slag is shown in Figure V-16(14). The effect of the time under a reducing slag on the
sulfur content of AISI 52100 steel is illustrated in Figure V-I 7( 12). The percent reduc-
tion in sulfur content for several high-carbon steels made in electric furnaces is shown
in Figure V-1S(13).
0.025
u. S. Steel Corporation
-
c
~ 0,020
'-
Q)
a.
~ 0,015
-
::I
en
.!!:? 0.010
1:1
c
-.J
0.0050
1.00
FIGUR~ V-IS.
RELA TIONSHIP BETWEEN CARBON CONTENT IN
THE STEEL AND SULFUR CONTENT IN THE STEEL
A sulfur balance for electric-furnace steelmaking (based on the material balance
given in Figure C-65 in Appendix C) is given in Table V-IS. Data on the sulfur content
in electric-furnace steelmaking off-gases are not available in the published literature.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
o 0'1
- 0
~ en 6
Q)
~ ~ 5
-
c:
.S .- 4
~
~ ;:,
..2 ~ 3
- ;:,
~ en "
2
V-6l
9
8
7
00 2 3 4 5 6 7 8 9
   Mole Ratio    
  CaO + MgO + MnO   
  Si02 + AI203 + 1P20~   
FIGURE V-l6. RELATIONSHIP BETWEEN SLAG BASICITY, IRON OXIDE
CONTENT OF THE SLAG, AND THE PAR TITION OF
SULFUR BETWEEN THE METAL AND THE SLAG
0.030
0.025
-
c:
~
"~
Q)
a.
0.020
~
;:,
.s::.
a.
;:,
en
Q)"
'0
o
....J
0.015
0.010
0.005
o
0.024   
 0.019  
Avg  0.014 
0.013 Avg Avg 
 0.011 0.011 -'-
43 heats  23 heats 
0.008 77 heats 
0.007 
 0.006 
1:00-1:30 1:30-2:00 Over 2:00
Time Under Reducing Slag, hours
FIGURE V -17. EFFECT OF TIME UNDER A REDUCING SLAG ON THE
SULFUR CONTENT OF AISI 52100 GRADE STEEL
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-62
75
o
-
c:
Q)
~ 70
a.
Q)
'5
.3
o
-
65
-

~ 60
I
c:
.Q
-
g 55
"C
Q)
a::
o
...
::J
-
""5
CJ)
50
45 0.025
0.030
Melt Sulfur, percent
0.035
FIGURE V -18.
SULFUR REDUCTION IN ELECTRIC FURNACE STEELMAKING
TABLE V-I5. SULFUR BALANCE FOR ELECTRIC FURNACE STEELMAKING
USING A CHARGE OF COLD STEEL SCRAP AND OXYGEN
PRACTICE (FROM FIGURE C-65 IN APPENDIX C)
(Based on the Production of One Net Ton of Raw Steel)
 Amount,
Item pounds
Steel scrap 2136
Coke breeze 6
Burnt lime 99
Ferroalloys 14
Sulfur
Content, per cent
Amount of

Sulfur , pounds
Input
0.020
0.700
0.43
0.04
0.060
0.070
0.06
0.01
Total
0.54
Steel and scrap
2060
140
Output
0.020
0.080
0.42
O. 11
Slag
Off-gas
Total
0.01
O. 54
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-63
Summary Balance of Sulfur for
Iron and Steel Industry
The amounts of sulfur-bearing raw materials and fuels consumed in the U. S. iron
and steel industry were tabulated from The Statistical Yearbook (1967) of the AISI. The
average sulfur content of the materials was estimated to compile a tentative sulfur
balance for the U. S. steel industry. The resulting tentative summary balance for sulfur
is given in Table V -16. The tentative and uncertain nature of this balance should be
recognized. In particular, gaseous emissions were obtained by difference. This is a
hazardous procedure, but the best available for estimation using the presently available
data. Some additional research on this subject undoubtedly could produce balances with
a higher degree of authenticity.
Nearly 76 percent of the sulfur input to the system came from the coal used to
produce coke and to generate electric power. Iron ore accounted for 9.8 percent of the
sulfur input. Steel scrap, fluxes and ferroalloys resulted in an input of about 4 percent
of the sulfur. Fuels for the open hearth furnace operation (oil, tar, and pitch) accounted
for about 3.3 percent of the sulfur input. Providing that the coke-oven gas was con-
sumed in the coke oven, blast furnace stoves, and soaking pits; almost all of the sulfur
input would be in the iron and steelmaking area.
The output of sulfur in the blast furnace slag is about 47 percent of the total sulfur
entering the system. The steel and steelmaking slag contain about 6 percent of the
sulfur output. According to this calculation, 39 percent of the sulfur would be in off-gas
or in products of combustion, and 8 percent unaccountable.
In the heating of steel for rolling or forging, there is no change in the sulfur con-
tent of the steel. When heating with natural gas with a low sulfur content or with electric
power, little or no sulfur would appear in the products of combustion.
If sulfur-bearing fuel oil or gas is used for heating or for reheating furnaces, the
sulfur in the form of S02 will be in the exhaust gas. During processing, the scale on
steel may absorb sulfur in an amount up to 1 percent when sulfur is present in the
exhaust gas.
CO Balances for the Iron and Steel Industry
Three steelmaking process segments normally produce an off-gas containing sub-
stantial amounts of CO: (1) coke-oven gas contains from 4.5 to 6.9 percent CO, depend-
ing upon the coal used during the coking proces s; (2) the blast-furnace top gas may have
a CO content from about 20 to 30 percent, depending upon operating practice; and (3) the
.... OF gas before combustion with aspirated air may contain from 74 to 90.5 percent CO,
and after combustion with aspirated air from 0.0 to 0.3 percent.
Off-gas from other steel-plant processes will not normally contain any appreciable
amounts of CO as the combustion process is meant to go to completion. However, such
gas may contain from 0.3 to 0.5 percent CO. .
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-64
TABLE V-16. ESTIMATED SUMMARY OF SULFUR BALANCE FOR THE U. S. IRON
 AND STEEL n~DUSTRY (1967)  
   Estimated 
  Amount, A verage Sulfur Total
  million Content, Sulfur
Item  NT percent Content, NT
 Input   
Coal  82.7 0.95 (786,000)
Coking    
Coke oven gas 196,000 NT   196,000
Coke 455,000 NT   455,000
Tar and pitch 21,000 NT   
Other by -products 114,000 NT   
 786,000 NT   
Other  6.9 1. 00 69,000
Iron ore  133.3 0.07 93,000
Steel scrap  65.4 0.03 20,000
Ferroalloys  2.5 0.07 2,000
Fluxes  31.4 0.05-1.00 21,000
Fuel oil    
Blast furnace  0.3 1.80 5,000
Other use (fuel)  4.7 1. 80 85,000
Tar and pitch    
8last furnace  0.2 0.60 1,000
Other use  1.3 0.60 8,000
Liquid petroleum gas  0.1 Nil 
Natural gas  12.2 Nil 
 Total   955,000
 Output   
Blast furnace slag  30.0 1. 50 450.000
Coke 431,000 NT   
Iron ore 93,000 NT   
Flux 15,000 NT   
Fuel oil 5,000 NT   
Tar and pitch ~NT   
 545,000 NT   
Sulfur to hot metal -27.000 NT   
 518,000 NT   
Raw steel and scrap  131.7 0.03 39,000
Steelmaking slag  13.6 0.10 14,000
Fuels    
Coke oven gas 196,000 NT   
Fuel oil 77, 000 NT   
Tar and pitch 8,000 NT   
Coke breeze 24,000 NT   
Coal 69,000 NT   
 374,000 NT   374,000
Possible loss from slag (518,000-450,000)   68,000
Unaccounted for    10,000
 Total   955,000
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-65
Coke -Oven Gas
A conventional size of slot oven in the United States carbonizes about 24 tons of
coal per day. Table V-9 shows that about 15,000 cubic feet of coke-oven gas is produced
for each ton of coke. This is about 10,000 cubic feet of gas for each ton of coal that is
carbonized.
A base of carbonizing 1 ton of coal per hour was selected to illustrate the flow of
coke -oven gas and its CO component:
Item
Volume Gas
CO Content
Source
Coke -oven gas
10, 000 £t3
550 ft3
---------------------------------------------------------
Use
Coke-oven gas to ovens
Other plant use
4, 000 £t3
6,OOO£t3
220 ft3
330 rt3
---------------------------------------------------------
End Product
Products of combustion

from coke -oven gas
44, 800 ft3
Nil
The CO component of coke -oven gas using the above basis may vary from 490 to
690 cubic feet per hour. Under normal conditions, raw coke -oven gas will not be
released to the air. However, leaks in the collecting system would release both CO and
sulfur-bearing gas to the atmosphere. .
Blast Furnace Top Gas
Based on Figure C -24 in Appendix C, the material balance shows the production
of 6450 pounds of blast-furnace top gas per net ton of hot metal. The approximate typi-
cal disposition of the top gas and its CO component is tabulated below:
Item
Amount, pounds
CO, pounds
Source
Blast furnace top gas
6450
1612.5
-----------------------------------------------------------
Use
Soaking pits or other
plant use

Blast furnace stove s
4850
1600
1212.5
400.0
-----------------------------------------------------------
End Product
Combustion products
from stoves
2460
Nil
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-66
The amount of CO produced per net ton of pig iron may vary from 1290 to 1935
pounds. The gas is collected and cleaned before use, and under normal operating con-
ditions is not exhausted to the atmosphere.
BOF Process
Based on Figure C-65 in Appendix C, the BOF process produces about 168 pounds
of off-gas per ton of steel. The typical and approximate disposition of the top gas and
the CO component is tabulated below:
Item
Amount of
Gas, pounds per
net ton of steel
CO, pounds per
net ton of steel
Source
Off-gas

Aspirated air added
(theoretical)
168
717-877
124.3-152
---------------------------------------------------------------
End Product
Total off-gas
885-1045
0-2.6
The off-gas from a BOF operation based on the example above will produce about
124 to 152 pounds of CO per ton of steel. After aspirated air is added to the off-gas,
the weight of gas will increase to about 885 to 1045 pounds per ton of steel and will con-
tain frorp. 0 to about 2. 6 pounds of CO.
During the normal operation of an iron and steel plant, only small amounts of CO
would reach the atmosphere from this source.
Fluoride Emis s ions
The detrimental effect of fluoride emissions on vegetation has been documented
and will not be discussed further here. In years past, uncontrolled emissions of fluo-
rides from some localized steel plant operations have contributed to this situation.
Because fluorspar (a calcium fluoride mineral) is used extensively in the making of
steel, the industry has been suspect as a contributor to fluoride emissions. This doubt
has been further advanced because no data are available concerning the fluoride emis-
sions generated in the different steelmaking processes. A recent report on work done
in Germany at August Thyssen-Hutte AG, Duisburg-Hamborn, has shown that fluoride
emissions generated during steelmaking in basic oxygen furnaces and electric furnaces
become bonded to the other particulates generated and are removedduririg the cleaning
operations of existing emission control equipment(l5). Analyses for fluoride emissions
were made continually during the course of two, 90-ton BOF heats. The fluoride con-
tent of the emissions to the atmosphere varied from 0.000045 grain per cubic foot (STP)
to 0.0005 grain per cubic foot (STP) in the first heat; and from 0.000045 to 0.0009 grain
per cubic foot (STP) in the second heat. Similar tests carried out with 20-ton electric-
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-67
arc furnaces showed fluoride particulate contents of O. 023 to 0.250 grain per cubic foot
(STP) and gaseous fluoride contents of O. 003 to 0.025 grain per cubic foot (STP) in the
raw gas. These values were reduced to 0.004 to 0.04 grain per cubic foot for par-
ticulates and 0.002 to 0.02 grain per cubic foot for gaseous fluorides. Collection
efficiency for the particulate fluorides was reported to be 90 percent, while collection
of gaseous fluorides was reported to be only 65 percent. The l~w collection efficiency
for the gaseous fluorides was suggested to be due to moisture in the off-gases which
combined with the gaseous fluorides to form hydrofluoric acid which in turn decomposed
on the gas probes to give high gaseous emissions in the off-gas. Fluoride contents of
sinter-plant waste gases were determined to vary from 0.0003 to 0.002 grain per cubic
foot for particulates, and 0.0009 to 0.01 grain per cubic foot for gaseous fluorides.
Nitrogen Oxide Emis sions
The amount of nitrogen oxide emissions generated in the steelmaking and allied
processes is not available in the published literature. Based on results obtained in
other industries, nitrogen oxide emissions probably occur in the steel industry. Be-
cause natural gas is used quite extensively in the iron and steel industry, this alone
would indicate a source of nitrogen oxide emissions. Gas analyses, specifically for
nitrogen gases, apparently have not been considered as a matter for even sporadic atten-
tion by the steel industry.
References
1. Faith, W. L., "Air Pollution Control, Chapter 5, Gases", John Wiley and Sons,
Inc., New York, 1959, pp. 142-174.
2. Wolfson, D. E., and Birge, G. W., "Carbonizing Properties of Allegheny County,
Pa., Coals", Bureau of Mines Report of Investigations 5455, 1959, 16 pp.
3. Birge, G. W., et al., "Carbonizing Properties of Coals from Wyoming and Mercer
Counties, W. Va. ", Bureau of Mines Report of Investigations 6615, 1965, 21 pp.
4. Wolfson, D. E., and Ortuglio, C., "Carbonizing Properties of Coals from Fayette,
Green and Washington Counties, Pa.", Bureau of Mines Report of Investigations
7131, 1966, 18 pp.
5. Wolfson, D. E., et al., "Relation of Properties of Coke Produced by BM-AGA and
Industrial Methods", AIME Blast Furnace, Coke Oven and Raw Materials
Proceedings, 20 (1961), pp 387-398.
6. DeCarlo, et al., "Sulfur Content of United States Coals", Bureau of Mines Informa-
tion Circular 8312, 1966, pp 40-44.
7. Woolf, P. L., and Mahan, W. M., "Fuel Oil Injection in an Experimental Blast
Furnace", Bureau of Mines Report of Investigations 6150, 1963, P 13.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
V-68
8. Mayo, W. H., "Training Personnel in the Control of Sulfur in the Open Hearth",
AISI Regional Technical Meeting, Philadelphia, 1952.
9. Lightner, M. W., "Current Concepts of Open Hearth Slag Control", AIME Open
Hearth Proceedings, 40, 1957, pp 304-314.
10. Trentini, B. A., et al., "Sulphurization from Fuel Oil", Iron and Steel, 29 (12),
December 1956, pp 537-542, ~ (1), January 1957, pp 11-17.
11. Macnamara, J., Discussion to "The Effects of Lime Properties on Basic Oxygen
Steelmaking", AIME Open Hearth Proceedings, 48, 1965, pp 79-82.
12. Queneau, B. R., and Klimas, C. V., "Sulfur Control in Electric Furnace Steel-
making", AIME Electric Furnace Steel Proceedings, ~, 1953, pp 281-289.
13. Van Voris, F. E., Discussion to "Sulfur Control in Electric Furnace Steelmaking",
AIME Electric Furnace Steel Proceedings, ~, 1953.
14. Bishop, H. L., Jr., et al., "Equilibria of Sulfur and Oxygen Between Liquid Iron
and Open Hearth Type Slags", Transactions AIME, 206 (1956), pp 862-868.
15. Graue, G., and Nagel, H., "Detection and Removal of Fluorine
of a Steelworks and Measuring Air Quality in Its Surroundings",
Luft, 28 (1), pp 9-17 (January, 1968).
in the Waste Gases
Staub-Reinhalt,
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-l
SECTION VI
ANALYSIS OF APPLIED CONTROL SYSTEMS
An explicit analysis of the various air-pollution control systems as they are applied
in the integrated iron and steel industry would of necessity be based on complete, ac-
curate, first-hand information pertaining to the various systems as they perform their
function in various iron and steelmaking operations. Unfortunately, information with
these qualities has not beei-I readily available in Phase I. Therefore, analysis as it
applies to this study must be based to a considerable degree on (1) what the capabilities
and limitations of various systems appear to be, (2) information available in the pub-
lished literature, and (3) information obtained by visitations to selected plants during
the study.
An analysis of several comparable pieces of equipment, processes, process com-
ponents, or materials usually is made with reference to some common standard or
boundary conditions. A well-defined, generally accepted standard of reference for com-
paring performance of equipment for controlling metallurgical emissions to the air is
not yet available. One standard that is used as a means of rating some pollution-control
systems is highly dependent on personal evaluation. This is the "Ringelmann Chart"
illustrated in Figure 1. (1)':' Development of the Ringelmann Chart was done on an arbi-
trary basis by Maximilien Ringelmann (of France) and was introduced to the technical
community of the United States in 1897 via a technical news publication. The Techno-
logic Branch of the U. S. Geological Survey .(predecessor to the Bureau of Mines) used
the chart in studies of smokeles s combustion in 1904, and by 1910 the chart was recog-
nized officially in Boston smoke ordinances. (2) Over the years, the Ringelmann Chart
has become accepted in the United States as a means of determining emis sion levels.
It has been only recently that equipment has been developed to monitor emis sions from
a steelmaking process - the basic oxygen furnace. (3) Monitoring of emissions by this
new method is based on grain loading per cubic foot of gas Equipment manufacturers
are reported to be unwilling to supply pollution-control equipment to any Ringelmann
requirement for metallurgical emissions, but do supply equipment to meet the various
grain-loading requirements. (4)
Equivalent to
20 per cent black.
Equivalent to Equivalent to
40 per cent black. 60 per cent black.
Equivalent to
80 per cent black.
FIGURE VI-I.
RINGELMANN SMOKE CHART
. References for Section VI-l are given at the end of the Section.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-2
Comparison of reported operating efficiences for the various emission-control
installations covered in this study is done on the basis of grain loadings per cubic foot
of exhaust gas. This method of comparing operating efficiences is used because there
appears to be no better method of making comparisons.
Even with the use of specific operating data,
calculations are not considered to be sufficient for
following statements made in 1963(5):
comparisons based on theoretical

good comparisons, as shown by the
"A rational theoretical equation covering all mechanisms in filtration
does not yet exist nor is there an exact overall equation for electro-
static charging and separation for all sizes considering all properties
of particles in an electrostatically charged field. "
"For the more complex cases we will still depend on Ipilot plant'
evaluations because of the economic risk in selecting a collector
if only a limited number of parameters can be resolved by the best
measuring techniques. "
According to technologists knowledgeable in this area of air -pollution control
technology(6), these same statements hold true in 1968.
The selection of an air-pollution control system for a particular ironmaking or
steelmaking process or process segment is not a straightforward, "off the shelf" type
of selection. Process variables, space availability, operating costs, inherent control-
equipment advantages, and future allowable-emission regulations must all be considered
in selecting emission-control equipment. (7) The factor of future allowable-emission
levels is of major concern because no one knows what these standards may be, and steel
companies are in a position of installing equipm.ent without knowing what future require-
ments will be. Usually, equipment installed to meet one level of control cannot be up-
graded inexpensively to meet a stricter level.
Many ordinances of a local nature establish requirements for allowable emissions
at the level of 0.05 grain per cubic foot of gas. However, this is not to say that this will
continue to be the acceptable level of allowable emissions. Color of emissions and other
factors are rapidly becoming points of concern to the steel industry and regulatory
agencie s.
The effect of particulate emissions on weather patterns is currently under study by
the National Center for Atmospheric Research. A recent report contains the following
statements with respect to the generation of ice nuclei and their possible effect on in-
creased rainfall(8):
"A search was made for strong local source s of ice nuclei. Apart from
dust blown up by strong winds, steel mills have proven to be the only
source. 11
"The test on 15 August was made to locate the source of the nuclei-
which was traced to two orange plume s emanating from two stacks of
a Gary steel mill. "
BATTELLE MEMORIAL INSTITUTE ~ COLUMBUS LABORATORIES

-------
VI-3
From the situations presented by these statements, it is apparent that the level of
allowable emissions is not stabilized, and the definition of an air pollutant has not been
completely resolved.
A factor affecting the e'conomics of air -pollution control in the iron and steel in-
dustry is the amount of allowable emissions based on the weight of materials processed.
The Los Angeles County, California, Air Pollution Control District permits a maximum
discharge weight of 40 pounds per hour for a process weight of 30 or more tons per
hour(9). This is based on the discharge from a single stack. If two or more steelmaking
furnaces, each making 30 tons per hour of steel, are operated simultaneously and the
emissions collected by a manifolding system for collection, the allowable discharge to
the atmosphere is still only 40 pounds per hour. This means that the collection system
would have to be de signed and constructed to a higher level of efficiency at an increased
cost.
This analysis of the various types of emission-control equipment used in the inte-
grated iron and steel industry deals primarily with the types of equipment as they apply
to the various process segments; additional analytical consideration is given to process
segments that have problems associated with the process itself, rather than with the
collection equipment (such as coke making). As a' point of information, a preliminary
tabulation of the types of equipment used in the various processes is given in Table VI-I.
This tabulation is not complete, and is based mainly on information in the published
literature. The tabulation shows, however, that electrostatic precipitators have found
the greatest application in the iron and steel industry; followed by scrubbers, mechan-
ical collectors (which are primarily cyclones), and fabric filters (bag houses).
TABLE VI-I.
REPRESENTA TIVE EMISSION -CONTROL APPLICA TIONS IN THE
INTEGRATED IRON AND STEEL INDUSTRY
Iron or Steelmaking
Segment

Sinte r plant
T'Olast furnace (a)
Open-hearth furnace
Basic oxygen furnace
Electric furnace
Scarfing
Type of Emission-Control Equipment
Mechanical Scrubbers Precipitators
Fabrics
17
13 (b)
o
o
o
4
2
51
6
15
5
4
9
108
93
23
1
3
3
o
o
o
29
2
(a) Final control equipment.
(b) Dust collectors followed by other equipment are not considered.
Electrostatic Precipitators
The removal of dust particles (primarily iron oxide) from exhaust gases of iron-
making and steelmaking processes is accomplished by passing the dust-laden gas be-
tween a pair of electrodes; one being a discharge electrode at 'a high potential and the
other an electrically grounded electrode (which is the collecting electrode), The poten-
tial difference must be great enough to establish a corona discharge around the discharge
electrode. Gas ions formed in ,the corona move rapidly toward the collecting electrode
under the action of the electrical field, and transfer their charge to the dust particles by
collision with them. The electrical field interacting with the charge on the particles
causes the particles to drift toward the collecting electrode and to be deposite'd on the
collecting electrode. (10) , . -,' ,

BATTELLE "MEMORIAL INSTITU,TE - COLUMBUS LABORATORIES

-------
VI-4
Electrostatic precipitators are of two types, (1) wire-in-tube and (2) wire-and-
plate (see Figures B-5 and B-6, in Appendix B). The wire-and-plate type is generally
used for the control of emissions in the integrated iron and steel industry.
The electrical theory used to design electrostatic precipitators is well developed,
and will not be discussed here. Interested readers can refer to some of the texts on the
subject. (10,11,12) In addition to the electrical environment within the precipitator,
other factors that affect the performance of electrical precipitators include (1) resis-
tivity of the dust, (2) its moisture content, (3) the gas-flow rate, and (4) the distribution
of gas flow within the precipitator.
Resistivity must be taken into account in the initial design phases of the precipita-
tor. This information must either be obtained by experimental methods or be available
from studies on similar operations. The variability of resistivity for processes produc-
ing iron oxide dusts is illustrated in Figure VI-2. (11,13) The wide range of resistivities
that can be. encountered makes it preferable to determine resistivities for the specific
process than to rely on speculation..
13
10
12
10
Open
hearth
II
E 10
u
I
E
.c.
o 10
10
~
":;
~
.~ 109
a:
C
~
c
a.
a. 8

-------
VI-5
again, the difference in resistivities for the fume for two different open hearths is
apparent, and illustrates the neces sity for knowledge about the characteristics of
specific fumes. The great difference in resistivities between limestone dust and the
open-hearth dust points up a factor ~hat can cause operating problems in the operation
of electrostatic precipitators for collecting open-hearth and BOF fume. Dust resulting
from the use of limestone (CaC03) as a fluxing agent in open hearths can cause less than
optimum performance of the electrostatic precipitator because of this large difference
in resistivities. Similar problems are associated with the use of lime (CaO) as a fluxing
agent in BOF practice.
IC"
E
(,)
I
E
.c
o
10"
10"
>.
...
:~
...
I/)
'x: 10'0
a::
...
c
G)
~
o
Q,
Q,

-------
99.9
-
c:
Q)
()
:u 99
c.
>.
()
c:
Q)
()
;;: 90
-
W
FIGURE VI-4.
VI-6
00
2 3 4
A (Collecting Surface Area), ft2
V (Gas Volume Flow), feet per second
5
6
PRECIPITA TOR EFFICIENCY AS RELA TED TO
COLLECTING-SURFACE AREA, GAS-FLOW
RA TE, AND DRIFT VELOCITY
The effect of variations in gas -volume flow on precipitator efficiency is illustrated
in Figure VI-5. The decrease in efficiency as the rated volume of the precipitators is
exceeded is quite evident. .
+-
c:
Q)
o
....
Q)
a. 90
>-
o
c:
Q)
o
'+-
'+-
W 80
FIGURE VI-5.
100
Designed efficiency
98 percent
Designed eff ic iency
95 percent
o
0.50
1.0
Ra ted Vol ume
1.5
2.0
EFFECT OF GAS VOLUME ON PRECIPITATOR EFFICIENCY
Electrostatic precipitators are used for a variety of processes in the iron and
steel industry, ranging from their use in sinter plants to their application in scarfing
operations.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-7
Sinter -Plant Applications of
Electrostatic Precipitators
The very nature of a sinter plant (with its multitude of transfer points for materials,
and discharge points for receiving, cooling, and screening the sinter) creates a severe
emission problem. Various types of pollution-control equipment are used for a single
plant; or if the operation can be sufficiently enclosed, a central pollution-control instal-
lation may suffice. Electrostatic precipitators are used as secondary air-cleaning units
in sinter -plant operations for the treatment of dust-laden gases coming from sintering-
strand windboxes. The only reported information located on dust loadings for electro-
static precipitators on sintering machines is that for the Inland Steel Company sintering
machine in East Chicago, Indiana. This installation was reported to handle an input dust
loading of 2.5 grains/ scf of gas at 457,000 cfm, and yield an output dl.lst loading of
0.038 grain/ scf gas; (16) an efficiency of 98.5 percent. However, since the pollution-
control system was installed, the materials charged to the sintering machine have
changed from straight ore fines to ore, flue dust, and lime. The characteristics of the
ore used has also changed. These changes in materials have resulted in an increase of
output dust loading to 0.25 grains per cubic foot, and a decrease in collection efficiency
to 90 percent.
Installation of most sintering machines in the United States was done at a time
when the advantages of self-fluxing sinter as blast-furnace burden had not been well
established. However, with the advancement of sinter technology, the use and produc-
tion of self-fluxing sinters became the rule rather than the exception. Lime additions
required in the production of self-fluxing sinters created increased dust problems for
the dust-collecting systems, with the result that additional electrostatic precipitator
capacity was required( 17), or use was made of other types of equipm~nt that were not
as vulnerable to such changes in operating procedures.
Blast-Furnace Applications of
Electrostatic Precipitators
The use of electrostatic precipitators for cleaning blast-furnace gas has come
about because of the requirements for cleaner gas for the hot-blast stoves. Trends to-
ward the use of higher blast temperatures required the use of checker bricks with
smaller holes, which in turn dictated the requirement for cleaner blast-furnace gas to
prevent plugging of the holes. Plugging would present major problems in efficient oper-
ation of the blast stoves. In all recorded installations, electrostatic precipitators have
been added to the existing emission-control systems on blast furnaces.
Operating problems have not been a point of major concern for application of
electrostatic precipitators to blast-furnace emissions. There are two possible reasons
for this trouble -free operation. First, the blast furnace is an almost continuous pro-
ducer of gas, except for the comparatively short intervals when the blowing rate of the
blast furnace is lowered during the time slag is flushed or iron is cast. Second, a
high percentage of the particulate emissions are removed by the wet-scrubbing systems
that had previously been used to clean the gases. In addition to performing a significant
cleaning operation, the wet scrubbers serve to condition the gases prior to their entry
into the electrostatic precipitator.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-8
Open-Hearth Applications of
Electrostatic Precipitator s
Use of oxygen lancing in the refining of steel in open-hearth furnaces made the use
of some type of pollution-control equipment mandatory. Electrostatic precipitators were
used first in the treatment of this emission problem. They were a rather logical choice
because some existing open hearths made use of waste -heat boilers to recover heat and
to generate steam from the exhaust gases, so no additional conditioning equipment was
required.
Electrostatic precipitators can be applied to open hearths in various ways. The
first installations in 1953 were essentially a precipitator for each open hearth. Kaiser
Steel Corporation, Fontana, California, installed one precipitator for each of nine open
hearths (l8), and in the same year the U. S. Steel Corporation (at its Fairless Plant)
installed twin, parallel precipitators for each of nine open hearths(l9). In 1959, U. S.
Steel installed four electrostatic precipitators to treat emissions from 11 open
hearths(20), followed by Bethlehem Steel's (Sparrows Point) installation of six pre-
cipitators for seven furnaces(21), and Weirton's 1965 installation of one precipitator for
two open hearths (22). This sampling of installations appears to suggest indecision on
the part of the various companies as to what type of precipitator installation best suited
their particular open-hearth shops. This however, was not necessarily the case.
Availability of space was a major factor in many cases in determining whether to use
one precipitator per furnace or not. Most of the earlier installations did have sufficient
space available, and there were technological uncertainties about the suitability of the
precipitators. These factors encouraged the use of one precipitator per furnace. U. S.
Steel's installation of four precipitators to serve 11 furnaces was the first attempt at
manifolding the gas off takes from the furnaces into a common collecting main that in
turn channeled the gases into four precipitators.
Manifolding of the exhaust gases from 11 open hearths served two purposes:
(l) it provided a mixing of the waste gases so that the temperature would not exceed
600 F, even though two waste-heat boilers could be by-passed with the discharge of
waste gases at 1200 F into the collecting main, and (2) a diluting of the fume and dust
took place with the result that the gas entering the precipitator had a more uniform dust
loading. (20) Diluting and mixing of the dust was a particular advantage because the
open-hearth furnaces in the shop were at different stages of processing the heats at any
given time, and at any instant the generation of emissions was different from each fur-
nace. One of the most-recent precipitator installations (1968) for open hearths uses
manifolded gas collection. Inland Steel Company's seven-furnace open hearth shop at
Indiana Harbor, East Chicago, Indiana, has one large precipitator. (23) The new instal-
lation at Youngstown Sheet and Tube Company's, Campbell Works will make use of
essentially a one-to-one installation with six precipitators for seven furnaces. (24) The
first unit was placed into operation in October, 1968.
Two major problems that have faced steel companies and equipment manufacturers
in the installation of electrostatic precipitators for open hearths have been (1) the design
of the ducts used to carry the gases from the open hearths to the precipitators, and
(2) the design of the gas -distribution systems at the entrance to the precipitators. Even
though a great deal of theoretical knowledge is available on the design of ducts, the use
of transparent models is considered to be almost a necessity in the practical design of
ducting. This approach was used in the design of the precipitator system at Kaiser in
1953(18), at Bethlehem - Sparrows Point in 1961(21), and at Weirton in 1965(22).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-9
The major problem with re spect to actual efficiency of electrostatic precipitators
on open hearths is the open-hearth process itself. The problem sterns from the varia-
tions in fuel used during the open-hearth refining of steel, which in turn affects the
moisture content of the gases. (25,26) A dry-gas condition occurs shortly after the hot-
metal addition, and lasts for about 15 to 20 minutes. The low moisture content is caused
by a low fuel-firing rate, low use of atomizing stearn, and a low initial oxygen-lancing
rate. (26) The moisture content drops as low as 2 percent, which is a basic cause of
poor efficiency, because low moisture levels result in higher resistivities with accom-
panying higher power requirements to achieve collection of the fume. In some case s,
the situation may be corrected in two ways: (I) power input to the precipitator can be
increased, or (2) a stearn-injection system can be installed to supply the desired mois-
~u.re. Increased power may be ineffective, as in the case of lime which causes back
ionization.
Basic -Oxygen-Furnace Applications of
Electrostatic Precipitators
Electrostatic precipitators were not the first type of pollution-control equipment
installed on basic oxygen furnaces. The first system was a wet system that was com-
bined with a distintegrator at McLouth Steel Corporation in 1955. (25) The first electro-
static precipitators were placed into service in 1957 by Jones and Laughlin at their
Aliquippa plant, followed in 1958 by the Kaiser Steel Corporation at Fontana,
California. (26) A list of the basic oxygen steelmaking shops in the United States, with
their respective pollution-control equipment is given in Table VI-2. The compilation is
based on the latest available information pertaining to current installations, plus re-
ported plans for new installations.
Problems associated with applications of electrostatic precipitators to basic oxy-
gen furnaces are basically those of variability in gas flow and the moisture content and
temperature of the entering gases, (which are functions of the process) and of mainten-
ance. Gas-flow rates for the process must be determined on the basis of theoretical
calculations(27), or on data obtained from similar operations. Calculation of theoretical
gas volumes is quite straight-forward and can even be developed as a nomograph, as
shown in Figure VI-6, that was developed as part of this study. As illustrated by the
dashed line in Figure VI-6, the off-gas volume for a 220-ton BOF heat using 70 percent
hot metal at 4. 0 percent carbon, an exce s s air factor of 100 pe rcent, and a blowing time
of 20 minutes would be 53,000 dm. However, elimination of carbon from the hot metal
is not the only source of carbon monoxide, and it appears that reactions in the hot metal
and slag contribute additional gases. This is illustrated in Figure VI-7, which shows
gas evolution from three different plants as they compare to the theoretical maximum
values. (28)
A significant design problem that is encountered in the design of electrostatic pre-
cipitators for new basic oxygen furnace installations is the potential production rate of
the BOF. The existing state of technology may predict a certain rate of production, and
the electrostatic pollution-control system may be designed for a nominal increase in
capacity; but should the BOF technology develop (as is quite likely) so as to result in a
larger increase in production, the electrostatic pollution-control equipment may soon be
inadequate. Production increases as high as 20 percent can be realized, as shown by
a 150 -ton BOF plant in the Chicago area that is now producing 205 net tons per heat. (29)
Some of the increased productivity is undoubtedly due to increased oxygen-blowing rates.
What effect such increases may have on the amounts of particulates is not known, but it
can be assumed that there will be at least a proportionate increase in iron-oxide fume.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
 TABLE VI-2. BASIC OXYGEN FURNACE INSTALLATIONS AND ASSOCIATED AIR-POLLUTION CONTROL EQillPMENT  
       Annual Capacity,    
      Net Tons net tons   Electrostatic  
     Number Per Heat March 1969 Future Startup Date Precipitator High-Energy Wet 
 Alan Wood Steel Co. Conshohocken, Pa. 2 140 1, 250, 000  1968 X  
 Allegheny-Ludlum Steel Corp. Natrona, Pa. 2 80 500,000  1966  X 
m Armco Steel Corp. Ashland, Ky. 2 160 1,400,000  1963 X  
»   
-i  Middletown, Ohio 2 200  2,000,000 1969  X 
-i Bethlehem Steel Corp. Bethlehem, Pa. 2 250 2,500,000  1968 X
l'll  Burns Harbor, Ind. 2 250  1,800,000 1970  X 
r    
r  Lackawanna, N. Y. 3 290 4,700,000  1964 -66  X
l'll  Sparrows Point, Md. 2 200 2,500,000  1966  X 
3: CF & I Steel Corp. Pueblo, Colo. 2 120 1,100,000  1961 X
l'll Crucible Steel Corp. Midland, Pa. 2 90 1,250,000  1968  X 
3:   
0 Ford MotOr Co. Dearborn, Mich. 2 250 2,500,000  1964 X  
a! Granite City Steel Co. Granite City, Ill. 2 225 2,200,000  1967 X
» Inland Steel Co. East Chicago, Ind. 2 255 3,000,000  1966  X 
r     2 210  2.000,000 1973  X 
Z Interlake Steel Corp. Chicago, Ill. 2 75 730,000  1959 X  
(I) Jones & Laughlin Steel Corp. Aliquippa, Pa. 2 80 1,000,000  1957 X
::!     3 200 3,000,000  1968 X  
-i  Cleveland, Ohio 2 225 2,250,000  1961 X  <:
C            H
-i Kaiser Steel Corp. Fontana, Calif. 3 110 1,440,000  1958 X  I
I'll McLouth Steel Corp. Trenton, Mich. 2 110   1958 X  ......
    o
I     1 110 2,800,000  1960 X  
n     2 110   1969 X  
0 National Steel Corp.           
r           
C Great Lakes Steel Div. Ecourse, Mich. 2 300 3.500,000  1962 X  
3:     2 200  2,000,000 1970 X  
m       
C Weitton Steel Div. Weirton, W. Va. 2 325 3,400,000  1967  X 
(I) Republic Steel Corp. Buffalo, N. Y. 2 100  1,000,000 1970 X  
r  Cleveland, Ohio 2 240 2,400,000  1966 X
»  Gadsden, Alabama 2 190 1,500,000  1965 X  
m  Warren, Ohio 2 180 1,600,000  1965 X  
0    
:u United States Steel Corp. Braddock, Pa. 2 220  2,250,000 1972  X
»  Duquesne, Pa. 2 215 2,400,000  1963  X 
-i    
0  Gary, lnd iana 3 200 3,700.000  1965  X 
a!  Lorain, Ohio 2 220  2,250,000 1970  X
l'll  South Chicago, Ill. 3 150  3,000,000 1969  X 
(I)    
 Wheeling -Pittsburgh Steel Corp. Monessen, Pa. 2 200 1,500,000  1964 X  
  Steubenville, Ohio 2 250 2,000,000  1965  X 
 Wisc.onsin Steel Div.           
 International Harvester Co. South Chicago, Ill. 2 140 1,200,000  1964 X  
 youngstOwn Sheet & Tube Co. East Chicago, Ill. 2 265  2,400,000 1969 X  
      Total 57,320,000 18,700,000  23 15 

-------
VI-II
100
20,000 30,000 40,000 50,000 70,000
Theoretical Total Off-Gas to the Atmosphere, scfm
(One percent carbon monoxide)
100,000
10,000
FIGURE VI-b.
THEORETICAL TOTAL OFF-GAS VOLUME FROM BOF FURNACES
AS INFLUENCED BY HEA T SIZE, PERCENT HOT METAL, AND
EXCESS COMBUSTION AIR FOR A 4.0 PERCENT CARBON HOT
METAL AND 20-MINUTE BLOWING TIME
Notes:
(a) For other carbon contents in hot metal, multiply off-
gas volume by the ratio: new carbon content/4. O.
(b) For other blowing times, multiply off-gas volume by
the ratio: 20 minutes /new blowing time.
,
OJ
E
'"
g
Shop B

COz- maximum
thearetical fram
0z blawing rate --...... .
COz- "flow
(as measured)
Shop B
~
LL
U
(/)
Shop A

-<;o:COz - flow(as measured)
('"COz-maximum
- - .~ theoretical from ~
0z blowin grate u
(/)
Blowing Time
Blowing Time
Blowing Time
FIGURE VI-7.
THEORETICAL AND ACTUAL GAS RATES DURING BLOWING
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-12
Maintenance problems with electrostatic precipitators are generally associated
with the precipitator proper which includes rappers, vibrators, and insulators. The
hoods over the BOF are a necessary part of the collection system, and can result in
operating problems. The gap between the BOF and the hood is usually dictated by the
anticipated operating conditions and the anticipated buildup of a skull on the mouth of the
furnace. Excess buildup can restrict the flow of air required for combustion of the car-
bon monoxide, with the re sult that a significant amount of carbon monoxide may reach
the electrostatic precipitator with possible disasterous results. The explosion hazard
with an electrostatic precipitator is a reality, and not just an anticipated possibility, as
attested by an explosion in 1968 at the Monessen, Pa., plant of the Wheeling-Pittsburgh
Steel Corp. (30) All iron and steel production was stopped for 1 week, and only partially
re sumed for the second week while repairs were completed.
Electric-Furnace Applications of
Electrostatic Precipitators
Only one known installation of an electrostatic percipitator with an e1ectric-
furnace plant is in operation. This installation is the electric-furnace shop of the Jones
and Laughlin Steel Corporation, Cleveland, Ohio. The electrostatic precipitators are
considered to be operating satisfactorily. Bethlehem Steel Corporation installed elec-
trostatic precipitators on the electric furnace plant at Los Angeles, California in 1955.
They were replaced by bag houses in 1967.
Wet Scrubbers
Wet scrubbers of various types have been used in the integrated iron and steel
industry for many years. The first wet scrubbers were simple spray towers used to
clean blast-furnace gas. However, as cleaner gas became a requirement for firing
blast-furnace stoves to higher temperatures, other types of wet scrubbing were used.
The introduction of fixed-orifice, then variable-orifice, and finally venturi scrubbers
was a natural for blast-furnace operation. Advances in blast-furnace technology and
improvements in burden materials started to reach a point where improvements in the
flow of reducing gases up through the burden became a necessity. One method of ob-
taining such improvement was the use of blast-furnace top pressures that were above
atmospheric pressure. In essence, the blast furnace system became a pressurized
system. The installation of orifice, variable-orifice, or venturi scrubbers in the gas
system was a practical way of obtaining cleaner gas. The pressure required to achieve
the necessary cleaning action was already in the blast furnace, and little additional
auxiliary equipment was required.
High-energy scrubbers (i. e., those capable of pressure. drops of 30 inches of
water or higher) are used in steel-plant applications. High-energy scrubbers are used
for controlling emissions from sinter plants, open hearths, BOF's, as well as from
blast furnaces. In some operations where particulates in blast-furnace gas must be
lowered to 0.005 grain per cubic foot, electrostatic precipitators have been installed
in series with wet scrubbers to obtain final cleaning.
One of the principal advantages of high-energy wet scrubbers is their ability to
handle variations in gas volumes, while still maintaining the required operating effi-
ciency. This characteristic of venturi scrubbers is illustrated in Figure VI-8(31),

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
FIGURE VI-B.
t
+-
c:
~
Q)
a.
>-
o
c:
.~
o
-
-
w
c:
o
+-
o
Q)
o
U
VI-l3
Water /Gas Ratio,gal./IOOOff-
OPERATING.CONDITIONS FOR A VENTURI SCRUBBER
The effect of water rate at a constant throat velocity on the output dust loading of
a venturi scrubber handling blast-furnace gas is shown in Figure VI-9. (32)
2 4 6 8 10
Water Level, gallons/ 1000

EFFECT OF WATER RATE ON OUTPUT DUST LOADING
FOR A VENTURI SCRUBBER
FIGURE VI-9.
0.12
- 0.10
o
(/)
(/)
c:
'5 0.08
~
0'1
g' 0.06
"C
c
..3 0.04
+-
(/)
::J
o 0.02
o
o
14
Theories that can explain the various mechanisms of collection involved in the per-
formance of wet scrubbers and can serve as a basis for comparison are not completely
developed, as indicated from the following statement from a recently published manual
on the subject. (33)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-14
where it can be seen that for a given throat velocity of the gases, the efficiency is in-
creased by increasing the water/ gas ratio, or simply by pumping water at an increased
rate. However, the maximum and minimum values of water and/or gas rates for high-
energy scrubbers used in the iron and steel industry are unknown.
"The theories of the various mechanisms involved in wet collection have
not been completely developed; some, such as the electrostatic effect and
humidification are more presumed than understood, and the air cleaning
tasks to which wet collectors are applied rarely involve a simple, uniform}
nonreactive particulate dispersed in a simple carrier gas. The wet col-
lectors themselves are typically not single mechanism units but usually
function on the basis of several collection mechanisms. : This makes
clear-cut classification of equipment impossible and imposes difficulties
in selecting a collector for a given task without knowledge of their pre-
vious application. "
An empirical method has been developed to correlate scrubber efficiencies.
method is called "The Contacting-Power Concept", and is defined as follows (34):
This
"In the gas -liquid contacting process, power is dissipated in fluid turbulence
(in gas and liquid phases) and, ultimately, as heat; it is this power, expressed
as power per unit of volumetric gas flow rate, that is the criterion of scrubber
efficiency, and it has been designated 'contacting power'. "
It should be noted that the power referred to excludes power consumed by motors, fric-
tion, and mechanical losses. A conclusion of the initial investigators which led to the
development of the contacting-power concept is as follows(35):
"When compared at the same gas power consumption, all scrubbers give
substantially the same degree of collection of a given dispersed dust,
regardless of the mechanism involved and regardless of whether the
pressure drop is obtained by high gas flow rates or high water flow
rates. The collection efficiency increases as the pressure drop in-
creases, the increase being especially rapid for pressure drops over
lO-in. water. II
Mathematically the contacting-power concept can be expressed as follows:
Nt = exP;
where
Nt = the number of transfer un:its
1
Nt = 2. 3 log 1 -E
E = collection efficiency
PT = total contacting power.
\
While the calculation of Nt is quite simple, similar calculations for PT re9,uire data on
the gas-flow rate, water-flow rate, and water feed pressure. . When the required data
is plotted on log-log coordinates, a straight line correlation is evident as shown in
Figures VI-lO and VI-ll(34). The coefficient ex is the value of the intercept where PT is

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
 4
+- 
Z 3
VI 
:= 
c 2
:::>
98
95
90 "E
a>
o
L-
80 ~
70 :>.
o
c
60 .~
o
....
50 W
L-
a>
....
VI
C
o
L-
t- 1
.... 0.9
00.8
a; 0.7
E 0.6
~ 05
c::. .
0.4, 2 3 4 5 6 78910

Contacting Power, hp/(IOOO cuft min}
40
.r IGURE VI-IO. PERFORMANCE OF A
VENTURI SCRUBBER ON A METAL-
T TJRGICAL FUME
VI-IS
7
6
99.9
99.5
99 +-
c
98 ~
L-
a>
a.
95
>.
o
c
90 .~
;0=
....
w
80
70
FIGURE VI-II. PERFORMANCE OF A
VENTURI SCRUBBER ON OPEN -HEAR TH
FUME
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
z+- 5
~ 4
c
:::>
L-
.&
VI
c
o
.=
....
o
3
2
Oxyge n.
in use
L-
a>
.Q
E
:J
Z
No oxygen used
one, and 'Y is the slope of the line. The coefficients are essentially functions of the dust
and/ or fume, and are not influenced by the way the contacting power is applied to the
scrubber. Some industrial data were used in analyzing the contacting power concept;
however, full-scale industrial tests in steel plants were not made.
I .
2 3 4 5 6 7 8 9 10
Contacting Power, hp/(IOOO cu ft min}
A characteristic of wet scrubbers over other types of emission-control equipment
is that gases such as carbon dioxide and sulfur dioxide will dissolve in the water. (36)
This can be particularly advantageous where sulfur dioxide is in concentrations that
would exceed the concentrations permitted by regulations. However, this plus in per-
formance is offset by the severe corrosion problems resulting from the formation of the
respective acids, and by the effects the acidified water has on the operation of the water-
treatment facilities. If the maintenance problems involved are disregarded, the wet
scrubber may be an effective means of reducing the emission of sulfur dioxide to the
atmosphere. It has been stated that there is a significant removal of sulfur dioxide by
wet scrubbers(37), but also that there are no data on the input and output of sulfur dioxide
for wet scrubbers(38).
Application of Wet Scrubbers in
Sinter Plants
Of about 40 sinter plants at various steel plants, only two are known to have wet-
scrubber installations. One uses a venturi scrubber to treat emissions from the wind-
box of the machine(39), and the other uses flooded-disk scrubbers at the discharge end
of the sintering machine(40). The output loading of the flooded-disk scrubbers is report-
edly O. a I grain per cubic foot.
Early application of wet scrubbers to sinter plants resulted in operating problems
which were traced to erosion and imbalance of the fan blades on the exhaust-system
blowers. These are the blowers that provide the draft through the sinter bed required

-------
VI -16
to ignite the fuel. Erosion of the blade s has been a problem even with dry pollution-
control systems. However, the imbalance occurring in the fan blades is aggravated in
sinter plants having wet pollution-control systems because the dust that is carried over
to the fan is moist, and has a greater tendency to accumulate on the blades. The im-
balance problem has been reported to be caused by uneven buildup of the dust and also by
the breaking off of large amounts of built-up dust which places severe unbalanced loads
on the blades. Both situations have caused severe vibrations and sometimes major
breakdowns in the blowers. This situation is minimized by constant preventive main-
tenance to remove the dust buildup.
Application of Wet Scrubbers to
Blast Furnaces
During 1967, about 170 blast furnaces were producing hot metal, and of this num-
ber about 51 were using wet scrubbers as the principal method of cleaning blast-furnace
gas. Of these, about 33 were high-energy scrubbers. In addition, about 35 blast fur-
naces were equipped with high-energy scrubbers as cleaning units preceeding electro-
static precipitators which serve as final cleaning units.
The first high-energy scrubbers were installed in 1955. They were simple, fixed-
orifice plates installed in the gas lines, with water introduced into the main at some dis-
tance upstream from the orifice plate. These scrubbers operated at pressure drops of
30 to 50 inches of water3 with resulting output loadings varying from 0.01 to 0.03 grain
per cubic foot. (41,42,4 ) The orifice scrubber, however, had the major disadvantage
that it could not handle the variations in gas flow, and consequently could not meet the
required emission limits during certain phases of blast-furnace operation when the
velocity of the gases coming from the blast furnace was lowered.
The need for high-energy wet scrubbers that could handle variations in gas flow
from a blast furnace led to the development of variable-orifice scrubbers that attempted
to cope with the variability of gas flow by adjusting the size of the orifice opening. The
performances of a variable-orifice scrubber and a fixed-orifice scrubber are illustrated
in Figure VI-12. (44) The effect of a reduction in wind rate on the two types of equip-
ment is quite evident.
During the same period that the orifice and variable-orifice scrubbers were re-
ceiving attention from blast-furnace technologists, the venturi scrubber was also under
investigation as a possible means of cleaning blast-furnace gases. The first application
of venturi scrubbers to blast-furnace gas was reported in 1955(45), with other installa-
tions reported in 1956(46) and 1960(47). Technical data relating the output dust loading
to the water rate and pressure drop of a blast-furnace venturi scrubber are shown in
Figure VI-13(45). The relationship between clean-gas dust loading and pressure drop
is shown in Figure VI-14(47), that illustrates the low output dust loadings that can be
achieved with this equipment. This high performance, however, can be achieved only
if the blast furnace is operating at a high-enough top pressure to provide the required
pressure drop. This level of high top pressure has not been achieved, as evidenced
by the large number of blast furnaces operating with electrostatic precipitators as the
final gas -cleaning unit. (See Table VI-l that shows 108 blast furnaces operating with
electrostatic precipitators. )
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-I?
0.12
't 0.10
If)
"'-
~
~ 0.08
0\
c
-g 0.06
.3
1;; 0.04
:J
o
0.02
o
o
20 40 60
Reduction in Wind Rate, percent
FIGURE VI-12. EFFECTIVENESS OF GAS CLEANING BY A FIXED-
ORIFICE SCRUBBER AND A VARIABLE-ORIFICE
SCRUBBER WHEN GAS -FLOW RATE IS VARIED
.~ ..
0.12
0\1f)-
CO,+-
15 (.!):J 0.08
ouu
o """'-
...J ~ If) 0 04
1;; o.S .
:JQ)o
OU ~ 0
+-=' ~ 8
80\
1: '0 6
f--
o'+-
-:J 4
OU
.~o
0-
Q: ';;; 2
~ C
Q)o
0= 0
~ g 0
1969
operation
10 20 30 40 . 50
Pressure Drop Across Venturi Section, inches water
FIGURE VI-B. OPERATING CHARACTERISTICS OF A BLAST-
FURNACE VENTURI SCRUBBER
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-I8
100

0(\1
:J:
.E
a.
o
...
CI
ell
...
::J
VI
VI
ell
...
a..
10
...
::J
-
c:
ell
>
1
0.001 0.01 0.1 1.0
Clean Gas Dust Loading, grains per standard cubic foot
FIGURE VI-14.
CALIBRATION CURVE FOR A BLAST-FURNACE
VENTURI SCRUBBER
Application of Wet Scrubbers to
Open-HeariliFurnaces
Wet washing of open-hearili gases was first considered to be economically ex-
pedient for shops that were to be operated only during high peak demands for steel,
under which conditions the low capital cost for ilie wet system was considered an ad-
vantage. (48) However, some open-hearili shops iliat were considered to be fairly new
found that wet scrubbers were economically attractive when ilie shop either had no waste
heat boilers or ilie existing boilers could not lower the gas temperatures enough to war-
rant the installation of electrostatic precipitators or bag houses. (49, 50) The first open
hearili installation was made in 1959(48) at U. S. Steel's Edgar Thomson Works, and
others subsequently followed. Output gas loadings of 0.01 to 0.05 grain per cubic foot
have been reported for ilie installations, again wiili the cleaning efficiency relating di-
rectly to the pressure drop of the scrubber. The relationship between clean-gas dust
loadings and pre ssure drop for an operating open hearth installation is illustrated in
Figure VI-15(48J. Oxygen lancing was used during the refining period. The figure is
representative of open-hearth practice with low oxygen-blowing rates and is not neces-
sarily representative of present-day practice using higher oxygen-blowing rates.
Application of Wet Scrubbers to
Basic Oxygen Furnaces
Wet scrubbers were first installed in 1954 on a basic oxygen furnace at the
Hamilton Plant, Dominion Foundries and Steel Ltd., Ontario, Canada. The first in-
stallations in the United States was made at the Duquesne Plant of the U. S. Steel Cor-
poration in 1963. The number of high-energy wet scrubbers installed over the years as
compared to the number of electrostatic precipitators is shown in Figure VI-16. One of
the principal reasons for selecting electrostatic precipitators over high-energy wet
scrubbers is the existence of a water-treatment problem in a plant. Problems can in-
clude inadequate water-treatment facilities, or the lack of sufficient water. Even though
there are 15 BOF plants in the United States with high-energy scrubbers as the primary
pollution-control equipment, there is a notable lack of published information concerning
their operation. It appears that problems associated with wet scrubbers in other appli-
cations apply al so to BOF installations.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
    VI-19      
........  0,10         
en          
c  0,08         
C+-         
~ 0          
010 0,06         
 -         
010     Ore and lime boil 
c.- 0.04    
.- .D    and working period 
'0 :J    
C 0     
0           
..J'O          
 ~          
+- C 0.02         
en '0          
:J C          
0 C  Charging, melt down      
 +-       
 en  and hot meta I       
+- ~        
:J Q) 0.01         
,2-a.         
:J  0,008         
0          
  26 28 30 32 34   36 38 40
    Pressure Drop, inches of water 
FIGURE VI-IS.
RELATIONSHIP BETWEEN CLEAN-GAS DUST LOADING AND
PRESSURE DROP FOR A WET SCRUBBER ON AN OPEN-
HEAR TH FURNACE (OXYGEN LANCING USED DURING THE
REFINING PERIOD)
30
25
5
Electrostatic
prec i pi tators
en
(; 20
+-
o
~
o
U
15
o
1956
,,"
--
/-
/
/
/
,,"
,,"

1""- High energy wet
1 scrubbers
/
"
,,"
,.
-
o
~
Q)
.D
E 10
:J
Z
1958
1960
1962
1964 1966
Year
1968
1970
1972
1974
FIGURE VI-16. INSTALLATION OF ELECTROSTATIC PRECIPITATORS AND
HIGH-ENERGY SCRUBBERS FOR AIR-POLLUTION
CONTROL AT BOF STEELMAKING PLANTS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-20
Problems associated with improper performance of wet scrubbers can be attrib-
uted to a lack of materials to withstand the abrasive and corrosive nature of the dust-
laden water, or a misapplication of construction materials. Design of wet scrubbers
does not appear to be a contributing factor to deficiencies in scrubber performance.
Application of Wet Scrubbers to
Electric Furnaces
High-energy scrubbers for electric steelmaking furnaces are known to be used in
only two shops; both owned by the Armco Steel Corporation. The first installation was
made at the Butler, Pa., plant on a 70-ton furnace in 19S8(S1), and the second the
Armco plant in Houston, Texas. The Butler plant (currently in the process of expan-
sion) will soon include three ISO-ton furnaces(S2) that probably will be serviced with
high-energy wet scrubbers. When expansion plans are completed in 1969, the Houston
plant will be equipped with high-energy wet scrubbers operating at a pressure drop of
60 inches. (S3) No other electric furnace shops are known to be using high-energy wet
scrubbers.
Fabric Filters
Fabric-filter installations (or baghouses as they are more commonly called) have
their biggest steel-industry application in the control of emissions from electric
furnaces -- a total of 29 installations. There also are three known applications at sinter
plants, and two applications on scarfing machines. Fabric filters are used on BOF's
in Europe(S4), but no installations of this type have been made in the United States.
The performance of fabric filters has been well developed on the basis of the-
0retical principles, and numerous descriptions are available in the published literature.
A reduction of the theoretical concepts to a simplified mathematical form results in the
following equation(SS):
r Lt I v2
H' =
7000
where
H' = filter resistance increase in inches of water
r = specific resistance of the dust (determined experimentally) in inches of
water gage per pound of dust per square foot of filter cloth area per foot
per minute of filtering velocity.
L = input dust loading to the filter in grains per cubic foot of air
t' = operational time in minutes
v = filtration velocity in feet per minute.
This equation leads to the conclusion that resistance of the filter to flow is dire ctly pro-
portional to (1) the square of the superficial face velocity, (2) the weight of the dust
collected on the fabric, and (3) the time of operation since the last cleaning. (S4)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-2l
A more recent method of evaluating fabric-filter performance is based on '~filter-
drag". The development of the method is based on similarity to electrical circuitry and
to heat transfer. The equation for filter-drag evaluation is as follows:
v = ~ ' or in electrical analogy I = ~
where
v = velocity in feet per minute
6P = pressure drop in inches of water
S = filter drag in inches of water per foot per minute.
.Lne factor V is similar to the air-to-cloth ratio of more common usage. Filter drag is
independent of the size of the unit, the filter ratio, the type of dust, the style of fabric,
and all other specifics of application. (56) The method has been used to evaluate the per-
formance of fabric materials prior to full-scale installation in a bag house.
Application of Fabric Filters
to Sinter Plants
Typical of the use of fabric filters in sintering plants is the collection of dust
generated at the discharge and screening locations at the Bethlehem Steel Corporation's
plant at Bethlehem, Pennsylvania(57), and the U. S, Steel Corporation's plant at Gary,
Indiana(l7). Both units are sectionalized, with sufficient capacity to permit shutting
down sections for maintenance without affecting the cleaning efficiency of the units.
Pertinent statistics for the two units are given in Table VI-3.
Application of Fabric Filters to
Open-Hearth Furnaces
An experimental baghouse was installed at the Lackawanna plant of the Bethlehem
Steel Corporation in February, 1960. (58) The development work was done with an
oxygen-lanced open-hearth furnace, and results were satisfactory. A production bag-
house was placed into operation by Bethlehem Steel at the No.2 Open-Hearth Shop at
Sparrows Point in 1963. (59) The only change that had to be made in melting practice
was the elimination of fluorspar as a flux. The baghouse had 10 hoppers of 80 bags
each. The bags were 11.5 inches in diameter and 34 feet long. Design capacity was
145,000 cfm at 500 F. With an input dust loading of 1.4 grains per cubic foot, the out-
put loading was 0.0007 grain per cubic foot, for an efficiency of 99.95 percent. The
open-hearth furnace and baghouse were operational in January, 1966, but the results of
subsequent operations are unknown.
Application of Fabric Filters to
Electric Furnaces
Fabric filters have been successfully applied to the control of emissions from
electric furnaces ranging up to 100 and l50-net-tons capacity, and for multiple-furnace
shops as well as one -furnace shops. (60, 61, 62, 63, 64) However, because electric
BATTELLE MEMORIAL INSTfTUTE - COLUMBUS LABORATORIES

-------
VI-22
TABLE VI-3.
DESIGN AND OPERATING DATA FOR SINTER-PLANT
FABRIC FILTERS ON SINTER STRAND DISCHARGE
Design or Operating Variable
U. S. Steel Corp.,
Gary, Indiana
Bethlehem Steel Corp. ,
"Bethlehem, Pa.
Volume of Air, cubic feet per minute:
172,000 at 255 F
240,000 at 350 F
Suction, inches of water:
12
n. a.
Pressure Drop Across Bags,
inches of water'
Total Bag s, numbe r
4 n. a.
10 16
88 72
880 1152
Hoppers in Unit, number
Bags per Hopper, number
Bag size:
Diamete r, inche s .
Length, feet'
11. 5 12
32.2 28
Fiberglas s Fiberglass
17 36
12 -20 n. a.
2. 17 2.29
Bag Type
Bag Life, months
Bag Permeability, cfm per foot2 of cloth
Air-to-Cloth Ratio (Normal), cfm per
ft2 of cloth
Air-to-Cloth Ratio
(One Compartment Cleaning),
cfm per ft2 of cloth
Theoretical Design Efficiency, percent
2.41 2.44
1 75 to 3 0 0 200 to 500
99+ 99+
Air Temperature, F
n. a. - not available.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS. LABORATORIES

-------
VI-23
furnaces have design and operating characteristics that can vary from furnace to
furnace and shop to shop, the development of pollution-control facilities is not as
straight-forward as for other steelmaking operations. One of the major problems in
the design of electric-furnace pollution-control systems is the design of methods that
will completely capture the fumes. The majority of electric-arc furnaces are top
charged, which means the roof is removed during charging. As a re suIt, the emissions
generated during the charging operation are difficult to capture and contain. Capture of
fumes by hoods and by direct-extraction techniques during melting and refining have not
completely solved the problems of collection and containment, and extraction of fumes
through the plant roof has been developed to control emissions through the entire
plant. (64, 64)
Cyclone Dust Collectors
One other type of dust cleaning equipment that has found extensive use in the con-
trol of emissions from ironmaking and steelmaking operations is the cyclone dust
collector. The cyclone separates particles from the gas by means of a centrifugal
force exerted on the particles in a vortex flow that drives the particles toward the wall
of the body of the collector. The particles at the wall move toward the discharge open-
ing of the cyclone as a result of an axial component of the vortex flow, aided in the
case of cyclones used in the iron and steel industry by the force of gravity. The magni-
tude of the radial forces acting on the particles depends on the nature of the vortex flow
in the different sections of the cyclone. Radial gas velocities tend to act as counter-
acting forces, and tend to offset the separating forces generated in the cyclone.
Cyclones are suitable for collecting medium and coarse dusts, but are not suited
for very fine dusts or metallurgical fumes. Their advantages are that there are no
moving parts, there is a wide choice of construction materials, and maintenance costs
are low. Power costs can be quite high because a high degree of efficiency is required.
Cyclones find their principal application as pre cleaners for other types of
emission-control equipment. Some of the applications apparently are deliberate,
others occurred in a transition from lower to higher colle ction efficiencies.
while
Cyclone s find application in pelletizing plants and in lime stone plants for the col-
lection of the large-size dust generated in certain of the operations. As pre cleaners,
they are used in series with dust catchers, wet scrubbers, and electrostatic pre-
cipitators for cleaning blast-furnace gas; as precleaners for electrostatic precipitators
handling the dust and gas from a sinter -plant wind box; and as part of the series of
equipment used in open-hearth-furnace -emission control. No information has been
located in the literature or during this study concerning the efficiencies of cyclones in
the various iron and steel plant operations, or concerning any particular problems that
have arisen with their use.
Cost of Applied Control Equipment
As part of the companion study for the Division of Economic Effects Research of
NAPCA, estimates were made of the capital costs and operating costs of the principal
types of emission-control equipment used in the integrated iron and steel industry. The
types of equipment considered were electrostatic precipitators, high~energy wet

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-24
scrubbers, and fabric filters. Cost estimates were based on established industrial
estimating techniques, with cost information supplied by certain steel companies and
equipment manufacturers.
For the purposes of the estimates made, a control system was considered to be
made up of all the items of equipment and their auxiliaries which are used solely for
the general abatement of atmospheric pollution in the neighborhood of the steel works.
Typically this will include a collecting hood or gas -collecting pipe at the furnace,
ductwork, spray cooler, dust collector, fan and motor, and stack. Included also will
be structural steel, foundations, control instruments, insulation, piping, water treat-
ment, and electric power supply facilities for the entire gas-cleaning system. Excluded
are those equipment items which, while they may contribute to the functioning of
pollution-control equipment, would be used for process or economic reasons even if
there were no pollution-control requirements. For example, an open hearth furnace is
usually connected to a stack whose primary task is to supply draft for causing gas flow
through the heat regenerative stoves. In this study, the cost of the original stack is at-
tributed to the steelmaking proce ss because the furnace cannot operate without it. How-
ever, any increase in stack height, or other modifications nece ssary when air pollution
equipment is installed, is charged as a cost of pollution equipment.
The cost of land occupied by pollution-control equipment has not been included. It
is recognized that such land has a real value, but a satisfactory method for estimating
it has not been established. Costs associated with preparation of the site, start-up
operations, and working capital also are not included. Certain portions of a control sys-
tem occupy or utilize parts of steel plant buildings and, therefore, might be charged
with a share of general building costs. This item has not been estimated here. In cal-
culating operating costs, no attempt was made to allocate a portion of general overhead
to control systems.
Capital and operating costs in the estimates are based upon collectors whose
efficiency can be relied upon to produce an outlet dust loading of O. 05 grain/scf of gas.
The estimates of capital costs include facilities for loading the collected dust or
sludge into trucks for transportation elsewhere. No by-product values have been
assigned. Central engineering costs, overheads, and fees were based upon a standard
sliding scale generally used by contract engineers. Labor cost was calculated at the
nominal value of $5. OO/man hour, including all welfare and fringe costs.
It is believed that the general precision of the capital cost estimates is such that
most specific plant situations will fall within :1:15 percent of the estimated values. In
more statistical terminology, it might be suggested that the standard deviation is about
:1:10 to 12 percent. It is to be expected that any specific plant location which presents
unusual cost problems associated with layout, structure, power supply, etc. might fall
outside these limits. In such cases, a detailed plant design and estimate must be pre-
pared if accurate capital cost data are required. The accuracy of operating cost
values is influenced by many factors which may vary considerably from one ,company
to another. The selection of control equipment should not be based upon small differ-
ences in estimates of operating cost.
Operating cost estimates include costs for the following items:
(1) Electric Power

(2) Maintenance

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-25
(3) Depreciation
(4) Capital Charges
(5) Operating Labor.
Electric energy is calculated at a standardized rate of l/:. per kilowatt hour.
Maintenance is taken at a nominal cost of 4 percent of the total investment. Deprecia-
tion is calculated on a straight-line method using total investment with an expected life
of ten years. Other studies of depreciation have suggested longer service life times,
but these are considered to be greater than average plant experience will confirm.
Advancing technology and rising standards give importance to the factor of technical
obsole s cence.
Capital charges are taken at 10 percent per annum. It is believed that this will
be reasonable in the light of rising interest rates and local taxes.
Annual calculations are based on 330 operating days per year, 24 hours per day.
This gives a total of 7,920 operating hours per year.
The estimated capital and operating costs are summarized in the form of graphs
in Figures VI-17 through VI-24. For further details on these estimates, reference
should be made to the companion report (A Cost Analyse s of Air Pollution Controls in
the Integrated Iron and Steel Industry); especially to Appendix C of that report.
Cost-Effectiveness of Applied Systems
The three principal types of equipment used to control iron oxide fumes from the
various process segments in the integrated iron and steel industry can meet current
air-quality requirements. This holds if the equipment has been correctly designed and
constructed and is properly maintained. The particular equipment selected must per-
form its designed function at a cost that will not create an excessive increase in the
cost of the finished steel products. Many factors are involved in determining the
actual cost of air-pollution control to any given steel plant. Factors such as the local
level of allowable emissions, power costs, availability of water, ease of dust disposal,
and possible reclamation or marketing of the recovered iron oxide dust enter into the
final determination of costs. Estimated operating costs for air -pollution-control
equipment on open-hearth furnaces, basic oxygen furnaces, and electric furnaces are
illustrated in Figure VI-25. The wide range possible in these costs is quite evident.
As regulations governing allowable emis sions become stricter, the costs for con-
trol increase, but not as a simple direct relationship. This can be illustrated in the
case of electrostatic precipitators as shown in Figure VI-26. (66) As collection ef-
ficiencies increase beyond 95 percent, the cost increases at a rapid rate. The effect of
collection efficiency on the cost of electrostatic precipitators only, as well as installed
cost, is shown in Figure VI-27 for installations in the iron and steel industry. Costs are
1968 costs. The one type of equipment whose costs are not drastically affected by in-
creased requirements for efficiency is the fabric filter or baghouse.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
6.0
5.0
(/)
L-
o
o
'U
-
o 4.0
(/)
c:
o
E
~
+-
(/)
o
U
E 3.0
a.
o
U
'U
Q)

o
+-
(/)
c:
H
'U 2.0
Q)
+-
o
E
+-
(/)
w
1.0
0.0
20
VI-26
Legend

Electrostatic Precipitator

- - High- Energy Wet Scrubber
-- --- Fabric Filter
/
/
/
BOF /
./
I "
,
II I

/ II
/

Open hearth / Open h~arth

/ Electric~ // Open hearth'
/ Furnace#,/ /
// .h II' '"
~ ///"///

~ ,a "';//
----:: ~,...~./ Electric furnace
~ ,..."" '"
....... ,...
- - ---.....
---
.i
30
40 50
100 150 200 300 400 500
Design Capacity, ACFM
1,000
ESTIMATED INSTALLED CAPITAL COSTS (1968) OF AIR-
POLLUTION-CONTROL EQUIPMENT AS RELATED TO
DIFFERENT STEEL-MAKING PROCESSES, ON THE BASIS
OF DESIGNED ACTUAL CUBIC FEET PER MINUTE (AT
TEMPERATURE) OF GAS FLOW RATE
FIGURE VI-I?
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
  2.6
  2.4
01  2.2
»  
'-i :2: 
-i 
m LL 2.0
u
r  
C 
en .~ 
:! c 
~ 1.2
-i Q)
C Q. 
-i O
m '0 1.0
I ~
C
o 'C 
.0 
-------
  6.0
 ~ 
 c 
 0 
 +- 
 0 
 U 
 Q) 5.0
 ....
 a.
m Q) 
» 0 
-i "0 
-i c 
111 o
r I/) 
r Q) 
111 2' 
~ 0 4.0
111 ~~ 
~ ulJ... 
0 OU 
;u .4.:: 0 JLU" ON-CON" ~O.. ~QUIPMENT ;'0 t S"!: !:LMAKIl' G :> tOC !:SS!:S

-------
m
»
....
....
III
r
r
III

~
III
~
o
:u
»
r
en
~
o
"0
"0
-
o
en
.~ 1.5
.-
E
+-
en
8
o
+-
'0. 1.0
8.
z
(II
:f
....
c
....
III
I
o
o
r
c
~.
m
c
(II
r
»
m
o
:u
»
....
o
!
III
(II
"0
.!!
o
+-
en
c::
-
"0 0.5
Q)
+-
o.
E
+-
en
W
2.0
/

,

/
/
.. l ~
, l
/ - ,'/
2:/'
, ~.
/h
, 0
//
..~
~
Number of
furnaces
\3
Fabric filter, 275 F
I
I
I
I
I
I
I
I
I 1
1 1
1 II
1 II
I II
1
31 //
1 I'
I //
/ //
// 2 //
/ ///
/ /r/
/ ",/
'" '" ",'" I
",'" , "
" ...."""
-" ....' ....
- ,""" "",,'"
-- -.--::,..-"""
---
<
......
I
N
~
High-enery
wet scrubber, 180 F
Electrostatic
precipitator, 500 F,
100,000
4OOpoo
00 '
4O/XX)
100,000
200/XX) ?l:)Qpoo
- 40,000
4OpOO
100,000 200.000 '&YJ,OOO .
Designed Capacity. ACFM
2OOpoo
FIGURE VI-20. ESTIMATED INSTALLED CAPITAL COSTS OF AIR-POLLUTION-CONTROL EQUIPMENT
INSTALLED ON ELECTRIC-ARC STEELMAKING FURNACES. CONTROL EQUIPMENT
DESIGNED TO HANDLE EMISSIONS FROM ANY ONE FURNACE ATONE TIME

-------
5.0
m
>
-i
-i
III
r
r
III
3:
III
3:
o
!
>
r
If)
~
.Q
(5
"0
b 4.0
If)
c
.Q
E
00-
:3 3.0
u
C
00-
"5-
c
u
"0
Q) 2.0
c
00-
If)
C
......
Z
III
j
-i
C
-i
III
I
o
o
r
c
3:
m
c
III
r
>
m
o
;n
>
-i
o
;n
III
III
"0
~
c
E
+: 1.0
If)
w
0.0
3OpOO
I

,

/3

.

/
2#
D'
,~
If
yJ'
/'

'Ii'
<:
......
I
VJ
o
High - energy
wet scrubber, 180 F
Electrostatic
precipitator, 500 F
Fabric filter, 275 F
~umber of
furnaces ~

3
./
I
31
I
I
I'
I.~
2/.7
~'/
/~'/
/J':"/
~~~
,,,7
I.".~"
."~.,,
~."
,'"
--
100,000
500,000
500pOO
20,000
Designed Capacity. ACFM
FIGURE VI-21. ESTIMATED INSTALLED CAPITAL COSTS OF AIR-POLLUTiON-CONTROL EQUIPMENT
INSTALLED ON OPEN-HEARTH FURNACES. CONTROL EQUIPMENT DESIGNED TO
HANDLE EMISSIONS FROM FURNACES OPERATING AT THE SAME TIME

-------
VI
L-
.2
o
"0
'0 1.6
VI
c
o
- 1.4
E
~ 1.2
u
o
~ 1.0
o
U
"0
Q) 0.8
o
.-
VI
c
H 0.6
"0
Q)
.-
o
E 0.4
.-
VI
W
0.0
. 20,000
2.0
1.8
0.2
VI-31
Legend
Electrostatic Precipitator
--- Low- Energy Wet Scrubber
----.;.;;. Fabric Filter .
--- - Cyc lone
~
Sinter-plant
wind box
at 325 F
Sinter-plant

material handling
at 135 F ~
~...
--...--... ---. ,/
.--- - -- --... "
-------- _Pellet ~

-- plant ... ~ ...'" Pellet plant
----...
.
50,000
100,000 . 200,000
Designed Capacity, ACFM
400,000
800,000
FIGURE VI-22. ESTIMATED INSTALLED CAPITAL CQSTS OF AIR-POLLUTION-
CONTROL EQUIPMENT USED IN SINTER AND PELLET PLANTS
BATTELLE MEMORIAL .INSTITUTE - COLUMBUS LABORATORIES'

-------
1_.. -.---------
2.0
~
lL..
U
cd:
'-
Q)
a.
~ 1.5
o
<5
'U
~
-
If)
o
U
01
C
VI-32
\
Sinter plant
material handling
at 135 F
Legend

Electrostatic Precipitater
-- - Low- Energy Wet Scrubber
------ Fabric Filter
--- - Cyclone
..
\
,

\

\

"
''"'

,
"-
',,-
~
o
:J
C
c

-------
VI-33
450
-
I/)
o
U

o
-
'0.
o
u
'0
~
Capital cost
350
Electrostatic
precipitator,
100F
400
o
-
~ I/) 300
......""
o
'0-
Q)-
_0
0'0 250
.~o
-0
1/)0
w-
High - energy
wet scrubber,
100 F
~
-
I/)
o
U
01
.S
-
o
....
~ 0.80
0:E
°lL
::Ju
~
-------
VI-34
1.80
c
S2 - 1.60
+- c
Q) .2
c+-
L- .2
~ ~ 1.40
L-
VI Q.
L- Q)
.S!'O
o '0 1.20
"t:! c
~ 0
+- VI
VI Q)
8 ~ 1.00
010
c.c.
:;:: 0
0-
~.E 0.80
g'~
"t:! 0
~ ~ 0.60
0'0
E :J
:;::0
VI c
W ~ 0.40
0.20
o
1.0 1.5
Annual Production of Raw Steel.
2.0
2.5
3.0
millions of net tons
FIGURE VI-25. RANGE OF ESTIMATED OPERATING COSTS FOR AIR-POLLUTION-CONTROL EQUIPMENT PER NET TON OF RAW
STEEL - OPEN-HEARTH FURNACES, BOFS, AND ELECTRIC FURNACES (TWO-FURNACE OPERATIONS)
3
        IJ
        V
       J 
       V 
     ~ /'  
   0..,    
    V ~   
   ~    
 - -      
-        
FIGURE VI-26. ELECTROSTATIC-PRECIPITATOR COSTS
AS AFFECTED BY COLLECTION EFFICIENCY
E
.....
o
o
o
o
o
o
LO
01
E .EO: 2
..... '0
o Q)
Q)
~ ~
Q.Q)
~ ~
o E
0.2
'0 ~
":01
~~
U"t:!
C
o
.c.
c
o
"t:!
Q)
VI
o
m
80 82 84 86 88 90 92 94 96 98 100
Efficiency, percent
BATTEL.L.E MEMORIAL. INSTITUTE - COL.UMBUS L.ABORATORIES

-------
15.0
10.0
9.0
8.0
VI 7.0
~
o 6.0
(5
"0
- 5.0
o
:g 4.0
o
E 3.0
-
VI
o
U
2.0
1.0'
80.0
VI-35
~~
/
/
Steel plant precipitators,.
installed cost
1968 costs
/
/
/
/
I
/
/
/
/
/ Cost trend.
I,precipitators
/ only, 1968 costs
II (Ref. 66)
/
90.0
95.0 98.0 99.0
Collection Eff i ciency , percent
FIGURE VI-27. INSTALLED COST FOR STEEL PLANT
ELECTROSTATIC PRECIPITATORS AS
AFFECTED BY COLLECTION EFFICIENCY
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI - 3 6
Effect of Efficiency Specifications Greater Than Current Legal
Requirements on the Cost of Air-Pollution Control Equipment
The remainder of Section VI is adapted from Appendix C in the com-
panion report, "Final Economic Report on A Cost Analysis of Air -Pollution
Controls in the Integrated Iron and Steel Industry", dated May 15, 1969,
and is a result of work done by the Swindell-Dressler Company, Division of
Pullman Incorporated, the subcontractor on the study.
Because of the dearth of experimental and empirical data on the relationship be-
tween collection efficiency and the cost of air-pollution control equipment, resort was
taken to estimation of this relationship from theoretical considerations. The estimates
based on theoretical considerations then were evaluated against the limited amount of
empirical data that could be obtained.
In many localities, current legal requirements specify a permissible particulate
emission at the stack of not more than 0.05 grain per dry standard cubic foot of gas (or
equivalent) exhausted to the atmosphere.. Some facilities have met this requirement (or
even exceeded it) with even fine, submicron sized steelm,aking dust by using high-
efficiency filters, scrubbers, and precipitators. Manufacturers have been able to guar-
antee this performance with their equipment in a variety of applications. Also it is noted
that blast-furnace gas has been cleaned to as low as 0.005 grain/ DSCF':' when nece s sary
for reuse of the gas in high-energy burners and fine checkerwork of the blast stoves
(although this is a coarser dust than from steelmaking).
This index "0.05" is not necessarily an ultimate measure of the effluent quality
that can be obtained. It came into use in the early 19601s, on the basis that an open
hearth furnace stack plume containing fume at such a concentration had an 'Iacceptable"
appearance in many steelmaking areas. The value "0.05" correlated approximately
with the maximum efficiency of electrostatic precipitator s normally offered by manu-
facturers at that time for collecting this fume. However, the rapid growth in the use of
oxygen lancing of steelmaking furnace s has led to larger quantitie s of finer fume in their
waste gases today. .
A stack plume cleaned to this level is not an invisible plume. The very fine steel-
making fume escaping at the stack causes a larger degree of scatter of transmitted light
than the larger particles previously encountered(13), and thus may be visible even in
low concentrations. Yet, visibility of an exhaust plume persists as a means of checking
collector performance, because it involves a simple comparison of "equivalent opacity"
of the plume against the Ringelmann Smoke Chart.
Local code limitations based on Ringelmann opacity judgments may find a concen-
tration of 0.05 grain/DSCF of steelmaking fume unsatisfactory. Where local codes are
based on a schedule of allowable fume emission weight per ton throughput of processed
material, the permissible fume rate customarily decreases for larger production equip-
ment, so that above 30 to 40 tons of proce ss weight per hour, the O. 05 level of control
may not be adequate.
°DSCF = dry standard cubic foot.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORAT,ORIES

-------
VI-37
Thus, the widespread use of 0.05 grain/DSCF. as a general limiting level for
emissions led to its choice as a basis for calculating the size and cost of collectors for
each process in the tabulations in Appendix D of this report. But, in recognition of the
use of more restrictive enforcement methods in some steelmaking areas, and because of
the trend in promulgating air-quality criteria which may suppress the emission sources
in an area to an increasing degree, the following indications are drawn of the difference
in cost for fume-collecting systems capable of an efficiency beyond the currently prac-
ticed or currently attainable level.
Performance Equations
The performance equations of gas cleaners, as currently understood and applied to
select the size and operating parameters for a particular cleaning application have this
in common - they are of the form:
Y) = 1 - e-F(x)
where
Y) = collection efficiency
or l-Y) = penetration, dust loss, or outlet concentration as a
fraction of the inlet concentration to the gas cleaner.
It corresponds to some figure like O. OS, for example;
0.05 (grains/ DSCF)
l-Y) == inlet conc. (grains/ DSCF)
In (l-Y)) = -F(x)
F(x) is a function of the
of the colle ctor .
size and operating parameters
Fabric Filters. The equation for the performance of fabric (bag) filters, has been
shown to be of the form(67):
S Dr
Y) = l-e - 0-
D
D' f . f~
D is the target efficiency and is a unctlon 0 Vf'
where
D = fiber diameter
g = gravitational constant
v = velocity of gas at filter face
- Q - flow rate of gas
- A - area normal to flow = face velocity
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-38
f = settling velocity of particle, as from Stoke's Law.
S = total projected area of all fibers in the filter
o cross section of filter bed '
both normal to the gas flow.
The relationship between the target efficiency ~' and the function W is shown in
Figure VI-Z8.
100
,  
,  
,  
\\ 
 \ 
 \~ ~Spherical obstacles
  01/0 (Spheres) = zero -
  " at Og/Vf = 24
  ,~ 0'/0 (Cylinders) = zero -
Cylindrical!>::""I... at Og/Vf = 16 \-
obstacles ,. ~...:::.. ~ J ,
  I -
0'/0
00 1
2 3 4 5
16
Q9..
Vf
FIGURE VI-Z8.
. ,
RELATIONSHIP BETWEEN TARGET EFFICIENCY(~)
AND ~ FOR FABRIC FILTERS .
Electrostatic Precipitators. The equation for the performance of an electro-
static precipitator takes the form of the Deutsch equation(lS, 68):
y) = l-e-A'fl/Q
where
AI = collecting surface area
~ Z(k-l)] rEZ
fl = C + (k+Z) J 67TJ.l = drift velocity
k = dielectric constant of particles
E = ele ctric field strength
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-39
r = particle radius
f.l = gas viscosity at temperature.
Wet Scrubbers.
A correlation established for wet scrubbers is

-ofp p )"1 -a(P )\Y
7) = l-e \: G + L = l-e \: T
as fullows(69):
whe re
P G = contacting power of gas stream
=O.157FS
FS = pressure loss across scrubber, in. water, exclusive of
loss due only to velocity changes or friction losses across
dry portions of the equipment .
PL = contacting power of liquid stream
qL
= O. 583PF Q
PF = liquid feed pressure, psig
qL = liquid feed rate, g.p.m.
PT = PG + PL = HP/lOOO CFM, based on Q
Q = actual gas flow at the scrubber, CFM
a, "I = constants for a particular dust, related to particle size
and size distribution.
Theoretical Factors Affecting Performance
To increase the efficiency of a collector, whose performance is describable by
this logarithmic-de cay-type function, it is necessary to increase F(X). The variable
flow and equipment parameters comprising F(X) for a particular dust are respectively:
Bag Filter So, ~ = So, ~A
AIE2 2L W n E2
Precipitator ~ = A V f.l
=
2L W n E2 2LE2
n b W V f.l = bVf.l
where
W = collector surface span normal to flow
L = collector surface length in direction of flow
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI -40
n = number of collecting ducts
b = separation of collecting surface s
Scrubber FS and PF(qL/QI)
where
qL/Q' is the liquid/gas ratio (gal/lOOO CF).
Effects of Propertie s of Particle s
(l) Increasing f increase s YJbag filter
f = 4g r2 p
l8f-.L
where p is the density of particle.
So larger, denser particles are collected more easily.
(2) Increasing r increases YJprecipitator' Again, larger
particles are more easily attracted to the collector. In.
creasing the dielectric constant of the particulate, and
decreasing re sistivity by pre -conditioning via temperature
and humidity (or S02 addition) increases YJprecipitator.
(3) Increasing a or 'Y increase s YJscrubber, as can be
established(69). Both increase with particle size.
Effects of Geometry of Dust Collector
(1) Increasing filter thickne ss or mat density, or decreasing
air/ cloth ratio by using larger bag surfaces will increase

YJbag filter'
(2) Increasing precipitator length in the flow direction, or de-
creasing plate spacing or tube diameter (within limits of
electrical stability) will increase YJprecipitator' Because
the dust loading decreases in the flow direction, it is
possible to achieve an economy by successive stages of
precipitation, each optimized electrically for maximum
efficiency at the respective loading it will see, rather than
simply extending the fir st- stage field.
(3) Decreasing the throat area of a scrubber increase s its
pressure drop and increases YJscrubber. This can be done
by variable geometric arrangement or by increased water
rate.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-41
Effects of Utility Parameters
(1) A partially blinded filter will be more efficient, but at a
cost of higher pressure drop and higher fan horsepower.
(2) Increasing electric field strength increases YJprecipitator
within the limits imposed by the geometry of the collector
and dust properties with respect to sparking. This limit
can be approached more closely with safety if automatic
controls are used to regulate the discharge. Energy use
ri se s.
(3) Venturi Scrubber. Increasing water usage or delivery
pressure in a scrubber increases 7Jscrubber. Increased
gas-pressure drop gives improved efficiency at the cost of
higher fan horsepower.
Effects of Flow

(1) Even though an increase in face velocity (~) gives a higher

theoretical efficiency in the inertial effect range, the effect
is reversed in dealing with small particles «1J..L). For a
filter with a fixed pressure drop and fixed cleaning routine,
the dust buildup will dominate, so that if increased loading
blinds the filter, causing spillage and less net cleaning, then
the following holds. Decreasing the quantity of gas treated
or using a larger filter for lower face velocity increase s
7Jbag filter'
(2) Decreasing the amount of gas treated by lowering precipitator
velocity and increasing residence time increases 7Jprecipitator
if distribution of the gas is maintained uniform between the
plates.
(3) Increasing the quantity of gas treated or increasing throat
velocity increases 7Jscrubber by increasing pressure 10ss
across the constriction, with an increase in PG and fan
horsepower.
Effect of Temperature on Viscosity
(1) Increasing temperature increases J..L gas, decreases
7Jbag filter, and decreases YJprecipitator'
(2) Increasing temperature increases the quantity of gas handled,
again lowe ring the se efficiencie s.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI -42
(3) Besides altering flow and settling or drift velocity, tempera-
ture also endangers the bags, structures, and mechanisms of
the collectors. Filters and dry precipitators must have an
inlet temperature above the water vapor (and sulfuric acid)
dew point to avoid corrosion and dust caking on the collector,
and causing dust handling problems in disposal conveyors.
(4) Temperature affects the scrubber mainly in increasing the gas
flow and increasing the saturation-water requirement.
Control System Cost Change s
It is a property of decay functions of the aforementioned type that, at high effi-
ciency, an increasingly large change in the exponent is required for a small incremental
increase in efficiency.
Electrostatic Precipitators. For example, for an electrostatic precipitator, it
has been stated that the precipitator unit size increases with respect to efficiency
change s as follows(70):
Overall Efficiency
for a Particular Dust
90 percent
99 percent
99. 9 pe rcent
Outlet Loading With 5.0
Grains/ DSCF Input Loading

0.5
O. 05
O. 005
Size of Precipitator
Box and Unit Cost
X
2X
3X
This tabulation excludes ductwork, water sprays, hood with its cooling auxiliaries,
and stack; but include s the precipitator and its electrical components. The fan and
motor size and cost, for a IX increase in precipitator size would be affected by an in-
crement corre sponding to an increased static pre s sure (S. P. ) of about 1-1/2 inche s of
water (the loss through Box X), with the volume remaining unchanged.
For a precipitator increment of X:
Horsepower increment = S. P. + 1. 5 x H P
S.P. ..
Fan-pressure increment = S. P. + 1.5 x S P
S. P. ..
for the total system fan.
Fan volume unchanged.
The effect of increasing the size of an electrostatic precipitator on operating effi-
ciency has also been reported in the literature for an electrostatic precipitator collecting
open-hearth emissions. (71) The relationship developed between the collection efficiency
and size of the precipitator (as shown by the square feet of collecting surface) is shown
in Figure VI-29. (71) The results of the study have shown that removing the dust from
315,000 cubic feet per minute of open-hearth waste gas required 58,300 square feet of
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-43
collecting surface area for an efficiency of 95 percent. An increase in the collecting
surface area to 96,500 square feet (an increase of 66 percent) resulted in an increase in
efficiency of only 4. 3 percent to 99.3 percent.
100
90
+- 80
c
~ 70
~
~ 60
~
~ 50
c
Q)
'0 40
:E 30
w
20
10

00
 ~.~   .- .~    
-.- ~  .- ..- -- ~-- -
   ...... I 
   ./ ' !    
  /   I    I
    I   
 )/   !    :
 /    I    !
 V    I    I
    1   
J     g!    ~r
I     rtI   
    m,   
I     1    I
I     !    !
0.5
1.0
1.66
o
o
8
8
o
(\J
~ ~ ~ ~ ~

Collecting Surface, sq ft
o
8
(X)
~
8
o
o
o
FIGURE VI-29.
RELATIONSHIP OF ELECTROSTATIC PRECIPITATOR
COLLECTING SURFACE TO COLLECTION EFFI-
CIENCY FOR OPEN-HEARTH EMISSIONS
The above variation in size corre sponds to Deutsch f sLaw:
(1-1')) =e
AfI'
--
Q
for a particulate of homogeneous size, shape, density, and composition.
inlet loading -R = 1-~. h R th t1 t 1 di
1') = inlet loading 1. L. ' were = e ou e oa ng.
R t t -constant2 x length
= cons an 1 x e
log R = constant3 + constant4 x length
for a given process and precipitator.
Case 1 Illustrative of Deutschf sLaw
R:
5 grains
SCFD
--
Box
X
--
0.5
--
Box
X
--
O. 05
--
Box
X
--
0.005
1')(percent): 90 90 90
net 1')(percent): 90 99 99.9
net size: X 2X 3X
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-44
However, real particulate varies in size, density, and susceptibility to charging
(depending on surface and compositional variables), so that the least collectable particles
remain after each treatment, lowering the efficiency of subsequent treatments. (72)
Case 2 below illustrates this with arbitrary efficiencies:
Case 2. Illustrative of Deutsch's Law
 5 -+ Box Box B'ox' Box Box Box  
R: -+ 0.5 -+ -+ 0.1 -+ -+ O. 03 -+ -+ 0.012 -+ -+ 0.006 -+ -+ 0.005
  X X X X  X 5/12X 
11 (percent): 90 80 70 60 50  40  
net 11 (percent): 90 98 99.4 99.76 99.88 99.9  
net size: X 2X 3X 4X 5X 5-5/12X  
A body of blast-furnace data(73) for a number of operating furnaces at various
precipitator loadings yields the following progre ssion, which shows this trend,
net y)(percent):
90
95
98
99
99.5
99.9
99.99
net size:
0.55X
0.86X
1.4X
2X
2.65X
4.5X
7.5X
Cost data from precipitator manufacturers indicate a close correspondence to
Case I in variation of cost (= constant x length) with efficiency. Guarantees are made on
efficiency rather than outlet loading, because the precipitator is not adequately adjust-
able for cleaning a higher inlet dust concentration to the same outlet level (say 0.05).
In fact, the higher loading may reach a point where spark-over occurs; so automatic
electrical controls are used to maintain the highest collection efficiency just short of
spark-over. (Large loading differences require design selection of plate spacing and
voltage optimized for the loading and dust properties of the individual process effluent).
The maximum guarantee is presently about 99. 5 percent, although higher efficiencies
(around 99.7 percent) can be reached.
The successive lowering of efficiency found with addition of identical precipita-
tion units can be compensated for. Because each successive unit sees a lower dust
loading, plates can be spaced more closely, and voltage optimized in each succeeding
section, while avoiding spark-over. Still, each type of dust must be te sted to deter-
mine its collectability as a function of precipitator length.
Tables VI-4 through VI-9 show some estimated cost changes for processes cleaned
by electrostatic precipitation to various outlet dust concentrations. The variation is
based on the Deutsch Law. Capital cost changes include:
Materials: precipitator + fraction of electrical.
Labor: corresponding to each of above at standard factors.
Engineering:
scaled fraction materials plus labor.
Annual operating cost changes include 0.24 (Capital change) + fraction of electric
power for precipitator and horsepower increments. Only small variations were noted
for capacity of the cleaner, so that only the central size of cleaners is included for
processes estimated in Appendix D at 0.05 grain per SCFD.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS. LABORATORIES

-------
TABLE VI-4.
VI -45
ESTIMATED COST DIFFERENCES FOR A SINTER
PLANT ELECTROSTATIC PRECIPITATOR AS
AFFECTED BY THE OUTPUT DUST LOADING
FROM THE WINDBOX
Plant Capacity
Gas Volume
Input Dust Loading:
6,000 net tons per day
630,000 ACFM at 325 F
0.8 grain per SCF(a)
Outlet
Loading
(R)
Capital Cost
Difference (6 KR)
KO. OS' percent
Annual Ope rating
Cost Difference (6 CR)

CO. OS' percent
Annual Direct Operating
Cost Difference (6 DR)

DO. OS' percent
O. 125
O. 050
- 29
o
O. 020
+ 29
R
log10 (R )
0.05
log10 (2.5)
- 23
o
+ 23
=
-6K
R
18%
=
-6C
R
15%
- 17. 5
O. 0
+ 17.5
=
-6D
R
10%
(a) 4 grains per SCF effluent precleaned by 80 percent efficient recovery cyclones.
ESTIMATED COST DIFFERENCES FOR A SINTER
PLANT ELECTROSTATIC PRECIPITATOR AS
AFFECTED BY THE OUTPUT DUST LOADING
FROM MATERIAL HANDLING
TABLE VI-5.
Plant Capacity
Gas Volume
Input Dust Loading:
6, 000 net tons per day
250,000 ACFM at 135 F
1. 0 grain per SCF
Outlet
Loading
(R)
Capital Cost
Difference (6 KR)
KO. OS' percent
Annual Operating
Cost Difference (6 CR)

CO. OS' percent
Annual Direct Operating
Cost Difference (6 DR)
DO. OS' percent
O. 125
0.050
- 18
o
0.020
+ 18
R
log10 (R )
0.05
log10 (2.5)
- 15
o
+ 15
=
-6K
R
18%
=
-6C
R
15%
- 10
o
+ 10
=
-.6D
R
10%
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-46
TABLE VI-6. ESTIMATED COST DIFFERENCES FOR A BOF
ELECTROSTATIC PRECIPITATOR AS AF-
FECTED BY OUTPUT DUST LOADING
Furnace Size
Gas Volume
Input Dust Loading: .
200 net tons
785, 000 ACFM at 500 F
4. a grains per SCF
Outlet
Loading
(R)
Capital Cost
Difference (6KR)

KO. 05' percent
Annual Operating
Cost Difference (6 CR)

CO. 05' percent
Annual Direct Operating
Cost Difference (6 DR)
DO. 05' percent
O. 125
O. 050
O. 020
- 9
a
- 10
a
- 11
a
+ 9
+ 10
+ 11
R
log10(R )
O. 05
log10 (2.5)
-6K
R
9%
-6C
R
10%
=
-6D
R
11%
=
=
.TABLE VI-7. ESTIMATED COST DIFFERENCES FOR AN OPEN
HEARTH ELECTROSTATIC PRECIPITATOR AS
, AFFECTED BY OUTPUT. DUST LOADING
Furnace Size
Gas Volume
Input Dust Loading:
200 net tons
85, 000 ACFM at 500 F
5. a grains per SCF
Outlet
Loading
(R)
Capital Cost
Difference (6KR)
KO. 05' percent
Annual Operating
Cost Difference (6 CR)
CO. 05' percent
Annual Direct Operating
Cost Difference (6 DR)
DO. 05' percent
O. 125
O. 050
0.020
- 10
a
- 9
a
- 7
a
+ 10
+ 9
+ 7
R
log 10 (R )
0.05
log 1 a (2. 5)
-6C
R
9%
=
-6D
R
7%
-6K
R
10%
=
=
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-47
TABLE VI-8. ESTIMATED COST DIFFERENCES FOR AN ELECTRIC
FURNACE ELECTROSTATIC PRECIPITATOR AS
AFFECTED BY OUTPUT DUST LOADING
Furnace Size 150 net tons(a)
Gas Volume: 185,000 ACFM at 500 F(b)
Input Dust Loading: 3. 0 grains per SCF
Outle t
Loading
(R)
Capital Cost
Difference (.6KR)

KO.05' percent
Annual Operating
Cost Difference (.6 CR)

CO.05' percent
Annual Direct Operating
Cost Difference (.6 DR)
DO.05' percent
b
O. 125
- 10
- 10
- 9
0.050 0  0   0
0.020 + 10  + 10   + 9
  (~) -.6KR -.6 CR -.6 DR 
  logl 0 R.05 
   - . = = 
  log10 (2.5) 10% 10% 9% 
(a) Two-furnace system.
(b) Assumes humidification of process fume is capable of maintaining particle resistivity in satisfactory collection
range.
TABLE VI-9. .ESTIMATED COST DIFFERENCES FOR A SCARFING
MACHINE ELECTROSTATIC PRECIPITATOR AS
AFFECTED BY OUTPUT DUST LOADING

Gas Volume : 100,000 ACFM at 100 F
Input Dust Loading: 1. 0 grain per SCF
Outlet
Loading
(R)
Capital Cost
Difference (.6KR)

KO.05' percent
Annual Operating
Cost Difference (.6 CR)

CO.05' percent
, Annual Direct Operating
. Cost Difference (.6 DR)

DO.05' percent
O. 125
- 19
- 18
- 17
0.050
o
o
o
0.020
+ 19
+ 18
+ 17
. (~)
log10 R.05
log10 (2.5)
-.6KR -.6CR
= 19% = 18% =
-.6 DR
17%
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI -48
Wet Scrubbers. In the case of the ventUri scrubber, it has been stated that the
following relate s capital cost to efficiency(70):
 Overall Efficiency for a Given   
 Dust at Inlet Load of   
 Inlet Grain/ DSCF, percent  Outlet Loading Capital
1.0 3.8 5.0 10 Grains/ DSCF Cost
90 97.4 98 99 0.10 X
96.2 99 99.24 99.62 0.038 1.43X
The operating expense s vary similarly for a venturi scrubber as efficiency is in-
creased. This is shown in Figure VI-30(48) for an open-hearth application where a de-
crease in outlet loading from O. 1 to 0.01 grain/SCFD results in more than doubling the
annual operating cost of the fan. For a given size adjustable venturi, the increased
efficiency requires an increase in available horsepower to the fan and selection of a
higher pressure fan. Operating power consumption increases directly with the pressure
drop.
g 0.10
~ '+- 0.08
C'lu
.~:.o 0.06
"0 :J
C u
.5"E 0.04
- c
1/)"0
:J c:
o.E
+- I/) 0.02
:J ...
Q.CV
-Q.
:J'-
o ~ 0.01
~ 0.008
C'I 26
Ore and lime boil
and working period
Charging, melt down
and hot meta I
30 32 34 36
Pressure Drop, inches of water
. . I
45 55 65
Fan Operating Cost (1960),
thousands of doHars per year
38
40
28
.'
35
75
FIGURE VI-30.
RELATIONSHIP OF OUTPUT DUST LOADING TO PRESSURE
DROP AND FAN OPERATING COST FOR A VENTURI
SCRUBBER OPERATING ON AN OPEN-HEARTH
FURNACE(48)
The above venturi-cleaned open-hearth application involves oxygen lancing during
the periods noted on the upper curve. Dust loading was low (0.82 to 0.87 grain/SCFD
during oxygen periods, and 0.35 to 0.45 grain/ SCFD during the charging, melting and
hot-metal addition periods). When these data are corrected to a typical peak of 5 grains/
SCFD loading for today's oxygen-lanced furnaces, it yields the following correlation:
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-49
(grains )
R, outlet loading SCFD
.6P, venturi pressure drop (in. w.)
O. 125
O. 05
O. 02
34.7
41. 0
48.2
6PR -0.178
.6 PO. 05 = (0.~5)
However, the contacting power concept(69) has been applied for the regression line
of a plot of non1anced open hearth gas-cleaning efficiency versus pressure dror at various
operation conditions(46). This give s for a peak 5 grains/ SCFD inlet loading:
R (grains/ SCFD)

O. 125
O. 05
O. 02
.6 P (in. w.)

44
78
136
.6 PR ( R )-0. 62
.6 PO. 05 = 0.05
The numerically higher exponent seems more in line with results from other steelmaking
fume.
Blast-furnace data give the following(46):
(grains )
R, outlet loading SCFD
.6P, venturi pressure drop (in. w.)
0.125
0.05
0.02
15.8
23.0
33.2
.6PR = (~)-O. 403
.6PO.05 0.05
Data for an electric furnace making 20 percent ferrosilicon show(46):
.6PR ( R )-1.53
.6PO.05 = 0.05
For the typical scrap-charged electric-arc furna.ce, wet scrubbing applications are
spar se, and data are not available for a scrubbing power -efficiency correlation.
Venturi gas cleaning data on the basic oxygen furnace have been developed as
follows(74):
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-SO
(grains )
R, outlet loading SCFD
~P, venturi pressure drop (in. w.)
O. l2S
O. OS
O. 02
27
41
60
gi ving
~PR
~ PO. OS
= (~)-O. 417
O.OS
Data from a pilot-size conventional venturi scrubber applied to clean scarfing-
machine effluent have been published as follows(7S):
(grains )
R, outlet loading SCFD
;~P, venturi pressure drop (in. w.)
0.12S
O. OS
O. 02
34
60
108
~PR = (~)-0.63l
~PO.OS O.OS
Tables VI-IO through VI-12 give some estimated cost differences for processes
cleaned by wet scrubbers of the high-energy types. The variation is based on the pre-
ceding scrubber -application data. Capital cost change s include:
Materials:
Fan and motor + fraction of electrical. The venturi
itself is assumed adjustable and of sufficient strength
for the higher pressure difference across its walls.
Water rate s are unchanged. .
Labor:
Corresponds to each of above at standard factors.
Engineering: Scaled fraction of materials plus labor.
Annual operating cost changes include 0.24 (capital change) plus electric power for
hor sepower increments.
An empirical relationship is indicated as follows:
~KR ~CR ~DR ~PR-~PO. OS ~PO. OS r ~PR ]
S.S%=9'%=1.2% = 60-41 = 19 LLPO.OS-l
~ -0.417]
= ~~ L(o,~s) -1
~K ~C ~D [ ~P ] ~ -0.417]
R - R - R - R -1 - ~ -1
l1.8%-19.4%-2S.9%- 6PO.OS - (O.OS)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-51
TABLE VI-I0. ESTIMATED COST DIFFERENCES FOR A BOF WET SCRUBBER
AS AFFECTED BY OUTPUT DUST LOADING
Furnace Size 200 net tons{a)
Gas Volume: 440,000 ACFM at 180 F
Input Dust Loading: 4.0 grains per SCF
Outlet
Loading
(R)
Venturi Pre s sure
Drop (.6 P),
inche s of water
Capital Cost
Difference (.6KR)
KO. OS, percent
Annual Operating
Cost Difference
(.6 CR) CO. OS,
percent
Annual Direct
Operating Cost
Difference
(.6 DR) DO. OS,
percent
0.125
0.050
27.5
41. 0
- 4.0
O. 0
- 6
o
- 8
o
O. 020
60.0
+ 5.5
+ 9
+ 12
(a) One furnace system.
TABLE VI-II. ESTIMATED COST DIFFERENCES FOR AN OPEN-HEARTH WET
SCRUBBER AS AFFECTED BY OUTPUT DUST LOADING
Furnace Size 200 net tons{a)
Gas Volume: 90,000 ACFM at 180 F
Input Dust Loading: 5.0 grains per SCF
    Annual Direct
   Annual Operating Operating Cost
Outlet Venturi Pre s sure Capital Cost Cost Difference Difference
Loading Drop (.6P), Difference (.6 KR) (.6 CR) CO. OS, (.6DR) DO.05,
(R) inches of water KO. OS, percent percent percent
0.20 32.6 -11.0 - 15.5 - 19.5
0.10 50.2 - 6.5 - 9.0 - 12.0
O. 05 78.0 0.0 0.0 0.0
.6KR .6CR .6DR .6PR-.6PO.05 .6PO.05 [.6PR l

-6.5%= -9% = -12% = 50.2-78 = -27.8 - .6P -J
0.05
It - O. 62 ~
= -2~~8 ~O. ~5) - J
.6KR .6CR .6DR f.6PR l ft R -0.62 ]
18.20//25.1% = 33.5% =[.6 PO. 05 -~ = ~O. 05) -1
(a) One-furnace system.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI - 5 2
TABLE VI-12.
ESTIMATED COST DIFFERENCES FOR A SCARFING MACHINE
WET SCRUBBER AS AFFECTED BY OUTPUT DUST
LOADING
Gas Volume
Input Dust Loading: .
100,000 ACFM at 100 F
1. 0 grain per SCF
    Annual Direct
   Annual Operating Operating Cost
Outlet Venturi Pressure Capital Cost Cost Difference Difference
Loading Drop (.6P) Difference (.6KR) (.6CR) CO. 05, (6DR) DO. 05,
(R) inches of water KO.05, percent percent percent
0.083 44.3 - 9 - 17 - 23
0.050 61. 0 0 0 0
0.030 81. 5 +11 + 21 + 28
.6KR .6CR .6 DR
11 % = 21% = 28% =
.6PR-.6PO.05
81. 5-61
=
.6P 0.05
20. 5
[.6P ]
R 1
.6PO.05-
61
= 20.5
[(OHOS) -0.631 -I]
.6K .6C .6D
R R R
32.6%= 62.5% = 83.4% =
~ .6P ]
R -1
.6P 0.05
=
[(O.HOS) -0.631 -I]
Doubling the venturi pre ssure drop would cause a 25.9 percent increase in direct
operating costs. Venturi loss of 41 in. w. is 85 percent of system loss, which accounts
for 40 percent of total horsepower (including an unchanged water pumping and treatment
system) in this case. .Power cost is about 72 percent of direct operating costs.
. Fabric Filtration. For an acceptable constant dust penetration through a fabric
filter, the face velocity or air volume -to-cloth area ratio must decrease with decreasing
particle size and density or with increased inlet loading. For a lower allowable penetra-
tion, the face velocity would similarly decrease. Thus, a more difficult or more thori
thorough cleaning job would involve increased cost to provide more filter surface area.
This is exemplified in the extreme case of a reverse jet-cleaned filter where face ve-
locities are the highest encountered. Nomograms that can be used to estimate the size
of reverse jet filters are shown in Figure VI-31 and VI-32. (76)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI - 5 3
  40
 ~ 50
 c:
 e 60
 v
A i 
0 70
 . 80
 o
 ;; 
 .D 
 ~ 90
 "" 
 co 
 aP 
  100
 30
 2'
 20
 U
C 
 10
..: 
U 
..... 
'" 
Q 
C) ,
z 
a 
« 
9 
~ 
on 
:> 
Q 
The following nO/11ograph is presented
as a convenient means of selecting
Filter Ratio for preliminal'y deter-
mination of the size Aerotul'n Dust
Collector' that will best satisfy the
needs of your installation.

In many instances the nomograph will
pl'Uvide cletenllination of the optimum
Filter Ratio. Because of the great val'i-
ety of possible service conditions and
the effect of the charactcl"istics of
specific dusts. final detenninat ions of
Filtel' Ratio will be made by Buffalo
Forge Company. This (Jl'OCedul'e pro-
vides the gl'catest assunmce of COITect
and economic selection of equipment
fOl' your installation.
IIOW TO I;SJ~
In OJ'der to select Filt.er Ratio, three
conditions pertaining to your specific
dust collection job ar~ needed. They arc:

a. The approximate percentage, by
weight, of dust particles 10 microns
or smalle~.

b. Dust content of the air enterin
-------
36.000
:".000
32.000
30.000
28.000
26.000
24.000
22.000
t
~ 20.000

~
U
~ 18.000
..
u
e-
o'
~
~
~
~
..
~
~
..
~
'"
; 16.000
ii:
14.000
12.000
10.000
8.000
6.000
2.000
°
FIGURE VI-32.
VI-54
FilTER SIZE
200 16.8
400 16.10
16.11
32.8
600 16.14
16.16

16.18
800 32.\0
16.20
~8.8
t
ci
!2
1.000 32.12  
1.200 )2.16  
 48.10 
   64-8'
 32-14  
1.400   
  48.12 
. 32-18  
1.600   64-10
 32-20  
  48.14 
1.800   
IIOW TO LOSE
This chart provides a conven-
ient and accurate means for
selecting the applicable size or
sizes of Aeroturn Dust Co 1-
Il'ctors whl'n Filter Ratio and
required Air Cleaning Capacity
are known.
1) Draw a line from the re-
quired Capacity through the
applicable Filter Ratio to in-
tersect the Filter Area scale.

2) From this pOint of inter-
section, draw a ho,.jzontal line
through blocks designating
Filter Size selections for de-
sired Capacity.
64.16
3) If horizontal line passes
through more than one Filter
Size, first size intersected will
be most economical. Subse-
quent selections will be less
economical.
.~
..
::
~
ii:
2.000
64.12
48-16
64-18
4) For capacities larger than
shown: Use % the required ca-
pacity in the above procedure.
Filter Size thus selected must
be doubled for full capacity.
2.200
48.1,8 6.-14
2.400
2.600
48.20
64.20
NOMOGRAM FOR ESTIMATING THE SIZE OF A
REVERSE-JET FILTER
2.800
3.000
3,200
3.400
3.600
A reverse -jet cleaned fabric filter calculated from the nomograms shown in
Figures VI-31 and VI-32 for dust having 70 percent of its weight less than 10 microns
in size and a specific gravity above 2.0 yields the following information on sizing of the
filter:
Efficiency Required for
0.05 Grain/ DSCF
Outlet, percent
Inlet Dust Loading,
grains/ DSCF

5
10
20
25
Filter Ratio,
cfm/ sq it

16.4
13.7
9.6
7.6
99
99.5
99. 75
99.8
Q/ A Filter
Area
X
1.2X
1.7X
2. IX
The effective filtering body is the dust cake layer on the bags.
this time seem amenable to treatment which will improve efficiency.
when new, and to a lesser extent when cleaned, holds little dust cake
This doe s not at
However, the bag,
so that the fabric,
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-55
with its small, dust-laden fibers is the basic filter until the filter cake layer reforms.
As the small fibers break in service, the bag loses filtration capability. Additionally,
the lower flow resistance of a cleaned bag passes a greater volume of air at reduced
cleaning efficiency than when it is dust-coated; but at a higher velocity, which improves
the collectability of larger particles and worsens the diffusiona.l efficiency dominating
small-particle colle ction.
An adequately designed baghouse will have a bag-cleaning cycle suited to the inlet
dust loading from the proce ss to which it is applied. This cycle is often automatically
adjustable, so that the filter maintains the same average (time-wise) efficiency with
variations in inlet dust loading and gas volume. The bag-cleaning period will begin when
the collected dust cause s the pre ssure drop through the filter to reach a set-point
pressure.
In addition, the fabric weave and material are chosen with the special character
of the process effluent in mind (such as particle size distribution). Economic factors
(bag life and initial cost differences) also enter this choice, but increased efficiency can
be achieved only by choosing from a group of fabrics which will give cleaning to the pro-
jected required level. Present practice usually gives efficiencies of 99 percent +, and
bag filters frequently give the highest efficiencies of the applicable cleaning devices con-
sidered for a process. Therefore, this selective optimization does not offer much
potential except as re search may reveal new materials and weave s.
As a case in point, a process having a generally large particulate may be ade-
quately cleaned by a certain bag to 0.05 grain/ SCFD. If lower outlet loading is required,
a suitable bag which gives similar results on a process with finer effluent may be sub-
stituted. The overall cost mayor may not be larger. The choices are presently limited
by limited test results on filtration properties of fabrics, and state-of-the-art in fabric
technology with regard to dust abrasion, flexural durability, and chemical and tempera-
ture re sistance. Electrostatic interactions of various fabrics with dust particle s may
prove to be significant.
No clear correlation has been advanced relating efficiency to operating parameters.
However:
A higher pressure drop may be expected to increase filtration action at
the cost of additional power, but the trend of such variation is not
known.
A higher filter -face velocity (higher air/ cloth ratio) theoretically
yields a higher efficiency of collection for particle s large enough to
be governed by inertial laws , but the se are ordinarily cleaned to
nearly 100 percent efficiency, so the filter size is governed by
loading. The small particles which escape collection migrate under
diffusional impulses, and efficiency here would increase with re-
sidence time (lower face velocity, lower air/ cloth ratio, thicker
filter media). The relative effects of the se coacting collection
mechanisms is not sufficiently understood at pre sent for use in
practical de sign.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LAE!ORATORIES

-------
VI-56
The se foregoing factors are insufficiently defined at pre sent for a useful definition
of the effect on costs of changed efficiency requirements of fabric filters.
Conclusions
A situation of diminishing returns is indicated by performance equations of the
exponential type. In many case s a O. 05 grain/ SCF D outlet concentration become s a
practical maximum level for improving efficiency, even though it is by no means an
absolute limit.
The state of the art, then, allows the gas-cleaner manufacturer to predict per-
formance, design a collector, and guarantee it with some confidence up to about 0.05
grain/SCFD outlet loading for particles greater than 2 microns in size. At lower output
levels, his experience is limited. The large change in size or operating parameters re-
quired for further small increases in efficiency would magnify the uncertainties known
to exist in these simplified exponential relations with their empirical constants.
Measurement techniques used to determine dust loading in the ducted stream be-
fore and after the collector leave much to be desired, especially where small concentra-
tions and even smaller changes in concentration are to be used as evidence of guaranteed
performance. Lack of homogeneity of most dusts from iron and steel processes make
the use of monitored data (light scattering or transmission, for example) difficult to
interpret, or the equipment difficult to calibrate, for all the variations in dust composi-
tion, size, gas flow rate, etc. caused by process changes during a heat cycle, or from
heat to heat. Isokinetic sampling (sampling at stream velocity) with traversing probes
involves much averaging (in time and space) with calculation and readjustment continuing
during the traver se. This is co stly and of que stionable accuracy. Null probe s, too,
operate with a significant degree of error in trying to balance small pressure differences.
Neither approach to isokineticity can give a time history of emission rate during the
course of a rapidly changing heat cycle because only two or three traverses can be run
at best in an hour. Gas density (composition and state) and moisture content data should
be monitored continuously and used as input to sampling-rate determinations during the
course of a sampling test. Deviations here can seriously affect the loading measured
as grains of dust per dry standard cubic foot of carrier gas.
From such quantitative data as can be obtained, control equipment is de signed
(often with a costly excess performance factor built in) and guaranteed somewhat con-
servatively. The guarantee is proven (or indicated) by standard sampling tests, and
no assurance is given that any particular level of Ringleman chart greyness will not be
exceeded. Research is needed to find a method to measure dust concentrations inexpen-
sively; and agreement is needed to correlate de sign and enforcement base s of
measurement.
Very fine particulate matter, because of greatly extended surface area, causes a
much greater scattering of light, even in small concentration. A Ringelmann compari-
son should thus in some way account for the nature of the emission being sampled. If
this correlation can be made, then this economical method of testing might be used to
obtain adequate design data.
In the case of very fine steelmaking dusts from open hearth, electric arc, and
basic oxygen furnaces; the collector performance is difficult to predict because of the
following:
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-57
(1)
The particle - size distribution is difficult to quantify with pre sent
methods for sampled dust, and the correlation of the se data to
"in situ" dust in the furnace effluent gas is in doubt. (Large dis-
crepancies in reported BOF dust sizing are a case in point.) The
smaller the size, the greater the difficulty.
(2) Agglomerative properties of the dust are not well established and
the effect of this on sampled dust sizing and on collection mechanisms
in the gas cleaners is not well understood.
(3) The mechanism of collection upon which the performance equations
are based (inertial and electrostatic forces) tend toward zero ef-
ficiency in the size range of the bulk of steelmaking dusts «2
microns), where molecular interactions dominate the motion of
particle s.
Any attractive interactions or agglomerative tendency would be beneficial to particle
collection on a clean collecting element, but joining of particles into larger interadhesive
masses would tend to blind a filter matrix (lessening gas handling capacity), or to inter-
rupt electrostatic precipitator field propagation about the wires and plates, and make the
collector surface difficult to clean and the dust hard to handle. This in some cases
necessitates close control of temperature and humidity.
For low-velocity collectors (inherently large and thus economically inefficient for
large -particle collection), a diffusional mechanism can give significantly high collecting
efficiencies. (The effect is greatest, in theory, near zero-micron size, and decreases
with increasing particle size.) A middle ground exists around 0.9 micron in a bag
filter where minimum efficiency can be as low as 10 percent; exactly in the center of
concentration of some 70 percent of steelmaking dust. This is shown in an efficiency-
particle size relationship in Figure VI-33. Reference should also be made to Fig-
ure VI-35 for a fabric filter, and to Figure VI-38 for an electrostatic precipitator, where
this effect also seems to be indicated.
2
I
      .,.  -
10        
 '\       
10 - ~    /  
  - /'   
.0 = '=  - -  - 
  = 
.        
1        
I        
.
,
"'''C\.I 1111 "'("0'"
s'
..
:
~
~
FIGURE VI-33.
RELATIONSHIP OF PARTICLE SIZE TO COLLECTION
EFFICIENCY FOR A FABRIC FILTER(78)
Further research is needed to:
(1) Develop technique s for reliable particle - size -distribution data
repre senting the dust as it exists in the effluent gas
(2) Determine the extent of agglomerative effects and their effect
in gas cleaning and effluent sampling mechanisms

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-58
(3) Utilize the diffusion mechanism for small particle s in an
optimum way while retaining economical and efficient inertial
mechanisms for large-particle collection. (If the valley of low
efficiency between the size ranges where diffusion and inertia
are effective cannot be narrowed by this development, then
another tack at development must investigate other gas-solid
interaction phenomena for possible use in gas cleaning.
Particle -interaction effects may be important here. )
(4) Develop economical methods to measure dust concentrations.
(Methods should be adequate for design purposes and well
correlated to methods used for obtaining enforcement data. )
In view of the foregoing difficulties, it is concluded that changes in legally required
efficiency levels (to outlet loadings below about 0.05 grain/ DSCF) would at this time have
to be based on questionable design measurement and theory (whose extension into this
range is also questionable.) The cost of such changes, as indicated by present under-
standing of the mechanisms of collection with proven equipment, would become in-
creasingly great for collection efficiency changes of very small magnitude i. e., changes
which can only be measured with an error of the same order as the change sought.
Technological Factors Affecting Gas-Cleaner Performance
(Adapted From Swindell-Dressler Report Given in Appendix C in
Companion Final Economic Report on Cost Analyses)
The processes in the iron and steel industry can and do depart from design capacity
and design operating conditions for a number of reasons that include the following:
(1) Economic pressures dictate the continued increase in productivity
of an installed furnace.
(2) Technological improvements make possible significant increases
in productivity (such as the introduction of oxygen blowing to open
hearth and electric furnace steelmaking) of a new or existing
facility.
(3) Batch handling of special heats, or runs of varying sizes and
treatments.
(4) Slack market conditions may require cutbacks in output.
Changes in rate of production cause effluent quantities to increase or to diminish
both in gas volume and loading. Operating conditions in the gas-cleaning system can
vary with the se conditions, as well as with the weather, gas -utilization program, raw
material charge, etc. Also, noncontinuous or batch-type metallurgical processes vary
during the course of a heat in both quantity and condition to the effluent.
To maintain satisfactory gas-cleaning performance under these conditions, it is
necessary to have anticipated these factors in designing the pollution-abatement system,
rather than specifying for average conditions. Maximum capacity should be anticipated,
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-59
or adaptation to additional capacity provided. Adjustable equipment can often be used to
optimize performance over a range of operations.
Provision should be made also in the initial installation to meet (or to add and
adapt equipment to meet) expected future requirements of the pollution-control code s
both as to dust content of effluent and treatment of objectional gas and solid chemicals
in the effluent.
Assuming proper design and selection of equipment, any variation or variability in
the process, control equipment, or performance would generally require an added cost.
Any unique feature of a particular gas-cleaning application (particle size, dust loading,
corrosion, etc.) would generally require a departure from a system designed for the
general case.
The following exerpt from the British literature in 1963 summarizes the perform-
ance factors required for effective particulate removal (74):
"The Qean Air Act and the increasingly wide use of oxygen in both the classical and the re-
cently developed top-blown converter processes have combined to create an urgent need for
highly efficient cleaning of high -temperature effluent gases containing submicron iron oxide
fume to the visibility threshold of O. 05 grains per cubic foot. In order to satisfy this need,
manufactUrers of gas cleaning equipment had first to find how collectors which had already
been well proved in other fields could be adapted to applications of which they had had no
previous experience. This entailed not only the establishment of the empirical design pa-
rameters concerned with efficiency, but also a very close consideration of the ability of each
type of collector to cope with unavoidable variations in gas volume, temperature, hu-
midity' solids concentration, etc.
The flexibility of any given type of collector (i. e. its ability to operate efficiently
without breakdown over a wide range of conditions) is much more important in practice
than its theoretical efficiency at constant flow -rate and temperature, etc.. and the best
unit for any given application will often not be the one which a comparison of efficiency
and cost based on idealized operating conditions would indicate.
Every manufacturer who can offer a complete range of equipmept must weigh very,
many factors before finally offering one particular type of collector. He may be handicapped,
particu1ar1yin the case of a completely new installation, by a shortage, of basic process data,
but he can usually arrive at a fairly accurate assessment of the relative strengths and weak-
nesses of the possible units. '
Although the size distribution and shape of dust or fume particles are of course the
factors which determine the fundamental suitability or otherwise of any given design of col-
lector for a particular application, other characteristics of the solids, the carrier gas, and
the process itself must be also carefully considered and their effect on the collection device
evaluated before a final selection is made. '
The agglomerating propensities of the solid particles are important because they de-
termine the size distribution of the particles presented to the collector. The extent to which
agglomeration into clusters or chains of particles will have proceeded, and hence what the
effective particle size will be immediately before the process of final collection is begun,
cannot be accurately predicted, and in practice allowance is made for it in the empirical
design constants used by equipment manufacturers. Agglomeration after collection affects
the caking properties of dry material, making it more easily released from filter fabrics,
less liable to re-entrainment during precipitator rapping, and more easily settled from
liquid effluent.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-60
The electrical resistivity of the material to be collected is of the utmost importance
if a dry precipitator is to be used.
If the collected material is not free-flowing when dry it may create dust handling
problems. Hygroscopic dust will give rise to similar difficulties in 'dry' collectors, unless
humidity and the temperature of solids and gas can be maintained at safe levels by control
of the process, lagging, external heating. warm air purging, or by a combination of these.
For the collection of dusts which are corrosive when wet the obvious choice is a dry
type of unit. unless there is a risk of condensation. If the waste gases contain water vapour
which comes from the process itself, or has been added for cooling or conditioning them,
and sudden temperature surges are likely, elaborate precautions against condensation may be
needed, and a more compact wet unit constructed from corrosion-resistant materials may be
more economical as well as more reliable.
The physical and chemical characteristics of the carrier gas must also be carefully
considered when a collector is being chosen. The effect of variations in gas temperature
and humidity. in particular, must be carefully investigated especially if, as is almost always
the case, they accompany or cause changes in gas volume and dust characteristics during
and after collection. These factors are affected by the method of hooding, cooling, and
volume and temperature control, but no matter how carefully these are engineered the char-
acteristics of the process may still cause the collector to be subjected to conditions which
are far from ideal and impair its operation either directly by affecting the collection pro-
cess, or indirectly by hindering dust discharge or causing structural damage. Collectors of
different types are more or less susceptible to different non -ideal conditions, as shown in
Table 1. The table is only intended to indicate some of the fundamental strengths and weak-
nesses of high -efficiency dedusters in relation to fluctuating operating conditions of one sort
or another. and is not intended to be a comprehensive summary; it does, however, demon-
strate the importance of factors which have nothing to do with the properties of particles. ..
Additionally, from the same literature source(79), the following Table I indicate s
operating conditions which affect the dust-collector efficiency at a peak level of equip-
ment maintenance, and factors which require regular attention (cleaning the collector
surface adequately to match dust loading, temperature control in dry collectors to mini-
mize moisture and heat deterioration and maximize dust removal and handling prop-
ertie s) to insure peak efficiency throughout the life of the equipment.
The effect of operating conditions on the effective performance of the gas-cleaning
function, the effect of those conditions which cause maintenance difficultie s and shorten
service life, and the effect of those conditions peculiar to a particular furnace type of
process on the design and selection of gas cleaning equipment are best judged in the light
of operating and design experience. The following quotations are discussions of emission
cleaning for three iron and steel industry proce s se s which pre sent difficult problems of
equipment selection, performance, and maintainability. These discussions were chosen
for their concise and comprehensive consideration from an application point of view of
the critical factors of equipment use. While they center on British practice (where raw
materials, processes, codes, etc. have some variance from general American practice),
the discussion of each factor remains pertinent (with perhaps some difference in degree)
to a consideration of a corresponding American plant. After considering all the condi-
tions existing on a particular job of equipment application, an engineer may find that the
particular situation with which he is dealing is somewhat more difficult or less difficult
than implied in the following quotations.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-61
TABLE I Effect on collector performance of ftuctuating operating conditions (a)
Dry plate precipitator Fabric filter Scrubber
Irrigated precipitator
TemperatUre
Humidity
Flowrate
Corrosive
solids or gas
Inlet
concentration
Normally up to 650°F
with standard. construc-
tion but momentary
peaks of I OOO°F can
be toJ.erated. Tempera-
ture must be selected
to suit electrical
characteristics of dust.
Normal maximum
temperatUre depends on
"fibre used. Up to say
275°F with organic
synthetics, 600°F with
fibreglass. Higher peaks
tolerable but reduce
bag-life dispropor-
tionately.
Operation below dew-
point leads to bag-
Cleaning troubles.
Chemical and physical
damage to fabric likely.
Dust disposal difficulties.
Normally below 200°F
with presaturation.
Surges can be prevented
if maximum water rate
always used in saturator.
Efficiency unaffected by
changes in humidity,
providing gas remains
near saturation.
Water-rate and/or throat
area must be adjusted to
compensate for changes
in inlet volume.
Alternatively volume
may be kept const:mt
by air-addition.
Special materials of
construction will prevent
corrosion. High-pressure
(high top speed) stainless
steel fan impellers can
give trouble.
Efficiency increased by
operation at lower
flowrates.
Insufficient moisture
may lower efficiency
by increasing dust
resistivity. High
humidity with low
temperature may cause
condensation, possible
corrosion, insulator
and plate cleaning and
dust disposal diffi-
culties. Accurate
control of spray
cooling essential.
Efficiency increased
if flowrate reduced,
although gas distribu-
tion may deteriorate.
Efficiency lime affected
by flowrate. Pressure
drop reduced as volume
falls.
Special materials of
construction eliminate
corrosion but price may
rule out.
Initial design must be
based on peak loading.
Initial design must be
based on peak loading.
Corrosion can be
avoided by accurate
temperature control,
insulation, auxiliary
heating, bypassing, or
corrosion-resistant
materials of construction.
Filter fabric may be
damaged. "
Initial design must be Efficiency not affected
based on peak loading. by increased loadings:
effect on pressure drop
depends on duration of
surges, but can be reduced
by temporary increase
"in Cleaning intensity.
(a) Table I taken from Reference 79.
"SINTER PLANT
MAIN STRAND GASES
The gases withdrawn from the main strand of a sinter machine present a fairly difficult gas
cleaning problem, not, as in most other iron -and steelmaking applications, because high
efficiencies must be achieved on very fine particles, but because of other characteristics of
the dust and the gases themsel ves.
Volumes are great and the use of medium and high pressure drop collectors would in-
volve large nonproductive power consumption.
The waste gases contain large quantities of both sulphur oxides and water vapour.
Consequently they have a high (acid) dewpoint so that condensation and corrosion are a con-
stant danger, aggravated by the wide fluctuations of temperature which occur from time to
time.
The coarser fractions of the dust burden are exceedingly abrasive.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-62
Hence, the ideal dust collector will have the following characteristics:
A pressure drop as low as possible.
Ability to operate efficiently over a wide range of temperatures without ill effect from
occasional dampness of dust and collector internal surfaces.
,
A construction which minimizes condensation, lends itself to reasonably economical
corrosion prevention, and is not susceptible to plugging during the occasional but inevitable
periods of operation below dewpoint.
Freedom from abrasion troubles, preferably by complete avoidance of high velocities,
otherwise by pre -collection of the coarse abrasive dust fractions prior to passing the gases
through any collector in which high velocities are used.
DUST CHARACTERISTICS
The particle size analysis of the dust content of sinter strand gases can vary between quite
wide limits. . . The type of dust to be dealt with depends on the mix fed to the strand, L e.
proportions of home and foreign ores and return fines, and also on whether or not the burden
is conditioned in a pelletizing drum. It must be remembered that changes in dust composi-
tion occur as the rate of sintering alters and the relationship between temperature, flame-
front penetration, and position on the strand varies.
DUS T LOADINGS
The general level of dust concentration is affected greatly by the nature of the material fed
to the machine and can vary from plant to plant between O' 1 and l' 0 grains per normal
cubic foot and may occasionally reach l' 2. The rate of solids emission is very sensitive to
variations in the progress of the sintering process along the length of the strand. Dust is
mainly generated early in the sintering process and again when the flame -front reaches the
bottom of the bed. It has been suggested that in the intermediate zone the increased moist-
ness of the lower part of the bed causes it to act as a crude filter and hence to pass less
dust. It has been found that as complete sintering approaches the discharge end of the
strand, Le. as the mean hottest windbox number increases, the dust loading rises noticeably.
G AS TEMPERA TURE
Gas temperatures usually fluctuate between 60°C and 200°C but lOO-150C is the most com-
mon range and 125 C may be taken as a reasonable average figure for UK practice.
(U.S. practice is in the range of 300 F to 400 F)
GAS COMPOSITION
The only constituents of the easte gases which are important from the gas cleaning point of
view are water vapour and sulphur oxides, both of which affect the frequency and severity of
condensation in the cleaning system. There will usually be about lcp/o water vapour by vol-
ume in the gases and the sulphur oxide content, expressed as S02' may be as high as l' 5
grains per normal cubic foot. Unfortunately, no acid dewpoint figures are available, but
water dewpoints as high as 50°C are encountered and acid dewpoints considerably higher than
this must therefore occur. So far as condensation and corrosion are concerned, the relatively
high proportions of water vapour and oxides of sulphur in the gases complicate the design and
selection of gas cleaning equipment. . . ./but they tend to facilitate electrostatic precipitation.)
CHOICE OF DUST COLLECTOR
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-63
In the authors' opinion the sulphur oxide content of the gases rules out wet methods of col-
lection, since these would result in difficult liquid effluent problems, and saturated gases
ha ving hardly any thermal lift and still containing some sulphur oxides would constitute an
air pollution problem worse in some respects than the original one.
The choice of a dry collector will be dictated by the quantity and size range of the
dust in the case under consideration, the space a vailable, the pressure drop which can be
tolerated and the outlet loading required. Generally speaking, particularly for dusts at the
coarser end of the normal range and if an outlet concentration of 10' 15 grains per normal
cubic foot is the highest acceptable outlet loading, or the dust is finer, or a settling chamber
cannot be accommodated within the space available, cyclones may be used, but their pres-
sure drop (up to 6 inwg) is a disadvantage and they must be speciall y constructed to with-
stand erosion by abrasive dust particles. For a stack loading of less than 0.10 grains per
normal cubic foot a more efficient type of collector must be used.
If. outlet loadings down to O' 05 grains per normal cubic foot are required, the only
suitable device is the electrostatic precipitator. Were it not for the constant danger of
condensation, the fabric filter would be a possibility, but a filter fabric would 'blind' when
operated under moist conditions. It is true that the silicone-treated fibreglass fabric (which
would have to be used in any case to withstand the high maximum temperature) is much
less susceptible to plugging than are the natural and organic synthetic cloths, and has been
found to regain its porosity on drying out, but there would always be a risk of the cloth be-
coming 'starched' with soluble salts and failing prematurely through what can only be de-
scribed as cracking. Fibreglass, which has poor flex resistance in the first place, is excep-
tionally vulnerable to this sort of trouble. This type of collector also compares unfa vour-
ably with the precipitator from the points of view of pressure drop, space requirement, and
maintenance cost, and would not be recommended for main strand gas cleaning.
Although, in common with all other collectors, a precipitator for this application has
to contend with occasional condensation, its operation is not unduly affected by moist condi-
tions' providing precautions are taken against corrosion and providing it has efficient rapping
gear which will clear any. . . build -up and prevent progressive deterioration in its perfor-
mance. The water vapour and sulphur oxides in the waste gases 'condition' the dust and, to-
gether with the relative coarseness of the dust,. . . (assist the precipitation process.)
The Head Wrightson sinter machine installed at the works of the Skinningrove Iron Co. Ltd.,
Saltburn-by-the-Sea, is provided with a Head Wrightson/Research Cottrell dry plate precipitator.
The machine was. designed to process a wide variety of home and foreign ore mixes, and experi-
ence indicated that the dust burden in the waste gases could be reduced by a simple settling
chamber from 1. 0 to 0.3 grains per normal cubic foot. (This is lower than typical American loading.)
The gas valume from the 16 x 6ft square windbox machine is 180000 per normal cubic foot.
The precipitator has two treatment zones, energized by a 15 kVA 230 mA transformer-
rectifier set and operates at a treatment velocity of 6.8 ft/s. In view of the expected
intermittent operation, it was thought advisable to fabricate the collector plates in copper-
bearing 'Corten' steel (O'IDfoc max., 0.1-0.3"7oSi, 0.5-1.oDJoMn, 0.3-0.5DJoCu, 0.5-1.5<'/oCr,
0.1-0.2"70 P) and these have withstood the adverse conditions very well without noticeable
deterioration. The interior of the precipitator shell is protected with gunned aluminous
cement and the whole unit is thermally insulated to minimize condensation. The precipi-
tator. . . was designed to operate at an average temperature of 300°F, and at an efficiency
of 86.7"70, corresponding to an outlet loading of 0.04 grains per normal cubic foot The
design performance has been. . . achieved and, although the sinter plant has worked on a
one or two shift per day basis and the precipitator has undergone an abnormal.LImber of start-
ups, there has been no deterioration of its internals. The sinter fan was inspected in August
1963, 20 months after commissioning and showed no sign of wear other than a general
smoothness over the faces of the blades; it is estimated that it will operate for at least another
3 -4 years without requiring maintenance. The machine had produced 250000 tons up to the
time of the inspection. Reduced fan maintenance and plant downtime are two useful in-
direct benefits of efficient main strand gas cleaning.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI - 64
The dust discharged from the precipitator hoppers is conditioned in a pelletizing drum
and the pellets produced are returned to the process via the return fines conveyor.
DISCHARGE END EXHAUST SYSTEM
The whole of the discharge end of the sinter machine is usually completely enclosed; 100
tons or more of dust per day may be released by the equipment in this area (i. e. the end of
the strand itself, the breaker, hot screen, and discharge to cooler). Air volumes vary with
the size of sinter machine and the completeness of hooding. and are between 30000 and
150000 cubic feet per minute, Gas temperatures are usually between 40° and 150°C. Both
the loading and the size range of the entrained dust are affected by the designs of hoods em-
ployed' and the exhaust volumes allocated to them, but dust burdens are typically in the
range 4-6 grains per normal cubic foot of which 8fJ1/0 might be <100].J m and IfJ1/o <10 ].Jm.
Careful hood design. combined with adjustment of individual exhaust rates during com-
missioning' can reduce both grain loadings and the proportion of coarse abrasive particles
carried in the gases. 1t is relatively easy to obtain collection efficiencies of 90 -95"/0 by
means of simple high -efficiency cyclones, and the stack discharge in such cases will con-
tain about 0.5 grains per normal cubic foot of dust, 9fJ1/0 of which is <10 ].Jm. At.this sort
of grain loading the stack plume does not appear offensive; all the same it represents a very
high rate of solids emission (up to 700 pounds per hour on a large plant), and more and more
interest is being shown in alternative higher-efficiency collection methods.
For cleaning the tip end emission the fabric filter and dry plate electrostatic precipi-
tator are two obvious possibilities. Wet methods can be employed (self-induced spray units
are fairly often used in the USA), but are notto be recommended because they introduce
a secondary (liquid) effluent problem, and are liable to suffer from wet-dry interface troubles
and sometimes from sludge discharge problems. At first sight, the fabric filter would appear
to be ideally suited to this application, providing it is designed so as to a void excessive
scouring of the bags by the abrasive dust, and properly maintained so that a small leak in
one bag cannot' grit -blast' a hole into an adjacent one and start a rapid and messy chain
reaction. The first requirement is quite easily satisfied, but the second is not so straight-
forward and a short period of neglect could have expensive and inconvenient consequences
in the form of extensive bag replacements and operation at reduced capacity. From the
point of view of efficiency and capital cost, the fabric filter is a 'good buy', but running
costs are a most important factor, and cannot be accurately forecast.
While the operating characteristics of the dry plate precipitator are quite predictable
for this application, discharge end precipitation is difficult because of the high resistivity
of the dust at the gas temperatures normally encountered, when the moisture content is less
than about 1. 5 % by volume. In cold, dry weather the water vapour content may be as low
as O. 5"/0 by volume, and under these conditions unstable precipitator conditions are liable
to occur at temperatures around 60°C.
The addition of relatively small quantities of water vapour, sifficient to raise the vol-
ume percentage to 2. 0, leads to a marked improvement in precipitator performance, as
does the addition of 100 ppm of S02. If a guaranteed efficiency. is to be maintained under
every circumstance, and at all times, and if water vapour or S02 cannot be added, the
precipitator will be perhaps three times as large as a unit which will operate satisfactorily
under all but the driest conditions. It is therefore well worthwhile either to mix in gases
from some other part of the sinter system or to add steam. If the problem of conditioning
can be overcome this is a very straightforward precipitator application,. . . (Application of a
dry system at the discharge end when water cooling at the sinter strand is employed could lead
to the reintroduction of moisture control problems.)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-65
QUENQ-l GASES
The quenching of hot fines in pug mill or drum gives rise to large quantities of fine dust,
particularly during periods of erratic plant operation.
In a typical installation the volume of gas vented from the drum was 7600 normal
cubic feet per minute at 40-120.C, containing between fP/o and 24= water vapour by volume.
It was found that the dust loading was greatly affected, not only by the quantity and distribu-
tion of spray water, but also by the quality of sinter being made. During normal operation
of the machine the loading was found to vary between 1. 3 grains per normal cubic foot when
sintering was complete, and 4.7 grains per normal cubic foot when incompletely sintered
material was being discharged from the strand. Shortly after commissioning, before the
sprays had been adjusted and while the operation of the machine was abnormally erratic,
the mean dust concentration had been 4'8 grains per normal cubic foot (corresponding to a
rate of discharge of nearly 400 pounds per hour) and the peak loading 33.2 grains per normal
cubic feet in gas volumes of 8500 -11000 normal cubic feet per minute. This illustrates the
effect of plant operation on stack emissions. The final emission rate averaged 140 pounds
per hour compared to 280 pounds per hour from the main stack and 135 pounds per hour from
the tip end cyclone stack.
The quench stack dust is rather fine (9f11/o <100].J m, 300/0 
-------
VI-66
be pressurized with warm air to keep the insulators dry. Dust should preferably only be
stored in the hoppers in an emergency because it tends to bridge, and it may be advisable to
heat the hopper sides. Some condensation is bound to occur at start-up, and it is advisable
to clear as much collected dust as possible from the interior of the precipitator while it is
shut down. If this is not done conveyers and dust discharge val ves may become clogged
with moist dust. If possible the precipitator should only be energized when it has reached its
normal operating temperature, so that little dust is collected in it when it is sweating. If
these. . . precautions are observed the dry plate precipitator will operate continuously. . .
if not, severe build -up, electrical and operating difficulties, and corrosion will be experienced.
A dry plate precipitator installation on a 250 ton tilting OH furnace. . . follows a waste
heat boiler and ID fan, and is designed to clean 78900 cubic feet per minute of furnace
gases at a maximum temperature of 280°C. The design inlet loading is 5 grains per normal
cubic foot during oxygen lancing and the outlet cleanness O. 04 grains per normal cubic foot.
The precipitator is insulated, the hoppers are steam -heated, and the insulator compartments
on top of the unit are pressurized with 600 cubic feet per minute of air at 200°F to prevent
outward leakage of dirty gas and to keep the insulators both dry and clean. The precipitator
has three treatment zones each of which is energized by a 21 kV A, 250 mA transformer
recitfier set. This is quite a good example of a precipitator fitted into a very restricted site,
utilizing turning vanes to reduce inlet and outlet duct sizes without detriment to gas
distribution.
The fabric filter may be used for OH gas cleaning but is more susceptible than the
precipitator to condensation troubles, has a much higher power consumption, and requires
more space. A filter serving one of the Ajax furnaces was reported to operate at a pressure
drop of 8 inwg and to have a bag-life on only 20 weeks. There seems to be no reason why
filters of modern design using improved high -temperature fabrics should not operate satis-
factorily at a pressure drop of 4-5 inwg with a baglife of a year or more, but prolonged
pilot -plant testing would be needed to prove the durability of the filter fabric.
Both the irrigated electrostatic precipitator and the high -energy scrubber are capable
of cleaning OH fume to O. 05 grains per cubic foot or better. but they would ha ve to be
constructed from expensive corrosion -resistant materials and would create secondary prob-
1ems of liquid effluent treatment and loss of stack gas buoyancy.
ARC FURNACES
FURNACE PRESSURE CONTROL
For consistently good fume control at minimum rates of extraction, automatic control of
furnace pressure is essential. The indicated pressure which it is necessary to hold within the
furnace depends on the position of the pressure pick -up. The accuracy of control required
is of the order of :1:1. 0 nwg for furnaces melting OH grades of steel but may be as fine as
:1:0.03 inwg for a furnace producing alloy steels. The control system used must have a
high speed of response if it is to cope with sudden fluctuations within the furnace.
GAS COOLING OR CONDITIONING
I-
I
Temperature at the outlet of the combustion chamber may be upwards of 1000 C and
the gases must be cooled before they can be cleaned. The methods available are air dilu-
tion' indirect cooling by heat exchanger, and evaporative cooling. It is considered that
the latter is often the best compromise on the grounds of simplicity, final gas volume,
space requirements, and initial cost.
However, the type of collection device used will often dictate the manner in which
cooling is carried out. With wet methods of collection, a comparatively'small spray tower
may be used (without fine control of the cooling sprays) and air dilution or indirect cooling
would be pointless. If dry precipitation is preferred, the gases must be conditioned (most
simply with water) and if a spray conditioning tower is required for this reason the gas will
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-67
be spray cooled to the desired precipitator operating temperature. The fabric filter does not
require pre-humidification of the gases for efficient operation and is, moreover, exception-
ally vulnerable to condensation. The preferred method of cooling in this case will depend
upon whether the filter fabric is organic-synthetic (e. g. Orlon or Tery1ene) and therefore
not suitable for operation at over 130°C, or fibre glass, which will withstand up to 250°C.
In the former case air dilution or indirect cooling may be used, but in the latter spray cool-
ing should present no difficulties providing a good control system is fitted.
SYSTEM CAPACITY AND SAFETY
. . . The details of safety require that. . . very conservative assumptions are made. The
problem of explosion hazards has been considered in recent papers.
Air may enter the system at the air break between elbow and fixed fume pipe and at
the combustion chamber, as well as through the furnace openings. The volume of air
entering by each of these routes is unimportant providing (a) that control of fumes is ob-
tained and (b) that the final waste gas volume is such that even if combustion has been in-
complete an explosive mixture cannot be formed.
The combined effects of combustion and dilution have been calculated for the lancing
period, and are shown in Table III. However, the rate of evolution of combustion, follow-
ing the addition of oily scrap cannot be predicted, and it must be remembered that in
practice the operation of a fume cleaning system must take second place to the production
of steel; allowance must also be made for occasional deficiencies in the standard of both
operation and maintenance of cleaning systems. Hence, although under ideal conditions
an 02 to waste gas ratio of 10: 1 would no doubt be adequate, it is recommended that a
ratio of not less than 15: 1 be used.
Ratio of waste
gas/oxygen
injection flowrate
TABLE III Effects of combustion and cWutionn

Approx. %
combustion for safe
operation (based on
100% oxygen
utilization)

Nil
Nil
5%
10%
32%
50%
55%
% Carbon
monoxide if no
combustion occurs
22: 1
16:1
15:1
14: 1
10: 1
6: 1
5: 1
9.1
12'5
13,3
14.3
20.0
33,3
40,0
Current understanding of the explosion problem is incomplete; explosions have been
reported even in conservatively designed systems following errors in operation, and it is
considered more prudent to use theory to predict the magnitude of apparent safety margins
rather than to reduce these to the point where (due to the intrusion of incalculable factors)
they do not exist, and a variation in the process or a mistake by an operator can cause
an explosion.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-68
GAS CLEANING
The furnace gases may be cleaned to 0.05 grains per cubic foot by precipitator
(wet or dry), fabric filter, high -energy scrubber, or combination scrubber -precipitator.
Dry plate precipitation is relatively straightforward providing the gases are properly
conditioned. It is therefore ideally suited to direct extraction systems but much less so for
hood or conventional hood vent installations.
(A) 75 ton furnace. . . has been fitted with direct extraction fume control equipment
and fume is to be collected by a dry -plate electrostatic precipitator ('B' unit referred to
below). The lancing rate of this furnace is 1200 cubic feet per minute and the volume
during lancing. after combustion and cooling 49200 cubic feet per minute. The precipitator
is designed to clean a total of 83200 cubic feet per minute from the existing furnace and
another which is to be added in the future, from 6.5 to 0.05 grains per normal cubic foot.
Furnace gases will pass through a water-cooled elbow and refractory-lined fixed duct
connected by a power-operated movable sliding sleeve. into a gas burner followed by a
combustion chamber. They will be cooled and conditioned in the rectangular spray
tower and will enter the precipitator at a temperature of 500'F.
The fabric filter is theoretically ideal. having a uniformly high efficiency irrespective
of throughput but it must be carefully designed and protected against condensation. Filtering
velocities may be as low as two feet per minute so that space limitations will often exclude
this type of cleaner.
The high-energy scrubber operating at a pressure drop of 30 inwg or more will do a satis-
factory fume-cleaning job (U.S. air pollution regulations would require about 4S inches water gage.)
and its compactness is a great advantage, particularly when the available space is limited.
Power may be saved by regulating the fan in an efficient manner to suit the rate of exhaust
required for fume control and the pressure drop needed at different periods of the melt to
give the statutory final gas cleanliness. but this is only practicable if the pressure drop of
the scrubber can be adjusted to the desired level over a wide range of flow-rates.
It must be stre ssed that in the long run regular maintenance and attention to op-
erating conditions affect the cost and effectivene ss of any gas-cleaning unit. The in-
corporation of automatic controls, operator-proof controls, scheduled preventive main-
tenance, anticipation of adverse process conditions and raw material possibilities are
important to the continued performance of gas-cleaning equipment after the guarantee
period.
Attention is now directed more specifically to the following parameters that affect

gas cleaner performance:
(1) Effect of gas-volume changes on collection efficiency of a dust
collector
(2) Effect of pressure drop within the gas cleaner on efficiency and
capacity of the collector
(3) Effect of dust loading, and effect of collector-surface renewal
on pressure drop, volume, and collecting efficiency
(4)
Effect of particulate as generated in each metallurgical proce s s
(particle density, particle density, particle size, size distribution)
on efficiency of each applicable dust-removal device
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-69
(5) Effect of temperature on efficiency of and gas volume to
collector, and required gas conditioning for cooling and
humidification before dust removal
(a) Gas analysis as it affects conditioning required
prior to cleaning and exhausting
(b) Corrosion and the use of water
(c) Abrasion and chemical effects of dust
(6) Adaptability of the particulate-removal system to removal
of gaseous pollutants.
""~fect of Gas Volume Changes
As previously indicated, the volume of effluent gas emitted by a metallurgical
proce ss may vary greatly during one heat, or according to changing production level of
the proce s s. Because the efficiency of dust removal change s when volume change s, it
becomes necessary to
-- -operate at constant volume with air substituted for effluent gas
deficiency,
---or, use a gas-cleaning device which adjusts itself to volume
changes, or is adjustable to satisfactory efficiency over a range
of volume.
Self-induced or orifice washers and certain fluidized-bed scrubbers can adjust
themselve s, essentially at constant efficiency. Adju stable -throat venturi s, orifice-
wedge and flooded disk scrubbers can be adjusted to suit a range of gas flow. The se and
other wet scrubbers can also be flooded (uneconomically) to achieve the same effect.
Multiple units (nested cyclones, parallel scrubbers, precipitator tubes or ducts,
multiple venturis, baghouse filter tubes) can be partially blocked off to maintain high
(design) efficiency at reduced volume, with economy of water and power use.
-- -or, de sign for maximum possible effluent volume, and lIover-cleanll
at reduced volumes.
The following quotation from the British literature in 1964 illustrates one sug-
gested method for determining good design in relation to gas flow(l3):
"It will be appreciated that the efficiency of a precipitator is greatest when the velocity of the gas
through the cross-section of the electrode system is uniform, and no gas is bypassing the electrode system.
This is ensured by the construction of . .. models... of the precipitator and inlet flue system. The flow
conditions in the model are adjusted to give the same Reynolds number as the full-scale plant, allowance
being made for scale factors, gas viscosity. and density. The flow pattern in the model is corrected using
splitters and baffles, their position being determined by experiment. Such model tests permit requirements
to be worked out in advance, and a void the difficulties in carrying out such work on site on the finished
plant. "
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-70
Effect of Pressure Drop
Because of large cross section and control of build-up conditions (temperature and
humidity) and regular rapping for dust removal, electrostatic precipitators usually will
show negligible change in resistance to flow in operation.
Bag filters, when new, have low resistance and low efficiency. Sometimes a pre-
coat of dust is applied to make the initial cleaning of process fume more effective. The
buildup of dust increase s both efficiency and pre ssure drop until the cleaning (by shaking
or reverse flow of air) cycle is initiated (often by a pressure signal). Then efficiency
will be at a lower (but still effective) level until the dust layer reforms on the fabric. (55)
Wet scrubbers increase in efficiency with increased resistance due to mechanical
constriction of the throat area or added water input. The proportionality of change as
attributed to Semraufs correlation is described later.
Efficiency of cyclones also depends upon pressure drop. Cyclones are used only
with coarse, easily collected dusts, however, and usually with a view to product re-
covery as much as to gas cleaning. As such, they may usually be regarded as process
equipment. The rules relating pressure drop, capacity, and efficiency are available in
the Air Pollution Engineering Manual. (80)
Effect of Dust Loading
An electrostatic gas cleaner is essentially a constant-efficiency device, so that
any change in inlet loading will be reflected proportionally in the outlet stream loading.
However, in actuality, changes in dust build-up occur and adversly affect the propaga-
tion of a uniform electric field. Plate spacing must be designed to accommodate the con-
dition of heavie st expected dust loading. Automatic controls are often required to main-
tain an optimum electric field without spark-over.
A well-designed bag filter will be unaffected by a change in inlet loading, except
that automatic cycling of the bag-cleaning system will adjust to the change (within cer-
tain limits of variability).
A wet scrubber will yield constant efficiency for a given pressure drop. There-
fore, a change in inlet loading will be reflected proportionately in the outlet loading.
However, wet-scrubbing systems can be adaptable to changing conditions, provided suf-
ficient power is applied. A venturi throat can be closed to maintain a given effluent
level with increased dust generation in the process. A process whose fume output varies
widely with time could be handled by making frequent adjustments of the cleaner to
maintain a constant acceptable output of fume.
Particulate Characteristics From Different
Proce s s Se gments
In Appendix C, "Characteristics of Emis sions", of this report, some data are
presented on the nature of particulate material generated by various processes and
conveyed by gases emitted from the process vicinity. This dust is generally nonuniform
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI -71
from one particle to another and from process to process. The differences may be
categorized as particle size, shape, density, and composition.
The mechanisms of particle collection on which gas-cleaning equipment are based
vary in collecting efficiency, generally with particle physical properties. The chemical
nature of the dust may affect its susceptibility to electric charging. This would pri-
marily affect electrostatic precipitation; but could be a second-order effect in wet
scrubbing and in fabric filtration. Solubility and chemical activity in water would affect
the water cycling, dust handling, and collector-surface maintenance in wet collectors.
Some general variations in the efficiency of collectors with these particle properties
can be drawn. Grade -efficiency curve s typical of industrial collectors in the mid-1950' s
have been presented for various types of dust-collecting equipment. (78) These show
the efficiency of collecting particle s of a given size. The curve s were based on te st
results using a standard dust (Table VI-13) with a 2.7 specific gravity. These curves
can be used to indicate in a general way the relative applicability of each type of equip-
ment to different process fumes. As shown in Figure VI-33, the efficiencies generally
are lower (often dropping abruptly) for finer grade s of dust. Some device s are more
economical to operate, but they generally do not clean fine particles from gases as
well as others.
TABLE VI-13.
GRADING OF W. C. 3 TEST DUST(a)
Size
of Grade,
microns
Pe rcentage by
Weight in
Grade
Percentage by
Weight Smaller
Than Top Size
of Grade
104-150
75-104
60-75
40-60
30-40
20-30
15-20
10-15
7-1/2-10
5-7-1/2
2-1/2-5
2-1/2
3
7
10
15
10
10
7
8
4
6
8
21
100
97
90
80
65
55
45
38
30
26
20
12
(a) From Reference 78.
Relationship of Particulate as Generated by
Different Proce sse s to Collecting Efficiency
Stairmand(78) has presented grade efficiency curves for various cleaning equip-
ment at specific conditions using a standard dust. These curves indicate the relative
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
1---
VI-72
applicability of each type of equipment to different dusts. Some devices are more eco-
nomical to operate but do not clean fine particulate from gases as well as others.
The curves are applied to various processes with their given particulate size
di stributions.
By making a density correction, the curves can be applied to dusts for which
particle-size distribution data (as in Table VI-13) are known. Some distribution data
are given in Appendix C to this report. No quantitative data are available upon which to
base corrections for particle shape, composition, and surface differences; so such an
application of the curves will not be quantitatively precise.
Particle - size -distribution data available for thi s kind of analysis are inadequate in
some measure. The size ranges reported are usually too large and require excessive
averaging in the region of greatest variation in efficiency on the grade-efficiency curve
(the fine-particle-size region). Steelmaking dust is largely concentrated in this region.
Whether or not averaging according to the log-probability distribution would be applicable
to distributions having given data ranges such as 0-1 micron or 0-5 microns is not
known.
The shape of the grade-efficiency curve may be affected somewhat by (1) process
variables which alter the properties of the dust, (2) conditioning of the dust by humidity
and tempe rature control (in the case of electrostatic precipitation), (3) collector
geometry (affecting treatment time) and (4) energy input (see Table I). The develop-
ment of a family of curves should be undertaken to develop grade -efficiency variations
at different levels of pertinent operating variables (such as scrubbing-energy level or
electrostatic-precipitation treatment duration). Note that steelmaking fume can gener-
ally be adequately removed with a scrubber pressure drop of 40 plus inches of water,
or by electrostatic precipitators whose geometry, control, and energization are spec-
ifically selected for that application. Using the given curves, however, to analyze the
effect of each type of treatment on actual proce ss dusts will give a rough comparison of
collectors for a given task, and a general comparison of dust-property effects on per-
formance of each collector. At least, relative indications may be drawn. Efficiencies
measured in the field of equipment that is collecting dust from the actual processes
dusts whose properties would also be tested under the collecting conditions) would allow
precise comparisons, more precise (and probably more economical) designing, and
dependable predictions of performance. Such data generally are not available, not very
good, or undisclosed. Therefore, the theoretical treatment while limited, is the be st
alternative available for relative comparisons.
The fabric-filter curve has been deleted from Figure VI-33 because it is based on
a theoretical calculation for a new filter. A more typical practical grade -efficiency
curve given in Figure VI-34(82) is based on actual test results. However, both indicate
that the efficiency would not go to zero as particle size approaches zero. This is dis-
cussed in the literature(67) in terms of a diffusional collecting mechanism which comes
into play at an increasing rate as particle size diminishes. The zero drop-off in the
other grade-efficiency curves represents the failure of the inertial impaction mechanism
to collect small particles. Discussions in the literature(67) include diffusion of particles
to water-droplet targots as well as to filter media. Therefore, this effect should also
apply to wet scrubbing. As indicated by the U-shape of curves in Figure VI-38, the
mechanism of diffusion seems also to apply to electrostatic precipitation. This mech-
anism, then, would suggest an upward alteration of the grade-efficiency curves.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
1---
SlOO
U
i 10

t
: .0
~
. .0
o
    .- '- 
   /'   
  /'    
 /     
 /     
      I
I      
I      
II      
      I
;
.. .0
:s
II
u
10 40 .0 10
'.UTIC",. 1111. N.C:AO...I.
.... II
Medium ~lfidcnC)'. 1b"'.d':rouehp'.IIC')'tlont.
Erficiency II , rn~ .. l1~.
.100
=
..
r 10
..
 ./ - - -  ,..-
I ,    I
I ,/     
I      
,      
I  - U"II'-AUO  
  ----- DRy ,- -
  I    
  , I I  
.
..
~.o
~
.00
.
~
... .0
a
..
10 10 JO 40
,....""C\,.I '.'U.,W'CAONI.
F.. 17
urcr Jiametu dry .and iuigucd c}'I.:Iont.
Edic~y II , nhnunl .. '7'/.
Fabric filter
curveomi tte d
for reasons
given in the
text.
SIOO
..
r 10
o
,.....- r
I 
I 
I 
 ,.....-
..
:60
u
~
.00
~
::; J 0
:s
~
. '0
'."TlCLI
IS 20
1111. "'CAO"'.
Fit- =
Sdl.induced \f'tIY cullector.
I!tkiency .1 , mn:ron. - 9 J ~~
VI-73
~IOO
~
r .0
.
..
~ .0
u
~
.'0
.
~
.. '0
:s
II
u
 I
I 
I 
I 
",,100
:,
...
% 10

.
u
~ 10
U
~
.00
o
 ---
I 
I 
I 
o
100
ao 40 60 10
MATICL. 1'11. WIGAONI.
Y.. 15
Hilh efficiency (Ion, cone) cyclone.
Efticicrw:y II , l!'i\,"fOfU - 7) '/,
.
g
u '0
:s
~
lOG,
10 20 )()
ItA"TlC:U 1111. WICRONS.

Fit- IS

Small di.meter. rubulu c)'clones.
ElRcicncy u.' ",i.:ronJ - &91.
'0
10
,..'00
=
u
~ 10
~
..
u
~ 60
o
  I --
  V 
 /  
/   
I  I 
  I 
  I 
,..100
=
u
r 10
/'  - - - - 'r- - - -
-   
,         
,         
,         
I         
    - UIAICAaO  
    __n 0" I   
    I     
    I I  
u
~
. '0
.
g
u '0
:s
g
10
10 20 JO 40
PARTICLI SIU. W'CRONS.
Fla. 18
Low prcssure drop cdlul" .:ydone..
Efficiency al , :ni.."TOfts - 042"l.
.
...
: 60
U
~
.00
.
g
u 10
~
II
u
o
10
5 10 II JO
MATltLI IUI!. I ~ ICR.?t4 s.
Fit- II
EI~tic preripiutOt. IrriS3Itd-(ffid-:n\:y ae " mkrons .. 9a~/.
Dry-efficienC)' al ~ microes .. 9: ~;.
~ 100 
~ I
~ 80
. 
u 
~'O 
~ 
.'0 
. 
2 
~ 
~ 30 
~ 
0 
I 10 I'
"'.TleLI Sill, NI(:-ONI.
""21
Spray lOWer,
Efficiency II , micron. .. 9<'%
10
..
~ 100 
= 
U 
~ 10 I
. 
u 
~ .0 
U 
~ 
.00 
. 
2 
~ 
uJO 
~ 
0 
.s 
I I' 1
'.A~TICLI 'III, ,..I(IIOH,.
Fit- u
"'" inpi"l..:mem scrubber.
Elftciel'Cy JI 1 mkroos - 97 :',
.0
..100
=
u
r 80
 - -
.'  
~'OO
u
i 10
 1/
I 
f 
I 
I 
.
..
~ 60
~
woo
.
g
u '0
:s
II
u
o
I 4 I . , .
'"""TlCLI 11'11. !tit r CROM I.
. F.:U
Vanu" tc:nJbbcr.
EfficiaKy 81 , mil,.Tons ... ~,9,6'1.
.
u ,
~.o
U
~
"'0
.
g
u '0
~
II
u
o
10
.1 " 50 . 7 .
P." TlCLI SUI. NICRONS.
. FIa- ZS
Di.intecrltOt ISS washer.
Effiocncy at " mi.:rons .. 981.
'0
..
FIGURE VI-34.
GRADE-EFFICIENCY CURVES FOR DIFFERENT TYPES OF
AIR-POLLUTION CONTROL EQUIPMENT(78)
(These curves were determined on the relatively coarse
dust described in the text, and do not necessarily apply
directly to metallurgical fume. )
BATTELLE
MEMORIAL
INSTITUTE - COLUMBUS
LABORATORIES

-------
VI-74
Variations in this effect will occur with temperature and particle concentration. By
designing low-flow-rate collectors, and by optimizing inlet conditions, one could take
advantage of thi s me chani sm with fine du sts.
99.99
After 10 shakes
c:
.
u
c:

-------
VI 75
TABLE VI-14. CALCULATED RELATIVE EFFICIENCY OF COLLECTING EQUIPMENT FOR VARIOUS PROCESS
 DUSTS AND COLLECTORS    
 (Results reported here are subject to the limitations given in the text. ) 
   Flux Fraction    
   (as in self-    
  Sinter fluxing sinter Basic Oxygen  Electric Arc Pressure Drop
  Strand making) Furnace Open Hearth Furnace (in. water)
Particle Specific Gravity 4.0 2.7 5.0 5.2 . 3.93 
grains  4  4-8 2-7 3-6 
Inlet Loading, SCFD   
Required Efficiency to Attain 98.73 98.9-99.38 97.5-99.28 98.33-99.17 
0.05 Grain per Standard      
Cubic Foot, U/O       
Computed Efficiency, U/o, for:      
Cyclones       
High Throughput  65 59  Not Applicable  3.7
High Efficiency  91 90    4.9
Multicyc10ne  98.5 98    4.3
Wet  97 96    3.9
Wet Scrubber       
Low Energy       
Spray  97 97    1.4
Wet Impingement 99.75 99.52    6.1
Self-Induced  98.25 98    6.1
High Energy       
Disintegrator  99.32 98.95 72 88 86 
Venturi  99.98 99.95 85(a) 94.5(a) 94(a) 22(a)
Electrostatic Precipitator      
Dry  99.00 95.5 64(b) 83(b) 81(c) 0.6
Wet  99.86 98.95   86(b) 0.6
Fabric Filter  99.99 99.99 (97.8) 99.5 99.5 4.0
(a) A pressure drop of 40+ inches of water gage normally is required to clean steelmaking fume to the specified 0.05
grain per standard cubic foot with a venturi scrubber.
(b) The codes in the mid-1950's (when these grade-efficiency data were typical in Britain) were less restrictive.
Precipitator designs today can be guaranteed to 99.5 percent efficiency if required, but the curves in
Figure VI-34 still can be used for comparative collection efficiencies.
(c) With proper humidification.
(d) Underlined values for computed efficiencies meet the required efficiency for outlet bonding of 0.05 grain per
standard cubic foot.
BAT.TELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-76
(1) The fan-power requirement is proportional to volume because
of this, many systems cool the gases to a minimum practical
temperature before entry into the fan, unless thermal lift
must be maintained to get rid of noxious gases in the effluent.
Since positive pressure ordinarily makes for simpler struc-
ture in a precipitator, and be st efficiency requires several
hundred degrees of temperature, this benefit is ordinarily
not available for a forced-draft precipitator fan. However,
in the wet scrubber, where efficiency depends on a high gas-
stream energy, at a high power-consumption level, the cooling
of gase s is particularly beneficial.
(2) Precipitators and especially bag filters are limited in the
temperature at which they operate. Thus, cooling as a
preconditioning step is required before entry to the gas
cleaner on many processes. The method of cooling also
affects the volume of gas to be handled, as shown in
Figures VI-36(l3) and VI-37(82).
1000 --.
900
800
~ 700

~: 600
~ ~oo
~~ 400
~v"


lool'
100
-L
o reo 100
----.--.-----.---
LU
a:
2 S
«
ITi
"-
~
<./),
« ,..~
t~ L-J 4
0'
CJ 5
'Z 0
o t)?
~
15 1-'
- «
h 0 3
) ~ 1.-:
1.-!.. <~
(, LJ..J
. CI,
ILJ t.:.
..,- L
J => 2
G V!
:> 
-------
VI - 77
Effect of Humidity
Additions of water to the gas stream or to the gas-cleaner system is a frequently
required preconditioning step., The humidity of the gas entering a baghouse must be
sufficiently below the dewpoint to preclude corrosion of the structure and clogging of the
bags with moist cake. On the other hand, quenching is an economical way to cool,
avoiding expensive radiation ductingor excessively large components to handle dilution
air.
Humidity is sometime s critical in the electrostatic precipitation of certain dusts
of high resistivity. Again, it can be coupled with cooling quench in the pre-conditioning
zone, but care must be taken to stay above the dew point temperature. No liquid effluent
results from these cases. It should be noted that in both of the above systems, the acid
dew point is also critical from a corrosion point of view in those cases where sulfur
dioxide is a significant process effluent component, such as the open hearth and some
coal-burning processes. While it has been noted that sulfur dioxide can be beneficial in
the precipitation of some proce s s dusts, this benefit is likely to be lost as sulfur dioxide
regulations take effect. ' .
Water flushing of elbows, fan blades, and other parts of the systems has been ef-
fectively used to inhibit impingement abrasion and to prevent dust buildup.
Corrosion, abrasion, dust buildup and excessive temperature are the most frequent
maintenance problems on a gas-cleaning system. Where water is used as a remedy,
careful pH control is important, as is control of solids buildup in a recirculating-water
system.
Except for this preventive maIntenance, water effluents usually are the result of
wet scrubbing or wetted-surface precipitation. This later finds application to electric-
arc-furnace fume-cleaning where satisfactory particle resistivity for collection is dif-
ficult to achieve, and in sinter plant and blast furnace applications where coarse,
abrasive dust must be removed.
""'ffect of Electrical Resistivity of Dust
The effect of electrical resistivity on the collection of dust is discussed as follows
in a British source published in 1964(13):
Much has been written on the subject of the effect of the resistivity of the dust on precipi-
tator efficiency and it is not proposed to go deeply into this aspect of the subject;. .. It can be
shown, however, that for dust of a very high resistivity, precipitation efficiency can be. seriously
reduced (see Figure VI -38 and section on particle sizing), and, for resistivities higher than 1011
ohm-cm, difficulties are likely to be encountered. There is some divergence of opinion between
different investigators on the value of resistivity at which difficulty is likely to be encountered
and it is thought that this is due to a number of factors difficult to control, such as the degree
of packing of the dust, so that in practice different forms of apparatus can disagree to a con-
siderable extent. At the same time resistivities measured by anyone form of apparatus, when
used by an experienced qperator, can be related to precipitator performance.
The electrical resistivity of most dusts and fume depends on the nature and condition of
the surface of the dust particles, rather than on the material of which the dust is composed; the
resistivity is in practice often determined by adsorbed layers of vapour, such as water, sulphuric
acid, or ammonia. These usually arise from reactions taking place in the furnace or vessel to
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-78
which the precipitator is attached; for instance, high-sulphur fuel oil used in firing OH furnaces
can produce sulphuric acid, and this in turn is adsorbed by the dust. Where the dust resistivity
is high, suitable layers to reduce the resistivity of the dust can be provided by the injection of
one of the conditioning agents listed above into the flue before the precipitatOrs. In practice,
however, in this country, it has not so far been found necessary to supply any artificial condi-
tioning agent to red oxide dust plant, although difficulties have been reported from abroad.
Figure VI-39 shows the resistivity plotted against temperature for fume originating from LD
converters, OH furnaces, arc furnaces, and ladle desiliconization processes.
It will be seen that the resistivity is below 1011 ohm-cm in all cases except for fume
from the arc furnace. In the case of OH furnaces the dust is normally 'conditioned' by the
water vapour and sulphur trioxide resulting from the combustion of the fuel used to fire the
furnace.
In the case of the arc furnace, there is normally no such supply of conditioning agent in
the gases leaving the furnace as curve I, which is the resistivity of a dust sample taken
immediately at the furnace outlet and is typical of a highly resistive dust without the condi-
tioning surface layer. When a precipitator is attached to an arc furnace it is necessary, in
view of the high temperatures involved, to cool the gases, usually by means of a water spray
tower, to an economical level for the precipitator. This has the effect of reducing the gas
volume to be treated; and at the same time, the dust is 'conditioned' by water vapour and the
resistivity curve assumes the shape shown for the other fume with the peak value below the
limit for efficient precipitation.
Effect of Particle Size on Precipitator Efficiency
The same British report(13) describes the following:
It can be shown from calculations on the forces acting on charged particles that the
efficiency of an electrostatic precipitatOr should decrease with decreasing particle size; this,
however, is not normally borne out in practice and many commercial applications of precipi-
tation are on processes in which much of the fume is submicron, as for instance, blast-furnace
gas cleaning and the red oxide fume evolved from oxygen blowing processes.
Figure VI-37 shows the relationship with precipitation efficiency and particle size for a
number of dusts, the first three relating to dry precipitation, and number 4 to a wet precipi-
tator. Curve No.1 was obtained on a dust whose electrical resistivity was of the order of
1013 ohm -em, and the precipitator was exhibiting signs of severe reverse ionization; it is
interesting to note that the fall-off in efficiency becomes increasingly serious for particle
sizings below 20 ].1m. Curve No.2 was obtained upon the same dust when the resistivity had
been decreased by the use of a conditioning agent, while curve No.3 was obtained on a dust
whose resistivity was well below the limit of 1011 ohm -cm quoted in the section on
resistivity.
Curve No.4, obtained on the wet electrofilter, has a fall-off comparable with the lowest
resistivity dry dust (curve No.3); in this case, although the dust was initially highly resistive, the
resistivity of the dust in the precipitator was reduced to a safe level by the cooling and natural
conditioning effect of the spray rower preceding the precipitator; in addition, since the particles
were deposited on a moving flow of water, the effect of high resistivity would be of no conse-
quence in any case. Since this fume consisted of non-magnetic particles of spherical form it
was possible, using the electron microscope, to continue the grading beyond the limit of most
forms of grading apparatus; these gradings indicated that tpere was no serious fall-off of precipi-
tator efficiency for particle sizings down to the order of 100 ].1m,
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-79
While the theory of the motion of the dust particles under the effect of the electric field
assumes that the dust is deposited as individual particles, there is in practice a strong tendency
for very fine fume to agglomerate into masses consisting of hundreds of fine particles, such
agglomerates behaving as single, much larger particles in the electric field, with the result that
efficiency is higher than would be theoretically calculated for such a fume. This is normally
considered to be one of the explanations why the precipitator fails to obey the basic theory. It
is also one of the difficulties of carrying out dust gradings and limits their value, since clearly
what is required of a dust grading apparatUs is the grading including the effect of agglomeration,
and one of the debatable points in dust grading methods is the energy which should be used to
disperse the agglomerates formed in the precipitators in a dust grading apparatUs. An interesting
featUre is the action of the conditioning agent, as illustrated by curves 1 and 2, since it would
appear that, in addition to reducing the resistivity of the dust, the agglomerating properties
are also materially improved. The authors consider that for efficient precipitation it is neces-
sary, particularly in a dry precipitator, for the dust to have the correct agglomerating properties
in addition to a suitable electrical resistivity value.
Adaptability of Particulate Removal Systems
to Removal of Gaseous Pollutants
Studies on the injection of dry, powdered limestone, dolomite, maganese dioxide,
alumina, and other metal oxides to process gases containing sulfur dioxide indicate that
some 30 to 60 percent of the SOZ can be absorbed by the additive and removed in the
particulate-removal system (as an added inlet loading).
A bag filter could do this effectively. A wet scrubber system gives the added
benefit of a liquid-:absorbtion stage, and has yielded good test re suIts. Alkaline solutions
may be used without the powder injection.
A catalytic oxidation process unit could be inserted in series following a precipi-
tator so the high temperature at which the precipitator operates could be used in the
oxidation. The process scrubbing would preclude following with a baghouse or
precipitator.
These systems are in the development phase, and their use is contingent on eco-
nomics and competitive-process developments.
BATTELLE MEMORIAL INST.ITUTE - COLUMBUS LABORATORIES

-------
FIGURE VI-38.
VI - 8 0
-.---"."..--...
I Arc furnace
1 lD conyett~r .
~ Open. hearth furnace,
AU11toilo
. Open-heonh furnace
~ Destllcon1zotlon
ladle proce n
o Open.heorth furnoce
90
,
l
o
~
>'
U
~
U
~
z
o
~
a:
U
~

8l
I
JL'

2 "
Hiqhly re.isti-.e dusl Od3ohm-cm)
qivinq serious reverse ionization
>-
~
;;
:;; .
~ to
'"
~
2 DJst as I after use of conditioninq
oqeni to reduce re.istivity
Dry type precipitator

3 Normal well conditioned dust
~
" Hiqhly re.isti-.e dust usinq wet
type precipitolor
10'
.1. J ....-!-.. _.J_~,. .1..--:- I -_.'--_J --- L
8 ~ ~ ~ ~ ~ M n ~
PARTICLE SIZE . Jim
J
b
100
200 JOO 400
TEMPERATURE. .C
100
600
VARIA TION OF PRECIPITATION EFFICIENCY
WITH PARTICLE SIZE
FIGURE VI -39.
ELECTRICAL RESISTIVITY OF RED OXIDE
FUME FROM VARIOUS OXYGEN-BLOWN
STEELMAKING PROCESSES
..BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-81
REFERENCES FOR SECTION VI
(1) Kudlich, R., (Revised by Burdick, L. R.), "Ringlemann Smoke Chart", U. S.
Bureau of Mines, Information Circular 7718 (1955).
(2) I'Maximilien Ringlemann - Man of Mystery", Air Repair, ~ (2), 4-6 (November
1952).
(3) McShane, W. P., and Bubba, E, "Automatic BOF Stack Monitoring", 33/The Mag-
azine of Metals Processing, ~ (5), 97-104 (March 1968).
(4) Communication to John Varga, Jr., Battelle Memorial Institute, Columbus Labora-
tories, December 20, 1968.
(5) Silverman, L., "Predicting Performance of Collector s in Air Pollution Control",
Journal of the Air Pollution Control Association, ]2 (12), 573 (December 1963).
(6) Communication from R. B. Engdahl to John Varga, Jr., Battelle Memorial Institute,
Columbus Laboratories, February 3, 1969.
(7) 11 Dust Collectors: A Look at Wet vs Dry Systems", 33/ The Magazine of Metals
Producing, 4 (6), 57-64 (June 1966).
(8) Vajda, S., "Blue Ribbon Steel With Blue Skies", Iron and Steel Engineer, 45 (8),
71-75 (August 1968).
(9) Langer, G., "Ice Nuclei Generated by Steel Mill Activity", Proceedings of the First
National Conference on Weather Modification", 220-227, April 28 - May 1, 1968,
Albany, New York.
(10) Gottschlich, C. F., "Removal of Particulate Matter from Gaseous Wastes-
Electrostatic Precipitators", American Petroleum Institute, New York, 42 pp.
(1961). ------
(11) White, H. J., "Industrial Electrostatic Precipitation", Addison-Wesley Publishing
Company, Inc., Reading, Mass. 1963.
(12) Stern, A. C., Editor, "Air Pollution" Volume III, Academic Press, New York,
437-456 (1968).
(13) Watkins, E. R., and Darby, K., "The Application of Electrostatic Precipitation to
the Control of Fume in the Steel Industry", Special Report 83, Fume Arrestment,
The Iron and Steel Institute, 24-35 (1964).
(14) Sproull, W. T., and Nakada,
Moisture and Temperature",
1350-1358 (June 1961).
Y., "Operation of Cottrell Precipitators-Effects of
Industrial and Engineering Chemistry, 43 (6),
(15) Lagarias, J. S., "Predicting Performance of Electrostatic Precipitators", Journal
of the Air Pollution Control Association, ]2 (12), 595-599 (December 1963).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-82
(16) Frame, C. P., and Elson, R. J., "The Effects of Mechanical Equipment on Con-
trolling Air Pollution at No.3 Sintering Plant, Indiana Harbor Works, Inland Steel
Company", Journal of the Air Pollution Control Association, 13 (12), 600-603
(December 1963). -
(17) Young, T. A., "Gary Steel Works Experience With Dust Control at Number 3 Sinter
Plant", Blast Furnace and Steel Plant, 56 (12), 1057-1063 (December 1968).
(18) Akerlow, E. V., "Modifications to the Fontana Open Hearth Precipitators", Pro-
ceedings, Semi-Annual Technical Meeting, Air Pollution Control Association,
Houston, Texas. 59-72 (December 1956).
(19) Speer, E. B., "Operation of Electrostatic Precipitators on O. H. Furnaces at
Fairless Works", Special Report No. 61, Air and Water Pollution in the Iron and
Steel Industry, The Iron and Steel Institute (1959), pp. 67-74.
(20) Schneider, R. L., "Engineering, Operation and Maintenance of Electrostatic Pre-
cipitators on Open Hearth Furnaces", Journal of the Air Pollution Control Associa-
tion, ~ (8), 348-354 (August 1963).
(21) Dickinson, W. A., and Worth, J. L., "Waste Gas Cleaning at Sparrows Point
Plantls No.4 Open Hearth", AIME Open Hearth Proceedings, 47, 214-225 (1964).
(22) Smith, W. M., et al., "The Use of a Flow Model in the Design of an Electrostatic
Precipitator", Blast Furnace and Steel Plant, 55 (12), 1097-1102 (December 1967.).
(23) "Inland Steel Completes $7 Million Air Pollution Control Facilities", American
Metal Market, p. 27, June 26, 1968.
(24) "Joy Building Youngstown Precipitators", American Metal Market, p. 8, February
7, 1969.
(25) Peterson, H. W., "Gas Cleaning for the Electric Furnace and Oxygen Process Con-
verter", AIME Electric Furnace Proceedings, ~, 262-271 (1956).
(26) Smith, J. H., 'lAir Pollution Control in Oxygen Steelmaking", AIME Open Hearth
Proceedings, 44, 351-357 (1961).
(27) Rowe, A. D., et al., "Waste Gas Cleaning Systems for Large Capacity Oxygen Fur-
nace Plants", Second Interregional Symposium on the Iron and Steel Industry, United
Nations Industrial Development Organization, Moscow, September 19 -
October 9, 1968. 35 pp.
(28) Wheeler, D. H., "Fume Control in L-D Plants", Journal of the Air Pollution Con-
trol Association, ~ (2), 98-101 (February 1968).
(29) "u. S. Steel BOFs at Gary Outstrip Their Design", Steel, 202 (10), 12
(September 5, 1968).
(30) "Pittsburgh Steel Co. to Maintain Regular Deliveries Despite Blast", American
Metal Market, p. 9 (May 15, 1968).
(31) Stairmand, C. J., 'IRemoval of Grit, Dust, and Fume From Exhaust Gase s
Chemical Engineering Processes'l, . The Chemical Engineer, No. 194,
pp. CE 310-CE 324 (December 1965). .
BATTELLE MEMORIAL INSTITUTE - COLlIMI=III~ I Al=ln~AT("'~I~~
From

-------
VI-83
(32) Campbell, W. W., and Fullerton, R. W., l1High-Energy Scrubbers Can Satisfac-
torily Clean Blast Furnace Top Gas", AIME Blast Furnace, Coke Oven, and Raw
Materials Proceedings, ~, 329-334 (1959).
(33) Air Pollution Manual, Part II-Control Equipment, American Industrial Hygiene
Association, Detroit, Michigan, p. 63 (1968).
(34) Semrau, K. T., "Dust Scrubber Design - A Critique on the State of the Art",
Journal of the Air Pollution Control Association, .!2 (12), 587-594, December 1963.
(35) Lapple, C. E., and Kamack, H. J., "Performance of Wet Dust Scrubbers",
Chemical Engineering Progress, ~ (3), 110-121 (March 1955).
(36) Bowman, G. A., and Houston, R.. B., "Recycled Water Systems for Steel Mills",
Iron and Steel Engineer, 43 (11), 139-147 (November 1966).
(37) Communication from R. B. Engdahl to John Varga, Jr., Battelle Memorial Institute,
Columbus Laboratories, February 28, 1969.
(38) Communication from P. D. Miller to John Varga, Jr., Battelle Memorial Institute,
Columbus Laboratories, February 28, 1968.
(39) Harris, E. R., and Beiser, F. R., "Cleaning Sinter Plant Gas With a Venturi
Scrubber", Journal of the Air Pollution Control Association, 15 (2), 46-49
(February 1965). -
(40) "McLouth Steel Using Scrubbers at Sinter Plant", Blast Furnace and Steel Plant,
54 (11), 1051-1052 (November 1966).
(41) Lowe, J. R., "An Orifice Gas Washer", AIME Blast Furnace, Coke Oven, and Raw
Materials Proceedings, .!i., 28-30 (1957).
(42) Reid, G. E., "Experience in Cleaning Blast Furnace Gas With the Orifice Washer",
Iron and Steel Engineer, E. (8), 134-137 (August 1960).
(43) Hipp, N. E., and Westerholm, J. R., "Developments in Gas Cleaning-Great Lakes
Steel Corp.", Iron and Steel Engineer, 44 (8), 101-106 (August 1967).
(44) Morgan, E. R., et a1., liThe Rejuvenated Blast Furnace", Blast Furnace and Steel
Plant, ~ (7), 625-631 (July 1962).
(45) Eberhardt, J. E., and Graham, H. S., "The Venturi Washer for Blast Furnace
Gas", Iron and Steel Engineer, 32 (3), 66-71 (March 1955).
(46) Basse, B., "Gases Cleaned by the Use of Scrubbers", Blast Furnace and Steel
Plant, 44 (11), 1307-1312 (November 1956).
(47) Morgan, M., "Industrial Waste Treatment-Steel Plants", Iron and Steel Engineer,
37 (7), 70-74 (July 1960).
(48) Bishop, C. A., et a1., "Successful Cleaning of Open Hearth Exhaust Gas With a
High Energy Scrubber", .Journal of the Air Pollution Control Association, .!..!. (2),
83 -87 (February 1961).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-84
(49) Johnson, J. E., IIWet Washing of Open Hearth Gases", Iron and Steel Engineer,
44 (2), 96-98 (February 1967).
(50) Broman, C. U., and Iseli, R. R., "The Control of Open Hearth Stack Emissions
With a Venturi Type Scrubber'l, Blast Furnace and Steel Plant, 56 (2), 143-147
(February 1968). -
(51) Pettit, G. A., "Electric Furnace Dust Control Systemll, Journal of the Air Pollu-
tion Control Association, .!2 (12), 607-621 (December 1963).
(52) "Swindell-Dressler to Furnish Armco Furnaces", 33/ The Magazine of Metal Pro-
ducing, ~ (7), p. 18 (July 1967).
(53) 11 Armco Plans Expansion Program at Houston", Iron and Steel Engineer, 43 (3),
p. 180 (March 1966).
(54) Finney, Jr., J. A., and DeCoster, J., IIA Cloth Filter Gas Cleaning System for
Oxygen Convertersll, Iron and Steel Engineer, 42 (3), 133-139 (March 1965).
(55) First, M. W., and Silverman, L., IIPredicting the Performance of Cleanable
Fabric Filters", Journal of the Air Pollution Control Association, 13 (12), 581-586
(December 1963). -
(56) Herrick, R. A., 11 Theory, Application of Filter Drag to Baghouse Evaluation",
Air Engineering, .!.Q (5), 18-21 (May 1968).
(57) 'ISinter Line Baghouse Collector Still Going Strongll, Iron and Steel Engineer, 45
(2), p. 124 (February 1968).
(58) Herrick, R. A., IIA Baghouse Test Program for Oxygen Lanced Open Hearth Fume
Controlll, Journal of the Air Pollution Control Association, 13 (1), 28-32
(January 1963). -
(59) Herrick, R. A., et al., "Oxygen-Lanced Open Hearth Furnace Fume Cleaning With
a Glass Fabric Baghouse", Journal of the Air Pollution Control Association, 16 (1),
7 -11 (January 1966). -
(60) Campbell, W. W., and Fullerton, R. W., IIDevelopment of an Electric Furnace
Dust Control Systemll, Journal of the Air Pollution Control Association, 12 (12),
574-590 (December 1962). -
,
(61) Bintzer, W. W., IIDesign and Operation of a Fume and Dust Collection System for
Two 100-Ton Electric Furnaces'l, Iron and Steel Engineer, 41 (2), 115-123
(February 1964). -
(62) Wood, R. M., and Burcham, J. 0., IIArc Furnace Steel Production, Kansas City
Works-Armco Steel Corp. 11, Journal of Metals, ~ (12), 1005-1008 (December 1964).
(63) Bintzer, W. W., and Kleintop, D. R., IIDesign, Operation and Maintenance of a
150-Ton Electric Furnace Dust Collection Systemll, Iron and Steel Engineer, 44 (6),
77-85 (June 1967). -
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VI-85 and VI-86
(64) Wilcox, M. W., and Lewis, R. T., "A New Approach to Pollution Control in an
Electric Furnace Shop", Iron and Steel Engineer, 45 (12), 113-120 (December 1968).
(65) Venturini, J. L., "Historical Review of the Air Pollution Control Installation at
Bethlehem Steel Corporation's Los Angeles Plant", Pre print No. 68-134, Air Pol-
lution Control Association Annual Meeting, St. Paul, Minnesota (June 23 -2 7, 1968),
19 pp.
(66) Stastny, E. P., "Choosing Your Electrostatic Precipitator", Power, 104 (1),
61-64 (January 1960).
(67) Stairmand, C. J., "Dust Collection by Impingement and Diffusion", Transactions
of the Institution of Chemical Engineers, 28 (1950) pp.
(68) Robinson, M., "A Modified Deutsch Efficiency Equation for Electrostatic Precipi-
tation", Atmospheric Environment, Permagon Press, Vol. 1, (1967) pp. 193 -204.
(69) Semrau, K. T., "Correlation of Dust Scrubber Efficiency", Journal of the Air
Pollution Association, 10 (3), (June 1960), pp. 200-207.
(70) Communication from the Pangborn Corporation to Swindell-Dressler Company.
(71) Elliott, A. C., and Lafreniere, A. J., "Collection of Metallurgical Fumes
Oxygen Lanced Open Hearth Furnaces", Journal of Metals, .!..!!. (6), 743-746
( June 1 96 6 ) .
From
(72) Penney, G., "Symposium on Gas -Solids Separation", Carnegie-Mellon University,
Pittsburgh, Pa., January 14, 1969.
(73) Berg, B. R., "Development of a New, Horizontal-Flow, Plate-Type Precipitator
for Blast Furnace Gas Cleaning", Iron and Steel Engineer, 36 (10) 93-101
(October 1959).
(74) Willet,. H. P., and Dike, D. E., "The Venturi Scrubber for Cleaning Oxygen Steel
Process Gases", Iron and Steel Engineer, 38 (7), 126 (July 1961).
(75) American Air Filter Company Bulletin 294-1 OM-3 -65 -CPo
(76) Buffalo Forge Company Bulletin AP650.
(77) Robinson, M., "Turbulent Gas Flow and Electrostatic Precipitation", Journal of 
the Air Pollution Control Association, .!..!!. (4) 235-239 (April 1968).
(78) Stairmand, C. J., "Design and Performance of Modern Gas -Cleaning Equipment",
Journal of the Institute of Fuel, 29 (181) 58-76 (February 1956).
(79) Punch, G., "Gas Cleaning in the Iron and Steel Industry, Part II: Applications",
Fume Arrestment, Special Report No. 83, Iron and Steel Institute, p. 10 (1963).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VII-l
SEC TION VII
PROBLEMS AND ASSOCIATED
OPPOR TUNITIES FOR RESEARCH
An objective of this study is to determine where resea'rch and development may be
undertaken to achieve the desired control of emissions to the air from the integrated
iron and steel industry. As a means for screening the various process segments with
respect to research needs, an identification key was developed to select the subjects
requiring research effort to resolve its air-quality problems, and to establish some
indication of priority. The key is given in Table VII-I.
TABLE VII-I.
IDENTIFICATION KEY FOR
PROBLEM IDENTIFICATION
Factor  Key 
 . ~ D QJ
Level of emission Severe Moderate Minimal 
Is it controlled? No Partially Yes 
Can it be controlled?  Partially Yes Uncertain
The key is explained as follows:
Level of emission -
Seve re -
Emission is (1) large in volume, (2) detrimental regardless
of the volume, or (3) particularly obnoxious with respect
to odor. As examples: (1) a large volume of emissions
would be the iron-oxide fume generated in uncontrolled
BOF operation, (2) detrimental in small amounts would
be applied to fluoride emissions, and (3) the evolution
of hydrogen sulphide from quenched blast-furnace slag
would be an example of a particularly obnoxious odor,
Moderate
- Emission is (1) modest in volume, (2) somewhat detri-
mental regardless of volume, or (3) slightly obnoxious or
disagreeable with respect to odor. As examples: (1) the
particulates generated during the quenching of coke,
(2) acid fumes from partially hooded pickling operations,
and (3) benzol odors downwind from a coke plant.
Minimal -
Emission is (1) at a low volume, (2) slightly detrimental
when in large volumes, but generally tole rabIe, and
(3) any odors classed as a minor nuisance. As examples:
(1) particulates generated during the indurating of pellets,
and (2) gases from mold coatings during ingot pouring.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VII-2
Is it controlled?
- This factor in the identification key is concerned with the
apparent average for the entire integrated iron and steel
industry.
No
The industry on the average does not control emissions to
the air from the process or from the process segment.
Partially - The steel industry generally does some control of emis-
sions to the air.
Yes
- The process or process segment usually appears to be
controlled adequately.
Can it be controlled? -
No
Present pollution-control technology is not adequate for
successful control.
Partially - Present pollution-control technology has some limitations
that interfere with successful control.
Yes
- Emissions can be controlled with present control-
equipment and technology.
Uncertain - Insufficient information available to determine if control
is pos sible.
It should be pointed out that under the factor "Can it be controlled? I', no process
or process segment is considered as absolutely uncontrollable with respect to emissions.
Priority for the conduct of research and development was evaluated on the basis
shown in Table VIl-2. The priority rating is based on a combination of the level of
emis sion, whether it is controlled, and whether it can be controlled. The first priority
is for processes or proces s segments having a severe level of emissions, no control or
only partial control, and the control technology is, for practical purposes, unknown or
only partially effective. Second or third priority is determined on the basis of emission
level, with the same considerations for control and the possibility of control. Any pro-
cess or process segment listed as controllable with respect to emissions by use of
present technology is not considered as an appropriate area for research and develop-
ment activities.
Ratings for the various process segments considered in this study are shown in
Tables VIl-3 through VII-8. Research and development priorities are determined on the
basis of the indicated ratings. Emissions from the process segments are designated as
follows: ~ - particulates, @) - gaseous, and ~ - aerosols.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VII-3
 TABLE VII-2. PRIORITY FOR RESEARCH AND
  DEVELOPMENT EFFOR TS 
Level of Is It Can It Be  
Emission Controlled? Controlled? Priority Rating
. . [?J  
. . ~ First
. ~ [?J  
~ . [?J  
~ . ~  
    Second
~ ~ [?J  
~ ~ ~  
0 . [?J  
0 . ~ Third
o ~ [2]  
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VII-4
TABLE VII-3. RESEARCH AND DEVELOPMENT EVALUATION FOR MAKING
PELLETS AND FOR RAW-MATERIAL HANDLING
          
-------
VII-S
TABLE VII-4. RESEARCH AND DEVELOPMENT EVALUATION FOR COKE MAKING
         iJ
        r?
        ~ 
        ~ 
        ~ 
        l:: 
        :£ 
        ~ 
Coke Plant Coal transfer - stOrage to coke plant 0 Wind . . OJ III-41, V-2, VII-12,
       A-4, C-16, C-26
 Internal transfer to crushing plant 0 Handling . ~ 0 III-41, VII-12, C-16
 Crushing and grinding 0 Process . ~ 0 III -41, IV -10, VII-12,
  C-16  
 Internal transfer to stOrage 0 Handling . ~ 0 III -41, VII-12, C-16,
       C-26  
 Coal transfer - storage to larry car m Handling . ~ 0 III-41 , V-3, VII-12,
       C-16  
 - larry car to coke oven 0 Handling . ~ IJJ III "41, IV-10, V-3,
 VII -12 
  ~ Open oven . ~ OJ V-3, C-16 
  @] Lids and ~ ~ E&3 III-41, IV-10, IV-n,
 Coking operation - oven door seals V-2, VII-12, C-14,
      C-18, C-27 
 - underfiring @] Fuel. rn ~ 0 III-41, V-2, V-31,
  C-18  
   Abrasion,    III-41, IV-10, V-3,
 Pushing coke 0 thermal . . OJ VII -12, C -19 
   draft    
  @] Incomplete . f2?J 0 C-19  
  coking   
 Quenching coke 0 Thermal ~ . ~   
 draft III -41, V-3, VII-12
 Transfer to stOrage 0 Handling 0 ~ IJJ III -41, V-4, C-20
 Crushing and screening m Process ~ 1221 0 III -41, VII-12, C-20
 Coke-oven gas system @] Leaks 0 0  III -41, VII-12, C-21
 By-product plant @] Leaks . ~ 0 III -41, V-4, VII-12,
  C-21  
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VII-6
TABLE VII-5. RESEARCH AND DEVELOPMENT EVALUATION FOR MAKING
SINTER AND MAKING IRON
         ~
         ~
         ...
-------
    VII-7      
TABLE VII-6. RESEARCH AND DEVELOPMENT EVALUATION FOR MAKING STEEL AND POURING INGOTS
          ,fJ
          u
          ~
          ~
          i}
          ~
          :::
         ~
         ~ 
Open-Hearth Refractory maintenance 0 Handling D ~ D V-l1 
Steelmaking    m  f22] ~ D  
 Charging scrap  Handling III-43, V-lO, C-54
         III-43, V-lO, V-38, VI-8,
 Preheating and melting m Material ~ l88! D VI-18, VI-2l, VI-26, VI-27,
  VI-28, A-13, C-62, C-66
    @] Fuel ~ .~ D III -43, V-lO, C-62
 Hot-metal addition  m Chern istry . ~ D III -43, V-lO, C-62
 Oxygen lancing  0 Process . ~ D III -43, V-l1, C-64
    @] Process ~ ~ D V-l1 
 Tapping  m Handling ~ . D \1-12 
    IT]  l22J 0  III-43, IV-15, VI-9, VI-l1,
BOF Charging scrap  Handling  VI-18, VI-26, VI-27,
Steelmaking         VI-28, A-16, C-66
 Charging hot metal  m Chemistry . ~ D III-43, V-12
 Oxygen lancing  m Process . ~ D III-43, V-12, C-69, C-70,
   C-7l. C-73
    @] Process . ~ D V-12, C-70, C-73, C-77
    m  ~ ~ D III-43, IV-16, IV-17, IV-18,
Electric - Furnace Charging scrap  Handling V-14, VI-2l, A-18, C-79,
Steelmaking         C-88 
 Charging hot metal (if used) 0 Chemistry . f82j D III -43, C -91
    m  ~ ~ D III-43, V-14, VI-12, VI-20,
 Melting  Process VI -26, VI -27, VI -28,
    o  . B23 D VI-29, C-9l
 Oxygen lancing  Process III-43, C-83
    @] Process ~ ~ D V-15, C-84
Pouring Ingots Pouring without hot tops 0 Oxidation D . D III -43, III -44, C -104,
 C-l06 
    @] Mold ~ . D V-12, A-24, C-l06
    coating
    @] Process D . D V-12, C-l06
 Pouring with hot tops  @] Material ~ . D C-l06, C-112
    m Material ~ . D C-l06 
 Leaded steel additions m Mater{al . D  C-l08 
 Ingot stripping  0 Process n  D C-l08 
 BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES  

-------
VII-8
TABLE VII-7. RESEARCH AND DEVELOPMENT EVALUATION FOR PRIMARY ROLLING,
CONTINUOUS CASTING, PRESSURE CASTING, AND CONDITIONING
       qJ
       ~
       ~
       ~
       ~
       !::
       £
       ~
Rolling Billets, Maintenance - soaking pits [E] Handling ~ ~ 0 
Blooms, and       
Slabs Firing - soaking pits @] Fuel ~ ~ 0 
 Charging and removing ingots [E] Process D . IT] 
 Primary rolling [E] Oxidation 0 0  
Continuous Flow of molten steel into machine [E] Oxidation 0 ~ 0 IV-19
Casting  @]  0 0  
  Oxidation  
 Torch cutOff 0 Process D . 0 C-110
  @] Process D . 0 
Pressure Casting Flow of metal into mold [E]  0 0  IV-19, C-1l2
 Flow of metal into hot top 0 Material D . 0 
  @] Material D . 0 C-1l2
 Torch cutting to remove riser 0 Process D . 0 C-113
Auxiliary Preparation of ingot molds [E] Process ~ . IT] 
Operations  [E]  ~ . IT] 
  Material C-I06
 Preparation of pressure casting mold 0 Process 0 0  
  m Material D . IT] 
 Pigging of molten iron 0 Chemistry . ~ IT] 
Conditioning Grinding and chipping [E] Process 0 ~ 0 
Semifinished       
Products Spot scarfing 0 Process ~ ~ 0 
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VII-9
TABLE VII-8. RESEARCH AND DEVELOPMENT EVALUATION FOR HOT ROLLING, COLD ROLLING, COATING
OF FINISHED PRODUCTS, WASTE INCINERATION, AND POWER GENERATION
         ,if
         (J
        ~
        'lJ"
        ;} 
        Qf 
        ::: 
        ~ 
        ~ 
  ~  ~ D  III-44, III-45 , IV-23,
Finished Heating for rolling Fuel  V-41, V-18, C-1l4
Products       IlI-44, V-18, VI-33, A-27,
 Hot scarfing prior to rolling 0 Process . ~ 0
 C-114, C-116, C-117
 Hot rolling IT] Oxidation ~ ~ 0 III-45 , lV-20, V-18, A-27,
 C-114, C-117
 Pickling for scale removal 0 Acid mist ~ 0  IV-23, V-18, C-117. C-118
 Shotblasting for scale removal 0 Proc ess ~ D  C -115  
 Cold rolling 0 Rolling oil 0 ~ 0 IlI-45, IV-22, V-18, C-117
 Hot galvanizing 0 Cover flux ~ 0  III-45 , IV-22, V-19, C-121
 Electro-gal vanizing IT] Cover flux D 0  IV-23, V-20, C-120
 Electro-tin plate 0  0 0  III -45, IV-22, V-20
 Paint coating 0 . Proc ess ~ 0  IIl-45 ,  I V -22 
 Plastic coating m Process ~ 0  IlI-45 , IV-22 
Waste Disposal of in-plant generated wastes 0  ~ ~ 0 A-28, A-29, C-124
Incineration         
  @]  ~ ~ 0   
Power In-plant generation of electric energy 0  ~ ~ 0   
Generation         
  @]  r?&J ~ IT]   
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VII-IO
First Priority for Research and Development
Process or process segments rated to warrant first-priority efforts are char-
acterized by the following key designations:
Level of Is It Can It Be
Emission Controlled? Controlled?
. . !2]
. . ~
. ggJ !2]
Examination of the processes in the tables shows the following to fall into the
first-priority classification:
(1) Unloading and transfer to storage of fine ore or pellets
(2) Unloading and transfer to storage of coal
(3) Coke plant - oven charging (particulates and gases)
(4) Coke plant - pushing coke
(5) Sinter plant - ignition and firing of sinter
(6) Ironmaking - (a) casting iron, (b) flushing slag, (c) slag disposal
(7) Pigging of molten iron.
Second Priority for Research and Development
Process or process segments rated to warrant second-priority efforts are char-
acterized by the following key designations:
Level of
Emission
Is It
Controlled?
Can It Be
Controlled?
~
~
~
~
.
.
~
~
!2]
[8B
!2]
~
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VII-ll
Examination of the processes in the tables shows the following to fall into the
second-priority designation:
(1) Unloading coarse ore
(2) Coke plant - gaseous emissions from lids and door seals
(3) Coke plant - particulates during quenching of coke
(4) Sinter plant - gaseous emissions from fuel
(5) Making iron - transfer of bulk materials from storage to stockhouse
(6) Preparation of ingot molds
(7) Gaseous emissions from in-plant generation of electric energy.
Third Priority for Research and Development
Process or process segments rated to warrant only third-priority efforts are
characterized by the following key designations:
Level of Is It Can It Be
Emission Controlled? Controlled?
D . IT!
o . ~
D ~ [2]
Examination of the processes listed shows the following to fall into the third-
priority designation:
(l) Transfer of coarse ore
(2) Unloading and transferring of limestone
(3) Coke plant - handling of coke
(4) Coke plant - coke-oven-gas system
(5) Sinter plant - transfer of sinter to storage
(6) Making iron - charging from skip hoist to blast-furnace top
(7) Charging and removing ingots from soaking pits
(8) Preparation of pressure-casting molds.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VII-12
Evaluation of Research and Development Priorities
The problems, possible solutions, and implications of the changes that are re-
quired and which may occur as a result of research and development, are discussed in
Table VII-9. Making of coke is considered first because it presents major problems
requiring solution at three levels of priority. Sinter-plant operation is considered
second because it also presents problems at all three levels of priority, but these
appear, overall, less severe than those for coke making.
Coke Making
The unloading of coal from barges and movement of coal to the storage areas pre-
sents a serious particulate-emission problem because of the fineness of the coal dust and
the ease with which it can become airborne once it is in a stockpile. The unloading op-
eration can possibly be improved by the development of continuous bucket unloaders that
would transfer the coal to shrouded conveyor belts for transport to storage areas. Such
an approach would keep the particulates contained, but in all probability the shrouded
conveyor belts would also have to be hooded and exhausted to remove any airborne par-
ticulates from the system. If this were not done, excessive wear of the conveyor-belt
rollers would probably occur, with resulting high maintenance costs. This type of sys-
tem would also have a tendency to restrict the expanse of storage space. Moisture
additions at the first transfer point, which would be stationary, should tend to minimize
dust generation at subsequent transfer points.
The containment of coal in the stockpiles and prevention of coal dust from becom-
ing airborne is a serious problem. Attempts have been made to keep dust down by
moisture additions to the piles, as well as by spraying the piles with plastic films.
Neither method has proved to be satisfactory. An ideal additive to the coal stockpiles
would be one that would keep a protective film over the stockpile, would heal itself if
broken, would not be detrimental to the crushing, grinding, and coking of coal, and
could be used to control the bulk density of the coal.
The transfer of coal from storage to the coal-grinding plant is a problem similar
to that of unloading and transfer of coal to stockpiles. The Japanese have reportedly
resorted to a reclamation system that utilizes underground systems to recover the coal
from the stockpiles, thereby eliminating the problem of agitating the coal pile.
Particulate emissions from the crushing, grinding, and transport of coal to the
storage bins above the larry cars usually are under control, and particulate emissions
from this source should not be present. Grinding mills and conveyor systems can be
serviced with efficient emission-control equipment. The recovered coal dust is a plus,
because any lost dust would mean a loss in yield for each ton of coal ground, with an
attendant increase in the cost of coke.
Charging of the coke ovens from the larry car is a persistent problem, and one of
constant concern to the coke-oven operator. All lids must be open during the charging
operation, with the result that dust which becomes airborne inside the coke oven due to
the thermal draft finds an easy path to the atmosphere via any of the open ports. Con-
stant maintenance of automated lid-lifting and closing equipment and rapid charging
have been the major means of minimizing emis sions from this source. The new
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VII-13
TABLE VII-g. ANALYSIS AND POSSIBLE SOLUTIONS TO COKE-PLANT AIR-POLLUTION PROBLEMS
     Process and
    Technological Equipment Factots
Process Segment Problem Possible Solution Advantages Disadvantages Requiring Solution
Making coke Insufficient control of Complete enclosure and Centralized control and Create industrial hygiene Design, construction,
 particulate and gas- building ventilation exhausting of emissions problems safety, and industrial
 eous emissions  to allowable limits Possible explosion hazard hygiene problems
Unloading of coal Particulate emissions Continuou's bucker un- Contain and minimize Restrict area of operation Design and construction
from barges and caused by clam -shell loading and discharge particu.lates and extent of storage 
transfer (0 storage unloading, free-fall to shrouded belt Permit all-weather area 
 dumping, and wind  operation  
Unloading of coal Particulate emissions Enclosure and exhaust- Contain and minimize Restrict area of operation Design and construction
from railroad cars, caused by car dumping ing of car-dumping particulates and extent of storage 
and transfer (0 or bottom dumping or bortom -dumping  area 
storage  area, and shrouded   
  belt to storage   
S rorage of coal Particulate emissions Unknown   Retention of particulates
 generated by wind    in srorage piles
Transfer of coal Particulate emissions Underground recovery Minimize or even elim- Extensi ve rebuilding of Design and construction
from storage (0 generated by clam -, and conveyor transfer in ate emissions due (0 coal-storage facilities 
coal crushing and shell recovery and to grinding plant coal recovery  
grinding plant free -fall loading of    
 cars    
Coke-oven charging P articulate and gaseous Double aspiration Reduce emissions during Unknown New charging methods
 emissions generated lines charging  
 during charging Shrouded charging Reduce emissions during Possible explosion or 
  pipes and collection charging fire hazard 
  equipment on   
  larry car   
  Pipe -line charging Eliminate emissions  
Co king Leakage of gaseous Improved maintenance Reduce emissions None Design, construction,
 emissions around Improved design in Minimize emissions None and maintenance
 doors, seals, and lids lids, seals, and doors   
Pushing of coke Particulate generated by Hooded exhaust over Contain and exhaust Possible interference Method and/or equip-
 abrasive action of coke push side ,particulate and gas- with pushing operation ment for capturing
 on oven brick, and  eo us emissions  and collecting
 thermal draft causing    emissions
 dispersion of particu - Fine spray nozzles to Reduce temperature of Possible interference Mechanics of quenching
 lates spray coke and re- the coke as it is with pushing operations and particulate
  duce temperature pushed, reduce the and create a safety emission
   intensity of the ther- problem 
   mal draft, and mini-  
   mize dispersion of  
   particulates  
Quenching of coke Emission of particulates Use of baffles Reduce particulates  Quench tower operation
 during the quenching Redesigned quench Minimize or even elim-  anct characteristics
 of hot coke tower inate particulates  
Coke-oven-gas Leakage of gas Improved gas-plant Minimize leakage  Design and maintenance
recovery system  components   
 High HZS content in Develop economical Minimize or even elim-  
 coke -oven gas method of removing inate sulfur oxide due  
  HZS to burning of coke-  
   oven gas  
By-product re- Leakage of gas and Improved by-product- Reduce emissions  Design and maintenance
covery plant ae rosols plant components   
Making coke Insufficient control of Develop new coke- Eliminate emissions Unknown New process and
 particulate' and gas- making process   equipment
 eo us e missions in    
 present coke plants    
BATTELLE
MEMORI AL
INSTITUTE - COLUMBUS LABORATORIES

-------
VII-14
European larry car which exhausts the emissions into a scrubber system located on the
larry car has not been fully successful because of fire and explosions which have re-
portedly occurred in two Canadian installations. >:< It has been reported that cooperative
work among several steel companies on the improved larry car-emission control system
has been somewhat successful. Details are not yet available, but are expected to be dis-
closed to NAPCA soon, probably by the AISI, which has informally expressed its inten-
tion to request research funding from NAPCA on this subject.
An American company (Allied Chemical Corporation, Wilputte Division) that de-
signs and builds coke ovens has under development a method of preheating coal followed
by charging into the coke oven by means of a pipeline system. The developers feel that
the method will control completely the emissions of particulates and gases to the atmo-
sphere during charging. The company is installing the system on a commercial battery
of coke ovens of its Semet-Solvay Division at Ironton, Ohio. Further information has
not yet been released by the company, but is expected to be released soon. The com-
mercial installation probably will be operational about mid-1970, unless unexpected dif-
ficulties are encountered. Assessment of the advantages and disadvantages of the sys-
tem must await operation on a commercial scale.
Leakage of gaseous emis sions around the lids, doors, and seals is a problem that
can be minimized only through constant maintenance. As coke ovens become older, the
leakage problem becomes aggravated. Coke-plant-construction companies have been
working on the development of improved designs, but under the conditions imposed by
the coking proces s, there has not yet been a substantial return on the effort put into
this part of coke -oven construction.
Pushing of coke results in particulates generated by the abrasive action of the
stove refractories on the coke as it is pushed from the oven. The thermal draft created
by the exposure of the hot coke to the atmosphere carries the particulates into the atmo-
sphere. Various types of hoods and exhaust systems have been tried on the European
continent, but none have operated successfully. The possibility of a prequenching
operation between the coke oven and quench car may be a means of reducing the temper-
ature of the coke, and thereby reducing the intensity of the thermal draft.
The quenching of coke is 'also a source of particulates; installation of baffle!) in the
quench tower has reduced this emission from 60 to 80 percent. Further work is being
carried out on this aspect of particulate control. Another subject of research could con-
cern itself with new designs for quench towers that would be a combination of quench
tower and heat exchanger to (1) prevent the steam generated from going into the atmo-
sphere, and (2) contain the particulates within the restricted area of the quench tower
itself.
New coking processes have been under development for many years, but none have
been able to compete with the existing by-product coke plant. However, the fact that the
coke by-product market has to a great extent been usurped by the petrochemical industry,
and markets for coal by-products have deteriorated, the attraction of the by-product
coke plant has diminished. Coking processes that do not recover the by-products are of
considerable interest in the integrated iron and steel industry today, and a number of new
coke-oven installations may in the future dispense with much of the by-product recovery
that has been traditional until now.
"This report has not been confirmed.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VII-IS
When new coking processes are developed, evaluation of them and of the usefulness
of the metallurgical coke they produce is usually hampered by two facets of scale. First,
the coking pilot plant often is too small to produce large amounts of coke for commercial-
scale evaluation, and second, the integrated steel industry includes only extremely lim-
ited capability to evaluate the performance of the new coke sufficiently to qualify it for
acceptance on a commercial scale. Without such acceptance, it usually is impractical
to authorize construction of a multimillion dollar coking plant based on a new proces s.
The blast-furnace operators usually are not satisfied with anything les s than fdl-scale
evaluation of a new coke. Experience has shown that such a test must be carried out for
a minimum of 1 week to provide valid results for comparison to existing practice. This
means that a blast furnace producing 2000 net tons per day of hot metal at a coke rate of
1200 pounds per net ton requires 1200 net tons of usable coke per day, which is 8400 net
tons for the complete test. Pilot facilities are seldom in a position to supply such re-
quirements. An experimental blast furnace operated by the Bureau of Mines was avail-
able for such tests and required only about 300 tons of coke to carry out the required
trial. Unfortunately, this furnace is no longer operative, and probably will soon be
declared surplus. This will leave in the United States only one small experimental blast
furnace suitable for such evaluations - the furnace owned by United States Steel.
For about a decade or more, the FMC Corporation has had under development a
new process for making coke, and at various times has operated a pilot plant at
Kemmerer, Wyoming, to produce their FMC coke. I This coke is made by a continuous
process involving fluidized beds and briquetting. One application of the coke has been
as a reducing agent in phosphorus furnaces - an application les s demanding than the
blast furnace in terms of mechanical strength of the coke. Recognizing the need for
new methods to produce coke for blast furnaces, United States Steel cooperated with
FMC to manufacture some metallurgical coke by this proces s and evaluate it in one of
the small experimental blast furnaces. The results were reasonably good, but were not
sufficiently convincing with respect to performance and cost to encourage U.S. Steel or
other steel companies to invest the much larger amounts of money needed to expand this
evaluation to a larger scale. Subsequently the Kemmerer pilot plant was closed down,
but there are today unconfirmed rumors that some thought is being given to reopening of
that plant for purposes that are unknown at this time.
At least one American coal company (probably two) have developed new methods
for making coke or char on a continuous basis under conditions that are thought to con-
trol emissions much better than those for conventional coke ovens. Island Creek Coal
Company is thought to have a process that might be of interest for application to the steel
industry, but little is known about this process at the present time. The activities of
Peabody Coal Company along these lines are of more interest. Peabody has a process
that involves pretreatment of coal on a traveling grate in a dutch oven, followed by
treatment in a shaft furnace. The coke is self-agglomerating during the process (like
coke in a conventional oven), rather than briquetted as is done for FMC coke. Through-
put time is understood to be between 1 and 2 hours, rather than about 16 hours as in a
conventional oven. Peabody installed in 1961 a demonstration plant rated at 100 tons
per day at Columbia, Tennessee; they use the plant for process development and sell
the coke for chemical purposes in the vicinity. Most of this coke was not intended to be
a metallurgical grade, but it is understood that some work was done on blast-furnace
coke. The results appeared good to Peabody, but have not been evaluated by blast-
furnace tests. The process looked good enough for some purposes that Monsanto
Chemical Company built a plant in 1966 at Decatur, Alabama. This plant is rated at
500 tons per day; and the coke is used as a reducing agent in phosphorus furnaces.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VII-16
Making Sin te r
Problems, possible solutions, and implications of changes in sinter-plant opera-
tions are given in Table VII-IO. Particulate and gaseous emissions are a problem in
the operation of a sinter plant, because each of the various proces s segments that are
included in the over-all operation has its own particular air-pollution problem that must
be handled. Sinter plants will undoubtedly be in use for many years to come, if for no
other reason than to recover the iron values that are generated as dusts in steelworks
operations.
Crushing and grinding operations associated with the preparation of materials for
sintering are straightforward, and particulate emis sions can be controlled with presently
available equipment. The transport of materials to and from sintering plants falls into
the same category. The areas of major importance in minimizing emissions is in the
ignition and sintering operation of the sinter strand. Particulates are drawn from the
sinter bed by the strong air flow required by the sintering operation. The primary fuel
used to make the sinter is coke in the form of fines. The sulfur content of the coke is
a major source of sulfur emissions to the atmosphere. Coke is the usual fuel in this
particular application, because the coke fines are to a great extent generated during
the crushing and sizing of the coke for blast furnace use. The fines are, therefore,
available at a reasonable cost for the production of sinter. Major problems exist in the
development of existing air-pollution control equipment to handle high-lime particulates
that are generated in the production of highly-fluxed sinters required to meet increased
demands for higher blast-furnace productivity. Wet scrubbers offer a possible solution
to the recovery of the high-lime particulates, as do bag houses. Wet scrubbers also
offer the possibility of recovering the sulfur gases, because the lime cont.ent of the par-
ticulates may act somewhat as a scavenger for the sulfur dioxide. The possibilities
could be explored as a combined method for recovering particulates and sulfur gases in
the same equipment. Work also is advisable to determine if other additives can be intro-
duced into the water to improve the recovery of sulfur dioxide by the lime. Particulates
generated during cooling, crushing, and screening of the sinter can be handled by exist-
ing air-pollution control equipment, and probably do not require further development in
their application. However, all of the foregoing depends to a considerable degree on
better knowledge of the nature and characteristics of particulates and gases evolved from
sintering plants. Therefore, the first step in further research aimed at better control of
emis sions from sintering plants is improved characterization of the amount and nature of
the emis sions. This improved characterization will require measurement, sampling,
analysis, and evaluation of emissions from several sintering plants being operated on a
commercial scale.
Raw Material Storage and Handling
Problems, possible solutions, and implications of change in the handling and
storage of the vast amounts of raw materials required in making iron and steel are
given in Table VII-II. The problem here is one of the most difficult to solve. Gener-
ation of particulates during the handling of raw materials from receiving to final transfer
for processing is a never-ending air-pollution problem. Particulates become air-borne
from the storage piles due to dry and windy weather conditions. The more-apparent
solutions such as complete enclosure and underground recovery systems are certainly
technically feasible, but appear to be economically impractical at this point in time.
Research and development may fruitfully be carried out to produce some type of wetting
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
VII-17
TABLE VII-10. ANALYSIS AND POSSffiLE SOLUTIONS TO SINTER-PLANT AIR-POLLUTION PROBLEMS
     Process and
    Technological Equipment Factots
'rocess Segment Problem Possible Solution Advantages D isad vantages Requiring Solution
Making sinter Insufficient control Complete enclosure Centralized control Limited access for Design, consrruction,
 of particulate and and building and exhausting of maintenance and maintenance
 gaseous emissions evacuation emissions to allow - Industrial hygiene Industrial hygiene
   able limits problem for working problem
    . personnel 
::rushing and Collection of iron    
grinding oxide, limestone,    
 and coke dust    
fr ansport to Iron oxide, limestone, Use of water sprays Minimize generation Actual level of control Design, construction,
sinter strand and coke dusts gener- and detergents at of particulates subject to variation and maintenance
 ated at transfer points transfer points   
  Complete enclosure Collection of particu - Creates secondary dust- 
  and evacuatiori of lates to allowable handling problem 
  transfer system limits  
:gnition and Iron oxide, limestone, Ducted control of Possible reduction of Restrict direct observation Redesign of sinter
sinterihg and lime d usrs combustion air to the volume of air of sinter strand -probably strand
  strand handled with re - require remote 
   suIting increased monitoring 
   collection efficiency  
  Oxygen enrichment Further reduction in Increase maintenance Effect of oxygen enrich-
  of combustion air volume of air handled requirements ment on process
   and further improve- S pace restrictions 
   ment in collection around sinter strand 
   efficiency  
 Sulfur dioxide gener- Reduce sulfur content Corresponding reduc- No known method for Reduction of sulfur
 ated by burning coke of coke tion in S02 reducing sulfur in in coal used to make
    coke coke
  Wet recovery system Minimize or eliminate Corrosion resistant Mechanism of S02
  with additives to S02 emissions to materials required removal
  combine with S02 atmosphere  
  Fabric collectors Minimize or eliminate Close control required Mechanism of S02
  with additives that S02 emissions to to monitor additions entrapment
  combine wirh S02 atmosphere  
Cooling, crushing, Generation of Can be controlled by   Design and construction
and screening particulates application of exist -   
  ing equipment   
Transfer to storage Generation of Enclosure and evacu- Eliminate particulate Increased maintenance Design, construction,
 particulates ation of transfer emissions problems and maintenance
  system   
BATTELLE
MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
In
»
-t
-t
!11
r
r
!11
~
!11
~
o
~
»
r
Z
III
:!
-t
C
-t
!11
I
o
o
r
c
~
In
C
III
r
»
In
o
:u
»
-t
o
:u
!11
III
TABLE VII-ll. ANALYSIS AND POSSmLE SOLUTIONS TO RAW MATERIAL STORAGE AND HANDLING AIR-POLLUTION PROBLEMS
Process Segment
Problem
   Process and 
  Technological Equipment Factors 
Possible Solution Advantages D isad vantages Requiring Solution 
Better methods of Minimize particulate  Design and 
unloading boats emissions  construction 
and barges    
Enclosure and Minimize particulate Increased maintenance,  <:
  ......
evacuation of emissions secondary dust handling  ......
 I
railroad car  problem  ......
  00
dumps    
Complete enclosure Eliminate emissions None Retention of par- 
Self-healing to the atmosphere  ticulates in 
coating Minimize emissions Unknown storage piles 
Underground re- Minimize particulates Increased maintenance Design and 
covery systems   construction 
Unloading and
transfer to
storage of coal,
fine ore, and
pellets
Storage
Recovery from
storage
Generation of particulates
due to handling
Generation of particulates
due to drying of mate-
rials and windy weather
Generation of particulates
due to handling

-------
VII-l9
agent that will minimize the generation of air-borne particulates during storage and
handling. Fortunately, gaseous emissions are not a problem during handling and stor-
age, except for the handling and storage. of blast-furnace slag, which is a special prob-
lem now receiving research attention by the AISI, and which is discussed in the next
paragraph under "Iron Making".
Iroq Making
Problems, possible solutions, and implications of change for emissions generated
during the making of iron are given in Table VII-l2. The most pressing problems
deserving of attention in the making of iron are (l) the generation of hydrogen sulfide
from the quenching of slag, and (2) improved methods for the control of "kish".
Some research has been conducted on the mechanisms controlling the generation
of hydrogen sulfide from quenched blast furnace slags. However, the problem is far
from solved, partially because of the limited funding that has been available to carry out
the research. Further research is required on rapid analytical methods for determining
the concentration of hydrogen sulfide in the air, as well as continued work to further
investigate the mechanisms of formation and possible methods for control.
The evolution of "kish" during the cooling of hot metal is a persistent problem in
the handling of hot metal. The greasy platelets of graphite are difficult to collect and
equally difficult to remove from collecting equipment. The only available solution
appears to involve hooding of the areas where kish is evolved, and connection of these
hoods to high-velocity ducts connected in turn to particulate-collection equipment.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
TABLE VII-12. ANALYSIS AND POSSffiLE SOLUTIONS TO IRON MAKING AIR-POLLUTION PROBLEMS
       Process and 
      Technological Equipment Factors 
 Process Segment Problem Possible Solution Advantages Disadvantages Requiring Solution 
 Transfer of bUl;: Generation of particulates Underground recovery Minimize emissions Increased maintenance Design, construction, 
m materials from due to handling     and maintenance 
»     
-i storage to       
-i       
III stockhouse       
r        
r        
III Stockhouse inter- Generation of particulate More complete applica - Minimize particulate Secondary dust handling Design and installation 
3: nal material emissions due to handling tion of available emissions problem  
III transfer to  equipment     
3:      
0 shiphoist       
:II       
» Material transfer Generation of particulate Enclosure and evacu~ Minimize emissions Increased maintenance, Redesign of blast furnace 
r 
 shiphoist to emissions due to handling ation of blast furnace  secondary dust handling top, dust handling, and 
z    
1/1 blast furnace top  top   problem, and potential safety problems 
-i      explosion hazard  
-i       <:
c:        H
-i Casting of iron Evolution of "kish" from Shorter troughs Reduction in time of None Redesign and reconstruc- H
III I
   N
I  cooling iron   iron exposure to  tion of cast house 0
o     atmosphere   
0        
r        
c: Flushing of slag Evolution of hydrogen Unknown    Mechanism of hydrogen 
3:    
m  sulfide     sulfide evolution and 
c:       means of control 
1/1       
r        
» Slag disposal Evolution of hydrogen Unknown    Mechanism of hydrogen 
m     
0  sulfide     sulfide evolution and 
:II       means of control 
»       
-i        
0 Transfer of iron Evolution of "kish" from Application of existing Minimize particulate None Design and construction 
:!! 
III to steelmaking cooling iron equipment  emissions   
1/1     
 furnace       
 Pigging of molten Evolution of "kish" from Redesign of pig machines Minimize particulate None Design and construction 
 iron cooling iron machines and appli- emissions   
   cation of available    
   pollution control    
   equipment     

-------
A-I
APPENDIX A
PROCESSES IN THE INTEGRATED IRON AND STEEL INDUSTRY
Because the manufacture of many products depends on the use of steel, the iron and
steel industry has grown to be one of the largest basic industries in the United State sand
the World. The production of steel consists of making metallic iron from iron ores, con-
verting the iron into steel, casting the molten steel into shapes that are solidified, and
then further processing of the solid shapes into semifinished products such as sheet,
strip, bar, rod, plate, slab, billet, bloom, or ingot. The following section provide s a
description of the various processes involved in the production of steel.
Manufacture of Iron and Steel
The principal steps in the manufacture of iron and steel are (1) preparation of
raw materials, (2) making iron, (3) making steel, (4) casting of steel, (5) rolling into
semifinished products, and (6) manufacture of finished products. Some of the descriptive
material in this section is taken from the work of J. J. Schueneman, et al. (1 )':, For more
detailed descriptive information on the various p:rocesses, it is recommended that such
well-established references such as "The Making, Shaping and Treating of Steel"(2) be
consulted.
Preparation of Raw Materials
The major raw materials used in the production of iron are iron ore (or agglom-
erates such as pellets and sinter made from ores), limestone, coke, air, and energy
in the form of heat. One factor that contributes to the economy of the integrated pro-
duction of iron and steel is that gases produced in the making of iron and coke fre-
quently are used to meet other energy requirements in the plant.
Iron Ore. Prior to the end of World War II, the United States iron and steel
industry was essentially self -sufficient in high-grade iron ores. However, because of
the high demand for steel, these high-grade ore deposits were depleted. The industry
instituted research to develop methods of utilizing low-grade taconite ores for blast-
furnace use. These developments were instrumental in the adoption of highly bene-
ficiated burdens in blast furnaces. The change in the makeup of iron constituents
charged to blast furnaces in the United States in the period from 1957 to 1966 is shown
in Figure A-I. During 1967 the iron and steel industry consumed a total of about 136
million net tons':":' of iron are and recycled mill scale and dust in the making of pig
iron. (3) Agglomerated materials such as pellets, sinter, and briquettes accounted for
about 97 million net tons, and the remaining 39 million net tons was in the form of
lump ore, most of which was imported.
"References for this appendix are given at the end of Appendix A.
~*Net ton = 2000 pounds.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-2
~ 90
~
8. 80
-
.s::.
.!2I 70
CI)
~
~60
::::I
CI)
~ 50
~
o
1957 58 59 1960 61
62 63 64 1965 66 67 68
Year
FIGURE A-I.
CHANGE IN BURDEN CHARACTERISTICS FOR
UNITED STATES BLAST FURNACES
Very little crushing and grinding of ore is done at the blast-furnace plant.
Usually the only crushing and grinding located at the steel-plant site is associated with
s inte r -plant ope rations. Pellets are made in plants located at the mine site s . High-
grade ores are crushed and sized at the mines within very narrow size ranges for
shipment to the blast-furnace plants(4). Fine materials that cannot economically be
processed further and used in the production of pellets near the mines are shipped to
steel plants for use in the manufacture of sinter.
Sinter. Sintering plants are dedgned to convert iron ore fines and blast-furnace
flue dust into a product more acceptable for charging into the blast furnace. This is
achieved by burning a mixture of ore-bearing fines plus a fuel consisting of coke dust or
coal. Combustion air is drawn through the flat porous bed of the mixture. The prin-
ciple of sintering is to supply just enough fuel to the material to be sintered so that a
sticky mass will be produced, but the material will not be melted sufficiently to run.
The bed is formed on a slow-moving grate composed of receptacle elements with per-
forated bottoms, known as pallets. The assembly of such pallets end to end in a hinged
or linked arrangement comprises an endless metal belt with large sprockets at either
end. The ignition furnace is either gas or oil fired, <:'l1d its purpose is to bring the fuel
in the charge to its kindling temperature, after which the down draft of air through the
bed keeps it burning.
The sintered material is dumped from the grate as it passes over the head
sprocket upon a screen, the undersized material becoming the return fines, and the
oversized material, which is still at a red heat in the center, passing to a sinter
cooler. The cooler can be a large rotating apron or a linear grate upon which the
sinter is deposited while cool air is blown through louvers located in the apron or grate.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-3
As the cooler reaches a certain position, stationary scraper bars push the sinter off
the apron into cars or conveyors. A sinter plant is illustrated schematically in
Figure A-2.
FIGURE A-2.
SINTER PLANT
Modern sintering plants have capacities ranging from 2000 to more than 6000 tons
of sinter per day. One plant of the latter capacity has a sinter -bed width of 12 feet and
a bed length of about 150 feet. In 1967 a total of about 51 million net tons of sinter were
made in the United States for use in the blast furnaces. By far the largest portion of
this sinter was made within the pe rimeter of blast -furnace plants. In contrast to fired
oxide pellets (which are very strong), sinter is relatively friable and does not stand up
well physically during shipment for long distances.
Oxide Pellets. The recovery of the iron components from taconite ores can only
be done if the ores are ground to a very fine powder. Sintering of the fine taconite con-
centrates was unsuitable as a method of agglomeration, so extensive efforts were
directed toward the development of pelletizing processes to agglomerate the concen-
trates into useful sizes. The first successful commercial pelletizing plant was placed
into operation by the Reserve Mining Company in 1955. The estimated annual pellet-
making capacity in the United States in 1968 was 56.3 million net tons.
The pelletizing process was originally developed to agglomerate the fine magnetite
concentrates, but since its initial development the process has been used for the
hematite ores as well. The process consists of two main operations: (1) rolling the
fine concentrates into damp balls of a suitable size (much like making a tiny snowball),
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-4
and (2) drying and firing the balls to make hard pellets. The pellets made are roughly
spherical in shape and about 1/2 inch in diameter. The types of equipment for making
the balls and for hardening them vary from plant to plant. One type of pellet plant,
such as used at the Empire Mine in Palmer, Michigan, is illustrated in Figure A-3(5).
CONCINnAn IIN
IY-IA!!
S1ACK
,',;, : ," ,,""'~,."
-~:
:,;;;.,;" . "',','"
~iR).,~ :~:.~~:::.
x':,.r~J '
fAN NO, 3 {COOUI} . -,'
---.-
(NO. I'AN 150 'I
FIGURE A-3.
GRATE-KILN PELLETIZING PLANT AT THE EMPIRE MINE
Limestone and Lime. Limestone (commercial CaC03) is the major fluxing mate-
rial used in producing metallic iron in a blast furnace. It is also used in some open-
hearth furnaces. The major role of limestone is to flux silica from ores, and to com-
bine with sulfur to lowe r the sulfur content of the iron or steel. To reduce the amount
of energy required to achieve the desired chemical reactions in the making of iron and
steel, technologists have adopted the use of burnt lime (commercial CaO) as an additive
in many steelmaking processes, but not in the blast-furnace process.
Limestone is crushed and screened to the desired size at the quarry site and only
the correctly sized stone is shipped to the blast-furnace plants. Burnt lime is also
prepared, usually at the quarry, by calcining the limestone to produce a high-quality
lime for use in basic oxygen furnaces, open hearths, and electric furnaces. There are
about eight different processes for making lime for use in the steel industry. (6) Many
steel companies have their own limestone quarries and related lime -producing facilities,
while others purchase their requirements from commercial operators.
Coke. Coke, the chief fuel used in blast furnaces, is the residue after distillation
of certain grades of bituminous coal. It is made in two types of ovens: (1) the beehive,
and (2) the recuperative or by-product oven. In either type of oven, the distillation or
coking process consists mainly of driving off certain volatile matter, leaving in the
residue a high percentage of carbon mixed with relatively srnall amounts of impurities.
The beehive oven is the older of the two types of oven, and on a national scale is
unimportant in comparison to by-product ovens. On a local scale in several small
areas, however, beehive ovens can be a matter of local concern. Their use at iron
and steel works has nearly disappeared, although a few are used occasionally during
times of maximum steel production when supplies of by-product coke may be in short
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-5
supply at the blast furnaces. The dome -like structure is built of refractory brick. It
has a flat floor sloping slightly toward the front. In the roof is an opening through which
coal is charged and the products of distillation and combustion escape. A door in the
front permits both the regulation of the amount of air admitted during the coking proces s
and the discharge of the coke after the process has been completed. A typical beehive
oven is about 12 feet in diameter by 8 feet high, and will hold about 6.5 tons of coal.
It is insulated with loam or clay to prevent loss of heat.
Beehive coke ovens are operated continuously to conserve the heat that has been
absorbed by the oven refractories. The ovens are charged as soon as practicable after
the coke from the previous cycle has been removed or "drawn" from the oven. The
heat stored in the oven refractories is enough to start the coking cycle for the following
charge of coal. The door to the beehive oven is partially bricked up and the coal
charged into the oven through an opening ("trunnel head") in the top of the oven. After
the total charge is in the oven, the coal is leveled to provide more or less uniform
treatment of the coal. The heat stored in the oven refractories starts the coking
process very soon after the charge has been leveled. Volatile matter from the coal is
driven off by the heat and starts to burn, thereby providing more heat to continue the
coking process. Coking takes place from the top to the bottom of the coal in the oven.
The rate of evolution of volatile materials and their subsequent combustion is controlled
by regulating the amount of air entering through the opening in the oven door. After the
coking process has been completed, the door is opened by removing the sealing brick-
work, and water is sprayed over the coke to quench or "water it out". By-products
are not recovered in the beehive process. Beehive ovens almost invariably are
located in coal fields; not within the perimeters of integrated steel plants.
In the by-product coking process, coal is heated in the absence of air. The vola-
tile matter is not allowed to burn away, but is piped to special equipment that extracts
its valuable ingredients. After the extraction process, some of the gas (heating value
about 550 Btu per cubic foot) returns to the ovens for use in heating the coking chambers
and for heating in other processes in the steel plant. These ovens are rectangular in
shape. Older ovens may be from 30 to 40 feet long, 6 to 14 feet high, and 11 to 22 inches
wide. Coke ovens built since 1967 are usually about 50 feet long, 16 to 17 feet high, with
coking chambers having an average width of 18 inches. As many as 100 of them may be
set together in a battery for ease in charging and discharging the coal and coke. A mod-
ern by-product oven can receive a charge of 16 to 20 tons of coal through ports at the
top. The ports are then sealed and coal begins to fuse, starting at the walls of the oven,
which may generate heat from 1600 to 2100 F. The fusing works toward the center of
the charge from both walls, and meets in the center, causing a crack down the middle of
the mass. This crack and the porous structure of the by-product coke are its distin-
guishing features. When coking is finished (16 to 20-hour carbonizing period), doors at
the ends of the oven chamber are opened, and the pusher ram shoves the entire charge
of coke into railway cars. The load is taken to a quenching station, where it is watered
by an overhead spray. After this, it is taken to a wharf to cool prior to screening.
The volatile products that have passed out of the ovens are piped to the chemical
plant where they are treated to yield gas, tar, ammonia liquor, and light oil. Further
refinement of the light oil produces benzol, toluol, and other complex chemical com-
pounds. However, in recent years, competition from the petrochemical industry has
made the recovery of coke by-product chemicals marginally economical or uneconomical,
unless coke-plant installations are of such a large size that processing costs can
compete with petrochemicals.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
1- --
A-6
The fundamental features of a coke battery cannot be changed during its lifetime,
which amounts to 20, 30 or more years. At the end of its life it is completely razed
and a new structure embodying current technological ideas is erected to replace it.
Fuels. An integrated iron and steel plant uses a great variety of fuels, some of
which are generated as part of the plant's own operations and some of which are
purchased.
Coke -oven gas is produced during the manufacture of coke from coal in the coke
ovens. The exhausted gas does not contain any particulates and does not require
separate cleaning because particulates are trapped in the by-product recovery system.
The major impurity is hydrogen sulfide, which in burning is changed to sulfur oxides.
For some in-plant use, the hydrogen sulfide is removed. Coke -oven gas typically has
a heat content of about 500 to 550 Btu per cubic foot.
Blast-furnace gas is a product from the ironmaking process in the blast furnace.
It is a rather low-heat-content gas, with an average value of about 80 Btu per cubic
foot. The gas when exhausted from the top of the blast furnace, is laden with particu-
late materials which are cleaned from the gas before its use in the steel plant. Blast-
furnace gas is usually burned to heat blast-furnace stoves, normalizing and annealing
furnaces, foundry core ovens, gas engines for blowing, firing of boilers, and gas engines
and gas turbines for power generation. Preheated blast-furnace gas combined with pre-
heated air has been used succes sfully for heating coke ovens, soaking pits, and reheating
furnaces. (2) The present trend in the use of higher blast temperature for blast furnaces
has increased the requirement for cleaner blast-furnace gas in the heating of blast
stoves. A cleaner gas is required to prevent clogging of the checker work in the blast
stoves, because clogging decreases stove efficiency and increases maintenance problems.
Tar, one of the by-products of the production of coke, is often used as a fuel for
firing open-hearth furnaces. The tar that is burned is not cleaned, and the sulfur con-
tained in it (usually about 0.60 percent) becomes sulfur oxide in the combustion process.
Commercial fuels (principally oil and natural gas) are used in the as -received
condition at the steel plant.
Air. Air is a necessary material in the production of metallic iron in the blast
furnace. It is used as it exists in the surrounding atmosphere without any treatment
except for preheating to temperatures varying from 1000 to 2000 F, before it is blown
into the blast furnace. Air requirements for the blast furnace may vary from 45, 000
cubic feet to 60, 000 cubic feet per ton of iron, depending on the type of practice used.
Heated air is used to supply thermal energy to the blast furnace, but it can also act as
a direct replacement for coke that would normally be burned to supply this thermal
energy inside the blast furnace. A typical blast stove is illustrated in Figure A-4,
with various types of checker brick used in it.
The use of high blast temperatures permits the blast-furnace operator to inject
auxiliary fuels (such as oil, natural gas, coal, or coke -oven gas) through the tuyeres.
The injection of these auxiliary fuels results in the generation of increased amounts of
carbon monoxide in the blast furnace. (Carbon monoxide is the active reducing gas in
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-7
ACtUS 00011
'--
~

II'
IiiP
~j
CH,M"'[YIJAlvyii
CO",,,'C'I'IO,,. ~
COLD flL"S1 '
CO......,(.TIO'"
FR£YN TYPE
CHlIiIINf.'I' "'oHVf.
COJrol"'[CTlO"
FIGURE A-4.
111

~ "16" HI.A' DuTT






i'l
"
::
:',
iil
!
::
;:1
:.;111

::
:'
.:1

:,1
~:1 5~"P' Dun
':1 5[W1' ~tLiC'"
;:1

:d
:'
::
:'
::

j'l
::!
':i
'i'l.

~ :
:!I
ii
:'
CO""U$1'IOIi
,........f...
1

SI(I",ULl
$ltllo WALL
SUPE" OuT"
''IIGH "LVIII!N'"

""I'"
:1
. - !
, ,
j ~
H
'I
,/ jl
. il
" IL"", '.He
II'!: i!
'I :1
j. Ii!
I II "'.IUL "'IOff

II~~.'..".'..'.""'"
j t~~~~~:LA" .".,
t:l:'r:.
n! r<,,'~~,
j:: .
Ii: .
I:: :
"
::
ii I
:: I
;: I
i jj Iii I
'I:: I
Ii ii,
;: I:
Iii ;
I':
f ~
~
ii

.:

II
TYPICAL BLAST FURNACE STOVE
BATTELLE
MEMORIAL
INSTITUTE - COLUMBUS
LABORATORIES

-------
A-8
the blast-furnace process.) This use of injected fuels also results in a lowering of the
amount of coke that must be charged to a blast furnace. Typical effects of fuel-oil
injection and coal injection on the amount of coke required to make one net ton of pig
iron is shown in Figure A-5(7) and Figure A-6(8). The injection of auxiliary fuels into
the blast furnace is economically attractive to blast-furnace operators. In 1968 in the
United States, about 50 blast furnaces were using injection of natural gas; about 15 were
using fuel oil; about 3 were using coke-oven gas; 3 were using tar; and 1 was using in-
jection of coal.
1400

~
8.c:
~ e 1300
C:'-
~.[

a;-- 1200
+-0
0c:
a:o
Q)+- 1100
~ +-
OQ)
UC:
1000
Blast Temper ure
0- 1697F
. --- 1877 F
.-- 2057F
"
" "'.........
" ....,.
....... ... ""-
.... " ....
~'" -..-
......-. --
.-- --.
-...
"".
970
o 40 80 120 160 200 240

Fuel Oil Injected, pounds per net
ton of pig iron
FIGURE A-5.
EFFECT OF FUEL-OIL INJECTION AND BLAST
TEMPERA TURE ON COKE RATE
c: 1250
o
~ 
- 
0 
c: 1150
o
- 
- 
Q) 
c: 1100
~
Q) 
a. 
Q) 
~ 1050
u 
- 
0 1000
VI 
"0 .
c: 
:J 
0 950
a.
~
Q)
C 900
a:
Q)
~
o
U
.
.

8500 50 100 150 200 250 300
Coal Injected at Tuyeres, pounds
per net ton of pig iron
FIGURE A- 6.
EFFECT OF COAL INJECTED AT THE TUYERES ON COKE RATE
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-9
Making Pig Irori
Pig iron has been manufactured in the United States for a great many years in
blast furnaces, and this is still the process used to make almost lOO percent of the pig
iron. Pig iron from a blast furnace is saturated with about 4 percent carbon. The iron
is tapped while molten from the blast furnace and usually is not allowed to solidify be-
fore it is delivered to the steelmaking process. The molten iron in steelmaking parlance
usually is called "hot metal'!. Only a small portion of molten iron from blast furnaces is
allowed to solidify into "pigs" for distribution in the solid state. The name "pig iron" is
generic and includes "hot metal'! and "iron pigs". Much research has been directed to-
ward the development of processes that would bypass the blast furnace as an iron pro-
ducer. The first such direct-reduction process for pig iron is expected to be operational
in the United States in 1969 near Mobile, Alabama.
Blast-Furnace Practice. The first step in the conversion of iron ore into steel
takes place in the blast furnace. The blast furnace is a large cylindrical structure about
100 feet high, lined with heat-resistant bricks. A blast furnace is shown schematically
in Figure A-7. Iron ore, coke, and limestone are charged through sealing "bells" at
the top, and heated air under pressure is blown into the lower section through the tuyeres
to burn the coke. The air is heated in stoves (described in the section on Air and illus-
trated in Figure A-4), which typically are 26 to 28 feet in diameter and over 100 feet
high. Three or four blast stoves are used per blast furnace, depending on the method
of blast heating developed in the various plants. As the solid materials (known coHec-
tively as the "burden") pass down the furnace from the top to the bottom, the reducing
gases (carbon monoxide and hydrogen) rising through the burden react with oxygen in
the ore to start the formation of iron. This reaction continues as the burden materials
flow toward the middle of the furnace, at which point the coke acts to take out still more
of the oxygen in the ore, and the limestone begins to crumble and react with impurities
in the ore and coke to form a molten slag. As the charge enters the zone of fusion, all
the materials but the coke become pasty or fused. The iron becomes a porous mass.
It then pas ses through the melting zone and becomes liquid. In this zone the ash from
the burned coke is absorbed by the liquid slag, while the iron absorbs silicon from the
slag and carbon from the coke.
The iron and slag form a molten mass in the hearth, the slag floating on a pool of
iron 4 or 5 feet deep. About every 4 or 5 hours iron and slag are drawn off. The slag
is removed more frequently than the iron. From 100 to 300 (or more) tons of iron are
drawn off at each time. The hot-metal or ladle cars which receive the iron range in
capacity from 40 to 160 tons. The ladle car usually is a special type of tank car that
makes it possible to deliver hotter iron to the steel works, even though it may be 20
miles away. Most of the metal produced in the blast furnace is used in molten form for
the manufacture of steel in open hearth and other types of steelmaking furnaces.
To produce 1 ton of pig iron requires, on the average, 1.7 t.ons of iron ore,
0.9 ton of coke, 0.4 ton of limestone, 0.2 ton of sinter, scale, and scrap, and 4. 0 to
4.5 tons of air. In addition to the pig iron, the furnace yields about 0.5 ton of slag and
about 6 tons of gases per ton of pig iron produced. Air constitutes over one -half of the
material entering the furnace, whereas gases constitute more than three -quarters of
the materials leaving the furnace. The difference is due to the fact that much of the
carbon and oxygen entering as solids, in the coke and ore, respectively, emerge as
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-10
:jl
II:
:' :11
~!:'Ii
:' ,I
~ !
"... :

If
~~'
~~
,";
,,'.
,.:'
:!:;
~':;
FIGURE A-7.
TYPICAL BLAST FURNACE
BATTELLE
MEMORIAL
INSTITUTE - COLUMBUS LABORATORIES

-------
A-II
gases. These gases, piped from the top of the furnace, are rich in carbon monoxide,
which can be burned. They are used to heat stoves and generate power. About 30 per-
cent of the gas is required to heat stoves, and the remainder is used for steam genera-
tion, unde rfiring of coke ovens, or he ating soaking pits.
Direct Reduction. Processes which bypass the blast furnace as a means of pro-
ducing pig iron usually make use of much smaller equipment and are not dependent on
handling such large quantities of materials as required in blast-furnace plants. The
first direct-reduction process to be constructed for commercial production of pig iron
in the United States is the Dwight-Lloyd-McWane (D-LM) process. The plant (under
construction near Mobile, Alabama) will be very small by steel-industry standards
(only about 200,000 tons per year). A typical flow sheet for the process is shown in
Figure A-8.
1- 0.. C..I PI..


~~~
~~:
~.


Q ~AAAAA/\A4
(........i..
..cJCI.
'.C.If;...
FIGURE A-8.
DWIGHT-LLOYD-McWANE DIRECT REDUCTION PROCESS
The D-LM Process makes use of a balling operation to prepare powdered ore,
coal, and flux for partial reduction on a sintering machine, after which the partially
reduced pellets are charged into an electric smelting furnace where the pellets are
further reduced and melted to make pig iron.
Although the D-LM Process is mentioned here by way of illustration, and because
it is a "first" for the production of pig iron in the United States by a means other than
the blast furnace, this particular process does not have any unique worldwide impor-
tance among processes of this general type. For example, by far most of the pig iron
that is produced throughout the world by means other than the blast furnace is made in
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-12
electric smelting furnaces of the Tysland-Hole type; sometimes with some type of
preheating or pretreatment of the burden. The largest installations of this type are in
Norway and Venezuela where prices of electrical energy are low in comparison to
prices for coal. Such processes for pig iron have not yet been adopted in the United
States, nor does there seem to be significant pressure to adopt them on a large scale
in the near future.
In addition to direct-reduction processes that produce molten pig iron (as dis-
cussed in the preceding paragraph), other direct-reduction processes perform the
reduction of ore to metallic iron without going through a melting step. Examples
include (1) the HyL Process now operational in several plants in Mexico, which uses
natural gas as the main source of heat and reductant, and (2) the SL-RN process now
being implemented by plant construction in New Zealand and Korea, which uses coal as
the main source of heat and reductant. Processes in this category produce "metallized
ore" or "sponge iron" (without going through the pig iron stage), and these solid
products are then melted and refined in steelmaking furnaces in a manner generally
similar to melting and refining of scrap. The first plant of this general type in the
United States presently is under construction near Portland, Oregon, using the Midland-
Ross "Midrex" process, but as in the cas,e of the D-LM plant, rated output will be
small in comparison to even a small blast-furnace plant.
Making Steel
Steel in the United States is made by three major processes, (1) open-hearth
practice, (2) basic oxygen (BOF) practice, and (3) electric-furnace practice. At one
time, steelmaking in the Bessemer converter was one of the predominate processes
used for making steel. By 1948, the production of Bessemer steel in the United States
had decreased to about 4.2 million net tons per year. This production decreased further
to about 1.4 million net tons by 1958, and by 1967 the total production of Bessemer steel
was only 0.3 million net tons. During 1968, Jones and Laughlin Steel Corporation shut
down the last Bessemer converters in the United States integrated iron and steel indus-
try. These converters were located at their Aliquippa, Pa., plant. The only Bes semer
converter remaining in operation in the United States is located at the A. M. Byers
Company, Ambridge, Pa., and is used in the manufacture of wrought iron. (9)
Bessemer converters are not considered in this study.
The production of carbon raw steel':' from 1954 through 1968 is shown in Fig-
ure A-9. The rapid increase in production of carbon steel in basic oxygen furnaces
(BOF) and the simultaneous decrease in tonnage made in open hearths is quite evident.
The relationship for the full year 1968 for the major steelmaking processes in
the United States was as follows:
°AISI definition. Raw steel is steel in the first solid state after melting and suitable for further processing or sale and includes
ingots, steel castings, and continuous or pressure-cast blooms, billets, slabs, or other product forms.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-13
Process
Production of Raw Steel in 1968
Percent
of Total
Millions of Net Tons
Open hearth furnace
Total
66. 1 50.4
48.6 37. 1
16.4 12.5
131.1 100
Basic Oxygen furnace (BOF)
Electric furnace
130
120
110
(/)
"0100
C'I
c
;: 90
o
(/) 80
c
o
~ 70
.....
~ 60
g 50
Carbon - Steel Ingots
~ 40
30
20
Oxygen Converter
10

o
Electric furnace
Bessemer
1965 1970
Year
1975
1955
1960
FIGURE A-9.
PRODUCTION OF CARBON RAW STEEL IN THE
UNITED STATES BY VARIOUS PROCESSES
Open-Hearth Steelmaking. The open hearth furnace at one time accounted for
about 90 percent of the steel made in the United States. During recent years, increa!:!.ed
use of the basic oxygen furnace and electric furnace has decreased the production of
steel in open hearths to about 55 percent of the industry total by early 1968.
Open-hearth steel is made usually from a mixture of scrap and hot metal in vary-
ing proportions, depending on relative cost and availability of these two main raw
materials. The object of the operation is to lower the impurities present in the scrap
and pig iron, which consist of carbon, manganese, silicon, sulfur, and phosphorus, to
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-14
the limits specified for the particular grade of steel. This refining operation is
carried out by means of a slag that forms a continuous layer on the surface of the
liquid metal. This slag consists essentially of lime combined with the oxides of
silicon, phosphorus, manganese, and iron, which are formed or added during the
operation.
Open-hearth furnaces are of two types, depending on the character of the refrac-
tory material that forms the basin holding the metal. Where the refractory material is
mainly siliceous fireclay or another silica-rich refractory, the furnace is described as
"acid", and where the basin is lined with dolomite (or magnesite), it is termed a
"basic" furnace. Acid open hearths are used mainly for making steel castings in the
foundry industry. Steel in the integrated iron and steel industry is made mostly in
basic open hearths.
Open-hearth furnaces in the integrated steel industry are large massive struc-
tures. The open-hearth furnace proper consists of a shallow rectangular basin or
hearth enclosed by walls and roof, all constructed of refractory brick, and provided
with access doors along one wall adjacent to the operating floor, as shown in Figure
A-IO. A tap hole at the base of the opposite wall above a pit is provided to drain the
finished molten steel into ladles. Fuel in the form of oil, coke -oven or natural gas, tar
from coke making, or producer gas (a gas rich in carbon monoxide manufactured by
blowing a limited quantity of air through a hot bed of solid fuel) is burned at one end.
The flame from combustion of the fuel travels the length of the furnace above the charge
resting on the hearth. Upon leaving the furnace, the hot gases are conducted in a flue
downward to a regenerative chamber called checkerwork or checkers. This mass of
refractory brick is systematically laid to provide a large number of passageways for
the hot gases. The brick mass absorbs heat, cooling the gases to about 1200 F. All
the elements of the combustion system burners, checkerwork, and flues are duplicated
at each end of the furnace, which permits frequent and systematic reversal of flow of
the flame, flue gases, and preheated air for combustion. A system of valves in the
flue effects the gas reversal so that the heat stored in checkers is subsequently given
up to a reverse -direction stream of air flowing to the burners. In some plants, the
Checker
Chambers
FIGURE A-IO. CROSS SECTION OF A BASIC OPEN-HEARTH FURNACE
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-15
gases leaving the checkerwork pass to a waste -heat boiler for further extraction of
heat so as to lower the temperature from around 1200 F to an average of about 500 or
600 F. Open-hearth furnace capacities span a wide range. The median is between 100
and 200 tons per heat (batch of finished steel), but there are many of smaller capacity
and an increasing number of larger capacity. Time required to produce a heat is com-
monly between 8 and 12 hours without the use of large amounts of oxygen.
The open-hearth process consists of se'veral stages: (1) tap to start, (2) charging,
(3) meltdown, (4) hot-metal addition, (5) ore and lime boil, (6) working (refining),
(7) tapping, and (8) delay. The period between tap and start is spent on normal repairs
to the hearth and plugging the tap hole used in the previous heat. During the charging
period, the solid raw materials (which usually include a combination of pig iron, iron
ore, limestone, scrap iron, and scrap steel) are dumped into the furnace by special
charging machines. The melting period begins when the first scrap has been charged.
The direction of the flame is reversed every 15 or 20 minutes. When the solid material
has melted, a charge of molten pig iron is delivered direct from the blast furnace in
large ladles and poured into the open hearth through a spout set temporarily in the
furnace door. This is the normal sequence for a. "hot-metal" furnace; but for a cold-
metal furnace, only solid materials (pig iron and/ or steel scrap) are added, usually in
two batch charges.
The hot-metal addition is followed by the ore and lime boil, which is a bubbling
action much like the boiling of water and is caused by the oxidized gases rising to the
surface of the melt. Carbon monoxide is generated by oxidation of carbon and is
characterized by a gentle boiling action called "ore boil". When carbon dioxide is
released in the calcination of the limestone, the more violent turbulence is called the
"lime boil".
The aims of the working pe riod are (l) to lower the phosphorus and sulfur content
to levels below the maximum level specified, (2) to eliminate carbon as rapidly as
possible and still allow time for proper conditioning of slag and attainment of proper
process temperature, and (3) to bring the heat to a condition ready for final deoxidation
in the furnace or for tapping. At the end of the working period the furnace is tapped,
with the temperature of the steel at approximately 3000 F.
The delay period includes waiting time during the heat cycle (e. g. equipment
breakdown, tapping equipment in use on another furnace, etc.) plus repair work not
usually done during the tap to start period.. For normal operation of a 10-furnace shop
as a whole, the following breakdown of the heat stages has been made:
Period
Percent of Time in
Indicated Period
Tap to start
Char ging
Meltdown
Hot-metal addition
Ore and lime boil
Working (refining)
Tapping
Delay
6
12
12
3
38
19
2
8

100
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-16
The use of consumable lances to inject gaseous oxygen into the bath during the
refining period and speed the oxidation reactions, shorten heat time, save fuel, and
increase production has become more or less standard practice over the last 10 to 12
years. Since 1957 water -cooled lances inserted through the furnace roof have come
into prominent use, Frequently, oxygen lances are used throughout the heat with the
exception of the charging and hot-metal-addition periods, By use of high oxygen flow
rates from hot metal to tap, production rates of 90 to 100 tons per hour are conceivable
in a 300 -ton furnace. Oxygen consumption under these conditions ranges from 600 to
1000 cubic feet per ton (900 to 1667 scfm during the period oxygen is being added),
There has been some experimentation with oxy-fuellances, i. e" the use of
oxygen in combination with the fuel. This procedure plus the substitution of burned
lime for limestone has increased the steel output of a 200 -ton furnace from 20 to
approximately 30 tons per hour.
Upright Basic Oxygen (BOF) Steelmaking. A process for refining molten pig iron
("hot metal") to make steel was developed in 1952 in Linz-Donawitz, Austria, in which
a top -blown oxygen converter was used to refine the pig iron. Although there are now
several variations in practice, the general technique worldwide is known as the "basic
oxygen" or. BOF proce s s. The furnace is a pear -shaped steel shell lined with refrac-
tory brick as shown in Figure A-II. The usual charge for this type of furnace consists
of hot metal (molten pig iron), steel scrap and flux. The ratio of hot metal to scrap
conventionally is about 70/30, The steel s crap can be replaced with iron ore or pre-
reduced iron pellets. A water -cooled lance is used to supply high-purity oxygen at
high velocity to the surface of the metal bath. The high velocity of the oxygen results
in impingement on the liquid-metal surface, which in turn produces violent agitation
and intimate mixing of the oxygen with the molten iron. Rapid oxidation of the dissolved
carbon, silicon, and manganese produces a heat of steel. In the blowing process,
some of the iron is oxidized as well and pas s e s off as fume, The B OF p roce s s diffe r s
from open-hearth practice in that external heat does not have to be supplied to facilitate
the refining of the iron. The only sources of heat are (l) the sensible heat from the
hot metal, and (2) the heat released by the exothermic reactions between the oxygen
and metalloids in the charge (primarily silicon and carbon), In March, 1969, there
were 27 steel plants in the United States with BOF installations with a total rated annual
capacity of 57 million tons. The 60 existing vessels have capacity ratings from 75 to
325 tons per heat. An additional 19 million tons of annual capacity is under construction
or planned for operation through 1970, The time required to make a heat of steel in
the BOF is much shorter than in the open hearth. Heat time for a typical 150-ton BOF
operation is as follows:
Charge scrap
Charge hot metal
Oxygen blow
Chemical tests
Tapping time
1 minute
2 minutes
20 minutes
5 minutes
5 minutes
Total time
33 minutes
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-17
TYPICAL SASIC OXYGEN FCE.
IS.QF)
FIGURE A-II.
BASIC OXYGEN FURNACE
~
~
~
Z
o
u
~
o
~
~
~
Z:
0,
~'
~
~
~
c
..
ZZ
00
~~
..~
Rotary Basic-Oxygen Steelmaking. A steelmaking process developed in Sweden
makes use of a rotating vessel that is operated in a nearly horizontal position as shown
in Figure A-12(10). The process is known as the Stora-Kaldo Oxygen Process.
Only one steel plant in the United States has a rotary-oxygen-furnace installation;
Sharon Steel Corporation has two Kaldo vessels with a nominal capacity of 150 net tons
of steel per heat. Operation of the Kaldo converters is somewhat similar in principle
to that for the upright BOF vessels, but there are significant differences that increase
heat time over that of the BOF and permit the use of more scrap than in the BOF.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-IS
OXYC:EN
CHARGING POSITION llM[ - ORt J7
-----"--
~." '- ;.:;:-....
A/' ,," -;'--:-...
(,'.:" , ." :' .,....~
. : \ . 1/
'.";"- '"'.'.c " J
.' '.."; ". ,'/. .
FIGURE A-12.
STORA-KALDO ROTARY OXYGEN CONVERTER
Electric -Furnace Steelmaking. Whereas the open-hearth and BOF steelmaking
processes conventionally use a charge that contains a high percentage of molten pig iron
(hot metal) obtained from blast furnaces, electric steelmaking furnaces conventionally
have no hot metal in their charge (although there has been some limited use of hot metal
in electric steelmaking furnaces). Generally, electric steelmaking furnaces depend on
metallic scrap for most of their charge. Because electric steelmaking furnaces permit
a high degree of control over their operations, expecially with regard to ability to hold
the steel for long refining periods and to control temperatures to high levels in the
furnace, they are generally preferred for the manufacture of alloy and stainless steels.
Although electric furnaces account for only about 12 percent of the total raw steel made
in the United States in 1967, they accounted for about 36 percent of the alloy and stain-
less steel. Of the total steel made in electric furnaces in the United States in 1967,
about 41 percent was in alloy and stainless grades. Comparable alloy and stainless
fractions for the other two major types of furnaces were about 9 percent of open-hearth
production and about 6 percent of BOF production.
The furnaces employed in electric -arc melting practices in the integrated steel
industry are refractory-lined cylindrical vessels with large graphite electrodes passing
through the furnace as shown in Figure A-l3. Electric energy is supplied to the elec-
trodes by transformers ranging in capacity from 4,000 to 85,000 kilovolt-amperes.(ll, 12)
The trend in recent years has been to provide electric-arc furnaces with larger trans-
formers than previously thought feasible. By installing larger transformer capacity,
the productivity of a given electric-arc furnace can be doubled. The largest installa-
tions to date are a 200 -ton direct-arc electric furnace powered by a 76, OOO-kV A trans-
former at the Laclede Steel Company, Alton, Illinois(l2); four 200-ton furnaces at the
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-19
Port for third electrode
FIGURE A-13.
DIRECT-ARC ELECTRIC FURNACE
Republic Steel Corp., Canton, Ohio, plant and a 250-ton furnace at Northwestern Steel
and Wire Company which was placed in operation in early 1969. The relationship be-
tween melting capacity of direct-arc electric furnaces in the United States and their
transformer capacities are shown in Figure A-14. (11, 12)
80
.
II)
~70
~
E
c 60
I
.
-
o
>
o 50
~
.
. .
.
g' 40
-
c
a:::
~30
E
~
o
~ 20
c
e
.....
.
.
..
.
.
.
..
10
..
.
I~....
:/'
.. .
..
o
o
FIGURE A-14.
50 100 150 200 250
Furnace Capaci t y, net loos

RELATIONSHIP BETWEEN ELECTRIC-ARC FURNACE
CAPACITY AND TRANSFORMER RA TING
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-20
Only basic -lined electric -arc furnaces are used to make steel in the United States
integrated iron and steel industry. Some acid-lined electric-arc furnaces are still in
use in the United States, but these are used to produce special steels in the foundry
indus try.
The metallic charge for direct-arc electric-furnace melting of steel usually con-
sists of mostly steel scrap, along with some cast-iron scrap or solid pig iron. Charge
preparation consists of the selection of the proper grades of scrap for the steel to be
made, and the addition of alloying elements to achieve the desired composition. High-
purity gaseous oxygen usually is used today to carry out refining more rapidly than
would be done with the older practice of using additions of iron ore as the source of
oxygen. .
Electric -induction furnaces are used in the integrated iron and steel industry only
to melt special alloys and stainless steels on a scale that is very small and scattered
when compared with methods for melting high-tonnage steels. Because induction-
furnace melting involves no products of combustion, and lacks the high-temperature
arc of the steelmaking electric-arc furnace, it is the cleanest method for melting steel.
If the scrap placed into an induction furnace is clean, emission from the furnace is
minor, with respect to both quantity and density, and is easily collected in simple equip-
ment. Because induction furnaces are used only in specialty situations, the scrap
charged to them almost invariably is selected with care as to composition and cleanli-
ness. In those cases where contaminated (e. g., oily) scrap is charged, the scrap
usually will emit fume and smoke until the contaminant is burned off. To minimize
fume emission during melting in such furnaces, contaminated scrap sometimes is pre-
heated, thus moving the point of evolution of fume from the melting operation back to
the preheating operation. In general, however, induction-furnace melting of steel is
a miniscule contributor to air-pollution problems.
Scrap Preheating. Preheating of steel scrap for charging into steelmaking fur-
naces is not a common practice, but when it is done it is accomplished by three tech-
niques: (1) heat exchange by the removal of sensible heat from gases not undergoing
combustion as part of the preheating cycle, (2) by the use of air-fuel burners, and.
(3) by the use of oxy-fuel burners. The use of air-fuel burners is most common. With
the use of such burners (as with the employment of noncombustion processes), tem-
peratures attained by the scrap are rarely above 1800 F. At these temperatures, the
only appreciable potential for particulate emission from the preheating step rests in
the presence of oil, paper, rubber, and other combustibles in the scrap. If the scrap
contains such combustibles, a considerable amount of fume can be generated during
preheating. In comparison to these first two methods of preheating, oxy-fuel preheating
adds additional p'roblems because of the higher temperatures that can be attained during
preheating. These higher temperatures extend into and past the melting range of the
steel.
A new steelmaking technique developed in the United Kingdom in 1962-63 made
use of an oxygen-fuel burner instead of an oxygen lance to achieve melting and refining
of low-carbon and low-alloy steels. (13) This fuel-oxygen scrap (FOS) process, has
not gained acceptance in the United States, but the idea of fuel-fired burners and oxygen-
fuel burners has been adapted to a limited degree in the United States for the preheating
of scrap in open-hearth, BOF, and electric steelmaking practices.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-21
Oxygen-fuel burners are used as preheaters to shorten the time required to melt
scrap in the steelmaking furnace, and thereby reduce the overall tap-to-tap time in open-
hearth practice, and to permit use of greater amounts of scrap in the BOF's. Tap-to-
tap times in open hearths have been decreased by 12 percent with a corresponding pro-
duction increase of 15 percent. (14) Scrap charges in BOF's have been increased from an
average of 28 percent of the metal charge to 36 percent of the charge. (15) Republic Steel
Corporation, at its Chicago Works, has four open hearths operating with oxygen -
natural gas roof burners (16); Inland Steel Company at East Chicago, Indiana, has in-
stalled oxygen - natural gas burners in all open hearths of their No.3 open-hearth
shop. (14) Wisconsin Steel Works, Chicago, Illinois, has used oxygen - natural gas
burners in their two 120.:.ton BOF's, (17) and the Pittsburgh Steel Company has used
oxygen - oil burners in their 200-ton BOF. (15)
Vacuum Degassing of Molten Steel
In the early 1950's several catastrophic failures of large electric-generator turbine
rotors were traced to the presence of hydrogen in the steel. These events were quickly
followed by research and development efforts directed toward developing methods of
. eliminating hydrogen from steel. In the United States, the first vacuum degassing instal-
lations were placed into operation in 1956. Steel technologists were not long in finding
that vacuum degas sing could be used also as a means of deoxidizing steels. The rapid
reaction between carbon and oxygen under reduced pressures produces a cleaner product
than when oxygen is removed by use of conventional deoxidizing additions such as silicon
and aluminum. This carbon-deoxidation technology was rapidly placed into practice by
the steel industry, as evidenced by the number of units installed in the years that fol-
lowed. The number of units installed from 1956 through 1967 is shown in Figure A-IS.
en
-
'2
::> 70
0'1
,~
~ 60
o
l50
I
E 40
::::I
::::I
°30
~
Cumulative
total installations
'020
~
~ 10
E
::::I
Z
1956 1957 195819591960 1961 1962 1963196419651966 1967
Year
FIGURE A-IS.
NUMBER OF VACUUM DEGASSING INSTALLATIONS
IN THE UNITED STATES
(Source: Battelle compilation. )
Vacuum degassing processes can be divided into three general groups:
degassing, (2) circulation degassing, and (3) ladle degassing.
(1) stream
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-22
Stream Degassing. The stream-degassingprocess was the first to be placed into
operation for the treatment of steel that is cast into large ingots for subsequent forging
into rotors for electric generators. In this process the vacuum-treated steel is collected
directly in the ingot mold that is inside the vacuum chamber as shown in Figure A-16.
If the degassed steel is collected in a ladle that is located inside the vacuum tank, the
OBSERVATION PORT
OR TV CAMERA

WATER
COOLING
STOPPER ROD

OBSI!RVATION PORT
OR TV CAMERA
VACUUM-
DEGASSING
CHAMBI!R
INGOT MOULD
FIGURE A-16.
INGOT STREAM DEGASSING
process (called ladle -stream degassing) is conducted, as illustrated in Figure A-17. The
ladle of vacuum-treated steel is removed from the tank and transported to the ingot-
pouring area where the steel is then cast into ingot molds. There are other variations
that perform the same type of operation by slightly different mechanical means, but the
end results are generally the same as for the processes illustrated.
To vacuum - -
system
FIGURE A-17.
LADLE-STREAM DEGASSING
Circulation Degassing. Some steelmaking technologists believed that the treat-
ment time in stream degassing was too short to take full advantage of the potential of
vacuum degassing. Several processes were developed to extend the vacuum-degassing
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-23
cycle and permit a longer treatment time. The circulation processes involve taking a
part of the molten steel from the ladle into the vacuum treatment chamber for treatment
and then returning the steel to the ladle. The cycle is repeated until the oxygen and
hydrogen contents of the steel are reduced to acceptable levels. Representative
circulation-degassing processes are shown in Figure A-18.
. -- ---
" if A
a.
D-H Process
Refractory lined
vacuum chamber
~
AltOON
P'RI!HI!AT
I!IUltNI!It
ASSI!MI!ILV
Process gas.
b.
Thermo-Flow Process
R-H Process
c.
FIGURE A-18.
CIRCULATION-DEGASSING PROCESSES
a.
Purge Degassing
Ladle Degassing. Ladle-degassing processes provide agitation or stirring the
molten steel in the ladle that is positioned in a vacuum tank. One process bubbles argon
gas through the steel to agitate the molten metal, and another process stirs the metal by
means of an induction coil. These processes are illustrated in Figure A-19.

Alloy
additions
Outlet to 1
vacuum p.JI'f1)S


Allo, t""" 1

Observation port
-'
..- -
Vacuum seol
Argon tu be
Furnace ladle
Sight
port
Vacuum
jets
::!J
Stainless
steel ladle
(non-magne-
tic)

Ladle-Induction Degassing
b.
FIGURE A-19. LADLE-DEGASSING PROCESSES
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A..24
Manufacture of Semifinished Products
An essential step in the preparation of molten steel for further processing into sal-
able products is the solidification of the molten steel into shapes that can be proces sed
into the desired products. The traditional method has been to pour (teem) the steel into
ingot molds, permit the steel to solidify, remove the ingots from the mold, reheat the
ingots, and roll them into the desired semifinished products such as billets, blooms, or
slabs. In recent years, two new methods have been developed and placed into USe by the
integrated iron and steel industry. These processes are (1) continuous casting and
(2) pressure casting, both of which eliminate much of the processing associated with con-
ventional ingot practice. Conventional ingot practice accounted for about 94 percent of
the raw steel produced in the United States in 1967. It is estimated that continuous cast-
ing accounted for 5.5 percent and pressure casting 0.5 percent.
The teeming of molten steel into conventional ingot molds at one time was accom-
panied by much evolution of smoke and fume, primarily because tar and other bitumens
were used as mold coatings. During the last decade the use of such coatings has been
drastically curtailed, so that visible emission during teeming is less than formerly.
However, under some teeming practices, evolution of air contaminants is high enough to
restrict visibility at the teeming station. This degree of evolution does not occur in the
newer continuous-casting and pressure-casting processes.
Some grade's of free- machining steels involve intentional additions of lead or sulfur
to the steel shortly before or during teeming. Because of the volatility of these elements
at the temperature of molten steel, fuming of the additives represents an emission prob-
lem. The tonnage of such steels represents a small fraction of total steel produced
nationally, but can be substantial at particular steel plants that make such grades.
Conventional Ingot Practice. Conventional casting and rolling require a large
amount of plant area to accommodate the teeming area, soaking pits, and roughing mill.
In addition, extens ive transport facilities are required to handle ingots and ingot molds.
Conventional ingot-casting practice is illustrated in Figure A-20. Molten steel is trans-
ported in ladles to the teeming station for pouring of the ingots. Molds are transported
~I

Teem Solidify


~ ,,,"'~, '''"'~~

,," ~ ",;,
lWJ Soaking pil
FIGURE A-20.
CONVENTIONAL INGOT-CASTING PRACTICE
to the same area on special cars holding two or three molds to the car. One string of
cars holds enough ingots to receive all of the molten steel from the heat. After the teem-
ing operation, the ingots are permitted to solidify in the mold and cool to a selected
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-25
temperature, which is dependent on the type of steel produced. After the proper hold-
ing time, the ingots are transported to the stripping station, where the molds are re-
moved. The ingots are then transferred to soaking pits where they are soaked to equal-
ize the temperature throughout the ingot and heated to the desired rolling temperature.
The heated ingot is removed from the soaking pit and transported to a roughing mill
where it is rolled into billets or blooms, or to a slabbing mill where it is rolled into
slabs. Billets, blooms, and slabs differ in size and shape. Billets are usually square
and measure 2 x 2 to 5 x 5 inches. Blooms are usually square or slightly oblong and
measure 6 x 6 to 12 x 12 inches. Slabs are always oblong and measure 2 to 12 inches
thick and 20 to 70 inches wide. Billets are used to produce bar and light merchant prod-
ucts, blooms to make heavier merchant products and structural products, and slabs to
make strip, sheet, and plate products.
Continuous Casting. For many years, steel producers recognized that continuous
casting was possibly the ultimate method for the conversion of molten steel to semi-
finished products. The first United States patent was issued to Sir Henry Bessemer in
1865. There was a period of experimentation on a pilot-plant scale in the United States
about 1940. Mechanical and material problems prevented early development of this
process, and it was not until 1943 that the first continuous- casting installation was suc-
cessfully operated in Germany. This was followed by further work in the United States
in 1946, Austria in 1947, the United States again in 1949, and Germany in 1950. In the
following years the Russian technologists devoted a great amount of effort to the process
and succeeded in placing several commercial plants into operation. Efforts in Europe
and the United States resulted in commercial installations in the 1950 IS. The estimated
capacity for continuous casting of raw steel in the United States in 1968 was about 7 mil-
lion net tons. Capacity now under construction is expected to increase this figure to an
estimated 14 million net tons in 1969, and 16 million by 1971.
Continuous-casting machines are of three general types: (1) vertical machines,
(2) vertical machines with bending rolls, and (3) curved-mold machines. These are
illustrated in Figures A-21 (18) and A-22(l9).
Pressure Casting. Pressure casting is a relatively new method for converting
molten steel into semi-finished products. The process was originally developed for the
manufacture of cast-steel freight-car wheels. Additional research and development led
to applic:ation of the proces s for making cast slabs. The first commercial installations
were for the production of stainles s steel slabs. Construction on the first plant des igned
to make plain-carbon-steel slabs was started in 1968 with start-up scheduled for early
1969. Total production of pressure-cast steel in 1967 is estimated to be 500,000 net
tons, all of it essentially stainless steel. With the addition of the carbon-steel facility
in 1969 this figure should increase to 1 million net tons annually.
The principal parts of a pressure-casting unit are shown schematically in Fig-
ure A-23. Operation consists of (1) placing a ladle of molten steel in the pressure tank,
(2) covering the tank with a special cover that includes a special preheated ceramic tube,
(3) positioning 'the mold over the tank and pouring tube, (4) introducing air pressure into
the tank and forcing the molten steel up the ceramic tube into the mold cavity, (5) seal-
ing the pouring tube with a refractory plug to prevent the flow of the molten steel from
the mold back into the ladle, (6) releasing pressure in the tank, and (7) moving the mold
from the tank and positioning another mold for another casting cycle. The mold is held
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-26
.,. -- .I..aollom-pour lad I.
J'
Siopper......
.. ..... Trou;h tun dish
.......u Woter-cool.d copper mold
..m.Osciliotin; mold tobl.
..''''''''.S.condory spray cool in;
.".'-""Withdrowol rolls
, ............Clomp
........Counterw.i;hltd cutoff from.

-------
A-27
in the closed position until the slab has solidified, after which the mold is opened, the
hot slab removed, and the mold prepared for another cycle of casting.
Entry for
pressur ized
air
Grophite molds
Steel slab
Mold-clomping
mechonism
Top of pressure
tank

Ceramic tube
Molten steel

Ladle
Base of pressure
tank
A 47"2
FIGURE A-23.
INSTALLATION FOR PRESSURE CASTING A SLAB
Manufacture of Finished Products
Rolling of ingots into billets, blooms, or slabs rarely yields a defect-free product.
Consequently, additional work must be done to condition the semifinished products be-
fore they are processed further. Conditioning is done by grinding, chipping, or scarf-
ing, depending on the type of steel involved and the kinds of defects that must be re-
moved. Grinding is done with conventional abrasive grinders, chipping with hand-held
chipping hammers or special equipment known as "peelers", and scarfing can be done
with hand torches, or with automated equipment that has recently become available.
Slabs are generally scarfed automatically before they enter the hot-strip mill.
The conditioned billets, blooms, or slabs are reheated to the required rolling
temperatures in special furnaces that are fired with fuel gas. After being reheated to
the desired temperature, the billets, blooms, or slabs are transferred out of the fur-
nace and transported to the hot mills where the rolling is done.
Billets and blooms are processed into bar, wire products, and structurals of vari-
ous weights and sizes. Air-borne emissions are not a problem in the production of
these products. Iron-oxide scale is formed during the time the semifinished products
are reheated for rolling, and is broken off by high-pressure water sprays as the steel
enters the first rolling stand of the mill. This scale is collected in scale pits and sent
to a reclamation plant for recycling (usually) to the blast furnace or sinter plant.
The processes that are of interest to this project are the surface-treatment opera-
tions such as acid-pickling lines, blast descaling, tin lines, galvanizing lines, plastic-
coating lines, and other coating operations. Thes e proces s eS are carried out continu-
ously with provision for the supply of steel via a "pay-off" coil at the process input end,
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-28
and a similar provision for removing the finished, coiled product from the "re- COil"
end. A horizontal continuous processing line that is typical of older plants is illustrated
in Figure A-24, and the vertical type of processing line that is representative of newer
processing lines is illustrated in Figure A-25(20). Equipment designs vary with the
process, space available in the steel plant, and preferences of steel-plant personnel.
.350'
RECOIL
FIGURE A-24.
NATURE OF HORIZONTAL PROCESSING LINE
120'
PAYOFF
,;
PROCESS
" "'~'\~"'~ "'~~~~~'" "''''''''''~~~~~'''''''''~~~~~'''~~~~~~'''~~'''~~'''~'
FIGURE A-25.
NATURE OF VERTICAL PROCESSING LINE
Auxiliary Operations
Two operations that usually are not considered as process segments in the making
or iron and steel, but nevertheless incidental to the manufacture of steel, are the found-
ries and incineration facilities associated with iron and steel plants.
Foundries. The foundry installations associated with iron and steel plants are pri-
marily used as a means of supplying castings for maintenance purposes, and in this
capacity are usually under the management of the maintenance department. In the large
majority of steelworks installations, the captive foundries have modest facilities for the
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-29
melting of nonferrous alloys such as aluminum, brass, and bronze. These facilities are
usually well ventilated and their emissions collected. Molten steel or iron required for
large castings sometimes is obtained from the existing facilities located at the steel
plant. Iron can be obtained directly from the blast furnace. Additions of steel or alloy-
ing elements are made to obtain certain metallurgical properties. Molten steel is trans-
ported to the foundry for pouring into molds prepared for the purpose. Some steelworks
foundries have cupolas or electric furnaces for melting ferrous metals for castings.
Two major foundry items required for the production of steel mayor may not be made
within the steelmaking complex. These are (1) ingot molds required for the transforma-
tion of molten steel into a solid product for additionai processing, and (2) iron and steel
rolls necessary for rolling ingots into finished products.
Most steel companies purchase ingot molds -from foundries that are in the business
of supplying this specialized item to the steel industry. In some cases, the ingot-mold
companie s have facilities adjacent to the steel plant and obtain the iron for casting the
molds from the steel-plant blast furnaces. If a s.ingle steel plant is large enough, or if
several steel plants of one company are centrally located, and in the final case, if a steel
plant is in an isolated location with respect to available outside ingot-mold sources, the
company may have its own in-plant facilities for making ingot molds.
Iron and steel rolls used in the rolling of ingots to semifinished and finished prod-
ucts are continually replaced to maintain desirable quality standards. The manufacture
of rolls is a much more specialized operation, technically and process-wise, than the
making of ingot molds. Because of this highly specialized requirement, few steel com-
panies make their own rolls, preferring to purchase them from companies specializing
in this item.
Incineration Facilities
The making of iron and steel requires the use of many materials that are not a part
of the ironmaking and steelmaking processes, but are necessary adjuncts. These ma-
terials include (1) wood from pallets used to ship refractory brick into a steel plant,
(2) paper or plastic bags used in the shipping, storing, and handling of various required
materials, (3) paper scraps that result from the various packaging and shipping opera-
tions, and (4) various other solid-waste materials that are generated in the iron and
steel plants. Considering the large amounts of such materials that are used, the prob-
lem of disposal of solid waste is considerable in the iron and steel industry. Therefore,
incineration is commonly practiced in most steelworks.
MAJOR REFERENCES FOR APPENDIX A
(1 )
Schueneman, J. J., et al., "Ai.r Pollution Aspects of the Iron and Steel Industry",
Public Health Service, Cincinnati, Ohio, pp 10-27, June 1963. PB 168 867, U. S.
Department of Commerce, Clearinghouse for Federal Scientific and Technical
Info rmation.
(2 )
"The Making, Shaping, and Treating of Steel", Eighth Edition, 1964, United States
Steel Corporation, Pittsburgh, Pennsylvania.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
A-30
(3) "Annual Statistical Report, American Iron and Steel Institute, 1968", American
Iron and Steel Institute, New York, N. Y., 1968, p. 80.
(4) Communication from the American Iron Ore Association, September 12, 1968.
(5) "Empire Mine is Dedicated - Production Exceeds 1,200,000 Tons of Pellets",
Iron and Steel Engineer, !!. (6), 179 (June 1964).
(6) "Lime for Steelmaking: Tailored to Fit New Demands", 33/The Magazine of
Metals Producing, ~ (1), 58-93 (January 1967).
(7) Ischebeck, P., et al., "Injection of Heavy Oil Into a Blast Furnace at High Blast
Temperatures and With Fully Beneficiated Burden", Stahl und Eisen, 83 (24),
1541-1546 (November 21, 1963). . -
(8)
Bell, S. A., et aI., "Coal Injection-Bellfonte Furnace", Journal of Metals, 20 (4),
85- 88 (April 1968).
(9)
"Bessemer Converters Fading Away - Almost", American Metal Market, May 28,
1968.
(10) Kalling, B., and Johansson, F., 'IStorals Kal-do Rotary Oxygen Steelmaking
Process", Blast Furnace and Steel Plant, 45 (2), 200-203 (February 1957).
(11 )
(12 )
(13 )
(14)
(15 )
(16 )
(17)
"Electric Furnace Round-Up", 33/The Magazine of Metals Producing, ~ (6), 71-80
(June 1966).
"Everybody Is Getting Into the 'High Power' Act", The Iron Age, 202 (2), 22-23
(July 11, 1968).
Metcalf, A., "Oxy-Fuel Steelmaking for the Fuel Oxygen Scrap (FOS) Process",
Steel and Coal, 1266-1268 (December 27, 1963).
Trilli, L. J., "Oxygen-Fuel Roof Burner Design and Operation-III at Inland Steel",
Journal of Metals, ~ (9), 1059-1060 (September 1966).
"Pittsburgh Ups the Ante on BOF Scrap",. 33/ The Magazine of Metals Producing,
~ (3), 69-82 (March 1968).
Gockstetter, G. J., "Oxygen-Fuel Roof Burner Design and Operation-I At Republic
Steel", Journal of Metals, ~ (9), 1055-1057 (September 1966).
Groen, R. G., "Scrap Preheating in The Basic Oxygen Furnace at Wisconsin Steel
Works", Journal of Metals, ~ (4), 478-483 (April 1966).
(18) Jaicks, F. G., et a!., "Review of Paper on Continuous Casting of Three Types of
Low-Carbon Steel", AIME Open Hearth Proceedings, 40, 67- 84 (1957).
(19) Shah, R., "Curved Mold Lowers Silhouette of Continuous Casting Line", The Iron
Age, 192, 58-59 (August 1, 1963). --------
(20) Foreman, A. R., "New Shapes in Processing Lines", Iron and Steel Engineer, 42
(9), 181-r83 (September 1965).

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
B-1
APPENDIX B
GENERAL DESCRIPTION OF
AIR-POLLUTION CONTROL EQUIPMENT
Many processes involved in the making of iron and steel create particulate and
gaseous emissions that result in air-pollution problems of varying degrees. Some of
the emissions are nothing more than simple dusts, occurring in rather small amounts,
that can be removed by equipment of somewhat simple design. Other particulate emis-
sions are more complicated in character and require more complicated equipment to
achieve desired levels of discharge to the atmosphere. Capture of different particulate
emissions generated in the making of iron and steel also requires different amounts of
energy (and involves different operating costs) to achieve acceptable dust loadings to the
atmosphere. One example of typical variations in energy requirements for various exit-
dust loadings is shown in Figure B-l. (l)'"
  60   
 Q) 50   
 CI    
A::J    
0.0    
001 40   
~    
c~    
 Q)    
 -    
~o 30   
::J ~   
~o    
Q)  20   
~cn   
Q..Q)    
 ..s:::    
 u    
 c: 10   
  0 .05 .10 .15
FIGURE B-1.
Exit Dust Loading t grains per standard cubic foot
EXAMPLE OF COMPARATIVE ENERGY LEVELS TO
MAINTAIN TYPICAL EXIT-DUST LEVELS
In addition to particulate emissions, some processes in the steel industry. produce
gaseous emissions that require the application of chemical methods if they are to be
captured.
One of the outstanding methods for control of emis sions, e specially for control
of gaseous emissions, involves the installation and application of automatic cont.rols
on combustion equipment. This method attacks the problem by inhibiting the forma-
tion of some undesirable components rather than by collecting them after they are
formed.
.References cited in this appendix are listed at the end of Appendix B.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
B-2
Although the design and nature of emission-control equipment is vitally important
to the effectiveness of collection, the behavior and efficiency of even the best and most
suitable equipment in a particular situation is affected drastically by factors such as the
age and condition of the equipment, the skill and attitudes of its operator s, and the
degree of attention given to regular and sufficient maintenance.

Particulate emissions in the integrated iron and steel industry can be classified
in three forms(2):
(l) Emis sions containing particulate matter in the form of relatively
coarse particles, a substantial portion in sizes above 10 microns.'~
This includes grit from combustion processes, kilns and calciners,
grinding and screening operations, and from driers.
(2) Emissions in which the majority of the particles are between 1 micron
and 10 microns, and arise from steelmaking processes. The fine air-
borne particles found in industrial atmospheres are in this size range.
(3) The third size range of emissions includes true fumes, which are pre-
dominantly below 1 micron. They may produce industrial haze, either
alone, in combination with normal atmospheric fogs, or after interac-
tion with other contaminants in the atmosphere. Typical of these
are fumes from oxygen- blown open hearths, electric-arc furnaces,
and basic-oxgyen (BOF) furnaces.
Equipment that can be used to control air pollution in the integrated iron and steel
industry can be classified into four general groups which are: (1) cyclone dust col-
lectors, (2) electrostatic precipitators, (3) bag filters, and (4) wet scrubbers, including
spray scrubbers. (2, 3)
Cyclone Dust Collectors
The principle of operation of cyclone separators is the imposition of a centrifugal
acceleration on gas-borne particles. This is usually achieved by admitting the dust-
laden gas tangentially to the periphery of a cylindrical vessel, resulting in a spiraling
flow pattern that causes the solid particles to be thrown outwards to the wall of the
ves sel, where they fall to a conical discharge pipe. The clean-gas exit is located on
the axis of the vessel. The configuration and relative size of the gas inlet and outlet
pipes is governed by the required characteristics of the unit. High-throughput units
differ from high-efficiency units by having larger inlet and exit areas for a vessel of a
given diameter. High-efficiency and high-throughput cyclones are illustrated in
Figure B-2. In theory, small-diameter cyclones will have an efficiency~":' superior to
larger units of similar proportions. High throughput combined with high efficiency is
sometimes achieved by nesting a n.umber of small cyclones into a single unit as shown
in Figure B-3. However, there is a hazard that units containing small nested cyclones
can develop poor performance characteristics because of blocking of the small solids-
discharge pipes, which results in uneven distribution of gas.
. 1 micron = 0.001 mm.
.. Efficiency refers to the amount of particulate matter removed by the control system, expressed as a percentage of the amount
of particulate matter in the entering gas.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
B-3
+EXIT
a.
High-Efficiency Cyclone
High-Throughput Cyclone
b.
FIGURE B-2.
TYPICAL CYCLONE CONFIGURATIONS
~

DUST lLADEN GAS
FIGURE B-3.
NESTED TUBULAR CYCLONES
-""/ TUBE PLATE \
CLEANED
GAS
~
FIGURE B-4. CELLULAR CYCLONE
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
B-4
. . The efficiency of a cyclone can be increased by introducing water into the system.
ThIs IS usually done with a ring main located just below the top cover plate of the
cyclone.
Medium-efficiency cyclones (such as the cellular cyclone illustrated in Figure B-4)
are. soz:ne~imes used to reduce the load on subsequent dust-removal equipment. Swirling
achon IS mduced by passing the gas through pitched vanes.
Electrostatic Precipitators
The action of electrostatic precipitators is based on the passage of a dust-laden
gas through an intense electrostatic field between electrodes of opposite polarity. The
particles take up an electrical charge from the discharge. electrodes, and are acceler-
ated towards the grounded electrodes or collector plates, which are at opposite polarity.
The dust particles give up their charge and are deposited as a layer on the collector
plates. The layers of dust are usually removed from the collector plates by periodic
rapping that causes the dust to fall into two collecting hoppers.
Two general types of electrodes are in general use: (I) the wire-in-tube system
as shown in Figure B-S, and (2) the wire-and-plate type shown in Figure B-6. There
are many refinements of electrode configuration employed by equipment manufacturers.
A schematic illustration of a full-size wire-and-plate electrostatic precipitator is shown
in Figure B-7.
GROUNDED
~ COLLECTOR
TUBES
GAS
FLOW
GAS FLOW t
DISCHARGE
ELECTRODES
FIGURE B-S. WIRE-IN-TUBE
ELECTROSTATIC PRECIPITATOR
FIGURE B-6. WIRE-AND-PLATE
ELECTROSTATIC PRECIPITATOR
When designing electrostatic precipitators, a great deal of attention must be paid
to the quantities of dust that must be handled, and large-scale pilot-tests are often re-
quired to obtain accurate assessment of precipitator performance. Lack of attention to
such items as the resistivity and stickiness of dusts has led to serious malfunctioning
of precipitator installations, because of excessive buildup of the voltage gradient across
the collected dust. This results in local areas of intense electrical discharge and leads
to the effect known as "back ionization". This back ionization results in current re-
quirements in exces s of the capacity of the electrical equipment, which causes a drop
in voltage and results in poor collection efficiency.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
B-5
FIGURE B-7.
INTERNAL CONSTRUCTION OF WIRE-AND-PLATE
ELECTROSTATIC PRECIPITATOR
Bag Filters
The filter cloth of bag filters consists of threads about 500 microns in diameter
which are spaced 100 to 200 microns apart, thereby forming a sieve with large openings.
However, the openings are criss-crossed by fine fibers about 5 to 10 microns in diam-
eter, which are the individual fabric fibers. The fine fibers form effective impingement
targets and can remove a high portion of submicron particles. (4) A diagram of a typical
filter fabric is illustrated in Figure B-3.
During the passage of a dusty gas through the fabric, particles will impinge
upon, and be retained by, these fine fibers and cause a buildup of a layer of solid
material on the fabric. If the gas velocity through the fabric is low enough, this solid
accumulation will be in the form of a loose floc that will effectively trap even submicron-
sized particles. The progressive accumulation of solid material on the fabric eventually
leads to an excessively high pressure drop in the system, or to a local breakdown of the
filter bed (i. e., the accumulated dust). The necessity for the periodic cleaning of the
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
B-6
filter fabric results in a tendency for the efficiency to vary in a cycle-like manner, as
the floc filter bed progres sively builds up, and then cyclically is removed. The opera-
tion of multiple- bag units enables collection efficiency approaching 100 percent to be
obtained. Care must be taken not to cool the entering gases to below their dew point.
The condensation of moisture on the bags leads to rapid blinding of the fabric and
usually forces shutting down the unit for cleaning.
\
\
t
Main strands, 500
Fine fibers, 5 to 10 microns
in diameter
microns in diameter
FIGURE B-8.
DIAGRAM OF A TYPICAL FILTER FABRIC
Bag-filter fabrics must have properties that will permit them to operate in vari-
ous atmospheres and at various temperatures. Operating atmospheres and temperature
limitations for some representative bag-house fabrics are listed in Table B-l(5J. The
service life of fiberglas s bags has been extended somewhat by treating the fabric with
silicone compounds.
TABLE B-1.
OPERATING CONDITIONS FOR TYPICAL BAGHOUSE FABRICS
Fabric
General Use
Maximum
Temperature, F
Cotton
Noncorrosive or mildly alkaline dusts and gases
180
Wool
Mildly acid conditions
200-215
Nylon
Alkaline dusts or gases unsuitable for acid conditions
200-215
Dynel
More acid- resistant than wool or Orlon
200-215
Orlon
Widely used for corrOSlve gases
250-275
Terylene
Widely used for corrosive gases
275-300
Fibe rglas s
Resistant to most gases except hydrogen fluoride
500-650
Bag filters are capable of handling gases with medium to high dust concentrations.
Three general types of bag filters are generally used: (1) low-velocity filters, (2)
(2) shaker-type bag filters, and (3) the reverse-jet filters.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
B-7
Low-Velocity Bag Filter
This type of bag filter consists of multiple fabric bags suspended vertically in a
box housing as shown in Figure B-9. A relatively simple bag-shaking gear is used. A
major disadvantage of the low-velocity bag filter is that, to prevent blinding of the fabric,
gas velocities are low (about 3 feet per minute). Higher velocities tend to drive particles
into the fabric making it difficult to remove the dust by simple shaking.
~' '. DUST-LADEN
, ,
, GAS
FIGURE B-9.
LOW-VELOCITY BAG FILTER
Shaker-Type Bag Filter
The shaker-type filter is similar in configuration to the low-velocity bag filter
except that it is equipped with an automatic rapping gear which may be actuated on a
predetermined cycle, or on reaching a certain pressure drop across the filter. The
.3haker-type filter typically is able to tolerate a face velocity of about 6 feet per minute
when dust loadings are low because of the more efficient method of bag cleaning. The
more cornman face velocities for bag collectors used in the steel industry are 3 feet
per minute or less. The shaker-type bag filter is illustrated in Figure B-lO.
Reverse-Jet Filter
In order for normal bag filters to operate at high efficiency, it is necessary to
provide large areas for filtration and avoid frequent cleaning, which in turn limits the
maximum dust concentrations that can be handled. The reverse-jet filter has design
and operating characteristics intended to overcome these limitations of normal bag fil-
ters. These characteristics of reverse-jet filters are as follows:

(1) A felt of compressed wool is used as the filter fabric and has a suf-
ficiently close texture that it can act as an effective filter without the
requirement that a floc must build up to aid in dust-retention capability.
(2) The deposited dust is removed by a reverse current of air from an
external blow ring which traverses the length of the bag, usually con-
tinuously. The construction of a typical reverse-jet filter is illustrated
in Figure B-ll.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
B-8
FIL TER BAGS
FIGURE B-10,
SHAKER-TYPE BAG FILTER
1 DUST-lADEN
, GAS
INLET MANIFOLD
FEl T Fll TER
TUBES
TRA VERSING
BLOW-RING
FRAME
REVERSE JET
AIR HOSE
CLEANED GAS
FIGURE B-ll.
REVERSE-JET BAG FILTER
Limitations on operating velocities are less than for normal bag filters, because
the requirement for deposited dust to improve efficiency is eliminated. Face velocities
of 10 feet per minute are common, and in some cases the face velocities can be as high
as 15 to 20 feet per minute. Efficiencies are close to 100 percent for particles down to
about 1 micron in size.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
B-9
A modification that has been developed is called a "pulse-jet filter". Bag-cleaning
is achieved by blowing strong jets of air into the bags from the "clean" side of the fabric,
thus causing a reversal in air flow and agitation of the bags. The bags may be cleaned
in sequence so as to maintain consistent collection efficiency. However, difficulty may
be encountered in the removal of material from the fabric, and a progressive buildup of
dust may result in an increased pressure drop in the system and a requirement for
periodic removal of bags for cleaning.
Wet Scrubbers
Wet scrubbers can be divided into two categories, (1) wet-impingement scrubbers
and (2) spray scrubbers. The principle of operation in these scrubbers is that when
dust-laden gas impinges on a liquid body, the gas will be deflected around or away from
the liquid body, but the dust particles (having greater inertia) will tend to collide with
the surface of the liquid and be subject to a retaining force.
Wet-Impingement Scrubbers
Wet-impingement scrubbers are dependent on a layer of water as the entraining
medium for dust particles. Irrigated-target scrubbers, orifice-plate scrubbers, and
disintegrators fall in this category.
Irrigated-Target Scrubbers. This scrubber functions by passing a gas upward
through a flooded perforated plate so that the liquid on the top of the plate is atomized
at the edges of the orifices. This atomization creates a dust-trapping spray directed
at targets located above the orifices, as illustrated in Figure B -12. An important fea-
ture of this design is its apparent freedom from choking of the holes in the orifice plate.
Scrubbers of this type may incorporate several orifice plates in series, two or three
being usual. Each plate imposes a pressure drop of about 3 inches of water gage.
TARGET
PLATE
WATER DROPLETS
ATOMIZED AT EDGES
OF ORIFICES
FIGURE B-12.
IRRlGATED-TARGET SCRUBBER
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
B-lO
Orifice-Plate Scrubbers. Scrubbers of this type operate on essentially the same
principle as the irrigated-target scrubber, except for the omission of the target above
each orifice. Scrubbers of this type have been used in the cleaning of blast-furnace
gas. (6)
Disintegrators. Disintegrator scrubbers have found use in the past in cleaning
blast-furnace gases. (7) In scrubbers in which a spray is generated, the energy re-
quired to atomize the liquid is obtained at the expense of a pres sure drop or by pumping
the liquid through nozzles. Because the collection efficiency increases with the in-
creased relative velocity between the liquid droplet and particle, higher efficiencies
are attained by increasing the energy input to the system. In the case of the disin-
tegrator, this increased energy is obtained by passing both the dirty gas and the
scrubbing liquid into the intermeshing vanes of a stator and high-speed rotor as shown
in Figure B-l3. Disintegrators are relatively inexpensive when their high performance
is considered. However, their energy and water consumptions are high.
DUST-LADEN GAS
FIGURE B-13.
DISINTEGRA TOR
Spray Scrubbers
Spray scrubbers are those in which the scrubbing liquid is broken into a spray to
form a large number of collection sites. The efficiency of spray scrubbers is improved
by increasing the relative velocity between the spray droplets and dust particles, thus
raising the collision rate between particles and droplets. Spray towers, venturi
scrubbers, and flooded-disk scrubbers fall into this class.
Spray Towers. The spray tower (as illustrated in Figure B-l4) has become some-
what obsolete because of its relatively high cost. However, where the structures exist
as a part of original blast-furnace installation, the spray towers sometimes are used as
precoolers for the large quantities of gas involved. Spray towers have the advantage that
no very close clearances are involved, and as a result, the unit can handle relatively
high dust concentrations without suffering from choking problems. In addition, because
very fine spray is not involved, the spray generators do not require fine jets and reli-
ability is improved. Also, the spray water often can be recirculated until it contains
quite a high concentration of solids.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
B-ll
FIGURE B-14.
SPRAY TOWER
Venturi Scrubbers. Venturi scrubbers are characterized by their high collection
efficiency over a large range of particle sizes and their correspondingly high pressure
drop.
Water is admitted from jets to the throat of a constriction in the duct carrying the
dusty gas. The high gas velocity atomizes the water, and the rapid acceleration of the
droplets of water leads to a very high particle-droplet collision rate. From the venturi,
the gas is then passed to a cyclone where the agglomerated particles that are now
relatively large can be separated easily. A simple venturi scrubber is illustrated in
Figure B-15. Optimizing of the performance of high-energy scrubbers (such as venturi
scrubbers) is important, because the power consumption of such scrubbers can be quite
high. Optimizing of venturi- scrubber performance has led to the development of
scrubbers whose pressure drop and performance can be controlled while the scrubber
is operating, either to deal with varying gas rates or to maintain a given efficiency
during a particular period of operation. A variable-throat venturi scrubber that oper-
ates in this manner is illustrated in Figure B -16. A sc rubber of this type can be oper-
ated for short periods of time when the proces s requirements demand it, at a pressure
drop that would be impractical or uneconomical for continuous operation.
PARALLEL THROAT

\ \l..J WATER TO RADIAL NOZZLES
\... DUST LADEN GAS INLET
TO CYCLONE
SEPARATOR ~
.AI"
FIGURE B-15.
VENTURI SCRUBBER
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
B-12
FIGURE B-16.
VARIABLE-THROAT VENTURI SCRUBBER
Another modification of a venturi scrubber is shown in Figure B-17. Fine nozzles
and jets are omitted, and slurries can be used as the cleaning medium. The liquid flows
down the sides of the scrubber toward the venturi throat where it cascades and forms an
atomized spray to perform the entrapping function on the particles.
CLEANED
GAS
FIGURE B-17.
VENTURI SCRUBBER USING SLURRIES
Flooded-Disk Scrubber. In the flooded-disk scrubber, an atomized spray is ob-
tained by positioning a rotating disk in the path of a dusty gas and flooding the surface
of the disk with water, as shown in Figure B-lB. An atomized spray generated in this
way has good particle-collection characteristics. Also it is possible to vary the posi-
tion of the disk in the tapered throat and allow for fluctuations in gas throughput.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
B-13 and B-l4
/ TO CYCLONE
SEPARATOR
AUTOMATIC
DISC POSITIONER
FIGURE B-l8.
FLOODED-DISK SCRUBBER
MAJOR REFERENCES FOR APPENDIX B
(1)
Borenstein, M., "Air Pollution Control for the Iron and Steel Making Processes",
Industrial Heating, 34(9), 1646-1648 (September 1967).
(2)
Stairmand, C. J., "Removal of Grit, Dust, and Fume From Exhaust Gases From
Chemical Engineering Processes", Chemical Engineer, No. 194, pp. CE310-
CE324 (December 1965).
(3)
Fox, M. R., "Dust Arrestment", W. S. Atkins Bulletin, No. 10, 29-40
(Summer 1966).
(4)
Stairmand, C. J., "The Design and Performance of Modern Gas-Cleaning Equip-
ment", Journal of the Institute of Fuel, 29 (181), 58-76 (February 1956).
(5)
Squires, B. J., "Fabric Filter Dust Collectors", Chemical and Process
Engineering, 43, 156-159 (April 1962).
(6)
Lowe, J. R., "An Orifice Gas Washer", AIME Blast Furnace, Coke Oven and
Raw Materials Proceedings, ~, 28-30 (1957).
(7)
Reid, G. E., "Experience in Cleaning Blast Furnace Gas with the Orifice
Washer", Iron and Steel Engineer, ~ (8), 134-13 7 (August 1960).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-l
APPENDIX C
CHARACTERISTICS OF EMISSIONS BY THE
INTEGRATED IRON AND STEEL INDUSTRY
The manufacture of iron and steel involves many different processes until semi-
finished or finished products are available for sale or further use. Some of the pro-
cesses can produce large quanitities of particulate and gaseous emissions, while other
processes are relatively free of air-pollution problems. Sources of emissions, their
characteristics, amounts generated, and types of equipment used to control emissions
from the various processes are discussed in this Appendix. Processes are considered
in the same order as described in Appendix A; "Processes in the Integrated Iron and
Steel Industry". However, as a guide to orientation, major sources of air pollution by
the integrated iron and steel industry include coke -oven plants, sintering plants, blast-
furnace operations, steelmaking furnaces (especially those using large amounts of
oxygen for steelmaking), and boiler plants.
Preparation of Raw Materials
Raw-material preparation within integrated steel plants includes the receipt,
stocking and de-stocking, sizing, and agglomeration of iron ores, fluxes, and miscel-
laneous charge materials. For the purposes of this study, the preparation of raw
materials is considered under the following operations: (1) receiving of raw materials,
(2) preparation of iron ore, (3) making sinter, (4) making pellets, (5) preparation of
limestone and lime, and (6) making coke.
Receiving of Raw Materials
Raw materials may arrive at an integrated steel plant via road, rail, or water,
and most are unloaded to open stockpiles. In the case of water transport, boats or
barges typically are unloaded to a receiving trough and then transferred to stock by a
gantry crane. The use of long-span gantry cranes with large clam- shell buckets (known
as ore bridges) is almost universal in the United States for handling bulk raw materials
at large steel plants.
Except for scrap iron and steel, coarse materials are generally broken to a maxi-
mum size of 2 to 4 inches. There is a trend within the industry to smaller sizes, and
also a trend toward performing the crushing and sizing operations at the mines.
Materials containing appreciable fines are screened, usually to remove sizes smaller
than 3/8 or 1/2 inch. This operation is also being done more at the mine, with a
resultant product shipped in a closely held siz.e range. Emissions from raw-material
receiving operations consist mostly of dust from handling th~ ores and fluxes.
The chemical nature of iron ores and fluxes is such that particulate dust from these
materials is mainly a nuisance, rather than a health hazard. Dusts from iron ores are
iron oxides combined with oxides of silicon, calcium, and magnesium. Dusts from
fluxes such as limestone and dolomite contain calcium and magnesium carbonates,
aluminum oxide, and silica. Fluorspar (another flux used in ironmaking and steelmaking)
contains calcium fluoride, calcium carbonate, and oxides of iron, aluminum, and
silicon.
BATTELLE MEMORIAL. INSTITUTE - COLUMBUS LABORATORIES

-------
C-2
Dust arises mainly from the free -fall handling of ores during unloading, transfer'
to stock, and transfer from stock to processing or to the blast furnace. Most in-plant
handling is via the overhead clam-bucket gantry (the ore bridge), which removes ore
from the receiving area to the storage piles, and from the piles to bottom-dump trans-
fer cars operating usually on an elevated section of railway. Because of its large scale,
the operation is virtually unshroudable, and often is elevated and exposed to wind. Good
control over dust as a nuisance depends upon careful ore-bridge operation and attention
to weather. One reason why ore manipulation gives little trouble is that the huge ore
bridges must be partly immobilized (for reasons of mechanical safety) when the wind
velocity rises to about 35 knots.
Storage piles are usually accumulated between the high piers of the ore-bridge
rails, hence accumulated dust tends to stay in place even in a stiff breeze. Two areas
which normally require special attention to housekeeping are the "high-line" (where
the transfer car runs) and the dock or receiving area. These places, if allowed to
accumulate fine dust, can be sources of particulate emission during windy weather.
Some materials are worse than others in terms of their natural tendency to form
dust. In the United States, there is a strong trend to the use of mas sive pre sized lump
ores and to artificial marble-size agglomerates of iron oXide in the form of pellets.
These materials seldom contain much dust. Those ores most likely to contain
appreciable dust (such as sinter fines) are sometimes wetted deliberately to obtain
some control over dusting. Spraying piles of ore and coal with oil or with plastics and
other coatings to minimize dust problems has been cited by some as an inadequate means
for control( 1)':', while other plants have found the method helpful.
Preparation of Iron Ore
Ores and fluxes are now usually crushed and screened at the mines before shipment
to the steel plant. This is a strong general trend in the American industry, but it has
recently been demonstrated that there often is a technological advantage to re-screening
of raw materials just prior to charging them to the blast furnace(2-7). A ~umber of
blast-furnace plants have installed final screening systems. Whether these final screen-
ing systems are located in the stock yard or within the charging stockhouses, they are
considered a part of ore preparation, and are, of course, potential sources of dust
emission. However, these dusts usually are confined to the screening installations by
the use of shrouding on the screens, closed reception and conveyance of fines, and
exhaust systems where they may be appropriate.
Making Sinte r
Originally, sintering--was used as a means of recovering fines (flue dust) produced
during the making of iron in blast furnaces. However, it is now recognized that fine ore
bought at lower cost than lump ores, then combined with flux and sintered, can be con-
verted into a useful and economical burden material for blast furnaces. As shown in
in Figure C-1, the use of sinter in the blast furnace burden in the United States has
increased over the years, but in 1964 the total usage started to decrease, mostly because
of increased use of pellets.
"References for Appendix C are given at the end of Appendix C.
BATTELLE MEMORIAL 'INSTITUTE ~ COLUMBUS LABORATORIES

-------
~
o
Q)
>-
a> 50
Q.
If)
C
o
:: 40
Q)
c
'+-
o
If)
c
o
E 20
c 
.Q 
- 
Q. 10
E
:J 
If) 
C 
0 
U 
FIGURE C-l.
C-3
60
30
/--
/
/
/"
/
I
I
",,__I Pellets
.,""
--
/
'"
_J
o
1958 1959 1960 1961 1962 196319641965/966 1967 1968
Year
CONSUMPTION OF SINTER AND PELLETS IN BLAST
FURNACES IN THE UNITED STATES
The main feed to most sintering machines is iron ore that is too fine in particle
size to be used directly in the blast furnace. The fuel usually is coke "breeze" that also
is too fine to be used directly in the blast furnace, although sometimes coarser coke is
crushed down to sinter-plant size if the plant has a deficiency of fine-size coke. The
foregoing is somewhat of an oversimplification, however, because many sintering plants
in integrated steel plants are fed with a wide variety of iron-bearing and carbon-bearing
materials useless for other purposes. Among materials of this type are mill scale
from rolling operations and certain types of dusts from steelmaking furnaces. In order
to be acceptable as feed to a sintering machine, dust from a steelmaking operation must
be low in zinc content. If the charge to the steelmaking furnace contains a substantial
amount of galvanized scrap, the dust will contain zinc. If the level of zinc content in the
dust is appreciable, the dust or sinter made from it would have an adverse effect on the
refractories in a blast furnace. In summary, .sintering machines generally accept and
proces s a wide variety of feeds, differing from plant to plant and sometimes from week
to week in each plant, and produce a considerable quantity of emissions of uncertain and
variable quantity and nature.
In modern, 'high-production sintering operations, the effective control and capture
of proces s dust is all but imperative even for material maintenance demands. Abrasive
dust has, in many instances, caused a major problem in maintaining operation of the
fans in the sintering machines(8, 9,10).
Identification of Emissions. Emissions from sinter-plant operations usually con-
sist of the following: (1) minor amounts of dust in the handling and grinding of raw
materials, (2) dust that is sucked through the grate bars, (3) combustion gases from
ignition and firing, and (4) dust generated in the cooling and screening operations. The
points of emission are designated in the flow sheet for a sinter plant in Figure C-2.
Circled numbers on the flow sheet are indexes to the circled numbers in the following
discussion.
BATTELLE. MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
Sinter fines
Steelmaking
dust
Coarse ore
Fine ore
Air
C-4
Limestone
@
MIXING
DRUM
Coke
@
Blast-furnace
dust
Sinter mix
SINTER MACHINE
Hot sinter
SINTER COOLER
BLAST FURNACE
Ignition fuel
@
G)
FIGURE C-2. TYPICAL FLOW SHEET FOR A SINTERING PLANT
Circled num bers refer to emission characteristics
tagged with the same num ber in the accompanying text.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-5
Enrissions from Crushing. The ores and limestone used in the making of sinter
are naturally moist or are w.etted before and during the crushing and grinding operation.
Possible emissions are identified as follows:
@~)0
CD
Iron Ore Dust - particles are rounded to elongated in shape and can have
a size as small as 2 nricrons. Larger particles are opaque, and red-orange
in top light. Individual small grains are transparent and blood red. (11)
Hardness:
5 (Mohs)
Specific gravity:
5.2
Chemistry: Usually mostly Fe203 or Fe304; some silica and limestone,
mostly soluble in HCl
@
Limestone Dust - nrineral name calcite.- It is c;olorless, with light-
transmitting characteristics varying from transparent to translucent.
Particles generally occur as rhombohedra because of their good
cleavage. Fragments may also occur as prisms. (11)
Hardness:
Specific gravity:
Chemistry:
3
2. 7
mostly CaC03
0)
Coke Dust - Particles are opaque, irregularly shaped, quite porous and
rough with some straight, sharp edges. They are gray-black in reflected
light. ( 11 )
Chemistry: About 90 percent carbon.
Q)
Combusion Products - The gases leaving a sintering strand are a result
of the combustion of the coke in the sinter mix and of the fuels used to
ignite the sinter mix. Fuel for ignition is frequently coke-oven gas.
Because an excess of air is used during the making of sinter to provide
an oxidizing atmosphere, it is unlikely that any unburned hydrocarbons
exist in the combustion products. However, coke-oven gas does contain
sulfur compounds that combine with sulfur in the sinter-mix coke and as a
result contributes to the presence of sulfur dioxide in the combustion
products.
Sinter Dust - Dust may contain particles of iron oxides, calcite, iron-
calcium silicates, and quartz. Iron oxide can be opaque, black, rounded
particles of magnetite (Fe304) with granular faces, and/or dense, rounded
elongated, and nearly spherical agglomerates of hematite (Fe203)' Calcite
occurs as smooth, rounded particles, and quartz as a transparent, rounded
particle. The iron-calcium silicates are transparent, vitreous, colorless
to yellow to green. Particles are irregularly rounded with smooth
surfaces. (11)
@
No detailed information is yet available on the compositions of dusts going
to the atmosphere, but they undoubtedly have the same characteristics -
as above ("sinter dust"), with the exception that the particles are finer.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-6
Typical size distributions for fine, medium, and coarse dusts from sinter plants
are shown in Figure C -3. (12) The samples were taken from the wind box of a sintering
24
22
20
C 18
Q)
~ 16
Q)
a. 14
"0 12
~ 10
~ 8
0:: 6
-
-a, 4
~ 2
o
. Fine dust
~ Medium dust
o Coarse dust
16
20
70
100
Size I microns
140 200 zro 325
40
100
Equivalent U.S. Screen Series
FIGURE C-3.
SIZE DISTRIBUTIONS OF VARIOUS SINTER-PLANT DUSTS
machine. Reported results on sampling tests of a sintering machine operating in the
United States provided data on amounts of dust generated and the average size distribu-
tion of the particulates. (13) These results are given in Tables C-l and C-2.
TABLE C-l. SINTERING-MACHINE STACK-EMISSION TEST DATA
ON EXHAUST GAS DUST LOADING AT IGNITION END
 Test Numbers 
Conditions 1 2 3
Standard Cu. Ft. of Gas/Minute 195,500 198,800 221,700
Actual Dust Loadings   
Pounds of dust/1000 pounds of gas 0.230 0.236 0.295
Grains/standard cu. ft. of gas (32 F) 0.126 0.130 0.166
Maximum Stack Gas-Emission Rates   
Pounds/min 3.39 3.49 4.98
Pounds/hr 203.4 209.4 268.8
TABLE C-2. SCREEN ANALYSIS OF PARTICULATE EMISSION FROM A SINTERING MACHINE
Screen Size,

microns
Weight Retained,
percent
Cumulative Weight,

percent
5
10
20
30
44
25.1
47.6
14.6
5.8
5.0
25.1
72.7
87.3
93.1
98.1
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-7
Shown in Figure C-4 are the dust emissions, gas flow, and dust concentrations
from a sintering machine 92 feet long operating in the United Kingdom. (12)
~
e
o
'+= >.
E-5 10
co.:
~r-: 8
52 6
u+-
t;~ 4
~~ 2
001
'?O 0
~ x 15
~ ~ >.
o '~.u 10
LLEo.:
If),.,' I-' 5
o +- '
<.9......2 0
~ 
e 
,Q 10
If) 
,5Q 
E 5
we
U)'E 
~....... 
o:e 0
FIGURE C-4.
1 3 5 7 9 II 13 15 17 19 21 23252728
FEED '
END --'--WI ndbox Leg Number
DUST-FLOW DISTRIBUTION ALONG A
92-FOOT SINTERING MACHINE
Sinter -Plant Emission-Control Equipment. There are about 48 operating sintering
plants in the United States. This total is made up approximately of 38 single-strand
machines, 7 two-strand machines, 1 three-strand machine, 1 four-strand machine, and
1 six- strand machine. Emission-control equipment varies from plant to plant and is
determined by local regulations with respect to allowable emissions, and with individual
company experience with the various types of emission-control equipment. Mechanical
precipitators, single cyclones, multiple cyclones, and, in some cases, simple dust
catchers have been used in the past to collect particulate emis sions from sintering
machines. (14) However as the need for lower allowable particulate emissions became
apparent, more efficient equipment has been installed, such as venturi scrubbers(9),
bag house s( 15), multicyclone -electrostatic precipitator combinations( 16), and a system
incorporating multicyclones, mechanical collectors, bag house, and electrostatic
precipitators. (17) Details of some of these representative systems are given in
Table C-3.
An additional practice that is used to maintain low dust levels in the handling of
sinter and sinter fines is to use water to wet the materials at the various transfer points
in the sinter-plant materials-handling system.(16, 17) Special wetting agents and
equipment to handle them are used as pad of such dust-suppression installations. This
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-8
technique appears to work well in some cases in
transfer points, and also permits more effective
control equipment by lowering the load on it.
maintaining low dust levels at the
operation of associated emission-
 TABLE C-3. CHARACTERISTICS OF SOME SINTER-PLANT EMISSION -CONTROL SYSTEMS
      Dust Loading,
    Gas Volume, Inlet grains per cu. ft.
Company Plant Equipment cfm Temp, F Inlet Discharge
Bethlehem Bethlehem, Pa. Bag house 225,000 200-400 n.a. n.a.
Inland  E. Chicago, Ind. Muiticyclone and 457,000 375 2.5 0.038
   Electrostatic    
McLouth Trenton, Mich. Flooded -disk 100, 000 100-150 n.a. 0.01
   scrubber    
United States Steel McKeesport, Pa . Venturi n.a. n. a. n. a. n.a.
   Scrubber    
  Gary, Ind. MuIticycIone and n.a. n.a. n. a. n. a.
   electrostatic    
   Muiticyclones n. a. n.a. n. a. n.a.
   Bag house 172,000 175-300 n. a. n. a.
Note: n.a., not available.      
Incorporating flux materials with the sinter to produce high-basicity (often- called
self-fluxed and super-fluxed) sinters has created some additional problems in the control
of particulate emissions at sinter plants. With reference to electrostatic precipitator
performance when operating on high-basicity sinter, the following has been stated. (17)
"The secondary gas cleaner is an electrostatic dry precipitator between the
end of the main and the induced draft fan. The entire system does an adequate
job of cleaning the exhaust gases, with routine maintenance of electrodes,
rectifiers, rappers, and wear areas, to protect the fan and to prevent a
community problem. "
"Very little more needs be said about this system, unless it would be that the
higher the sinter basicity, the less efficient the precipitator. It appears that
with super-fluxed sinter, larger precipitators may be needed to maintain a
clean discharge stack. II
The underlining has been added by the present authors. Other attempts to use electro-
static precipitators on sinter plants used to make self-fluxing sinter resulted in such
lack of precipitator stability that its use was abandoned, and efforts were directed
toward the application of bag houses and wet scrubbers. From this, it appears that
any steel company wishing to take advantage of high-basicity sinter to increase the
productivity of its blast furnace s (to remain competitive in dome stic markets and
foreign competition in those markets) will find that manufacture of high-basicity sinter
will involve air-pollution-control costs higher than thoseassociated with the manufac-
ture of conventional sinter.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
Making Pellets
C-9
The use of iron-ore pellets in blast furnaces for the making of pig iron has become
a significant mode of operation in the American integrated iron and steel industry, The
increased use of pellets from 1958 through 1967 is shown in Figure C-l. Pellet-making
plants in the United States and Canada are listed in Tables C-4 and C-S. (18) Canadian
plants are listed because many United States steel companies have interests in these
plants, In contrast to sintering plants that usually are located near the blast-furnace
plant (because sinter does not withstand shipment without degrading), pellet plants
usually are located near the are mine (or within several hundred miles of the mine),
The pellets, which are strong, are often shipped hundreds or thousands of miles to blast-
furnac e plants,
TABLE C-4. IRON ORE PELLET PLANTS IN THE UNITED STATES
Company
Location
Annual Capacity,
gross tons
Bethlehem Steel Corporation

Cornwall
Grace
The Cleveland-Cliffs Iron Company
Em pire Iron Mining Co. .
Humboldt Mining Co.
Marquette Iron Mining Co.
Marquette Iron Mining Co.
Pioneer Pellet Plant
The Hanna Mining Company
Butler Taconite
Groveland
National Steel Pellet Co.
Pilot Knob Pellet Co.
Inland Steel Company
Jackson County Iron Co.
Kaiser Steel Corporation
Eagle Mountain
Meramec Mining Company
Pea Ridge
Oglebay Norton Company
Eveleth Taconite Company
Pickands Mather & Co.
Erie Mining Company
Reserve Mining Company
E. W. Davis Works.
United States Steel Corporation
Atlantic City are
Minntac
Cornwall, Pennsylvania
Morgantown, Pennsylvania
Palmer, Michigan
Humboldt, Michigan
Eagle Mills, Michigan
Republic, Michigan
Eagle Mills, Michigan
N ashw auk, Minnesota

Iron Mountain, Michigan
Keewatin, Minnesota

lronton, Missouri
Black River Falls, Wisconsin
Eagle Mountain, California
Sullivan, Missouri
Eveleth, Minnesota
Hoyt Lakes, Minnesota
Silver Bay, Minnesota
Atlantic City, Wyoming
Mountain Iron, Minnesota
Total United States Annual Capacity
700,000
1,500,000
3,200,000
800,000
800,000
2,000,000
1,200,000
2,000,000
2,100,000
2,400,000
1, 000, 000
750,000
2,000,000
2, 000, 000
1, 600, 000
10,300,000
10,700,000
1,500,000
4,500,000
51, 050, 000
BATTELLE MEMORIAL INSTITUTE -COLUMBUS LABORATORIES

-------
C-IO
TABLE C- 5. IRON -ORE PELLET PLANTS IN CANADA
Company
Location
Annual Capacity
gross tons
Bethlehem Steel Corporation
Marmoraton Mining Co., Ltd.
Marmora, Ontario
450,000
The Cleveland-Cliffs Iron Company
Sherman Mine
Timagami, Ontario
I, 000, 000
The Hanna Mining Company
Carol Pellet Company
National Steel Cor. of Canada
Labrador City, Newfoundland
Capreol, Ontario
10,000,000
625,000
Inland Steel Company
Caland Ore Co.
Atikokan, Ontario
I, 000, 000
The International Nickel Company
of Canada, Limited
Sudbury Mine
Copper Cliff, Ontario
1,170,000
Jones & Laughlin Steel Corporation
Adams Mine
Kirkland Lake, Ontario
1,250, 000
Pickands Mather & Co.
Hilton Mines, Ltd.
Wabush Mines
Griffith Mine
Shawville, Quebec
Point Noire, Quebec
Iron Bay, Ontario
900, 000
6,000,000
I, 500, 000
Steep Rock Iron Mines Limited
Steep Rock Mine
Steep Rock Lake, Ontario
Total Canadian Annual Capacity
1,350,000
25,245,000
Identification of Emissions. Materials received at pellet plants include iron-ore
concentrates and a binder material, usually bentonite. The concentrates are received
usually in a moist condition, and the generation of dust during receiving is not generally
considered a problem. Bentonite is received in covered hopper cars and unloaded -into
special bins for metering into the pelletizing operation. The pelletizing operation is
conducted with some moisture in the ore-binder mixture to achieve the necessary
pellet formation in rotating drums or on rotating disks. The final stage in pelletizing
of the concentrates is indurating (heat hardening) to produce the required hardness that
will permit subsequent handling of the pellets during shipment and charging to a blast
furnace.
The concentrates as delivered to the pellet plants are magnetite, except for two
plants that are operating with hematite concentrates. During the indurating treatment
of the magnetite pellets, the magnetite is transformed (by oxidation) to hematite. When
particulate emi s sions are generated to any extent in the production of pellets, the
particulates are magnetite, hematite, or bentonite. The se particulates have the follow-
ing characteristics:
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-11
Magnetite Dust. All partiCles of this isometric mineral are opaque,
even those with dimensions less than 1 micron. The crystals fracture
unevenly and rarely show cleavage. The crushed fragments are attracted
to one another and cohere in a fluid mount because they are magnetized. (11)
Chemistry: mainly Fe304, with some gangue, mostly silica.
Hematite Dust. Particles are rounded to elongated and can be as small as
2 microns. Larger particles are opaque and red-orange in top light. Individual
small grains are transparent and blood-red. (11)
Hardnes s :
Specific gravity:
Chemistry:
5 (Mohs)
5.2
Mostly Fe203,
mostly silica.
Soluble in HCl.
Minor amount of gangue,
Bentonite. Transparent, colorless, apparently rounded particles that can
occur as small agglomerates. Because the particles are so small, their
agglomerates look almost fibrous. (11) Bentonite is a fine-grained clay,
usually of the montmorillonite type.
Pellet-Plant Emission-Control Equipment. The minor amounts of dust generated
in pelletizing plants are usually handled by simple cyclones. The indurating operations
are conducted under rather low air-flow rates. High air-flow rates, such as encountered
in sinter plants or blast furnaces, are not generally present, thus in-process formation
of particulate emissions is usually not substantial.
Preparation of Limestone,
Dolomite, and Lime
The use of limestone (mostly CaC03), dolomite (mostly CaC03' MgC03)' and
lime (mostly CaO) in the making of iron and steel is necessary to remove undesirable
elements from the iron ores (mainly silica, alumina, and sulfur, in the case of the blast
furnace) and from pig iron (mainly sulfur, phosphorus, and silicon in the case of steel-
making). The amounts of limestone ':' and lime consumed in the American integrated iron
and steel industry are shown in Figure C-5. (8) The rapid increase in lime consumption
from 1962 to date has been due to its use in BOF steelmaking. Limestone consumption
is influenced by the total production of pig iron as well as by improved blast-furnace
technology, which has resulted in a decrease in the amount of limestone consumed in
the production of each ton of pig. The trend in limestone consumption per net ton of pig
iron from 1958 through 1967 is shown in Figure C-6. (8)
In BOF steelmaking, lime is used instead of limestone because of the faster reac-
tion rates obtainable with the lime, and because the use of lime conserves heat in the
..... OF furnace. It has been determined that the characteristics of the lime itself have an
important effect on the reaction rates. (19,20) A soft lime, which is calcined at a lower
temperature and for a shorter time than hard lime, is the preferred material. Both
types have the same chemical composition, but the soft lime is less dense. (20)
Increasing demands for lime for BOF steelmaking have necessitated an increase in
calcining facilities.

*In the U. S. steel industry, statistics on "limestone" include dolomite.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
If) 
c 30
o
+- 
+- 
Q) 
C 25
-
o 
If) 
c 
.2 20
- 5
E
c 4
o
+- 
a. 
E 3
:J 
If) 
c 2
o
u 
0 
:J 
C 
c 

-------
C-13
Estimates have been made of the lime production that is captive to the integrated
iron and steel industry. (21) These estimates since 1965 are given in Table C-6. It
also has been estimated that by 1970 the captive lime production will be less than
20 percent of the 5.6 million net tons required annually by the steel industry.
TABLE C-6.
ESTIMATED LIME PRODUCTION
CAPTIVE TO THE INTEGRATED
IRON AND STEEL INDUSTRY IN
THE UNITED STATES
Year.
Net Tons
1965
1966
1967
1968
1970
158, 000
358, 000
600, 000
860,000
1,000,000
Identification of Emissions. Compositions of particulate emissions that may be
generated during the preparation of lime stone, dolomite, and lime are characteristic of
the minerals. Compositions vary from about 95 percent calcium carbonate and less
than 1 percent magnesium carbonate in high-quality limestones, to about 55 percent
calcium carbonate and 40 percent magnesium carbonate in dolomites. Dusts produced
during crushing operations are the main particulates generated during the preparation
of limestone and dolomites. However, little is actually emitted to the atmosphere.
Primary crushing is usually done with the minerals dry or at their natural moisture
content, and during this stage of preparation the generation of air -borne particulates is
small. In the stages of finer crushing, where dusts could be a problem, processing
usually is carried out wet. Quoting from a recent report on a new plant for the prepara-
tion of flux and BOF lime, the following is descriptive of these operations. (22)
11 Bin No. 2 provides 1, 100 tons of minus 4-inch feed to the second
principal proces sing section. This' section consists of four parallel identical
circuits, each having a 48-inch x 72-inch vibrating feeder, a single-deck
5-foot x 14-foot screen, an 8-foot x 20-foot rotary scrubber; two 5-foot x
14 -foot double-deck screens and a 460 Hydrocone tertiary crusher. 11
11 The feeders draw stone from the bin and distribute it to the single-
deck screens. This circuit is entirely wet processing, with all screens
equipped with spray bars. . ."
(Underlining has been added for the purposes of this present report. )
Emissions from lime kilns can include lime dust generated at the discharge end of
the kilns, carbon dioxide from the calcining of the limestone, and the products of com-
bustion from the fuels used to heat the kiln. Fuels used are usually IQ.w-sulfur fuel oils
and natural gas. Because lime is used to accolnplish desulfurizatton during the steel-
making process, it would be to the disadvantage of lime producers to use fuels that
would result in increased sulfur content in the lime.
Emissions from limestone preparation and calcination of limestone can be
identified as follows:
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-14
( 1)
Limestone Dust - Mineral name calcite. Colorless, with light-transmitting
characteristics varying from transparent to translucent. Particles
generally occur as rhombohedra because of their perfect rhombohedral
cleavage. Fragments may also occur as prisms. (11)
Hardness:
Specific gravity:
Chemistry:
3 (Mohs)
2. 7
Mostly CaC03
(2)
Dolomite Dust - Mineral name dolomite. In the pure state, dolomites
are colorless and the fine particles are generally transparent to translucent.
Electron-microscope studies of finely ground dolomite have shown the
presence of few good cleavage planes and an absence of perfect rhombs. (23)
Specific gravity: 2.85
Chemistry: Mainly CaC03' MgC03 in various combinations.
(3)
Lime Dust - Lime is usually white in color of varying intensities, but
some may have a light cream, buff, or gray cast depending on the
nature of the impurities in the lime. (24) . '
Specific gravity: 3.3
Chemistry: Mainly CaO
Lime-Preparation Emission-Control Equipment. As stated earlier in this Appendix,
the generation of air -borne emissions during the preparation of limestone and dolomite
is not a substantial problem because much of the processing is carried out wet. This is
not the case in the making of lime. .
Several types of equipment are used to collect dust from the discharge and cooling
locations of calciners. A scrubbing tower is reported to remove all dust and smoke
from a commercial lime plant operating with a rotary kiln. (25) Cyclones are used with
some circular-hearth calciners(26), and electrostatic precipitators are included in a
new lime plant placed into ope,ration by the Republic Steel Corporation in 1968. (27) A
recent report on emission control equipment placed into operation in 1968 at the Pueblo,
Colorado, plant of CF & I Steel Corporation, describes the use of multicyclones and a
bag house to recover dust from a vertical lime kiln. (28) No information has been found
yet in the published literature pertaining to the amounts or sizes of particulates that
may escape to the atmosphere from lime-burning equipment.
Making Coke
Metallurgical coke is the major fuel and reducing agent used in the production of
blast-furnace hot metal, and will probably be the major fuel and reductant for many
years in the future. Technological developments in the making of hot metal have
resulted in a decrease in the amount of coke ne'eded to make 1 net ton of hot metal.
(The amount of coke required to make 1 net ton of hot metal is referred to as the "coke
rate".) Even though the coke rate has decreased, the'total consumption of coke has
increased because of the increased production of hot metal.'. The trends in coke rate
and in total coke consumption in American blast furnaces from 1958 through 1967 are
shown in Figure C-7. (8) Much of the lowering of coke rates has been the result of

BATTELLE MEMORIAL INSTITUTE -COLUMBUS LABORATORIES

-------
C-15
improving the composition and physical form of the iron ore. The decrease in coke rate
that is possible with sinter and pellets in the blast-furnace burden is illustrated in
Figure C-8. (29) It is possible to reduce the coke rate to about 840 pounds (ideally) per
net ton of hot metal by the use of improved burden materials, injection of auxiliary fuels
(such as oil and natural gas), and the use of very high blast temperatures. (30, 31)
1600
Q)
..:JC.
o
<.)
-
o
1550
1500
IJ)
1:J
C
:J
o
Q.C
~O
Q)'~ 1400
+-en
0.-
0::Q.
Q) 15 1350
..:JC.
o c
U.8 1300
Q)+-
enQ)
e c 1250
Q)...
>Q)
<[Q.
1450
/1\
/ \
/ \
\
/-""\
/ \
I \
I--
I Total coke
I consumed
annually
I
I
/
/
62 U>
a
60 CD
c
58 ~
OIJ)
:J C
co
56 C +-
<[Ci)
1:JC
54 Q)-
EO
:JIJ)
52 g? 5
0=
u.-
50 Q)E
..:JC. ~
o~
48 U g
oc
+- ...
~~
1200 46
1958 1959 19601961 1962196319641965 196619671968
Year
TRENDS IN THE CONSUMPTION OF COKE AND IN COKE RATE
IN THE PRODUCTION OF HOT METAL IN AMERICAN BLAST
FURNACES
FIGURE C-7.
/700
...
Q)
Q.
IJ)
1:J
C
:J
o
Q.
~O
~ Q) 1500
°E
0::
O)+-
C5 2 1400
U-
Q)0
enC
e.8 1300
Q)+-
>Q)
<[C
1600
1200
o
All pellets
1000 2000 3000
Sinter and Pellets Charged to Blast Furnace,
pounds per net ton of hot metal
FIGURE C-8.
EFFECT OF AMOUNT OF SINTER AND PELLETS IN
BLAST FURNACE BURDEN ON COKE RATE
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LApORATORIES

-------
C-16
The manufacture of coke as discussed in this Appendix is concerned with by-product
coke ovens. Beehive coke is still made to a limited extent. However, during the last
10 years, production of beehive coke has not exceeded 1 percent of the total coke pro-
duction in the American integrated iron and steel industry. From 1965 through 1967,
this percentage has decreased to O. 7 percent. A beehive coke oven is difficult to adapt
to air -pollution control, and in all probability the industry will phase the beehive coke
ovens out of operation as rapidly as the required additional capacity can be made
a vailable through the construction of by-product coke ovens.
By-product coking is a process combining agglomeration and thermochemical
conversion. Bituminous coals that are too finely sized and too friable for direct use as
the major blast-furnace fuel are blended and baked at high temperature. The baking
process decomposes the long-chain organic polymers of the coal. Aromatic tars and
oils are vaporized and driven off along with various free radicals, and the carbon resi-
due reforms in massive chunks of strong coke suitable for use in smelting. Fuel gas,
tar, aromatic oils, and ammonia may be recovered as by-products from the crude gas.
Conventional coke plants emit both particulate materials and offensive gases in
the normal course of operation. (32) The particulates are mainly coal dust and coke
dust that become airborne during handling. The gases are mainly mixtures of ammonia
and aromatic vapors escaping from the ovens and from sumps, seals, and vents in the
by-product system. Because these emissions are considered to be particularly unplea-
sant, coke plants deserve very serious attention in planning for environmental control.
Identification of Emis sions. Emis sions are generated in several locations and
operations during the manufacture of coke and during the processing of coke by-products.
These are discussed under the appropriate headings as follows: (1) coal handling; (2) oven
charging; (3) oven operation, pushing and quenching; (4) coke handling; and (5) by-product
processing. The points of emission are designated in the flow sheet for a by-product
coke plant shown in Figure C-9. The circled numbers on the flow sheet are indexes to
similar numbers in the following text where emissions are discussed.
Coal Handling. Coking coals are received usually by hopper car, river barge,
or lake freighter. Because these coals are highly beneficiated at the mines, they
average a much finer size than steam coals. Shipment of high-grade coking coals at a
top size of 1/2 inch is a common practice. For this reason, coking coals are dusty
(unless wet), and each transfer point in the unloading and handling steps is a potential
emission site.
Although coking coals are often wetted before shipment, they are seldom wetted
upon receipt. The problem is that fine coal is hard to wet uniformly, and nonuniform
moisture interferes with both handling and blending. Overwetted coal will stick to
belts and foul the transfer points, while underwetted coal continues to dust. Blending is
based on dry weights and presumes uniformity of moisture in each individual coal used.
Fuel oil is sometimes sprayed on coal just prior to pulverizing or blending. The purpose
of the oil is to control bulk density of the blend - it also controls dusting.
Typical coal-handling operations include unloading, transfer to and from stock,
pulverizing, blending, and transfer to the coke ovens. Most handling is with rubber
belts for upward movement and by gravity chute for downward movement. There are
BATTELLE" MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-17
Cleaned Coal
3100 pounds
Air
@@@
(OO)@@

COKE OVENS
3085 ounds
Hot Coke
2220 pounds
@
Other plant use
Water +
Spent Liquor
FIGURE C-9. TYPICAL FLOW SHEET FOR A BY -PRODUCT COKE PLANT
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-18
many transfer points in most conveyor systems, and a typical particle may free-fall
as many as twenty to thirty times before it arrives at the coke ovens. Each occasion of
free-fall provides opportunity for fines to escape as dust. However, coking-coal trans-
fer points can be controlled by utilizing water-spray systems, plus detergent. A re-
cent report on an investigation of settled particles at a coke oven plant contains the
following statement(33): -
"On the basis of the data obtained during a 6-month study near a coke-making
operation it is concluded that material handling and stockpiling operations are
major contributors to settled particulate deposition, while coke oven charging
was not a major source. "
The extent of the coal-storage problem can be visualized by realizing that as much
as 3 million tons of coal is stored at times in one storage area. To some degree, the
problem of fugitive dusting can be minimized by limiting storage of coal to smaller
quantities. Deliveries of coal by unit trains directly from mine to coke-oven area lowers
the amount of coal that must be stored within the steel plant.
Oven Charging. The last step in handling of the coal is charging of the blended
coal fines (now pulverized to a top size of 1/8 inch) through ports in the top of the coke
oven. The oven itself is a slot-like retort about 10 to 18 feet in height, 30 to 60 feet in
length, and 15 to 20 inches in internal width. The sidewalls of the oven are at an
incandescent temperature at the time of charging. There are usually 4 or 5 coal-charging
ports, each about 10 to 14 inches in diameter. The charging vehicle is called a larry
car. It receives a weighed charge into separate hoppers (one for each port) and dis-
charges from the hopper bottoms.
The oven slot is vented to the by-product system at one or both ends via vertical
ascension pipes leading to a manifold called the collector main. Coke ovens are oper-
ated under slight negative draft in this main, but the oven pre ssure is quickly equalized
when the charging ports are opened. Modern plants have steam jets in the raw-gas
ascension pipes so that the draft on the opened oven can be increased during charging.
Even so, the charging operation is characterized by particulate and gaseous emissions.(34)
When the cold, damp coal falls against the hot walls of the coke oven, the moisture
is quickly changed to steam, and the breakdown of polymeric materials in the heated
coal proceeds very rapidly. The volume of the mixture of steam and raw-coal gases far
exceeds the capacity of the aspirated ascension pipe; and so it rushes out of the charging
ports as smoke. The escaping coal gas is usually ignited by incandescent coal fines,
which are also blown out of the oven by the reaction. Of course, this rush of gases
interferes with charging, and some fines are blown out of the descending coal. At the
end of charging, the steam, flame, and smoke puff upward through the ports of the larry-
car hoppers. .
A new development is being used in Europe to minin.ize emis sions during oven
charging. The larry car mechanically remeve-s the oven lid, drops coal through a
sleeve into the oven, collects the emissions produced during .charging, and treats them
in a wet scrubber. Adaptation of this system to existing American ovens presents major
problems because the oven structure usually would not necessarily be strong enough to
support the additional weight. Also, the larry car usually would be so high in the new
system that the car would not fit under existing coal bunkers. The new system, if it
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-19
works out consistently in practice has potential, for use on new American batteries.
There is some hope that work being done in the United States on pneumatic charging
may be succes sful enough to warrant inclusion in future American practice.
Oven Operation, Pushing, and Quenching. Coke ovens are heated by complex flue
systems between the slot retorts. The fuel is usually coke -oven gas or blast-furnace
gas, depending on the energy balance within the steel plant. Combustion in coke-oven
flues is accomplished with an excess of preheated air in such a way as to preclude the
presence of unburned fuel in the products of combustion. Accordingly, the emissions
from the coke-oven exhaust stacks are quite ordinary, unless coal gas from the retorts
leaks into the flues. This sometimes occurs in older coke plants.
After an oven is charged, the release of raw-coal gases continues at a high rate
for a considerable period of time, at least until the entire charge has been dehydrated
and steam emission stops. During this time, the oven is under negative pressure, but
localized pressures are generated within the coking bed and raw gas may leak out
around the ports and the end doors. This emission is normally a mere wisp when com-
pared to the gas released during charging, and no direct attempts have been made to
suppress or collect it. It is to the best interests of the operator to minimize door leak-
. age, because the cracks that allow gas to leak out in the first part of the cycle will allow
air to leak in (to consume coke) in the later stages. In any particular coke-oven bat-
tery, the amount of leakage is generally related to the age of the ovens, the level of
maintenance applied, and the skill and motivation of the operating crew. The idea of
collecting oven emissions by building an enclosure over the whole battery represents a
prodigous undertaking, but has the potential for great improvement of emission control.
During the actual coking operation (which typically requires 16 hours), usually
little gas or particulate matter is evolved into the atmosphere. Some escapes because
of ineffective sealing of the charging ports and end doors. When an oven is pushed
(emptied) at the completion of the coking cycle, the end doors are removed and a
mechanical pushrod is forced all the way through the oven to eject the coke. The open-
air release of 10 to 15 tons of incandescent coke produces a considerable induced draft
in the immediate surroundings, and fine particulates are blown high into the atmosphere.
However, at this point. the coke is both strong and massive, and only two kinds of fines
are likely to be present. Some of the air-borne particulates may be those that have
settled in the area because of prior coal handling, but the rest comes from incompletely
coked coal (adjacent to the cool coke -~ven doors) and from abrasion of the coke as it is
pushed from the oven. Particulate emissions are not great during the pushing operation
unless the amount of uncoked coal is considerable, under which circumstances coal
smoke may also be released for a brief period. With proper coking cycles and ove1I1-
heating practices, emissions generated from pushing "green coke" (coke with incom-
pletely coked coal) can be minimal.
The pushed coke is received into an open hopper car of special design, with a
sloped bottom and side gates made of grating. This car may be self-propelled or moved
by locomotive to a large brick chimney that fits over the open top of the car. Sprays in
the chimney deluge the hot coke with water to cool and quench it. During the quenching
operation, particulates are lifted into the atmosphere by the chimney effect created by
the rising cloud of steam. The particulates tend to fall out locally in the vicinity of the
quench tower, and usually are not carried great distances as a suspended dust plume.
The quenching time is about 2 minutes (depending on the practice in any particular plant),
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-20
and 87 percent of the particulate emission from quenching occurs during the first min-
ute of the quenching cycle, as shown in Figure C.,...lO. (35)
100
'+-
o
-c
Co
B 0 in 80
... I/)
If °E
_w' 60
'-g,~
°w.Q
3: B 40
Q) '';::
> ...
'+= 0
o a.. ° 20
:J-
E.E
8~
o
o
20 '40 60 80 100
Quenching Time, seconds
120
FIGURE C-lO °
PARTICULATE EMISSION DURING COKE QUENCHING
Coke Handling. The quenched coke is dumped from the quench car onto a sloping
brick wharf from which it is carried on conveyors to sizing and screening operations.
Dusting is not so much of a problem in the handling of coke as it is with coal handling.
Usually only a small amount of dust is generated in the handling operations. Transfer
points, screens, and crushing machinery are protected by steel and rubber shrouding to
minimize dust losses and reduce housekeeping maintenance. Also, the handling equip-
ment often is enclosed by galleries and buildings. Aside from these factors, the coke
itself is a massive and stable material, and fines generated during crushing and handling
are removed before the mainbody of coke is transferred to stock or to the blast furnace.
The fines usually are dumped into open hopper cars via chutes from the screening
plant. The fines have a higher moisture content than the massive, coke and typically
contain over 8 percent moisture. Dusting of this "breeze" coke usually is not a serious
problem except in windy weather. The small amount of volatile organic material
remaining in the coke is stable and is known to persist to temperatures of 2300 F or
higher within the blast furnace. Therefore, it is not emitted as an air contaminant
during the storage of coke.
By":Product Processing. The mixed gases driven off of the coking coals during
conversion to coke enter the by-product system via the ascension 'pipes at one or both
ends of each retort oven. These ducts lead into a collector main (or two) and are
maintained under negative draft by exhauster fans far downstream in the system.
The mixed gases contain organic (usually aromatic) compounds released by the
destruction of long-chain polymers in the coal,plus carbon monoxide, hydrogen,
methane, hydrogen sulfide, ammonia, and nitrogen from tramp air drawn into and
through the ovens. Among the prominent organic groups one may number anthracene
and other tarry compounds, benzene, toluene, xylene, naphthalene, phenols, and pitch.
The purpose of by-product processing is to condense, separate, absorb, distill, clarify,
and otherwise segment and recover the valuable substances in the raw gas. With the
BATTELLE MEMORIAL INSTI.TUTE - COLUMBUS LABORATORIES

-------
C-21
advent of high-volume petrochemical processing in the 1940's, the market prices of
many organic chemicals were lowered drastically, and by-product recovery from coke-
oven gas is not now,so economically attractive to steel companies as in the past,
especially in the case of small-capacity coke-oven plants. However, most American
coke plants have by-product systems, and these are still used to obtain credits to the
coking process. A flow sheet for a typical by-product plant is given in Figure C-ll.
Raw-coal gas (which includes steam) rises into the coke-oven ascension pipes and
passes into the collector main. Some ovens have mains and ascension pipes at both
ends of the ovens. The beginning of cooling and condensation is forced almo st
immediately by sprays of recycled water in the ascension pipes and along the mains, but
the final temperature at the end of the main is still well above the boiling point of water.
The mechanism of cooling is the evaporation of the sprayed water (called flushing
liquor); thus the raw gas is consider"ably diluted by steam. Shortly after emerging from
the collector mains, the gas temperature drops below 212 F and water condense s
rapidly from it.
With the condensation of water, droplets of tar are also condensed. This process
is hastened in primary coolers cooled normally with river water, and the mixed
tar and condensed liquor are separated in decanters. Some of the liquor is recycled to
the flushing-cooling system, and some is drawn off for recovery of ammonia values by
distillation and absorption in sulfuric acid. The partly cleaned gas goes through the
exhauster fans into the rest of the by-product system.
Emissions from the primary end of the system usually are minor in amount be-
cause the system is under negative pressure. There is some odor indicative of free
vapor at the tar collectors and decanters and wherever the liquor runs in lines that are
not fully enclosed. In particular, ammonia and organic fumes are strong at the sumps
where decanted liquor and other flush liquor is collected for recycling to the collector-
mains sprays., It is arguable that the flush liquor should be handled in closed, unvented
ducts. To minimize local nuisance, the flushing liquor sewers are usually fairly well
capped and covered, but there often is no sealing because these ducts become fouled
with tar and other "goop". Access for steam cleaning is essential, but is not always
used frequently.
Downstream of the exhausters, the detarred gas is still quite rich in ammonia,
both free and in combination. The liquor decanted from the tar at primary cooling is
also quite rich, and because of the moisture vaporized from wet coal this liquor exceeds
the requirements of the flushing-liquor system. Both the excess liquor and the gas are
stripped of ammonia, which is recovered usually as ammonium sulfate. (America.! s
largest coke plant is big enough to recover the ammonia on a commercial basis as
anhydrous ammonia), In the conventional sulfate process, the gas is reheated (to hold its
moisture in the vapor phase) and passed through sulfuric acid where ammonium sulfate
is precipitated. Ammonia vapor distilled from the surplus flushing liquor is also passed
through this precipitator. The residue after distillation, known as weak liquor, may be
processed further or disposed of via sewers or to the coke quench. The residue still
contains ammonia plus a number of soluble organics such as phenols.
Emission to the air from the ammonia system generally is quite minor, because
the ducts are closed piping for the most part, and leaks usually are promptly detected
and repaired. One activity that can cause trouble in the older plants is the addition of
strong sulfuric acid to the precipitator tank, an operation usually accompanied by
considerable fuming of the acid.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-22
COKE OVENS
Ammonia
Further Processing.
SCRUBBER.
Steel Plant Use
FIGURE C -11. BY -PRODUCT PLANT FLOW SHEET
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
G-23
In large coke plants, the ammonia-recovery system may be augmented by systems
for recovery of phenol (carbolic acid) from the weak liquor after ammonia distillation
and for recovery of pyridine bases which dissolve in the ammonium sulfate precipitator
tank from the gas. In smaller coke plants, one or both of these activities may be by-
passed where volume does not justify capital investment for a recovery system. Phenol
and pyridine systems are closed except for tank vents, and usually have no substantial
problems with regard to emission to the atmosphere.
The crude tar collected in the collector mains and in the tar decanters is aug-
mented by minor amounts of tar precipitating at the secondary coolers or other points in
the gas system. From this crude tar, pitch sludge settles to the bottom of the decanter
and is mechanically raked out for disposal. Settled crude tar is sent to a separate plant
for secondary processing by distillation.
Conventional tar proces sing yields pitch tar, creosote, and two or more weights
of tar oil that may be used in roadbuilding. Crystals of naphthalene are a by-product
(often discarded), and in larger plants the tar oils may be further refined to yield
salable fractions.
Tars are heavily loaded with polycyclic aromatic hydrocarbons (PAH), and are
generally considered hazardous. Steelmakers have recently taken steps to lead tar-
processing tank vents and storage tank vents through scrubbers that absorb or destroy
the fumes. The work has not met with great success yet because tars tend to condense
upon and foul the equipment. When the scrubbers get fouled, they will discharge
unscrubbed vent vapor.
Gas leaving the ammonia precipitation tank is still warm and contains a number of
light aromatics boiling between 200 and 400 F. These light oils are condensed from
vapor by cooling the gas to about ambient temperature in a final cooler. The condensate
is then scrubbed with a high-boiling wash oil (derived from petroleum) to dissolve the
aromatics, and the wash oil may be steam distilled to recover the principal aromatics
benzene, toluene, and xylene in commercial form. The gas is now completely stripped
and enters the plant fuel system.
The vapors of the light oils are considered hazardous both because of toxicity and
because of flammability. Nevertheless, the condensers in the distilling system and
some of the process tanks are vented. The vapors issuing from these vents often pervade
a large area with the sweet, almost pleasant smell characteristic of aromatic vapors.
Abnormalities (such as fire or major leakage) are rare in by-product systems
because the high hazard level prompts strong preventative measures. However, in the
ordinary course of pumping, straining, dewatering, and otherwise treating coke -oven by-
products, leakage and vapor loss is inevitable. There may be as many as 40 pumps in
the liquor and stitl systems; usually some of these have imperfect seals. Forced
ventilation in the pumphouse is a necessity.
Identification of particulate emissions from coke-plant operations has been limited
in the past, but as a result of research carried out by the integrated iron and steel
industry since 1950(36-43), methods have been developed that permit the identification
of particulate emissions generated in coke-plant operation. While this research has
been directed primarily toward obtaining a better understanding of the coking proces s,
and toward improvements in the properties of coke required by the advancing blast-
furnace technology, it has also provided a means of identifying air-borne particulates.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-24
Optical reflectance has been found to be the best method for identifying different types
of coal and their particulates. Figures C-l2 and C-13 illustrate some relationships
between reflectance, ultimate carbon in the coal, and volatile matter in the coal(36)j the
latter two materials are an indication of the "rank" or classification of the coals.
100
.
.. .
-
c:
~ 40
~
Q)
a.
..: 30
Q)
::.
~ 20
-
c:
~ 90
~
Q)
a.

c: 80
° .
.Q
~
8 70
~~.~ .

,
./
-
I
.
Q)
-
o
.S 60
-
::J
50
~
-
.2
g
50
o
1.0 2.0 3.0
Reflectance (Ro)' percent
10
o
o
1.0 2.0 3.0
Reflectance (Ro)' percent
FIGURE C-13.
RELA TION OF
REFLECTANCE TO
ULTIMATE CARBON
FIGURE C-l2..
RELA TION OF
REFLECTANCE TO
VOLA TILE MATTER
The research on reflectance properties of coals was extended to coke, and a relation-
ship was established between the reflectance of coal and the reflectance of the coke wall
(referring to the cell structure of coke) in the coke made from,that coal. This relation-
ship is illustrated in Figure C-14. (40) For such relationship to be valid, the final
coking temperature must be known.
c: 1.5
Q)
u
Cii 1.4
a.
rr.0 1.3
Q)
u
c:
o
-
u
~
.....
Q)
rr.
o
o
u 0.9
0.8
0.7
6
1.7
14
FIGURE C-l4. RELATION OF COAL REFLECTANCE TO
COKE-WALL REFLECTANCE
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
1.6
1.2
1.1
1.0
7 8 9 10 11 12 13
Coke Reflectance,Ro' percent

-------
C"725
The techniques developed extensively for the petrographic examination of coal
and coal have been applied to the identification of particulate emissions in a coke
plant. (33) Particulate emissions generated in a by-product coke plant are identified
as follows:':'
CD
Coal dust - Bituminous, or soft coal, is translucent in thin areas; it is
reddish-brown by transmitted light, and brownish-black with dull to mod-
erately high reflectivity in reflected light. The surfaces are slightly rough
with occasional indications of the original fibrous structure. These irreg-
ular chips have sharp edges, and in places show conchoidal surface
fractures. (11)
@
Coke balls - Identified as oval in shape with an unusual network-like internal
structure. It is suggested that coke balls are produced during the thermal-
drying stages of coal proces sing and are inherent in products leaving coal-
processing plants. Similar conditions occur during charging of by-product
coke ovens, where some coal fines are carried through the hot zone and
out adjacent, open charging holes. (33)
Q)
Char - Partially devolatilized coal particles that exhibit optical properties
between those of coal and coke. The partial devolatilization of coal particles
suggests that they have not been subjected to temperatures high enough
or for periods long enough to complete the coking process. (33)
o
Pyrolytic carbon - The tarry residue from the volatile organic portion of
coal. Two forms of pyrolytic carbon are identified. The first is an
aggregate of minute oval grains; each grain is relatively uniform in size,
extremely smooth in appearance, and exhibits extreme anisotropy in
polarized light. The second normally occurs as a crenulated band of
varying width and length, smooth in appearance, and strongly anisotropic
in polarized light. The size of these materials is extremely variable. (33)
@
By-product coke - The optical characteristics of particles of by-product
coke are controlled by the rank (reflectance) of the coal which is
carbonized (as illustrated in Figure C-14). Because coals of different
rank are usually blended to make an optimum mix, particles of coke
produced from these mixes have complex and highly variable optical
properties. Particulates of coke made from high-volatile and medium-
volatile coals may be granular in appearance, have thick coke walls, and
have few internal pores. Particulates from coke made with low-volatile
coals have distinctive ribbon-like graphitic textures, have thin coke walls,
and comparatively large internal pores. (33)
Results of a 6-month study have shown that the origin of particulates in a coke
plant can be divided on a weight basis as follows: 40 percent from coke, 30 percent
from coal, and 30 percent f~0m other sources, where other sources are classified as
road dust and other mineral dusts normally found in a steel plant. Screen analyses of
the particulates, discus sed in the study, found in the vicinity of one coke plant are
listed in Table C-7. (33) Screen analyses of particulates generated at one coke-plant
quench tower are given in Table C -8. (35) The average weight of particulate
emission during a 2-minute quench cycle was calculated to be about 6 pounds.
.Circled numbers in this text refer to points of emission bearing the same circled numbers in Figure C- 9.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
     C-26    
  TABLE C.7. SCREEN ANALYSES OF COKE-PLANT PARTICULATES, WEIGHT PERCENT 
Screen Size, Coal Coke  Pyrol ytic High-Temperature Mineral Fly 
microns Dust Balls Char Carbon Coke Matter Ash Total
 -74 17.4 1.2 1.8 2.0 2.8 13.8 5.1 44.1
+74 -250 11.8 1.3 4.9 3.4 6.4 5.9 0.5 34.2
+250 1.1 0.5 7.6 2.2 6.4 3.7 0.2 21. 7
TOTAL 30.3 3.0 14.3 7.6 15.6 23.4 5.8 100.0
TABLE C-8. SCREEN ANALYSIS OF QUENCH-TOWER PARTICULATES
Screen Size

Mesh Microns
Weight Percent
Retained Cumulative
6
16
30
50
100
200
-200
3327
1167
589
298
147
74
-74
o
1
9
35
39
13
3
o
1
10
45
84
97
100
Coke-Plant Emission Control. Control of air-polluting emissions generated in a
coke plant is difficult because of the nature of the process and the great amount of
material handling that is required in the making of coke.
Coke-plant operators seek to control dusting at transfer points because the coal
losses are costly and because the dust is a serious fire hazard. To this end, the
handling operations are usually well shrouded with steel plate and are conducted (mainly)
inside steel buildings that are swept or flushed regularly. The 'Ibug dust" (as it is .
called) can ignite spontaneously; spread of fire to a dusty area can produce a violent
explosion. These factors promote rigid house-keeping standards, at least within the
buildings. Dust escaping to the surrounding grounds is seldom reclaimed.
Coal-handling operations are not all shrouded, and most of the particulate
emission of a coke plant arises from the unguarded unloading, stocking, and reclaiming
of fine coal during dry, breezy weather. If the coal is unloaded or stocked by conveyor,
the discharge is onto an open pile with high exposure. If'the coal is recovered with a
gantry crane, the clam-shell bucket must dump into a rail car or receiving hopper from
an appreciable height because of clearances for the equipment. The only known control
over systems of this kind is careful operation and suspension of work in windy weather.
Newer plants avoid these kinds of transfer; either by underground reclaiming (as
in Japan) or by use of carriers similar to earth.:.moving equipment. Special tractor
carriers can carry coal out onto a low-profile pile and dump from a bottom slot with'
minimum free-fall and dust. During reclaiming, the slot opens as a scoop and no
free-fall is involved in the open-stock area. The coal storage space required is much
larger, however, and thus many steel companies resist'this practice.
The sealing of end doors on most postwar coke ovens is mechanical, and is
achieved by applying closure pressure between a thin stainless steel strip on the door

BATTELLE MEMORIAL INSTIT!JTE'- COLUMBUS LABORATORIES

-------
C-27
and a matching flat plate on the oven end wall. Older ovens may have doors sealed
("luted") with fire clay, but these require more labor. Either type of seal can be only
so good as the amount of maintenance attention given to it. Accretions on, or damage
to, the sealing surfaces must be corrected before leakage will stop.
The cast-iron lids on the charging ports seal under their own weight if the mating
surfaces on the ports are swept clean before the lids are replaced. As with doors, the
effectiveness of the seal is directly related to the attentiveness of the operator. Most
operators keep the top of the battery well-swept at all times.
The rush of gases generated during the charging of coal into a by-product oven
blows fines out of the coal descending from the larry car. At the completion of charging,
steam, flame, and smoke puff upward through the ports of the larry-car hoppers. A
tight fit between larry-car discharge and oven ports has been considered as a solution
to this problem, but because of the rush of gas, coal then would be blown out of the
larry-car hoppers themselves, and thus create an even more serious particulate-
emission problem. Close tolerances between equipment and oven are almost an
impos sibility because the larry car rides on widely separated rails, and even flexes
upward as coal is discharged. If aspiration on the standpipe were increased enough to
prevent the flow of gases out of the charging port, air would then be sucked into the
nearest open port and the coke oven could explode during charging.
A current promising approach to the control of charging emissions is with suction
applied on a shroud pipe arranged around the charging port and fitted to the rim of the
charging port. The exhausted air is passed through a disintegrator located on the
larry car, and flared to the atmosphere, hopefully yielding only carbon dioxide and
water vapor as final emission. Equipment of this type has been installed in France and
Canada.
There are severe problems even with the shrouding and scrubbing of gases
exhausted from coke ovens by this method. If the mixed coal and shroud air ignite
ahead of the scrubber, there may be a violent explosion because of the restricted space.
Such explosions have occurred with the installations in Canada. Fresh water for the
scrubber system must be taken on each time the larry car receives a charge of coal, and
the dirty water must be discharged. The volume of water required for each charging
cycle is about 500 to 1000 gallons per charge of coal, and the dirty water requires
elaborate waste treatment.
It has been found possible to reduce somewhat the amount of particulates generated
during the quenching operation by the installation of baffles in the quench tower. (35)
Particulate emissions to the atmosphere before the installation of the baffles in one
case was about 6 pounds per load of coke quenched. After installation of the baffles in
the quench tower, the emission of particulates was reduced to about 3/4 pound per
quench, or a reduction of over 1000 pounds per day of particulates to the atmosphere
from one quench tower.
Another problem may be created in the selection of the water used to quench coke.
Many coke-plant operators make use of phenol-laden and ammonia-laden waste liquors
mixed with the quench water as a means of disposing of the waste liquor. Some
materials (such as ammonia) are dispersed adequately to the atmosphere by evaporation,
but other materials either precipitate in the coke or are carried into the air by the
steam rising from the quench tower. Phenols and other organics entering the blast
furnace with the coke (as a result of the waste liquor - water quench) are probably
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-28
safely destroyed because of the high temperatures in the blast furnace and the closed-
circuit emission-collecting systems used with blast furnaces. The same compounds
rising from the quench tower would be dispersed to the atmosphere. The use of baffles
in a quench tower might also serve to lower significantly the fallout of liquid droplets
in the vicinity of the quench tower. (35)
Ironmaking
Historically, the making of pig iron, or hot metal asit is called in the industry,
has been done mostly in,blast furnaces. Processes which bypass the blast furnace
have been the subject pf much research and development for many years. They are
termed "direct reduction" processes and produce either (1) molten pig iron generally
similar to blast-furnace hot metal, or (2) solid "metallized" products such as sponge
iron for use directly in steelmaking furnaces. A plant of the first type (rated to produce
about 200,000 tons of pig iron per year) is under construction at Mobile, Alabama, by
the McWane Cast Iron Pipe Company. (44,45) A plant of the second type (rated to produce
about 500,000 tons of steel per year from sponge iron) is under construction in
connection with the operations of the Oregon Steel Division of Gilmore Steel Company,
in Portland, Oregon. (46,47) However, such processes account for only a very small
fraction of the production~of metallic iron in the United States (roughly 0.3 percent).
The production of blast-furnace pig iron in the United States has increased steadily
over the years to keep pace with market demands. Production from 1958 through 1967
is shown in Figure C -15. The increase in production has been achieved with a
 95
C 90
o 
... 
H 
C1 85
a: 
- 80
Oil)
CC 
00 75
,- ....
....
u"" 
:J<1> 
uC 70
0-
0:0 
II) 65
-c
g,Q
c= 
C.- 60

-------
~ 95
~
+-
OJ
C 90
'+-
o
(/)
,9 85
E~ 80
o
+-
OJ
~
+-
o
I
'0 70
c
o
'+=
g 65
-0
o
~
0... 60
o
:J
C
c

-------
C-30
The blast furnace is one of the largest chemical reactors used by man. New
furnaces commonly have a hearth diameter of 30 feet or more. The blast furnace acts
as a countercurrent reactor in which solid materials descend by gravity from the top,
react with gases generated near the bottom, and then are forced upward.
Iron ore, fluxes, and coke are charged into the top of the furnace through a
succession of two or three seals that serve to limit leakage of gas at this point. Pre-
heated air (sometimes augmented with oil, gas, oxygen, or steam) is forced through
ports (tuyeres) arranged radially near the bottom of the furnace and just above the
hearth. The incoming air and admixed additives react between themselves and with the
hot coke to generate a reducing gas rich in hydrogen and carbon monoxide, at a flame
temperature of up to 3500 F. The hot reducing gases liberate some of their heat to melt
the iron and slag, then continue upward to carry energy and chemical potential to the
unreduced ore in the upper part of the furnace. Molten iron and slag drip down into the
hearth and are tapped intermittently through special ports.
The ascending gas (with a typical calorific content of about 80 Btu per standard
cubic foot) is removed from the top of the furnace, stripped of dust, then used to fire
regenerative stoves for heating more air to be blown into the tuyeres. Surplus blast-
furnace gas (not needed to heat air) frequently is burned in the powerhouse to generate
electricity for the blowing engines (and other purposes), and less commonly is used to
heat the flues of the coke -oven plant.
The molten iron tapped from the blast furnace is put into special, large railroad
cars for transfer to steelmaking operations. The slag, also molten, may be granulated
with water, or may be put into steel pots for conveyance to a dump area. Dust trapped
in gas -cleaning operations frequently is recycled via the sinter plant.
Identification of Emis sions
The points at which emissions are generated in the production of pig iron in a blast
furnace are described with respect to the several processing steps involved. Flow
diagrams of blast-furnace operations having different modes of operation are given in
Figures C-18 through C-24. The range of burdens covered in these flow diagrams is
from unscreened ore (by today's technology, an unusual situation generally regarded as
"old hat") to "modern" burdens consisting mainly of sinter, pellets, or mixtures of
sinter and pellets. Several of the flow diagrams show the use of preheated air without
the injectionof any auxiliary fuel or steam, while two of the diagrams allow for the
injection of natural gas with the air blast. Many other arrangements of component
parts of these flow diagrams are possible. Combinations of burden, fuel and air used
to describe variations here, however, cover a wide range, from the conditions shown
in Figure C -18 where manufacture of 1 net ton of pig iron requires about 3000 pounds of
coke and 9750 pounds of heated air, to the conditions shown in Figure C-24 where
manufacture of the same weight of pig iron requires about 1025 pounds of coke and
4300 pounds of heated air.
In Figures C-18 through C-24, circled numbers are inserted at major points of
expected potential air-polluting emissions, and these circled numbers will be used later
to identify types of emissions with the sources as given on the flow diagrams.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
CD
@
C-31
@
BLAST FURNACE
@
Slag,
1,050 pounds
G)
Blast air,
9,7.50 pounds
(a) Some values disclosed by
various sources are 15, 59,
85, 100, 182, and 296
pounds.
@
@
Pig iron,
2,000 pounds
Other plant use,
10,950 pounds
Heated air,
9,750 pounds
Com bustion

products,
5,650 pounds
@
FIGURE C-18. TYPICAL BLAST-FURNACE OPERATION ON A BURDEN OF 100 PERCENT UNSCREENED ORE
- .
BATTELLE MEMORIAL INSTITUTE - COLUMB'US LABORATORIES

-------
C-32

-------
C-33
Q)
Q)@@
@
@
Iron are,
1,250 pounds
Sinter,
1,690 pounds
Coke,
1,370 pounds
Limestone,
350 pounds
Heated air,
4,880 pounds
BLAST FURNACE
@)
@
@
Slag,
784 pounds
Top gas,
7,300 pounds
Pig iron,
2,000 pounds
Other plant use,
5, 500 pounds'
G)
Com bustion air,
1,225 pounds
Blast air,
4,880 pounds
Heated air,
4,880 pounds
Com bustion
products,
2,770 pounds
@
FIGURE C-20. TYPICAL BLAST-FURNACE OPERATION ON A BURDEN OF LUMP IRON ORE AND SINTER
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
.

-------
@@@
@
Sinter,
2,600 pounds
Screened ore,
800 pounds
@)
Slag,
830 pounds
CD
C-34
@
Coke,
1,445 pounds
BLAST FURNACE
@
Limestone,
200 pounds
Heated air,
5, 850 pounds
@
@
Pig iron,
2,000 pounds
Other plant use,
6,600 pounds
Combustion air,
1,500 pounds
Atmosphere
Blast air,
5,850 pounds
Combustion
prod ucts.
3,380 pounds
Heated air,
5,850 pounds
FIGURE C-21. TYPICAL BLAST-FURNACE OPERATION ON A BURDEN CONSISTING MAINLY OF SINTER
BATTELLE MEMORIAL INSTITUTE ",:"CQLUMBUS I..ABORATORIES

-------
C-35
(i)
@
@
Heated air,
4,660 pounds
BLAST FURNACE
@
@
@
Slag,
440 pounds
Dust,
40 pounds
Top gas,
6,990 pounds
Pig iron,
2,000 pounds
Other plant use,
5,240 pounds
G)
Com bust ion air,
I, 190 pounds
Top gas,
1,750 pounds
Blast air,
4,660 pounds
Heated air,
4,660 pounds
Combustion
products,
2,695 pounds
@
FIGURE C-22. TYPICAL BLAST-FURNACE OPERATION ON A BURDEN OF 100 PERCENT PELLETS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-36

-------
C-37
G)@@ 
-------
C-38
Charging of Raw Materials. The immediate solid raw-material requirements of
the blast furnace are held in surge hoppers under the railway high-line used for the
transfer of raw materials from storage areas to the blast furnace. In the case of coke
(and sometimes sinter), these hoppers usually are filled by conveyor belts; in the case
of ore and flux, bottom-dump-transfer cars frequently are used. From the surge
hoppers (called "pockets"), the materials are drawn in a predetermined sequence into a
system that weighs the prescribed charges and deposits them into a skip hoist leading
up to the furnace top. In some newer installations, conveyor belts are used to charge
the raw materials to the blast furnace. (48) The use of a conveyor-belt system.for
charging requires more area than a s~ip-charging system, and this is a disadvantage in
established steel plants where ground space is usually at a premium. Future new blast-
furnace plants probably will make more extensive use of conveyor belts for charging.
. .. !
The transfer and weighing system can take.many forms. In conventional United
States stockhouses, coke i"s fed almost directly into the skip horst, whereas, ore and flux
are discharged from the pockets to ,a second transfer car that is scale-mounted.
Operations in the stockhouse usually "dusty", depending on the materials, the weather,
and the degree of shrouding of transfer points. Newer stockhouses rely on conveyor
belts replacing the transfer car, and can be considerably cleaner (but are not nece's sarily
cleaner) .
The skip hoist containing a component of the charge is ,hoisted to the topmost part
of the furnace and dumped into a receiving hopper. This transfer is highly exposed, but
partial shrouding is possible. From the first receiving hopper, the charge is dropped
stepwise through one or two more hoppers and closures into the furnace. The multiple-
closure system is used to contain the furnace gases as charge is added.
The degree of fit between hoppers and their closures (called "bells") is variable
with design, age of the equipment, quality of maintenance, and other factors. Generally
the top equipment tends to become leaky with use because of abrasion, wear, creep,
and other di storting factor s. The higher the operating pres sure of a furnace, the faster
leakage develops. In new high-pressure blast furnaces, three closures are usually used
and an artificial steam system is used to maintain back-pressure between the two closed
seals during each transfer.
The present state of affairs is that the t9P of an operating blast furnace is
considered to be an area of continuous gas hazard while the furnace is in operation. Men
who must do maintenance work on top are sent aloft in safety groups, with compressed-
air breathing packs or hose-connected masks. It is probably practically impossible for
blast furnaces, even new ones, to be operated without continuously discharging at least
some noxious gases to the atmosphere.
Two emis sions come from the top of a blastfurnace; top gas and the dust which it
entrains. The top gas is a mixture mai:nly of steam, nitrogen, carbon monoxide, and
carbon dioxide. On a dry basis, this gas may average 25 to 30 volume percent carbon
monoxide; thus it is toxic. Emissions to the atmosphere occur from leakage around
hoppers and seals. Top gas may leak from instrumentation such as ports for rods used
to determine the height of the charge materials inside the furnace. Dust entrained in the
top gas is a result of the abrasion sustained by the burden materials during charging and
during the initial stages of passage down the blast furnace. It is possible to minimize
particulate emissions by using good raw materials and sound operating practices and
thereby reduce the load on the dust-cleaning system. The effect of improving the burden,
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-39
by using increased amounts of sinter and pellets,
net ton of hot metal going into the dust-collecting
Figure C-25. (29)
on the amount of dust generated per
system is illustrated in
(/)  
"0  
C  
:J  300
o 
a.  
~  
E  
(1)  
-  
(/)  
>-  
Cf)  
CJI  200
c 
+= - 
U a 
(1)- 
-(1) 
<5 E 
u- 
-0 
~.c: 
0- 100
 o
.£c 
(1)0 
- - 
0- 
c:: (1) 
_c 
(/) ... 
:J (1) 
00. 
FIGURE C-25.
a 1000 2000
Amount of Sinter Plus Pellets Charged,
net ton of hot metal.
(Balance of burden is i ran are.)
EFFECT OF BURDEN IMPROVEMENT ON DUST RATE
FROM A BLAST FURNACE TO ITS DUST-COLLECTING
SYSTEM
High above the charging system on a blast furnace, at the uppermost part of the
gas-collection mains (called uptakes), two or more safety valves ("bleeders") are
located to relieve "unusual" gas pressures within the furnace. During abnormalities
in furnace operation, sudden movements ("slips") of the burden materials may occur,
During slips, the bleeders will open automatically to relieve the high pres sures, and
will discharge dust and gas to the atmosphere. The operating factors producing ab-
normal gas pressures also tend to increase the dust loading of the gas. Fortunately,
the use of improved raw materials to reduce the dust loading in a blast furnace also
tends to minimize abnormal operating conditions and lower the frequency of slips. With
modern burdens in use, the discharge of dust and gas during slips occurs only occa-
sionally. American plants have advanced considerably in this respect since 1950, and
for some furnaces the bleeders rarely open. A blast-furnace plant containing eight
furnaces .has reported that only one or two short-interval openings of the bleeders occur
during any given week for all eight furnaces. (49)
Smelting of Iron, The conventional blast furnace is a steel shell lined with re-
fractory bricks (see Figure A-7), and there may be numerous ports in the steel shell
to admit piping for water-cooled plates, Recently, blast-furnace designs have been
changing toward complete water jacketing with a closed shell. To the extent that the
furnace shell is pierced for instruments and coolers, gases generated inside the blast
furnace may leak back through the brickwork and to the exterior of the furnace shell
where they usually burn as small flames.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-40
Internal reactions, besides generating considerable carbon monoxide, can be quite
complex. The contact of .nitrogen in the blast air with hot coke, for example, conceiv-
ably can generate a variety of carbon-nitrogen compounds including cyanide. Tramp
elements in the burden (such as selenium) might be trapped, accumulated, and re-
fluxed in the furnace, or slowly emitted in the gas as some compound. Blast-furnace
burdens used in the United States usually are relatively free from known noxious
materials. However, as blast-furnace campaigns (the period between blast-furnace
relinings) lengthen so as to produce perhaps 4 million tons of iron, which requires
the smelting of about 6 million tons of ore, 2 parts per billion of arsenic in the ore
conceivably could accumulate as a ~4-pound slug of reflux arsenic in the furnace.
No data are available yet on these types of accumulations or on the possibilities for
their emittance.
Casting and Flushing. Throughout the preceding discussion, the sulfur charged
as part of the coke and as part of other materials has not been mentioned. Curiously
enough, almost none of this sulfur enters the top gas. Most of it is trapped and fixed in
the furnace by the fluxes, and is almost completely partitioned between the iron and the
slag in the hearth.
A modern American blast furnace produces between 400 and 700 pounds of slag per
ton of hot metal. On a volume basis, slag weighs about one-third as much as iron.
Therefore, the volume of slag is equal to or somewhat less than the volume of iron.
Except in very fast operations ,the' slag is flushed out of the furnace twice for each tap
of iron; once with the iron and once through a special slag notch about an hour before the
iron is cast. Some fast operations with frequent iron taps and low slag volume omit the
preliminary slagging operation.
The sulfur that is not dissolved in the iron is combined in the slag. The sulfur in
the iron is kept under very close control, rarely exceeding 0.6 pound per ton of iron,
and usually maintained at 0.5 pound per ton. The extent of sulfur control by the slag can
be illustrated by the following discussion.
If we consider a burden with 1100 pounds of coke (at 0.65, percent sulfur) per ton
of iron, ,and containing 1 pound of sulfur per ton of iron (from other sources such as ores
and sinter), the total sulfur load in the charge is just over 8 pounds per ton of iron. A
normal American range is 7 to 10 pounds. Deducting 0.,5 pound for sulfur in the iron,
one may deduce that the sulfur load on the slag is 6.5 to 9.5 pounds per ton of iron. The
amount of sulfur in slag normally ranges from 1 to 1. 8 percent.
When the hot slag is flushed from the blast furnace, the sulfur reacts with oxygen
in the air to form sulfur dioxide near the slag runners. In damp weather hydrogen
sulfide may also be formed by these reactions:
2 S-- + 302 -+ 2 S02 + 2 0--
S-- + H20 -+ H2S + 0-- .
The air can be fouled by these reactions during flushing and casting alike. At older
furnaces, the slag-runner system may extend for 100 feet or more, so that considerable
surface is exposed. In newer practices, slag is often run a short distance and then
quick-cooled and granulated with high-pressure water. This increases the effective
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
G-4l
surface area of the slag, and.the formation of hydrogen sulfide may continue in the
granulation pit at low temperature.
The molten iron as tapped from the blast furnace is saturated with carbon dissolved
from coke in the hearth of the blast furnace, and rejection of flakes of graphite begins
almost immediately as the iron emerges from the furnace. This graphite (called "kish")
is very much lighter than the iron, so it rises quickly to the surface where currents of
heated air sweep it into the atmosphere. Other emissions from the surface of the hot
iron include manganese vapor, which oxidizes upon escape to form a fine dust.
These emissions from the surface of iron generally are considered harmless by
the steel industry, and the amount of manganese vapor is usually small. But the
kish is a substantial dirt nuisance because it is eas.ily borne for long distances by light
breezes and has an oily tenacity that makes it hard to wash away from the articles it
settles on. At present, kish control is a matter of minimizing the amount precipitated.
This is done by running the iron short distances into closed" submarine" or torpedo>:'
ladles that are well preheated. The combination of minimal temperature drop and small open
surface is helpful in minimizing the formation of kish. Kish rising to the opening of the
ladle can usually be raked away and disposed of before it becomes airborne. Those
plants still using open-top ladles and oil-fashioned long runners have a substantial kish
problem which is worsened when the iron is transferred out of the ladles to a mixer or
pig machine. In a few instances, hooding has been installed with modest (but not good)
effecti venes s.
Kish also is released when hot metal is transferred from the submarine ladles to
the pouring ladle used to charge the steelmaking furnace. Special emission control
equipment is being installed in some steel plants to remove the kish at the hot-metal
reladling stations.
Burners and Stoves. A high proportion of the gases generated during the blast-
furnace smelting reactions is burned in stoves to heat refractory-brick checker work
(see Figure A-4), which in turn releases the heat to the air that is blown through the
blast stoves on its way to the blast furnace. The amount of blast-furnace gas generated
is primarily a function of the coke rate for any particular smelting practice, as illus-
trated in Figure G-26. (50) The distribution of gas between blast-stove use and other
in-plant uses is also shown. It should be noted that as blast-furnace practice improves
so as to lower the coke rate, availability of blast-furnace gas for other in-plant uses
decreases.
Blast-furnace top gas is thermally lean, and as blast-furnace operations have
improved, the gases tend to get still leaner. Whereas gas at 95 Btu per cubic foot was
common in 1950, the more usual heating value is now 80 to 85 Btu per cubic foot. The
calorific value of blast-furnace top gas as it is affected by the coke rate (which is an
indication of the degree of technology used in smelting) is shown in Figure C-27. (50)
Some gases are so lean that they must be mixed with coke-oven gas or natural gas to
assure adequate combustion and a sufficiently high flame temperature.
The burners for the blast stoves are fired alternately according to a predetermined
cycle and may be highly automated. One of the three or four stoves is always "on blast"
(i. e., heating blast air), while the others are "on gas" (being fired and heated). The
stoves are changed from blast to gas when they cool to a temperature at which they can
"The word "submarine" is derived from the shape of the ladle.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-42
no longer attain the required temperature of blast air. A conventional stove system ducts
all combustion gas to a common chimney, although separate -chimney installations still
exist. The drafting of a common chimney is better, and it can be built higher on the
same budget as several separate chimneys.
140
o
....
(1) 120
E
....
~o
~.c 100
<.9-
(1)0
(.) 5 80
g....
~ ~
~ ~ 60
1i)CU
o~ 40
(1) .~
..c
B 20
o
8
Gas available for
other use
o
1000 1200 1400 1600 1800 2000
Coke Rate, pounds per net ton of hot metal
FIGURE C-26.
2200
EFFECT OF COKE RATE ON VOLUME OF BLAST-FURNACE
GAS PRODUCED
125
+-
o
o
-
120
()
B
:J
()
...
(1)
a.
:J
-
(1)
i 100
o
>
()
-
...
o
o
U
+-
Q)
Z
FIGURE C-27.
"",,- --;
" .
" . ..
,fI1I'. .. .:. --
""...... .... ~.--
"" .:. .".--
~'I."'. :or"

A.I ...,
~ .,. ~
.-" .' "".
~. . "
,.-... '\.""
/ .:~, Y
/ . . ..)
/":t .,t
, ..C. ... /...
. ..~.
, ... .'"'
/ .'/
, . lit .. /
85 / /
./
. .. "
80 '
1000 1200

Coke Rate,
115
110
105
95
90
1400 1600 1800 2000
pounds per net ton of hot metal
EFFECT OF COKE RATE ON THE CALORIFIC VALUE OF
BLAST-FURNACE TOP GAS '
Heat release in the stoves is by the burning of carbon monoxide and hydrogen, with
traces of methane if natural gas is being injected at the tuyeres. These fuels are burned
with excess air, regulated by combustion controllers, often aided by excess -oxygen
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-43
sensors. The chimney effluent, then, generally is relatively free of objectionable
components (unless sulfur-bearing coke-oven gas is used to improve heat release).
Fuel Injection. The coke consumed in a blast furnace is a relatively expensive
fuel. A case can be made for the fact that in terms of effective heat released it costs as
much as electricity. To minimize coke consumption, heavy oil, tar, coal, coal-oil
slurry, coke-oven gas, and natural gas are in use as tuyere injectants. They are
introduced through blowpipes arranged to' discharge into the stream of hot blast air just
as the blast enters the tuyeres of the furnace.
Under normal conditions, these auxiliary fuels do not alter the types of emissions
from ironmaking. But under start-up conditions, and when the amounts injected are
large, auxiliary fuels may be incompletely burned at the lower level. As with any flame,
incomplete reaction can produce soot, and it is usual'that such' soot will blow on through
the furnace, foul the dust system, and overload the wet collectors until the situation is
recognized and corrective Ci.ction taken. Hopefully, as blast-furnace operators gain
experience with the use of injected auxiliary fuels, such temporary problems will become
less frequent.
Characteristics of Emissions. Emissions generated in the making of iron in the
blast furnace and in its immediate auxiliaries have major chara<:;te'ristics thatnow will
be described. The circled numbers accompanying these descriptions refer to locations
of emissions as identified by corresponding circled numbers on the flow diagrams that
make up Figure s C - 18 through C - 24. .
CD Iron-ore dust - Particles are rounded to' elongated in shape and can be as
small as 2 microns. Larger particles are opaque and red orange in top
light. Individual sma~l grains are transparent and blood-red. (11)
Hardness: 5 (Mohs)
Specific gravity: 5.2
Chemistry:
Usually mostly Fe203 with some Fe304.
in HCl. Contains some silica, alumina,
oxide s.
Mostly soluble
and phosphorus
o Coke dust - Particles are opaque, irregularly shaped, quite porous and rough
with some straight, sharp edges. They are gray-black in reflected light. (11)
Chemistry:
(Typical) 85-90 % fixed carbon, 2 % maximum volatile
matter, U.6 - 1.5% sulfur, 0.018 - 0.040 % phosphorus,
balance ash. (51) ,
CD Limestone dust - Mineral name calcite. It'is colorless, with light-transmitting
characteristics varying from transparent to translucent. Particles generally
occur as rhombohedra because of their perfect rhombohedral cleavage. Frag-
ments may also occur as prisms. (11)
BATTEL'LE MEMORIAL INSTITUTE - COLU'MBUS LABORATORIES

-------
C-44
Hardness
: 3 (Mohs)
Specific gravity: 2. 7
Chemistry: Mostly CaC03

o Flue dust - The blast-furnace flue dust typically contains 15 percent metallic
iron, 40 percent red iron oxide, 40 percent magnetic iron oxide, and 5 percent
limestone, (11) but many variations reust exist.
Iron - Fragments are opaque, black, and sharp, ma,gnetic, with finely
granular surfaces,( 11)
. Red iron Oxide (hematite) - Particles are transparent, rounded grains,
usually less than 2 microns in maximum dim~~sion. (11)
Magnetic iron oxide (magnetite) - Particles are opaque, black, rough
fragments, partially or completely coveredwithred iron oxide.(l1)
Limestone dust - Transparent, colorless rhombohedra, and rounded
particles. Many particles may also be covered with red iron oxide. (11)
Results of chem"ical anC!-lyses on flue dust generated in blast-furnace practice
are tabulated in Table C-9. (52, 53) The size analysis of the same sampling of flue
dust is given in Table C-10.
. -
Flue dust out of United States blast furnaces is reported to vary from 40 to 90
pounds per net ton of hot metal produced in a multi-furnace operation, with one new
furnace generating 40 pounds of dust per ton of iron. (1) A blast furnace in the United
Kingdom operating with a high-pellet burden is reported to generate only 20 pounds of
dust per ton of iron. (12) The dust loadings for the four blast furnaces of Great Lakes
Steel Corp., operating with a sinter and pellet burden were reported in 1967 to be as
shown in Table C-11. (54) These are dust loading to the dust-c~llecting system as
determined by actual collection. Specific data on particulate emissions to the atmo-
sphere from leakage at the top of the blast furnace or from slips are not available to
the project.
@ Top gas - The chemistry of blast-furnace top gas is determined by the
nature of the burden used in any particular furnace and the operating
variables such as blast temperature, injection ofa~xiliary fuels, and
the additionof moisture. The relationship betw,een moisture in the blast
air and the hydrogen content in the top gas is illustrated in Figure C-28. (55)
Top-gas analyses for several differ-ent types of burden situations in the United
States are listed in Table C-12. The compositions of the top gases are fairly consis-
tent, except for the burdens using unscreened ore, in which instances the carbon
monoxide contents are high and the carbon dioxide contents low. The unscreened ore
itself probably is not the:: caus~ of this difference in gas composition. The difference
probably is the result of other variations in practice that accompany the use of
unscreened ore.
BA.TTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-45
TABLE C- 9. CHEMICAL ANALYSES OF DRY, BMST-FURNACE FLUE DUST(52, 53)
 Weight Percent 
Component Range for Several Plants Midwest Plant
Fe 36.5 - 50.3 47.10
FeO n.a.  11.87
Si02 8.9 - 13.4 8.17
Al203 2.2- 5.3 1.88
MgO 0.9- 1.6 0.22
CaO 3.8- 4.5 4.10
Na20 n.a.  0.24
K20 n.a.  1.01
ZnO n.a.  0.60
P 0.1 - 0.2 0.03
S 0.2- 0.4 n.a.
Mn 0.5- 0.9 0.70
C 3.7 - 13.9 n.a.
n. a. - not available.   
TABLE C-IO. SIZE ANALYSIS OF BLAST FURNACE FLUE DUST FROM
U. S. BLAST FURNACES(52) .
Size
U. S. Series Microns Range, percent
20 833 2.5 - 20.2
30 . 589 3.9-10.6
40 414 7.0 - 11. 7
50 295 10.7 - 12.4
70 208 10.0 - 15.0
100 147 10.2-16.8
140 104 7.7 - 12. 5
200 74 5.3- 8.8
-200 -74 15.4 - 22.6
TABLE C-l1. DUST LOADINGS FOR GREAT
LAKES BLAST FURNACES(54)
Blast Furnace

Designation
Dust Loading,
pounds per net ton of hot metal
-------
----
A
B
C
D
60
39
28
36
-----------------.------
-----.-------------------'--
BATTELLE MEMORIAL INSTITUTE - COLUMBUS 'LABORATORIES

-------
C-46
4.0
  o
 o 0
  ~~
  B
8  0
8 0 8
  o
8 000
 c}o
  o
  o
o  
  0
-
c
Q)
o
L..
Q)
a.

Q)
E
:J
(5
>
o
3.6
o
3.2
o 000
o
o
II)
o
~
a.
~
2.8
. .
. .. .
. ...
,.. .
. .. ..
.. ...
..",.. .... .
00 .
Legend

Natural moisture
Added moisture
o
2.4
-
o
+-
C
Q)
+-
C
8
.
.
....,
I
o
c
Q)
01
o
L..
'0
:>.
I
.
:s..
1.2
12
grains per cubic
2 4 6 8
Moisture Content of Blast,
foot of air
FIGURE C-28. RELATIONSHIP OF MOISTURE IN BLAST AIR TO HYDROGEN IN BLAST-FURNACE TOP GAS
TABLE C-12. TOP-GAS ANALYSES FOR DIFFERENT BLAST-FURNACE BURDENS
-------------------
--------------- _._---- ----------------
-----.-.--,-.-----.-------- --------.---------.---
Burden
Auxiliary
Fuel Injection
Gas Content, volume percent
CO C02 H2
Unscreened Ore None 41. 5 7.1 3.3
Unscreened Ore None 32.9 7.3 3.6
Screened Ore None 27.2 12.8 3.1
Screened Ore Natural Gas 26.7 13.7 2.1
100 perc~nt Sinter None 23.7 14.8 3.5
100 percent Sinter Natural Gas 25.9 13.2 1.7
100 percent Sinter Natural Gas 23.5 16.3 5.7
50 percent Sinter +    
50 percent Unscreened Ore None 24.1 14.1 4.1
50 percent Sinter +    
50 percent Screened Ore None 25.0 14.2 2.6
50 percent Sinter +    
50 percent Pellets None 23.8 17.3 2.4
100 percent Pellets None 24.9 15.4 2 9
100 percent Pellets Natural Gas 23.8 19.0 3.6
100 percent Pellets Natural Gas 24.0 16.7 2.9
100 percent Pellets Coal 26.1 14.4 3.4
----------- --------'-------------'- ------
--------_._- ---.--- --'- -------_._--------------- - --.--
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-47
@ Kish - Carbon in the form of flaky graphite that is rejected by the molten
iron as it cools during flow from the blast furnace to ladles. Other types
of particles may be entrained with this kish. The graphite particles are
opaque, black, sharply angular flakes with smooth surfaces. Some are
in layered agglomerates, occasionally showing rounded lZ0-degree angles,
and even forming rounded hexagonal tablets. Other particles accompanying
the kish may consist of opaque, black, rather coarse, fragments of magnetic
iron oxide, and transparent, deep-red, rounded particles of hematite.
Traces of quartz and calcite may also be found with the kish. (11) Graphite
typically makes up about 90 percent of the emis sian, with the magnetic
oxide at 5 percent and hematite at 5 percent.
Chemistry: Graphite (C)
Magnetic iron oxide (Fe304)
Hematite (FeZ03)
Quartz (SiOZ)
Calcite (CaC03)
(2) Hydrogen sulfide - When blast-furnace slag comes in contact with water, a
reaction occurs that forms. small amounts of hydrogen sulphide. The reaction
takes place during granulation, when the slag is hot, as well as at ambient
temperatures when the slag is cold. No data are available as to the amounts
of hydrogen sulfide that escape into the atmosphere. Because hydrogen
sulfide is detectable in very low concentrations, and has an unpleasant odor,
it is considered to be an air pollutant. Complaints from the surrounding
neighborhood tend to emphasize the problem. This is particularly true when
the slag is hauled by truck through the neighborhood on a rainy day. Research
work is presently underway, sponsored by the American Iron and Steel
Institute, and is directed toward determining methods of suppres sing the
formation of hydrogen sulfide from this source. (56)
@ Combustion emissions - Carbon monoxide emission produced from the
burning of fuels in the firing of blast stoves is small because of the careful
controls maintained for combustion, and the fact that burning is carried out
under conditions of exces s air. Sulfur dioxide can be present in the products
of combustion, if coke-oven gas is used to heat blast stoves.
Blast-Furnace Emission-Control Equipment
The primary reason for cleaning blast-furnace top gas has been to make it suitable
as a fuel for heating blast stoves, and secondarily to provide clean fuel to other opera-
tions in the steel plant. Generation of a recycled fuel in the smelting of iron is one of the
features that contributes to the economy of the blast-furnace process. If blast-furnace
gas were not cleaned, the particulate matter would clog the holes in the regenerative
brickwork of the stoves, slagging reactions would be accelerated and might lead to
catastrophic failure of the large amount of brickwork in the stoves. Changes in the
technology of blast-furnace practice have led to the use of. higher blast temperatures
from an average of 1300 Fin 1960 to temperatures varying from 1550 F to 1850 F in
1969. This continuing trend toward higher blast temperatures has resulted in the design
and construction of blast- stove refractory tile with smaller holes and thinner walls,
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-48
which are necessary to improve the heat-transfer characteristics in the blast stoves.
It is interesting to note the contrast between statements made on this subject in 1956
and 1967:
1956 - "As a result of the processes described, the cleaned gas has a
dust content of less than 0.01 grains per cubic foot. Gas this clean permits
the use of smaller checkerbrick in the stoves, which provides greater heating
surface area per stove and makes possible the use of higher blast temperatures
with a consequent improvement in furnace efficiency. (57) .
1967 - "Future high-temperature operation at 2000 F or more will
necessitate the installation of high-energy burners and will create the need
for gas cleanliness in the 0.001 grain per scf range. (54)
The type of emission-control equipment used on blast furnaces is affected by
( 1) the operating blast temperature, which governs the openings required in the tile in
the blast stoves, and (2) the availability of space around the blast furnace. The various
systems used have certain common pieces of equipment such as dust catchers and
primary washers, while subsequent items of equipment may be dictated by other in-plant
use of the blast-furnace gas. Flow diagrams of some blast-furnace gas-cleaning sys-
tems are illustrated in Figures C-29, C-30, and C-31. (58, 59, 60,61,62,63,64,65)
The gas cleaning system on the newest blast furnace constructed in the United
States (Youngstown Sheet and Tube Co. 's No.4 blast furnace at Indiana Harbor) consists
of a dust catcher, two automatically adjusted venturi scrubbers, and a gas-cooling
tower. It is claimed that this system will clean the gases to a dust loading of 0.005
grain per cubic foot. (66)
BLAST STOVES
and
BOILERS
PLANT USE
COKE OVENS
Weirton Steel, Weirton, W. Va. - 1958
U. S. Steel, Geneva, Utah - 1959
FIGURE C-29. FLOW DIAGRAMS OF TWO EARLY BLAST-FURNACE GAS-CLEANING SYSTEMS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
PLANT USE
Kaiser Steel,
Fontana, Cal. - 1961
C-49
PLANT USE
PLANT USE
BLAST STOVES
OTHER
PLANT USE
CF & I.
Pueblo, Colo. - 1962
Jones & Laughlin,
Cleveland, Ohio - 1963
Armco Steel,
Ashland, Ky. - 1963
FIGURE C-30. FLOW DIAGRAMS OF BLAST-FURNACE GAS-CLEANING SYSTEMS
PLANT USE
U. S. Steel,
Fairless, Pa. - 1965
PLANT USE
PLANT USE
Great Lakes Steel,
Ecourse, Mich. - 1967
Bethlehem Steel,
Sparrows Pt., Md. - 1968
FIGURE C-31. FLOW DIAGRAMS OF THREE RECENT BLAST-FURNACE GAS-CLEANING SYSTEMS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
G-50
Steelmaking
Prior to 1960, by far most American steel was made in open-hearth furnaces,
but since then the introduction of basic oxygen steelmaking (BOF) furnaces and the
installation of larger and more efficient direct-arc electric furnaces has resulted in a
decline in the percentage of steel made in open-hearth furnaces. In 1967, about
55 percent of the steel produced in the United States was still made in open hearths.
However, one set of projections made in that same year indicated that by 1969 basic
oxygen furnaces may be producing 59.5 percent of the steel, open hearths 20.5 percent,
and electric furnaces 20.0 percent. (67) Battelle's present projections are given in Sec-
tion IV of the :nain body of this report.,
Open-Hearth Furnaces
Although the open-hearth furnace is well on its way to lowered importance as a
means for making steel, and some forecasters see virtual extinction by 1990(68), there
are still many open hearths in operation or on a stand- by basis to be placed into opera-
tion to meet peak demands for steel. About 467 open hearths are in this category as of
mid-1968; by 1973 this number will be reduced to 257. The number is decreasing as
new BOF steelmaking capacity is installed. The distribution of all United State s open-
hearth furnaces by nominal capacity for 1968 and 1973 is shown in Figure C-32, while
similar distributions for several geographical steelmaking districts are similarly shown
in Figures C-33 and C-34. It can be seen that in 1968 the highest concentration of open-
hearth furnaces is in the Chicago District with 102 furnace s (9 furnaces located in
Duluth, Minn., for a district total of 111), followed by the Pittsburgh District with 101
furnaces, Youngstown with 65, and the Northeast Coast District with 43. No open
hearth furnaces are in operation in the St. Louis district.
The open-hearth proces s (known outside of the United States as the Siemens-
Martin Process) was developed on the basis of the regeneration principle, where checker
chambers are used for regeneration of heat. Regeneration was a necessity for the classic
open-hearth furnace because a gaseous or liquid fuel burned with ambient-temperature
air produces a flame temperature only slightly above 3000 F, which is not high enough
to melt the charge materials. Transfer of heat from the flame to the charge is primarily
by radiation. By using regenerators to preheat the air to 1000 to 1200 F, the flame
temperature is raised high enough to melt the charge materials and carry out the making
of steel. With the availability of tonnage oxygen, checker chambers and the regenera-
tive principle became unnecessary, and the basic design of the furnace no longer was
completely satisfactory for the rapid production of steel from charges high in proportion
of hot metal.
Although many open-hearth furnaces operate today with oxygen for refining the
steel, information on the number of open-hearth furnaces operating today with and with-
out oxygen was not available during this study. (69) Conditions that would favor open-
hearth steelmaking without oxygen-refining include (1) use of cold-melt charges of scrap
steel, (2) manufacture of steel with high ~arbon contents, and (3) local economic or air-
pollution considerations.
Emission Identification. The conventional open-hearth process begins usually with
the charging of up to 50 percent home scrap and purchased scrap to the hearth of the
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
100
~
~ 50
E
:J
Z
90
C-51
Total of 467 open hearth furnaces in 1968.
1973 = 257 furnaces (forecast)
Shaded portion represents furnaces forecast to
be removed from operation by 1973.
125 175 225 275 325 375 425 475 525 575
100 150 200 250 300 350. 400 450 500 550 600

Nominal Open Hearth Furnace Capacity, net tons
FIGURE C-32. SIZE DISTRIBUTION OF OPEN-HEARTH FURNACES
IN THE UNITED STATES IN 1968 AND 1973
80
70
60
40
30
20
10
o
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
L
~
Q)
..CI
E 10
::J
Z
~ 20
Q)
..CI
E
::J 10
z
~
Q)
E 10
::J
Z
~
Q)
..CI I 0
E
::J
Z
~
Q)
..CI
~ 10
z
20
30
20
20
30
C-52
North East Coast district 43 furnaces in 1968
16 furnaces in /970
o
~ To be removed from operation
Pittsburgh District 101 furnaces in 1968
37 furnaces in 1973
o
,Buffalo District 28 furnaces' in 1968
6 furnace in 197 I
o
Cleveland District 18 furnaces in 1968
6 furnaces in 1971
o
20
Youngstown District 65 furnaces in 1968
42furnacesinl970
o
125 175 225 275 325 375 425 475 525 575
100 150 200 250 300 350 400 450 500 550 600

Nomina 1 Open Hearth Furnace Capaci ty I net tons
FIGURE C-33.
SIZE DISTRIBUTION OF OPEN-HEARTH FURNACES
IN SEVERAL GEOGRAPHICAL DISTRICTS FOR 1968
AND SUBSEQUENT YEARS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-53
 20
~ 
OJ 
..0 10
E 
::J 
Z 
 0
 40
 Duluth
 30
~ 
OJ 
..0 20
E 
::J 
Z 
 10
 o
 20
~ 
OJ 
E 10
::J 
Z 
 0
 20
~ 
OJ 
..0 10
E 
::J 
Z 
 0
Detroit District 12 furnaces in 1968
No furnaces after 1969
(QJ To be removed from operation
Chicago District
102 furnaces
9 furnaces - Duluth, Minnesota
III furnaces total in 1968
78 furnaces in 1973
District 27 furnaces in 1968
21 furnaces in 1970
Southern District 26furnaces in 1968
20
Western District 31 furnaces
27 furnaces
~
OJ
..0 10
E
::J
Z
o
125 175 225 275 325 375 425 475 525 575
100 150 200 250 300 350 400 450 500 550 600

Nomina I Open-Hearth Furnace Capacity, net tons
FIGURE C-34. SIZE DISTRIBUTION OF OPEN-HEARTH FURNACES
IN SEVERAL GEOGRAPHICAL DISTRICTS FOR 1968
AND SU BSEQUENT YEARS
(Size range as follows: ISO-ton furnaces includes open
hearths having nominal capacities from 150 to 174 net
ton s , etc.)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-54
empty furnace. ':' After a period of intense preheating of the scrap, hot metal from the
blast furnace is added, and firing is continued until the scrap is fully dissolved in the
molten charge. During this period of meltdown, limestone is added to form a basic
slag and to release carbon dioxide. The CaO component of the limestone and the C02
react with phosphorus, silicon, and manganese to form their respective oxides, which
in turn are absorbed by the slag. Another source of oxygen for these reactions is the
preheated scrap which becomes highly oxidized as it is melting. Iron ore may also be
added as a source of oxygen to help continue oxidation reactions and to remove carbon
from the hot metal.
Availability of high-purity oxygen has enabled steelmakers to shorten the refining
stages and combine them into one step by the use of oxygen lances. The lances extend
into the furnace through the roof and are activated immediately after the addition of the
hot metal. The rate of oxidation of metalloids and carbon is two to four times faster
than when limestone and ore are the main sources of oxygen.
Particulate and gaseous emissions from the open-hearth process originate from
(1) the physical action of the flame on charged materials and the resulting pickup of
fines, (2) the chemical reactions in the bath, (3) the agitation of the bath, and (4) the
combustion of fuel. (70, 71) The emissions include sulfur dioxide, carbon monoxide,
carbon dioxide, and fly ash from the fuels, plus iron oxide and other metallurgical fumes
from the steelmaking process. The metallurgical fumes include fine silica from the
burning of silicon monoxide released from the molten bath, manganese oxide from
manganese vaporized from the bath and subsequently oxidized, and iron oxide from the
rust on scrap, or in later parts of the heat, and from iron droplets oxidized in the
open area above the steel during refining.
Emis sions from open-hearth steelmaking come from the materials and fuels
used in the process. Variations in the types and amounts of emissions vary according
to the stage of the process. Some minor particulate emissions may be generated dur-
ing the charging of materials into the furnace and during tapping of the heat, but these
are of a minor nature. Flow sheets for various operating practices are given in Fig-
ures C-35 through C-41. While some of these practices may rarely be used, the flow
sheets are given as a matter of information. Figure C-4l (oxygen practice with
60 percent hot metal) is most typical of today' s operations in large steel plants.
Open-hearth furnaces generate four major types of particulate emission.
are indexed below with circled numbers keyed to Figures C-35 to C-4l.

CD Open-hearth dust - Charging period. Two components appear in the dust
generated during charging of the furnace. One is a magnetic iron oxide of
black, opaque spheres, and elongated, rough particles with sharp jagged edges,
all generally coated with red iron oxide. The second component comprises
transparent, rounded particles of red iron oxide, usually less than 2 microns
in dimension. They occur free or in simple agglomerates. (11)
These
*In tonnage steelmaking operations, the use of 40 percent scrap and 60 percent hot metal is common. Less common is the
"cold-melt" practice that uses no hot metal.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
G-55
Heated fuel oil,

and steam,
400 pounds
CD
@
@
Scrap yard,
60 pounds
@
FIGURE C-35. OPEN-HEARTH FURNACE OPERATING WITH A COLD-METAL CHARGE CONSISTING OF
30 PERCENT PIG IRON AND 70 PERCENT STEEL SCRAP (ORE PRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-56
Heated fuel oil,

and steam,

326 pounds

-------
Heated fuel oil
and steam,
275 pounds
C-57
Rolling
operation,
2.000 pounds
\.
CD
@
Scrap yard
@
@
FIGURE C-37. OPEN-HEARTH FURNACE OPERATING WITH HOT-METAL PRACTICE CONSISTING OF
50 PERCENT HOT METAL AND 50 PERCENT STEEL SCRAP (ORE pRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-58
Heated fuel oil

and steam.
258 ounds
Q)
@
@
@
FIGURE C-38. OPEN-HEARTH FURNACE OPERATING WITH HOT-METAL PRACTICE CONSISTING OF
60 PERCENT HOT METAL AND 40 PERCENT STEEL SCRAP (ORE PRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-59
Heated fuel oil
and steam,
267 pounds
Steel scrap,
647 pounds
CD

-------
C-60
Heated fuel oil
and steam,
183 ounds
Oxygen,
55 pounds
CD
@
@
@)
FIGURE C-40. OPEN-HEARTH FURNACE OPERATING WITH HOT-METAL PRACTICE CONSISTING OF
50 PERCENT HOT-METAL AND 50 PERCENT STEEL SCRAP (OXYGEN PRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-61
Heated fuel oil
and steam,
167 pounds
Steel scrap,
907 pounds
CD
@
@
@)
FIGURE C-41. OPEN-HEARTH FURNACE OPERATING WITH HOT-METAL PRACTICE CONSISTING OF
60 PERCENT HOT METAL AND 40 PERCENT STEEL SCRAP (OXYGEN PRACTICE)
BATTELLE MEMORIAL INSTITUTE';" COLUMBUS LABORATORIES

-------
C-62
o Open-hearth dust - Hot metal to lime-up. Three components make up the
dust from this period of open-hearth operation. (1) Loose agglomerates of
tiny transparent grains usually less than 1 micron in dimension. It is a
hydrated iron oxide such as HFe02. Individual grains and agglomerates are
yellow under top light. (2) Tiny, rounded, transparent, red grains of iron
oxide usually less than 1 micron in dimension. (3) Opaque, black spheres
and rounded particles of magnetic iron oxide. Some particles are covered
the hydrated iron oxide and/ or the red iron oxide. (11)
o Open hearth dust - tap to charge. The same as Item CD with the addition of
black, opaque, frothy, rounded particles of coke. (11)
@ Particulates in combustion product. About 85 percent of the material is
transparent, deep red, rounded grains, of iron oxide, usually less than
1 micron in dimension. The remaining 15 percent is black, opaque spheres
3 to 5 microns in dimension of magnetic iron oxide. The smaller grains
are orange in top light and tend to form simple agglomerates or loose
lumps.
All lime dust does not occur as a visually apparent particulate, it is present
in open hearth dust in very small quantities as shown by chemical analyses.
Sulfur in the form of sulphates also occurs in open hearth dust, but informa-
tion is not available in the published literature on visual characteristics.
No data have been located in the United States literature on the composition of the
products of combustion, but these do contain sulfur compounds originating from the
sulfur contained in the open-hearth fuels (such as oil, tar, and coke-oven gas). It has
been reported that the type of fuel (differentiating between tar and oil) has an effect on
the rate of dust emission, with average dust loadings being higher for tar and lower for
oil. (71) Results from an investigation conducted in Germany for similar fuel combi-
nations are shown in Figure C-42. (72) These data confirm results reported in the
United States. (71) Data from the same German investigators on the amounts of carbon
dioxide and sulfur dioxide in the waste gas during various stages of an oxygen-lanced
heat are shown in Figure C-43.
The amount of dust generated during the open-hearth steelmaking process varies
according to the different stages of the process (see Figure C-42) and according to the
practice. Oxygen lancing produces more particulate emission than open-hearth
practice without lancing. Pertinent information on the amounts of dust generated during
various stages is given in Table C-13. (70, 73, 74, 75)
The dust loadings for the various stages of the process vary over quite a range of
values. For example, the difference between the first two examples in Table C-13 is
quite large. However, the differences can be attributed to the time at which oxygen was
introduced into the process and the length of time it was used. In the first example,
oxygen was not used until after the hot metal had been added. In the second example,
the practice was to use oxygen from the time scrap was charged into the furnace until
the required levels of carbon in the steel were reached in the refining period. Similar
information on an 80-ton open-hearth operation in Germany, using a tar-oil fuel mixture
in combination with oxygen lancing, is shown in Figure C-44.
Typical dust loadings per net ton of raw steel are estimated as about 20 to 22
pounds for open hearths operating with oxygen lancing. (49,75) German literature

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
0.30
-
g 0.25
..-
<.>
:g 0.20
<.>
'0
~
o 0.15
'0
c
o
-
III
~
~ 0.10
III 0.09
c
'0 0.08
~
01 0.07

.~ 0.06
'0
g 0.05
...J
-
~ 0.04
o
FIGURE C-42.
C-63
40% Crude tar +
60 % Coke oven gas

Fuel oil
40% Fuel oil +
60% Coke oven
gas
!--Charging+-Melt - Down~-+ Lime-Ore Boil+ Refining-1
Time --
EFFECT OF DIFFERENT FUELS ON DUST LOADINGS FROM AN
80-TON OPEN-HEARTH FURNACE OPERATING WITHOUT
OXYGEN INJECTION (GERMAN PRACTICE)
-
111°
c.E 0 016
(!) u .
~:Q
~ ~ 0.012
3:~
cQ)
.- a. 0.008
NIII
OC
U)'o
~ 0.004
0.000
o
FIGURE C-43.
,,'
- - ..," \
,
\
\
CO2 ,,---
........... /'
"t, r, J
I \ 1\ I
, \ "
I \ , , I
\ I
, \ I 'I
, ,I \1
, 'J \I
(.1. "
,,,
5 III
o
(!)
4~
III
o
3:
3 c
2 3
Time, hours
N
20
u
4
5
SULFUR DIOXIDE AND CARBON DIOXIDE CONTENTS OF AN
OXYGEN-LANCED.OPEN HEARTH FIRED WITH A TAR-OIL
FUEL (GERMAN PRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-64
TABLE C-13. DUST GENERATION IN OPEN-HEARTH STEELMAKING
---------------.----------.
-------"- - --------_._-------------
------------------------------
---------------
Plant
Practice
Year
of

Data
Dust Loading, grains per standard cu. fr.
Charge to Hot Metal Lime-Up ';ap to
Hot Metal to Lime-Up to Tap Charge
--_._-
"--------------.--------.---.----.-----
U. S. Steel Corp.
Edgar Thomson Works

National Steel Corp.
Weinon Steel Div.
Oxygen 1959 0.35 0.45 0.82 0.87
Oxygen 1965 0.78 1.90 2.70 0.21
Oxygen 1967 0.25 0.65 1.61 n.a
No oxygen 1963 0.56 0.61 0.18 0.11
Youngstown Sheet & Tube
Co.. Indiana Harbor Works

Steel Company of Canada
Hilton Works
----.-------.
----- ----------.------.- ---'----'------'---.-
'---'-'--------------'---
----------------------
Note: n. a. - data not available.
'0 5.0
~o
0' '+-
C
:u.~ 4.0
c.c
0:;)
-.Ju
- ~ 3.0
en 0)
~a.
o
~ 2.0
c
~
0' 1.0
lancing --+Tar-oil firing~
6
Sulfur as
sulphate
~
0)
5;
.s=.
a.-
4 ~ c
(j)0)
u
en ~
C 0)
a.
3--
en.s=.
~O'
0.-
2 .~ ~
3
Time, hours

FIGURE C-44. DUST LOADING DURING OXYGEN-LANCED OPEN-HEARTH
PRACTICE WITH A TAR-OIL FUEL 1,fiXTURE (GERMAN
PRACTICE)
contains a reported value 'of 11 pounds per net ton of raw steel(72), but here again, the
time at which oxygen is introduced into the process (and its duration) has an effect on
the amount of dust generated. An open hearth operating without oxygen injection had a
reported value of 7.95 pounds of dust per net ton of raw steel. (73)
00
~
I~
~
(j)
o
Data on the size distribution of particulate emissions from open-hearth furnaces is
almost nonexistent; information having been located from only one published source in
the literature. (70) Figure C-45 shows the size distribution of particulate emissions
from the No.4 Furnace of the Edgar Thomson Works of the United States Steel Corpora-
tion during the lime boil, as well as a composite sample. It should be noted that the dust
particles for the lime-boil sample average smaller in size than the composite sample.
About 77 percent of the particles in the lime -boil sample were smaller than 5 microns,
while the composite sample had only 46 percent of its particles smaller than 5 microns.
Chemical compositions of open-hearth particulate emissions are listed in
Table C-l4. (70,71,72,73,76,77) .
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
--
c: 25
Q)
u
~ 20
a..
-- /5
L:.
01
'(ii 10
3:
FIGURE C-45.
C-65
35
30
Composite sample
5
o
1-11-2 1-5 1-101-15 1-201-251-301-401
+2 +5 +10 +15 +20 +25 +30

Particle Size, microns
SIZE DISTRIBUTION OF OPEN-HEARTH PARTICULATE EMISSIONS
. FOR OPERATION WITH OXYGEN LANCING (D. S. PRACTICE)
TABLE C-14. CHEMICAL COMPOSITIONS OF OPEN-HEARTH PARTICULATE EMISSIONS,
OXYGEN LANCING, WEIGHT PERCENT
El em em       
or  U. S. Steel Corp. Steel Co. of Canada United Kingdom Germany  
Compound Edgar Thomson Homestead Hilton Works United Steel Co. Dillingen U. S. Plant
Fe203 89.07 88.70 n.a. 88.5 79.65 n.a.
FeO  1. 87 3.17 n.a. 2.2 0.31 n.a.
Total Fe 63.70 n:a. 63. 5 - 68. 0 Q.a. 55.90 59.40
Si02  0.89 0.92 1.16 - 1. 56 0.4 0.47 2.00
A1203 0.52 0.67 0.15 - 0.44 0.4 0.52 0.48
CaO  0.85 1. 06 0.68 - 1. 06 0.9 0.88 1. 85
MgO  n.a. 0.39 0.32 - 0.44 1.5 1. 86 1.12
MnO  0.63 0.61 n.a. n.a. 0.61 n.a.
Mn  n.a. n.a. 0.43 - 0.55 n. a. n.a. 0.28
CuO  n.a. 0.14 n.a. n.a. n.a. n.a.
Cu  n.a. n.a. 0.11 - 0.16 n.a. n.a. 0.08
ZnO  n.a. 0.72 0.26-2.04 n.a. n.a. n.a.
Zn  1. 70 n.a. n.a. n.a. n.a. 0 - 3.0
PbO  n.a. n.a. n.a. n.a. n.a. n.a.
Pb  0.50 n.a. 0.05 - 0.95 n.a. n.a. n.a.
Sn02  n.a. n.a. n.a. n.a. n.a. n;a.
Cr  n.a. n.a. 0.06 - 0.11 n.a. n. a. n.a.
Ni  n.a. n.a. 0.03 - 0.05 n.a. n. a. 0.07
P205  0.47 1. 18 n.a. n.a. 1. 52 n.a.
P  n.a. n.a. 0.06-'0.12 0.3 n.a. 0.15
S  0.40 0.92 0.34-0.70 1.4 2.69 2.78
Alkalies 1. 41 n.a. 0.56-1.71 n.a. 2.72 2.88
Note: n.a. - data not available.      
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-66
Open-Hearth Emission-Control Equipment. Open-hearth furnace installations
operated without the use of oxygen lances were in themselves fairly efficient dust coHec-
tors because of the auxiliary units needed to achieve efficient operation. These
auxiliary units are the slag pockets, checker chambers, and flues to the waste -heat
boilers. Dust-laden gases pass out of the open-hearth chamber through the downtakes
into the slag pockets, where there is a considerable decrease in the velocity of the gases
and a change in direction. This results ina deposition. of large particles in the slag
pockets. The waste gases then pass on to th~ checker chambers (the heat regenerators) .
where, because of additional changes in direction, additional deposition of dust particles
takes place. This is followed by further changes in direction as the gases pass into the
flues that conduct them to the waste heat boilers, where still more dust settles out of
the gases. However, the advent of oxygen la~cing in open-hearth furnaces resulted in a
need to install dust collectors to eliminate the large volumes of fume generated during
the lancing operation. .
The first installation of an electrostatic precipitator on an open hearth was made by
the Kaiser Steel Corporation in.the late 1940's. (78) This was a commercially available
unit which had not been designed specifically for use with open hearths. Many design
and operating problems were encountered and resolved as a result of this installation.
The first electrostatic precipitators designed specifically for open-hearth use were
installed in 1953 at the Fairless Works, United States Steel Corporation and by Kaiser
Steel at Fontana, California. (79,80)
Electrostatic precipitators have been the principal choic'e for emission control on
open hearths. However, the use of venturi scrubbers and bag houses has also been
investigated for the collection of open-hearth particulate emissions. In 1955, pilot-
plant work was started on the application of venturi scrubbers to open hearths, and
resulted in 1959 in the installation of a full-sized scrubber on the No.4 open hearth of
of the Edgar Thomson Works, United States Steel Corporation. (70) An experimental
bag-house program was initiated in 1960 at the Lackawanna Plant, Bethlehem Steel Cor-
poration, and resulted in the installation of full-sized equipment at the Sparrows Point
plant in 1963. (81, 82) Flow diagrams of dust-cleaning systems for open hearths ar~
illustrated in Figures C-46 and C-47. (70, 71, 75, 82-85)
It should be emphasized that the slag pockets, checker chambers, and waste-heat
boilers also act as part of the dust-collection system and are common to all systems in
open hearths. The waste -heat boiler and heat exchanger are shown in the flow sheet
(Figure C-47) with the bag house at the Sparrows Point plant because the heat exchanger
is required to assure that the gases entering the bag house are below the maximum
service temperature of the bag fabric.
Basic Oxygen (BOF) Furnaces
Basic oxygen furnaces are becoming the principal means of making steel in the
United States, and are expected by 1969 to be producing well over half of American
steeL The trends in installed BOF capacity and actual production are shown in Fig-
ure C-48. (86, 87) The lag between rated installed capacity and actual production from
1961 through 1965 can be attributed to the learning that must take place when new tech-
nology involving large, massive equipme!lt is brought on stream, and also to a lag in
additional sales necessary to absorb the new extra'steelmaking capacity. The lag be-
tween rated installed capacity and actual production decreased from 1965 through 1967,
as can be expected when operators become more knowledgable in a new technology. If

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-67
Six
electrostatic
precipitators
Four
electrostatic
U. S. Steel Corp.
Fairless, Pa. - 1953
Bethlehem Steel Corp. United States Steel Corp.
Sparrows Pt.. Md. - 1961 Homestead Works - 1963
FIGURE C-46. FLOW DIAGRAMS OF OPEN-HEARTH DUST-COLLECTING SYSTEMS
USING ELECTROSTATIC PRECIPITATORS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
U. S. Steel Gorp.
Edgar Thomson Works- 1959
C-68
Youngstown Sheet & Tube Go.
Indiana Harbor Works - 1963
Republic Steel Gorp.
Buffalo Disuict- 1964
Bethlehem Steel Gorp.
Sparrows Pt.. Md. - 1963
FIGURE C-47. FLOW DIAGRAMS OF OPEN-HEARTH DUST-COLLECTING SYSTEMS USING SCRUBBERS AND BAG HOUSES
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-69
the trends of instal.led capacity and actual production continue to 1970, it may be antici-
pated that by that year BOF furnaces will be producing over 70 million net tons of raw
steel per year. .
80
10
          -- .."
          --. 
         ,  
         ,  
         ,  
   Future    .'  
   planned   '   
   installations  i/   
   I    
      ~~    
  Rated   7 '/     
  installed.      
- -- capacity  IJ      
  I ''/- -Production 
   "..~ I       
  ~        
  "..' "..,        
 ~ .."".        
  I I  I I   I I 
VI 70
c
o
+-
Q) 60
c
-

~ 50
c
~
"E 40
Q)
Q) 30
en
~
fi. 20
o 1958 1960 1962 1964 1966 1968 1970 1972
Year
FIGURE C-48.
TRENDS IN INSTALLED CAPACITY AND RAW -STEEL
PRODUCTION IN BOF FURNACES
Emission Identification. In bas"ic oxygen steelmaking, the need for a large-surface
bath, (such as is required in open-hearth steelmaking) is overcome by forcing a jet of
high-purity oxygen below the surface of the metal under pressure. The jet also provides
violent agitation, and therefore increases the area of the slag-metal interface. The
BOF process is exothermic to the extent that up to 30 percent (or more) of steel scrap
can be melted using as fuel only the carbon and other metalloids dissolved in the metal.
No conventional fuel is added. In terms of emis sions, the sulfur dioxide and unburned
hydrocarbons associated with open hearths are nonexistent with BOF furnaces.
In the initial stage of basic oxygen steelmaking, the charging of carbon- saturated
hot metal upon cold scrap results in a release of kish as the molten iron is rapidly
cooled. Only a part of this kish is contained by the furnace vessel. The initiation of
oxygen blowing is marked briefly by a heavy dark-brown smoke' (caused by the direct
burning of iron) which persists until the metalloids begin to oxidize and refining begins.
J.ne first elements burned are silicon, manganese, and phosphorus; their. oxides enter
the furnace slag, but absorption is imperfect, and some white silica and lime fume with
minor amounts of manganese enter the fume-exhaust system.
As most of the metalloids .become oxidized, the oxidation of carbon increases in
rate to consume the rest of the oxygen blown, and the volume of gas leaving the furnace
mouth increases noticeably. An. excess of air often is permitted to mix with the exhaust
gases as they pass into the fume-exhaust system. This is taken as a safety precaution
to prevent the existence of a high carbon monoxide content in the flue system and elimi-
nate a possible explosion hazard. An analysis of operating data relating the amount of

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-70
oxygen blown to the volume of exhaust gases (C\.t temperature of combustion) generated
has shown that the volume of exhaust gases is about 25 times greater than the volume of
oxygen blown, This relationship is shown in Figure C-49. (88)
-
Q)'
Q)
-
u
:0 16
~
u
g 14
o
~ 12
c::
~ 10
CD
c:: 8
Q)
C\
~ 6
o
- 4
o
Q) 
E 2
~ 
0 
> 0
 o
FIGURE C-49.
18
Vessel tonnage,
nominal rating

. '"
165T
25 SCFM gas = I CFM02
50
100 150 200 250 300 350 400 450
Exhaust Gas Volume,IOOO cubic feet
RELATIONSHIP BETWEEN THE VOLUME OF OXYGEN
BLOWN AND VOLUME OF EXHAUST GASES
During the oxidation of carbon, the fuming appears to be limited to iron dust either
from iron vaporized from the bath or as iron droplets ejected by the carbon monoxide
rising from the molten bath. Factors that determine the amount of fumes generated
during the blowing process include the type of oxygen lance used, the velocity of the
oxygen, the carbon content of the iron, and the temperature of the iron. An effect of
the number of holes in an oxygen lance is shown in Figure C-50. (89) The s'ingle-hole
lance shows a higher pE'!ak of gas emission ,than a multiple-hole lance, but the total
amount of gas evolved is about the same for both designs. The s,ingle-hole lance is no
>.
-
+-
c:
C
:::J
a
If)
c
(!)

Q)
+-
If)
C
~
FIGURE C-50.
"',
i \
i \
j \ Single-hole lance
-- I, ,',
I , '
8' I -,
: \. .L' -----..l....Multiple-holelance
: "', ...-- \ \
'I ' ,
., \ ,
! i ~
,/ "',
10 15 20 25
Blowing Time. minutes
30
35
EFFECT OF NUMBER OF HOLES IN OXYGEN LANCE
ON EMISSIONS DURING OXYGEN BLOWING
BATTELLE MEMORIAL .INSTITUTE - COLUMBUS LABORATORIES

-------
C-71
longer used, because the multiple-hole lances provide other better operating charac-
teristic s in addition to having lower peaking volumes than the single-hole lances.
The effect of velocity on the amount of emissions obtained from laboratory experi-'
ments is illustrated in Figure C-51. (90) The effects of increasing the velocity of the
oxygen stream are quite pronounced, but at the higher velocities, noise becomes a prob-
lem in the vicinity of the ve sse 1.
c
- 1000
~
o
5 10 50 100 5001000 5000
Calculated Oxygen Velocity, feet per second
8000
0'
E 7000

-t:i 6000
QJ
>
05
>
w 400
QJ
-a 3000
E
C/) 2000
Blown with oxygen
at 0.14 cubic foot
per minute
"'C
c
~
~u.
00
;>.10
:-=~
~-
Q) C
>Ic
Q)IQ)
- CI
~I~
.)( 10
°lc
~.-
0.1.
'$....
I
FIGURE C-51.
EFFECT OF VELOCITY ON EMISSION DURING OXYGEN
BLOWING OF BOF FURNACE (LABORA TORY RESULTS)
The effect of carbon content in the iron on the amount of emissions generated is
illustrated in Figure C-52 from two independent laboratory investigations. (90,91)
 10    
()     
Q)     
en     
...     
Q)     
a. 8    
CI     
E     
Q) 6    
-    
C     
a::     
c     
,2 4    
:;    
(5     
>     
w     
Q) 2    
.JC    
0     
E     
(/)     
 00 1 2 3 4
  Carbon, percent 
:t 1.5
~
.c
Q)
:;
c:
'E
,
~ 1.0
...
Q)
a.
,
,
,
t
,
.
I
I

~
en
c:
o
...
~
c:
.2 0.5
-
~
o
>
W
Q)
.JC
o
E
(/)
o
A
o
o
2
Car bon I percent
6
FIGURE C-52.
EFFECT OF CARBON CONTENT ON EMISSIONS FROM
BOF FURNACE (LABORATORY RESULTS)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-72
The effect of metal temperature is shown in Figure C'- 53, again from labo rator)"
investigations. (90)
 12     
u      
Q)      
en 10 .    
.....    
C'      
E      
Q)      
- 8     
c     
a:::      
c::  .,    
0      
- 6     
~     
0      
>      
w      
Q) 4     
...:      '
o      
E      
(f) 2     
      .
 o     
 2300 2400 2500 2600 2700 2800 .2900 . 3000
   Temper'ature I F 
FIGURE C-53.
EFFECT OF METAL TEMPERATURE ON GENERATION OF
EMISSIONS FROM BOF FURNACE (LABORATORY RESULTS)
From the effects of the various factors illustrated in Figures C-49 to C-53, it is
evident that the amount of evolution of emissions. from a BOF furnace is dependent upon
the interaction of numerous factors. These factors are controlled mainly in such a way
as to make steel most effectively and economically and are controlled secondarily to
inhibit generation of emi s sions. .
The predominant particulate emission is brown iron oXide, and the only gas of
concern is carbon monoxide. A predominance of submicron sizes in the oxide dust
makes it especially difficult to trap and collect: The calculated mean diameter for dust
obtained from specific surface measurements was reported to range from l..s to 2.9
microns. (92) More recent work using electron-microscope counting techniques has
indicated a particle-size distribution with a count median diameter of 0.012 micron. (93)
If galvanized scrap is used as part of the scrap charge to the basic oxygen furnace, the
fume will contain 7.inc ferrites from vaporization and oxidation of the zinc. In some
plants, zinc oxide may comprise from 5 to 8 percent of'the BOF dust and render it
useless for recycling to the blast furnace via the sinte~ plant. (Zinc is a recognized
destructive agent of blast-furnace refractories.) The problem of handling galvanized
scrap is a considerable one for companies manufacturing large tonnages of galvanized
product. If open hearths are available, it is considered preferable to divert galvanized
scrap to the open hearth, rather than use it in basi~ oxyg~n fu~naces.
Many systems have been designed to handle the carbon monoxide generated. The
newer ones seem to emphasize minimizing the amount of aspirated air because of prob-
1ems of overheating of tubes and other parts of the hood and duct work. Many basic
oxygen furnaces are equipped with waste-heat boilers; a few try to recover chemical heat
from the carbon monoxide by burning it; some seek only to flare the gas safely to the
atmosphere.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-73
The Kaldo rotary steelmaking process (used by only one plant in the United States)
has the same characteristic emissions as the BOF steelmaking process, except for the
amounts and size of emissions. The carbon monoxide content of the exhaust gases is
much lower from the Kaldo process than from the BOF steelmaking process, and the
size of particulate emissions is reported to be larger. (94) . .
Typical £lowsheets for the BOF steelmaking process and for the Kaldo process are
given in Figures C-54, C-55, and C-56.
The major emissions from BOF furnaces are described below. The circled num-
bers below refer to emission locations indexed with the same numbers in Figures C-54
through C-56.
CD "Kish" - Carbon in the form of graphite is rejected by the molten iron as it
cools during charging into a BOF steelmaking vessel on top of cold steel
scrap. The graphite particles are opaque ,black, sharply' angular flakes
with smooth surfaces. Some are in layered agglomerates, occasionally
showing rounded 120-degree angles, and even forming rounded hexagonal
tablets. Other particles may consist of opaque, black, rather coarse,
fragments of magnetic iron qxide and transparent, deep red, rounded
particles of hematite. Trances of quartz and calcite may also be found
with kish. (11)
Chemistry:
Graphite (C)
Magnetic iron oxide
Hematite (Fe203)
Quartz (Si02)
Calcite (CaC03)
(Fe304)
o Silica fume - Approximately 50 to almost 100 percent silica, often containing
small quantities of iron, manganese, magnesium, and carbon. Color of the
collected material is gray to off-white. Its bulk density is about 10 to
12 pounds per cubic foot. (95)
Chemistry: Si02
Size distribution of silica fume as determined by counting techniques using
1500 X photomicrographs is given in Table C-15.
TABLE C-15.
SIZE DISTRIBUTION OF SILICA FUME
Particle Size,
microns
Particle Larger,
number percent
Particle Larger,
cumulative percent
0.28
0.20
O. 16
0.11
0.085
0.059
0.040
0.022
-0.022
0.40
0.25
2.35
7.00
17.00
21.00
27.00
20.50
4.50
0.40
0.65
3.00
10.00
27.00
48.00
75.00
95.50
100.00
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-74
BASIC OXYGEN FURNACE
@
@
Raw steel,
2, 000 pounds
Slag, ,
298 pounds
Dust, ,
47 pounds
Off gas,
192 pounds
Rolling
operation,
2, 000 pounds
Scrap yard,
78 pounds
(a) Charge scrap plus cooling scrap = 456 pounds
FIGURE C- 54. BASIC OXYGEN FURNACE OPERATING WITH 80 PERCENT
HOT METAL AND 20 PERCENT STEEL SCRAP
BATTELLE- MEMORIAL INSTITUTE":' COLUMBUS LABORATORIES

-------
C-75
Oxygen,
138 ounds
BASIC OXYGEN FURNACE
@
@
Ra w steel,
2,000 pounds
Off gas,
168 pounds
Rolling
operation,
2,000 pounds
Scrap yard,
77 pounds
Dust collector
(a) Charge scrap plus cooling scrap = 678 pounds.
FIGURE C-55. BASIC OXYGEN FURNACE OPERATING WITH 70 PERCENT
HOT METAL AND 30 PERCENT STEEL SCRAP
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-76
Steel Scrap,
988 pounds
Hot metal,
I, 208 pounds
Ferro alloys..
14 pounds
Oxygen,
159 pounds
CD
ROT ARY OXYGEN FURNACE
@,
GY.
Scrap,
70 pounds
.Dust,
22 pounds
Off gas,
181 pounds
Scrap yard,
70 pounds
Dust collector
FIGURE C- 56. ROT ARY OXYGEN FURNACE OPERATING WITH 55 PERCENT
HOT METAL AND 45 PERCENT STEEL SCRAP
BATTELLE MEMORIAL INSTITUTE - COLUMBUS l:ABORATORIES

-------
C-77
Q) Basic oxygen process dust - Tiny (less than 1 micron in dimension) rounded,
transparent particles of red iron oxide, which tend to agglomerate. Shiny
black spheres of magnetite covered with red iron oxide. (11) The particle size
of this dust is discus sed in more detail below.
There is a great discrepancy in the reported size analysis of BOF dust. Reports
have been made of 95 percent less than 1 micron in size(96), as well as 99 percent less
than 0.2 micron in size. (97) These reports show median sizes of 0,45 and 0.065 mic-
rons respectively. A more recent report states a count median diameter of 0.012
micron. (93) The Kaldo proce ss is reported to have a particle distribution in which
6 percent of the material is less than 1 micron in size, which makes the Kaldo dust
larger than the BOF dust. (94) The larger particulate size for the Kaldo process is ex-
plained as a result of agglomeration of finer particles.
Data on grain loadings per cubic foot of off gas are limited to a single source which
reports that the dust concentration ahead of a precipitator varied from 2.02 to 4.96
grains per cubic foot and averaged 3.59 grains per cubic foot. (98) Data on the amount
of dust generated per net ton of raw steel produced by the BOF process varies over a
large range. In 1959, the dust generated was reported to be between 14.5 and 27.4
pounds per net ton. (99) In 1965, results for an operation in Europe were reported to
vary from 19.6 to 46.6 pounds of dust per net ton of steel. (100) In 1968, an average
figure of 40 pounds of dust per net ton of steel was reported for an operation in the
United States. (101) Dust generation for steel made by the Kaldo process was reported to
be 10 pounds per net ton of raw steel. (102)
Chemical compositions of basic oxygen furnace dusts are given in Table
C-16. (12,88,103) The high zinc content in the dust from two plants is due to the use
of galvanized- steel scrap in the BOF charge. A variatiOn in the composition of BOF
dust will exist from plant to plant, depending on the particular compositions of hot
metal and scrap charged. Variations will also exist within any given plant, depending
on the particular type of steel produced in each heat.
Compositions of some off gases from ,BOF and Kaldo steelmaking ve ssels are given
in Table C-17(88, 102, 104), and a log of off-gas composition before combustion with
aspirated air is given in Figure C-57. (104)
BOF Emission-Control Equipment. The first BOF steelmaking furnaces were in-
stalled in 1954 by the McLouth Steel Company at their T renton, Michigan, plant. Air-
pollution-control equipment consisted of a wet washer and disintegrator. (105) A BOF
steelmaking plant placed into operation the same year in Canada by Dominion Foundries
and Steel Ltd. made use of a wet-washing system with venturi scrubbers. (106) The sec-
ond and third BOF plants installed in the United States were at the Jones and Laughlin
Aliquippa Works and at the Fontana plant of Kaiser Steel Corporation, where electro-
static precipitators were used for dust collecting. (98)' The choice of electrostatic pre-
cipitators at Fontana was influenced to some degree by the fact that Southern California
is a water- short area, and the water system at the plant was a recirculating system
that would have required extensive expansion to handle the additional load from a wet-
washing system. Flow diagrams for typical gas-cleaning systems are shown in Fig-
ure C-58 for wet-cleaning systems(107-109) and in Figure C-59 for electrostatic
systems. (110, Ill)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-78
TABLE C-16. CHEMICAL COMPOSITIONS OF BASIC OXYGEN
FURNACE STEELMAKING DUST, WEIGH:T
PERCENT
Element 0 r  BOF Dust from
Compound Typical U. S. Plant s
FeO 1.5 n. a. n. a. n. a.
Fez03 90.0 80.00 n. a. n. a.
Fe n. a. n. a. 56.0 57.68
Mn304 4.4 n. a. n. a. n. a.
Mn n. a. 0.35  I.Z 1. 54
SiOZ 1. Z5 z.oo  1.9 1. Z9
AlZ03 O. Z O. 15  0.4 O. 13
CaO 0.4 5. 10  3. 1 3.59
MgO 0.05 1. 10 n. a. 0.63
S  n. a. O. lZ  0.09 O. lZ
P  n. a. O. 10  O. Z 0.09
PZ05 0.3 n. a. n. a. n. a.
Cu n. a. 0.04  0.03 n. a.
Zn n. a. Trace  1. 93 4.80
Note: n.a. - data not available.     
TAB LE C - 17.
COMPOSITION OF OFF GASES FROM OXYGEN
STEELMAKING PROCESS, VOLUME PERCENT
Gas
BOF Process
Before Combustion After Combustion
With Aspirated With Aspirated
Air Air
Kaldo Process
Before Combustion
With Aspirated
Air
COZ
CO
NZ
5.0-16.0
74.0-90.5
0.7-13.5
0.0-0.3
7Z.9
ZZ.7
3.0-8.0
74.5-78.9
4.4
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
   C-79    
 95        
 90        
 85        
+- 80        
c:         
<1>         
u 75        
...        
<1>         
Q.         
 70        
<1>         
E 20        
:J        
.g 15 ',Carbon dioxide   
 10  """"""""'-     
    ........    
 5  Nitrogen """'- -- -- 
  Oxygen      
 0 ......       
 0 2 4 6 8 10 12 14 16
   Time, minutes   
FIGURE C-57.
OFF-GAS ANALYSIS FROM A 60-TON BASIC OXYGEN
CONVERTER (BEFORE COMBUSTION)
Electric Steelmaking Furnaces
The price of electric energy has declined during the last decade. This and greater
availability of steel scrap have made electric furnaces attractive economically for the
production of plain carbon steels, as well as for the manufacture of alloy and stainless
steels. During the last 10 years, the annual tonnage of steel produced in electric fur-
naces in the United States has doubled. In 1957, electric furnaces accounted for 7 per-
cent of all the steel produced in the United States; this increased to 11 percent in 1967.
Electric-furnace steel production from 1957 through 1967 is shown in Figure C-60. (8)
The trend in recent years has been toward larger and more powerful electric furnaces
in the integrated iron and steel industry. The distribution of furnaces by size, as of
1968, is shown in Figure C-61. (112, 113, 114) Of the total of 195furnaces, 40 percent
have capacities less than 50 tons, 36.4 percent have capacities between 50 and 90 tons,
and the remaining 23.6 percent have capacities greater than 100 tons. Electric furnaces
are combined with continuous casting machines in some of the newer installations.
These range in size from 25 -ton furnaces used in mini steel plants that cast billets to
150-ton furnaces that are used for producing the steel for multi strand continuous cast-
ing machines such as are being installed at Jones and Laughlin's Aliquippa Plant and
National Steel's Great Lakes Division Plant at Ecourse, Michigan.
Emission Identification. Emissions generated during electric -furnace steelmaking
originate from the physical nature of scrap used, the cleanliness of the scrap, the na-
ture of the melting operation, and oxygen lancing. Thin steel scrap will oxidize easily
and result in heavy fuming and a high metal loss during melting in electric -arc furnaces.
For this reason, thin steel scrap is considered generally undesirable by electric -furnace
operators. Dirty scrap is a major source of emissions. Dust emissions as high as
30 pounds per net ton of steel can result from the use of particularly dirty scrap. (115)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
Inland Steel Company
Indiana Harbor, Ind. - 1965
C-80
Bethlehem Steel Corp.
Sparrows Point, Md. - 1965
Wheeling Steel Corp.
Steubenville, O. - 1965
FIGURE C-58. EXAMPLES OF WET-CLEANING SYSTEMS FOR BOF STEELMAKING FURNACES
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
generating
hood
Great Lakes Steel
Ecourse Mich. - 1962
C-81
Wisconsin Steel
South Chicago 111. - 1964
FIGURE C-59. EXAMPLES OF ELECTROSTATIC-PRECIPITATOR GAS-CLEANING SYSTEMS
FOR BOF STEELMAKING FURNACES
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
FIGURE C-60.
FIGURE C-61.
C-82
~ 16
o 
Q.I 
:>. 15
~ 
Q.I 
a. 14
I/)
C
o 13
+-
+- 
Q.I 
C 12
-
o 
I/) II
C
o 
E 10
C 9
o
+- 
U 
:J 
"tJ 
0 
~ 
a.. 
6
1957 1959 .1961 1963 1965 1967
1958 1960 1962 1964 1966 1968
Year
ANNUAL PRODUCTION OF RAW STEEL IN ELECTRIC
FURNACES IN THE UNITED STATES
28
26
24
III 22
~ 20
g 18
~ 16
IL.. 14
15 12
~ 10
.a 8
E
:J 6
Z 4
2
o
10 20 30 40 50 60 70 80 90 150 200 250
15 25 35 45 55 65 7585 100 175 225

Furnace Capacity. net tons
SIZE DISTRIBUTION OF ELECTRIC STEELMAKING FURNACES
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-83
Figure C-62 shows the reported dust emissions (in pounds per net ton of steel) for 22
operations using various combinations of normal scrap (scrap with little to moderate
rust), dirty scrap (scrap with heavy rust), and oxygen landng.(llS-122) It can be seen
that electric-furnace melting with dirty scrap can generate as much dust as an electric-
furnace heat using oxygen lancing. It has been estimated that 20 percent of the emis-
sions are produced during oxygen lancing, with the remaining 80 percent attributed to
the meltdown period. (119) Dust loadings during the oxygen-blowing phase of a 40-ton
electric -furnace heat are shown in Figure C -63. (119)
- 25
Q)
Q)
+-
(I)
.....
o
c: 20
o
+-
+-
Q)
!=
~
~ 15
(I)
"'0
c:
:::J
o
a. 10
(I)
c:
o
(I)
(I)
E
w
+-
(I)
:::J
o
FIGURE C-62.
FIGURE C-63.
3
100
. Normal scrap- with oxygen lanCing
o Normal scrap- no oxygen lancing
o Dirty scrap - no oxygen lancing
.
o
5
....6
. "
....'
....""
Q.""
....""
....
....
....
....
....

. . ....~"

. .....
. . "
"
0'
...,0"
-...0'"
.
.

o
o
o
o
o
00
30
10 15 20 25
Melting Rate, net tons per hour
DUST EMISSIONS DURING ELECTRIC-FURNACE
MELTING OF STEEL
4
~ 3
~ -
-
-
-
-
-
--
01 U
c: .-
.- .Q
"0 ::J
C u 2
o ...
...I Q)
Co
-
-
-

-------
C-84
Flow diagrams for electric-furnace melting are shown in Figures C-64, C-65,
and C -66. All practices shown are for cold-melt practice. Hot metal (as a part of the
electric-furnace charge) has been used on occasion in the making of electric-furnace
steel(l23, 124), but this is not a routine type of operation and is used mainly when short-
duration scrap shortages may occur in an integrated iron and steel plant that has access
to hot metal from a blast furnace. Depending upon the practice and raw materials used,
principal emissions from electric-arc steelmaking furnaces are dust from the furnace
itself, dust from scrap preheaters, and furnace off gas. These are discussed separately
below using circled numbers to key the descriptions to location of emission in
Figures C-64 through C-66.
CD Electric-furnace dust - Opaque, rounded grains that are peach to reddish in
color in top light. Small agglomerates are present, but are not common. (11)
The chemical composition of electric -furnace dusts will be influenced by the com-
position of the steel being melted. Because of this, optical characteristics of the dust
may also vary because of the different alloying-element oxides that may be present.
Because electric furnaces are used for melting a wide range of alloy and stainless steels,
the chemical composition of any particular dust will reflect the composition of the alloy
melted. Some reported compositions of dust emis sions are given in
Table C-18. (125-128)
TAB LE C - 18 .
CHEMICAL COMPOSITIONS OF ELECTRIC-FURNACE DUSTS
,
WEIGHT PERCENT
Element or   Sample Designation  
Compound A B C D E F
FeO 4.2 n.a. n.a. n.a. n. a. 4 - 10
Fe203 35.04 50.55 52.62 52.05 50.05 19 - 44
Cr203 0.00 0.56 0.00 0.15 13.87 0 - 12
MnO 12. 10 12.22 5.34 1.29-2.58 n. a. 3 - 12
NiO 0.30 n.a. tr tr 3. 18 0 - 3
PbO n.a. n. a. 3.47 0.81-1.08 n. a. 0 - 4
ZnO n.a. n.a. 8.87 1.24-2.48 n. a. 0 - 44
Si02 8.80 5.76 6.78 3.85 5.50 2 - 9
A1203 12.90 5.85 2.55 14.61 n. a. 1 - 13
CaO 14.90 2.60 6.72 1.40-4.20 9.80 5 - 22
MgO 7.90 7.78 3.49 1.66-4.98 6.64 2 - 15
S 0.26 tr 0.59 n.a. n. a. 0 - 1
P 0.10 0.28 n. a. n.a. n.a. 0 - 1
C 2.30 n.a. n. a. n. a. n. a. 2 - 4
Alkalies 1.20 4.76 n.a. n.a. 2.50 1 - 11
Note: n. a. - not available, tr - trace
Sample A - Single 20-ton furnace. Plant specializing in tool and die steels.
Sample B - Representative sample from plant with four 75-ton and two 200-ton furnaces producing
low-alloy and stainless steels.
Sample C - Single IOO-ton furnace producing low-alloy steels for plate.
Sample D - Single IOO-ton furnace producing low-alloy steels for plate.
Sample E - Single 70-ton furnace producing stainless steel.
Sample F - Representative samples from multiple-furnace shop. Furnaces vary in size from 4 to 200-ton, producing
low-alloy and stainless steels.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-85
ELECTRIC FURNACE
@
CD
Rolling
operation
FIGURE C-64. EXAMPLE OF ELECTRIC-FURNACE STEELMAKING USING A CHARGE OF
COLD STEEL SCRAP (ORE PRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS' LABORATORIES

-------
C-86
Atmosphere
ELECTRIC FURNACE
@

-------
C-87
Atmosphere
FIGURE C-66. EXAMPLE OF ELECTRIC-FURNACE STEELMAKING USING A CHARGE OF COLD STEEL SCRAP
(OXY-FUEL BURNERS FOR MELTDOWN; OXYGEN PRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-88
The high contents of zinc and lead oxide in some of the above analyses can probably be
attributed to the use of galvanized scrap and scrap containing lead (terne plate). The
change s in composition of electric -furnace dust during various periods of making a
particular heat of low-alloy steel are given in Table C-19. (129)
TABLE C-19.
CHANGES IN COMPOSITION OF ELECTRIC-FURNACE
DUST DURING A SINGLE HEAT
   Composition, weight pe rcent  
Period Fe203 Cr203 MnO Si02 CaO MgO Al203 P205 S02
Melting 56.75 1. 32 10.15 9.77 3.39 0.46 0.31 0.60 2.08
Ore oxidation 66.00 1. 32 5.81 o. 76 6.30 0.67 0.17 0.59 6.00
Oxygen lancing 65.37 0.86 9. 17 2.42 3. 10 1.83 O. 14 o. 76 1.84
Refining 26.60 0.53 6.70 Tr 35.22 2.72 0.45 0.55 7.55
Tr - trace.         
The size of particulate emissions from electric -furnace melting in the United
Kingdom has been reported to be 30 percent by weight below 10 microns for furnaces
operating without oxygen lancing, and 40 percent by weight below 3 microns for furnaces
operating with oxygen lancing. (130) Size distributions for some electric steelmaking
furnaces operating in the United States are given in Table C-20. (122)
TABLE C-20. SIZE DISTRIBUTION OF PARTICULATE
  EMISSIONS FROM ELECTRIC STEEL-
  MAKING FURNACES, PERCENT 
Particle Size,  Electric Furnace Size, net tons 
microns 3.5 4 14 17 19 50
o to 5 57.2 63.3 59.0 72.0 43.3 71.9
5 to 1 0 37.8 17.7 33. 1 10.5 17.7 8.3
10 to 20 3.4 8.0 4.9 2. 7 6.4 6.0
20 to 40 1.6 8. 1 3.0 4.7 14.6 7.5
greater than 40 0.0 2.9 o. 0 10. 1 18.0 6.3
o Dust from scrap preheaters - Information on the physical and optical
characteristics of dust generated during the preheating of scrap has
not been located. However, it can be as sumed that the composition of
the dust will be influenced mostly by the cleanliness of the scrap, its
content of volatile matter, and presence of surface coatings on some
of the steel.
Information on dust generated during scrap preheating is not available for el~ctric-
steelmaking furnaces in the United States. Investigations carried out in Norway have
resulted in data relative to dust emissions during scrap preheating. (131) Work done
with heat sizes varying from 6.6 to 26.4 net tons has shown that particulate emissions
varied from 0.040 to O. 111 grains per cubic foot, which is equivalent to 0.07 to 0.27
pound per net ton of scrap charged. Although dust generation varied considerably, no

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-89
correlation was found between dust loading and any of the operating variables. The
variations in grain loadings are shown in Figure C-67, and the size distribution of the
particulate emis sions is shown in Figure C - 68.
0.12
(; 0.10
- 0
0'-
.!:: u 0.08
"'C .-
0..0
o ~
....J u 0.06
~
+-Q)
enQ.
c5 en 0.04
c
o
~0.02
--,
I
~-----..,
L_-
---
S5

-H3 Test I
Numbe~
H2
S4
-----
0.000
40
FIGURE C-67.
DUST LOADINGS DURING PREHEATING OF SCRAP IN NORWA Y
24

22

"'C 20
Q)
.: 18
o
Q) 16
a:::
+- 14
c
~ 12
~
~ 10
+-
.J::. 8
0'
Q) 6
~
4
2
o
850 200 100
400 150 75
Size, microns
I 40 I 100 I 200 I 325 I
20 70 140 270 -325
Equivalent U.S. Series
55 -45
45
FIGURE C-68.
SIZE DISTRIBUTION OF DUST PARTICULATES
DURING SCRAP PREHEATING IN NORWAY
According to this information, while the dust generated from melting in an electric
steelmaking furnace varies from 82 to 100 percent less than 40 microns in size, 98 per-
cent of the dust generated during preheating of scrap is larger than 45 microns in size.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
+-
C 80
Q)
u
~
Q) 70
c.

E 60
::J
g 50
+-
~ 40
+-
C
<3 30
(/)
o
(!)
C-90
100
Ore addition
Start of boil ~
~, Slag Off. ~ ~ ~
'0- -q ~
, , ,
A 1\
N2 \ ,\ if
, \ /,
\ ,. 'r/'
\' ,
\' ,
\ ,
9--...,
I ........
I ........ _0
I '0.. -(Y'" - -
I N2
I
I
J
Oxygen lancing
90
20
10
o
o
2
3
Time, hours
4
5
FIGURE C-69.
"U
Q) Q) 800
>+-
-::J
°c
>.-
wE
Q)
"U~
.- Q)
xc.
o
g (U 400
~Q)
-
Cu
.8:0 200
~ ::J
Ou
U
FIGURE C-70.
GASES GENERATED DURING THE PRODUCTION
OF A BALL- BEARING STEEL IN A 22- TON
ELECTRIC FURNACE
1000
600
16
CARBON MONOXIDE EVOLUTION DURING OXYGEN LANCING
IN A 16. 5-TON ELECTRIC STEELMAKING FURNACE
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-9l
o Electric -furnace off gas - The off gas is made up mainly of carbon dioxide,
carbon monoxide, oxygen, and nitrogen. The composition of the off gas
varies somewhat with the slag practice used, the stage of the heat, and the
use of oxygen for lancing. The change in composition of the gas during the
course of one heat is illustrated in Figure C-69. (129) Carbon dioxide stayed
below 15 percent during the entire course of the heat, while the carbon
monoxide content varied inversely with the nitrogen content. The carbon
monoxide content of the off gas during oxygen lancing went as high as
85 percent.
Carbon monoxide evolution during the oxygen-lancing period is shown in Fig-
ure C-70 for a 16. 5-ton electric furnace producing high-chromium steels. (132) For
the same furnace, it also was possible to relate the evolution of particulate emissions
with that of carbon monoxide, as shown in Figure C-71. (132)
~ 5.0
::J
c:
E
~
~ 4.0
tJ)
"'0
c:
::J
~ 3.0
E
w
100 200 300 400 500 600
Carbon Monoxide, cubic feet per minute
700
FIGURE C-71.
RELATIONSHIP BETWEEN THE EVOLUTION OF DUST
PARTICULATES AND CARBON MONOXIDE IN A 16.5-
TON ELECTRIC FURNACE
Electric-Furnace Emission-Control Equipment. Control of emissions from
electric steelmaking furnaces is affected by the design of the furnace. Two designs of
electric-arc furnaces used in the integrated iron and steel industry are (1) the movable-
roof top-charging furnace, and (2) the fixed-roof door-charging furnace. The top-
charging furnace has an advantage of permitting more rapid charging, while the door-
charging furnace permits somewhat better control of melting practice and improved
refractory life because of an essentially closed system and lower impact on the furnace
hearth during charging.
Emissions from electric -arc steelmaking furnaces are controlled and collected
by three main types of systems: (1) collection of emissions by the use of hoods over and
around the furnace at points of emission, (2) direct extraction from the furnace interior,
and (3) shop-roof extraction and collection. Emissions leave an electric furnace around
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
L---
C-92
the electrode ports in the roof of the furnace, the tapping spout, slagging door, and in
the case of top-charged furnaces, through the open furnace top during charging.
Hooded Collection. A hooded collection system consists of close-fitting hoods at
the points of emission to collect the particulate and gaseous emissions and carry them
via a duct system to the dust collector. Several hoods are required, and must be
movable in the case of top-charging furnaces. A hooded system tends to obscure
visibility from the crane -operator's cab, and results in added operational hazards.
Direct Extraction. This system consists of an exhaust opening located in the roof
of the furnace so as to draw off the emissions and direct them to the collector. The
system is operated in such a manner that air flows into the furnace and out the exhaust,
thus minimizing the discharge of emissions through the various doors and electrode
ports. This method has shown a tendency to cause the loss of some alloying elements,
as shown in Figure C-72 for manganese. (133) Difficulties have also been reported in
1:: 90
Q)
~ 80
Q)
c: 70
>-
Qj 60
>
8 50
Q)
a::
Q) 40
en
~ 30
a
:? 20
a
~ 10
o
FIGURE C-72.
0.05 0.10 0.15 0.20
Carbon at Tap, percent
0.25
MANGANESE RECOVERY IN STEEL AS AFFECTED
BY THE METHOD OF EXTRACTION OF ELECTRIC-
FURNACE EMISSIONS
operating with special carbide slags in refining of special steels. (120) Otherwise, there
are no apparent metallurgical difficulties associated with direct-extraction methods. (120)
Advantages of direct extraction with respect to furnace operation have been reported to
include decreased electrode consumption and increased roof life. Electrode consumption
has been reported to decrease 8 percent and roof life to increase 16 percent. (118) An
increase in roof life from 170 heats to 320 heats per roof reline has been reported for
Roanoke Steel on a 20-ton furnace. (134) Direct-extraction systems usually require
lengthy duct work or heat exchangers to cool gases to a safe temperature before they
enter a bag house. It has been determined that a minimum of 600 feet of ducting is
required to cool direct-extracted off gases by radiation before their entry into a bag
house. (135) -
Shop-Roof Extraction Systems. Emission collection by this method consists of
making the shop building itself a large collecting hood for all emissions generated in a
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-93
multiple-furnace shop. Special exhaust ducts and hoods must then be utilized to ensure
proper removal of the emissions from the shop.
Flow diagrams of examples of direct-extraction and furnace- shell extraction sys-
tems with bag-house collectors are given in Figure C-73(l20, 127, 136, 137), dust-
collecting systems using wet scrubbers in Figure C-74(l28, 138,139), and shop-roof
extraction systems in Figure C-75(l34, 140, 141).
A new operating technique that is making an appearance in electric-furnace melt-
ing is the continuous charging of metallized pellets into the electric furnace through the
roof of the furnace. The technique is also possible with fragmentized scrap, which can
conceivably develop into a system for preheating scrap and for exhausting emissions
from an electric furnace.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
m
»
~
~
111
r
r
111

~
111
~
o
::0
»
r
Z
III
~
~
c
~
111
,
o
o
r
c
~
m
c
III.
r
»
m
o
::0
»
~
o
::0
111
III
Lukens Steel Co.
Coaresville, Pa. - 1964
Lukens Steel Co.
Coatesvi1le, Pa. - 1966
U. S. Steel Corp.
Chicago, 111. - 1961
()
I
-..0
~
Bethlehem Steel Corp.
Seattle, Wash. - 1959
FIGURE C-73. EXAMPLES OF DIRECT-EXTRACTION EMISSION-CONTROL SYSTEMS WITH BAG HOUSES

-------
C-95
Armco Steel Corp.
Butler, Pa. - 1959
Armco Steel Corp.
Houston, Texas - 1966
FIGURE C-74. EXAMPLES OF ELECTRIC-FURNACE DUST-COLLECTING SYSTEMS USING WET SCRUBBERS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-96
Ingot
pouring
area
Jones & Laughlin Steel Corp.
Warren, Mich. - 1966
Top-charge
.furnaces,
one 70 ton
two 100 ton
Bethlehem Steel Corp.
Los Angeles, Cal. - 1966
FIGURE C-75. EXAMPLES OF ELECTRIC-FURNACE SHOP-ROOF EMISSION-CONTROL SYSTEMS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-97
Vacuum Degassing
Vacuum degas sing of molten steel (as discussed in Appendix A) was first developed
as a means of removing hydrogen from steel, and later was developed further to use the
carbon- oxygen reaction to deoxidize the molten steel. These technological developments
resulted in cleaner steels with improved properties.
Lmission Identification
The use of vacuum-degassing processes results in off gases that contain hydro-
gen, oxygen, nitrogen, carbon dioxide, carbon monoxide, and methane. Dust is also
generated because of the violent agitation of the metal in the vacuum chamber, and the
combined effects of high vapor pressures of the metallics and low pressure in the
vacuum chamber. The off gases are combustible, as indicated by a report that when
removing 3 parts per million of hydrogen and O. 04 percent carbon at a treating rate of
5 tons of steel per minute, the gases leaving the steam- ejector system when ignited pro-
duced a flame 10 feet high. (142) When the same amounts of hydrogen and carbon are
removed at a treating rate of 30 tons of steel per minute, the exhaust gases have a net
combustion value of about 250,000 Btu per minute. The dust particulates will vary in
composition depending on the alloying elements in the steel and their respective vapor
pressures at the pressures in the vacuum vessels. Flow diagrams for the two general
types of vacuum-degassing processes are shown iri Figures C-76 and C-77.
The major emissions are described below, using circled numbers to key the
descriptions to the locations in Figures C-76 and C-77.
(.0 Gases liberated from the molten steel during degassing are principally
(a) the hydrogen that has been dissolved in the steel, and (b) a carbon
monoxide - carbon dioxide mixture resulting from the reaction between
carbon and oxygen in the steel. Although nitrogen occurs in the off gas,
its source is air in the system or aspirated air. Small amounts of
water, oxygen, methane, and argon may also be present. The compo-
sition of gases released will vary during the course of the degassing
treatment as shown in Figure C-78. (143)
The variation in the composition of off gases during vacuum- stream degassing for
different grades of rotor steel, at variou9 times and levels of vacuum, are given in
Table C-21. (144) ..
BATTELLE MEMORIAL INSTITUTE --COLUMBUS LABORATORIES

-------
C-98
Molten steel
STEAM EJECTOR
@
@
Atmosphere
Dust.
Vacuum-de assed molten steel
LAD LE
INGOT MOLD
FIGURE C-76. VACUUM-STREAM DEGASSING OF MOLTEN STEEL
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-99
Molten steel
VACUUM -DEGASSING CHAMBER

-------
I
I
C-IOO
-
c::
Q)
o
~ 40
a.
70

60' ~\
50 \ I ',\Yd'Ogen

\! '\,

~

30 /\
,
-
c::
~
c::
8
en
o
C>
15
20
25
30
Time I minutes
FIGURE C-78.
VARIATION IN GAS CONTENTS DURING
VACUUM-STREAM DEGASSING
  TABLE C-21. COMPOSITION OF OFF GASES DURING V ACUUM-STREAM DEGASSING OF STEEL  
 Typical Steel Analysis, weight percent  Time, Pressure,  Gas Analysis, mole percent 
C Mn Si Ni Cr Mo V W minutes microns C02 CO H2 02 H20 N2 A
0.24 0.60 0.28 2.75  0.30 0.08  0 80 1.6   12.9 34.9 50.1 0.5
        5 320 1.9 21.7 40.3 1.5 3.5 30.9 0.2
        10 360 0.6 26.0 59.0 0.5 2.3 11.5 O. 1
        15 340 2.5 38.4 31. 4 0.8 1.7 25.0 0.3
        22 1000 1.8 49.7 24.5 0.6 0.6 22.6 0.3
0.38 0.70 0.75 2.80 1. 20 0.50 0.18  0 175 3.2 18.5 1.6 9.0 6.7 68.2 0.6
        5 400 0.7 26.1 62.1 0.1 0.1 10.8 0.1
        10 410 0.4 19.6 67.2 0.2 0.6 11. 9 --
        16 300 2.1 27.0 50.0 4.4 1.8 14.7 --
        24  4.3 38.2 24.0 0.1 0.3 32.6 0.4
        26 590 4.1 33.0 19.6 0.5 1.7 40.5 0.5
0.25 0.88 0.25 0.88 12.00 0.88 0.20 0.88 0 80 0.4   21.1 1.9 75.6 1.0
        5 540 1.7 19.5 65.6 0.0 0.2 12.9 0.1
        14 550 4.6 44.3 37.0 0.0 0.5 13.3 0.2
        16 550 6.6 47.0 26. 1 0.2 1.0 18:8 0.2
0.28 0.88 0.28 0.25 1. 05 1. 25 0.25  0 160 0.3 1.2 -- 20.7 10.4 66.5 0.8
        5 620 2.4 29.8 41.4 0.1 3.6 22.5 0.2
        11 680 1.1 37. 5 41. 5 0.1 1.6 18.0 0.1
        19 600 1.1 48.3 33.7 0.2 1.8 14.9 0.1
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-IOl
Examples of composition during vacuum- stream degassing of ball- bearing steels
are given in Table C-22. (145)
TABLE C-22.
OFF-GAS COMPOSITION DURING VACUUM-STREAM
DEGASSING OF BALL- BEARING ST EELS
 Steel Analyses, weight percent  Gas Analyses, volume percent
C Mn Si Ni Cr Mo V C02 CO H2 02 N2 CH4
0.23 0.75 0.14 0.26 O. 33   2. 7 15.8 29.5  50.7 1.3
O. 32 1. 31 0.14 O. 36 0.37 0.27  1.2 36.4 42. 0 1. 1 15. 6 3. 7
O. 38 1.35 0.22 0.04 0.26 0.26  1. 3 47.9 27.0 1.8 20. 1 1.5
0.40 O. 60 0.86 0.22 4.91 1.40 O. 91 7.4 24. 7 10. 1 O. 7 51. 2 5.9
O. 33 1.34 O. 16 O. 18 O. 37 0.29  0.8 40.4 28. 7 1.3 27.3 1.5
0.32 0.46 0.20 O. 17 .3.58 0.53  1.3 44. 0 19.9  24.9 9. 9
@ The compositions of the off gases emitted to the atmosphere are essen-
tially the same as those of the gases evolved from the steels in the
vacuum-treating units. Some solution of the gases in water may occur
as the gases are exhausted with steam and pas s through intercondensers
and a hot-well before being exhausted to the atmosphere. Analyses made
of the gases as they exhaust to the atmosphere from D-H vacuum-
degassing installations have shown them to have an average composition
of up to 80 percent carbon monoxide, up to 15 percent hydrogen, and up
to 20 percent carbon dioxide. (146) .
G) Metallic dusts - are deposited on the walls of the steel chambers used
@ in stream degassing. Descriptions of the size consist of the dusts
have not been located in the literature. However, the dusts are very
finely divided and are pyrophoric in nature to the extent that safety
precautions must be exercised in opening the tanks after the steel has
been processed. (144, 147)
Chemical composition of one deposited dust has been reported to include the follow-
ing: 2. 1 percent carbon, 1. 1 percent silicon, 0.4 percent aluminum, 78.0 percent
manganese, and 12. 3 percent iron. (144) Analyses of other dusts generated during
vacuum- stream degassing and analyses of the metal in the ladle after treatment are
given in Table C - 23. (147)
TABLE C-23.
DUST AND METAL ANALYSES FOR VACUUM-
TREA TED STEELS
   Elements, weight percent  
Material C Mn Si Ni Cr V Mo Cu Fe
Steel in ladle 0.33 0.73 0.25 .2.86 o. 99 0.22 0.53 O. 17 
Dust 1.66 46.30 1. 63 O. 38 o. 36 0.01 O. 05 1. 60 17.60
Steel in ladle 0.33 0.83 0.26 O. 17 1. 01 0.23 1.21 O. 14 
Dust 1.69 47.70 1.40 O. 13 0.38 O. 04 O. 09 1. 20 15.50
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-102
@ Dusts generated during vacuum degassing are not emitted to the atmos-
phere because of the dust-washing characteristics of the steam ejectors
used to create. the vacuums required. All of the vacuum-degassing
installations used in the integrated iron and steel industry in the United
States use steam ejectors. Detailed reports on the amount of dust gen-
erated during vacuum treatment have not yet been located in the pub-
lished literature. A single report does state that about 10 pounds of
dust are generated during a single cast and collected in the vacuum
system, and because the installation is designed to handle lOO-ton heats,
the generation of dust is about one-tenth of a pound per ton of steel. (148)
Vacuum Degassing Emission-Control Equipment
As has been stated in the preceding discussion, practically no particulate emissions
enter the atmosphere from vacuum-degassing installations. This can probably be attrib-
uted to the fact that the steam ejectors used for creating the necessary vacuum are to a
certain extent venturi scrubbers. A vacuum-degassing installation may have from 4 to
6 ejectors and related intercondensors, depending on the amount of steel to be treated
and the vacuum levels required. A typical steam- ejector system is illustrated in
Figure C-79. (144) An ejector is shown schematically in Figure C-80. (149) Its simi-
larity to a simple venturi scrubber shown in Figure B-15 is apparent.
XH"'un
r VACUATOIt
#. IT[AW-EJECTOR PUMP
II 2 INTIACONDENSIUt
HOT WILL
If J INTEACONOfNIER
E"ICTOIt
11 I 'TE'M ..nCTOR
TO' 0' CHARGINO 'J.9!!.
-.-
AtUUM
TANK
1iiiT'TOiiO,- PIT
FIGURE C -79.
TYPICAL STEAM-EJECTOR SYSTEM FOR VACUUM DEGASSING
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C -1 03
STEAM CHEST
DISCHARGE
FIGURE C~80.
CROSS SECTION OF A STEAM EJECTOR
Gases passing through the steam- ejector system are usually emitted to the atmos-
phere, except where they may be passed through a water hot-well and be a hazard to
working personnel, in which case they are ignited and flared to the atmosphere. (142)
.l ne amount of gases which may be discharged during vacuum degassing is influenced by
the carbon content in the steel and the amount of hydrogen to be removed. The weight
and volume of gases, as equivalent air, for a typical situation are listed in Table
C~24. (142)
TABLE C-24.
EFFECT OF TREATING RATE AND AMOUNT OF CARBON
REMOVED AS CARBON MONOXIDE ON THE GAS LOAD
DURING VACUUM TREATMENT WITH STEAM
EJECTORS(a)
Carbon Removed,

percent
Gas Load, equivalent air per hour
5-tons-per-minute Treating Rate 30-tons-per-minute Treating Rate
Pounds Cubic Feet Pounds Cubic Feet
0,00
0, 01
0, 05
13
156
760
161
1,933
9,416
80
940
4, 360
991
11, 647
54,020
(a) Treating rate is in terms of tons of steel per minute. Gas loading is based on removal of 4 ppm of hydrogen from the steel.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C -1 04
Manufacture of Semifinished Products
Semifinished products, as considered in this study, include ingots, billets,
blooms, and slabs. These are the intermediate shapes that are produced in the manu-
facture of the various end products of steel. However, even the semifinished products
as considered here are a marketable item accounting for 3.3 percent of American steel
shipments in 1967(8). Production of billets, blooms, and slabs from 1958 through 1967
is shown in Figure C-81. The data shown are based on the assumption that tube ~ounds,
light merchant shapes, reinforcing bar, joint bars, tie-plate bars, wire rods, and
forging billets were produced from billets; piling, rails, and heavy structural shapes
from blooms; and hot flat-rolled products and skelp from slabs. The production data
include manufacture by conventional ingot casting and rolling, by continuous casting,
and by pressure casting.
If) 70
c:
o
If)
+-
g 50
"0
o
~
a..
"0 40
Q)
L:.
(/)
:~ 30
......
-
'E 60
E
Q)
(f) 20
......
o
c:
o
---~----
Billets --...... --
. ---------
........---------- .-
10
+-
U
::J
"0
o
I-
a..
Blooms
----- --
-
-
o
1958
1959
1960
1961
1962
1963
Year
1964
1965
1966
1967
1968
FIGURE C-81.
ANNUAL PRODUCTION OF SEMIFINISHED STEEL
PRODUCTS
Ingot Casting and Rolling
The casting of ingots long has been the conventional method for converting molten
steel into a solidified shape suitable for further proces sing into semifinished products.
Many sizes, types, and designs of ingot molds are used in the production of ingots. The
particular type of mold is influenced by the end product to be made and by the chemistry
of the grade of steel cast. Steel is classified into three types: (1) killed steel, (2) semi-
killed steel, and (3) rimmed steel. Schematic illustrations of their respective ingot
structures are given in Figure C-82.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-I05
Hot top
r------l

I I
QQotOlVllvPIIVII",
\~ e' 0 6 Q " . ' 0 '" D'
o 'IIO'O~ ,
0'0:0,.0,1
Do, ~
0\.11'1 ,'"
00,.0',
, 6"00'
/JODO,\)\) ,.
P(J " ""00
"01100 '
di P 01) 0 ,,, ..
, '0 9000
" 000 0 '<:I 0
o 0
0011 00
~ 0
b '"
Po
Gas
holes
Killed Steel Ingot
Semikilled Steel Ingot
FIGURE C-82.
r--------'
I I
I ,
I I
J \
;' .....
:"~,, ;~OD
, 0 II 0
d D q'\> "0 00
, "0,0 I
II Q". 0 '11'00
:0::'09/10
, 0°0 ~
, ,,, 6
o 0 ,
II
o , 1:1
o ,,0 0
II ~
II 0
o 0
o 0
o ,
o ..
o :
o 0
o
o 0
o II
o '
o ,
o 0 0 Q
CI> D 0 't:.
c:::I" fJ ~
;;; Do"" " . ~
<==' C>
c::> ~
<=> "'"
~ 0
-;00 D6 00 00 OOOOOOgOo6lloai
Rimmed Steel Ingot
INGOT STRUCTURES FOR DIFFERENT TYPES OF STEEL
Kill"ed steels are generally used where a homogeneous structure is. required in the
finished product. Alloy steels, forging steels, and carburizing steels are typical ex-
amples of killed steels. Steels with more than 0.30 percent carbon are generally made
as killed steels. The term l'killed'l means that the molten steel has been thoroughly
deoxidized with various elements (frequently aluminum) so that it will be quiet when
poured into a mold. Ceramic hot tops are placed on the top of ingot molds poured with
killed steel. The function of the hot top is to delay solidification of the steel at the top
of the mold so as to supply molten metal to the solidifying ingot. If the hot top is not
used, the shrink cavity will penetrate deep into the body of the mold and will result in
less usable steel from the ingot. Steel sheet (as for example, automobile body stock)
conventionally is cast as rimmed steel.
Once the ingots have been poured, they are retained in the ingot mold for a spec-
ified period of time depending on the size of the ingot and the chemistry of the steel.
This period is usually referred to as "track time" because the ingots are usually
handled on special railroad flat cars. At the proper time, the ingots are removed from
the molds or the molds are removed from the ingots, depending on whether the ingots
are of "big-end-up" or "big-end-down" configuration. Ingots are next transferred to
soaking pits where they are heated to equalize the temperature throughout the ingot and
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-106
increase the temperature of the ingot to that required for rolling. The combustion
characteristics of the soaking pits are often controlled so that a certain thickness of
scale is developed on the surface of the ingots. This is done so that when the scale
breaks off during the first pass in the rolling operation, it takes with it certain minor
surface defects and the result is a better surface on the ingot. This practice results
in a major source of "mill scale".
Ingots, depending on their size and the desired end product, are rolled in blooming
mills or slab mills. Mills that perform both functions are often called universal mills.
The billets, blooms, and slabs that are the semifinished products in this stage of pro-
duction are cooled to ambient temperature and transferred to a storage area where
they are inspected, and where the surfaces are prepared or "conditioned" for the follow-
ing rolling operations.
Emis sion Identification. When molten steel is exposed to air, some type of fume
IS generated. This is particularly true during pouring when the molten steel is subjected
to turbulent action during the time it is flowing into the mold. Gaseous emissions occur
during the heating of ingots in the soaking pits, and a minor amount of particulates are
released during the primary rolling operations. A flow diagram illustrating the casting
and rolling operations is shown in Figure C-83, in which numbers in circles refer to
the source s of emis sion discus sed below.
CD
Fume - Minor amounts of iron oxide fume are generated in the tapping
of molten steel into ladles and during pouring of the molten steel from
the ladle into the ingot molds. Data concerning amounts, size, and
chemical composition are not available; however it can be assumed
that the particulates are of a very fine size and are primarily iron
oxide.
@
Gases from Mold Coatings (frequently carbonaceous) - Mold coatings
are applied to the inner surface of ingot molds to minimize certain
types of defects that are detrimental to the .quality of the final prod-
uct. One type is a nonvolatile coating that relies on surface texture
to accomplish the desired surface improvement. The other general
type is a coating that relies on volatilization to shield metal splashes
from the mold wall and produce an improved surface on the ingot.
The volatile type of mold coating include s coal-tar products,
petroleum derivatives, or naturally occurring materials such as
Gil sonite( 150).
Q)
Gases from Hot-Top Materials - Depending on the materials used to
make hot tops for use with ingots of killed steel, gases of varying
amounts and composition can be generated. Data on the chemical
compositions of the gases and their amounts are not available. Hot
tops can be divided into three classifications: (1) permanent, (2) in-
sulating, and (3) exothermic(l51-156) Permanent hot tops are made
of castable or preformed refractories. Amounts of gases generated
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
I~
~r
Slab,
pounds
STORAGE
C-107
Molten steel,
2020 pounds
Fume
LADLE
Solid ingot,
2000 pounds
(Heating)
SOAKING PIT,
2000 pounds
G)
o
o
~
G)
Scrap,
390 pounds
Atmosphere in vicinity
of operation
o
Combustion products
Atmosphere
Heated ingot,
2000 pounds
Mill scale,
10 pounds
Billets,
pounds
Scrap,
390 pounds
Scrap,
390 pounds
Bloom,

pounds
FIGURE C-83. CASTING OF STEEL INTO INGOTS; AND ROLLING TO SLABS, BLOOMS, AND BILLETS
STORAGE
(Numbers in circles refer to sources 'of emissions described in the text. )
BATTELLE MEMORIAL. INSTITUTE - COLUMBUS LABORATORIES

-------
C-108
by these in use are very small, and their compositions are harmless.
Insulating hot tops contain materials that have a tendency to char and
become porous. Smoke may be generated when molten steel comes in
contact with the insulating materials. Exothermic hot tops are gen-
erally composed of mixtures of aluminum, iron oxide, oxidizers, and
inert materials. Contact with molten steel starts the exothermic reac-
tion in the hot top, and causes generation of smoke and fumes. Some
compositions of exothermic hot tops cause copious fumes. Commer-
cial exothermic hot-top materials are proprietary and their composi-
tions are not publicized. Data on th.e amounts and chemical composi-
tion of emissions generated during pouring of ingots are not available.
Usually the emissions generated are treated as a nuisance, but their
contribution to air pollution can be substantial.
@)
Lead Fumes (Lead oxide fumes generated during the addition of lead
to free-machining steels) - Certain types of free-machining steels
contain lead which acts as a chip breaker during machining. The
lead is added in the form of shot during the time the molten steel
is poured into the ingot mold. Because lead fumes are considered
to be a health hazard, facilities for their collection usually are
provided where this type of steel is produced.
0)
Combustion Products (Carbon dioxide, carbon monoxide, nitrogen,
and sulfur dioxide) - Ratios of carbon dioxide to carbon monoxide
are determined by specific combustion practices in each steel
plant. The amount of sulfur dioxide is dependent mostly on the
amount of coke-oven gas used and on the degree t~ which sulfur has
been removed from the coke-oven gas. The use of coke-oven gas
as a fuel in the firing of soaking pits and the amounts used is deter-
mined by the energy availability and energy requirements in a spec-
ific steel plant. Tolerance of hydrogen sulfide in coke-oven gas
is determined to some extent by the types of steels produced and
whether the existing level of hydrogen sulfide is considered a factor
in air pollution. Some steel companies use a minimum amount of
coke-oven gas or desulfurize it before use( 107). Others use the
coke-oven gas as it is received from the coke plant without any
special desulfurizing. (12,49,157) If the sulfur content of the coal
used for making coke is low enough, the sulfur content of the coke-
oven gas is correspondingly low and the steel company does not
consider desulfurization to be necessary. (158)
Primary rolling operations do not contribute emissions to air pollution, unles s
steam generated from the flow of water in the primary mill is considered as an emis-
sion. Heavy iron-oxide scale (mill scale) that is formed on the ingots during heating
in the soaking pits is broken from the ingot during the first pas ses in the mill and is
removed from the mill by water which carries it to scale pits from whence it is further
transported to scale-recovery systems that are associated with the water-treatment
facilities of the steel plant.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-109
Ingot Casting and Rolling Emission Control. The emissions generated during the
pouring of ingots can be minor or major, depending on the practice in the particular
plant. When detrimental emissions are generated (such as in making leaded steels,
using volatile coatings on the molds, or using some exothermic hot tops) the necessary
facilities for exhausting the emissions should be provided. Bag houses sometimes are
used for these applications. To some degree, fuming of the steel during teeming can
be minimized by shrouding the stream of molten metal with an inert gas such as
argon. (159, 160) This technique has found use in the production of high-quality vacuum-
degassed steels, but the economiCs at this time do not favor its use for tonnage-steel
applications. In some electric-furna.ce shops where the entire building is considered
as an emission-control system, the fumes generated during ingot casting are ex-
hausted by the cornmon emission-control system for the entire operation. (139)
Continuous Casting
Continuous casting eliminates the need for ingot molds, soaking pits, and rolling
mills to produce billets, blooms, and slabs; and in some situations is thought to pro-
vide an economic advantage over the conventional method of casting ingots. High-
tonnage production of continuous-cast products is just starting in the integrated iron
and steel industry. The estimated annual production from 1962 through 1968 is shown
in Figure C-84 in comparison with the total product.ion of billets, blooms, and slabs
by all methods. It should be pointed out that the estimated actual production for 1968 of
100
95
Total production of
billets, blooms, and
slabs
III
g 90
+-
+-
Q)
c:
..... 85
o
III
c:
o
.- 80
:E
+-
U
::::J
"0
o
...
a..
10
5
Estimated continuous-:-cast "-
production -\..--"
---
. --
--
o
1962
1963
1964
1965
Year
1966
1967
1968
FIGURE C-84.
CONTINUOUS-CAST PRODUCTION AS COMPARED WITH TOTAL
PRODUCTION OF BILLETS, BLOOMS, AND SLABS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-110
about 4. 5 million net tons of continuous-cast steel is far short of the 1968 estimated
capacity of 7 million tons. This difference is caused mainly by delays encountered in
the start-up of the larger continuous casting machines.
Emission Identification. The generation of emissions during furnace tapping and
filling of the ladle for continuous casting is the same as for conventional casting of in-
gots. However, the mechanics of the continuous-casting proces s and equipment design
tend to minimize emissions in the actual continuous-casting operation. A flow diagram
showing the points of emission in the continuous casting of steel is shown in Fig-
ure C-85, in which circled numbers refer to the following discus sion.
CD
Fume - Minor amounts of iron oxide fume are generated during the
filling of the ladle and during further handling of the molten steel
on its way to the continuous-casting machine. Data are not available
concerning amounts or size consist of the fume. Fume generated
during the actual pouring of the molten steel into the continuous-
casting machine is minimal. This can be attributed to the fact that
the molten steel is falling from the ladle to the tun-dish and finally
'into the continuous-casting mold over very short distances. This
reduces its exposure to the air to a very short period of time, and
minimizes the opportunity for iron oxide fume to form. The tun-
dishes sometimes are covered or blanketed with inert or reducing
gases which tend to further minimize the formation of oxide fumes.
@
Carbonaceous Gases - Lubricants are used in the continuous-casting
mold to prevent seizure between the solidifying steel and the mold.
Rape seed oil is the usual lubricant. Data concerning the amounts
and compositions of the gases generated are not available. However,
the amounts generated are so small that this emission is not con-
sidered as a problem by steel companies.
Q)
Fume - Oxide particles that are generated during the cut-off opera-
tion. No data are available on amounts. Chemical analysis would
show that the particulates are mainly iron oxide.
Continuous -Casting Emis sion Control. The continuous -casting proces s by its de-
sign tends to minimize the generation of emissions, and emission-control equipment,
as such, usually is not incorporated into the process. Some exhausting systems may
be employed to evacuate the steam that is generated in the secondary cooling system
by the impingement of water sprays on the hot steel( 161). Cutoff torches used in cut-
ting the continuous-cast steel into manageable lengths generate a small amount of
particulate emissions. In the production of plain carbon steels, this is not considered
a problem by steel companies. In a few installations in which- stainles s steels or free-
machining leaded steels are made, emission-control systems using filters are used to
collect the fumes generated(l6l). Many installations have hydraulic shears to cut con-
tinuous billets or blooms into the desired length. This in effect eliminates the emission
problem. In comparison to the casting of billets, the casting of large slabs creates a
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
Molten steel,
2060 pounds
CONTINUOUS-CASTING
MACHINE - SLABS
G)
o
o
Solidifying slab.
2013 pounds
Scale,
S pounds
Continuous solid slab.
2005 pounds
C-lll
LADLE
G)
o
CONTINUOUS -CAS TING
MACHINE - BILLETS AND BLOOMS
MECHANICAL SHEARS
Billet and bloom.
2005 pounds
FIGURE C-S5. CONTINUOUS CASTING OF STEEL
(Circled numbers refer to emissions discussed in the text. )
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-1l2
somewhat greater amount of fume in the cutoff operation involving torches, simply be-
cause a thicker cross section of steel must be cut. Shears could be used to cut the
continuously cast slab, but the high cost for such a shear usually is considered to make
this method of cutting economically unattractive. It has been estimated that the cost
of such a shear would be in excess of $1 million. (162)
. Pres sure Casting
Pres sure casting, like continuous casting, eliminates much of the equipment re-
quirements associated with conventional ingot casting and rolling. Pressure casting
usually is not economically competitive with continuous casting in the production of
billets and blooms, but in the casting of slabs it doe s appear to be competitive for some
situations. More information on the competitive situation will undoubtedly be forth-
coming when a 500, OOO-net-ton-per-year plant for pressure casting of carbon steels is
put into operation in 1969 in Oregon. Production figures for pres sure-cast slabs are
given in Table C-25(l63).
TABLE C-25.
PRODUCTION OF PRESSURE
CAST SLABS
Year
Annual Production, net tons
1964
1965
1966
1967
1968
35,000
50,000
70,000
85,000
150,000
Emission Identification. Like other steelmaking processes, fumes are generated
during the tapping of the furnace and filling of a ladle. From that point on, emissions
from pressure casting probably are less than for continuous casting. Points of emission
are shown in the flow sheet in Figure C-86, in which circled numbers refer to the fol-
lowing discussion.

-------
C-113
Q)
Fume - Iron oxide fumes generated during torch cutting of the
riser and its removal from the slab. These are considered
insignificant in contributing to an emission problem from the
steel plant, and can be readily collected by a suitable exhaust
system.
, Molten steel.
2150 pounds'
G)
o
TORCH CUT OFF
. FIGURE C-86. PRESSURE CASTING OF STEEL SLABS
(Circled numbers refer to emissions discussed in the text. )
Manufacture of Finished Products
The manufacture of finished products, for the purposes of this study, is con-
sidered in two steps. The first is the rolling of billets, blooms, and slabs into hot-
rolled and cold-rolled products. The second is the coating operations associated with
the production of finished products. .
Rolling Operations
Rolling operations for the production of finished products are generally carried
out on continuous mills. Billets are usually processed into products such as reinforcing
bar, hot-rolled bar, cold-rolled bar, wire for cold-heading operations (this may be as
large as 1-1/4-inches in diameter), high-quality bar products depending on the chemis-
try of the steel, and small angles and channels. Blooms are used to make larger
BATTELLE MEMORIAL INSTITUTE - COLUMBUS' LABORATORIES

-------
C-114
angles, channels, and structural products. These are normally furnished in the condi-
tion they are produced and without any further finishing or cleaning operations. Flat-
rolled products, on the other hand, are supplied hot-rolled, cold-rolled, tin-plated,
galvanized, terne-coated, painted, or plastic-coated to commercial users of the
products.
Hot-Rolling Emission Identification. As the hot-rolling of steel proceeds toward
finished products, the generation of air-polluting emissions is minor. Flow sheets for
the rolling of finished products are shown in Figure C-87 for bar and merchant products
from billets and blooms, and in Figure C-88 for flat-rolled products from slabs. Con-
ditioning of semifinished products is generally done at the primary-mill facilities before
billets, blooms, or slabs are delivered to the finish-rolling mills. However, this will
vary from plant to plant.
Conditioning at the primary mill is primarily spot conditioning that is done with a
variety of equipment. Hand grinding and chipping hammers are used along with special
equipment, known as "peelers", that mechanically removes defective portions of the
billets or blooms. Scarfing with hand torches is also done. The type of conditioning
treatment is determined by the chemical composition of the steel and by the quality re-
quirements specified by the consumer for the finished product. Grinding, chipping, and
peeling techniques do not generate emissions as considered in this study. Grinding
operations do produce some particles, but these are collected at the grinding station.
Generally these particles are of such a size that they do not become extensively air-
borne, and settle in the vicinity of the grinding operation. The following circled numbers
refer to points of emission as marked on Figures C'-87 and C-88.
CD
Fume - Iron oxide fume generated from hand scarfing of the billets
and blooms. No information is available concerning the size or the
amount of fume generated per net ton of billets or blooms. How-
ever, it has been reported that the loss in yield for scarfing billets
varies from 3 to 6 percent depending on the type of steel (164). Most
of this metal loss is in the form of metal splatter rather than fume.
Hand scarfing of slabs poses the same proht'ems as hand scarfing of
billets and blooms. However, hand scarfing of slabs usually is
limited to the removal of deep defects that would not be removed by
machine scarfing prior to rolling.
@
Products of combustion - Mixtures of carbon monoxide, carbon
dioxide, and nitrogen resulting from the firing of reheat furnaces
prior to rolling of billets, blooms, or slabs. The semifinished
products in this stage of manufacture are at a point where quality
control is getting to be a very large factor in the process. Re-
heating must be done under closely controlled conditions so the
semifinishe~ products are at the proper temperature for rolling
and are heated uniformly throughout the thicknes s of the product.
Natural gas is the usual fuel for firing the reheat furnaces) and,
under the closely controlled combustion practices, pollution is not
considered to be a problem. As the semifinished products proceed
toward a finished product, any losses in material contribute more to
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
BILLET AND BLOOM
STORAGE
(0
o
(0
Pickled
Grit
blasted
FIGURE C-87. TYPICAL HOT-ROLLING OF BAR AND
MERCHANT MILL PRODUCTS FROM
BILLETS AND BLOOMS
(Circled numbers refer to discussion
in text. )
C-115
SLAB STORAGE
8
CD
G)
Cold
rolling
FIGURE C-88. TYPICAL HOT-ROLLING OF SHEET AND
S TRIP FROM SLABS
(Circled numbers refer to discussion
in text. )
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-116
economic loss than those of any preceeding steps. Induction heat-
ing of large, continuous-cast slabs is being implemented at the
McLouth Steel Corp., to provide uniform heating of slabs as well
as to reduce the amount of scale formed on the steel. Needles s
to say, this method of heating steel for rolling is a pollution-free
method.
Q)
Fume - Iron oxide fume generated during machine scarfing of slabs,
blooms, or billets immediately prior to hot rolling. Fume is re-
ported to be generated at a rate of 2 to 3 grains per standard cubic
foot of gas, and the dust-collecting systems typically are designed
to operate between 75,000 and 135,000 C£m during the short dura-
tion of the scarfing operation(1, 101, 157). Several factors that enter
into the amount of metal that is removed during machine scarfing in-
clude: (1) speed of the semifinished product through the machine,
(2) oXYyen pres sure at the scarfing head, and (3) temperature of the
steel. ( 65) Metal removal by machine scarfing results in yield
reduction which can normally vary between 0.85 to 1.54 percent,
but can go as high as 2.5 percent for slabs, and as high as 7 per-
cent for blooms. (166-168) Machine scarfing is increasingly used
as a means for surface conditioning prior to hot rolling. The trend
in the tonnage of steel that is machine scarfed is shown in Fig-
ure C - 89. (168-170) It has been estimated that this trend will level
off at 50 percent of all tonnage, and that based on information re-
ceived from Europe, a considerable amount of continuously cast
steel will have to be machine scarfed. (169) Steel technologists in
the United States are directing a great deal of effort toward improving
the surface quality of continuously cast steel so as to eliminate the
need for extensive surface conditioning.
50

45
40
     1968~
      ~
      /
     / 
    / 
    /  
    r  
   I   
  ~ ""   
 .-    
- V     
-      
35
30
....
~ 25
&. 20
15
10
5
o
1935
1940
1945 1950 1955
Year
1960 1965
1970
FIGURE C-89.
TREND IN INGOT TONNAGE OF STEEL THAT IS
MACHINE SCARFED

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-1l7
@)
Fume - Iron oxide fume generated during rolling in last few finishing
stands of a hot-strip mill. The emissions are submicron in size,
but no further information is available as to size distribution and
amounts. (157)
Hot-Rolling Emission Control. The principal emission associated with hot rolling
is iron oxide fume generated during the scarfing operation. Generally emis sion-
control equipment is not used with hand scarfing operations, because this is usually
an intermittent type of operation that is combined with chipping and grinding. How-
ever, some steel mills, especially those concerned with meeting increasing customer
quality requirements, may scarf almost all of the billets and blooms by hand. In such
situations, exhaust hoods are used to remove the fumes from the scarfing area. Infor-
mation concerning the collection and/ or disposal of the fumes and dust collected is not
available.
Fume-control from machine scarfing of slabs varies from no control to complete
control. (1,12,49,101,157,158) Both electrostatic precipitators and high-energy
scrubbers are used to collect the fume generated during machine scarfing. (1,157,158)
The small amount of iron oxide fume generated in the finishing stands of a hot
mill is reported to be collected in one plant by a high-energy scrubber. (157)
Cold-Rolling Emission Identification. Particulate and gaseous emissions asso-
ciated with preceding steps in the manufacture of steel are not encountered in the cold-
rolling operation. The major emis sions associated with cold-rolling operations are
acid fumes generated in the pickling of hot-rolled steel strip or dust resulting from
mechanical cleaning operations that may be used to prepare the surface of hot-rolled
strip for cold rolling. A flow sheet illustrating the cold-rolling process is shown in
Figure C-90, in which circled numbers refer to the following discussion.
Pickling of hot-rolled strip is required to provide a clean metal surface for
cold rolling.
CD
Acid Fumes - Either sulfuric or hydrochloric ac.id fumes depending
on the individual plant situation with respect to product require-
ments and availability and cost of acid. Sulfuric acid has been used
traditionally for pickling operations to prepare hot-rolled steel for
cold rolling. The increasing cost of sulfuric acid, combined with
problems of disposal of spent sulfuric acid liquor has prompted a
trend to hydrochlo ric acid pickling. (171, 172) It has been estimated
that, by 1970, hydrochloric acid and sulfuric acid will each be used
for 50 percent of the steel-pickling operations. (171)
~
Water-Oil Mist - Mixtures of water with water-soluble oils generate
mists during the cold- rolling operations.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-1l8
Hot -rolled coils
ACID PICKLING TANK
G)
Re -coiled hot -rolled
strip
Pickled hot -rolled
coils
COLD -ROLLING MILL
o
FIGURE C-90. COLD ROLLING OF STRIP STEEL
(Circled numbers refer to discussion in text. )
Cold-Rolling Emission Control. Acid fumes from the pickling tanks generally
have been controlled by steel plants. The pickling lines are hooded and exhausted to
fume-control systems. Wet scrubbers and packed towers are used to remove the acid
mists from the pickling tanks. (1,13,101,157) Where tonnage requirements do not
economically warrant the installation of acid-fume collection system1 the fumes may
be exhausted to the atmosphere by building roof-exhaust systems. (15 ) Collection of
mists generated by cold-rolling operations is done with mechanical mist eliminators or
with wet s c ru b be r s ( 1 57, 1 73) .
Coating of Finished Products
Surface coating of steel fo r protection or appearance is tending toward continuous-
line proces sing so that operating economies can be realized. The continuous lines also
lend themselves to control of emissions that may be generated. A listing of the number
of continuous and batch units in the integrated iron and steel industry is given in
Table C-26. In addition to the facilities listed, there are 3 batch-galvanizing facilities
for pipe and 15 continuous-galvanizing facilities for wire. The various facilities are
located in 43 steel plants throughout the country. Annual production of several different
types of coated products from 1958 through 1967 are shown in Figure C-91. The produc-
tion of hot-dipped tin and terne plate which are made primarily in batch operations that,
from pollution aspects, are more difficult to control, has decreased from 450,000 net
tons in 1958 to 30,000 net tons in 1967.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
(/) 
c: 
0 
- 
- 
Q) 
c: 
'0 
(/) 
c: 
0 
E 
~ 
c: 
0 
+= 
0 
::J 
"0 
e 
a.. 
 O.
 O.
 o.
 o.
 o
C-1l9
12.0
              /  - 
             /   
        Total   /    
     J~           
 .......  /            
 '" I              
      Electrolyte tin plate   ,- 
     / ...--...,     ---- --- ~" 
       -'--I.... ,   
    /'       ~"".. " ~  "" 
 '....  I            
  ', I          ~   
           ~     
      /" Galvanized sheet    
   ----             
.,.:>                >
5                 
 ~....   -,',           
4  ....            
  ~   \Hot-dipped tin and terne plate   
      \           
3       \          
      ,          
        "',    k? ~  ~ 
          '.... -    
2   Long terne  ~ --~ ~    
 /"      " ---    
      --- -,"  "''''....   
.1   -           
-'~             "  
 ~  Other metallic coated    ...  
      I I I     --- 
11.0
10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
0.0
1958
1962 1963
Year
1965 1966
1967 1968
1964
1959
1960 1961
FIGURE C-91. PRODUCTION OF SHEET STEEL COATED PRODUCTS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C -120
TABLE C-26.
SURFACE-TREA TMENT FACILITIES FOR SHEET
PRODUCTS IN THE INTEGRA TED IRON AND
STEEL INDUSTRY
Type of Facility
Continuous
Batch
Acid pickling
Galvanizing
Electrolytic tin plating
Long terne sheets
Aluminum coating
Chromium coating
Nickel coating
Copper coating
Painting

Total (Excluding pickling)
124
62
40
6
2
4
2
2
6
42
12
15
6
2
o
o
o
o

35
124
Emission Identification. Emissions associated with the surface-treatment lines
are not the particulate and gaseous emissions associated with other steelmaking prac-
tices. Emissions originate mainly from the acid-pickling operations and from the coat-
ing operations. Many of the coating processes are mechanically similar - the differ-
ences being in the type of coating applied. Processing is similar for the batch opera-
tions and for the continuous operations, except that the continuous lines are designed to
perform the coating operation around the clock rather than on an intermittent, stop-
and-go basis as for batch treatments. The continuous operations usually lend them-
selves to better control of emissions than do the batch operations. A flow sheet for
typical galvanizing operations is shown in Figure C-92, in which circled numbers refer
to the following discussion. .
CD
Products of Combustion - Normal mixtures of carbon monoxide,
carbon dioxide, and nitrogen as sociated with the combustion of
natural gas in heat-treating furnaces.
@
Acid Fumes - Sulfuric or hydrochloric acid fumes. One steel plant
has reported that hydrochloric acid fumes are emitted to the atmo-
sphere from a tower scrubber at a rate of 0.5 gallons per hour when
treating steel at the rate of 100 tons per hour. (174) This rate of emis-
sion is equivalent to 108 grains per ton of steel treated.
Q)
Flux Emissions - Emissions generated from the cover flux which can
be ammonium chloride or zinc ammonium chloride. No data are avail-
able on the chemical composition of the emissions. However, data
from galvanizing job shops provided the compositions given in Ta-
ble C-27. (175) In the absence of data from iron and steel industry
installations, these data may be considered as typical of emissions
from batch operations in steel-plant facilities. Particulates from

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-121
batch galvanizing installations have an average particle size of
about 2 microns as they evolve from the pots, but have a tendency
to agglomerate. A size analysis of par'ticulates taken from a bag
house showed that 23 percent by weight of the particulates were
greater than 250 microns in size. (175)
(2)
(2)
HEA T -TREA TING
FURNACE
HEA T -TREA TING
FURNACE
G)
o
G)
o
FIGURE C-92. TYPICAL GALVANIZING PROCESSES
G)
HEAT-TREATING
FURNACE
G)
(Circled numbers refer to discussion in text. )
TABLE C-27.
CHEMICAL COMPOSITION OF GAL V ANIZING EMISSIONS
Component
Source One, weight percent
Source Two, weight percent
H20
ZnCl2
ZnO
Zn
NH4CL
NH3
Oil
Carbon
Not identified
2.5
3.6
15.8
4.9
68.0
1.0
1.4
2.8
0.0
100.0
1.2
1-5.2
6.5
0.0
23.5
3.0
41. 4
0.0
9.2
100.0
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
,(

-------
C -122
Coating installations for' the application of other metals', paints,. and plastics are
essentially the same as those illustrated in the flow sheets in Figure C-92. They con-
sist of a preparation section followed by plating, dipping, or spraying, which is fol-
lowed by cleaning and drying operations.
Auxiliary Operations
Two auxiliary operations considered in this study are (1) the foundry facilities
associated with steel-plant operations, and (2) the incineration facilities required for
the disposal of solid combustible wastes generated in the manufacture of iron and steel.
Foundry Facilities. Foundry facilities associated with steel plants are primarily
used to supply needed replacement parts as a maintenance function. Some companies
have facilities for making their own ingot molds (neces sary for the conversion of molten
steel to solidified steel) and rolling mill rolls (required to make semifinished and fin-
ished products). Foundry facilities for steel plants in the United States are given in
Table C-28, with data on ~elting facilities and types of metals melted. (176)
Three major companies that supply ingot molds to the steel industry are:
(I) Shenango Incorporated, with plants at Neville Island and Sharpsville, Pennsylvania,
and Buffalo, New York, (2) Valley Mold and Iron Corporation, with plants at Chicago,
Illinois, Cleveland, Ohio, and Hubbard, Ohio, and (3) Vulcan Mold and Iron Company,
with plants at Lansing, Illinois, Latrobe, Pennsylvania, and Trenton, Michigan.
Emission Identification. Emissions from the aluminum-melting operations are
generally limited to chlorides that are generated by the use of chlorine gas in the re-
moval of gases from the molten aluminum. Emissions generated from the melting and
casting of brass are generally limited to zinc oxide fume generated from the oxidation
of zinc in the alloy. No specific data are available on emissions from aluminum and
bras s melting, but it is reasonable to expect that they have the same characteristic s
as those generated in the respective foundry operations in ordinary commercial
operation.
Specific data on emissions from cupolas also are not available with respect to
specific steel-plant operations. Again,. it can be expected that the emissions generated
will have the general characteristics of emissions generated in commercial gray iron
foundry operations. It is understood that a project has been initiated by the National
Air Pollution Control Administration to investigate the problems of iron and steel
foundry emissions in detail. An indication of the size characteristics of particulates
emitted from gray iron cupolas is shown in Table C-29. These data are from litera-
ture sources on the gray iron foundry industry. (179,180) Chemical analyses of par-
ticulates from a cupola are given in Table C-30. (181) The combustibles noted in
Table C-30 are primarily particles of coke that are blown from the cupola. The many
operating and material variables in cupola melting influence the characteristics of
emissions from gray-iron melting cupolas. These variables include (1) the ratio of
coke to metallics charged,. (2) characteristics of the coke, (3) type of scrap, (4) clean-
liness of scrap, and (5) the rate at which air is blown into the cupola.
BATTELLE MEMORIAL INSTITUTE - COLl:JMBUS LABORATORIES

-------
C-123
TABLE C-28. FOUNDRY FACILITIES LOCATED AT STEEL PLANTS IN THE UNITED STATES
Company Plant Metals Melted Special Items Melting Facilities
Armco Steel Corp. Ashland, Ky. Brass, aluminum,   
  gray iron   
 Torrance, Cal. Steel    
Bethlehem Steel Co. Bethlehem, Pa. Brass, gray iron,  Three 25-tOn electric
  ductile iron  induction furnaces
 Foundry and Machine Gray iron and ductile Rolls  
 Division, iron    
 Bethlehem, Pa.(177,178)     
 JohnstOwn, Pa. Brass, gray iron   
CF & I Corp. Pueblo, Colo. Brass, aluminum, gray   
  iron, steel   
Crucible Steel Corp. Midland, Pa. Brass, gray iron, steel   
Ford MotOr Co. Dearborn, Mich. Gray iron Ingot molds  
Granite City Steel Co. Granite City, ill. Gray iron Ingot molds  
Inland Steel Co. East Chicago, Ind. Gray iron Ingot molds  
Jones & Laughlin Pittsburgh, Pa. Brass, aluminum, gray  Two cupolas, 9 tons per
Steel Corp.  iron, steel  hour each 
Kaiser Steel Corp. Fontana, Cal. Brass, aluminum, gray Ingot molds  
  iron, ductile iron, steel   
Lone Star Steel Co. Lone Star, Texas Gra y iron Ingot molds Two cupolas, 30 tons per
     hour each 
Repu bUc Steel Corp. Gadsden, Ala. Brass, aluminum, gray  Two cupolas, 10 tOns per
  iron, steel  hour each 
United States Steel Roll and Machine Works, Gray iron, ductile iron Rolls Two cupolas, 9 tons per
Corp. Canton, Ohio    hour each 
 South Chicago, ill. Aluminum, gray iron,  One cupola, 10 tons per
  ductile iron  hour 
 Gary, Ind. Gray iron, ductile iron,   
  steel    
 Fairfield, Ala. Brass, aluminum, gray   
  iron, steel   
 Johnstown, Pa. Gray iron, ductile iron,  One cupola, 12 tons per
  steel   hour 
 Lorain, Ohio Gray iron, steel  One cupola, 9 tons per
     hour 
 Braddock, Pa. Gra y iron Ingot molds  
 Provo, Utah Brass, aluminum, gray   
  iron, steel   
Youngstown Sheet and Indiana Harbor, Ind. Gra y iron   
Tube Co.      
BATTELLE MEMORIAL INSTITUTE '- COLUMBUS LABORATORIES

-------
C -124
TABLE C-29. SIZE CHARACTERISTICS OF PARTICULATES EMITTED FROM GRAY IRON CUPOLAS
Particle Size  Screen Size, Test I, Test 2,
Range, microns Weight Percent microns weight percent weight percent
o - 4 4 - 10 +833 0.7 2.9
5 - 9 2 - 15 +252 12.0 19.6
10 - 24 4 - 15 +147 18.1 23.8
25 - 49 5 - 15 +97 16.6 17.5
50 - up 45 - 85 +74 9.3 8.8
   +47 12.7 11.5
   -47 30.6 15.9
TABLE C-30. CHEMICAL ANALYSIS OF PARTICULATES
FROM A GRAY IRON CUPOLA
Component Range, percent
Si02 20 - 40
CaO 3 - 6
Al203 2 - 4
MgO 1 - 3
FeO, Fe203' Fe 12 - 16
MnO 1 - 6
Combustibles 20 - 50
Emissions generated in the melting of steel for castings in a steel plant are the
same as those described in the sections on steelmaking emissions, because the steel
used for castings is made by the facilities already located in the steel plants. Infor-
mation on these emissions can be found on pages C-50 through C-93 of this Appendix.
Incineration. The large amounts of miscellaneous combustible solid materials
that must be disposed of in a steel plant create a disposal problem. In older steel
plants, several small incineration facilities may be located in various parts of the
steelworks complex to handle the disposal of wastes. However, with the emphasis now
placed on the control of emissions, many of the multiple-site incineration facilities
probably will be consolidated into one major incineration facility with the required
emis sion-control equipment as part of the facility. This trend is taking place with the
new plants under construction and during modernization of existing steel plants. No
specific data are available on emissions generated in steelworks incinerating operations.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
C-125
REFERENCES FOR APPENDIX C
(1)
Ziercher, J. L., Report on Official Travel, Battelle Memorial Institute, Columbus
Laboratories, October 4, 1968.
(2)
Tsujihata, K., et al., "Developments in Ironmaking at Yawata Iron and Steel Co. ,
Ltd. ", Blast Furnace and Steel Plant, 53 (3), 242-248 (March 1965).
(3) Send, A., and Wimzer, G., "Trends in Burden Preparation and Their Effect on
Blast Furnace Operation in Germany", Journal of Metals, 19 (7), 58-64 (July
1967). -
(4) Knepper, W. A., and Sciulli, C. M., "Operation of an Experimental Blast Furnace
with Sized Coke", AIME Ironmaking Proceedings, 24,38-40 (1965).
(5 )
White, R. H., "The Effect of Coke Sizing on Blast Furnace Operation", Blast
Furnace and Steel Plant, 54 (3), 241-245 (March 1966).
(6) White, R. H., and Meyer, V., "Blast Furnace Operation With Washed Burden
Materials", Journal of Metals, ..!:1 (6), 52-54 (June 1967).
(7) Nitchie, C. M., "Effect of Screened and Sized Sinter on Blast Furnace Operation",
AIME Ironmaking Proceedings, 26, 15- 19 (1967).
(8) Annual Statistical Report, American Iron and Steel Institute, New York, N. Y.
(1962, 1967).
(9) Harris, E. R., and Beiser, F. R., "Cleaning Sinter Plant Gas With Venturi
Scrubber", Journal of the Air Pollution Control Association, 15 (2), 46-49
(February 1965). -
(10)
Baranyi, J. F., "Results of Design Changes in Sinter Plant Fans", Iron and Steel
Engineer, 42 (12), 85-90 (December 1965).
(11) McCrone, W'. C., et al., "The Particle Atlas", Ann Arbor Science Publishers,
Inc., Ann Arbor, Michigan (1967), pp 153-236.
(12) Young, P. A., et al., II The Generation and Treatment of Sinter Plant Dusts",
AIME Blast Furnace, Coke Oven and Raw Materials Proceedings, 20, 299-313
(1961). -
(13)
Ziercher, J. L., Report on Official Travel, Battelle Memorial Institute, Columbus
Laboratories, October 17, 1968.
(14) "Symposium on Sinter P\ants", Iron and Steel Engineer, 36 (6), 101-122 (June
1959).
(15) Chapman, H. M., "Experience With Selected Air Pollution Control Installations in
the Bethlehem Steel Company", Journal of the Air Pollution Control Association,
~ (12), 604-606 (December 1963).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
(16)
(17)
( 18)
(19)
(20)
(21 )
( 22)
(23)
(24)
(25)
(26)
(27)
(28)
(29)
(30 )
C-126
Frame, C. P., "The Effects of Mechanical Equipment on Controlling Air Pollu-
tion at No.3 Sintering Plant, Indiana Harbor Works, Inland Steel Company",
Journal of the Air Pollution Control Association, 13 (12), 600-603 (December
1963). -
Young, T. A., "Gary Steel Works Experience With Dust Control at Number 3
Sinter Plant", Preprint, Blast Furnace and Coke Plant Association Meeting,
Chicago, Illinois, October 4, 1968. 18 pp.
"Iron Ore News Highlights qf 1967", American Iron Ore Association, January 15,
1968. pp. 20- 23.
Behrens, et al., "The Effects of Lime Properties on Basic Oxygen Steelmaking",
AIME O~en Hearth Proceedings, 48, 74-82 (1965).
Tartaron, F. X., and Ruschak, J. D., "Effect of Lime Structure in Oxygen Steel-
making", Report of Investigations 6901, U. S. Bureau of Mines, 1967. 41 pp.
"Lime for Steelmaking: Tailored to Fit New Demands", 33/The Magazine of
Metals Producing, ~ (1), 58-93 (January 1967).
Thomas, F. H., "The Production of Blast Furnace and Sinter Plant Flux and
BOF Lime", Journal of Metals, .!.2 (6). June 1967. pp. 86-89.
Graf, D. L., and Lamar, J. E., "Properties of Calcium and Magnesium Car-
bonates and Their Bearing on Some Uses of Carbonate Rocks", Economic Geology,
Fiftieth Anniversary Volume, Part II, The Economic Geology Publishing Company,
(1955), pp. 639-696.
Boynton, R. S., and Gutschick, "Lime", Industrial Minerals and Rocks, AIME
Seeley W. Mudd Series, Third Edition (1960), pp. 497-517.
"New Lime Plant Begins Production", Iron and Steel Engineer, 42 (12), 162
(December 1965).
MacNamara, J., "Lime Plant at Algoma", Iron and Steel Engineer, 44 (10), 126-
130 (October 1967).
"Republic Steel Plans Limestone Plant on Lake Erie", Mining Engineering, 19
( 1 0 ), 2 9 - 3 2 (0 c to b e r 1 967 ) .
"CF & I Smoking With a Filter and the Air's a Pure Delight", Metalworking News,
p. 14 (October 21, 1968).
Uys, J. M., and Kirkpatrick, J. W., "The Beneficiation of Raw Materials in the
Steel Industry and Its Effect Upon Air Pollution Control", Journal of the Air Pollu-
tion Control Association, g (1), 20-32 (January 1963).
Ramm, A. N., "Minimum Theoretically Possible Coke Oven Consumption for
Pig Iron Production Under Modern Conditions", Stal, No. 10, pp. 860-871
(October 1964).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
(31 )
(32)
(33 )
(34)
(35)
(36)
(37)
(38)
(39)
(40)
(41)
(42)
(43)
(44)
(45)
C-127
Nakatani, F., et al., "Theoretical Consider.a tion on Blast Furnace Coke Rate",
Transactions of the Iron and Steel Institute of Japan, ~, 263-280 (1966).
"Practical Suggestions for the Reduction of Emissions, Dust, and Grit at Coke
Ovens", Special Publication No.5, British Coke Res earch As sociation,
Chesterfield, Derbyshire. (1962) 12 pp.
Herrick, R. A., and Benedict, L. G., "A Microscopic Classification of Settled
Particulates Found in the Vicinity of a Coke-Making Operation", Paper No. 68-
137, Annual Meeting of the Air Pollution Control Association, St. Paul, Minnesota.
June, 1968. 23 pp.
Sellars, J. H., and Hornsby-Smith, M.P., . "Smoke Emissions During the Charging
of Coke Ovens", Coke and Gas, 23, 411-420 (1961).
Fullerton, R. W., "Impingment Baffles to Reduce Emissions from Coke Quenching",
Journal of the Air Pollution Control Association, 17 (12), 807-809 (December
1967). -
Schapiro, N., 'and Gray, R. J., "Petrographic Classification Applicable to Coals
of All Ranks", Proceedings of the Illinois Mining Institute, 68, 83- 97 (1960).
Schapiro, N., et al., "Recent Developments in Coal Petro'graphy", AIME Blast
Furnace, Coke Oven, and Raw Materials Proceedings, 20, 89-109 (1961).
Benedict, L. G., and Berry, W. F., "Further Applications of Coal Petrography",
American Conference on Coal Science, Advances in Chemistry Series, Vol. 55,
American Chemical Society (1966), pp. 577-601.
Thompson, R. R., et al., "The Use of Coal Petrography at Bethlehem Steel
Corporation", Blast Furnace and Steel Plant, 54 (9), 817-824 (September 1966).
Bayer, J. L., and Denton, G. H., "Applications of Coke Microscopy.to Plant
Problems I', Blast Furnace and Steel Plant, 54 (12), 1133-1142 (December 1966).
Denton, G. H., et al., "Progress and Problems in Routine Petrographic Evalua-
tion of Coals for Coke Plant Use", Journal of Metals; .!.1 (5), 88-92 (May 1967).
Benedict, L. G., et al., "Relationship Between Coal Petrographic Composition
and Coke Stability", Blast Furnace and Steel Plant, 56 (3), 217-224 (March, 1968).
Thompson, R. R., and Benedict, L. G., "Goals, Accomplishments, and Limita-
tions of Petrographic Methods of Coal Evaluation'l, Journal of Metals, 20 (3), 79-
84 (March 1968). -
"McWane Company to Market Merchant Iron", American Metal Market, October 13,
1966.
Ban, T. E., and Violetta, D. C., "D-LM New Commercial Ironmaking Process",
Iron and Steel Engineer, 45 (9), 101-111 (September 1968).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
(46)
(47)
(48)
(49)
(50)
(51 )
( 52)
(53)
( 54)
(55)
(56 )
(57)
(58)
(59)
(60)
(61 )
C-128
"Gilmo.re Steel Corporation to Construct Integrated Steel Producing Plant in the
Pacific Northwest", Blast Furnace and Steel Plant, 55 (12), 1123 (December 1967).
"Some Economic Tests Coming at Oregon Steel", Steel, 163 (14), 40 (September 30,
1968).
"Phase II of Burns Harbor Project Started", American Metal Market, (March 15,
1968), p. 5.
Hoffman, A. 0., Report on Official Travel, Battelle Memorial Institute, Columbus
Lab.oratories, October 23, 1968.
Heynert, Von G., et al., "Charge Preparation and Its Effect 'on Operating Results
of the Blast Furnace", Stahl und Eisen, ~ (1), 1-12 (January. 5'., 1961).
"The Making, Shaping, and Treating of Steel", Eighth Edition, .1964, United States
Steel Corporation, Pittsburgh, Pennsylvania.
"Dust Recovery Practice at Blast Furnaces", Steel Industry Action Committee,
Ohio River Valley Water Sanitation Commission, 36 pp. (January, 1958).
Gaffney, L. J., and Holowaty, M. 0., "Inve stigation of the Effects of Burden
Constituents on Blast Furnace Refractory Linings'., AIME Ironmaking 'Proceedings,
~, 11 - 14 (1966).
Hipp, N. E., and Westerholm, J. R., "Developments in Gas Cleaning - Great
Lakes Steel Corp. ", Iron and Steel Engineer, 44 (8), 101-106 (August 1967).
Carney, D. J., et al., "Continuous Analysis of Iron Blast Furnace Top Gas",
AIME Blast Furnace, Coke Oven, and Raw Materials Proceedings, ~, 142-157
(1954).
Woehlbier, F. H., and Rengstorff, G. W. P., "Preliminary Study of Gas Formation
During Blast-Furnace Slag Granulation With Water", Paper No. 68-136, Annual
Meeting of the Air Pollution Control Association, St. Paul, Minnesota, June 1968.
Weise, W. H., "Blast Furnace Flue Dust Treatment Facilities", Sewage and
Industrial Wastes, 28, 1398-1402 (November 1956).
Ess, T. J., "Weirton Steel Company", Iron and Steel Engineer, 35 (11), W-42 -
W-61 (November 1958).
Ess, T. J., "United States Steel's Geneva Works'., Iron and Steel Engineer, 36
(6), G-2 - G-27 (June 1959).
Ess, T. J., "Kaiser Steel - Fontana Plant", Iron and Steel Engineer, 38 (2), K-1 -
K-23 (February 1961).
Crawford, C. C., "CF & I Steel at Pueblo", Iron and Steel Engineer, ~ (4), P-2 -
P-27 (May 1962).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
(62)
(63)
(64)
( 65)
(66)
(67)
(68)
(69)
(70)
(71 )
(72)
(73)
(74)
(75)
(76 )
(77)
C-129
Longenecker, C., and Lassen, E. G., IIJones and Laughlin Rebuilds and Expands
Purchased Plant, and Converts It Into a Modern Producer", Blast Furnace and
Steel Plant, ~ (8), 652-668 (August 1963).
Brady, J. L., IIAmanda Blast Furnace", AIME Ironmaking Proceedings, ~ (1964).
Sieger, E. W., and Dean, A. F., IINo. 1 Blast Furnace Reline'l, Iron and Steel
Engineer, 42 (2), 105-113 (February 1965).
Jewell, C. J., IIOver the 4000 Ton Barrier - Sparrow s Point' J'II, Blast Furnace
and Steel Plant, ~ (1), 59-63 (January 1968).
IINew Blast Furnace Operating at Indiana Harbor", Iron and Steel Engineer, 45
(1),161 (January 1968).
Stone, J. K., IIL-D Steelmaking at Mid-1967'1, Journal of Metals, .!1. (7), 10
( J ul Y , 19 6 7) .
Sims, C. E., IIWhat is Ahead in the Next 25 Years for Electric Furnace Steel-
making", Journal of Metals, 20 (2), 44-50 (February 1968).
Varga, J., Jr., Communication with the American Iron and Steel Institute,
November 7, 1968.
Bishop, C. A., et aI., IISuccessful Cleaning of Open-Hearth Exhaust Gas With a
High-Energy Scrubber", Journal of the Air Pollution Control Association, 11 (2),
83-87 (February, 1961). -
Schneider, R. L., IIEngineering, Operation and Maintenance of Electrostatic
Precipitators on Open Hearth Furnaces", Journal of the Air Pollution Control
Association, ~ (8), 348-353 (August 1963).
Zimmer, K. 0., IIDust-Laden Waste Gases Emitted in the Basic Open Hearth
Process With the Usual Melting Practice and With Oxygen Injection; Also Re-
moval of the Dust From Them", Stahl und Eisen, 84 (17), 1070-1075 (1964).
Elliot, A. C., and Lafreniere, A. J., IICollection of Metallurgical Fumes From
Oxygen Lanced Open Hearth Furnaces'l, Journal of Metals, 18 (6), 743-747
June, 1966. -
Smith, W. M., and Coy, D. W., IIFume Collection in a Steel Plant", Chemical
Engineering Progress, 62 (7), 119-123 (July 1966).
Broman, C. U., and Iseli, R.
With Venturi Type Scrubber",
(February 1968).
R., liThe Control of Open Hearth Stack Emis sions
Blast Furnace and Steel Plant, ~ (2), 143-148
Thomas, F. A., IIOxygen Lancing Means Open Hearth Waste Gas Cleaning Sys-
tems, II. Venturi Gas Scrubbers", Journal of Metals, ..!...2. (3), 264-266 (March
1965).
Jackson, A., IIFume Cleaning in Ajax Furnaces", Fume Arrestment, Special
Report 83, The Iron and Steel Institute, pp. 61-64 (1964).

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
(78)
(79)
(80)
(81)
(82)
(83)
(84)
(85)
(86 )
(87)
(88)
(89)
(90)
(91 )
(92)
(93)
C-130
Akerlow, A. K., "Modification to the Fontana Open Hearth Precipitators",
Journal of the Air Pollution Control Association, 7.. (1), 39-43 (May 1957).
"Smoke Eaters for Nine Fiery Furnaces", Air Repair, ~ (3), 113 (February 1953).
Akerlow, A. K., "Design and Construction of Fontana Open Hearth Precipitators",
Iron and Steel Engineer, 34 (6), 131-138 (June, 1957).
Herrick, R. A., "A Baghouse Test Program for Oxygen Lanced Open Hearth
Fume Control", Journal of the Air Pollution Control As sociation, 13 (1), 28-32
(January 1963). -
Herrick, R. A., et al., "Oxygen-Lanced Open Hearth Furnace Fume Cleaning
With a Glass Fabric Baghouse", Journal of the Air Pollution Control As sodation,
~ (1), 7-11 (January 1966).
Speer, E. B., "Operation of Electrostatic Precipitators on O. H. Furnaces at
Fairless Works", Air and Water Pollution in the Iron and Steel Industry, Special
Report No. 61, The Iron and Steel Institute. (1958), pp. 67-74.
Dickinson, W. A., and Worth, J. L., "Waste Gas Cleaning System at Sparrows
Point Plant's No.4 Open Hearth", AIME Open Hearth Proceedings, 47, 214-225
(1964). -
Johnson, J. E., "Wet Washing of Open Hearth Gases", Iron and Steel Engineer,
44 (2), 96-98 (February, 1967).
"L-D Process Newsletter", Kaiser Engineers Division of Kaiser Industries,
Newsletter No. 46, October 21, 1968.
"Annual Statistical Report, American Iron and Steel Institute, 1967", American
Iron and Steel Institute, New York City, 1968.
Gaw, R. G., "Gas Cleaning", Iron and Steel Engineer, 22. (10), 81-86 (October
1960) .
Koenitzer, F., and Zimmermann, K. A., "Gas Cooling in the Oxygen Steelworks
of August Thyssen-Hutte A. G. ", Engineering Experience in Oxygen Steelworks,
Publication 98, The Iron and Steel Institute. (1966), pp. 8-9.
Rengstorff, G. W. P., "Factors Controlling Emissions From Steelmaking Pro-
cesses", AIME Open Hearth Proceedings, 45, 204-219 (1962).
Kosaka, M., and Minowa, S., "Effect of Rate of Carbon Elimination Upon the
Formation of Oxide Fumes in Oxygen Blowing", Tetsu-to-Hagane, 50 (11), 1735-
1738 (1964). -
Watkins, E. R., and Darby, K., "The Application of Electrostatic Precipitation
to the Control of Fume in the Steel Industry", Fume Arrestment, Special Report
83, The Iron and Steel Institute (1964), pp. 24-35.
McShane, W. P., and Bubba, E., "Automatic BOF Stack Monitoring", 33/The
Magazine of Metals Processing, ~ (5), 97-104 (May 1968).

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
(94)
(95)
(96)
(97)
(98)
(99)
( 100)
( 101)
(102)
(103)
(104)
(105)
( 106)
( 107)
( 108)
( 109)
C-131
Kalling, B., et al., "Metallurgical Characteristics of the Kaldo Oxygen Steel-
making Proces s", Preprint, AIME Open Hearth Meeting, Chicago, Illinois
(April 1960), 19 pp.
Yocom, J. E., and Chapman, S., "The Collection of Silica Fume With a Venturi
Scrubber", Air Repair, ! (3), 155-158 (November 1954).
Trenkler, H., and Hauttmann, H. F., "LD-Process of Steelmaking With Oxygen
Jet", Metals Progress, 69 (1), 49-56 (January 1956).
Behrendt, A., "Gas Cleaning in Relation to Oxygen Pre-Refining and the Rotor
Proces s at Oberhausen", Air and Water Pollution in the Iron and Steel Industry,
Special Report 61, The Iron and Steel Institute (1958), pp. 90-96.
Smith, J. H., "Air Pollution Control in Oxygen Steelmaking", Journal of Metals,
~ (9), 632-634 (September 1961).
Loughrey, D. R., "The Basic Oxygen Process at Jones and Laughlin", AIME
Open Hearth Proceedings, 42, 274-285 (1959).
Massobrio, G., and Santini, F., "Some Starting and Operating Experiences With
the 300-Ton Oxygen Furnaces at the Taranto Works", AIME Open Hearth Pro-
ceedings, 48, 115-119 (1965).
Ziercher, J. L., Report on Official Travel, Battelle Memorial Institute,
Columbus Laboratories, October 10, 1968.
Dormsjo, T. 0., and Berg, D.
Its Metallurgy and Economics",
1959) .
R., "The Kaldo Oxygen Steelmaking Proces s -
Iron and Steel Engineer, 36 (4), 67-77 (April
Krijgsman, M., "Recovery and Utilization of Dust From the Basic Oxygen Steel-
making Processl1, Blast Furnace and Steel Plant, ~ (4), 44-62 (April 1964).
Rudnitskii, Ya. N., et al., "Determining the Amount of Gases Evolved in Con-
verters and the Time for Which Oxygen Should be Injected in an Oxygen-Blown
Converter", Stal, No.1, pp. 15-20 (January 1968).
Vajda, S., "Symposium on Basic Oxygen Furnaces - Equipment Layout", Iron and
Steel Engineer, 37 (10), 73-78 (October 1960).
McMulkin, F. J., "Oxygen Steelmaking in Canada", AIME Open Hearth Pro-
ceedings, 38, 241-254 (1955).
Nelson, F. D., "Progres s of New Basic Oxygen Shops - 1. At Inland Steel Co. ",
Journal of Metals, ~ (7), 785-787 (July 1965).
"Sparrows Point, Another Basic Oxygen Shop for Bethlehem Steel'., Journal of
Metals, ~ (5), 556-557 (May 1966).
Holloway, W. P., "Progress of New Basic Oxygen Shops - II. At Wheeling Steel
Corp.", Journal of Metals, ~ (7), 788-792 (July 1965).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
( 110)
(111)
(112)
( 113)
( 114)
( 115)
(116)
(117)
( 118)
(119)
(120)
( 121)
( 122)
(123 )
( 124)
( 125).
( 126)
C-132
Jones, M. A., "Engineering Aspects of the Basic Oxygen Furnace Plant at Great
Lakes Steel", Iron and Steel Engineer, !!.. (3), 123-128 (March 1964).
Nickel, M. E., "Progress of New Basic Oxygen Shops - IV. At Wisconsin
Steel Works, International Harvestor Co. ", AIME Open Hearth Proceedings, 48,
131-135 (1965). -
"Electric Furnace Round-Up", 33/The Magazine of Metal Producing, i (6),
pp. 71-80 (June 1966).
"Everybody is Getting Into the 'High-Power' Act", The Iron Age, 202 (2), pp. 22-
23 (July 11,1968).
Bennett, K. W., "Steelmakers Turn to Electrics", The Iron Age, 201 (26), 56-57
(June 27, 1968).
Coulter, R. S., "Smoke, Dust, Fumes, Closely Controlled in Electric Furnaces",
The Iron Age, 173 (2), 107-110 (January 14, 1954).
Kane, M., and Sloan, R. V., "Fume Control-Electric Furnace Melting Furnaces",
American Foundryman, ~ (11), 33-35 (November 1950).
Brief, R. S., et al., "Properties and Control of Electric-Arc Steel Furnace
Fumes", Journal of the Air Pollution Control Association, .~ (4), 220-224
(February 1957).
Hohenberger, A., "Dust Removal From Electric Arc Furnaces", Stahl und
Eisen, ~ (15), 1001-1005 (1961).
Davies, E., et al., "The Control of Fume From Electric Arc Furnaces", Journ::ll
of the Iron and Steel Institute, 201 (2), 100-110 (February 1963).
Bintzer, W. W., "Design and Operation of a Fume and Dust Collection System
for Two 100-Ton Electric Furnaces", Iron and Steel Engineer, 41 (2), 115-123
(February 1964). -
Baum, A., "Removal of Dust From Electric Furnace Waste Gases", Stahl und
Eisen, 84 (23), 1497-1500 (1964).
Danielson, J. A., "Metallurgical Equipment", Air Pollution Engineering Manual,
Public Health Service Publication No. 999-AP-40. (1967), pp. 235-257.
Hoff, W. A., "Use of Hot Metal in Electric Furnaces", AIME Electric Furnace
Proc eeding s, .!.i, 293 - 295 (1956).
Schmudde, A. W., "Use of Hot Metal in Electric Furnaces", Journal of Metals,
~ (4), 501-503 (April 1966).
Finkl, C. W., "Dust and Fume Control in a Modern Melt Shop", AIME Electric
Furnace Proceedings, .!.i, 272-278 (1956).
Peterson, H. W., "Gas Cleaning for the Electric Furnace and Oxygen Process
Converter", AIME Electric Furnace Proceedings, .!.i, 262-271 (1956).

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
(127)
( 128)
(129)
( 130)
(131 )
(132)
(133 )
( 134)
(135)
( 136)
(137)
(138)
( 139)
( 140)
C-133
Campbell, W. W., and Fullerton, R. W., "Development of an Electric-Furnace
Dust-Control System", Journal of the Air Pollution Control Association, 12 (12),
574- 590 (1962). -
Pettit, G. A., "Electric Furnace Dust Control System", Journal of the Air
Pollution Control Association, ~(12), 607-609, 621 (December 1963).
Harms, F., and Riemann, W., "Measurement of Fumes and Dust Volumes From
70-Ton Electric Arc Furnaces Operated Partially on Oxygen", Stahl und Eisen,
82 (20), 1345-1348 (1962).
Hipkin, A. S., "Cleaning of Fume From Arc Furnaces", Air and Water Pollution
in the Iron and Steel Industry, Special Report 61, The Iron and Steel Institute
(1958), pp. 108-114.
Nestaas, 1., and Romslo, R., "Measurement of Particulate Emissions From
Scrap Pre-Heaters", Report for Schjelderups Industriovner A/S, Oslo, Norway,
by The Engineering Research Foundation, Technical University of Norway (1968),
10 pp.
Kahnwald, H., and Etterich, 0., "Determination of the Volume, Composition,
and Temperature of the Waste Gas and the Dust During Meltdown and Oxidati~n
by Oxygen Lancing in a IS-Ton Electric Arc Furnace", Stahl und Eisen, 83 (17),
1067-1070 (1963).
Walker, W. S., and Harris, T. H., "Some Operational Details of a Large Elec-
tric Melting Shop", Journal of the Iron and Steel Institute, 198, Part 1, 5-12
(May 1961). -
"How Electric Arc Furnaces Pay Off at Roanoke Steel", Carbon and Graphite
News, .!.Q (1), 6-7, 1966 (Metal Progress, 90 (6), December 1966).
Varga, J. Jr., Report on Official Travel, Battelle Memorial Institute, Columbus
Laboratories, August 27, 1968.
Jenison, R. E., "Lukens Steel Co. 's Electric Melt Shop Complex", Journal of 
Metals, .!.2. (6), 41-43 (June 1967).
"How Electric Arc Furnaces Pay Off at Bethlehem-Seattle", Carbon and Graphite
News, .!.Q (1), 2-3 (1966). (Metal Progress, 90(6), December 1966).
Ess, T. J., "Armco at Butler", Iron and Steel Engineer, 38 (8), A-2 - A-20
(August 1961).
Rankin, W. M., "Electric Furnace Steel Production, Houston Works, Armco
Steel Corp. ", Journal of Metals, 20 (4), 104-107 (May 1968).
"J & L Combating Air Pollution at Stainless and Strip Div. ", Iron and Steel
Engineer, 45 (3), 139-140 (May 1968).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
( 141)
(142)
(143 )
( 144)
( 145)
( 146)
( 147)
( 148)
(149)
( 150)
(151)
(152)
( 1 53 )
(154)
(155)
(156)
C-134
Venturini, J. L., "Historical Review of the Air Pollution Control Installation at
Bethlehem Steel Corporation's Los Angeles Plant", Preprint No. 68-134. Air
Pollution Control As sociation Annual Meeting, St. Paul, Minnesota (June 23 -27,
1968), 19 pp.
Hornak, J. N., "Vacuum Degassing - Why and How?", Iron and Steel Engineer,
42 (6), 73-79 (June 1965).
Hornak, J. N., and Orehoski, M. A., "Vacuum Casting of Steel'l, Preprint of
paper published in Journal of Metals, .!2. (7), 471-475 (July 1958).
Hornak, J.
Properties
(1958).
N., and Orehoski, M. A., "Effect of Vacuum Stream Degassing on
of Forging Steels", AIME Electric Furnace Proceedings, ~, 68-84
Hobson, J. D., "Results From a Pilot Plant for Vacuum Stream Degassing, and
Some Theoretical Consideration of the Process", Hydrogen in Steel, Special
Report 73, The Iron and Steel Institute (1962), pp. 30-61.
Schempp, E. G., Communication to J. Varga, Jr., Battelle Memorial Institute,
Columbus Laboratories, October 14, 1968.
Lehman, A. L., "Vacuum Stream Degassing", AIME Electric Furnace Pro-
ceedings, ~, 84-93 (1958).
Forster, G. B., "R-H Degassing'l, Journal of Metals, ~ (4), 628-633 (May
1966) .
Wilson, L. H., and Unick, T. F., "Ladle To Ladle Vacuum Stream Droplet
Degassing Facility and Operations At Sharon Steel Corporation", Blast Furnace
and Steel Plant, ~ (9), 823-832 (September 1965). .
Geogiadis, J. F., and Hendrick, R. B., "Method for Evaluation of Volatile Mold
Coatings", AIME Open Hearth Proceedings, 45, 326-337 (1962).
Bunting, R. L., "Lukens Cement Hot-Top Practice", AIME Open Hearth Pro-
ceedings, !!., 338-346 (1958).
Roloff, D. V., and Smith, K. V., "A Study of Hot-Topping Practice", AIME Open
Hearth Proceedings, 42, 18-29 (1959).
Bayers, W. E., and Boyle, C. D., "Exothermic Sideboard Hot Tops", AIME
Open Hearth Proceedings, 44, 430-443 (1961).
Wayne, T. J., "Automated Preparation of Low Volume Hot Tops", AIME Open
Hearth Proceedings, 47, 267-269 (1964).
Selky, J. L., and Bingham, R. C., "Disposable Hot Tops'l, AIME Open Hearth
Proceedings, 47,272-287 (1964).
Lamont, J. A., et al., "Investigation of Hot-Top Materials", Journal of Metals,
~ (6), 738-742 (1966).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
( 157)
( 158)
( 159)
(160)
(161)
( 162)
(163)
(164)
(165)
(166 )
( 167)
(168)
(169)
(170)
( 171)
C-135
Hoffman, A. 0., and Ziercher, J. L., Report on Official Travel, Battelle
Memorial Institute, Columbus Laboratories, November 14, 1968.
Hoffma~, A. 0., and Ziercher, J. L., Report on Official Travel, Batte).le
Memor-ial, Institute, Columbus Laboratories, December 19, 1968.
Hoffman, M. F., et al., "Argon Casting for Improving Steel Quality"', AIME

Electric Furnace P'roceedings, 43, 375-386 (1960).
\ -
Wilson, W., "Argon'Teernfng of Degassed Steel", Journal of Metals, Q (4), 350-
352 (1961).
Communication from K. L. Backhaus, Concast, Inc., to J. Varga, Jr., Battelle
Memorial Institute, C6lumbus Laboratories, January 31, 1969.
Communication from T. Sullivan, Mesta Machine Co., to J. Varga, Jr., Battelle
Memorial Institute, Columbus Laboratories, September 9, 1968.
Communication from V. Navis, Amsted Research Laboratories, to J. Varga, Jr.,
Battelle Memorial Institute, Columbus Laboratories, December 16, 1968.
Discussion to article: Glossbrenner, A. B., "Tirnken Steel and Tube Division's
Approach to Bloom and Billet Conditioning", AIME Metallurgical Society Confer-
ences, Vol. 13. Bar and Applied Products, ?:J.., Interscience Publishers, New
York, N. Y. (1961).
Trilli, L. J., "Hot Machine Scarfing of Semi-Finished Carbon Steels", AIME
Metallurgical Society Conferences, ~, Flat Rolled Products II: Semi-Finished
and Finished, 3-17 (1960).
Whittaker, R., and Long, R. L., "Factors Affecting the Yield of Free-Cuttil1;g
Steels at Park Gate", Optimization of Steel Product Yield, ISI Publication 107,
The Iron and Steel Institute. (1967), pp. 47-55.
Keefe, J. M., "Optimization of Yield in Wide Strip Rolling, Part 2: Ingot to
Pickled Coil", Optimization of Steel Product Yield, ISI Publication 107, The Iron
and Steel Institute. (1967), pp. 64-72.
McLean, C. J., "Control of Defects - Flat Rolled Products", Blast Furnace and
Steel Plant, 54 (3), 231-240 (March 1966).
Communication from A. L. Hodge, Linde Company, to J. Varga, Jr., Battelle
Memorial Institute, Columbus Laboratories, January 30, 1969.
Glossbrenner, A. B., "Timken Steel and Tube Division's Approach to Bloom
and Billet Conditioning", AIME Metallurgical Society Conferences, Vol. 13.
Bar and Applied Products, 21-25. Interscience Publishers, New York, N. Y.
(1961).
"Trends in Steel Pickling and Waste Acid Treatment", 33/The Magazine of Metals
Producing, i (3), 65-76 (March 1966).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
I
(172)
(173 )
( 174)
(175)
(176 )
(177)
(178)
(179 )
(180 )
(181 )
C-136
Thompson, H. J., "Conversion from Sulphuric to Hydrochloric Acid in a Hori-
zontal Pickle Line", Iron and Steel Engineer, 45 (2), 102-108 (February 1968).
Communication from P. R. Klauss, Swindell-Dressler Company to J. Varga, Jr.,
Battelle Memorial Institute, Columbus Laboratories, October 18, 1968.
Miltenberger, R. S., liThe Use of Hydrochloric Acid in Conventional Pickling
Facilities", Blast Furnace and Steel Plant, 53 (9), 833-836 (September 1965).
Lemke, E. E., et al., "Air Pollution Control Measures for Hot Dip Galvanizing
Kettles", Journal of the Air Pollution Association, .!.Q. (1), 70-77 (February 1960).
Penton's Foundry List, 1967-68, Penton Publishing Company (1967).
"Iron Roll Foundry Expansion Under Way at Bethlehem Plant", American Metal
Market, p. 5, September 8, 1964.
"Takes Delivery of 3 25-Ton Furnaces", Metal Working News, p. 13, October 11,
1965.
Sterling, M., "Foundry Air Pollution Problems ", paper presented at the National
Conference on Air Pollution, December 1966, ,Washington, D. C.
"Foundry Air Pollution Control Manual", American Foundrymen' s Society, Des
Plaines, Illinois (1956), p. 39.
Cowen, P. S., "Roundup on Air Pollution", Gray and Ductile News, Gray and
Ductile Iron Founders' Society, pp. 5-11 (December 1967).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

-------
D-l
SUMMARY AND CONCLUSIONS
In this appendix, the cost of air pollution control equipment is estimated
for various processes of the integrated iron and steel industry. Factors
affecting the performance of control equipment and the effect of performance
level on cost are discussed.
The purpose of the cost estimating presented here is to. provide average
data on the emission control cost per production unit, as an initial step in
the formulation of a model for calculating the national cost of air pollution
control in the industry. A technique of estimating costs for this purpose is
described and used. The more detailed technique for estimating the cost of a
specific installation to suit the particular conditions and specifications is
not called for here. Without that detail, as outlined below, these estimates
cannot and should not be used to determine the cost of control for a specific
installation. An engineering analysis for that purpose, using classical esti-
mating techniques, is available from a number of engineering firms, and should
be used if the cost of control is desired for, e.g., company A's furnace B in
ci ty C.
The costs presented in the tabulations here are based on process parameters
representative of average modern practice. .Costs will vary from plant to plant
depending on such specifics as raw material properties, details of the process
as applied, product properties desired, unusual materials of construction or
unusual combinations of equipment occasioned by special corrosion or abrasion
conditions, details of integration of equipment into plant lay-out, etc. Typical
methods of control which are, or could be, applied to the average process are
cost estimated. Typical options for gas cooling are incorporated in each system;
this choice, together with the average process effluent level (both gas quantity
and particle concentration), determines the capacity and power ratings of
equipment.
The effect on cost of unusual arrangements for combined control systems
and for area ventilation is discussed. The effects on cost of unusual adapta-
tions of equipment to existing plants and facilities are discussed.
Also presented here is an indication of the theoretically determined
difference in cost for a more effective control system of the types typically
used. This is not the cost of altering an existing installation, with its
specific needs. These cost differences also are intended as input to the
model for national costs. The development of the model with cost data from
systems as used today, or as designed, will yield the ultimate tool for
determining national cost, and provide an average .comparison to the initial
input data presented here.

Certain nominal unit costs have been established as bases for calculating
operating costs. These costs will vary with monetary fluctuations over the
life-span of equipment. In adjusting field data as input to the model devel-
opment calculations, these costs would be normalized. Electrical energy is
standardized at $50/installed HP per year (based on a standardized 330 oper-
ating days per year x 24 hours/operating day = 7,920 operating hours/year).
This corresponds generally to 3/4~/KWH for large motors. Labor cost is set
at $5.00 per manhour including all welfare and fringe costs. Ratios of real,
local costs to these standardized values may be used as factors for adjusting
reported operating costs.

-------
D-2
PRIMARY CATEGORIZATION OF COSTS
Many costs arise during the life span of an industrial project, from
the earliest planning to the final demolition of the obsolete plant.
These costs are commonly assembled into three categories:
1.
Capital Costs: Cash outlays associated with
planning, engineering, purchasing, construction
and startup of the installation. Such costs occur
only once during the life of the installation.
2.
Operating Costs: Charges associated with the
operation, maintenance and financing of the plant
during its period of productivity. These costs
are repetitive in nature, constituting a continual
flow of cash away from the operating organization.
3.
Demolition and Salvage Costs: Cash transactions
arising while the facility is being dismantled and
sold off. Some of the cash flows are expenses and
some are income. The algebraic sum constitutes
either the Demolition Cost or the Salvage Value
depending upon the direction of the net cash flow.
These items occur only once during the life of the
plant, in which respect they are related to the
Capital Costs in (1) above.
THE GENERAL STRUCTURE OF CAPITAL COSTS
The capital cost of any plant or facility is the sum of many separate
cash payments made to suppliers of component parts, to workers of many kinds
for their labor, to consultants and contractors for their services, to
shippers for transportation of components to the construction site, etc.
This can be expressed as
CT = cl + c2 + c3 + ...... + cn
where CT is the total capital cost and cl' c2' c3'
cash payments as described above.
. . . .. c
n
are individual
In most industrial projects, the total number of these payments, n, is
very large. To simplify accounting and cost analysis, they are usually
grouped into a relatively small number of categories. Each category contains
the amounts of all payments related to some recognizable sub-division or
functional aspect of the total project. If these categories are called

Cl' C2' C3' ...... CM'
Cl = cl + c2 + c3
C2 = c4 + Cs + c6 + c7
C = C + C + C
M e m n

-------
D-3
The sum of
individual cash
the categories.
the categories equals the total cost, provided
payment appears once and only once, in one and
That is,

CT = Cl + C2 + C3 + ......+ CM
that each
only one of
In a typical situation, the total number of payments, n,
the range of thousands, but M, the number of cost categories,
10 to 20. Various categorizing schemes may be, and have been
categories may be functional in nature:
might be in
might be only
used. Some
Cl =
materials cost
C -
2 -
field labor cost
C3 =

C4 =
engineering labor cost
freight cost
etc., etc.
In such a scheme, Cl is the sum of all materials costs on the project
while C2 is the total cost of all labor required for field erection of all
materials.
Another system of categorization can be based on major component parts
of the installation. Each category would include all costs (materials,
labor, freight, engineering, etc.) associated with one part of the plant.
C =
1
cost.of furnaces
C =
2
cost of gas cleaning equipment
C =
3
cost of water supply system
C =
4
cost of buildings
etc., etc.
The choice of categorization scheme is usually a matter of tradition
or convenience, as long as each cost item appears once and only once in the
array.
As was already noted, these cost items occur only once in the life of
the project. Moreover, these costs are clearly defined cash transactions,
determined by market action. Unlike certain components of operating cost
(discussed elsewhere), they are generally unaffected by accounting practices,
tax procedures and other constraints arising from policy de~isions.
The planning, engineering and construction of an industrial plant
often encompasses a time span of two to four years. During these years,
CUrreDG~ inflation may be great enough to have a significant effect on
costs arising in the later phases of the overall program.

-------
D-4
CATEGORIZATION SCHEMES
A typical categorization scheme for industrial projects divides
capital costs into the following ten categories:
1.
Material: This includes the purchase cost of all materials,
machines, component parts, cement, structural steel, etc.
which are required in the field to make up the complete
operating installation. Net purchase costs are often F.O.B.
point of origin.
2.
Erection Labor and Supervision: This item includes wages
and salaries, payroll taxes, welfare benefits, etc. for all
persons employed at the construction site in the installation
of the material items listed above. Some estimating pro-
cedures involve the preparation of separate estimates for
the labor force and for a supervisory group of engineers.
This procedure may be needed when supervision is supplied by
an organization that is not responsible for employing the
general construction personnel.
3.
Freight: This covers the cost of transporting all of the
materials from their respective points of origin to the
construction site.
4.
Special Tools: This cost category includes rental and
transportation charges for special tools or equipment that
may be needed at the construction site. Excavating equipment
and large hoisting machinery are often rented for brief periods
of time during a construction project because the amount of
work to be done by them does not warrant their outright
purchase for one job.
5.
Taxes and Insurance: This covers the payment of necessary
sales taxes, permits and charges for insurance protection
as required during the course of the project.
6.
Engineering: In this category are collected all of the costs
associated with the central engineering and design aspects of
the project. This includes the services of engineers and
other personnel involved in the design, purchasing, and
general management of the project. The costs of these
services are usually calculated at a standard rate which
provides for salaries, fringe benefits, occupancy, and
departmental overhead. In addition, this category can
logically include general overhead and fee to the central
engineering organization.
7.
Client Engineering and Coordination: During the design and
construction of a new plant installation, the company or other
organization which will operate the finished plant usually
participates in the engineering and general management of
design and construction. This may involve the preparation of
specifications, review of design drawings, review of purchase
orders, participation in development of field schedulea, etc.
Charges in this category should include those directly
associated with the personnel involved together with an
appropriate share of overhead.

-------
D-5
8.
Startup:
operation
it into a
This covers costs associated with preliminary
and test of the new equipment in order to bring
cQndition of reasonable operating efficiency.
9.
Inventory:. This working capital item covers those moneys
which must be tied up in inventory of raw materials, goods in
process, maintenance supplies, etc.
10.
Land: This covers the fair value for the land area assigned
to the installation. In general, it includes not only the
land occupied by processing or manufacturing equipment but
also land utilized for the storage of materials and products
immediately preceding and following the processing unit. In
building a new installation, there is often a cost arising
from the preparation of the land site. This may involve
grading, filling, removal of old structures, etc., and must
not be omitted from the calculation of total investment.
The total investment is the sum of all of the foregoing items. This
total will not be the total payment to contractors, suppliers, and construct1on
labor because of the presence of items 7 through 10 above. Nevertheless, all
of these items (1 - 10) make up the total investment required to erect the
new industrial installation and bring it into normal working condition. It
is this total investment cost which enters into the cost computations else-
where in this project.
The detailed arrangement and tabulation of individual cost items in
the total as defined above can be handled in a variety of ways. The
particular method selected is largely a matter of individual preferences.
This may be based upon accounting practices well established in earlier
projects or upon cost classifications needed for tax or operating control
purposes.
In one method of tabulating costs, the costs are arranged according to
the piece of equipment which is concerned; the cost of a single item of
equipment would include material, erection labor and supervision, freight,
engineering, etc. This leads to an estimate composed of a group of cost
figures which are the total installed costs of individual pieces of equip-
ment or of operating subsections.
Another approach to the. problem is based upon functional 1ines~ Costs
are arranged in accordance with the ten categories given above. This
method does not display the total installed cost of a single piece of
equipment but does reveal the total cost of each function. In particular,
it discloses the total cost of field labor and the total cost of engineering
services. Control of these two .functions is often considered to be an
important matter by the managers of engineering contracts. In general, this
latter scheme will be used in the present investigation, with some reduction
in the number of categories, because of the generalized nature of the
estimates involved.

-------
D-6
THE GENERAL PROBLEM OF CAPITAL COST ESTIMATING
The estimator of capital costs wishes to predict the total cost, ~,
of a new installation to be designed and constructed at some later time.
His principal working technique is that of extrapolation into the future,
using design data on the new plant and past cost experience with similar
facilities. Various estimating procedures are available, differing in
the amount of work they entail, and in the accuracy of the resulting
estimate. In general, the more laborious methods are needed to achieve
the more accurate results. As a result, a choice must be made in any given
case between estimating precision and estimating cost.
The most economical estimating procedures generally involve a direct
estimate of CT, total capital cost of a new plant, derived from a historical
record of the cost of similar plants. In practice, however, such methods
give very rough, imprecise estimates because the design of the new plant
usually differs in major respects from its predecessors. Since many com-
ponent parts of plants undergo only small changes with time, it is usually
possible to obtain more precise results by making separate estimates of
component costs.
Component cost estimates may be taken from historical records, or from
recent market quotations. Both of these sources unavoidably contain poten-
tial errors which pass along into the total cost, CT' The probable error
in the total is a weighted average of the probable errors in all of the
components. The weighting factors are based on the relative costs. of the
individual components, and the average is calculated as a root-mean square.
As might be expected, this procedure assigns greatest importance to the
most expensive components, with proportionately less emphasis on the cheaper
items. In practical estimating, therefore, high-priced components must
receive a great deal of attention if the final estimate .is to be accurate.
Smaller, inexpensive parts of the plant may be treated in a more approximate
manner without seriously distrubing the total cost.
One consequence of this situation is a trend toward increasing the
number of components to be estimated separately. It is argued that this
will eliminate high-priced components and improve estimating accuracy
because no single error will have a la~ge weighting factor associated
with it. Experienced estimators know tnis to be true, to some extent, but
that a limit exists which cannot be passed. Even though the number of
components separately estimated becomes very large, each component price
estimate (however small it may be) still contains its error. The percent
error in the total is always a weighted average of the percentage errors
in all the component prices.
The cost of preparing an estimate increases as the number of estimated
components increases. This acts to restrain the tendency toward enlargement
of the number of estimated components, which is reinforced by the unavoidable
total error even with a large number of components, as described in the
previous paragraph.

-------
D-7
It is important to realize that some design work must be done before
the actual estimating can begin. This design work, which is expensive,
must generate enough data about each component in the estimate to permit
the setting of a component price. For a fully detailed estimate, the
design cost may be almost as great as that needed for actual construction.
For example, a detailed estimate of the cost of foundations in an
industrial plant might cover the following items:
Earthwork:
Machine Excavation
Trench Excavation
Hand Excavation
Trucking and Hauling
Backfill
Deep Foundations:
Bearing piles
Sheet piles
Walers
Formwork:
Buildings, mats and piers
Spread footers
Grade beams
Footings
Walls - below grade
Walls - above grade
Heavy equipment
Heavy mats
Elevated slabs
Shored slabs and columns
Earth slab paving
Reinforcing:
Bar
Mesh
Miscellaneous
Steel:
Anchor bolts
Embedded steel
Embedded railroad tracks
Miscellaneous steel
Checker plate and grating
Concrete:
Buildings, mats and piers
Spread footers
Grade beams
Continuous footings
Walls - above grade
Walls - below grade
Heavy equipment
Heavy mats

-------
D-8
Concrete: (continued)
Elevated slabs
Shored slabs and columns
Earth slab paving
Fine grading for paving
Batch plant
Waterproof walls and piers
Vapor barrier and waterstop
Joint materials
Color, sealer and grout
Finishes:
Steel Trowel
Screed finish
Brick
Floor hardener
Separate estimates are to be made of labor and materials for each
of the above items. It is evident that this array requires the making of
a very detailed design together with an equally detailed compilation of
historical data.
The design and estimating of a complete plan on this basis is very
expensive, and will be done only under extremely competitive conditions.
Moreover, this kind of analysis is not possible until a specific plant
site has been selected and its characteristics have been determined.
The objectives of the present study can better be served by using a
small number of components. In fact, the elaborate detail set forth above
would be inappropriate because this work is not directed toward any single
plant location.
THE GENERAL STRUCTURE OF OPERATING COSTS
Two classes of transactions enter into the operating cost of a
manufacturing plant. One is composed of direct cash expenditures for
labor, raw materials, fuel, electric power, maintenance supplies, etc.
The other is a group of costs whose magnitudes are determined in part
by managerial policy decisions of various kinds. Depreciation charges and
gene~al overhead burden are of this latter type.
Like capital costs, the multitude of individual operating cost items
is usually arranged into a small number of categories. Some of these have
been mentioned in the preceding paragraph. Since operating costs, unlike
capital cost, arise continually over the many operating years of the plant's
life span, they are usually collected and reported for comparatively short
periods of time. Most often, the year is the time interval chosen for
steady-state analysis, in order to eliminate season effects and daily
fluctuations caused by minor events in plant operation. The total
operating cost over this period can then be related to the total production
during the same time to arrive at a useful value for the unit production
cost.

-------
D-9
Thus, if 0* is the unit production cost, ° is the total operating
cost, and W is the total number of units of proauction, all during a
given year,
°T
0* =-
W
0T = 01 + 02 + 03 + 04 + 05 + ...... + ON
where 01 =
operating labor cost
° =
2
raw materials cost
maintenance labor and materials
°3 =

°4 =
electric power cost
° =
5
depreciation charges
° =
6
working capital charges
etc., etc.
During the operating life-span of a typical plant, there will be
substantial changes in technology, administrative techniques, social
practices, markets, state regulations, interest rates, currency values, etc.
These evolutionary changes may lead to substantial modifications in the
unit production cost.
THE GENERAL PROBLEMS OF ESTIMATING OPERATING COSTS
The estimation of the cost of the two types of transactions described
on the previous page brings two different problems to the estimator. The
first of these, the direct cash outlays for labor, maintenance, power, etc.,
can best be handled with the help of historical records of actual plant
experience. Cost records are generally kept by operating companies for
organizational cost centers based on considerations of product management,
administration structure, etc. When these cost centers cover the equipment
of interest to the estimator, plant records can supply directly the historical
basis needed for close estimating. When the cost centers do not correspond
to the operating area being examined by the estimator, the direct con-
struction of an historical base is not possible. In that case the required
cost components must be arrived at by theoretical calculations, personal
recollections, intuition, etc.
Unfortunately for the present study, steel companies do not generally
maintain cost centers around their pollution control activities. It is
the general practice to use cost centers which include both production units
and control facilities within a single perimeter. At this time there are
only scattered cost data available on operating labor and maintenance for
air pollution control installations in the steel industry.

-------
D-IO
The second type of component in operating cost is that whose
magnitude is established by policy decisions relating to depreciation,
overhead, capital charges, etc. In principle, depreciation should be
based in a simple, non-controversial way on the actual life of the
equipment and its ultimate salvage value. In practice, the prediction
of the life and salvage value of pollution control equipment is uncertain.
Operating conditions are usually severe, maintenance practices vary, and
the danger of obsolescence is great. Depreciation rates therefore are
strongly influenced by policy.
The allocation of corporate overhead involves even more difficult
policy questions. Charges for the use of capital in control equipment
require predictions of interest rates, profitability of alternative
investments, and future credit rating. The estimator is clearly working
in a very imprecise area when he considers these problems.
As a result, operating cost estimating is inherently uncertain and
must not be expected to lead to results of high precision. The selection
of optimum or preferred pollution control equipment or processes should
not be based upon small differences between the operating cost of
alternative designs.
COSTS OF CONTkOL SYSTEMS
A control system is considered to be made up of all the items of
equipment and their auxiliaries which are used solely for the general
abatement of atmospheric pollution in the neighborhood of the steel
works. Typically this will include a collecting hood or gas collecting
pipe at the furnace, ductwork, spray cooler, dust collector, fan and
motor, and stack. Included also will be structural steel, foundations,
control instruments, insulation, piping, water treatment, and electric
power supply facilities for the entire gas cleaning system. (Water treat-
ment includes all those items required for gas cleaning water uses and
sufficient for avoiding a water pollution problem.) Excluded are those
equipment items which, while they may contribute to the functioning of
pollution abatement equipment, would be used for process or economic
reasons even if there were no pollution abatement requirements.
The cost of land occupied by pollution abatement equipment has not
been included. It is recognized that such land has a real value but a
satisfactory method for estimating it has not been established. Costs
associated with preparation of the site, start-up operations and working
capital are also not included. Certain portions of a control system
occupy or utilize parts of steel plant buildings and, therefore, might be
charged with a share of general building costs. This item has not been
estimated here. In calculating operating costs no attempt has been made to
allocate a portion of general overhead to control systems.
Capital and operating costs in the following tabulations are based upon
collectors whose efficiency can be relied upon to produce an outlet dust
loading of 0.05 grains/SCF of gas. A later section of the report contains a
discussion of the relationship between cost and collection efficiency. In
general, higher efficiency is more costly.

-------
D-ll
EFFECT ON COST OF MULTIPLE FURNACE INSTALLATIONS
Pollution abatement equipment becomes increasingly cheaper as the
number of furnaces that can be connected to one common control system
increases. The largest saving in multiple furnace situations can be
realized from the alternate scheduling sequence of furnace operation.
Furnace shops using arc furnaces and BOF vessels which have high and
low gas emission periods seldom reach peak conditions at the same time.
This is generally due to material handling which limits the furnaces in
a common shop to being charged and teemed in succession. These furnaces
can be lanced or blown alternately which will permit designing the
multiple furnace control system for a reduced thermal capacity and reduced
air volume. A single collecting system sized on this basis for two furnaces
will handle approx. 1~7 times more fume than for a single furnace appli-
cation and a three furnace system could be designed to handle 2.5 times
more. This results in a large reduction in capital cost, although there
is a loss in operating flexibility. Where flexibility is important, a
design compromise is possible by using a single collector whose size will
permit peak operations on all furnaces at the same time. When more than
three units are to be served by a common control system, it is best to
assume that several furnaces will have to be at peak load together.
The following tabulation is suggested as a rough guide for estimating
the effect of combinations on capital cost; for all processes and types
of control equipment.
1.
Separate collectors
on each furnace
1 2 3
100% 200% 300%
100% 170% 250%
Number of Furnaces
2.
One collector to handle
peak loads on all
furnaces at once.
3.
One collector to handle
only one peak load at a
time.
100%
140%
200%
For example, the capital cost for control equipment serving two
furnaces together, with both able to run at peak loads, would be 170%
of the cost for a single furnace installation.
SPECIAL PROBLEMS ENCOUNTERED WHEN INSTALLING
NEW CONTROL EQUIPMENT IN EXISTING PLANTS
Existing plants can and have been revamped to accommodate modern dust
collecting equipment. A grass roots plant affords the flexibility of selecting
and installing cleaning equipment and duct work for maximum efficiency and
minimum capital cost. Providing and adapting fume abatement equipment for an
existing plant can in many cases be very expensive, especially if satisfactory
land is not available for locating this new equipment. In many cases the fume
collecting equipment must be located on top of the building roof which requires

-------
D-12
strenghtening all the supporting columns and trusses. A more serious situation
would require placing the fans and motors at roof level. For wet scrubbers in
such a case the weight due to the much higher scrubber horsepower requirements
can become so costly that it might be necessary to use an inferior type of
collector. The second-best collector may ultimately cost the customer more in
capital expenditure, maintenance, operating cost and efficiency.
The most costly aspects of designing a new fume collection system for
an operating facility often involves unusual and unorthodox arrangements of
fume pickup at the furnace. This is in part due to lack of space for supports
and interference with existing structures or obstructing personnel, vehicle
and crane approaches. To avoid these conditions may require an alternate type
of pickup at the furnaces, which, in turn, could dictate the type of apparatus
used for separation of the fume and particulate matter from the gases. In one
existing arc furnace shop, for example, it was most desirable to employ a direct
shell tap extraction from the arc furnaces, but in this shop very little free
area would then have been left for water cooled ducts, spark boxes, or cooling
chambers. Consideration was given to running ducts under the teeming building
but this alternative would have been extremely expensive and would have caused
considerable shut-down time. Moreover, the added furnace roof loading would
have required expensive revamping to support the extra weight of the water-
cooled elbow. The only practical solution was to install roof-truss hoods over
the furnaces. This system had the inherent disadvantage of moving 4 to 5 times
the air volumes required by the direct shell tap. The additional air volume
resulted in much more capital investment on fans and cleaning equipment. This
fume collection system was a compromise design forced by the limitations of
existing facilities. A newly designed plant could include direct shell taps on
the furnaces, resulting in much less expensive equipment and in a considerable
reduction in operating expense.
Limited space around an existing plant may require locating new control
equipment on the roof, as previously discussed, or possibly in a location so
remote as to make the duct runs much longer than would otherwise be needed. The
capital cost increases because of the extra ducting and the added air friction
losses. The increased static pressure requires a larger fan of greater horse-
power. The result will be increased investment and higher operating costs for
the life of the system.
Older control equipment at existing plants has often influenced the
selection of a new type of collector system. A plant already operating with
wet scrubbers will, if at all possible, try to adapt the new system to similar
cleaning equipment especially if there is enough reserve built into the plant
slurry system to handle the additional loading. Maintaining the same pattern
of equipment will not necessarily be the best selection for the process involved
and may increase the capital outlay as well as operating costs of the new
installation. A grass-roots plant is not generally as much influenced by such
continuity factors.
The amount of shut-down time required to install the fume collectors
will add to the cost of the installation by the amount of lost production. The
amount of down time that can be tolerated will influence the type, location and
duct routeingof the system. Any deviation from the most direct design such as
would be practical for a grass roots plant will therefore add cost to the install-
ation.

-------
,
D-13
Detailed design and estimate studies will usually be needed if cost
estimates are desired for new control equipment in existing plants. The
general cost data given in this report will not be reliable in such cases.
DATA SOURCES. PROCEDURE. PRECISION OF RESULTS
The primary data used in the preparation of cost estimates were
taken from a number of sources. These were:
1.
Estimate files of the Swindell-Dressler Company.
These files ranged in age from the immediately
current back to 1962. All costs were adjusted for
price escalation to bring them to present levels.
2.
Cost information supplied by certain steel companies
relating to their own plant facilities. Such data were
generally in the form of total costs for complete
installations. These were adjusted for inflation
by means of the same factors used on the estimate
data from Swindell-Dressler files.
3.
Cost information supplied by certain manufacturers
of pollution control equipment. This information
was presented in response to specific requests by
Swindell-Dressler and came in the form of budget
figures.
The Swindell-Dressler file data are detailed estimates prepared to meet
the requirements of particular competitive situations. Each estimate
assembled the costs for a specific location at a specific moment in time.
and this was done in much detail based upon a combination of firm price
quotations from suppliers and historical data about labor productivity at
the location in question. Each estimate, therefore, contains cost elements
which are influenced by local conditions not necessarily applicable to
other plants and geographical locations. A simple compilation of these
estimates would not have been adequate for the needs of the present investi-
gation.
The data from these many particular cases have been rearranged into
a more generalized form. The primary tactic used here has been that generally
followed by other government agencies and students of cost estimating.
This is to segregate the capital costs of the principal items of equipment
in the installation and to prepare smoothed, adjusted values for these
equipment items over a wide range of operating capacities. As is
well known. such smoothed values usually form a straight line graph on log-
log paper, with the slope of the line being related to certain characteristics
of the type of equipment involved. In this study smoothed material cost
data did give satisfactory linear graphs with slopes that were reasonably
related to the type of equipment.

-------
D-14
Following the practice of others in this field, the minor and bulk
materials were estimated on the basis of ratios applied to the costs of
the principal items. Labor costs were estimated for each principal item
and bulk category by the use of standard Swindell-Dressler factors
relating labor to material costs. The resulting labor figures apply to
the Pittsburgh area but may be adjusted for other locations through the
use of regional labor indices.
The capital costs included only facilities for loading the collected
dust or sludge into trucks for transportation elsewhere. No other disposal
costs or by-product values have been assigned. Central engineering costs,
overheads and fees were based upon a standard sliding scale generally used
by contract engineers.
It is believed that the general precision of the capital cost estimates
is such that most specific plant situations will fall within %15% of the
tabulation values. In more statistical terminology it might be suggested
that the standard deviation is about %10 - 12%. It is to be expected that
any specific plant location which presents unusual cost problems associated
with layout, structure, power supply, etc. might fall outside these limits.
In such cases a detailed plant design and estimate should be prepared if
accurate capital cost data are required. As previously noted, the accuracy
of operating cost values is influenced by many factors which may vary
considerably from one company to another. The selection of control equip-
ment should not be based upon small differences in operating cost estimates.
Operating cost estimates present tabulated costs for the following
items:
1.
Electric Power
2.
Maintenance
3.
Operating Labor
Direct Operating Cost - The sum of above three items.
4.
Depreciation
5.
Capital Charges
The items included in Direct Operating Cost are those cost elements
which are the direct cash outlays discussed on page C-8. They are, to some
extent, under the control of the plant operating management. The costs
assigned for Depreciation and Capital are, as noted on page C-lO based upon
policy decisions not generally under the control of management at the plant
operating level.
Electric energy is calculated at a standardized rate of 3/4~ per kilowatt
hour. The cost of make-up water is not included as such. Operating labor
cost was calculated at the nominal value of $5.00/man hour including all
welfare and fringe costs.

-------
D-15
Maintenance is taken at a nominal cost of 4% of the total investment
shown in the capital cost tables. This figure has been the subject of
considerable discussion and has been retained in this analysis because it
is believed to represent a reasonable value over the total life of the
equipment. Most steel plant maintenance cost records do not make a
separate accounting for each pollution control installation. It is,
therefore, difficult to arrive at an exact, numerical evaluation of total
maintenance during the life of a piece of control equipment. In actual
practice, maintenance expenditures are not uniform from year to year.
In many cases major maintenance outlays occur only after the passage of
several years of operation. Moreover, as might be expected, maintenance
costs usually increase during the service life of an item of control
equipment. There have been some cases reported where major costs were
experienced early in the life of a control installation. It is believed
that these cases should properly be attributed to inadequate engineering
rather than standard maintenance. These incidents were more common some
years ago when knowledge of pollution control engineering was not as .
extensive as it is today. There have also been cases reported in which
major modifications were made to control equipment after it had been in
service for some years. Some times these episodes were caused by changes
in the operating practice of the process segment which placed greater
burdens on the control equipment. This type of cost is not considered to
be a part of maintenance. The 4% figure is retained in this study be-
cause it is believed to represent a reasonable value for good maintenance
in well designed equipment when calculated over the entire life of the
installation.
In those cases where filter bag replacement represents a major
maintenance cost item, the system, less bags, is given the 4% maintenance
charge; and the cost in material and labor for bag replacement at a reason-
able average rate (18 months for Sinter plants, 2 years for steelmaking
shops) is added for the net maintenance cost listed.

Depreciation is calculated on a straight line method using total
investment with an expected life of ten years. Other studies of depreciation
have suggested longer service life times, but these are considered to be
greater than average plant experience will confirm. Advancing technology
and rising standards give importance to the factor of technical obsolescence.
Capital charges are taken at 10% per annum. It is believed that this
will be reasonable in the light of rising interest rates and local taxes.
Annual calculations are based on 330 operating days per year, 24 hours
per day. This gives a total of 7,920 operating hours per year.
ALTERNATE SYSTEMS
In the tabulations which follow, several alternate control systems are
included, for most processes. Aside from cost, other factors enter into the
selection of a system. The nature of the process to some extent dictates or
precludes the use of a particular type of control. For example, some collected
dusts can be reused directly in the process or used as burden in the plant

-------
D-16
following agglomeration; the wet or dry state of the collected dust may afford
a convenience to disposition of the dust according to the current practice of
a particular plant.
The gas may be reusable as a process material in the plant, where changes
in its temperature and humidity by the cooling system would have to be consid-
ered in the total plant energy economy. The local cost of treated water, space
requirements for retreatment facilities, and possible difficulties in using
water near the process vessel may affect the choice of wet or dry systems.
The particle size and concentration of the effluent determine whether
high efficiency gas cleaning equipment (high-energy wet scrubber, electrostatic
precipitator, or baghouse) is needed, based on particle size vs. efficiency
experience data for different types of collectors. This data is largely in
the form of proprietary design curves in the files of equipment manufacturers.
For the purposes of this study, the dividing line between low-and high-energy
wet scrubbers is 12 inches of water pressure drop across the collector, with
high-energy applications generally using several times this pressure drop.
The nature of the dust collector equipment to be used largely determines
the extent and method of cooling the process gas. The process effluent may
vary in temperature from IOO°F for material transfer point ventilation to
3000°F or higher for furnace exhaust. This gas may be quenched by air dilution
or water sprays to a lower temperature, or undergo a heat exchange to cool
without adding material to the effluent stream. If the gas is combustible,
it may initially be burned, with an excess of air to insure completness of
combustion, where a fire or explosion hazard would exist. A wet-type cleaner
may treat water-quenched or hot gases directly. The electrostatic precipi-
tator used on highly (electrically) resistive particles requires a degree of
cooling and humidification control to be effective and of economical construc-
tion. On the other hand, over-cooling or over-quenching can result in conden-
sation with resultant corrosion, collector surface fouling, and dust handling
problems in dry precipitators, and baghouses as well. Thus, in general
- excess air is added to the effluent stream where combustion occurs,
in a water-cooled or other heat-exchange vessel;
- water addition completes the cooling for a wet system;
- indirect cooling by heat exchange will provide the most economical
cooling to about 500°F in dry systems;
- added humidification by water sprays usually completes the treatment
of gases prior to electrostatic precipitation;
- air dilution for bag temperature control usually completes pre-
baghouse cooling.
Finally, to achieve economical fan power levels, the gas volume is kept
low by gas cooling especially with high-energy wet scrubbers. This ultimate
effluent gas, if sulfur oxides persist in significant degree to this point in
the system, must have sufficient lift in the form of thermal or mechanical
energy, or stack height to disperse in the atmosphere.

-------
D-17
SINTER PLANTS
The following tables contain capital and operating cost data for two
siz~~ of sinter plants. Sinter plant control systems are usually designed so
that one control unit handles gases coming from the windbox while a separate
control unit receives dust collected at several points in the material'
handling system.
The attached estimate gives separate figures for the windbox and
materials handling operation. Various combinations of types of collection
equipment are used on sinter plants and it is therefore necessary to offer
separate values for these two zones of collection. The total cost for a
given sinter plant will be the sum of the cost for the windbox and the cost
for the materials handling.
The tables do not include system components through recovery cyclones
or windbox fans. Booster fans are included. Modifications only are in-
cluded in the windbox stack item cost.
The capacity of the sintering machine for the tabulated gas volume
and dust collection cost is based on the average nominal capacity, making
normal sinter. Variations will occur with differences in the burden. For
example, self-fluxing sinter capacity may be as much as 35 percent higher
than a machine's normal 'capacity. (Symposium on Sinter Plants, Discussion,
Iron and Steel Engineer, June, 1959.) In other reports no such change is
noted. With self-fluxing sinter, more particulate matter passes through
the cleaner.
The addition of oily turnings and borings to the burden generates oil mist
in the windbox gas. One solution to this is the use of.a very high energy
wet scrubber system.
The temperature of the windbox gases is determined
process used on the machine. The gases are moist. Too
can result in corrosion -causing condensation and tacky
problems in dry collectors.
by the sintering
low a temperature
dust handling

-------
D-18
SINTER PLANT (WINDBOX) - WET SCRUBBER (LOW ENERGY)
Gas Volume - ACFM @ 325°F
Plant Capacity - TPD
1. Materia1*
2. Labor 
3. Central Engineering
4. Client Engineering
  TOTAL
105,000
1,000
CAPITAL COST
$193,000
100,000
72 ,000
18,000
$383,000
OPERATING COST ($/Yr.)
1.
Electric Power
$ 20,000
2.
Maintenance
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
15,000
30,000
$ 65,000
38,000
38,000
$141,000
Note:
1)
Prices: 1969 base.
For items included in materials, see pages C-10 and C-17.
*
630,000
6,000
$880,000
440,000
245,000
61,000
$1,626,000
$110,000
66,000
80,000
$256,000
163,000
163,000
$582,000

-------
D-19
SINTER PLANT (WINDBOX) - ELECTROSTATIC PRECIPITATOR 
Gas Volume - ACFM @ 3250F
Plant Capacity - TPD
1.
Materia1*
105,000
1,000
CAPITAL COST
$180,000
88,000
72 , 000
18,000
$358,000
OPERATING COST ($/Yr.)
$ 12,500
14,500
20,000
$ 47,000
Direct Operating Cost
Note: 1) Prices: 1969 base.
36,000
36,000
$119 , 000
* For items included in materials, see pages C-10 and C-17.
2.
Labor
3.
Central Engineering
4.
Client Engineering
TOTAL
1.
Electric Power
2.
Maintenance
3.
Operating Labor
4.
Depreciation
5.
Capital Charges
TOTAL
630,000
6,000
$800,000
385,000
225,000
56,000
$1,466,000
$ 77,000
 59,000
 30,000
$ 166,000
 147,000
 147,000
$ 460,000

-------
D-20
SINTER PLANT (WINDBOX) - FABRIC FILTER
Gas Volume - ACFM @ 3250F
105,000
1,000
Plant Capacity - TPD
CAPITAL COST
1. Materia1*
2. Labor 
3. Central Engineering
4. Client Engineering
  TOTAL
$154,000
72,000
58,000
14,000
$298,000
OPERATING COST ($/Yr.)
1.
$
8,500
Electric Power
2.
Maintenance
18,000
3.
Operating Labor
20,000
Direct Operating Cost
46,500
4.
Depreciation
30,000
5.
Capital Charges
30,000
$ 106,500
* For items included in materials, see pages C-IO and C-17.
Note :l)This system is rarely used.
2) Prices: 1969 base.
630,000
6,000
$800,000
340,000
222,000
55,000
$1,417,000
$ 45,000
 87,000
 30,000
 162,000
 142,000
 142,000
$ 446,000

-------
D-21
SINTER PLANT (MATERIAL HANDLING) - WET SCRUBBER (LOW ENERGY)
Gas Volume - ACFM @ 1350F
Plant Capacity - Tons/Day
CAPITAL COST
1.
Materia1*
48,000
1,000
$194,000
146,000
81,000
20,000
$441,000
OPERATING COST ($/Yr.)
$ 36,000
Note:
1) Prices: 1969 base.
2.
Labor
17,600
15,000
$ 68,600
44,000
44,000
$156,600
* For items included in materials, see pages C-lO and C-l7.
3.
Central Engineering
4.
Client Engineering
TOTAL
1.
Electric Power
2.
Maintenance
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
250,000
6,000
$420,000
480,000
184,000
46,000
$1,130,000
$
95,000
45,500
40,000
$
180,500
113 , 000
$
113 , 000
406,500

-------
D-22
SINTER PLANT (MATERIAL HANDLING)-ELECTROSTATIC PRECIPITATOR
Gas Volume - ACFM @ 1350F
Plant Capacity - Tons/Day
1,000
250,000
6,000
48,000
CAPITAL COST
1. Material'\-
2. Labor 
3. Central Engineering
4. Client Engineering
  TOTAL
$149,000
$420,000
91 , 000
195,000
60,000
133,000
15,000
$315,000
33,000
$781,000
OPERATING COST ($/Yr.)
1.
Electric Power
$ 7,000
$ 41,000
2. Maintenance   12,500 
3. Operating Labor   15,000 
   Direct Operating Cost  $34,500 
4. Depreciation   31,500 
5. Capital Charge   31,500 
   TOTAL   $97,500 
*For items included in materials, see pages C-IO and C-17.
Note: 1) Prices: 1969 base.   
31,000
20,000
$92,000
78,000
78,000
$248,000

-------
D-23
SINTER PLANT (MATERIAL HANDLING) - FABRIC FILTER
Gas Volume - ACFM @ 1350F
48.000
1.000
Plant Capacity - Tons/Day
CAPITAL COST
L.
Materia1*
$120.000
2.
Labor
68,000
3.
Central Engineering
49.500
4.
Client Engineering
12.500
TOTAL
$250,000
OPERATING COST ($/Yr.)
1.
$
9,000
Electric Power
2. Maintenance 12,800
3. Operating Labor 15.000
 Direct Operating Cost 36 , 800
4. Depreciation 25,000
5. Capital Charges 25.000
 TOTAL $ 86,800
*For items included in material, see pages C-IO and C-17.
Note:
1)
Prices: 1969 base.
250.000
6.000
$350,000
166,000
116,000
29.000
$661,000
$ 38,000
36,500
20.000
94,500
66,000
66.000
$ 226,500

-------
D-24
PELLETIZING PLANTS
The following tables contain cost data for a pelletizing plant
of 1,500,000 tons per year. This is a commonly used plant
capacity with larger output being achieved through use of parallel
units. It is not likely that many pelletizing plants will be
built whose capacity is less than that shown here. It is believed
that the costs presented are reasonably typical of the several
types of moving grate equipment now in use.
Like the sinter plant, several control systems are used at different
points on the unit. The total cost is the sum of the cost at the
dryer exhaust and the materials handling dust points.
Many pelletizing plants hold to the shaft furnace design, using
multiples of the 60 ton/hr. furnace. A system of cyclones is
included in this section for the cleaning of the process gas
leaving the furnace. And also, air from the cooling unit and material
handling points at the discharge station is cleaned separately.

-------
D-25
PELLETIZING PLANT (MOVING GRATE - DRYER EXHAUST) - CYCLONE
Gas Volume - ACFM @ 2500F
320,000
1,500,000
Plant Capcity - Tons/Year
CAPITAL COST
1.
$ 180,000
95,000
Materia1*
2.
Labor
3.
69,000
Central Engineering
4.
Client Engineering
17,000
$ 361,000
TOTAL
OPERATING COST ($/Yr.)
1.
$ 20,000
 14,000
 30,000
$ 64,000
 36,000
 36,000
$ 136,000
Electric Power
2.
Mainten~lnce
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
*
For items included in materials, see pages C-10 and C-24.
Note:
1)
Prices: 1969 base.

-------
D-26
PELLETIZING PLANT (MOVING GRATE ~ MATERIAL HANDLING)
- WET SCRUBBER (LOW ENERGY)
o
Gas Volume - ACFM @ 70 F
55,000
Plant Capacity - Tons/Year
1,500,000
CAPITAL COST
1.
Material*
$
80,000
2.
Labor
54,000
3.
Central Engineering
36,000
4.
Client Engineering
9,000
TOTAL
$ 179,000
OPERATING COST ($/Yr.)
1.
Electric Power
$ 22,000
 7,000
 10,000
$ 39,000
 18,000
 18,000
$ 75,000
2.
Maintenance
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
*
For items included in materials, see pages C-IO and C-24.
Note:
1)
Prices: 1969 base.

-------
D-27
PELLETIZING PLANT (SHAFT FURNACE - PROCESS EXHAUST)-CYCLONES
Gas Volume - ACFM @ 4600F
125,000
60
Plant Capacity - Tons/Hr.
CAPITAL COST
1.
$135,000
Materia1*
2.
Labor
82,000
3.
Central Engineering
55,500
4.
Client Engineering
13,500
$286,000
TOTAL
OPERATING COST ($/YR.)
1.
Electric Power
$ 23,000
2.
Maintenance
11,300
3.
Operating Labor
15,000
$ 49,300
Direct Operating Cost
4.
Depreciation
28,600
5.
Capital Charges
28,600
.$106,500
TOTAL
*
For items included in materials, see pages C-IO and C-24.
Note:
1)
Prices: 1969 base.

-------
D-28
 PELLETIZING PLANT (SHAFT FURNACE - MATERIAL HANDLING) 
       CYCLONES AND WET SCRUBBER
         (LOW ENERGY)
Gas Volume - ACFM @ 700F  30,000  19,000
Plant Capacity - Tons/Hr.   60 
      CAPITAL COST   
1. Materia1*    $ 45,000  $35,000
2. Labor     32,000  22,000
3. Central Engineering  23,000  19,000
    /
4. Client Engineering  6,000  5,000
       $106,000  $81,000
" TOTAL    $187,000 
 OPERATING COST ($/Yr.)   
1. Electric Power $ 4,000  $ 7,500
2. Maintenance  4,000  3,500
3. Operating Labor  5,000  5,000
  $ 13,000  $16,000
 Direct Operating Cost   $ 29,000 
4. Depreciation  10,500  8,000
5. Capital Charges  10,500  8,000
  $ 34,000  $32,000
    $ 66,000 
* For items included in material, see pages C-IO and C-24.
Note: 1)
15000ACFM capacity for once a week cleaning routine.
2)
Prices: 1969 base.

-------
D-29
COKE OVEN
Cost data are not presented for control of emissions from coke ovens.
The engineering problems involved are still being investigated, both in
the U.S.A. and abroad. Satisfactory control equipment, with proven
industrial performance, has not yet been developed.

-------
D-30
BIAST FURNACE
The attached table presents cost information for a typical modern large
blast furnace. It is anticipated that most future blast furnaces in the
United States will be of this size or greater. They will probably have
the type of wet scrubbing system shown here. Older units with combinations
of several types of control equipment are not likely to be copied in the
future.
Blast furnace gas cleaning costs should be divided between emission control
and normal plant operation. In the absence of an industry consensus, it is
suggested that an equal share be allocated to each of these accounts. The
portion of the top gas which is fine-cleaned for the blast furnace stoves
is shown for a number of plants on pages C-3l to C-37 of the technical
counterpart of this report, entitled, "Final Technological Report on a
Systems Analysis Study of the Integrated Iron and Steel Industry," May 15, 1969.

-------
D-31
BLAST FURNACE - WET SCRUBBER (TWO STAGE. HIGH ENERGY)
Wind Rate - SCFM
150,000
210,000
Gas Volume - SCFM
CAPITAL COST
1.
$1,427,000
Materia1*
2.
Labor
636,000
3.
Central Engineering
360,000
4.
Client Engineering
90,000
$2,513,000
TOTAL
OPERATING COST ($/Yr.)
1.
$ 20,000
 100,000
 40,000
$ 160,000
 251,000
 251,000
$ 662,000
C-IO and C- 30.
Electric Power
2.
Maintenance
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
*
For items included in materials, see pages
Note:1)These are total costs of cleaning the furnace top gas
of particulate matter. Since this operation serves
the ends of both emission control and plant operational
requirements (material recovery and fuel conditioning
for re-use), a share of the cost should be apportioned
to each account. It is suggested, in the absence of an
industry consensus, that the shares be equal.
2)Prices: 1969 base.

-------
D-32
BASIC OXYGEN FURNACE
The following pages contain cost data on several sizes of basic oxygen furnaces.
They assume that a new plant is being designed and that the pollution control
equipment is included in the original design. The figures cover a single furnace
only. It is recognized that various combinations of multiple units are used
in actual practice. The influence of this is discussed on page C-ll.
Heat extracting hoods ar~ included. These are total combustion systems
for typical oxygen-blow rates, with excesa air used for a portion of the
cooling. Water additions for saturation in wet scrubber systems and for
humidification (considerably less than for saturation) in electrostatic pre-
cipitator systems completes the cooling typically. For baghouse systems, the
gas would be kept dry, using air dilution at the hood and before the baghouse
with heat exchange means between.

-------
D-33
BASIC OXYGEN FURNACE - WET SCRUBBER (HIGH ENERGY)
Gas Volume - ACFM @ 1800F   220,000  440,000  660,000
Furnace Size - Tons   100  200  300
  CAPITAL COST    
1. Materia1*    $910,000 $1,460,000 $1,960,000
2. Labor     490,000  790,000  1,060,000
3. Central Engineering   250,000  390,000  470,000
4. Client Engineering   60,000  100,000  120,000
  TOTAL $1,710,000 $2,740,000 $3,610,000
  OPERATING COST ($!Yr.)    
1. Electric Power  $ 207,000 $ 432,000 $ 664,000
2. Maintenance    68,000  110,000  145,000
3. Operating Labor   40,000  60,000  80,000
  Direct Operating Cost $ 315,000  602,000  889,000
4. Depreciation    171,000  274,000  361,000
5. Capital Charges   171,000  274,000  361,000
  TOTAL $ 657,000 $1,150,000 $1,611,000
* For items included in material, see pages C-IO and C-32.
Note:
1)
One Furnace System. For effect on cost of combined cleaning
systems on multiple furnace shops, see page C-ll.
2)
These estimates cover full combustion systems in which all of
the gas leaving the converter is mixed with an excess quantity
of air. All of the carbon monoxide is therefore burned to carbon
dioxide. This is the common industry practice in this country.
Systems have been designed which collect this gas in a substantially
unburned state. Such non-combustion systems may offer certain
economies. The exact extent of these economies has not yet been
generally recognized in the industry.
3)
Prices: 1969 base.

-------
D-34
BASIC OXYGEN FURNACE - ELECTROSTATIC PRECIPITATOR
a
Gas Volume - ACFM @ 500 F
375,000
100
785,000
Furnace Size - Tons
200
1,200,000
300
CAPITAL COST
3.
Central Engineering
250,000
$1,600,000 $2,250,000
800,000 1,100,000
410,000 550,000
100,000 140,000
$2,910,000 $4,040,000
1.
Material'\'
$ 900,000
2.
Labor
450,000
TOTAL
60,000
$1,660,000
4.
Client Engineering
OPERATING COST ($/Yr.)
1. Electric Power $ 90,000
2. Maintenance  66,000
3. Operating Labor  20,000
 Direct Operating Cost $ 176,000
4. Depreciation  166,000
5. Capital Charges  166,000
 TOTAL $ 508,000
$ 210,000 $ 310,000
 116,000  162,000
 30,000  40,000
$ 356,000 $ 512,000
 291,000  404,000
 291,000  404,000
$ 938,000 $1,320,000
* For items included in material, see pages C-IO and C-32.
Note:
1) One Furnace System. For effect on cost of combined cleaning systems
on multiple furnace shops, see page C-ll.
2) Prices: 1969 base.

-------
D-35
   BASIC OXYGEN FURNACE - FABRIC FILTER  
Gas Volume - ACFM @ 2750F 288,000 600,000 892,000
Furnace Size - Tons 100 200 300
    CAPITAL COST  
1. Materia1* $ 660,000 $1,280,000 $1,840,000
2. Labor  360,000 690,000 990,000
3. Central Engineering 200,000 340,000 470,000
4. Client Engineering 50,000 90,000 120,000
   TOTAL $1,270,000 $2,400,000 $3,420,000
  OPERATING COST ($/Yr.)    
1. Electric Power $ 43,000 $ 89,000 $ 130,000
2. Maintenance   59,000  112,000  160,000
3. Operating Labor  20,000  30,000  40,000
 Direct Operating Cost $. 122,000 $ 231,000 $ 330,000
4. Depreciation   127,000  240,000  342,000
5. Capital Charges  127,000  240,000  342,000
  TOTAL  $ 376,000 $ 711,000 $ 1,014,000
* For items included in material, see pages C-IO and C-32.
Note:
1)
One Furnace System.
on multiple furnace
For effect on cost of combined cleaning systems
shops, see page C-ll.
2)
This system is used in Europe, but so far has not had an American
application.
3)
Prices: 1969 base.

-------
D-36
OPEN HEARTH FURNACE
The following tables contain cost data for open hearth furnaces. They are
based upon the addition of gas cleaning equipment to an existing furnace
shop. It is not likely that many new open hearths will be built in the
future. The figures shown are for a single furnace. The effect upon cost
of multiple furnace combinations are discussed on pageC-llof this report.
It is assumed that waste heat boilers and boiler fans are existing at the
furnaces, and needed stack modifications are included in the estimates,
along with booster fans. Waste heat boilers, while they contribute to
pollution control by cooling the gases (without adding additional material
to the gas stream which would increase size and cost of subsequent equipment),
also serve the plant energy economy, and have been in general use on open
hearth furnaces having no abatement equipment. Thus, they are not included
in the cost of air pollution control and no credit is assigned for steam
produced.
The estimates are based on averaged data for current oxygen-blown furnaces
of different sizes, charged typically with 50% hot metal, 50% cold scrap.
The typical gas cleaning equipment begins with boiler exhaust gas at 5000F
and 18% mOisturQ.(Steam augmentation is assumed during the dry gas period
after hot metal addition when fuel and atomizing steam rates are low, and
during low-rate initial oxygen lancing when gas temperature is low and the
checker water cooling sprays are not used.) Thus, temperature and humidity
control are minimized for dry gas cleaning systems. This gas volume per ton
furnace capacity diminishes on the average with increasing furnace capacity,
and is cleaned directly in an electrostatic precipitator system. The gas
is cooled by air dilution before a baghouse collector. The wet scrubber
saturates the gas.

-------
D-37
OPEN HEARTH FURNACE - WET SCRUBBER
(HIGH ENERGY)
Gas Volume - ACFM @ 1800F
30,000
90,000
240,000
Furnace Size - Tons
60
200
600
     CAPITAL. COST    
1. Materia1~'c'    $160,000 $430,000 $1,000,000
2. Labor      85,000  230,000  540,000
3. Central Engineering   60,000 .140,000  280,000
4. Client Engineering    15,000  35,000  70,000
   TOTAL   $320,000 $835,000 $1,890,000
     OPERATING COST ($/Yr.)    
1. Electric Power $ 24,000 $ 77,000 $ 210,000
2. Maintenance    13 , 000  33,000  76,000
3. Operating Labor    40,000  60,000  80,000
  Direct Operating Cost $ 77,000 $ 170,000 $ 366,000
4. Depreciation    32,000  83,500  189,000
5. Capital Charges    32,000  83,500  189,000
   TOTAL $.141,000 $ 337,000 $ 744,000
* For items included in materials, see pages C-IO.and C-36.  
Note: 1) One Furnace System. For effect on cost of combined  gas cleaning
   systems on multiple furnace shops, see page C-ll.  
2)
A variance from this design and cost is noted for a tar-fired
furnace with no waste heat boiler. Gas volume at higher temp-
eratures before and after saturation, and other factors lead
to a 60% higher cost.
3)
For a discussion of unusual problems encountered when installing
new collecting equipment at existing furnace shops, and an in-
dication of cost variances, see page C-ll.
4)
Prices: 1969 base.

-------
D-38
OPEN HEARTH FURNACE - ELECTROSTATIC PRECIPITATOR
o
Gas Volume - ACFM @ 500 F
29,000
85,000
225,000
Furnace Size - Tons
60
200
600
CAPITAL COST
70,000
$320,000 $700,000
170,000 380,000
110,000 200,000
30,000 50,000
$630,000 $1,330,000
1. Materia1*
2. Labor 
3. Central Engineering
4. Client Engineering
  TOTAL
$130,000
52,000
13,000
$265,000
OPERATING COST ($/Yr.)
1.
Electric Power
$
5,000
$ 15,000 $ 45,000
25,000  54,000
30,000  40,000
$ 70,000 $ 139,000
63,000  133,000
63,000  133 , 000
$196,000 $ 405,000
2. Maintenance  11 , 000
3. Operating Labor 20,000
 Direct Operating Cost $ 36,000
4. Depreciation  26,500
5. Capital Charges 26,500
 TOTAL $ 89,000
* For items included in material, see pages C-IO and C-36.
Note:
1)
One Furnace System. For effect of combined gas cleaning systems
on a multiple furnace shop, see page C-ll.
2)
For a discussion of unusual problems encountered when installing
new collectors at existing furnace shops, and an indication of
cost variances, see page C-l1.
3)
Prices: 1969 base.

-------
D-39
  OPEN HEARTH FURNACE - F ABRI C FILTER  
Gas Volume - ACFM @ 2750F 45,000 135,000 350,000
Furnace Size - Tons 60 200 600
  CAPITAL COST  
1. Materia1* $75,000 $210,000 $530,000
2. Labor  40,000 120,000 300,000
3. Central Engineering 36,000 80,000 180,000
4. Client Engineering 9,000 20,000 45,000
  TOTAL $160,000 $430,000 $1,055,000
1.
Electric Power
OPERATING COST ($/Yr.)
$ 8,000
2.
Maintenance
7,700
$ 22,000 $ 54,000
21,000  51,000
30,000  40,000
$ 73,000 $ 145,000
43,000  105,500
43,000  105,500
$159,000 $ 356,000
3.
Operating Labor
20,000
Direct Operating Cost
$ 35.700
4.
Depreciation
16,000
TOTAL
16,000
$ 67,700
5.
Capital Charges
* For items included in material, see pages C-IO and C-36.
Note:
1)
One Furnace System. For effect on cost of combined gas cleaning
systems on a multiple furnace shop, see page C-ll.
2)
This system is not currently in general use, but it has been
successfully applied in the U.S.
3)
For a discussion of unusual problems encountered in installing new
collecting equipment at existing furnace shops, and an indication
of cost variances, see page C-ll.
4)
Prices: 1969 base.

-------
1----
I
D-40
ELECTRIC ARC FURNACES
The following pages contain cost data relating to electric arc furnaces de-
signed for production of carbon steel. The figures are for completely new
installations. The special problems encountered when installing new control
equipment in existing plants were discussed on page C-l1. Each cost value
applies to a system of two furnaces with a common gas cleaner capable of
handling only one furnace at peak loads at any given time. For effect on
cost of a different system of multiple furnace control see page C-ll.
The volumes listed are based on typical oxygen blowing rates used in furnaces
making carbon steel from cold scrap. Oxygen and exhaust rates may be con-
siderably higher when making stainless heats. An excess of air would typ-
ically be added to the furnace gases for complete combustion of carbon
monoxide and hydrocarbons (the latter, during the melt-down of oily scrap),
and for cooling. These mixed gases would then be water quenched in wet
scrubbing, and also in pre-conditioning of the particles before electrostatic
precipitation, though less water would be used in the latter system to avoid
condensation and to optimize the temperature and humidity conditions for
effective precipitation. Although, for baghouse collection, these gases also
could be water quenched to some extent, effecting an economy in collector
size,it is more typical to cool by use of a radiating exchanger, and to.
finish the cooling with controlled air dilution just before the baghouse.
.~

-------
D-41
ELECTRIC ARC FURNACE - WET SCRUBBER (HIGH ENERGY)
o
Gas Volume - ACFM @ 180 F
36.000
25
137.000
150
210.000
250
Furnace Size - Tons (each)
CAPITAL COST
1. Materia1* $173,000 $511,000 $723,000
2. Labor 93,000 277,000 388,000
3. Central Engineering 67,000 162,000 215,000
4. Client Engineering 17,000 40.000 54.000
 TOTAL $350,000 $990,000 $1,380,000
   OPERATING COST ($/Yr.)  
1. Electric Power $ 40,000 $174,000 $265,000
2. Maintenance 14,000 40,000 55,000
3. Operating Labor 40,000 60,000 80.000
 Direct Operating Cost $ 94,000 $274,000 $400,000
4. Depreciation 35,000 99,000 138,000
5. Capital Charges 35,000 99,000 138 , 000
 TOTAL $164,000 $472,000 $676,000
* For items included in materials, see pages C-IO and C-40.
Note:
1)
Two Furnace System, alternating peak loads. For effect on cost of dif-
ferent combinations of furnaces per cleaning system see page C-ll.
2)
See also the section on Special Problems Encountered When Installing
New Control Equipment in Existing Plants, page C-ll.
3) A variation from these costs is noted in a case of a single furnace
cleaning system where, after correction for the savings in a 2-
furnace system, the cost would be 40% higher than indicated here.
Remote placement of the scrubber is one factor in this variation.
4)
Prices:
1969 base.

-------
D-42
ELECTRIC ARC FURNACE - ELECTROSTATIC PRECIPITATOR
Gas Volume - ACFM @ 5000F
48,000
25
185,000
150
280,0()()
Furnace Size - Tons (each)
250
CAPITAL COST
1.
Materia1~'(
$159,000
2.
Labor
85,000
$465,000 $652,000
251,000 352,000
151,000 197,000
38,000 49,000
$905,000 $1,250,000
3.
Central Engineering
61,000
TOTAL
15,000
$320,000
4.
Client Engineering
   OPERATING COST ($/Yr.)   
1. Electric Power  $ 8,000  $ 30,000 $ 60,000
2. Maintenance   13 , 000  36,000  50,000
3. Operating Labor   20,000  30,000  40,000
  Direct Operating Cost $ 41,000  $ 96,000 $ 150,000
4. Depreciation   32,000  90,500  125,000
5. Capital Charges   32,000  90,500  125,000
   TOTAL  $105,000  $277 ,000 $ 400,000
~'( For items included in materials, see pages C-10 and C-40.  
Note:
1)
Two Furnace System, alternating peak loads. For effect on cost of
different combinations of furnaces per cleaning system see page C-ll.
2)
See also section on Special Problems Encountered When Installing New
Control Equipment in Existing Plants, page C-ll.
3)
Prices: 1969 base.

-------
D-43
ELECTRIC ARC FURNACE - FABRIC FILTER
Gas Volume - ACFM @ 2750F
60,000
25
230,000
150
350,000
250
Furnace Size - Tons (each)
CAPITAL COST
1.
$120,000
$441,000 $654,000
209,000 321,000
140,000 196,000
35,000 49,000
$825,000 $1,220,000
$ 40,000 $ 52,000
39,000  57,000
30,000  40,000
$109,000 $ 149,000
82,500  122,000
82,500  122,000
$274,000 $ 393,000
* For items included in materials, see pages C-IO and C-40.
Material*
2.
Labor
60,000
3.
44,000
Central Engineering
4.
Client Engineering
11 , 000
$235,000
TOTAL
1.
Electric Power
OPERATING COST ($/Yr.)
$ 8,000
2.
11 , 000
Maintenance
3.
Operating Labor
20,000
$ 39,000
Direct Operating Cost
4.
23,500
Depreciation
5.
23,500
$ 86,000
Capital Charges
TOTAL
Note:
1)
Two Furnace System, alternating peak loads. For effect on cost of
different furnace c9mbinations per cleaning system see page C-ll.
2)
See also section on Special Problems Encountered When Installing
New Control Equipment in Existing Plants, page C-ll.
3)
Prices: 1969 base.

-------
D-44
ELECTRIC ARC FURNACE

Combination Direct Evacuation Control and Furnace Canopy-
Type Area Ventilation System - Fabric Filter
Gas Volume - ACFM @ 1400F
Shop Size - 2 Furnaces @ Tons (each)
125.000
20
750.000
120
CAPITAL COST
1.
Materia1*
$240,000
$1,200,000
2.
Labor
102,000
480,000
3.
Central Engineering
96,000
353,000
4.
Client Engineering
TOTAL
24,000
$462,000
88.000
$2,121,000
1.
Electric Power
OPERATING COST ($/Yr.)
$ 18,500
$ 100,000
2.
Maintenance
21,000
98,000
3.
Operating Labor
Direct Operating Cost
30,000
$ 69,500
40,000
$ 238,000
4.
Depreciation
46,000
212,000
5.
Capital Charges
46,000
212.000
$ 662,000
TOTAL
$161,500
* For items included in material see pages C-IO and C-40.
Note:
1)
Two Furnace System, alternating peak loads. For effect on cost
of different furnace combinations per cleaning system see page C-ll.
2)
See also the section on Special Problems Encountered When Installing
New Control Equipment in Existing Plants, page C-ll.
3)
Prices: 1969 base.

-------
D-45
S~ING
The following table presents cost data on scarfing units. These
units are of two different sizes. The smaller size is usually
employed when the billets to be handled are n~ver larger than about
50 inches. Larger billets will require the larger gas cleaning
equipment. The material cost excludes the cost of the Smoke
Tunnel. In wet cleaning systems on a scarfer, the water circuit
is normally coupled to an existing slab mill water treatment system,
so that slurry treatment is excluded in this case.

-------
D-46
SCARFING - WET SCRUBBER (HIGH ENERGY)
Gas Volume - ACFM @ 1000F
50,000
100,000
CAPITAL COST
TOTAL
$114,000 $176,000
66,000 96,000
48,000 68,000
12,000 17,000
$240,000 $357,000
1.
Material'\-
2.
Labor
3.
Central Engineering
4.
Client Engineering
  OPERATING COST ($/Yr.)  
1. Electric Power  $ 38,000 $ 75,000
2. Maintenance  10,000 14,000
3. Operating Labor  5,000 7,000
  Direct Operating Cost $ 53,000 $ 96,000
4. Depreciation  24,000 36,000
5. Capital Charges  24,000 36,000
  TOTAL  $101,000 $168,000
,\-
For items included in materials, see pages C-IO and C-45.
Note:
1)
Prices: 1969 base.

-------
D-47
SCARFING - ELECTROSTATIC PRECIPITATOR
Gas Volume - ACFM @ 100°F
CAPITAL COST
1.
Materia1*
2.
Labor
3.
Central Engineering
4.
Client Engineering
TOTAL
OPERATING COST ( $ /Yr . )
1.
Electric Power
2.
Maintenance
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
50,000
$135,000
85,000
57,000
14,000
$291,000
$ 8,000
12,000
5,000
$ 25,000
29,000
29,000
$ 83,000
*
For items included in materials, see pages C-IO and C-45.
Note:
1)
Prices: 1969 base.
100,000
$204,000
112,000
76,000
19,000
$411,000
$ 18,000
16,000
7,000
$ 41,000
41,000
41,000
$123,000

-------
D-48
HCL PICKLING LINE - WET WASHER
The following table presents cost data on a spray washing system for an
HCL Pickling Line acid fume removal system. Most modern lines are now
sized for 80 inch strip. Fiberglass material is used for all duct and
stack work. Fume is scrubbed by successive spray and eliminator units.
For optional acid brick lined tunnel (6 ft. sq.) to outside fume collectors,
add $184 per foot of length to capital cost total, and $44 per foot of
length to annual operating cost total.

-------
D-49
HCL PICKLING LINE - WET WASHER
Gas Volume - ACFM @ 100°F
130,000
Line Capacity
80 inch at 1,000 FPM
CAPITAL COST
1. Materia1*
2. Labor 
3. Central Engineering
4. Client Engineering
  TOTAL
$ 81,000
30,000
23,000
6,000
"$140,000
OPERATING COST ($/Yr.)
1.
Electric Power
$ 10,000
2.
Maintenance
6,000
3.
Operating Labor
Direct Operating Cost
5,000
$ 21,000
4.
Depreciation
14,000
5.
Capital Charges
TOTAL
14,000
$ 49,000
*
For items included in materials, see pages C-IO and C-48.
Note:
1)
Prices:
1969 base.

-------
D-50
COLD ROLLING MILL - MIST ELIMINATOR
The following table presents cost data for an eliminator system to remove
the palm oil and water mist emission at roll stands of a typical, large,
five stand tandem cold rolling mill. The suction of the system picks up
mist from closure plate enclosed areas at each stand, carries it through
a tunnel to two mist eliminators and fans. The ventilation air thus
cleaned is disqharged up a stack. The treatment of collected oil for re-
use or disposal is not included.

-------
D-51
COLD ROLLING MILL - OIL MIST ELIMINATION
Gas VA lume - ACFM @ 1l0oF
200,000
Mill Size
80 inch, 5 stand tandem
CAPITAL COST
1. Material*
$ 85,000
2.
Labor
62,000
3.
Central Engineering
29,000
4.
Client Engineering
TOTAL
7,000
, $183,000
OPERATING COST ($/Yr.)
1.
Electric Power
$ 18,000
2.
Maintenance
7,000
3.
Operating Labor
7,000
Direct Operating Cost
$ 32,000
4.
Depreciation
18,000
5.
Capital Charges
TOTAL
18,000
$ 68,000
*
For items included in materials, see pages C-IO and C-50.
Note:
1)
Prices: 1969 base.

-------
D-52
POWER PLANT BOILERS
The following pages contain cost data on several sizes of in-plant boiler
houses. They assume that smoke and fly ash control equipment is being in-
stalled on an existing coal-fired boiler. The figures cover a single boiler
only. Various combinations of multiple boiler-collector units are used in
actual practice, with savings in larger sizes dissipated in additona1 duct,
dampers and complicated setup. The stack is considered to be already existing.
Multiple cyclones, when used for primary collecting, are included as they
yield no process advantage to the boiler. Booster fans are included.
Mechanically fed coal-fired boilers may achieve acceptable fly ash control
with multi-cyclones alone. However, large modern boiler houses. in integrated
steel plants would usually use pulverized coal firing for efficiency and quick
regulation of firing rate as well as ease of combined or auxiliary firing with
blast furnace or coke oven gas. Pulverized coal's higher percentage of fly ash with
a finer size grading requires the use of high efficiency control equipment,
of which the electrostatic precipitator is almost solely used (often in con-
junction with a mechanical primary co1lector~ as it is more economical than
wet scrubbing. The exhaust gas usually contains a significant amount of sulfur
dioxide, which promotes effective cleaning with a smaller precipitator than
would be required without it. The hot, buoyant gases leaving the precipitator
disperse more readily than if cooled by scrubbing or for baghouse cleaning.
Sulfur dioxide emission suppression, using limestone injection with baghouse
collection or absorbtive solution scrubbing, currently undergoing tests for
public utility application, may eventually displace electrostatic precipitation
of fly ash. But the trend in steel plant boilers is toward relatively po11ution-
free fuels, particularly gas and oil. Combustion devices to prevent carbon
monoxide emissions are considered 100% process beneficial, and not funded as
pollution control equipment. The formation mechanism and control techniques
for nitrogen oxides emissions are currently under study; a preventive method will
likely be sought for their limitation. The development of acceptable soot
build-up removal means remains a problem.

-------
D-53
POWER PLANT BOILER
Mechanically Fed, Coal Fired Boiler-Multicyclone Collector
o
Volume, ACFM @ 600 F
Boiler Size, pounds steam/hr.
CAPITAL COST
1.
Material~'(
2.
Labor
3.
Central Engineering
4.
Client Engineering
TOTAL
32,000
50,000
$20,000
10,000
10,000
2,500
$42,500
OPERATING COST ($/Yr.)
$ 2,300
1.
Electric Power
2.
Maintenance
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
*
1,700
7,000
$11,000
4,300
4,300
$19,600
For items included in materials, see pages C-IO and C-52.
Note:
1)
2)
Prices: 1969 base.
One Boiler System.
96,000
150,000
$ 60,000
30,000
24,000
6,000
$120,000
$
7,000
5,000
15,000
$ 27,000
12,000
12,000
$ 51,000

-------
D-54
POWER PLANT BOILER
Pulverized Coal Fired Boiler - Electrostatic Precipitator
Volume, ACFM @ 3000F
100,000
200,000
CAPITAL COST
/
1.
Material'"
$260,000
$440,000
2.
Labor
140,000
230,000
3.
Central Engineering
100,000
170,000
4.
Client Engineering
TOTAL
25,000
$525,000
45,000
$885,000
1.
Electric Power
OPERATING COST ($!Yr.)
$ 28,000
$ 55,000
2. Maintenance  21,000
3. Operating Labor  30,000
  Direct Operating Cost $ 79,000
4. Depreciation  52,500
5. Capital Charges.  52,500
  TOTAL  $184,000
36,000
40,000
$131,000
88,500
88,500
$308,000
*
For items included in materials, see pages
C-IO and C-52.
Note:
1)
2)
One Boiler, Two Precipitator System.
Prices: 1969 base.

-------
D-55
SAMPLE CALCULATION - OPERATING COST ($/Yr.)
The sample illustrates the calculations performed in arriving at the
operating cost for a fabric filter installation on a 150 ton Electric Arc
Furnace. Electric power costs (@ 3/4~ per kwh) is obtained by calculating
the total horsepower of all motors (plus power to lights and instruments)
and multiplying by a cost per horsepower factor. Power to a fan motor is
calculated by applying an efficiency to the power required for reversible
adiabatic compression. This latter quantity is called "Air H.P." Power
to a water pump motor is similarly calculated, the reversible pumping power
requirement being called "Water H.P."
1.
Electric Power (*)
Basis: $50/HP/Yr 800 HP Motor
$50/HP/Yr x 800 HP =
$ 40,000/Yr
2.
Maintenance
Basis:
4% of Capital Cost
Capital Cost $825,000
0.04 x $825,000 =
$ 33,000/Yr
6,000/Yr**
3.
Depreciation
Basis:
10% of Capital Cost
Capital Cost $825,000
0.10 x $825,000 =
$ 82,500/Yr
4.
Capital Charges
Basis: 10% of Capital Cost
Capital Cost $825,000
0.10 x $825,000 =
$ 82,500/Yr
5.
Operating Labor
Basis: 3/4 Man/Shift or 18 Manhours/Day
$5.00/Manhour
18 MH/Day x $5.00/MH x
330 Opr.Day/Yr =
$ 30,000/Yr
TOTAL
$274,000/Yr
*
See following page for notes.
** The difference from the 4% standard maintenance
cost with bag replacement cost figured as
described on page C-15.

-------
D-56
*1.
Electric Power Cost @ $0.0075/KWH
Operating Days = 330 Days/Yr
1 HP = 0.746 KW
$0.0075/KWH x 330 Days/Yr x 24 Hr/Day x 0.746 KW/HP
x .89 Mo~or Eff. $50/HP per Yr
*2.
Air HP = 0.0001575 PQ
P = Static Pressure, in. water
Q = Volume, CFM
Motor HP = Air HP
Eff.
Eff. = Efficiency - Range 60 to 70%
GPM x H
Water HP = 3,960
*3.
GPM = Gallons per Minute
H = Head, in Ft.
Motor HP = Water HP
Eff.
Eff. = Efficiency - Range 75 to 85%
METHOD OF DETERMINING EXHAUST GAS VOLUMES IN SIZING
COLLECTING SYSTEMS FOR PRICING
The following sample illustrates the method of calculating the capacity"
of the collector in each estimated system. In general, the exhaust gases are
cooled in transit through the system, so that successive items of equipment
in the system will have different volumetric capacities due to gas volume
changes with temperature and with the material additions (dilution air or water
vapor) added to effect cooling.
The starting point is to determine a typical process exhaust gas composition
and volume rate per unit of process throughput (as SCFM/ingot ton). In some
cases this is determined solely by the oxygen lancing rate which generates the
maximum exhaust volume during a steelmaking heat. In the open hearth case,
since fuel and air are customarily added to the furnace during lancing, and
waste heat boilers are generally used for cooling the exhaust gases, typical
volumes of gas At the boiler outlet condition we~e selected as a starting volume
for the gas cleaning system.

-------
D-57
In the blast furnace,scarfing, sinter plant windbox, pelletizing (dryer or
process), and power plant cases typical modern practice was used as a basis
for determining the process exhaust volume. In materials handling and mist
pick-up cases, where in-drawn ventilation air entrains particles, mist and
vapors to be controlled, typical modern systems were studied to determine
ventilation rates for adequate emission containment and to ensure the inclusion
of sufficient pick-up points to contain a plant's effluent according to the extent
that current technology can meet current standards.
Sample of volume determinati~n method:
Peak oxygen rate to process = 1500 SCFM at 32°F
1.)
2. )
Combustion with air of carbon monoxide
Carbon monoxide (maximum) = 3000SCFM
2CO + 02 + ?1. N2 = 2C02 + 79 N2
21 21'
produced.
+ excess air
Combustion products
CO 3000 SCFM
2

N2
+ Excess air
1 x 79 x 3000 SCFM = 5640 SCFM
2 21 '
Excess air = 500% in a typic~l case
= 5 x 100 x 5640 SCFM = 35,700 SCFM
79
Total 44,300 SCFM
44,300 SCFM x 1.7 (=Factor for two furnaces with alternating peak loads.:
"" 75,500 SCFM
3.)
Cooling the gases
The combustion occurs in a water-cooled, double-wall duet where
cooling occurs by radiation and convection of heat to the ~alls. The
gases leaving this section will typically be at about 1200 F. The
size of such indirect heat exchanger will be determined by combining
heat transfer and heat balance equations in an iter~tive calculation,
based on certain reasonable assumptions of water ~emperatures, gas
velocity, and water circuit capacity. Optimizing the total cooling
and gas cleaning system is an extensive design task, so that typical
equipment for each system has been selected for this study's estimates.

The cooling by air or water additions to the gases at l2000F involve:
a heat balance for calculating resultant volume.
, ' .

;Mml\n + MwHw '"" ~~~
M = pound moles of each component.
H ;::: enthalpy of each component at conditions.
m = each component of uncooled gas.
n = each component of cooled gas. ,
w ;::: water at spray water temperature.

-------
D-58
For a final temperature of 500°F, suitable for an electrostatic precip-
itator, about 20% moisture is required by this analysis.

75,500 SCFM f .8 = 94,500 SCFM

94,500 SCFM x (560 + 460)OR -
492°R -
185,000 ACFM @ 500°F
CAPITAL COST BREAKDOWN
The following tables illustrate the relative importance of various
components in total material costs.
This is a very rough breakdown, and variations occur due to capacity
and type of system. However, the relative orders of magnitude are well
maintained. Certain conclusions can be drawn from this tabulation con-
cerning the sensitivity of the total to local conditions. Foundations and
structure may change considerably without having a marked effect on the total.
Very often, a local requirement which tends to increase structure will simul-
taneously reduce foundations. The figures used for these two components are
based upon simple structures supporting the collector near grade, and a soil
bearing value of 4,000 1bs. per square foot.
The stack and fan components are rather closely related to gas volume
and collector type. They are therefore relatively well defined. Electrical,
while an important component, is predicted with comparative certainty from
horsepower.
The key cost element is the collector itself, and it is to this item
that the estimator gives the greatest attention. Generally this will involve
obtaining a price quotation from a reliable manufacturer, although the published
literature also contains useful information.
The second category, labor et a1, is estimated on the basis of
anticipated labor costs for each of the components in Total Material.
Typical rules for this calculation are:
(a)
(b)
Collector:
Labor is about 35% of Material
Fan, motor and starter:
Material
Labor is about 15% of
(c)
(d)
Stack:
Labor is about 100% of Material
Ductwork:
Labor is about 100% of Material
(e)
(f)
Steel:
Labor is about 30% of Material
Foundations:
Labor is about 130% of Material
(g)
Electrical:
Labor is about 150% of Material

-------
  D-59   
  MATERIAL BREAKDOWN   
  "   
Sinter Plant - Windbox Gas Cleaning   
     Wet Electrostatic Fabric
     Scrubber Precipitator Filter
1. Foundations    4 3 4
2. Ductwork and Stack  2 5 4
 Modifications     
3. Collector    30 68 71
4. Fan and Motor    7 7 10
5. Structural    2 3 3
6. Electrical    7 9 5
7. Water Treatment & Piping 46 1 1
8. Controls    2 4 2
  Total 100% 100% 100%
Sinter Plant - Material Handling -   
Dust Collection      
1. Foundations    4 3 4
2. Ductwork and Stack  12 18 21
3. Co llec tor    28 47 46
4. Fan and Motor    7 5 7
5. Structural    4 7 10
6. Electrical    8 17 9
7. Water Treatment & Piping 35 1 1
8. Controls    2 2 2
  Total 100'70 100% 100%

-------
D-60
MATERIAL BREAKDOWN
Pelletizing Plant (Moving Grate) - Dust Collection
1. Foundations  
2. Ductwork and Stack
3. Collector  
4. Fan and Motor  
5. Structural  
6. Electrical  
7. Water Treatment and Piping
8. Control  
 Total  
Pelletizing Plant (Shaft Furnace)
Cyclone
Wet
Scrubber
5
2
15
5
45
40
14
20
7
5
11
8
1
18
2
2
100%
100%
    Process Exhaust Material Handling
    Cyclones Cyclones Wet Scrubbe...
1. Foundation   2 4 2
2. Ductwork and Stack 10 20 12
3. Collector   41 34 39
4. Fan and Motor   18 13 18
5. Structural   11 12 7
6. Electrical   12 13 8
7. Water Treatment & Piping 2 1 12
8. Controls   4 -2. 2
 Total  100% 100% 100%

-------
D-61
MATERIAL BREAKDOWN
Blast Furnace
1. Foundations
2. Ductwork and Stack
3. Collector
4. Fan and Motor
s. Structural
6. Electrical
7. Water Treatment and Piping
8. Control
 Total
Two Stage Venturi
Scrubber System
3
10
46
7
6
25
3
100%

-------
D-62
   MATERIAL BREAKDOWN  
Basic Oxygen Furnace     
     Wet Scrubber Electrostatic Fabric
     (High Energy) Precipitator Filter
 1. Foundations   4 3 2
 2. Ductwork and Stack 30 36 37
 3. Collector   10 31 32
 4. Fan and Motor   9 5 6
 5. Structural   6 6 5
 6. Electrical   7 8 7
 7. Water Treatment and Piping 31 7 7
 8. Controls   3 4 4
   Total 100% 100% 100%
Open Hearth ,Furnace    
   Wet Scrubber Electrostatic Fabric
   (High Energy) Precipitator Filter
1. Foundations  3 2 2
2. Ductwork and Stack 21 25 25
 Modifications    
3. Collector  15 40 42
4. Fan and Motor  6 2 4
5. Structural  9 9 9
6. Electrical  11 10 8
7. Water Treatment and Piping 33 9 7
8. Controls  2 3 3
  Total 100% 100% 100%

-------
D-63
   MATERIAL BREAKDOWN  
Electric Arc Furnace (Direct     
Extraction Fume System)     
     Wet Scrubber Electrostatic Fabric
     (High Energy) Precipitator Filter
1. Foundations   4 3 2
2. Ductwork and Stack 22 .31 35
3. Collector   15 35 34
4. Fan and Motor   9 6 7
5. Structural   7 6 7
6. Electrical   10 10 9
7. Water Treatment and Piping 30 5 2
8. Con t ro1s   .2 4 4
   Total 100% 100% 100%
Electric Arc Furnace (Combination Direct
Evacuation Control and Furnace Canopy- Type
Area Ventilation System)
    Fabric Filter
1. Foundations   3 
2. Ductwork and Stack 37 
3. Collector   27 
4. Fan and Motor   10 
5. Structural   10 
6. Electrical   7 
7. Water Treatment and Piping 1 
8. Controls   5 
  Total 100% 

-------
D-64
  MATERIAL BREAKDOWN 
Scarfing    
   Wet Scrubber Electrostatic
   (High Energy) Precipitator
1. Foundations  2 2
2. Ductwork and Stack 12 20
3. Collector  30 55
4. Fan and Motor  24 7
5. Structural  5 3
6. Electrical  16 7
7. Water Circuit  7 2
8. Controls  4 4
  Total 100% 100%
Power Plant Boiler   
    Electrostatic
   Cyclone Precipitator
1. Foundations  6 3
2. Ductwork  13 33
3. Collector  40 20
4. Fan and Motor  19 2
5. Structural  8 19
6. Electrical  12 17
7. Water Treatment and Piping 0 1
8. Controls  2 5
  Total 100% 100%

-------
, D-65
GASEOUS POLLUTANTS
The present report does not present cost data on equipment for the
control of gaseous pollutants. Methods for the chemical treatment of gases
for the removal of sulfur.' and nitrogen. oxides are s till under development.
Reliable plant cost data will ~ot be available for some time.
Volatiles emitted during the processing of coke oven by-products can
generally be controlled by careful operating control of leaks, drips, drains,
and vents. Any waste gases from flare stacks will probably contain sulfur
oxides, for which treatment methods are not commercially available.
AREA VENTILATION AND EMISSION CONTROL
While the technology for cleaning of effluent material contained in
exhaust ducts from enclosed processes has reached a state of development
where clearly defined practices and equipment can be specified, the means
to clean areas where process materials enter or leave'the process enclosure
and to clean the ventilated air from shop structures and outside handling
areas is only now developing. Until the sizing and alternate methods have
been tested by sufficient application, and competitive pricing has evolved,
a definitive estimate of. the cost and performance of truly adequate control
means is premature. .
The ventilation air vol~mes may be ~any times the volume of the gases
cleaned in ducted exhaust circuits from the process; and explosion hazards
at times occur with the influx of air. An example estimated here at the
current level of development is the electric arc furnace melt shop with a
combination of direct evacuation control at the furnace and a canopy above
the furnace. This system provides containment and control during all phases
of the heat cycle when the furnace roof is in place, and, in addition, good
control in preventing fume from escaping from the building during those
operations' when there is no local containment at the furnace, such as
charging, teeming,andslagging. 'The added volume to the collector is 1 to
2.5 times gre'ater than with furnace flue gas treatment alone, or 2 to 3.5
times greater for the total system. In a case where the canopies are
installed higher, at the roof truss, with no direct furnace evacuation,
the volume is 4 to Stimes greater than it would be if shell evacuation'
alone were to be uSed. The basic oxygen furnace fume system, sized for
peak volumes during oxygen lancing, could be fitted with auxiliary hoods
and dampers to accommodate the hot metal charging and teeming area at low
level, utilizing this peak evacuation capacity 'for area ventilation during
off-blow periods. The same external operations at the open hearth would
require added exhaust capacity, used in turn on each furnace of the shop.
Drafts in the shop seriously effect the '''catch'' of open hoods, especially
high-lofted canopies as applied to arc furnaces. .
Alternate means ,are being applied for exhausting and fume removal.
These include various pickup devices:

-------
D-66
1.
Close fitting hoods (with relatively low volume
required) applied to pickling tanks and roll stands
for mist pickup.
2.
Low auxiliary hoods and partial enclosures applied
to pouring operations of hot iron or steel, or the
crushing, screening, loading and discharging of dry
materials (sinter, ore, coal, coke, fluxes and other
chemicals).
3.
Tunnels as applied to scarfing units and conveying
lines.
4.
High canopies with isolation dampers for selective
ventilation of high concentration dust areas, and
total building air-change systems are currently
being evaluated at a few melt shops. Buildings
to enclose extensive areas of material handling
and open processing with many dust generation
points or discharges that are difficult to control
at the source, are used to some extent now (at
crushing and screening stations, for example).
In principle, the enclosure of such an area with cleaning and possibly
recycling of the ventilation air therefrom could effect a reduction in volume
and system complexity compared to that for many high pickup canopies. In
practice, however, while emissions to the atmosphere could be significantly
reduced, hazards would in many cases accompany returning air from the collector
discharge to the workspace, limiting application of this principle. The
magnitude of the task suggests the need for less costly, more effective,
close-to-source control means. The volume required for. adequate entrainment
of emissions varies greatly, becoming much larger and less effective when
pickup devices are farther removed from the source.
And while concentrations of pollutant material can be measured at points,
the open-air distribution of concentration cannot be adequately profiled. The
concentration of an air borne material beyond the plant area is subject to the
weather and fall-out variables. Therefore, research is needed to quantitatively
evaluate an area atmosphere by means that could be used for design criteria by
equipment manufacturers and would give correlated information in performance
guarantee tests and the abatement inspector's spot check.
With enough application of engineering design, less expensive means of
controlling presently uncontained volumes will evolve; Plant design can
accomplish some grouping of.high dust areas to reduce ventilation requirements.
Process change and new equipment design will increasingly consider pollution
problems as a factor. Building design, currently based on natural ventilation
means, could undergo changes to reduce the extent and facilitate the means of
ventilation. And with optimization of means, a more realistic cost level will
in time evolve.

-------
D-67 and D-68
Some prior cost tables give estimates of costs for ventilating dust
and mist areas and cleaning the captured air around several processes. The
estimates represent the most adequate systems currently being applied or
quoted for process ventilation needs to supplement ducted process gas
exhausting and cleaning:
Sinter plant material handling,
Arc furnace canopy-type area ventilation,
Scarfing tunnel evacuation,
Pickling line mist removal,
Rolling mill mist pickup.
Note: The report by Swindell-Dressler Company appears in its entirety
as Appendix C of the companion "Final Report on a Cost Analysis of
Air-Pollution Controls in the Integrated Iron and Steel Industry", dated
May 15, 1969; and an adaptation of that portion not included in this
Appendix D is presented in pages VI-36 to end of Section VI of this
technological report.

-------