Iron Making by Tupkary

MODERN
IRON MAKING
HANDBOOK
R. H. Tupkary, PhD
V. R. Tupkary
MLI Handbook Series

Sarhan Musa, PhD
Prairie View A&M
(Series Editor)
Mercury Learning and Information

Dulles, Virginia
Boston, Massachusetts
New Delhi
Copyright ©2018 by Mercury Learning and Information LLC. All rights reserved.
Original Title and Copyright: Introduction to Modern Iron Making, 4/E. © 2016 by Khanna Publishers.
This publication, portions of it, or any accompanying software may not be reproduced in any way,
stored in a retrieval system of any type, or transmitted by any means, media, electronic display
or mechanical display, including, but not limited to, photocopy, recording, Internet postings, or
scanning, without prior permission in writing from the publisher.
Publisher: David Pallai

Mercury Learning and Information
22841 Quicksilver Drive
Dulles, VA 20166
[email protected]
www.merclearning.com
1-800-232-0223
R. H. Tupkary and V. R. Tupkary. Modern Iron Making Handbook.

ISBN: 978-1-68392-136-3
The publisher recognizes and respects all marks used by companies, manufacturers, and
developers as a means to distinguish their products. All brand names and product names
mentioned in this book are trademarks or service marks of their respective companies. Any
omission or misuse (of any kind) of service marks or trademarks, etc. is not an attempt to
infringe on the property of others.
Library of Congress Control Number: 2017948166
171819321 Printed in the USA on acid-free paper.
Our titles are available for adoption, license, or bulk purchase by institutions, corporations, etc.
For additional information, please contact the Customer Service Dept. at 800-232-0223(toll free).
All of our titles are available in digital format at authorcloudware.com and other digital vendors. The
sole obligation of Mercury Learning and Information to the purchaser is to replace the book,
based on defective materials or faulty workmanship, but not based on the operation or functionality
of the product.
LICENSE, DISCLAIMER OF LIABILITY, AND LIMITED WARRANTY
By purchasing or using this book (the “Work”), you agree that this license grants per-
mission to use the contents contained herein, but does not give you the right of owner-
ship to any of the textual content in the book or ownership to any of the information or
products contained in it. This license does not permit uploading of the Work onto the
Internet or on a network (of any kind) without the written consent of the Publisher. Du-
plication or dissemination of any text, code, simulations, images, etc. contained herein
is limited to and subject to licensing terms for the respective products, and permission
must be obtained from the Publisher or the owner of the content, etc., in order to re-
produce or network any portion of the textual material (in any media) that is contained
in the Work.
Mercury Learning and Information (“MLI” or “the Publisher”) and anyone in-
volved in the creation, writing, or production of the companion disc, accompanying
algorithms, code, or computer programs (“the software”), and any accompanying Web
site or software of the Work, cannot and do not warrant the performance or results that
might be obtained by using the contents of the Work. The author, developers, and the
Publisher have used their best efforts to insure the accuracy and functionality of the
textual material and/or programs contained in this package; we, however, make no war-
ranty of any kind, express or implied, regarding the performance of these contents or
programs. The Work is sold “as is” without warranty (except for defective materials used
in manufacturing the book or due to faulty workmanship).
The author, developers, and the publisher of any accompanying content, and anyone in-
volved in the composition, production, and manufacturing of this work will not be liable
for damages of any kind arising out of the use of (or the inability to use) the algorithms,
source code, computer programs, or textual material contained in this publication. This
includes, but is not limited to, loss of revenue or profit, or other incidental, physical, or
consequential damages arising out of the use of this Work.
The sole remedy in the event of a claim of any kind is expressly limited to replacement
of the book, and only at the discretion of the Publisher. The use of “implied warranty”
and certain “exclusions” vary from state to state, and might not apply to the purchaser
of this product.
CONTENTS
1. HISTORICAL ........................................................................... 1
1.1 Ancient Ironmaking ............................................................................2
1.2 Historical Ironmaking Processes ........................................................3
1.3 Developments that Led to Modern Ironmaking ...............................5
Coke Furnace......................................................................................6
Hot Blast Generation..........................................................................7
Blast Furnace Gas Cleaning ...............................................................7
Charging Devices................................................................................8
Pig Casting Machine ...........................................................................9
1.4 Modern Ironmaking ...........................................................................9
1.5 Alternative Methods of Iron Production .........................................10
1.6 History of the Iron and Steel Industry in the United States ...........13
1.7 Twenty-first Century Trends (U.S. and the World) .........................15
U.S. Production of Selected Mineral Commodities in the
First Quarter 2017 ............................................................................16
1.8 Concluding Remarks ........................................................................17
2. MODERN BLAST FURNACE PLANT AND PROCESS ...... 19

2.1 Introduction ......................................................................................19
2.2 Modern Blast Furnace......................................................................20
Furnace Details ................................................................................21
vi • Modern Iron Making Handbook
2.3 Hot Blast Stove .................................................................................24

2.4 Gas Cleaning .....................................................................................24
2.5 Raw Materials Storage and Handling ..............................................24
2.6 Liquid Products Disposal .................................................................25
Hot Metal Ladles ..............................................................................26
2.7 Blast Furnace Operation and Control .............................................27
2.8 Essentials of the Blast Furnace Process ..........................................28
Stack or Shaft ....................................................................................29
Bosh ...................................................................................................30
Tuyere or Combustion Zone ............................................................30
Hearth ...............................................................................................31
2.9 Efficiency of Blast Furnace ..............................................................32
3. RAW MATERIALS FOR IRONMAKING .............................. 35

3.1 Introduction ......................................................................................35
3.2 Iron Ores ...........................................................................................38
World Deposits of Iron Ores ............................................................45
3.3 Metallurgical Coal.............................................................................52
Metallurgical Coal Reserves .............................................................53
3.4 Limestone and Dolomite as Flux .....................................................53
4. MAIN BLAST FURNACE FUEL: COKE .............................. 57

4.1 Introduction ......................................................................................57
4.2 Functions of Coke.............................................................................59
4.3 Quality Requirements of Coke.........................................................60
Reactivity ...........................................................................................61
Size ....................................................................................................63
Strength and Abrasion Resistance....................................................64
Micum Test .......................................................................................64
ASTM Tumbler Test (U.S.) .............................................................66
Preparation of BF Fuel ....................................................................66
4.4 Preparation of Coke (India) .............................................................68
Coal Ranking (U.S.) .........................................................................71
4.5 Modifications in Coke Making Practices .........................................72
Contents • vii
Stamp Charging of Coke Ovens .......................................................72

Economizing Coke-Charge-Mix ......................................................74
Use of Waste Plastic in Coke Making ..............................................75
4.6 Alternative Fuels ...............................................................................76
4.6 Ferro-coke.........................................................................................77
5. DISTRIBUTION OF TRADITIONAL BURDEN ................. 79

5.1 Introduction ......................................................................................79
5.2 Elements of Distribution ..................................................................82
5.3 Factors Affecting Distribution .........................................................83
Factors Affecting Distribution Inside the Furnace .........................83
Angle and Size of the Big Bell ..........................................................83
Angle of Repose and Shape of Particles ..........................................86
Order of Charging.............................................................................91
5.4 Limitations ........................................................................................92
6. BURDEN PREPARATION: TREATMENT OF

IRON ORES ........................................................................... 95
6.1 Introduction ......................................................................................95
Lumps: ..............................................................................................96
Coarse Fines: ....................................................................................96
Very Fine Fines: ................................................................................96
6.2 Burden Qualities ...............................................................................97
Physical ..............................................................................................97
Chemical ...........................................................................................98
6.3 Crushing and Sizing ..........................................................................99
6.4 Beneficiation of the Iron Ores........................................................102
Magnetic Concentration .................................................................103
Gravity Methods .............................................................................103
Flotation Method ............................................................................104
Electrostatic Separation..................................................................105
Magnetic Roasting ..........................................................................105
Washing ...........................................................................................105
Drying .............................................................................................105
viii • Modern Iron Making Handbook
6.6 Calcining/Roasting ..........................................................................106

6.7 Blending ..........................................................................................106
6.8 Shipping ..........................................................................................107
6.9 Products of Ore Processing ............................................................108
6.10 Treatment of Iron Ores (India) ......................................................108
Ore Treatment at Tata Steel ...........................................................110
Alumina Problem and Beneficiation ..............................................112
Treatment of Goa Ores ...................................................................112
Kundremukh Iron Ore Project ......................................................113
6.11 Agglomeration of Iron Ores ...........................................................114
Briquetting ......................................................................................114
Nodulizing .......................................................................................115
Vacuum Extrusion ...........................................................................115
Sintering ..........................................................................................116
Pelletizing ........................................................................................116
6.12 World Resources (2017 USGS): .....................................................116
Substitutes .......................................................................................116
6.13 Concluding Remarks ......................................................................117
7. BURDEN PREPARATION: AGGLOMERATION

AND SINTERING ................................................................ 119
7.1 Principle of Sintering ......................................................................119
7.2 Process Variables.............................................................................121
7.3 Sinter Quality ..................................................................................135
7.4 Mechanism of Sintering .................................................................137
7.5 Raw Material Requirements ..........................................................140
Size ..................................................................................................140
Fuel Content ...................................................................................140
Moisture ..........................................................................................142
Circulating Load .............................................................................142
Performance of Sintering Machine ................................................143
Economics .......................................................................................144
7.6 Recent Trends in Sintering Practice ..............................................145
7.7 Alumina Problem in Sintering........................................................146
Indian Scenario ...............................................................................146
Contents • ix
8. BURDEN PREPARATION: AGGLOMERATION

AND PELLETIZATION....................................................... 151
8.1 Introduction ....................................................................................151
8.2 The Pelletization Process ...............................................................152
8.3 Theory of Bonding ..........................................................................153
8.4 Mechanism of Ball Formation........................................................155
Nucleii Growth Region...................................................................156
Transition Period.............................................................................156
Ball Growth Region ........................................................................156
Growth by Assimilation ..................................................................156
Growth by Layering ........................................................................157
8.5 Feed Size, Shape, and Porosity ......................................................157
8.6 Green Ball Production ....................................................................158
Disc Pelletizer .................................................................................159
Drum Pelletizer ..............................................................................162
8.7 Additives ..........................................................................................164
8.8 Induration of Pellets .......................................................................165
Drying .............................................................................................165
Preheating and Firing .....................................................................166
Firing Units .....................................................................................167
8.9 Comparison of the Pellet Firing Machines....................................171
8.10 Indian Pelletization Plants ..............................................................172
8.11 U.S. Production (Pelletization) ......................................................172
9. CHARACTERIZATION OF BF CHARGE MATERIALS ... 175

9.1 Introduction ....................................................................................175
9.2 Room Temperature Physical Properties ........................................176
Shatter Tests ....................................................................................176
Tumbling and Abrasion Test...........................................................177
Compression Test ...........................................................................179
Porosity ............................................................................................179
9.3 Reducibility .....................................................................................179
9.4 Physical Behavior of Materials During Reduction
at High Temperatures.....................................................................180
Decrepitation ..................................................................................181
x • Modern Iron Making Handbook
Low-Temperature Breakdown Test (L.T.B.T.) ..............................181

Reduction Degradation Index Test (R.D.I) ...................................182
Reducibility Test (R.I.) on Iron Bearing Materials .......................183
Coke Reactivity Test (C.R.I.) and Coke Strength
after Reaction (C.S.R.) ...................................................................183
Hot Compression Strength ............................................................183
Swelling ...........................................................................................184
Softening of Materials in the Blast Furnace ..................................184
High Temperature Permeability Under Reducing Conditions
and Applied Load ...........................................................................185
10. PHYSICAL-THERMAL-CHEMICAL PROCESSES IN A

BLAST FURNACE............................................................... 187
10.1 Introduction ....................................................................................187
10.2 The Blast Furnace Profiles .............................................................188
Thermal, Physical, and Chemical Profiles .....................................188
Physical Structure ...........................................................................189
10.3 Thermodynamic Equilibria ............................................................193
Carbon-Oxygen Equilibria .............................................................193
Equilibria in Fe-C-O and Fe-O-H Systems ..................................194
Elements of Stoichiometry .............................................................197
10.4 Physical Chemistry of Blast Furnace Process ................................199
Reactions in the Tuyere Zone.........................................................199
Reactions in Stack ...........................................................................202
Kinetics of Iron Oxide Reduction ..................................................205
Physical Nature of Ore and Its Previous History ..........................206
Nature and Composition of Gangue in the Charge ......................207
Temperature and Carbon Monoxide Content of Gases ................207
Chemical Nature of Iron Oxide .....................................................208
Pressure ...........................................................................................208
Solid-Solid Reaction .......................................................................208
Reactions in the Bosh ....................................................................209
Reactions in the Hearth..................................................................211
Sulphur Reaction: ...........................................................................214
10.5 Alumina Problem ............................................................................217
Indian Ores .....................................................................................220
Contents • xi
10.6 Kinetics of Reactions in Bosh and Hearth .....................................221

10.7 Efficiency of the Blast Furnace Process ........................................223
Elements of Enthalpy Balance .......................................................223
Process Zones in a Blast Furnace...................................................225
Thermal Efficiency .........................................................................227
10.8 Quality of Hot Metal.......................................................................228
Silicon Content of Hot Metal .........................................................229
10.9 TiO2 Addition in Burden.................................................................230
10.10 U.S. and World Events, Trends, and Issues.
Source: USGS 2017 ........................................................................231
Recycling .........................................................................................232
Import Sources (2012–15) .............................................................233
Trends and Issues............................................................................233
World Mine Production and Reserves ..........................................234
11. CONSTRUCTION OF A BLAST FURNACE AND

ITS ACCESSORIES ............................................................. 235
11.1 Blast Furnace Refractories .............................................................236
Stack Lining ....................................................................................236
Hearth Lining .................................................................................237
Hearth Walls ...................................................................................239
Bosh Lining .....................................................................................239
Carbon-lined Furnaces ...................................................................240
11.2 Blast Furnace Structure .................................................................241
11.3 Blast Furnace Cooling Arrangements ............................................242
Shaft Coolers ...................................................................................243
Hearth and Bosh Coolers ...............................................................245
Cooling of Hearth Bottom..............................................................246
11.4 Tap Holes ........................................................................................247
11.5 Cast House ......................................................................................249
11.6 Tuyere Assembly .............................................................................251
11.7 Raw Material Section......................................................................252
11.8 Charge Hoisting Appliances ...........................................................253
11.9 Top Charging System......................................................................255
11.10 Blowers, Boilers, Pumps, etc. .........................................................255
xii • Modern Iron Making Handbook
11.11 Slag-granulation ..............................................................................256

11.12 Instrumentation and Control .........................................................256
12. BLAST FURNACE PRODUCTS: BLAST FURNACE

GAS CLEANING AND UTILIZATION .............................. 259
12.1 Uptake, Downcomer, and Bleeder.................................................260
12.2 Dust Catcher ...................................................................................262
12.3 Semi-Fine or Primary Cleaning .....................................................263
Scrubbers or Spray Towers .............................................................263
Venturi Washer ...............................................................................265
12.4 Fine or Secondary Cleaning ...........................................................265
Electrostatic Precipitator (ESP) .....................................................266
12.5 Utilization of Blast Furnace Gas ....................................................267
12.6 Hot Blast Stove ...............................................................................268
Construction of the Stove and Its Accessories...............................269
13. BLAST FURNACE OPERATION ....................................... 275

13.1 Blowing-in .......................................................................................275
Drying .............................................................................................275
Filling ..............................................................................................276
Lighting and Operating Until Routine Practice is Established ....277
13.2 Banking ...........................................................................................278
13.3 Blowing-out .....................................................................................279
13.4 Tapping ............................................................................................279
13.5 Fanning ...........................................................................................280
13.6 Back Drafting ..................................................................................280
13.7 Conclusions .....................................................................................280
14. IRREGULARITIES IN BLAST FURNACE

OPERATION AND THEIR REMEDIES ............................ 281
14.1 Hanging ...........................................................................................282
Remedies.........................................................................................283
14.2 Scaffolding.......................................................................................283
Contents • xiii
14.3 Slip ...................................................................................................284

14.4 Chilled Hearth ................................................................................284
14.5 Pillaring ...........................................................................................284
14.6 Breakout ..........................................................................................285
14.7 Choking of Gas Offtake ..................................................................285
14.8 Flooding and Coke Ejection Through the Tap Hole ....................285
14.9 Leaking Tuyeres, Tap Holes, and Coolers .....................................286
14.10 Channeling ......................................................................................286
14.11 Salamander Formation ...................................................................287
15. BLAST FURNACE PRODUCTS-II SLAG AND METAL ... 289

15.1 Introduction ....................................................................................289
15.2 Composition of Pig Iron .................................................................291
15.3 Slag Composition ............................................................................293
Slag Volume.....................................................................................294
How to Calculate the Slag Volume?...............................................294
Disposal of Slag ...............................................................................295
15.4 Production of Ferro-Manganese ....................................................297
15.5 The Sulfur Problem ........................................................................297
15.6 Basic Burdening of Blast Furnace and External
Desiliconization ..............................................................................299
15.7 Acid Burdening of Blast Furnace and External
Desulfurization ...............................................................................300
Indian Practice of Acid Burdening ................................................301
15.8 Disposal of Metal ............................................................................302
16. IMPACT OF BURDEN PREPARATION AND

DISTRIBUTION ON BLAST FURNACE PERFORMANCE...305
16.1 Introduction ....................................................................................306
16.2 Burden Distribution .......................................................................307
16.3 Burden Trajectory and Profiles ......................................................308
16.4 Burden Distribution Patterns.........................................................309
xiv • Modern Iron Making Handbook
16.5 Blast Furnace Performance............................................................310

16.6 Effect of Sizing................................................................................312
16.7 Use of Agglomerate as Burden.......................................................313
Sinter ...............................................................................................313
Pellets ..............................................................................................314
16.8 Sinter vs. Pellets ..............................................................................316
16.9 Agglomerate vs. Natural Lumpy Ore .............................................318
16.10 Sizing of Coke and Its Quality ........................................................318
16.11 Flux Quality.....................................................................................319
17. MODERN BLAST FURNACE DESIGN

AND PRACTICE ................................................................. 321
17.1 Introduction ....................................................................................322
17.2 Large-Capacity Blast Furnaces ......................................................323
17.3 Coke Quality ...................................................................................327
17.4 Burden Preparation ........................................................................330
17.5 Improved Distribution of Charge ..................................................331
Stock-line Armor .............................................................................332
17.6 High Top Pressure ..........................................................................334
17.7 Charging Devices for High Top Pressure ......................................337
Tops with Only Bells .......................................................................338
Paul-Wurth/CRM Top ....................................................................338
NKK’s Four Bell Top ......................................................................339
Tops with Bells and Valve Seals ......................................................340
Bell Construction and Operation ...................................................343
Top with Only Valve Seals (No Bells) ............................................344
Gimble Top Distributor ..................................................................345
17.8 Higher Blast Temperature and Driving Rate ................................346
17.9 Oxygen Enrichment of Blast ..........................................................347
17.10 Humidification of Blast...................................................................349
17.11 Fuel Injection .................................................................................351
Pulverized Coal Injection ...............................................................355
17.12 Lime Injection ................................................................................356
17.13 Pre-reduced Ore as Burden ...........................................................359
17.14 Injection of Reducing Gas at the Bottom of the Stack..................362
Contents • xv
17.15 Utilization of Plant Iron-Bearing Wastes .......................................362

18. BLAST FURNACE OPERATIONAL CONTROL ............... 367

18.1 Introduction ....................................................................................367
18.2 Process Variables.............................................................................369
Variety of Operational Variables .....................................................369
18.3 Probes for Measuring the Process Variables..................................371
18.4 BF Operational Models ..................................................................372
18.5 Process Control Strategies..............................................................373
Overall Control Strategy .................................................................374
18.6 Central Working Index ...................................................................376
18.7 Advantages of Process Control .......................................................377
19. ALTERNATIVES TO BF I – LOW-SHAFT AND

CHARCOAL FURNACE ..................................................... 379
19.1 Introduction ....................................................................................379
19.2 The Low-Shaft Furnace .................................................................380
19.3 Mini-Blast Furnaces (MBFs) .........................................................385
Special Features of MBF ...............................................................385
Modern MBFs ................................................................................386
Chinese MBF..................................................................................387
First MBF in India .........................................................................387
Charcoal Blast Furnace ..................................................................389
20. ALTERNATIVE ROUTES OF IRON PRODUCTION II –

ELECTRO-THERMAL PROCESSES.....................................391
20.1 General ............................................................................................391
20.2 Historical .........................................................................................392
India ................................................................................................393
20.3 The Submerged Arc Furnace Plant ...............................................393
20.4 Operation ........................................................................................396
20.5 Smelting Practice ............................................................................397
20.6 Modern Trends ...............................................................................400
xvi • Modern Iron Making Handbook
20.7 Special Features..............................................................................400

20.8 Irregularities in Operations ............................................................401
21. ALTERNATIVE ROUTES OF IRON PRODUCTION III –

SPONGE IRON PRODUCTION .............................................403
21.1 Introduction ....................................................................................403
21.2 Physical Chemistry of DR Processes .............................................405
21.3 Sponge Ironmaking Processes........................................................406
HyL Process ....................................................................................407
21.4 Midrex Process................................................................................411
Other Gaseous Reduction Processes .............................................413
Gas-based Processes .......................................................................413
21.5 Fluidized Bed Processes .................................................................415
21.6 Coal-Based DR Processes ..............................................................416
General Design Features ...............................................................418
Raw Materials .................................................................................418
Product ............................................................................................420
Ring Formation in a Rotary Kiln ....................................................420
Coal-based Process Designs ...........................................................420
Process Control ...............................................................................421
21.7 Coal-Based Shaft Furnaces ............................................................422
21.8 Gasified Coal-Based Processes .......................................................422
21.9 Sponge Iron Production in India ...................................................423
21.10 Metallurgical Properties of Sponge Iron .......................................423
21.11 Use of Sponge Iron .........................................................................425
22. ALTERNATIVE TO BLAST FURNACE IV – SMELTING

REDUCTION (SR) PROCESSES ............................................429
22.1 Introduction ....................................................................................429
22.2 Classification of SR Processes ........................................................432
22.3 Thermo-chemical Model ................................................................435
22.4 Vertical Shaft Furnace Processes ...................................................438
Corex Process ..................................................................................438
Sumitomo SC Process ....................................................................440
Contents • xvii
Kawasaki Process (SR) ....................................................................441

Finex Process ..................................................................................442
Technored Process ..........................................................................443
22.5 Electrical Processes ........................................................................443
INRED Process ..............................................................................443
ELRED Process .............................................................................445
COMBISMELT Process ................................................................447
Plasmasmelt Process .......................................................................447
Fastmelt Process .............................................................................448
22.6 Converter Processes .......................................................................448
Krupp–COIN Process ....................................................................448
PCIG Process ..................................................................................448
HI-Smelt Process ............................................................................448
Reactor Steelmaking Process .........................................................449
AISI-HyL Direct Steelmaking .......................................................450
22.7 Bath Smelting Processes.................................................................450
Romelt (Earlier FLPR) Process .....................................................450
DIOS (Direct Iron Ore Smelting Process) ....................................451
22.8 Rotary Hearth Furnace Processes (RHF) .....................................452
IT mk3 Process ................................................................................452
22.9 TUPKARY-SR Process ...................................................................453
APPENDIX A ............................................................................. 459

APPENDIX B ............................................................................. 465
APPENDIX C ................................................................................475
APPENDIX D ............................................................................ 481
INDEX ....................................................................................... 487

CHAPTER
1
HISTORICAL
In This Chapter
● Ancient Ironmaking
● Historical Ironmaking Processes
● Developments that Led to Modern Ironmaking
● Modern Ironmaking
● Alternative Methods of Iron Production
● History of the Iron and Steel Industry in the United States
● Twenty-first Century Trends (U.S. and the World)
● Concluding Remarks
I
ron is the fourth most abundant rock-forming element and composes
about 5% of the Earth’s crust. Astrophysical and seismic evidence indi-
cate that iron is even more abundant in the interior of the Earth and
has apparently combined with nickel to make up the bulk of the planet’s
core. Geologic processes have concentrated a small fraction of the crustal
iron into deposits that contain as much as 70% of the element. The princi-
pal ore minerals of iron are hematite, magnetite, siderite, and goethite. An
estimated 98% of the ore shipped in the world is consumed in the manu-
facture of iron and steel. The remaining 2% is used in the manufacture of
cement, heavy-medium materials, pigments, ballast, agricultural products,
or specialty chemicals. As a result, demand for iron ore is tied directly to the
production of raw steel and the availability of high-quality ferrous scrap.1
1
Peter Kuck, Iron Ore Statistical Compendium. USGS.
2 • Modern Iron Making Handbook
1.1 Ancient Ironmaking

Although evidence of the use of iron is found in Egypt dating as far back
as 6000 years B.C., it is believed that these were the remains of meteoric
origin and not the product of smelting carried out deliberately or unknow-
ingly by man. It is interesting to note that the word iron is derived from
words meaning “something hard from the sky.” This indicates that the
ancients believed in the extra-terrestrial origin of iron. The antiquity of iron
has been the subject of considerable discussion amongst pre-historians and
archeologists. The word ayas has been interpreted by some to mean iron,
and it closely resembles eisen which means iron in German language. The
earliest archeological reference for iron is found in the eleventh-century
B.C. and not earlier.
It is most likely that iron, as product of smelting, was first found in ashes
of a fire near some outcrop of an iron rich mineral. In the history of human
civilization, the intentional heating of iron mineral with charcoal heralded
the beginning of iron ore smelting or ironmaking. Subsequently, its fabrica-
tion into implements and weapons marked the emergence of the Iron Age
from the prior Bronze Age. It should also be noted that in ancient times the
term steel was not used. The Iron Age means the iron and steel age in the
modern sense of the terms. It may appear strange that although rich iron
ores are widely distributed and even readily and abundantly available as
outcrops, the Iron Age followed the Bronze Age. The anomaly can however
be easily understood by the study of metallurgy of iron vis-à-vis copper.
Copper is nobler and has a lower melting point as compared to iron. It is
and it must have been easier to smelt copper ore than iron ore, producing a
malleable product. The process of smelting iron ore to produce a malleable
product developed later. This product which is now known as wrought iron
did not attract enough attention until it was realized that repeated heating,
in contact with charcoal, and hammering followed by quenching, hardened
the metal to an extent very much superior to bronze. Ironmaking flourished
rapidly thereafter.
The primitive ironmaking involved construction of a furnace using
stone and clay. Its operation consisted of charging a mixture of ore and
charcoal; after initial ignition, air was blown through the charge continu-
ously until the operation was over. A simple, inexpensive and easy-to-work-
with furnace passed successively through various stages of developments,
which finally culminated into a giant modern blast furnace complex.
Historical • 3
The modern blast furnace (B.F.) is a very expensive operation which

uses a computer to control all ironworking activities. This simultaneously
increased the production rate from merely a few kilograms per batch to
several thousand tons (the biggest blast furnace today is Nippon Steel
Corporation’s Oita No. 2 B.F. which produces on an average 13500 tons of
molten iron per day (thm/d)). Increasing the volume of production from
individual ironmaking blast furnaces has also improved upon the specific
fuel consumption in making iron. The earlier furnaces used a kg per kg iron,
and the consumption has now decreased to less than half a kg per kg of iron
in these large blast furnaces.
1.2 Historical Ironmaking Processes

Study of ironmaking would be incomplete without understanding the suc-
cessive stages of development in ironmaking technology over the last few
hundred years; the last few decades are particularly important.
The primitive iron smelter constructed his smelting furnace either
against a hillock or as bowl in the ground using clay and stones. Remains of
the bowl design have been discovered dating around 400 B.C. Charcoal was
used as a fuel and natural draught supplied the blast for burning the fuel.
Oxygen of the blast burned the charcoal to produce heat, and the carbon
monoxide that was generated reduced the iron oxide to iron. The size of
such a primitive furnace was 30–60 cm in diameter and 50–60 cm deep.
These bowls produced a spongy mass of iron weighing 1–3 kgs. The prod-
uct was called a bloom and the place of production was called a bloomery
in Europe.
Once it was realized that the poor yield was due to the short time of
contact of carbon monoxide with the ore it led to the construction of a shaft
above the bowl. The increased height allowed longer contact of ore with the
reducing gases and a higher recovery of iron was achieved. Such a furnace
was constructed, with respect to the wind direction, in such a way that it
operated on natural draught, with an opening provided at the bottom of the
furnace to admit air in and the exhaust gases to escape from the open top.
Later, the natural draught was replaced by forced air blast generated by the
use of crude bellows made from animal hides. The designs persisted until
the thirteenth-century.
The next development took place in the thirteenth-century in the
Spanish province of Catlonia where a smelting unit was invented that
increased the production rate and decreased the fuel consumption. The
smelting unit came to be known as the Catalan Forge. The air bellow was
replaced by a falling water device known as a trompe to force air blast
through the furnace. A pair of trompes was used for each furnace to keep
an uninterrupted supply of blast to the furnace. In such a furnace about
0.5 t of 40–50% iron ore yielded about 100 kg of metal consuming nearly
1.5 t of charcoal, and the operation lasted nearly 12 hours.
Supporting
structure
Arch
Bellow To Water
Wheel
Tuyere
FIGURE 1.1 Stukofen, the progenitor of the modern blast furnace.
In Europe, as the demand for iron grew, the size of the Catalan furnace
progressively increased resulting in the evolution of the German Stukofen as
shown in Figure 1.1. This is considered to be the progenitor of the modern
blast furnace. It was about 3.0–4.5 m in height, built of stone and clay and
had a round or a rectangular cross-section. Air blast was supplied through two
tuyeres by water-driven bellows and a false door was provided for the extrac-
tion of the blooms. The output of such a furnace was about 45–50 t per year.
In the course of development, using more dense charcoal, increasing
the height of the furnace, and increasing maximum blast pressure eventu-
ally resulted in the generation of adequate temperature in the furnace such
Historical • 5
that the product was molten iron. The molten iron flowed out from the
furnace as distinct from the slag impregnated blooms obtained from the
earlier furnaces. The product, on solidification, was not malleable, but it
was soon realized that it could be cast into useful shapes. This perhaps is the
origin of ferrous foundry.The product of the Stukofen was run into a series
of parallel channels, where it cooled and solidified, somewhat distantly
resembling a nursing litter of pigs. This is how the term pig iron originated
for the metallic product from an iron ore smelting furnace. The pig iron,
after solidification, was broken up, re-melted and used for the production
of castings. Thus, cast iron was discovered.
Once the product was obtained in molten condition, the furnace opera-
tion could be made continuous by charging the material from the open
top and tapping the liquid product from the bottom without interruption.
A simple and inexpensive process was thus superseded by a complex pro-
cess to increase the production rate, and the same process continues to
dominate this field today. The data relevant to these different types of the
furnaces are given in Table 1.1.
TABLE 1.1
Weight of iron Fuel consumption Recovery % Production
per smelting % of the ore t/week
operation (kgs)
Blooming hearth 60–70 450 20 0.03–0.15
Catalan Forge 140–160 360 31 0.5–0.7
Stukofen 300–900 250 40 1.0–3.0
The solidified product of Stukofen, i.e. pig iron, was not workable in con-
trast to the malleable bloom secured from the blooming hearth and the
Catalan Forge. The Stukofen could not replace the earlier iron ore smelt-
ing processes until the development of puddling process by Henry Cort in
1784, which converted pig iron into a malleable wrought iron.
1.3 Developments that Led to Modern Ironmaking

After the successful adoption of blast furnace for ironmaking in the fifteenth
century, the process continued to be used for nearly 200 years without any
significant changes in its design and practice. Charcoal was universally used
as a fuel although it suffered from some disadvantages. Because of its fri-

able nature, charcoal as a fuel imposed severe limitations on the height of
the furnace shaft. This resulted in restricting the size of the furnace and
consequently its production rate. Charcoal was also expensive because the
conversion of wood to charcoal accounted for nearly half of the total operat-
ing cost. The enormous amount of forest wood that was being consumed for
this purpose finally led to the imposition of restrictive legislation on wood
burning for charcoal.
These reasons forced the iron smelters to look for an alternative to
charcoal; attention was focused on coal which was far more widely avail-
able. Although the sulfur content of coal had earlier prevented its use in
bloomeries, new approaches had to be developed to find a way out of the
sulfur problem; the presence of sulfur makes iron brittle during hot work-
ing. Another major difficulty was how to produce air blast at pressures capa-
ble of penetrating a column of material 6–8 m in height with water-powered
bellows? After several failed attempts, Dud Dudley successfully smelted
iron ore with coal at the ironworks in Worcester, England in 1621.
Coke Furnace
A real breakthrough in ironworking had to wait until the advent of coked
coal, which had good strength and cellular structure. In 1709, Abraham
Darby successfully smelted iron ore with coked coal, which is now called
coke. His success was mainly due to the employment of a large furnace and
a powerful blast to cause ignition of the coke. This was made possible by
the invention of the steam engine by James Watt and its ingenious adoption
by John Wilkinson for producing a powerful blast. Being sufficiently strong,
the use of coke permitted construction of much bigger furnaces for it could
withstand a much taller overlying burden in the furnace. With the coke’s
open cellular structure, penetration of blast for efficient combustion was
also achieved.
Until the middle of the nineteenth-century, the furnaces were open at
the top, and the waste gases used to burn at the top when they came in con-
tact with atmospheric air. In 1845, Bunsen and Playfair demonstrated the
enormous waste of heat in allowing the gases to burn at the top. They also
concluded that these gases could be conducted elsewhere without interfer-
ing with the furnace operation for burning of lime or brick kilns.
The effective use of gas was however far from easy since it contained
high amounts of dust. The enlargement of the down-comer in the beginning
Historical • 7
served as a dust catcher, and the gas once cleaned could be used for pre-
heating purposes with some difficulties.
Hot Blast Generation

Several important discoveries, which led to improvement in thermal effi-
ciency of the blast furnace, were made in the first half of the nineteenth-
century. The observation that blast furnace output was better in winter
months than in summer was quite erroneously attributed to its lower tem-
perature in winter months than in summer. Neilson, a Scottish engineer,
conceived correctly in 1829 that the thermal efficiency of a blast furnace
could be improved by using a preheated blast. Cast iron pipes were used
to preheat the blast. The mechanical ingenuity of Cowper in 1857 finally
led to the development of a stove to preheat the blast. The hot waste gases
from the furnace-top were drawn off to preheat the stove checker bricks,
which in turn heated the blast later. This led to a very large saving in fuel
consumption. The most modern blast furnace stove still has several of the
features originally developed by Cowper himself.
A minimum of three stoves, preferably four, one of which is on the
blast while the others are being heated up, are attached to a modern blast
furnace. The clean gas lends itself to effective utilization and this contrib-
utes in no small measure to the economy of blast furnace operation. The
successful adoption of the Cowper stove for a blast furnace was possible
mainly due to the development of the cone and bell (or cup and cone)
charging arrangement in 1850. In this arrangement, a conical hopper was
fitted on the furnace top. The hopper was closed or opened for charging by
the bell. The exhaust gases could thus be carried through pipes to wherever
required. The single bell design was later on improved to a double bell
arrangement which continued to be used until recently.
Blast Furnace Gas Cleaning

The use of a Cowper stove for preheating the air blast necessitated the blast
furnace exhaust gases to be thoroughly cleaned to avoid the dust contained
in the gas from clotting the checker work in the stove. This led to the addi-
tion of an elaborate gas cleaning system to the blast furnace complex. In
the beginning, the coarse particles in the exhaust gases were removed by
passing the gases through a large brick-lined chamber in which the velocity
of the gases was reduced to allow their settling. This has now been devel-
oped into the dust-catcher. The finer particles which still remained in the
gas were removed in wet scrubbers wherein the gas was subjected to water
sprays. Obstruction to the upward flow of gases by means of baffles ensured
intimate contact of gas and water. The use of scrubbers could decrease the
dust to a level of few grains per cubic meter. The cleanliness had to be
improved still further in order to ensure smooth stove operation without
clotting. The electrostatic precipitator was developed for this purpose by
F.G. Cotrell in the early part of the twentieth-century. An assembly of dust
catcher, scrubber, and an electrostatic precipitator forms the gas cleaning
system even in the modern blast furnace plants.
Charging Devices
The blast furnace could not have increased its production rate without the
parallel developments in mechanical methods for handling of raw materials
and the products of the furnace. By 1870, vertical hoists were used to lift
the charge barrows to the furnace top where they were wheeled on to the
edge and the contents were dumped on to the bell. The job at the furnace
top and pig casting bed had then been described as tasks beyond human
endurance. In 1883, the skip hoist was introduced on the famous Lucy fur-
nace near Pittsburg, PA. The bucket hoist also came into vogue in Germany
around the same time. It was not a difficult matter to obtain the desired
uniform distribution of the charge by manual charging. The mechanical
charging, however, gave rise to a serious problem of distribution of charge
in the furnace, a factor so vital to the smooth running of any furnace. The
mechanical charging therefore led to the development of Brown, Neeland,
and Mckee distribution systems to distribute the charge more evenly in
the furnace. Although the distribution systems continue to improve in their
design, the principles are essentially the same. The mechanical charging
led to the development of stock bins, the high line and increasing level of
automation in charging and weighing, so much so that in a modern furnace
these operations are carried out nearly fully automatically.
By the beginning of the present century, daily production of an indi-
vidual furnace had reached nearly 100 tons of pig iron per day. Such a fur-
nace required nearly a ton of air blast per minute (i.e. nearly 1000 m3/min
at atmospheric pressure). This was possible only by using centrifugal fan
blower for it could only force such a large volume of blast across the resis-
tance offered by the stove, the tuyeres, the charge, etc. A 100 t/day furnace
required nearly 200 t of ore, 100 t of coke, 50 t of limestone, etc. to be
charged in the furnace daily. The earlier bucket charging arrangement had
to be dispensed with to make way for a two-skip hoist charging arrangement.
Historical • 9
Pig Casting Machine

Until the end of the nineteenth-century, the liquid pig iron was cast in a
sand bed. The furnace was generally tapped once every eight hours and this
allowed enough time for solidification, removal of solid pigs and prepara-
tion of the cast house for the next tapping. As the capacity of the furnace
increased, the amount of metal tapped at any time and the frequency of tap-
ping increased. In consequence, the casting operation became too arduous
and costly to be carried out manually. The pig casting machine was invented
in 1894 by Uehling to expedite and mechanize the casting operation. The
introduction of basic open hearth hot-metal practice around the same time
for conversion of molten iron into steel required the metal to be kept in stock
and the mixer was developed. This eliminated much of the casting operation
and only the foundry grade pig iron required a pig casting machine.
1.4 Modern Ironmaking

The general outline of the modern blast furnace including its accessories
like the stove, the gas cleaning system, the raw material storage facilities,
charging system, etc. had been developed by the beginning of the present
century. The increase in production rate by this time had been mainly due
to the increase in the size of the furnace rather than any other significant
innovation and changes in the furnace design and its operation. By 1900,
the blast furnace was able to produce about 400 thm/day spending nearly
2500 kg of coke per t of hot metal. This improved to 1500 t at 1000 kg coke
rate by 1950 and is now 7 to 13000 t at much less than 500 kg coke rate.
The works of Carnot, Clapeyron, Clausius, Helmohltz, Gibbs, Stirling
and several others led to the establishment of the science of thermodynam-
ics by the 1850s. Being well-versed in the physical sciences and having a
remarkable mechanical ingenuity, Bessemer, Kelly, Thomas, the Siemens
brothers, Martin, Heroult and several others set the pace of large scale steel-
making by the end of the previous century. The advent of twentieth-century
marked the beginning of systematic investigations towards understanding
of the physical and chemical principles underlying the iron and steelmaking
processes. The blast furnace operation underwent several modifications in
its design and practice in the light of this understanding. The advantages of
using screened raw materials, particularly sized ore and coke, became quite
apparent. The agglomeration of ore fines as sinter or pellets proved as an
economic burden in spite of the additional processing cost involved in these
operations. These changes in quality resulted in a considerable increase in

production rates from the same furnaces.
The pace for developing the blast furnace operation to improve its
production rate and decrease the fuel consumption, in particular that of
coke, was set during and after World War II. The use of fluxed and super-
fluxed sinter and/or pellets proved to be technically feasible and qualita-
tively a better burden than the sized ore or un-fluxed agglomerate. Blast
furnaces are now generally operated with at least some proportion of sinter
and/or pellets. The proportion of sinter and/or pellets has been progres-
sively increased since the 1950s so much so that in some of the most mod-
ern Japanese furnaces the entire ore charge is in the form of sinter and/
or pellets.
The use of oxygen enriched blast and humidified blast is increasingly
being used and has become a common practice. Similarly, blast furnaces
are now being operated under high top pressure to the tune of 2.0–2.5 kg/
cm2 gauge, to improve gas-solid reactions in the stack. Since World War II,
in order to maintain the increasing rate of production, good quality coking
coal has become scarce and therefore a costly commodity in relation to its
consumption. Since coke serves three functions as a spacer, a heat producer,
and a reducing agent, attempts are underway to substitute at least a part of
the coke requirement by other available fuels. Various liquid, gaseous and
fluidized fuels have been successfully injected through the tuyers to substi-
tute as much of the coke requirement as is possible without impairing the
production rate. All these attempts along with other innovations and modi-
fications in furnace design and operation have succeeded, and the coke rate
has been brought down to an average of 0.7 t/t of pig iron produced with
lowest value of 0.4–0.5 t/t in some of the most modern Japanese furnaces.
It may be of interest to recall here that the coke consumption was nearly
2.5–3.0 t/t of iron production nearly a century ago and about 0.9–1.1 t/t of
iron around 1950. The present low coke rate value is in itself a substantial
single achievement in the development of the blast furnace technology.
1.5 Alternative Methods of Iron Production

Coking coal is a valuable and scarce commodity on which the smelting of
iron ore in a blast furnace inescapably depends. The availability of the right
type of coking coal is the chief obstacle for installation of larger blast fur-
naces. Blast furnace sizes have increased over the years, but this was possible
Historical • 11
because of matching improvements in quality of coking coal. Attempts

were taken to establish alternative processes of ironmaking bypassing the
traditional route of the blast furnace. These methods fall broadly in the fol-
lowing four categories:
1. Use of electric-arc furnace to partly replace role of coke by electrical

power, so that coke is required only for reduction purposes;
2. Adoption of low-shaft or small blast furnaces such that use of poorer
quality coke and small size coke is possible for producing iron;
3. Production of sponge iron or DRI, HBI, etc. using gaseous or cheap
coals as reluctant and fuel to produce iron in solid state;
4. Development of reduction-smelting processes wherein poor quality
coal may be very efficiently used for molten iron production.
The history of these alternative processes of ironmaking is almost a century
long saga. Many of these processes were developed to the extent that they
could compete with the traditional blast furnace route, at least, under some
specific local conditions. The first such process to have been adopted in
practice was the Tysland-hole electric furnace for iron ore smelting. It was
introduced in the early 1930s in the Scandinavian countries where cheap
hydro-electric power was available and no coking coals were present. This
alone led to the development of the submerged electric arc smelting fur-
nace later in Norway. This basic design was used for electric smelting of iron
practically all over the world. As an alternative process, the low shaft fur-
nace was developed in Germany during World War II and was extensively
used. It used iron ore with thermal coal as charge. One such proto-type
was set up at the National Metallurgical Laboratory, Jamshedpur in the
1970s to assess its feasibility for iron production in India. It however proved
unsuccessful.
A number of different techniques for direct reduction of iron ore have
been developed over the last one hundred years since the idea was first
put forth in 1873 in England. The Wiberg and Hoganas processes were
successful in Sweden solely due to the peculiar local conditions. The first
viable direct reduction operation that attracted widespread interest, as an
alternative process of iron production, was the HyL-Process plant set up
in Mexico in 1954 on a commercial level. Thereafter, several other pro-
cesses employing either a vertical shaft or an inclined rotating kiln have
since been developed and put into commercial practice all over the world.
In all these processes the iron ore is reduced in a solid state. It thereby pro-
duces iron only in solid state with a honeycomb-type porous structure, com-
monly known as sponge iron or pre-reduced iron or directly reduced iron
or DRI for short. The processes for the production of such a product are
often called as DR processes. Hot Briquetted Iron or HBI for short is only
a specialized variety of the same product. The technology has now reached
a level where any one of the direct reduction processes coupled with the
electric furnace process for steelmaking can compete with the traditional
established route of blast furnace and oxygen steelmaking under most of
the local conditions.
The coal based rotary kiln-type plant and the MIDREX gas based plant
was adopted more widely in India.
In 2006, the world production of DRI from all such processes was
59.8 Mt against the total steel production of 1245 Mt from all processes.
The production of DRI/HBI in India in from 2005–2006 was 16.3 Mt from
a total of around +250 different units adopting different technologies. The
dominant technology was coal based rotating kilns, but included 3 gas-
based units of total 6 Mt capacity. It aims to produce 33% of total steel
production, using alternative methods of iron production in India by the
year 2020. The number of DRI units is bound to increase further. India is
the largest sponge iron producer in the world today.
The Smelting-Reduction processes, developed only in the last couple
of decades, are advanced versions of DRI technology. In contrast to the
product being in a solid state in DRI processes, the SR processes pro-
duce liquid pig iron without using coke in a blast furnace. The reducing
gases generated in the melting operation, in these processes, are intelli-
gently used for solid state reduction as a prior step. These are known as SR
Processes and at least six different commercially viable SR-Technologies are
now available for commercial exploitation. The first to be developed was
the COREX process, adopted in India for the first time at the JSW Steel
Plant at Vijayanagaram in Karnataka. However, it has not been able to use
entirely ordinary coal as fuel, as was envisaged in the original design. It uses
a substantial proportion of coke as fuel like the blast furnace. Although vari-
ous trials have been carried out using coal as fuel, it is yet to be established
on a routine basis. However, the SR processes based on non-coking coals
are bound to come up in the future. The prediction is that hot metal would
be produced in equal proportion, from the blast furnace and the SR tech-
nologies in the near future.
Historical • 13
1.6 History of the Iron and Steel Industry in the United States2
The U.S. iron and steel industry has paralleled the industry in other coun-
tries in technological developments. In the 1800s, the U.S. switched from
charcoal to coke in ore smelting, adopted the Bessemer process, and saw
the rise of very large integrated steel mills. In the 20th century, the U.S.
industry successively adopted the open hearth furnace, then the basic oxy-
gen steelmaking process. Since the American industry peaked in the 1940s
and 1950s, the U.S. industry has shifted to small mini-mills and specialty
mills, using iron and steel scrap as feedstock, rather than iron ore.
240
220
Pig Iron
200
Steel
180
Millions of Metric Tons
160
140
120
100
80
60
40
20
0
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020
FIGURE 1.2 Graph of U.S. iron and steel production, 1900–2014. Data from USGS.
U.S. production of iron and steel peaked in 1973, when the U.S. indus-
try produced a combined total of 229 million metric tons of iron and steel.
But U.S. iron and steel production dropped drastically during the recession
of the late 1970s and early 1980s. From a combined iron and steel produc-
tion of 203 million tons in 1979, U.S. output fell almost in half, to 107 mil-
lion tons in 1982. Some steel companies declared bankruptcy, and many
permanently closed steelmaking plants. By 1989, U.S. combined iron and
steel production recovered to 142 million tons, a much lower level than in
the 1960s and 1970s.
Iron manufacture before the 19th century required charcoal, and
Britain’s once-vast forests could no longer supply enough charcoal for the
2
Source: USGS
nation’s increasing need for iron. By 1700, Britain was becoming increas-
ingly dependent on iron imported from its sometimes-adversary Sweden.
Britain looked to the seemingly limitless forests of its American colonies
to supply Britain with iron. British investors started an iron furnace near
Perryville, Maryland, which in 1718 started exporting iron back to Britain.
That success prompted formation of more companies, which built numer-
ous iron furnaces around Chesapeake Bay, supplied by bog iron ore, which
was widespread. By 1751, Virginia and Maryland were exporting 2,950 tons
of pig iron to Britain each year; at the time, British iron production was
about 20,000 tons per year.3
While the Chesapeake Bay furnaces were established for export, iron
furnaces were established in the 1700s throughout the American colonies
for domestic consumption. Iron furnaces were located along rivers to sup-
ply water power. Also required were forests for charcoal, iron ore, and lime-
stone for flux. In addition, the furnace needed to be close to a major market
or close to water transport.
British business interests were split on colonial iron: manufacturers
appreciated the lower prices due to colonial imports, but the British iron
and steel industry objected to the competition. Parliament compromised in
the Iron Act of 1750, which eliminated the import duty on colonial pig iron,
but barred the manufacture of steel or of iron plate in the colonies. The law
was widely ignored by colonial governments.
By 1776, up to 80 iron furnaces throughout the American colonies were
producing about as much iron as Britain itself. If one estimate of 30,000 tons
of iron each year is accurate, then the newly formed United States was the
world’s third-largest iron producer, after Sweden and Russia.
The movement away from charcoal in U.S. iron smelting began in
1827, when a puddling furnace in Phoenixville, Pennsylvania started
using anthracite coal. Blast furnaces continued to use only charcoal until
about 1840, when coke from coal started replacing charcoal as the fuel
and reducing agent. Coke has a higher crushing strength than charcoal,
allowing larger smelting furnaces. Because iron and steelmaking at the
time consumed more coal than iron ore, the steel mills moved closer to
the coal mines to minimize transportation costs. A problem of coke was
that it carried impurities such as sulfur, which degraded the quality of the
3
https://en.wikipedia.org/wiki/History_of_the_iron_and_steel_industry_in_the_United_States
Historical • 15
steel. Although coke quickly became the dominant fuel for iron-smelting,
in 1884 charcoal was still used to make ten percent of iron and steel in the
U.S.. The use of charcoal for steelmaking survived in the U.S. on a small
scale until 1945.
U.S. production of iron and steel peaked in 1973, when the U.S. indus-
try produced a combined total of 229 million metric tons of iron and steel.
But U.S. iron and steel production dropped drastically during the recession
of the late 1970s and early 1980s. From a combined iron and steel produc-
tion of 203 million tons in 1979, U.S. output fell almost in half, to 107 mil-
lion tons in 1982. Some steel companies declared bankruptcy, and many
permanently closed steelmaking plants. By 1989, U.S. combined iron and
steel production recovered to 142 million tons, a much lower level than in
the 1960s and 1970s.
1.7 Twenty-first Century Trends (U.S. and the World)

Within less than a couple of decades since 1991, a phenomenal change
and in the steel industry, in particular, has taken place. The striking feature
has been the dominance of NRIs in the iron and steel industry all over the
world. This period also witnessed a global consciousness demanding the
production of quality steels and improvement in production economy. The
World Steel Dynamics in June 2001 declared Tata Steel (India) as the best
steel plant in the world, producing one of the cheapest steels. This ranking
was based on 17 different parameters, which included all the areas of oper-
ations of an integrated steel plant from production to marketing, customer
care, research and development efforts, and resource management. The
Tata Business Excellence Model acted as a key driver to achieve this quality
level. In its earlier form as TISCO and in its newly christened form as Tata
Steel, it has the oldest blast furnace and is yet very efficient by present day
world standards. Against the global benchmark of 40 t/m2/day sinter plant
productivity, they have already achieved 35 t/m2/day as compared to their
own figure of 25 t/m2/day a decade ago. In the same period, the specific raw
material consumption has come down from 4.66 t to 3.33 t/t of hot metal.
Similarly, they have revolutionized their coke making using stamp charging
which allows non-coking coal in the charge-mix while making good grade
metallurgical coke. Their latest BF produces 2.0 to 2.25 t/m3/day with a
coke rate of only 450 kg/thm and an additional 150 kg of pulverized coal
injection.
Against the world total steel production of almost 1250 Mt in 2006 and
1343 Mt in 2007, the Indian contribution was only nearly 55 Mt. This was
only 33 Mt until very recently. The redeeming feature, however, is that it is
going to be nearly 120 Mt by the year 2012 and 180 Mt by 2020. 2020 has
been earmarked as the benchmark year for attaining total overall self suf-
ficiency in many other respects as well (based on the TIFAC report). The
steel production will have to be matched by equivalent production of hot
metal from BF or other processes and DRI from different processes.
U.S. Production of Selected Mineral Commodities in the First

Quarter 2017
U.S. mine and plant production data for 14 selected mineral commodities
are provided monthly (or quarterly) by the U.S. Geological Survey to the
Board of Governors, Federal Reserve System (FRS), for use in preparing its
index of industrial production and the related capacity indexes and capacity
utilization rates. These measures cover manufacturing, mining, and electric
and gas utilities, and they are among the key economic indicators moni-
tored by the FRS for guidance in determining national monetary policy.
Base Metals
Production of copper, iron ore, and zinc decreased by 7%, 2%, and 6%,
respectively in the first quarter of 2017, whereas secondary aluminum
production increased by 2% compared with the fourth quarter of 2016
(Figure 1.3).
FIGURE 1.3 U.S. production of selected base metals from the first quarter of 2012 through the first
quarter of 2017, indexed to the first quarter of 2012. Source: U.S. Geological Survey.
Historical • 17
1.8 Concluding Remarks

Iron production is the inevitable first step necessary for steel production
using natural resources. Iron can be produced by any one or more of the
three dominant alternative processes: BF, SR, and DRI routes.
The blast furnace has now become synonymous with extraction of iron,
although it is also used for the extraction of metals like lead, copper, etc.
The modern blast furnace process is the consummation of accumulated
experience and developments over several centuries, if not millennia. The
dominant role of blast furnace in the production of iron is obvious from the
share of molten iron in the overall charge in finally producing steel. The
SR processes are expected to contribute more and more proportion of hot
metal and proportionately reducing the role of blast furnace. The share
of BF is continuously falling and the share of DRI/HBI and the share of
hot metal from SR processes is increasing on global level. This is shown in
Table 1.2.
Table 1.2 Share of Different Metallics as a Percentage of the Total Used

for Steelmaking
Iron source 1995 2000 2005 2010
Hot metal 58.9 57.2 55.0 52.5
DRI/HBI 3.4 5.2 6.6 8.1
Scrap (recirculating) 12.8 9.6 8.8 8.0
Scrap (collected) 24.9 28.0 29.5 31.4
The table clearly shows that the dominating role of blast furnace is
slowly decreasing; this is due to a shortage of coking coal and its attendant
ascending cost. For producing steel, the proportion of metallic produced by
the alternative methods is therefore continuously increasing; it is becoming
relatively more economical and locally suitable.
Uses of Iron and Steel Alloys

Iron and steel are used widely in the construction of roads, railways, other
infrastructure, appliances, and buildings. Most large modern structures,
such as stadiums and skyscrapers, bridges, and airports, are supported by a
steel skeleton. Even those with a concrete structure employ steel for rein-
forcing. In addition, it sees widespread use in major appliances and cars.
Despite growth in usage of aluminum, it is still the main material for car
bodies. Other common applications include shipbuilding, pipelines, min-
ing, offshore construction, aerospace, white goods (e.g. washing machines),
heavy equipment such as bulldozers, office furniture, steel wool, tools, and
armor in the form of personal vests or vehicle armor (better known as rolled
homogeneous armor in this role).
Modern Standards
Most of the more commonly used iron and steel alloys are categorized
into various grades by standards organizations. For example, the Society of
Automotive Engineers has a series of grades defining many types of steel.
The American Society for Testing and Materials has a separate set of stan-
dards, which define alloys such as A36 steel; the most commonly used struc-
tural steel in the United States. The JIS also define series of steel grades
that are being used extensively in Japan as well as in third world countries.
CHAPTER
2
MODERN BLAST FURNACE
PLANT AND PROCESS
In This Chapter
● Introduction
● Modern Blast Furnace
● Hot Blast Stove
● Gas Cleaning
● Raw Materials Storage and Handling
● Liquid Products Disposal
● Blast Furnace Operation and Control
● Essentials of the Blast Furnace Process
● Efficiency of Blast Furnace
2.1 Introduction
The modern blast furnace plant consists of the following important sections:
1. Blast furnace proper
2. Hot blast supply equipment
3. Gas cleaning system and gas storage
4. Raw material storage and handling
5. Liquid products disposal

6. Process control equipment
The schematic arrangement of the various sections in a modern blast fur-
nace plant is shown in Figure 2.1. A typical lay-out of the blast furnace plant
is shown in Figure 2.2.
FIGURE 2.1 Schematic arrangement for various sections of a blast furnace plant.
2.2 Modern Blast Furnace

A modern blast furnace is a nearly 30 m (about 100 feet) tall welded plate
construction with circular cross section of varying sizes at different levels.
It is illustrated in Figure 2.3 to show the essential metallurgical design fea-
tures and various important parts of the furnace. The cross-sectional area
increases from the top or the throat downwards, the maximum area is at
the bosh level (mantle) and it decreases downwards thereafter. The struc-
ture of the furnace essentially consists of a massive foundation, the hearth,
the bosh, the mantle and columns, the stack, the raw material hauling and
charging facilities, and the top as shown in Figure 2.1. For more details see
Reference 11.
Modern Blast Furnace Plant and Process • 21
The typical set of dimensions of a furnace producing nearly 3000 thm/d

with an effective inner volume of about 1500 m3 are shown below:
Hearth diameter 8–9 m

(26–28 ft)
Bosh diameter 9–10 m
(28–30 ft)
Hearth height 3–3.5 m
(10–12 ft)
Bosh height 3–4 m
(10–14 ft)
Stockline diameter (throat) 5.5–6.5 m
(19–22 ft)
Stack height 18–20 m
(60–65 ft)
Total overall
Useful height 30–32 m
(100 ft)
Working height 21–24 m
(tuyere line to throat) (70–80 ft)
Total useful volume 13–1600 m3
(45–55000 ft3)
Stack batter one in twelve.
The outer welded steel shell is lined from inside with refractory lining to stand
the smelting conditions during its operation. Previously the furnace was sup-
ported by a box-cage like structure, but now it freely stands only on its founda-
tion without lateral support.
Furnace Details
Foundation. It is a massive steel reinforced concrete mass partially
embedded below the ground level. It should be sufficiently strong to
stand the loaded furnace weight, which may be as much as 10,000 t for
a 2000 t/day furnace. It may be about 15 m in diameter and 6–8 m thick
upon which is placed the furnace bottom consisting of about 4–5 m thick
of fire bricks.
FIGURE 2.2 Layout of a modern blast furnace plant.
Hearth. It is a receptacle to collect the liquid slag and metal and is also
referred to as a crucible. The old practice of using fire bricks for hearth con-
struction is almost universally replaced by carbon blocks. Water cooled copper
or steel plates are laid in the side walls to protect the lining. The carbon wall
may be more than a meter of uniform thickness or a stadium-type construc-
tion. The hearth wall contains a tap hole for iron, 12–15 cm in diameter and
about 0.3–0.6 m above the hearth bottom level, and a slag notch 1.2–1.6 m
above the iron tap hole level, but staggered through a right angle in the hori-
zontal plane. These holes are closed with clay when not in use and are opened
for tapping slag and metal as and when necessary. The tuyeres are located at
the top level of the hearth uniformly distributed over the entire cross-section.
Bosh. The top of the bosh has the maximum diameter of the furnace
and it is the zone of intense heat. It is a stadium-type (stacked cut bricks)
construction with steel reinforcement. Water cooled copper or steel plates
are inserted at regular intervals in the furnace lining in this zone to effect
protection against high temperature. The intensive cooling of a thin-walled
bosh forms a layer of solidified slag mixed graphite on the lining which pro-
tects the brickwork from molten metal and slag.
Mantle and Columns. The furnace structure above the bosh level is sup-
ported on a heavily braced steel ring encircling the furnace at the top of the bosh.
This is called the mantle which is supported by uniformly spaced upright heavy
columns that are firmly anchored in the concrete foundation at the bottom.
Stack. It is a frustum of a huge cone mounted on the mantle that extends
to the top of the furnace. The furnace top, that is the bell, the charging
arrangement, the gas off-takes, etc. are mounted on top of the stack. The
stack is welded steel plate construction lined from inside to a thickness of

about a meter of hard firebricks. Flat water cooled plates are inserted in the
lower part of the stack lining. The top 2–3 m height, which is the stockline,
is protected from abrasion caused by the falling charge by providing armor
plates on the inner surface of the lining.
Tuyere and Bustle Pipe. The tuyeres are located immediately above
the hearth and hot air blast is blown through the tuyeres for fuel combus-
tion (see Figure 11.10). The number of tuyeres vary with the size of the fur-
nace. Usually it is an even number between 10–20, and they are uniformly
spaced all over the periphery.
Air from hot blast stoves is supplied to a huge circular pipe encircling
the furnace at the bosh level. This is called the bustle pipe. The individual
tuyeres are connected to the bustle pipe, which by virtue of its enormous
size, equalizes the pressure of the blast at all the tuyeres.
Bell and Hopper. A bell and hopper or the cup and cone arrangement
was once commonly used in blast furnaces for charging the solid charge (see
Figure 5.5). It allowed smooth charging without the off-gas leaking out into
open. The off-gas, which has some fuel value by virtue of its CO content,
and which is available in large amounts could thus be collected, cleaned and
utilized as a by-product fuel, for pre-heating the blast and the still leftover gas
could be used elsewhere in the plant. This design lasted for a long time, with
or without some additional facilities to improve upon the charge distribution
inside the blast furnace. A modified bell-less top with distributing chute known
as Paul-Wurth Distributor (BLPWD for short) slowly replaced the traditional
double bell arrangement since the 1980s . The replacement was nearly com-
pleted by the beginning of twenty first century, except for some exceptional or
the small furnaces. This improved the burden distribution inside the furnace
and thereby improved the blast furnace efficiency considerably.
Towards the later half of the first decade of the twenty-first century
a still better modified and more efficient top has been designed and put
into use. This is known as gimble top, which has been developed by the
Siemens-Voest-Alpine group. New blast furnaces or those being modern-
ized have started adopting it as a better alternative.
When high top pressure was introduced several complex seals had to
be used on the top with double bell arrangement. The Paul-Wurth bell-less
top proved very effective from this point of view as well. The gimble top is
thought to be still more efficient from the point of view of adopting high
top pressure, and the designers claim it is more effective for better charge
distribution and easier mechanical maintenance.
Off-takes. There are four exhaust pipes which are connected to the
furnace top evenly at four points. These rise vertically up above the furnace
top and then join to a bigger single pipe known as the down comer which
delivers the gas to the gas cleaning system (i.e. dust catcher).
2.3 Hot Blast Stove

The air blast is preheated to a temperature of 700–1300 C in Cowper
regenerator stoves. A set of three or four regenerators is provided for each
furnace. The stove is about 6–9 m in diameter and 30–35 m in height.
Special thin walled bricks are used to construct the checker work in the
stove. While working one stove is on-blast, heating it, while the remaining
two or three are on-gas, getting heated by burning the cleaned blast furnace
gas. The earlier stove had nearly 5000 m2 as the checker surface area for
heat exchange, but the modern stove may have about five times this much.
The thermal efficiency of such a stove is around 80–90%. Several valves
are provided on the stove assembly to carry out the changeover from gas to
air and vice versa smoothly. Steam driven centrifugal blowers are generally
employed to supply the air blast at a uniform rate to the furnace. These are
located in the blower house near the furnace.
2.4 Gas Cleaning

The blast furnace gas is cleaned thoroughly before being used as a fuel. The
down comer opens up in a dust catcher where coarse particles contained in
the gas settle down by virtue of the change in their direction of flow. The
coarse cleaned gas then passes through wet or dry scrubbers and electro-
static precipitators. Nearly 25% of the gas obtained from the furnace is
consumed in the stove for producing hot preheated blast. The nearly 75%
of the remaining gas is consumed elsewhere in the plant. This is known as
BF gas. A gas holder may be provided to store the surplus gas.
2.5 Raw Materials Storage and Handling

To produce one ton of pig iron nearly two or three tons of solid charge
materials are charged in the furnace. It means that a furnace producing
5000 t of pig iron per day would need nearly 10000–15000 t of solid charge-
materials per day. The successful operation of such a furnace is therefore
a function of regular supply of these raw materials to the large scale fur-
nace. A sufficiently large and readily available reserve stock of these raw
materials at the furnace site is essential to avoid any interruption in the
production. Equally large storage facilities in the form of an ore yard and
storage bins are always provided in a blast furnace plant for these materials.
Facilities for unloading, blending, etc., particularly in the plants located on
waterfronts and dependent on imported ores, are also provided. Efficient
charging and accurate weighing facilities are required in any modern plant
to achieve smooth operation.
Weighing charge materials on such an enormous scale is a challenging
task. The accuracy and efficiency of weighing is always in question. Load
cells are very commonly used for such weighing but the accuracy always
remains questionable and a matter of dispute between the supplier and the
buyer departments.
2.6 Liquid Products Disposal

Three different products are generated while smelting iron ore in a blast
furnace. The gas comes out of the furnace from top and is processed in the
gas cleaning plant to produce clean BF gas as a fuel for reuse. Since the iron
ore contains some impurity oxides like silica, alumina, calcia, magnesia, etc.,
along with the ash arising out of the coke, it produces a molten oxide liquid,
with or without the help of a flux like limestone and which flows out of the
furnace as a thin oxide liquid called slag. Slag is lighter than metal, and it
floats above the metal layer in the hearth. Iron produced after reduction of
iron ore get saturated with carbon, and containing some other impurities,
collects at the bottom of the hearth from where it is tapped periodically.
The slag and the metal are the two liquid products that are to be removed
from the furnace hearth periodically.
A sufficiently large, raised platform is provided in front of the furnace
at a level just below that of the metal tap hole. Firebrick channels are con-
structed in this platform to allow liquid metal and slag to flow separately
into the respective ladles, moving on bogies on rails. A skimmer arrange-
ment separates slag from the metal and diverts it to slag side. An over-
head crane is provided on this platform for maintenance. This is called
the cast-house.
Broadly, the slag and the metal are separated and ultimately these flow
in two different channels, to be collected in two different ladles. Slag was
once sent to slag dumps as a waste product. Steel plants accumulated this
slag from ironmaking as well as steelmaking and created piles of such waste
materials. It also occupied precious plant land. Persistent efforts have led
to develop the use of BF slag as a raw material in cement plants. Cement
is nothing but essentially complex calcium and aluminum silicates. BF slag
is mainly composed of these chemicals, and it is being safely added as the
raw material for cement. BF slag is granulated by pouring it in stirred water.
Now, almost all BF slag is used in this manner.
The hot metal may be disposed of as:
1. Cast in pig beds;
2. Cast in pig casting machine;
3. Sent to steelmaking shop in hot metal transfer ladles, directly or via
a mixer.
Pig iron in its classical sense is no longer produced, and sand bed casting
in the cast house is only of historical importance. The pig casting machine
may be located near the blast furnace or little away from it. It may form a
part of the BF plant or a separate section. Most of the pig iron these days is
converted into steel in an integrated steel plant. The molten metal from the
BF is sent to the steelmaking shop in hot metal transfer ladles. The foundry
grade iron, if produced, is sent to the pig casting machine for producing
pigs and sold for cast iron founding.
Hot Metal Ladles

Two types of ladles are generally employed for transferring hot metal
from the blast furnace to the steelmaking shop, directly or via the inac-
tive mixer, acting as a buffer. Out of these two, the open-top ladle was
the first to be used. It suffers from more heat losses to the atmosphere
and as a result the temperature of hot metal falls. This is a net loss of
thermal energy and can be critical in steelmaking. These ladles are still
fairly widely in use. Lately, the hot metal is externally treated commonly
to remove S or Si and may be even P. The open top ladle is far better
suited for such operation.The second type is known as torpedo ladle. This
was specifically designed to conserve the temperature of the hot metal
during its transit from the blast furnace to the steelmaking shop. The
shape is like a torpedo wherein the surface area of hot metal exposed to
atmosphere is relatively very small. The open top ladle is a inverted frus-
tum of a cone, with wide open top; whereas the torpedo is more horizon-
tal with a small opening in the center at the top. This is the reason why it
does not lose that much heat to the surrounding area. Their shapes are
shown in Figure 2.3.
FIGURE 2.3 Shapes of hot metal transfer ladle. The open top ladle is shown in (a) and the torpedo
ladle is shown in (b).
2.7 Blast Furnace Operation and Control

A blast furnace once started is expected to run for no less than five years
before it is due for complete relining and major repairs. During the cam-
paign charge, materials are regularly put in the furnace from the top and
the liquid metal and slag are tapped from the bottom periodically.
The charge materials are drawn out from the storage bins in required
amounts and in desired sequence and are weighed as accurately as is feasi-
ble without impairing the rate of feed. The weighed materials are put in the
charging skips which raises them to the furnace top and discharges them
through the distribution chute. The ore, coke, and limestone are never
premixed. These are charged separately, one skipful at a time. Depending
upon the requirements, sometimes many skips of the required material are
charged.
The blast furnace operation is controlled from the top by trying to dis-
tribute the solid charge more uniformly in the furnace. It is simultaneously
controlled from the bottom by the temperature and driving rate (rate of
blast feed) of the blast. In order to exercise this control, a series of instru-
ments are provided at various levels in the blast furnace complex to mea-
sure and control various parameters. Computers are used in order to do this
job more efficiently.
2.8 Essentials of the Blast Furnace Process

In the blast furnace, solid charge materials like iron ore (or agglomerated
ore), coke and limestone are charged in the vertical shaft of the furnace at
the top. Hot air blast is blown through tuyeres located at the bottom level.
The oxygen from the preheated blast combines with the carbon of coke
and generates heat and carbon monoxide. The gas phase containing mainly
nitrogen and carbon monoxide ascends upwards through the charge which
reacts with and extracts heat from the gas phase. Eventually the charge
melts and the metal and slag produced get stratified and are separated to
obtain the metal. Fundamentally, the blast furnace is a counter-current
reactor.
The main chemical reactions are the reduction of iron oxide to iron and
the burning of coke. The iron oxide may get reduced as:
2 〈 Fe2O3 〉 + 3 {CO} = 3 {CO2} + 4 〈 Fe 〉 (2.1)
〈 Fe2O3 〉 + 3 {C} = 3 {CO} + 2 〈 Fe 〉 (2.2)
The actual mechanism shall be discussed later. Since the iron oxide is
reduced directly by carbon as in equation (2.2), it has traditionally been
referred to as direct reduction of iron oxide. The CO generated by burning
of the fuel at the tuyeres or by direct reduction of iron oxide as in Eq. (2.2)
reduces iron oxide. This also occurs in equation (2.1) and is referred to as
indirect reduction. The CO required for indirect reduction is essentially
generated at the tuyere level where carbon of coke burns as
2 〈 C 〉 + {O2} = 2 {CO} (2.3)
As the gases rise and carry out the reduction the CO2 content, the gas
increases and a reaction of the type
{CO2} + 〈 C 〉 = 2 {CO} (2.4)
may take place thereby regenerating the reducing gas. This is known as a
solution loss reaction.
Limestone and dolomite, if added as a flux in the charge get calcined as:
〈 CaCO2 〉 → 〈 CaO 〉 + {CO2} (2.5)
〈 MgCO3 〉 → 〈 MgO 〉 + {CO2} (2.6)
in the temperature zone of 800–1000 C. The resulting basic oxides combine
with the gangue oxides to form the slag.
Reduction of the oxides of iron are practically complete while it is in
solid state. This reduced iron is impure because of associated gangue con-
stituents of the ore. Separation of iron from the associated gangue can only
be brought about by melting the whole charge. As the charge descends
in the furnace, its temperature is increased and finally it is melted in the
bosh region. Two liquid phases form as a result of melting, one is an iron
phase and the other are molten oxides called slag. Those oxides which are
not reduced, in spite of melting, join the slag phase and those which are
reduced to elemental state join the metal phase. The slag thus contains
mainly silica, alumina, alkali-alkaline earth oxides, etc. with some minor
other oxides and sulfides. A part of the SiO2, MnO, P2O5, etc. get reduced
and the metal is contaminated by Si, P, Mn, S, etc. The metal also gets satu-
rated with carbon because of its direct contact with coke inside the furnace.
Therefore, the iron melt may contain up to 8 wt% of impurities like C, Si,
Mn, P, S, etc. as the metallic product. The slag may contain mainly SiO2,
Al2O3, CaO and other minor oxides and sulfides. Stratification of metal and
slag is best achieved if the slag is thin, that is of low viscosity. Addition of
flux in the charge helps, to a large extent, in achieving the correct nature of
the slag.
The gas composition and the temperature vary over the vertical as well
as the horizontal cross section of the furnace. The chemical processes are
therefore continuous from bottom to top and throughout the whole vol-
ume of the furnace. The furnace, however, can conveniently be divided
into four main zones with respect to the physical processes occurring in
the furnace.
Stack or Shaft
This extends from the stockline down to the mantle level and into the area
where the burden is completely solid. The charge gets heated from 200 C
at the stockline level to nearly 1100–1200 C by the time it descends to the

bottom of the stack. In order to ensure free fall of the charge material,
since it expands progressively with progressive rise in temperature, the
cross section of the furnace is uniformly increased to almost double the
size from stockline to the mantle level. This furnace batter is generally one
in twelve but the actual value essentially depends on the thermo-physical
properties of the charge. Much of the iron oxide reduction occurs in the
stack. The success of the blast furnace process therefore depends on the
efficiency with which the countercurrent gas-solid interaction is brought
about in the stack. It is this requirement that has led to the developments
in burden preparation like sintering, pelletization, ore preparation, etc.
The agglomeration plant may therefore form an auxiliary to the blast fur-
nace plant.
Bosh
The charge materials begin to soften and fuse as they come down into the
bottom of the stack. The next lower zone is called bosh in which melt-
ing of the burden, except that of coke, takes place. The gangue and flux
combine to form the slag. The furnace walls in this region are parallel to
some extent and then taper down, or they taper entirely down to reduce
the sectional area by about 20–25% to accommodate the resultant decrease
in the apparent volume of the charge. The burden permeability in this
region is mainly maintained because of the presence of solid coke. This
dictates that only coke of adequate strength and size should be used to
achieve this end for efficient operation. Any degradation of coke, leading
to decreased permeability in the bosh region is bound to affect the blast
furnace operation adversely.
Tuyere or Combustion Zone

By the time the charge descends down into the area near the tuyeres,
except the column of coke, the entire charge is molten. The oxygen of
the blast burns coke to CO, and a number of combustion zones, one
in front of each tuyere, exist in the tuyeres zone. A runway or race-
way exists in front of each tuyere. The flame runs through these areas
and the gas-flow smoothly expands over the entire cross-section of the
furnace. The raceway is first horizontal and then smoothly changes its
direction to vertical while expanding over the entire cross-section of
the furnace.
FIGURE 2.4 Structure of tuyere zone and flame area.
This is shown in Figure 2.4 which illustrates the structure of tuyere

and the flame. The temperature in this zone is often referred to as raceway
adiabatic flame temperature (RAFT for short). This name is based on the
near adiabatic conditions that prevail in front of the tuyere.
Hearth
Although most of the coke burns at the tuyere level, a certain fraction
is believed to descend into the hearth where it dissolves in the metal to
near saturation. The entire charge is molten and tends to stratify into
slag and metal layers in the hearth; these are tapped separately. The
cross-section of the furnace below the tuyere decreases since permeabil-
ity is no longer required, nor can it be maintained in the liquid charge.
The walls of the well are parallel and smallest cross-section of the fur-
nace exists in the hearth.
2.9 Efficiency of Blast Furnace

The performance of a blast furnace is assessed essentially in terms of its
productivity and coke rate. The higher is the productivity and lower is the
coke rate, the better the furnace performance. The productivity has been
described quantitatively in various ways as follows:
Basis or related to Expressed as Best world record Where commonly

used
1. Hearth area Tons per unit area of the hearth 60 t/m2/day USA
per unit time
2. Furnace volume Working volume required to 0.4 m3/t/day USSR

produce one ton of iron per
unit time
Number of tons of pig iron pro- 2.5 t/m3/day Japan, India and
duced per unit working volume many others
of the furnace
3. Hearth diameter Blast furnace output index UK

P(B + 10)
B.O.I. = –160
72(D – 11)
where P = output t/day
B = burden wt, kg/t
D = hearth dia, m
The productivity of a blast furnace mainly depends upon the rate of

blast that can be pushed through the furnace. This is called the driving rate
of a blast furnace. For higher productivity, the furnace is driven hard. The
higher the rate of the blast, the more the oxygen is pushed in for combus-
tion of coke. Thus, the rate of coke burning is higher and consequently
more heat is generated. This will result in a higher rate of melting of the
charge as well as higher rate of stack reduction. Harder driving is possible
only if the charge is permeable inside the furnace. Permeability can be
improved by preparing the charge burden, i.e. making it more suitable, and
obtaining better distribution inside the furnace so that the overall burden
permeability through the stack is better.
The furnace is driven hard, that is maximum volume of blast is blown
in the furnace. The charge quality that maintains higher permeability in the
furnace is therefore considered better for higher furnace productivity.
The blast furnace can be run more efficiently by charging lumpy

charge of close size range and the lumps should be such that they main-
tain maximum possible bulk permeability during their descent in the
blast furnace.
Presently the average productivity of blast furnaces is in the range of
1.0 t/m3/day in old style blast furnaces to 1.5 t/m3/day in modified furnaces
to 2.5 t/m3/day in recently installed modern furnaces like the G furnace of
Tata Steel in India.
There are obvious limitations in improving the old furnaces, but
modern furnaces with adequate burden preparation and operational
improvements have comparable performance with those of the best in
the world.
Similarly, the coke rate of old furnaces is in the range of 700–800 kg/
thm but that of the modern furnaces is around 500 kg/thm or even less
which is comparable with the better furnaces in the world.
On the whole, performance of a BF is not judged only on the basis of its
productivity, but on many other parameters like fuel rate (energy efficiency),
campaign life (i.e. years of continuous operation before major repairs), major
maintenance cost, silicon content of the hot metal, etc. The most advanced
modern furnace designs and performance data are as follows:
Parameters Approx. Dimension
Working height 25 m
Inner hearth dia., 15 m
Working volume 5000 m3
Production 12000–13000 thm/d
Productivity up to 2.6 thm/d/m3 of volume
Fuel rate (coke) 350–450 kg/thm (including PCI)
Campaign life 10 years
Si content of HM 0.5% but as low as 0.2%
It is generally believed that near-peak performance has been achieved

and no dramatic improvements are possible in the near future.

The modern BF plant, besides the blast furnace complex, may have coke
oven batteries, an iron ore bedding facility, an agglomeration plant (more
often a sinter plant), a refractory relining facility, a pig casting machine, etc.
It should be kept in mind that ironmaking is the primary stage in inte-
grated steel production. It accounts for the bulk of energy consumed and
therefore governs the overall economics of steel production beginning with
iron ores as the raw starting material. The modern ironmaking furnace
thus continues to improve upon its efficiency of iron production by way of
improved mechanical designs and metallurgical operations.
CHAPTER
3
RAW MATERIALS FOR
IRONMAKING
In This Chapter
● Introduction
● Iron Ores
● Metallurgical Coal
● Limestone and Dolomite as Flux
3.1 Introduction
Iron ore and coke, i.e. coking or metallurgical coal, are the two most impor-
tant raw materials required for iron production. In the alternative methods
of iron production, a cheaper substitute of coke is used as a fuel depending
upon its availability. These will be discussed separately in a later chapter.
The present discussion pertains to blast furnace smelting only. The smelt-
ing operation generally requires external flux such as lime to render the
gangue associated with the ore and coke ash. The flux is generally siliceous
in nature, fusible at a relatively low temperatures so as to separate readily
from the molten iron. Limestone is also required as the next important
raw material but in a much smaller proportion. It is technically essential
and otherwise economical to use dolomite along with limestone as a flux.
Earlier, manganese ore or some other source of Mn was considered as an
essential component of BF charge, but it is no longer considered essential.
Manganese ore used to be added to control sulfur in hot metal.
The raw materials required for ironmaking are:

1. Iron ore, including sinter, pellets, etc.
2. Coking coal or readymade coke
3. Pulverized coal/tar/natural gas, etc. for tuyere injection
4. Limestone, dolomite etc. as flux
5. Quartzite, dunite, etc. to adjust alumina/silica ratio
Air is also an important raw material but it is available anywhere at no cost.
Blast Furnace Smelting Materials (India)

The approximate quantities of charge materials required for production of
one ton of iron under Indian conditions are given below:
Material Quantity in kgs

Iron ore/sinter/pellets (+60% Fe) 2000–2500
(the sinter contains lime)
Coke 450–550
Pulverized coal (for injection) 50–200

Limestone 300–400
Air 4000–5000
The total charge weight per ton of iron produced has continuously
decreased from around 5 tons to 3.5 tons over the last few decades. Of the
total iron units charged in the blast furnace, iron recovery in the form of
hot metal is close to 99% or more, disregarding the dust losses which are
minimal. The actual weight of iron ore to be charged to produce one ton of
hot metal can be readily predicted beforehand.
Nearly 3.5–4.0 t of solid charge materials are required to make one ton
of pig iron. This primarily depends on the iron ore quality, proportion of
ore and sinter, quality of limestone, etc. This is also shown in Figure 3.1,
where only iron ore lumps are put in the charge. As compared to this fig-
ure, the average solid charge consumption in many of the other countries
is closer to 2.5 tons. The reasons for this are many, but they are essentially
related to the quality of raw materials. It should also be noted that for a one-
million-ton steel plant, assuming that iron yield in steelmaking is 90%, and
Raw Materials for Ironmaking • 37
assuming nearly 70–80% of charge is hot metal, nearly 1.8–2.2 Mt of raw

materials are used annually to run the plant. Such a charge would contain
nearly 150–200 kg of gangue per thm to be fluxed by limestone to produce
slag which will be around 300–500 kg per thm (ton of hot metal).
FIGURE 3.1 Simplified material flow in and out of the blast furnace.
In 2007, the world production of crude steel was around 1300 Mt, out
of which nearly 20–25% came from recycling of steel scrap. The remainder
was created by first producing either molten pig iron or DRI followed by
its conversion to steel. If the average iron content of iron ores is assumed to
be nearly 60%, then 1500–1600 Mt of iron ore will be smelted annually to
achieve the present level of steel production. All steel producing countries
do not possess the required iron ore from domestic sources. The produc-
tion and utilization, including national and international marketing of iron
ores, therefore, forms a very important economic activity all over the world.
For the smelting of iron ore at this level, an equally large amount of
fuel in the form of metallurgical coal, ordinary coal, other hydrocarbon gas-
eous fuels, etc. will be required. Assuming a coke rate of 500 kg/thm and
taking into account the efficiency of conversion of coal into coke, nearly
600–700 Mt of metallurgical coal would be required annually. Being a rare
commodity, metallurgical coals are quite costlier than thermal coals. Their
value goes up with decreasing ash content. The current prices (2017) of
good quality iron ores with +62% Fe is around $64 USD and Rs.4200/t
at site and in 2016 good quality coke with 10–12% ash brought $224/mt.
Countries with abundant sources of iron ore and or good quality coke can
export these and earn valuable foreign exchange currency.
3.2 Iron Ores

The naturally occurring materials containing iron are known as minerals
of iron. The mineral deposits from which iron can be extracted economi-
cally are referred to as ores of iron. Iron constitutes about 4.6% of the
earth’s crust and it is one of the most widely distributed and abundant
elements in nature. Being a fairly basic metal ion, it is not generally found
in native state except in the masses of meteoric origin. Then it is usually
associated with nickel and cobalt. Native iron, though rarely, is found in
eruptive rocks. The red coloration of soil is due to the presence of iron.
The remains of most plants and animals contain iron. The principal iron
minerals are:
Oxides.......... Hematite and magnetite
Hydroxides.......... Limonite, goethite, etc.
Carbonates.......... Siderites, spathose iron, etc.
Sulfides.......... Pyrite and pyrrhotite
Complex.......... Ilmenite, silicate, etc.
The mineralogical characteristics of the important iron minerals are

shown in Table 3.1.
Hematite. This is the economic iron bearing mineral which has prob-
ably the widest distribution in the world. It is weakly magnetic but there
exists a variety known as maghenite or Fe2O3 which is magnetic (like mag-
netite) and is found in small amounts in many ore bodies.
Magnetite. This is the second most widely spread economic iron bear-
ing mineral in the world. It is rare that a hematite body does not contain
a few percent of magnetite and vice versa. As the name suggests this min-
eral is highly magnetic and as such occupies a particular place in iron ore
treatment.
Goethite and Limonite. These generally occur as a cap on the origi-

nal iron-bearing mineral body because of weathering and form flakes and
needles occurring as intergrowth with the original constituent of the ore.
These are converted to oxide by calcining. These are important economic
minerals in some places.
Siderite. The siderite deposits occur as large masses, lenses, and veins in
many parts of the world and have been worked for many years. Some coun-
tries still find this as a good source of iron. It produces oxide ore on calcining.
Pyrite and Pyrrhotite. These are mainly a source of sulfur and the
left over iron oxide, after sulfur recovery, is a by product. In spite of the
fact that the residual iron oxide is quite impure, considerable tonnages of
these residues continue to be used as a cheap source of iron. These along
with other sulfides are generally present in most iron ore bodies and which
decrease the natural value of the ore.
Ilmenite. The same remarks as for pyrites apply here. Ilmenite should
truly by classified as a titanium ore which is frequently associated with vana-
dium. The iron residue that is recovered is often as a byproduct. Ilmenite
is isomorphous with hematite and is generally present in small quantities in
most hematite ore bodies whose value is hereby reduced.
Silicates. These are present in most ore bodies in varying proportions.
In high grade hematite and magnetite type of deposits their proportion is
fairly low but they are one of the major constituent of the brown hema-
tite deposits. The presence of silicate gives rise to the oolitic structure.
The oolite is a spherical nodule of 2–3 mm but more commonly less than
0.5 mm in diameter consisting of concentric and radiating crystals. Iron
may be present in these crystals or in the matrix. Silicates of iron, if present,
do pose problems during beneficiation of the deposit.
Although a large number of iron minerals have been found, only a few
of them are important for ironmaking. The various iron minerals mentioned
above do not occur in nature in pure form. Iron ore consists essentially of
a mixture, in varying proportions, of iron bearing minerals and other sundry
minerals called waste or gangue which contains essentially silica, alumina, lime,
and magnesia. These can be present in substantial proportions. Iron minerals
also contain other elements, either as constituents of iron bearing minerals
themselves or as separate entities. These are generally present in quantities
ranging from the fraction of one percent to a few percent. Many of these,
when present in quantities greater than specified limits, which are generally
low, have a deleterious effect on the quality of metal produced. They must
be eliminated during the preparation of ore, in the blast furnace, or during
subsequent steelmaking processes. They are known as impurities. For an iron
mineral to become iron ore, the cost of its mining transportation and smelting
must be such that the cost of the final metal remains competitive with the cost
of metal derived from other ores. In other words, the iron mineral deposit is
called iron ore only if it is possible to produce iron from it economically.
Many times the impurities in ores have a value of their own. If they exist
in sufficient amount to justify their recovery, the economy of the process is
favorably altered due to the recovery of such by-products.
If the processing can be justified from the point of view of recovery
of such impurities, because of their adequate proportion, the ore can no
longer by itself be called iron ore. It may be referred to as an ore of that
element and iron may be recovered as a by-product.
Classification of Iron Ores

Iron ores can be classified simply in terms of their appearance as:
Black ore .......... Magnetite
Red or reddish black ore .......... Hematite
Brown ore .......... Limonite
Grey or whitish ore .......... Carbonate
Iron ores may also be classified according to the quality of gangue asso-
ciated with them as:
Nature o f ore .......... Characteristics o f associated gangue
Siliceous .......... Predominantly silica
Aluminous .......... Largely alumina
Argillaceous .......... Chiefly clayey matter
Calcareous .......... Predominantly lime
Bituminous .......... Large amount of bituminous or coaly matter
Titaniferous .......... Large amount of ilmenite (FeO.TiO2)
The presence of elements like phosphorus, sulfur, arsenic, zinc, lead,

nickel, vanadium, copper, chromium, titanium, etc. in the ore has a con-
siderable bearing on the operation of ironmaking process and the quality
of the product. These are termed as impurities. If the ore is associated
with a substantial amount of these impurities, the ores are often referred
to as phosphoric (containing phosphorus), sulfurous (containing sulfur),
arsenical (containing arsenic), titaniferous (containing titanium) etc. An
ore may be both phosphoric and sulfurous if it is associated with phospho-

rus and sulfur.
Iron ores may also be conveniently classified according to their mode of
geological origin as follows:
Sedimentary. This is the largest and most important group of ores
occurring broadly in four different types. In one the iron oxide and silica
are in alternate bands and are known by different names in different parts
of the world viz. banded hematite quartzite (B.H.Q.) in India, Calico-rock
in South-Africa, itabirite in Brazil and taconite or jaspillite in the USA. It is
porous, fragile and less pure but where secondary deposition of iron oxide
has taken place it is dense, massive and more pure. To the second type
belongs the Minnette ores of Western Europe. sedimentary ores of UK,
the Alabama ores and the Wabana ores of New Foundland. These are pre-
cipitated ores from shallow water around granules, often of foreign matter.
The third type is the ironstone of U.K. The fourth, mainly found in India
and Brazil, is in the form of powder (purity often more than 60% Fe) and
is known as blue dust.
Igneous. It is derived from igneous material of volcanic origin and
occurs mainly as magnetite. The Kiruna ores of Sweden is its typical exam-
ple. The magnitnaya ores are only modified forms of the Kiruna type ores.
Lateritic. It is formed in tropical conditions of alternating dry and wet
seasons resulting in leaching away of silica and alkalis from the mother igne-
ous rock, leaving behind mainly iron oxide and alumina.
Replacements. The circulating water earlier removed limestone
depositing iron carbonate in its place. Subsequent surface weathering has
oxidized part of the iron carbonate to oxide or hydrated oxide.
Structure and Texture of Iron Ores

The constituent minerals in the iron ores are not necessarily distributed at
random throughout the ore body. They frequently follow a pattern which
has been determined by the geological, metamorphic and geochemical
influences which the ore body has undergone. This pattern can have a bear-
ing on the course which ore preparation should follow and on the subse-
quent behavior of the ore in the blast furnace. The following principal types
can be distinguished as:
Oolitic ores. Concentric layers or as radiating crystals and iron may be
present in these or in the matrix rock.
Lateritic ores. These are formed by surface action at the expense of

underlying rock rich in iron and consist of intermingled mass of iron hydrox-
ides, alumina hydroxides and other minerals. Their beneficiation raises par-
ticular problems which in many cases have eluded a economic solution.
Banded ores. These are characterized by alternate bands usually a few
mm thick but sometimes up to several cm. Some bands may have predomi-
nance of iron minerals while the other may have a predominance of quartz
and silicates. The marked success achieved in the beneficiation of these
deposits has given an impression that all banded ores are easy to benefici-
ate. In reality they present a great diversity of types as regards the thickness
of their banding, grain size, degree of liberation, constituent minerals, etc.
Each type must be considered strictly on its own merits.
Hard massive ores. The constituents are inter-grown into a homog-
enous mass showing no preferential direction or pattern except occasional
very slight banding. It is a fine grained material and offers high resistance
to crushing and abrasion. These are difficult to reduce in the blast furnace.
Soft ores. These can be massive but the crystals are more loosely knit
together to a banded structure. These are more porous and are liable to
produce more fines during handling. Better porosity however can be uti-
lized in a furnace if other properties are tolerable.
Powdery Ores. These can not be charged as such in the furnace. They
must be agglomerated before charging. Being powdery in nature, concen-
tration operation can be readily effected before agglomeration.
Magnetite ores. Magnetite ores can be divided, according to their
origin, into three main categories:. sedimentary-metamorphic, contact-
metasomatic and magmatic.
In sedimentary-metamorphic ores the ore minerals are usually fine-grained
and texture may be banded. Ores of this type generally contain martite,
hematite, iron carbonates, and silicates. The gangue is mostly quartz and
sulfur, and phosphorus contents are generally low. The examples of this type
of ores are the deposits of Lake-Superior in the USA and Krivoi-Rog ore
fields in the USSR. In contact-metasomatic ores the magnetite generally
occurs in grains of different sizes and the amount of sulphides of iron and
other metals like copper and zinc can be high. And considerable amounts of
CaO are present with siliceous gangue. The deposits in eastern parts of the
USA and Western Siberia belong to this category. Ores of magnetic origin
differ widely in grain size and composition. The phosphorus content is also
generally high. Magnetites are liable to oxidation and these ores contain
areas of mixtures of martite and hydroxides with magnetites.
Hematite ores. These are most widely distributed in the world and
from which a major portion of iron is produced. These are usually of the
sedimentary-metamorphic type. In view of the differences in conditions of
formation of the metamorphic processes, the granulometry, the structure
and the iron content of these ores very considerably. They include the hard,
homogeneous, fine-grained hematite-taconites from the Lake Superior
region of U.S.A., hard quartz banded ores of Sweden and Labrador, coarse
grained ores which are easy to crush and grind, and high grade ores (both
soft and hard) containing large inclusions of pure hematite from the Mesabi
range of U.S.A. and the Krivoi-Rog ore fields in the U.S.S.R. The sulfur
and the phosphorus contents of hematite ores are usually low. The gangue
mainly consists of silica but alumina can also be appreciable.
Brown hematite ores. The term brown hematite ores is used to cat-
egorize a variety of ores mostly sedimentary in origin where the material
moved in solution from its initial point of formation. The main ore minerals
in the brown hematite category are the various iron hydroxides, limonites,
hydrogoethites, silicates, and siderites. The gangue is generally siliceous but
occasionally lime is present along with a high silica content. In the acidic
oolitic ores, more than 30% of the total iron may be present as iron silicates
of diverse composition. The sulfur content is usually low but high phos-
phorus contents are frequent and can exceed 1%. It is not unusual to find a
good amount of manganese as well. The Kershenski ores are characterized
by the presence of arsenic as an impurity. Although these ores are common
all over the world, they are principally worked in France. Luxemburg, U.K.,
Germany, etc. where these account for bulk of the reserves.
Siderites (Spathicores). These are a carbonate type of ores and con-
stitute a very small proportion of the total world availability, but in countries
like Canada, Austria and Germany these are or have been of considerable
importance. Siderites are prone to oxidation and the ore often contains
considerable amounts of hydroxides. Near the surface these ore bodies are
covered by a siderite-limonite-hematite cap.
Titanomagnetites. The most important of the complex ores are the
titanomagnetites. The main ore mineral is magnetite with titanium present
as ilmenite or as titanomagnetites. Some of these ores contain vanadium
which is recovered from these ores along with iron as a by-product. The
composition varies considerably and may contain 60–70% magnetite and
20–27% ilmenite.
Laterites. These are produced due to surface alterations under tropi-
cal conditions. The final result is the formation of an iron cap on hard ore
extending a few meters below the surface. Such ores can contain 50–60%
iron in the form of hydroxides and a high proportion of alumina which has
not been leached out. These ores many times contain impurities like chro-
mium, nickel, cobalt, etc., in such high proportions that it is worked for their
recovery. Laterites are difficult to process and still largely remain unused.
U.S. Production and Use

In 2015, mines in Michigan and Minnesota shipped 98% of the usable iron
ore products in the United States — the remaining 2% of domestic iron
ore was produced for non-steel end uses — with an estimated value of
$3.8 billion. Twelve iron ore mines (nine open pits and three reclamation
operations) and three iron metallic plants, including direct-reduced iron
(DRI) and iron nugget producers, operated during the year to supply steel-
making raw materials. Each open pit mine site included a concentration
plant and pelletizing plant. A stand-alone pelletizing plant in Indiana used
iron ore fines from reclamation plants in Minnesota. The United States
was estimated to have produced and consumed 2.5% of the world’s iron
ore output.
Events, Trends, and Issues

U.S. iron ore production decreased in 2015 owing to an approximate 30%
increase in steel imports in 2014 that persisted through 2015. This increase
in low-cost steel imports led to reduced demand for domestic steel, reduc-
ing iron ore consumption from domestic iron ore mines. Major declines
in seaborne iron ore prices in the current and preceding years coincided
with the rising rate of low-cost steel imports in the United States. During
the year, six iron ore mines in the United States had either been idled,
reduced production, or closed permanently. As of November, three opera-
tions remained idled indefinitely and an open pit mine and its associated
iron nugget facility were to be idled through the first half of 2017. A DRI
facility in Louisiana increased its production rate early in 2015 after equip-
ment failures in 2014. Construction resumed on a 7-million-ton-per-year
iron ore operation in Minnesota; production was scheduled to begin in the
second half of 2016. Construction continued on a 2-million-ton-per-year

hot-briquetted iron facility in Texas; production was scheduled to begin in
the first quarter of 2016. One iron ore operation was scheduled to be idled
at the end of 2016. One company planned to use the current idling of an
iron ore facility to reconfigure equipment to produce an alternate flux pellet
to meet contract requirements.
World Deposits of Iron Ores

The major iron ore producing countries in the world are indicated in
Table 3.2. These are chiefly China, Australia, Brazil, Canada, Iran, India,
Kazakhstan, South Korea, Mexico, Russia, South Africa, Sweden, United
Kingdom, Ukraine, USA, and Venezuela. Out of an estimated 40 billion t
of deposits in Australia, nearly 95% production comes from the Hamersley
basin of Western Australia. The ore is rich hematite with 64–68% Fe and
0.02 to 0.14% P. It produced 152 Mt iron ore in 1996 and raised it to 262 Mt
in 2005. It is the major iron ore exporter next to China. The major iron ore
deposits of Brazil are located in states like Minas Gerais, Para Mato Grosso,
and Sao Paulo. The important deposits are located north-west of Rio de
Janeiro. Brazil produced 263 Mt in 2004, 280 Mt in 2005, and 428 in 2015.
It is one of the chief exporters of iron ore in the world. These deposits are
quite rich in iron contents. In Canada the iron ore is mainly found in the
Quebec-Labrador region, Ontario, and British Columbia. It is also an iron
ore exporter country. China is the largest consumer of iron ore in the world
since it produces nearly 30% of the world steel production. Hence, it is
the major importer of iron ore from various countries. The main deposits
of iron ore in China are similar to the Lake-Superior deposits in the USA.
The ore deposits of Belgium, Germany, and France are brown hematites
and carbonates with high phosphorus contents. This is the reason why they
produce Thomas grade iron containing more than 1–5% P. The Minnette
iron ore of France contain only 35% iron but occurs both as calcareous as
well as siliceous grades. This is the reason why these are properly blended
to produce a self-fluxing burden and used even at such low iron contents.
In Iran there are three main deposits, namely, Choghart, Chadormalou
and Sangan which are currently mined to feed their own steel plants which
produced 9.4 Mt of crude steel in 2005. Iran exports a good amount of iron
ore. South Korea is the fifth largest steel producer. It consumed nearly
75 Mt of ore in 2005, and only 20 Mt was produced locally. The impor-
tant iron ore deposits in Russia are found in the Saksagam iron ore fields.
Russia produced 95 Mt of iron ore in 2005 out of which some 13 Mt was
exported. These are very high quality ores. Russia is one of the top four
steel producers in the world. In South Africa, rich irons are available in
the form of banded iron ore formations. The ore is hard massive hematite
with 66–68% iron. Large reserves of oolitic ores also occur in the Transvaal
region. It produced 40 Mt in 2005 out of which 25 Mt was exported. The
most important iron ore deposits in Sweden are found around Kiruna and
Malmberget. These are the largest underground iron mines in the world.
Current production from these mines is 22 Mt of ore annually. The famous
Bog iron ore deposit is rich magnetite containing negligibly small amount
of phosphorus. The iron that is produced from these was once known as
Swedish Iron or Hematite Iron. In the United Kingdom, no substantial iron
ore is mined. UK depends entirely on imported iron ore for running their
steel plants. It has been importing around 15–17 Mt of ore annually in the
last few years to produce 13–14 Mt of steel. Ukraine has the largest iron
ore deposits in the world. These are mainly located at Krivoi Rog basin in
Denpropetrosk region. These deposits occur as lenticular masses of hema-
tite with 63% iron. With an annual production of 39 Mt of crude steel in
2005, Ukraine is considered one of the major steel producing country in
the world. The iron ore deposits of the USA consist mainly of intermixed
limonite and clay with small quantities of fine-grained quartz and musco-
vite. The important deposits come from the Lake-Superior region along
with regions like Alabama, Kentucky, Virginia, and Pennsylvania. The USA
produced nearly 55 Mt of iron in 2005, and it produced only 55 Mt of ore.
It relies more on recycled scrap for its steel production. The extremely
fine taconite deposit of the Mesabi-Range once dominated the design and
practice of BF in USA.
Because of the high level of world iron production, it has become
imperative to look to iron ore reserves from a nationalistic point of view. As
a result, every country thinks about iron ore production from a future per-
spective. This is why countries like China, USA, UK, and others, although
possessing large reserves of iron ore, import iron ores and keep their own
ores as reserves for future consumption. This is also why the prices of iron
ores and metallurgical coal or coke have been continuously and drastically
rising over the last few years when steel production steeply rose.
Japan does not possess any iron ore or coal deposits and yet produces
a large proportion of the world steel production by entirely importing the
required raw materials. The total world reserves of iron ore are estimated
to be nearly 800 billion t of crude ore containing nearly 230 billion t of
iron. China produces nearly a third of the total world steel production. It
does not produce the equivalent iron ore from within China. Consequently,
China is on the look out to buy iron ore from any other country to sustain
its dominance in steel production. It is, therefore, a dominant player in the
international iron ore trade. The production of iron ores per country and
the corresponding reserves are shown in Table 3.2.
World Resources
U.S. resources are estimated to be 110 billion tons of iron ore containing
about 27 billion tons of iron.
(mainly low-grade taconite-type ores from the Lake Superior district
that require beneficiation and agglomeration prior to commercial use).
World resources are estimated to be greater than 800 billion tons of crude
ore containing more than 230 billion tons of iron.
TABLE 3.2 World Mine Production and Reserves:

Mine production Reserves
2014 2015 Crude ore Iron content
United States 56 43 11,500 3,500
Australia 774 824 54,000 24,000
Brazil 411 428 23,000 12,000
Canada 44 39 6,300 2,300
China 1,510 1,380 23,000 7,200
India 129 129 8,100 5,200
Iran 33 33 2,700 1,500
Kazakhstan 25 25 2,500 900
Russia 102 112 25,000 14,000
South Africa 81 80 1,000 650
Sweden 37 37 3,500 2,200
Ukraine 68 68 6,500 2,300
Other countries 153 125 18,000 9,500
World total (rounded) 3,420 3,320 190,000 85,000
Iron Ore Grading (India)

In India, the cut-off grade for estimating the hematite ore resources is
more than 55% Fe. A proposition has been made to alter this to 45% Fe
so that the resource position shall automatically improve. This would be
possible in view of modern trends in technology for iron ore beneficiation
and utilization. It should be noted that during 1980–1990, nearly 470 Mt

of iron ore was mined and the resource position was estimated on 1990-
01-04 to be 5223 Mt. In 1990–2000, nearly 656 Mt was mined and during
2000–2005, nearly 532 Mt ore was mined, but the reserves were estimated
on 2005-01-04 to be 3141 Mt more than the previous estimate. This is all
due to additional discoveries of iron ore deposits. There still exists hope to
find new deposits in India.
At an estimated steel production of 110 Mt by the year 2012, iron ore
need would rise to 176 Mt every year. With these demands, the present
reserves can not last for more than 150 years. According to the estimates
made by The Joint Plant Committee in their Bulletin on Iron and Steel,
January 2005, the Indian reserves are expected to last for nearly 200 years.
One of the characteristic features of the Indian iron industry is that some
mines are captive mines of steel plants, whereas the others cater to open
domestic sale and export. Much of the ore requirements of sponge iron
plants come from these non-captive mines.
The mining operations are fairly extensively mechanized in India, and
as a result nearly 60% of the mine production comes in the form of iron
ore fines. Additional fines, about 10%, are generated during subsequent
handling of the ore. As a result, nearly 70–75% of mine output comes in
the form of iron ore fines. There is great necessity to consume these fines
either for export or via agglomeration for ironmaking in India. 40–45 Mt of
fines are lying accumulated at the captive mines of SAIL which could not
be fully utilized. Efforts were made to use them as feed for sinter plants.
The sinter plant uses generally +1 to 10 mm size and hence the 1 mm size
is left unutilized. This needs urgent attention since it is already creating pol-
lution problems. This 1 mm fraction can only be pelletized and then used
in ironmaking. Roughly, the economical size of a pelletization plant is 3 Mt
annual capacity, and it presently costs nearly Rs.500 crores in India, which
proves prohibitive in the present economic scenario.
Sponge iron units require +65% Fe grade with closely sized fractions of
5–18 mm hard lumps. This is rather scarce in India.
The dominant requirement for steel is in the infrastructural and con-
structional sector. If this is buoyant, it leads to increased demand of steel
and for iron ore. The general GDP growth is an indication of the rising
demand for steel and for iron ore as well. The present GDP rise has been
and is expected to remain for the next several years in the range of 8–10%;
this is a good omen for steel industry.
Valuation of Iron Ores

The value of an iron ore deposit depends upon several factors. Simple for-
mulas have been put forward by many workers for the valuation of an iron
ore deposit. The factors affecting the value of an iron ore deposit can be
conveniently summarized as follows:
1. Richness
2. Location
3. Composition of the gangue
4. Treatment and preparation needed before smelting
5. End use
Richness means the percentage of metallic iron in the ore: the richer the
ore, the less amount of ore is required to produce a unit weight of pig iron.
For example, in order to produce a ton of pig iron, about one and half tons
of ore is required in Australia (68% Fe), about two tons are required in
India (55–60% Fe) and nearly three tons are required in the U.K. (30–35%
Fe). The fuel and flux required in all these three cases would be consider-
ably different. For richer ores, since the gangue is less, the fuel and flux
consumption will be less than that for leaner ores wherein gangue is more.
The value of an ore is therefore decided not directly by its iron content
alone but by its iron content and other associated minerals together.
The location of an ore, both geographically and geologically, is a very
important factor in its evaluation. Geological location is related to whether
the deposit is underground, as an outcrop, or is in the form of a hill. When
the ore can be obtained by quarrying or open cast mining with a shallow
overlying burden, the cost of obtaining the ore at the ground level railway
station is much lower than that involved in underground mining, which
requires expensive excavating machinery. Many important iron ore depos-
its, for example, some Indian and Australian deposits, are located as hills
and that allows the ore to be obtained at the ground level railway station
under gravity. The mining operation in this case is also cheap. Geographical
location is related to the mode of transport required to bring the mined
ore to the smelter and includes the capital expenditure on installation of
railways, harbors, ore carries, etc.
The evaluation of an ore needs careful consideration of the size of the
ore reserves, its location, possible difficulties in mining the ore, and the
distance and the mode over which it is to be transported. In opening a

new mine, such capital expenditure is warranted only if the reserves are of
sufficient magnitude to ensure at least ten years operation to amortize the
investment made in developing the mine and the necessary transport sys-
tem. The development of the Goa mines in India and the Hammersley mine
in Western Australia by the Japanese are typical examples of exploitation of
iron ore deposits by a needy customer from a distant place. Similarly, the
magnetite ore deposit of Kundremukh (Karnataka) was developed with the
support of the Iranian Iron and Steel Co. for its beneficiation and agglom-
eration and their contractual export to them for a fixed period.
The composition of gangue associated with an ore may reduce the value
of an otherwise rich ore. In some cases, it may even enhance that of a lean
ore. The effect of associated gangue is related to its nature i.e. whether it
entirely stays with the slag, partially with the slag and remainder with the
metal, or entirely joins the metal phase under the conditions of smelting.
The entire amount of phosphorus in the burden (except that blown as dust)
is reduced and joins the pig iron. The presence of phosphorus affects the
properties of cast iron adversely and complicates the process of steelmak-
ing. In general, the phosphorus content of an ore therefore reduces the
value of that iron ore.
The presence of oxides, like alumina, lime, magnesia, and alkalis stay
with the slag entirely during iron smelting. Alumina raises the melting
point of slag and would require additional fuel to raise the temperature
in order or make the slag a free flowing liquid. Alkalis tend to eat away
the furnace lining and cause difficulties in running the furnace smoothly.
Under no circumstances is the alkali content of the burden allowed to
exceed a safe level, which is extremely low. The presence of lime and
magnesia is, however, an asset since it can be mixed with siliceous ores
to obtain a self-fluxing blend. The total weight of an ore blend in such
a case may in fact be less than that (ore plus external flux) to produce
a unit weight of pig iron. It saves the cost otherwise incurred in using a
flux in the burden. The total charge weight per ton of iron produced may
also decrease in such a case. On the whole, the value of an ore is drasti-
cally reduced by the presence of alkali oxides, reduced to some extent by
the presence of alumina, and is in fact enhanced by the presence of lime
and/or magnesia. Silica, manganese oxide and sulfur in the burden are
partially reduced during smelting so that some of the silicon, manganese,
and sulfur join pig iron. The amount of these in the pig iron depends on
the manner of burdening and operation of the furnace. The presence of

sulfur, in particular, reduces and the presence of manganese may, if at all,
enhance the value of the ore. A reduction of 0.1% sulfur in the ore may
reduce the coke rate by 2% and limestone consumption by 1–2%. Arsenic
needs to be eliminated from the ore by a prior reduction-volatilization
roasting, otherwise it is bound to join pig iron. Zinc in the burden does not
pass into the pig iron but is volatilized during reduction. It thus builds up
in the stack and may cause scaffolding which impairs the smooth operation
of the furnace. It may also penetrate the refractory and destroy it. Lead is
reduced in the furnace, but being immiscible and heavier, it accumulates
at the bottom of the hearth and eventually may penetrate the refractory
joints and destroy them.
The treatment and preparation which the ore needs, before it is charged
in the blast furnace, determines the cost of the ore at the furnace. Dense
ores may need crushing or breaking to render them more porous. Fine ores
will have to be agglomerated and carbonate ores will have to be calcined
to render them fit for charging in a blast furnace. Certain ores need some
while the others need more elaborate ore dressing and/or agglomeration
treatments in order to render them in the proper richness and/or physical
state to be useful as a burden. All this adds to the cost and thereby decreases
the value of the ore deposit.
The end use of iron also decides its value. There are only two broad
ways of using iron ore for iron production: production of iron in molten
state using BF or SR processes and solid state reduction of ore to produce
DRI, without in any way affecting the gangue constituents of the ore. The
gangue constitutions play a major role in smelting operations producing
molten iron. The amount and mutual proportion of the gangue constitu-
ents, to a large extent, decide the value of the ore for such operations. In the
BF route, the nature and proportion of gangue, that is mainly the content
of alumina and silica, decides the flux quality and quantity required for the
smelting operation. The quality of sinter, its reduction-degradation behav-
ior, is improved by lowering the alumina content and may require benefi-
ciation of the ore fines. Every attempt is made to bring the alumina content
down below 2% to finally improve the BF operation. These treatments
decrease the value of the original ore in that proportion. The alumina also
indirectly decides the volume of slag produced in the operation and thereby
influences the productivity as well. It also indicates the fuel requirements
of the smelting operation.
In DRI production the gangue does not take part in the reduction
operation as such, but it is to be removed subsequently during steelmaking
operation by its assimilation in the slag. The gangue decides the nature and
amount of flux required in steelmaking. It also decides the power consump-
tion since more power is consumed for melting when the amount of gangue
is higher. On the whole, gangue quality and quantity affects the subsequent
steelmaking operation in terms of its efficiency and economy.
3.3 Metallurgical Coal

Natural coal is too dense and/or fragile to be used as a fuel in a modern
blast furnace. Neither the blast will be able to penetrate it quickly for burn-
ing, nor is it strong enough to stand nearly 25 m (80 ft) of burden lying
over it inside the furnace. However, certain coals when heated out of con-
tact with air result in a carbonaceous, nearly volatile free matter which is
strong and porous. This mass is called coke, and it is just the right type of
fuel needed to run blast furnaces. The coal which can be so converted into
coke is called metallurgical coal or coking coal, and it is the raw material
required for ironmaking. Although coke is practically universally used as a
fuel, charcoal is still used at a few plants, particularly in Brazil, but these are
exceptional cases.
Coke costs make up nearly 20–30% of the overall cost of iron and steel
production. The quality of coke is therefore at the center of all consider-
ations of an ironmaking operation. If iron ore quality improves by just 1%
of iron content, it reduces the coke consumption by nearly 0.8–1.2% and
improves production by about 1.5–2.5%. The coke quality and ore qual-
ity must be as superior as possible. The quality of coke used for ironmak-
ing has a fundamental effect on the operation and output of the furnace.
Chemically coke can only be as good as the coal from which it has been
made. Hence the coal itself should be uniformly low in ash and low in dele-
terious impurities like sulfur and phosphorus, and high in fixed carbon. The
coke should contain no more than 2% volatile matter, no more than 10%
ash, and no less than 85% fixed carbon. The phosphorus content may vary
from 0.018–0.04% and the sulphur content from 0.6–1.5% (usually around
1%). Coke is the chief source of sulfur in the blast furnace.
A reduction of 0.1% sulfur in coke can reduce coke rate by 2% and
limestone consumption by 1.5–2.0% with a proportionate increase in pro-
ductivity of the furnace.
The physical properties of coke can be greatly improved by modifying

the carbonization process. Those coals which can give coke good physical
properties are known as prime coking coals. The others, in the descending
order, are known as medium coking and low to weakly or blendable coking
coals. Although the world coal reserves are substantial, the metallurgical
coal reserves are relatively scarce and are of high value. Very few coun-
tries in the world are endowed with good quality metallurgical coals and
these are the USA, USSR, UK, France, Belgium, Australia, South Africa,
Canada, etc.
Metallurgical Coal Reserves

The Australian or the Chinese coking coals methods are far superior to
other coking coals methods, as these contain around 9–10% ash as against
24% of the standard coking coal methods. These are regularly imported to
substitute a part of the coking coal requirements in several countries.
Taking into consideration the mining losses (nearly 50–60% of the in-
situ reserves are considered extractable) and further losses on washing of
coal (almost all metallurgical coals require washing to reduce ash content
below the level of 17%), only 40% of the estimated reserves of metallurgi-
cal coals would be available for actual use. For optimum utilization of the
prime coking coal, it is therefore blended with medium and weakly coking
coals to the tune of about 50%. At the envisaged rate of consumption of
coke in iron and steel industries (annual rate of growth of steel production
at 7%), the total metallurgical coal in the country will be exhausted within
the next fifty years. Nearly 75% of the coal fed to the coke ovens is finally
available as coke after carbonization. Another important sector of metal-
lurgical coal is its conversion to coke in merchant cokeries (beehive coke)
which consume 3.5 Mt of coal. It will go up to 5 Mt in the near future. Small
amounts of metallurgical coal is directly fed to the coke ovens for carbon-
ization, since it is of better quality. In years to come, all the coal will have to
be processed. Due to its decreasing quality, the coal will have to be washed
before it is put in the coke ovens.
3.4 Limestone and Dolomite as Flux

A flux is a substance added during smelting to bring down the soften-
ing point of the gangue materials. It reduces the viscosity of the slag and
decreases the activity of some of its components to make them stable or
unstable in the slag phase. The molten oxide product of smelting is known
as slag, which is formed by the combination of gangue and the flux. The
gangue arises in a blast furnace from the ore and the ash of the coke. Flux
itself may contain similar gangue oxides. In general, silica and alumina pre-
dominate as gangue constituents in an iron ore. Coke ash is invariably of the
same character. Limestone and dolomite are therefore required as fluxes.
Some of the continental ores are calcareous. When mixed with siliceous
ores, they can produce a self-fluxing blend and an addition of an external
flux then becomes unnecessary.
The liquid temperatures of an innumerable number of slags have
been experimentally determined and are plotted as ternary and high
order phase diagrams (both real and pseudo) on which these tempera-
tures are superimposed. The reader can refer to standard texts for this
information. The amount and type of flux required in a blast furnace is
determined according to the composition of the burden vis-a-vis the soft-
ening point obtainable from the relevant diagrams. Once the temperature
of the slag is fixed, the oxide equilibrium diagram gives all the possible
composition ranges which will have this as its softening point. The basic
fluxes include limestone dolomite, in some cases rock phosphate, and
basic steelmaking slags.
Lime or magnesia content of a flux gives a fallacious impression about
the quality or value of the flux. The value of a flux is expressed in terms
of available base since it may itself contain oxide impurties similar to the
gangue which it aims to flux away. The available base value depends upon
the basicity (ratio of base to acid content) of the slag under operating condi-
tions. In a normal blast furnace, the basicity (CaO + MgO)/SiO2 is usually in
the range of 1.0–1.3. The available base is therefore equal to
= % (CaO + MgO) – % (SiO2).B
where B is the basicity. This much will be available for fluxing the gangue
of the ore and the ash.
The lower the softening point of the slag, the lower its viscosity at a
given temperature. Hence, the flux is needed in such a way that slag of the
lowest possible softening point is formed. In an iron blast furnace, the slag
is also supposed to retain as much of the sulfur content of the burden as is
possible. A basicity of 1.0–1.3 and adequate temperature of slag can achieve
this efficiently.
Limestone and dolomite are available abundantly all over the world but
the quality is not sufficient in many areas. It should contain minimum phos-
phorus, sulfur, silica, and alumina. The lower these impurities, the higher
its value as a flux. In general, CaCO3 content should not be less than 95%,
SiC2 should be less than 5%, and MgO and Al2O3 each should be below 3%
depending upon the SiC2 content. Nearly 0.5 t of limestone is required for
making a tonof pig iron. If the insoluble content is high, limestone benefi-
ciation may become necessary. The beneficiated product will be fine in size.
This form is not suitable as a charge directly in the furnace, but it is better
as a feed for making fluxed sinter.

The production of iron and steel has been continuously increasing all over the
world. As a result, the production of iron ores forms an important economic
activity. Since many countries rely more on imports of such ore to maintain
their iron production, international trading of iron ores has become very
important. It also results in obtaining valuable foreign exchange for many
countries. The same is true for metallurgical coals. Many prefer to sell coke
than coal. Hence, coke is also an important internationally traded item.
This has led to the study of all the aspects of iron minerals and ores as
they occur in nature and their possible economic uses. It has also led to
the development of more economical blends of iron ores and metallurgical
coals to be used in ironmaking. The importance of limestone and dolomite
as flux, and its place in producing good quality super-fluxed sinters has led
to the detailed study of these fluxes. This study has discovered cheaper
alternative fluxes for iron and steelmaking with the end goal of obtaining
slags of low liquid temperatures.
CHAPTER
4
MAIN BLAST FURNACE
FUEL: COKE
In This Chapter
● Introduction
● Functions of Coke
● Quality Requirements of Coke
● Preparation of Coke (India)
● Modifications in Coke Making Practices
● Alternative Fuels
● Ferro-coke
4.1 Introduction
The developments in blast furnace technology over several centuries can
conveniently be divided into the following specific areas based on the type
of fuel used:
1. The charcoal period;
2. The anthracite period;
3. The rock coal period;
4. The beehive coke period;
5. The by-product coke period.
Being very weak, uneconomical and problematic from ecological view-

point, charcoal could not continue universally as a blast furnace fuel for
long as the production level increased. The naturally occurring coal, in
the form of anthracite, because of its good purity, was used as an alter-
native. Similarly, the abundantly available rock coal also was used as an
alternative fuel to charcoal. Being weak, anthracite and rock coal could not
solve the problem of the increasing size of the blast furnaces. The right
alternative was found in coked coal i.e. coke. Coke is obtained by heating
metallurgical coal out of contact with air; the volatile matter is removed
and carbon particles join each other to form a porous cellular mass with
sufficient strength. It is nearly free of moisture and volatile matter. Coke
has good strength and yet a porous structure which enables it to react with
oxygen readily for burning. Not all coals can be so converted into coke.
Those which can be converted are called metallurgical coals. There are
different ranks of metallurgical coals. Some are very good at coking and
give strong and cellular mass on coking; these are called prime variety
coking coals. There are semi or weakly coking coals; these produce less
strength in resulting coke. Other coals fall in between and have intermedi-
ate strengths and porosity; these are called the medium coking variety. This
method of classification holds true under Indian conditions. The Australian
classification divides the coking coals into only two varieties: hard coking
(HCC) and semi-soft coking coals (SSCC). Their HCC are comparable to
the prime coking variety.
Coke was made in beehive ovens in the early days. The technology
still continues to be in use, but it produces a lower grade of coke meant
for foundries and such applications but not for blast furnaces. The blast
furnace coke for iron production is now only made in by-product coke
ovens. Several such ovens are put together in the form of a battery,
and it forms an inalienable part of steel plant. These ovens are meant
to recover all the by-products of the coking process and thereby affect
greater economy. Tar is one of the important by-products of this pro-
cess. This process also improved the quality of coke produced as com-
pared to the earlier beehive ovens practice; the chief reason was that
coking is carried out at much higher temperature in these by-product
coke ovens.
Coke is the only solid fuel used in a blast furnace for smelting iron ores
efficiently. The quality requirements of this fuel form the subject matter of
the present chapter.
Main Blast Furnace Fuel: Coke • 59
4.2 Functions of Coke

Coke fulfills five major roles in a blast furnace:
1. It is a fuel providing heat to meet the endothermic requirements of
chemical reactions and melting of slag and metal.
2. It produces and regenerates reducing gases for the reduction of iron
oxides.
3. It provides an open permeable bed through which slag and metal pass
down into the hearth and hot reducing gases pass upwards.
4. It supports the overlying burden, more so at the lower levels/regions.
5. It carburizes iron, thereby lowering its melting point and consequently
the working temperature of the BF is reduced.
These five roles place obvious requirements on coke. The carbon content
of coke should be maximized for it to be a good fuel; it should contain a
minimum of ash and other deleterious impurities.
As a regenerator of reducing gases and as a heat producer, it should
have a high reactivity with oxygen, carbon dioxide, and water vapor. There
is, however, no evidence to show that the reactivity of coke imposes any
constraints on its combustion rate at the tuyeres.
Permeability of the charge, particularly in the bosh region, where
everything else except coke is either plastic or molten, is maintained by the
coke alone. The coke should be of a narrow size range and suffer minimal
breakdown in its passage to and through the furnace until it burns at the
tuyere level. The role of coke as a fuel and as a regenerator of reducing gas
can be and has been considerably reduced by the use of a higher tempera-
ture blast and injection of auxiliary fluid fuels in the furnace. This has been
developed to the extent of achieving a coke rate as low as nearly 400–500
kg/t of hot metal in some of the most modern furnaces as compared to the
average value of 700–900 kg/t. It has also been estimated that this value
could not be reduced below 300–350 kg/t without seriously reducing the
permeability of the charge.
The function of coke as a spacer, particularly in the bosh region, can-
not be dispensed with nor can it be substituted by anything else in a blast
furnace. This function of coke is of relatively greater importance than its
heat generating and ore reducing functions. In other words, a modern blast
furnace cannot be run without a certain amount of coke as a charge mate-

rial. The amount could be reduced to a value as low as is feasible without
impairing the permeability of the charge. The quality of coke is related to
the quality of coal, its processing and subsequent carbonization process.
There variables need to be adjusted, to whatever extent feasible, to obtain
the required quality of coke so that the stipulated functions of coke inside
the blast furnace will be optimum.
4.3 Quality Requirements of Coke

The efficiency of a blast furnace is directly dependent on the permeability
of the charge in the furnace, which in turn directly depends on the quality
of coke. It is expected that the quality of coke should be such that it causes
a minimum of operational difficulties and maximum of production rate at
minimum coke consumption. The properties which determine the value of
coke as a blast furnace fuel are:
1. Chemical composition as determined in terms of the contents of fixed
carbon, ash, sulfur, phosphorus, and other deleterious impurities.
2. Reactivity as determined by its physical structure i.e. cellular structure
and porosity.
3. Size range.
4. Thermal stability at high temperatures.
5. Strength and abrasion resistance essentially assessed in terms of CSR.
Chemically, the useful component of coke is its fixed carbon which is the
fuel as well as the reducing agent in a blast furnace. The balance is made
up of ash (the inorganic residue after burning), volatile matter, other
impurities, etc. As the ash content increases, the available carbon of coke
decreases. Consequently, the heat supply to the furnace per unit weight
of coke used decreases. The amount and quality of ash in the coke has a
considerable bearing on the blast furnace operation. Ash usually contains
refractory oxides like SiO2, Al2O3, CaO, etc. Phosphorus and sulfur are usu-
ally present in ash in the form of inorganic compounds. The effects of such
refractory gangue constituents of iron ore have already been discussed in
an earlier chapter, and the same arguments hold true for the ash as well.
An increase in ash content of coke by 1% over a critical limit results in a
decrease in production by 3–6% and involves an increase in coke consump-

tion by about 4–5% in a normal blast furnace operation. The sulfur present
in coke requires additional flux for its removal or the sulfur content of the
iron goes up. All the phosphorus in coke and in the entire burden joins pig
iron. The total amount of sulfur and phosphorus in coke is usually small.
Even if the sulfur content increases, to some extent, it can be taken care of
by adjusting the slag composition. Attention is, therefore, focused on to the
ash content alone.
The ash content of the coal blend for coke ovens is therefore adjusted
by blending and/or washing to produce coke with the lowest ash content
that is economically feasible without impairing the coke’s other physical
properties.
Most of the sulfur and part of the phosphorus in pig iron generally
come from coke; the remaining phosphorus comes from the ore. Their
presence beyond a certain limit unnecessarily affects the steelmaking pro-
cess adversely. The adverse effect of sulfur is more pronounced than that
of phosphorus. The ash of coke containing excess sulfur has to be regarded
from the overall economic viewpoint of iron and steelmaking. The tertiary
coal available in the northeastern part of India is of good coking quality, but
since it contains 2–7% sulfur (a major portion as inorganic) it is not eco-
nomical to use it for making coke and producing iron.
The chemical composition, although important in determining the qual-
ity of coke, is relatively less important than most of the other requirements
of coke which, directly or indirectly, affect the permeability of the burden
in the blast furnace and consequently its overall performance. These are
the strength properties of coke under a variety of conditions.
Reactivity
Reactivity may be defined as the rate at which the carbon atoms of coke
react with oxidizing gases like oxygen, CO2, H2O, etc. This is of interest in
BF to assess the ease with which coke burns and reacts with the above men-
tioned gases, which are invariably present in the blast furnace. The actual
rate of this reaction is a function of coke surface characteristics, its area
exposed and the chosen conditions. A highly reactive coke may react more
readily with the CO2 in the ascending gases in the BF and thereby get not
only consumed in the stack but lose its strength as well. Both of these are
detrimental to BF operation. A low reactivity may delay its burning in front
of the tuyeres which is not good for efficient BF operation.
Reactivity is measured under standardized conditions of experimenta-

tion wherein a known weight of coke is subjected to react with a known
composition of oxidizing gas under prescribed physical conditions to evalu-
ate the coke reactivity index (CRI) and also to assess the coke strength after
reaction (CSR). The Japanese method developed by Nippon Steel is uni-
versally adopted for determining the above properties. These values have
a great bearing on the efficient working of the BF because coke does react
with ascending gases in a BF and thereby gets degraded to some extent.
The CRI and CSR vary inversely with each other. A high value of CSR is
required for efficient BF operation. Under Indian conditions, the CSR is
generally in the range 55–65, whereas, for global best practices it is as high
as 75–80 measured under standard conditions. The rate of burning of coke
controls the rate of production from a blast furnace. In the absence of any
other data, the rate of burning is directly proportional to: the area of fuel
exposed to the blast, the temperature and pressure of the blast, and the
affinity of the particular type of carbon for oxygen.
A more open and well-developed cellular structure of coke will expose
the maximum area of carbon to the blast; this is what is desired without
impairing its strength and abrasion resistance. If other parameters are the
same, coke with higher porosity is a better fuel. An example quoted in the
literature shows that an increase in porosity by about 7% resulted in saving
of nearly 100 kg of coke per ton of pig iron produced.
Fortunately, experimental evidence indicates that the combustion of
coke imposes no limitations on the blowing rate because of its cellular
structure. This means it usually has adequate porosity. High reactivity of
coke with carbon dioxide, catalyzed by the presence of alkalis, however, is
detrimental to the furnace operation. This is because it results in degrada-
tion of coke during its descent in the furnace. A high porosity is detrimental
in this respect. The consideration of better combustibility is, however, more
dominant than the degradation effect due to solution loss reaction and a
higher porosity is always preferred.
Different cokes have different reactivities. This is related to the quality
of coal blend used and the carbonization process. There is little doubt that
the temperature of carbonization has a considerable influence on the prop-
erties of coke, particularly on its ignition temperature. The usual practice
is to measure the ignition temperature of coke, and the blast temperature
must be maintained in excess of the ignition temperature. The blast tem-
perature in modern furnaces is very much in excess of this temperature in
order to minimize the fuel consumption. This is also the reason why the
combustion of coke does not limit the blowing rate of a furnace. In general,
a high porosity of coke, coupled with adequate strength, is what is desirable
during coking.
Size
Of all the qualities of good metallurgical coke, the most important are its
size range and strength, which are also mutually interdependent. The size
range is known to affect the distribution of materials inside the furnace and
consequently the gas flow which has a direct bearing on the production
rate. The size of coke is generally chosen to match the size of other raw
materials. It is much bigger than the size of the iron bearing materials to
ensure maximum bed permeability for smooth furnace operation.
In spite of every effort, some degradation of coke, from the original
size, does occur when coke finally descends into the tuyere region. In order
to maintain adequate permeability in the bosh region, the original size of
the coke should be enough to accommodate this degradation.
Extensive investigations have been carried out to determine the influ-
ence of different coke sizes on the blast furnace productivity and coke rate.
The investigations have shown that the coke size has to be sufficiently lumpy
i.e. the proportion of +40 mm size in the charge coke should be at least 80%
to ensure a sufficiently high permeability of the charge. It was found that
for a mean size of burden of 13 mm, the mean size of coke should not be
less than 53 mm and this corresponds to an average coke to burden size
ratio of about 4:1. In general, it has been suggested that the optimum size
of coke should be 3–5 times that of the iron bearing material. The +80 mm
fraction is also detrimental since this size has lower stability than that of
–80 mm fraction. The +80 mm fraction should also be removed as much as
possible to provide coke of more uniform size.
During its descent in the furnace, coke gets progressively heated to
1500–1600 C before it burns in front of the tuyeres. When coke gets heated
beyond the temperature at which it was made in coke ovens, the process
of carbonization restarts and the coke begins to contract. The temperature
gradient in an individual piece of coke sets in differential contraction and
expansion, particularly where large mineral matter is present, resulting in
high local stresses and its tendency to degradation increases. It has been
shown that the thermal stresses experienced by coke in the blast furnace
are greater the larger the lump size of the feed coke. In other words, the
smaller size is better for better thermal stability. The following factors are
therefore believed to favor high thermal stability: absence of large lumps
in feed coke, uniformity of coke texture, minimum inert inclusions of large
sizes, high carbonization temperature and heat soak, prior mechanical con-
ditioning, and low chemical reactivity.
Strength and Abrasion Resistance

Considering the simplified life history of coke, from the coke oven where it
is made to its final demise in front of the tuyere in the blast furnace, coke
has to stand the rigors of handling and charging which include drops and
flows at several places before landing in the furnace. It then has to stand
the abrading action of the surrounding particles until it descends into the
tuyere region. In addition to this, coke has to stand high temperatures and
the nearly 20–25 m tall burden lying over it when it reaches the tuyere
level. A coke that breaks down into fines under these conditions will affect
the furnace permeability adversely. In fact, the height of the modern fur-
nace is controlled mainly by the strength of the available coke.
It is an established practice to estimate the suitability of coke for blast
furnace, in terms of its strength and abrasion resistance, by measuring its
shatter and abrasion indices empirically, as laid down by the various stan-
dard specifications. The shatter test gives resistance of coke to impact
whereas the tumbler test is designed to measure the resistance of coke to
degradation by a combination of impact and abrasive forces. In principle,
the shatter test consists of dropping the carefully obtained sample of cer-
tain fixed weight of coke from a standard height on a standard quality floor.
The shatter index is expressed as percentage retained on various sieves of
certain fixed sizes. In principle, the abrasion index is measured by putting
a standard weight of sample in a drum and rotating it for a fixed number of
revolutions at a fixed speed. The abrasion index is expressed as the percent-
age of material retained on various sieves of pre-fixed sizes.
Micum Test
The most popular tumbler test is the Micum test as per the German
Standard DIN 51712 of 1950. In the test, 50 kg of air dried coke sample of
+50 mm size is charged in a drum, one meter long and one meter in diam-
eter, and rotated for 100 revolutions at the rate of 25 rpm. The product is
screened on 50, 40, and 10 mm screens, and the results are expressed as
M40 (% of the total charge retained on 40 mm screen) and M10 (% charge
passing through 10 mm screen) values. It is believed that the M40 index gives
an indication of the hardness of coke and the M10 its resistance to abrasion.
The drum employed herein has longitudinal angle iron baffles inside. The
Indian Standard specifications IS: 1854 has adopted the same Micum test as
the standard test. The British Coke Research Association has developed and
extended Micum test wherein the revolutions of the drum are increased.
The ASTM standards D 294.64 (reapproved in 1972) and D 2490.70
have been devised to assess the hardness and stability factors of coke. In the
294.64 test, 35 kg of coke of 50 to +15 mm size is charged in a half Micum
drum (914 mm in length and 475 mm in diameter) and rotated for 1400
revolutions at 25 rpm speed. The fraction retained on a 25 mm screen is
termed as the stability factor and the cumulative percentage retained on
6.3 mm screen as the hardness factor. In the modified D 2490.70 specifica-
tion, the drum is rotated for 700 revolutions instead of 1400 and 34 kg of
coke of 50 to +15 mm size is taken as the charge. The results are reported
in an identical manner.
BS:1016 part 13:1961 describes a test by which the abrasion resistance
of coke can be estimated from the percentage retained on 1/8th inch screen
after 28 lbs. of coke of +2 to 3 inch size has been rotated for 100 revolutions
in a half Micum test.
A summary of the data pertaining to similar tests for iron ore are shown
in Table 9.1 (Chapter 9).
The value of these two Micum indices depends on the following factors:
1. Quality of coals(s) used;
2. Blending procedures adopted, if an;
3. Carbonization temperature;
4. Carbonization time;
5. Quenching procedure and subsequent handling;
6. Moisture content of the coal mix vis-a-vis the bulk density;
7. Extent of crushing and degree of oil addition;
8. Mechanical compaction used.
The examination of coke samples removed from in front of the tuyeres
of operating furnaces have demonstrated that there are major anomalies
between the quality of coke as measured by conventional strength indices

at the point of charging and the behavior of that coke in the furnace. While
these indices may be of value for day-to-day assessment of coke quality at a
particular plant, they do not provide a satisfactory measure of the ability of
the coke to resist breakdown inside the furnace.
The strength of coke under reducing conditions and after gasification
is more likely to provide the required information to assess its behavior
inside the blast furnace. If there is evidence that furnaces with alkali con-
centrations in the burden show increased breakdown, probably by affect-
ing the gasification mechanism, the effect of alkali concentrations on coke
breakdown must be examined in detail. The development of appropriately
simulated tests, coupled with planned examination of coke samples directly
drawn from the operating furnaces, will permit the establishment of more
satisfactory relationships between coke quality and its performance in the
furnace. This will ultimately lead to the production of coke more suited for
the modern high productivity furnaces. An excellent account of low blast
furnace productivity as adversely affected by the poor quality of coke has
been discussed with reference to the furnace operation in TISCO.
ASTM Tumbler Test (U.S.)

A 10 kg representative sample of the –75 mm +50 mm square hole coke is
placed in the specified tumble drum and rotated for 1400 revolutions. The
test is based on the American standard ASTM D 3402.
• Stability – the percentage of coke +25 mm after 1400 revolutions
• Hardness – the percentage of coke +6.3 mm after 1400 revolutions
Generally the higher both indices the stronger the coke. The test is com-
monly used in North America to evaluate coke strength.4
Preparation of BF Fuel
The blast furnace requires coked coal or coke; as a fuel. Coke is more porous
and yet has much better strength and other properties more suited for blast
furnace operation. Coke is produced from certain coals, called metallurgical
coals, by the process of carbonization, which essentially means the heating
of coal out of contact with air, when it fuses and forms a cellular structure.
It is nothing but destructive distillation of coal under controlled conditions.
4
http://www.coaltech.com.au/Coketests.html
During carbonization, the volatile matter contained in the coal escapes and
may or may not be collected as by-products. In the old style Beehive Ovens,
much of these by-products were not collected, whereas in the modern
By-product Coke Ovens or Slot Ovens, as they are called, these by-products
are collected. One of the important by-products is tar. A variety of beehive
oven-designs, developed in various countries, are available on the market.
The quality of coal is assessed in terms of reflectance of constituents of coal
called macerals which consists of three main groups: vitrinite, exinite, and
inertinite, out of which vitrinite is predominant. The reflectance of vitrinite
is measured under a microscope and this value determines the rank of the
coal and thereby the coking tendency of the coal.
Other properties of interest to assess the coking properties of coal are
the free swelling index, grindability index, fluidity, crucible swelling num-
ber, and dilatometric properties. These decide the quality of coke that will
be finally produced under certain conditions of carbonization. In principle,
when carbon in the form of coal is heated out of contact with air, the coal
mass almost fuses; particles join each other and form a more compact mass.
The loss of volatiles (which are generally 20–35%) make the structure more
porous. The conditions of carbonization have only marginal influence on
coke properties. The value of coke is assessed in terms of its:
1. Room-temperature strength;
2. High temperature strength;
3. Reactivity (RI);
4. Ash content;
5. Chemistry;
6. Strength after reduction (CSR).
The cost component of coke in hot metal production is nearly 70%. Coke
quality has a tremendous influence on blast furnace productivity and its
economics as determined by the fuel consumption for producing a ton of
hot metal i.e. coke rate. Out of these, the CSR and the RI are by far the
most influential properties of coke from this point of view.
The quality of coke primarily depends on the quality of coal or coal blend
used for making coke. Improvements in coke quality, to some extent, can be
brought about by pre-carbonization techniques. The essential design of the
by-product or slot coke-oven has been the same since the nineteenth-century.
Several modifications, however, have been made to improve its efficiency of

carbonization. The oven is usually 400–600 mm wide, 1300–1400 mm long
and may vary from 4 to 7 m in height. Several such chambers (15–55 in Nos.)
are put together to form a coke oven battery. The gap between the two adjoin-
ing chambers is used as the common heating chamber to externally heat the
chambers on either side. They are charged from the top and the doors are
sealed after charging is over. The chambers are usually lined with silica bricks.
The charge is usually a fine ground coal blend with 80% of the charge of 3
mm size. It takes 16–20 hours for carbonization, at the end of which, the red-
hot coke, approximately in size of that of the chamber, is horizontally pushed
out and quenched either in water or in nitrogen stream. As it comes out of
the oven, it breaks. It further breaks on quenching. However, it still requires
crushing and is therefore crushed and screened to separate the fraction, usu-
ally +50–80 mm that is ideally required as a blast furnace charge. The sil-
ica lined ovens are continuously kept hot to prevent their spalling. Being an
important raw material for ironmaking, coke-oven batteries form an inalien-
able part, like a sinter plant in most integrated steel plants using BF route.
A good quality coke can be made from a good quality metallurgical
coal by an appropriate carbonization process. Since not all the metallur-
gical coals have the required chemistry, the charge for carbonization is
obtained by blending two or more types of coals and/or beneficiation of
the low grade coal. The latter operation is commonly known as washing of
coal. With the increasing consumption, the reserves of good quality of coal
are slowly depleting. The proportion of beneficiated or washed coal used
in making coke is therefore steadily increasing all over the world. In fact,
in some countries coal washeries are considered essential adjuncts to coal
mining operations. Significantly, coal washing has been considered a must
for countries which have been exporting good quality coals as they pay rich
dividends. Washing of coal is adopted chiefly to obtain a useful fraction of
coal with acceptable low ash content.
4.4 Preparation of Coke (India)

Although India has a vast reserve of coal, the proportion of metallurgical coal
is far too small as shown earlier in Table 3.8. The poor quality is with respect
to their rich, inert characteristics, both organic as well as inorganic, and poor
coking characteristics. It means they are low in vitrinite and reflectance val-
ues (poor ranks). The ash content has steadily increased from 16 to 24% over
the years and is still on the rise. Washing of coal is therefore imperative to
obtain a coal blend of 15% ash to finally produce coke of up to 22% ash con-
tent. This is the main reason why the productivity of Indian BFs is very poor.
Indian coals are well known with respect to their washability charac-
teristics because the ash is finely and evenly distributed in the coal matrix
thereby rendering washing very difficult. Generally, heavy media separation
is adopted for 75 +25 mm fraction and cyclone or jig for the 25 + 6 mm
fraction after the run-of-mine is all crushed below 75 mm size. The cyclone
washer design adopted from the Dutch State mines has proved quite suit-
able for washing Indian coals. It may be mentioned here that the annual
coal washing capacity now exceeds 60 Mt annually.
In order to conserve the prime coking coal and to optimize its use, the
Central Fuel Research Institute (CFRI), Dhanbad has developed a coal
blend of coking coals as:
Prime coking coal .......... 40–50% max.
Medium coking coal .......... 30–40%
Semi or weakly coking coal 10–20%
This division is based on the caking index of coal. A CI of above 20 is

considered as prime coking, a CI of 17–20 as medium, and below that as
weakly coking coals. The blend that was then recommended was:
1. Low volatile prime coking coal of Zaria 45–50%
2. Medium volatile medium coking coal of Bokaro 30–40%
3. High volatile semi to weakly coking coal of Dishergarh 10–20%
The resulting coke was of poor quality with respect to CSR (50) and
CRI and ash content (around 25–28%), but it had to be tolerated since no
import of good quality coal was allowed. As a result, the productivity of
Indian BFs were 1.0 thm/m3/d only. The BF sizes were also very small as
compared to the better ones overseas.
Since prime coking coal reserves in India were too meager, maximum
use of coals other than prime coals were to be used for ironmaking. Special
coal blending and preparation techniques including selective crushing, oil
blending, etc. were developed at the CFRI and were adopted until the lib-
eralization in 1991.
The Indian standard specifications for BF coke were laid down in
Table 4.1. The situation has undergone significant change after 1991 when
the import of good quality coal became readily possible.
TABLE 4.1 Indian Standard Specifications for Blast Furnace Coke

(IS: 1354–1954 revised)
Characteristics Grade I Requirements of Grade III
Grade II
Proximate analysis (dry basis)
Ash percent maximum 20 24 28
V.M. percent maximum 2 2 2
% Sulfur 0.70 0.70 0.70
%Phosphorus 0.30 0.30 0.30
Shatter index
Over 38 mm, % maximum 85 85 –
Over 13 mm, % maximum 97 97 –
Micum index
Over 40 mm, % minimum 75 75 –
Through 10 mm, % maximum 14 14 –
Stability factor
On 25 mm, % minimum 40 40 –
Porosity, % 48 35–48 –
The rank of the coal also determines the coking properties. For good
coking the reflectance of vitrinite should be in the range of 1.15–1.25 and
the fluidity of the coal blend should be at least above 250 ddpm. The above
division was based on getting the stipulated metallurgical specifications
of coke for Indian conditions, though not of desired quality. Now import
of coking coal every plant has developed their own coal blend, incorpo-
rating the imported high value, low ash, good quality coking coal, to suit
their individual quality requirements vis-a-vis the economics involved. The
Indian standard specifications for blast furnace coke once adopted under
the socialistic controls are no longer valid. Every attempt is being made to
use coke as close to international standards as possible, using imported coal
in part and developing the coking process suitably.
With these modifications, it is now possible to obtain CSR in the range
of 58 to 65 which is the bare minimum needed for good blast furnace pro-
ductivity. A productivity of 2.0–2.25 t/m3/d has thus been achieved at Tata
Steel. In some of the very good blast furnaces on a global level, the CSR is
close to 80. This is one of the main reasons why their blast furnace sizes are
large and the productivity is 2.5 thm/m3/d.
Coal Ranking (U.S.)5

A major factor in determining coal quality is coal rank. Rank refers to steps in
a slow, natural process called “coalification,” during which buried plant mat-
ter changes into an ever denser, drier, more carbon rich, and harder material.
The major coal ranks, from lowest to highest, are lignite (also called “brown
coal” in some parts of the world), subbituminous coal, bituminous coal, and
anthracite. Each rank may be further subdivided, as shown in Figure 4.1.
The rank of coal is determined by the percentage of fixed carbon, moisture
(inherent water), volatile matter, and calorific value in British thermal units
per pound (Btu/lb) after the content of mineral matter and sulfur have been
subtracted from the total. Fixed carbon is solid, combustible matter left in
coal after the lighter, volatile, hydrogen-rich compounds are driven off during
coalification. Volatile matter is slowly removed from coal during coalification
but may be rapidly removed during destructive distillation. The volatile mat-
ter contains the raw materials from which the organic chemicals are obtained.
In the United States, the tests to determine the amounts of the above-men-
tioned substances and the rank of the coal are performed using standards
published by the American Society for Testing and Materials (1999).
Meta-anthracite
Anthracite
PERCENTAGE OF FIXED CARBON (DRY, MINERAL-MATTER-FREE BASIS)
Semi-
anthracite
Low-
volatile
bitumi-
nous
Medium-
volatile
bituminous
High-volatile C bituminous
High-volatile A bituminous
High-volatile B bituminous
Subbituminous B
Subbituminous C
Subbituminous B
or
Lignite
GROSS CALORIFIC VALUE (BTU/LB ON A MOIST,

MINERAL-MATTER-FREE BASIS)
FIGURE 4.1 Diagram showing classification of coals by rank in the United States. Rank is a measure of the
progressive alteration in the series from lignite to anthracite. Higher rank coals are generally harder, contain
less moisture and volatile matter, and have higher calorific values. Modified from Trumbull (1960). For a
more detailed explanation of coal-rank determination, see American Society for Testing and Materials.
5
https://pubs.usgs.gov/pp/1625f/downloads/ChapterC.pdf. USGS.
In general, the higher the rank of a coal, the more deeply it was buried,
and, therefore, the higher the temperature it was subjected to during and
after burial. Older coals tend to be of higher rank because they are more
likely to have been buried more deeply for longer periods of time than
younger coals. To give a sense of the effects of increasing rank, the follow-
ing example may be used: lignite is soft, dusty, and can ignite spontaneously
under the appropriate conditions, whereas anthracite is quite hard, clean to
the touch, and must reach a temperature of about 925 °F (496 °C) before it
will ignite. Furthermore, anthracite contains about twice the calorific value
of lignite (about 15,000 Btu/lb, 3,780 cal/g and 7,000 Btu/lb, 1,764 cal/g,
respectively) because lignite contains more moisture and oxygen and less
fixed carbon than anthracite. Subbituminous and high-volatile bituminous
C coals have oxygen and moisture contents and calorific values that range
between those of lignite and anthracite. Bituminous coals of higher rank
than high-volatile A (Figure 4.1) have calorific values that may exceed those
of anthracite.
4.5 Modifications in Coke Making Practices

Any developments and modifications in coke making practices is chiefly
meant to improve the CSR and CRI properties of resultant coke. In this
respect, improvements in the design and operation of slot-ovens have
almost reached its pinnacle and not much improvement is expected. All
the improvements that are brought about are therefore directed towards
improvements in pre and post carbonization technologies. These generally
include:
1. Preheating;
2. Briquette blending;
3. Stamp charging;
4. Admixing oil in the charge;
5. Admixing water in the charage;
6. Using a better coal blend.
Stamp Charging of Coke Ovens

Several technologies like briquette-blending, binderless-briquetting, pre-
heating, and stamp-charging have been developed in recent years for
improving the quality of coke. Their effect on improving the coke-proper-

ties are shown in Table 4.2.
TABLE 4.2 Comparison of Important Parameters in Different Pre-

carbonization Technologies
Parameters Conventional Briquette- Binderless- Preheating Stamp

Top charging blending briquetting charging
Bulk density of coal 700–750 750–800 750–800 800–850 11–1200

charge kg/m3
M10 of coke 10–11 8.5–9.5 9.0–9.5 8–9 5.5–6.5
CSR of coke 30–40 35–45 35–45 45–50 60–65
Oven throughput 100 105–110 105–110 110–115 112–115

index
Maintenance 100 105 100–105 150–175 110–115

requirement index
It has been accepted that coke strength from the same quality of charge
can be improved by compacting coal particles to improve the bulk-density
of the charge inside the oven. This increases the particle-particle contact
and develops stronger bonds when coal softens at carbonization tempera-
ture. To achieve this more effectively, stamp-charging has been developed.
This is a pre-carbonization technological improvement. A solid cake of coal
blend almost similar in size and shape to that of the oven chamber is formed
outside to achieve bulk density of nearly 1150–1200 kg/m3. This cake is
inserted in the oven horizontally from the ram side door. The rest of the
process of carbonization remains the same. The data in Table 4.2 proves the
superiority of this technique.
In stamp-charging, the coal-mix charge of coke ovens is prepared by
compacting it by drop hammer such that the bulk density of mix increases
as compared to when the mix is just charged in the oven. This was adopted
by Tata Steel in India for the first time. The compacting leads to achieve
bulk density of the order of 1150 kg/m3. Addition of moisture of 8–9% of
the mix with coal charge size of 90% below 3.2 mm size is found to be bet-
ter. This increased density is obtained by dropping heavy weight on the
mix, and the cake thus formed is charged in the ovens. The dimensions of
this pre-formed cake have to be developed through detailed experimen-

tations. The compacting presses the coal particles closer to each other
and form bonds more readily and at a higher strength. All the variables
are adjusted to obtain maximum strength in resultant coke. All the coke
oven batteries at Tata Steel were then fed with stamp-charged coal mix.
The coke CSR therefore improved from earlier 35 to 58–65 and it obvi-
ously led to improvements in blast furnace productivity from the earlier
1.3 to 2.25 t/m3/d. The use of stamp-charged coke made from a blend,
using in part imported coal and using powdered coal injection from the
tuyeres, finally reduced the coke rate at Tata Steel in their BFs from
650–700 to 460–470 kg/thm. However, they have yet to achieve the CSR
of 65–80 achieved by BSC, Sumitomo, Usinor, BHP, and other modern
steel plants.
The advantages claimed by Tata Steel for stamp-charging are adopting:
1. Higher yield of sized coke for BF usage;
2. Higher bulk density of final coke; an advantage for skip charging of BF;
3. Better M10 index-signifying higher stability under ambient conditions;
4. Better coke strength after reaction; leading to improved BF operation;
5. Improvements in M10 and CSR values are not sensitive to coal blend;
lower rank coals can be used;
6. Substantial cost effectiveness.
The improvement in coke making also included a few other developments
along with the introduction of stamp-charging like reduction in ash content
of charge-coal, incorporation of cheaper coal in the coke-mix, and use of
moisture and oil in the mix.
Economizing Coke-Charge-Mix
In the ultimate analysis, even after all improvements, the cost of coke must
also be low and competitive. Coals are divided into coking i.e. hard cok-
ing coal (HCC), semi-soft coking coal (SSCC) and thermal or non-coking
coals by the major international suppliers like Australia. The HCC is much
costlier than SSCC, which is costlier than the thermal coals. In India, these
are generally known as prime coking, medium coking and semi to weakly
coking coals.
At Tata Steel, good quality, low ash, important coal, to the extent of
50–60%, is being incorporated to produce good quality of coke. The extra
cost thus incurred has been off-set, to a large extent, by incorporating
5–10% of non-coking coal in the mix in the coal blend used for coke making.
Stamp-charging has made all this technologically possible. This improved
the overall BF economics. The use of 8–9% moisture and 0.4–0.6% oil
(LDO) has also improved the flowability and bulk density of the mix and
improved the resulting coke quality. All these improvements resulted in 4%
shrinkage in cake volume during carbonization, which is considered desir-
able from the point of view of oven life.
Use of Waste Plastic in Coke Making

Japan Iron and Steel Federation (JISF) have successfully developed a pro-
cess wherein plastic can be converted into coke for BF. The plan aims to
substitute 1.5% of the energy requirement of BF with plastic.
Initially, plastic in fine form was injected into the tuyeres like any other
injectant fuel. The chlorine contained in plastic poses a threat to the BF
lining life. Hence, initially low chlorine plastic was injected with the rea-
sonably good results. Later on, the process for de-chlorination of plastic
was developed, and then the dechlorinated plastic was injected in the BF.
It influenced the raceway immensely. The method of preparation showed
great influence on the quality of combustion and gasification behavior of
plastic in the raceway. The coarse plastic gave high combustion and gas-
ification efficiency compared to fine plastic and pulverized coal. The CO2
gasification rate of unburnt char derived from plastic was much higher than
that of pulverized coal. Hence, waste coarse plastic may even be useful as a
reducing agent in BF.
The JISF has reached plastic consumption in BF to as high as 1 Mt/
year. This will be nearly 10% of the total plastic waste generated. Plastic was
incorporated in the coal mix for coking and then processed through coke
ovens. The results have been quite encouraging. Tata Steel has also under-
taken trials on a laboratory scale and has confirmed that a certain amount
of plastic can be incorporated in the coal blend for coke making without
affecting the CSR and the RI adversely. For the plastic to be incorporated
in the coke oven charge-mix, it must first be freed of metallic, ceramic,
PVC (for chlorine) and other foreign matter. It must also be pulverized.
The JISF is recycling 100,000 t of such plastic annually for ironmaking. It is
also coming up as an important technology for waste plastic utilization and

reduces CO2 emissions and saves coal.
4.6 Alternative Fuels

Since coke is a scarce and dear commodity, every attempt is being made to
decrease its consumption by substituting, to whatever extent possible, with
other alternative fuels. No other solid fuel can be substituted for coke in
the charge. The alternative fuels are either liquid, gas or pulverized fuels,
which cannot be used as a constituent of the burden. These can only be
injected in the furnace through the tuyeres along with the blast. Successful
use of fuels like natural gas, oil, coal-tar, coal-slurry, etc. have already been
made at a number of plants, and it is fast becoming a routine practice.
Some of these fuels e.g. coal slurry, pulverized coal, etc. need to be pre-
pared properly before they can be conveyed to the blast furnace and finally
injected through the tuyeres. Charcoal was also used as an alternative fuel
to replace the coke entirely, using small furnaces. It has also to be prepared
from right type of forest wood. The charcoal should also be as strong and as
porous as is feasible. The proportion of iron made by charcoal furnaces is
practically negligible and charcoal preparation has not been discussed here.
Another important fuel to substitute for conventional coke is the
recently developed formed coke. Several processes have been developed all
over the world to produce molded fuel from non-coking lower rank coals.
These molded solid fuels are designated as formed coke. The formed coke
that has the requisite physical and chemical properties can only be used in
the blast furnace. An excellent account of the processes of making formed
coke is given by Lahiri and co-workers who claim the following potential
advantages of formed coke over the conventional coke:
1. Cheaper non coking coals are used to make the formed coke;
2. The choice of coals to make formed coke can be made more in terms
of their sulfur and phosphorus contents;
3. Relatively lower capital and running cost for its manufacture;
4. More consistent quality of the product from a given process;
5. Continuous operation of the plant unlike the coke ovens;
The possibilities of producing formed coke from Indian raw materials

have been investigated at the CFRI where a 12 t/day capacity pilot plant
has been installed and operated. The properties of the formed coke thus
produced by them are shown in Table 4.3.
TABLE 4.3 Properties of Formed Coke from Low Rank Indian Coals
Size: 38–63 mm
Shape: Ovoid
Proximate analysis:
Moisture: 4.5%, Ash: 25%,
Volatile matter: 1.8%, Fixed carbon: 68%
Micum index:
% on 40 mm (M40) 78.6, % below 10 mm (M10) 16.4
Shatter index:
% on 38 mm (S1.5) 93.6, % on 13 mm (S0.5) 97.5
Porosity: 46.9
Point crushing strength: 275 kg
Sulphur: 0.33%
4.6 Ferro-coke
A carbonized lump produced from a mixture of iron bearing fines and non-
metallurgical coal is known as ferro-coke. The proportion of iron bearing
fines in ferro-coke may vary from a few percent to 15–20% at best. It is
produced by carbonizing the mixture much in the same way as that for
conventional coke making. The product is found to possess all the neces-
sary qualities required for its use as a blast furnace fuel. The advantages
claimed are:
1. Non-metallurgical coal is used indirectly as a fuel in the blast furnace;
2. Total fuel requirement inside the furnace is reduced since certain pro-
portion of iron is charged as pre-reduced iron;
3. Coke consumption in the blast furnace is decreased;
4. Blast furnace productivity is improved;
This has been tried at many plants, but it has yet to be developed to the
stage of universal applicability.
This is of great potential interest to India because of the meager and
poor quality metallurgical coals.

Coke making has progressed to a considerable degree to decrease the cost
of coke and reduce its consumption in a BF per ton of hot metal produc-
tion. It is also clear that a certain amount of coke is necessary to run a
BF to maintain the permeability of the burden for hot reducing gases to
pass through. The better the quality of coke in terms of CSR and CRI, the
more efficient and economical the BF operation is. The reduction in coke
rate has been achieved by using substitute fuels injected through the tuy-
eres or incorporated in the coal-mix to make coke. The ferro-coke or the
formed-coke technology has not progressed enough in spite of few decades
of development to produce and use them. From this point of view, the use
of plastic in coal charge mix looks more promising for coke making and will
be followed with great interest by the ironmaking industry.
CHAPTER
5
DISTRIBUTION OF
TRADITIONAL BURDEN
In This Chapter
● Introduction
● Elements of Distribution
● Factors Affecting Distribution
● Limitations
5.1 Introduction
The blast furnace is essentially a furnace working on the counter-current prin-
ciple. The descending solid charge meets a current of ascending gases; reduc-
tion of iron ore along with its progressive heating has to take place during this
passage. Of all the reactions taking place inside the blast furnace, reduction of
iron ore is the most important and the most difficult. The production rate of a
blast furnace is directly determined by two important factors:
the rate of reduction of iron oxide and the rate of heating of the burden.
These two factors are not independent, but they are related to the
quantity of blast that goes through the furnace, i.e. its driving rate. The
efficiency with which both of the above functions are met in the fur-
nace stack predominantly determine the furnace productivity. The stack
is essentially a diffusion controlled counter-current gas-solid reactor in
which the ascending hot reducing gases reduce the iron oxide and heat
up the descending burden. The rate of reduction and heating of the
burden depends upon the degree and time of contact of gases with the
burden. The burden inside the furnace should therefore have uniform
and good bulk permeability. The uniformity of distribution of reducing
gases throughout the horizontal and the vertical cross-sections of the fur-
nace is an essential feature for efficient reduction and heating of the
burden. Since the gases are passing through the furnace at a tremendous
speed, the only way of achieving uniform gas-solid contact is by obtaining
uniform resistance of the charge to the ascending gases over the entire
cross-sections of the burden. In practice, this can only be achieved by
obtaining a uniformly permeable burden inside the furnace. A more uni-
formly permeable burden distribution also allows a large amount of blast
to be forced through the furnace, other conditions remaining the same.
The initial distribution of charge in the throat usually persists through-
out the furnace shaft. Hence, it should be controlled to obtain burden
distribution of uniform permeability or at least one which offers equal
resistance to the passage of gases all through the cross-section.
The gas utilization, i.e. the coke rate, is a function of the burden
distribution. It has been found that each additional percent of CO uti-
lized for reduction means a saving of almost 7 kg of coke per ton of iron
produced. The question of distribution of charge materials in the throat
of the furnace, under given conditions, plays a very dominant role in the
performance of a furnace. Burden distribution is a generic term used to
denote radial ore or coke distribution as well as the particle size distribu-
tion in a blast furnace.
If the charge particles were of similar size and shape, the burden would
be a uniformly permeable burden irrespective of the way the materials
were charged in the furnace. The furnace charge is however far from this
ideal situation, even from the point of view of certain metallurgical require-
ments, leaving aside the practical limitations. The blast furnace charge con-
sists of different sizes of coke and ore, sinter, pellets, and limestone, with
different physical properties.
It is difficult to distribute them at the top of stock column, so that
the entire vertical and horizontal cross-sections of the furnace offer
equal resistance to the gas flow. This is due to the fact that a consid-
erable amount of pretreatment of the charge materials is done with
this object in view. Deliberate additional attempts have to be made to
Distribution of Traditional Burden • 81
obtain a uniformly permeable burden inside the furnace, or one offer-

ing more uniform resistance to the passage of gases, with the help of
charging devices and charging procedures. In fact, control of the blast
furnace operation begins with control of the distribution of the charge
in the throat.
FIGURE 5.1 The two bell charging arrangement at the top of the blast furnace.
Until recently, blast furnaces in general were constructed with a

double bell charging arrangement at the top. The most popular design
was the McKee revolving top. In this design, the materials are first
charged in the small bell hopper, i.e. on the small bell which is then
lowered to allow the material to drop on to the lower or the big bell. The
big bell remains closed while the small one is being lowered, and then
the small one is closed and the big bell is lowered to drop the charge
inside the furnace. Such an arrangement prevents almost all of the fur-
nace exhaust gases from escaping into the atmosphere while charging as
5.2 Elements of Distribution

A mass of material consisting of varying sizes, shapes, and densities fall
inside the furnace with various trajectories. In general, dense, small, and
irregular particles remain where they fall, building ridges. Light, large, and
smooth particles roll into troughs. This situation is shown in Figure 5.2.
FIGURE 5.2 Segregation of fines and coarse particles form in a pile when material is allowed to freely
fall from a hopper.
This results in non-uniform permeability in the bed. The areas contain-

ing the coarser materials offer less resistance to the passage of the furnace
gases. This results in higher temperature and better reduction in these areas
which tend to descend faster than areas containing finer particles inside the
blast furnace. In general, the coarser particles tend to segregate in the cen-
ter of the furnace shaft, and the fines segregate at or near the wall depend-
ing upon the clearance between the bell and the furnace wall. Any means
of distributing the charge, whereby the zones of loose packing in the center
are minimized, will result in more uniform gas flow through the burden.
More uniform gas flow throughout the bed, however, results in increased
resistance to the gas flow which may necessitate higher blowing pressures.
As much as possible, iron ore should be present in the areas of maxi-
mum gas flow for efficient reduction. The system of charging, the level
of the charge and the size of the charge are to be selected so as to have
maximum utilization of gas, is subject to ensuring smooth performance of
the blast furnace. It has been observed that the burden moves down in the
furnace in slight jerks even when the operation is considered to be smooth.
The individual particles are under various pressures inside the column: the
overlying weight of the burden pushing it down, the frictional drag with
respect to surrounding particles which varies in amount and direction, the
pressure of ascending gas which varies considerably form particle to par-

ticle, the jet effect of the gas when the resistance to its passage is more
because of lower permeability, etc. All this leads to loosening of the bed
wherein the gas flow acts as lubricant. A narrow ring-shaped gas flow along
the walls is beneficial. It acts like a lubricant to help move the burden down-
wards overcoming the frictional drag of walls. The high permeability and
consequent rapid gas flow through the central zone help the lateral move-
ments of bigger particles and removal of surplus gas.
5.3 Factors Affecting Distribution

A number of factors affect the distribution inside the furnace, and their
careful control can minimize the unevenness of gas flow and lead to a more
uniform temperature and rate of reduction of the charge. The factors that
affect the distribution of charge inside the furnace have been summarized
below:
Factors Affecting Distribution Inside the Furnace

(a) Design of the blast furnace and its (i) Angle and size of the big bell.
charging device (effect of these factors is (ii) Additional mechanical device(s) used for
constant). obtaining better distribution.
(iii) Speed of lowering large bell.
(b) Inconsistency in physical properties of (i) Size range of the various charge materials
charge materials (deficiencies caused by (ii) Angle of repose of raw materials and other
this should be eliminated by improving physical characteristics of the charge.
quality of the burden. (iii) Density of charge materials.
(c) Level, system and sequence of charg- (i) Distribution of charge on the big bell
ing, program of revolving the distributor (ii) Height of the big bell from the stock-line
(conditions determining major means of i.e. charge level in the furnace throat.
blast furnace process control from top). (iii) Order and proportion of charging of various
raw materials.
Angle and Size of the Big Bell

The angle and the size of the big bell are of considerable importance.
For optimum operational results, it is generally agreed that an angle of
50–53 degrees from the horizontal should be maintained. This allows rapid
discharge from the bell, minimizing the possibility of build up of wet mate-
rials on the bell and its consequent interference with regular distribution.
The size of the bell should be viewed in relation to the throat diam-
eter of the furnace. The way the distribution is affected by the clearance
between the bell and the furnace wall is shown in Figure 5.3.
FIGURE 5.3 Scheme of location of crest and the contour-types as affected by the clearance of the big
bell from the furnace wall.
The larger the clearance, the crest (where fines pile up) accumulates
further away from the wall. The smaller the clearance, the more the fines
accumulate nearer the wall. Too much or too narrow a gap between the bell
corner and the throat wall tends to segregate the sizes and prevents normal
distribution from being achieved. Studies have also confirmed that there
exists a optimum ratio of throat to bell diameter which leads to optimum
segregation of the burden. The normal gap in the existing furnaces that are
working satisfactorily is nearly 800–1000 mm. The experience at TISCO is
that there is improvement in blast furnace performance with increased gap
between the bell and the throat. Experience in other countries indicated
that a good throat-bell diameter ratio seems to be 1.36. The actual ratio
should be selected depending upon the contents of fines in the charge.
While using lean ores of low density, calcined ores or ores of fine texture, it
is preferable to keep this gap on the higher side.
The V and the M type contour formations at the stockline, which are
also called as hoppers, depend upon the bell clearance from the wall as
The behavior of particles immediately after their impact on the stock-
line, to a large extent, controls the nature of segregation (i.e. uniformity of
permeability) in the furnace burden.
FIGURE 5.4 Location of crest and the contour-types affected by the stockline level i.e. distance from the
bottom of the big bell to the stockline.
The behavior depends upon the angle of the bell and the height of
fall of the material from the bell on the stockline (i.e. the height of the big
bell from the stockline). Although the stockline varies during the furnace
operation, there is a limit to which it is allowed to go down or come up with
respect to the designed optimum level. The effect of height of the big bell
above the stockline on the distribution of the charge is shown in Figure 5.4.
The crest is formed at three different heights. The lowering of the stockline
to the point where the trajectory of the falling material comes against wall
of the furnace results in fines segregating near the wall. This is true for any
system of charging and for any size range of the charge. This is shown at
position h1 and as the level goes up the crest moves away from the wall as
shown at positions h2 and h3. At the higher levels, say h2 and h3, the angle of
inclination becomes steeper and the hopper formed in the center becomes
deeper and the crest moves away from the wall. The permeability of the
whole improves with crest reasonably away from the wall. In other words,
it is necessary to keep the clearance between the bell and the throat wall as
large as possible (i.e. having a larger throat to bell diameter ratio).
The speed of lowering and raising of the large bell is of considerable
importance, since the faster rate of drop tends to throw the materials
towards the wall and alters the contour as well as distribution. The speed of
lowering should be as close to maximum as possible because slower speed
tends to segregate the fines. The speed of raising the bell back to its seat is
relevant only to the extent that it should not strike the seating too hard. This
problem is aggravated in a high top pressure operation.
The density of three important raw materials: the ore, the coke, and the
limestone, are quite different. The heaviest is iron ore with around 5–6 g/cc.
The lightest is coke with a density of around 1.5 g/cc, and the limestone is
intermediate with a value of density around 3.0.3.5 g/cc. This means that
the rolling tendency of coke particles is maximum and that of the ore is
minimum. Since the density values can not be altered, the sizes may be
chosen so that their differential rolling tendencies are offset to some extent.
The problem of very dense ores is serious due to their sluggish reduction
rates rather than their tendency towards segregation. Such ores are invari-
ably crushed and sintered to obtain more porous agglomerates before they
are charged in the furnaces.
Angle of Repose and Shape of Particles

When a multi-particle material is allowed to gently fall on a horizontal plane
it tends to form a conical heap. The base angle of this cone is known as
angle of repose of that material. This angle depends upon the particle size,
surface characteristics, moisture content, shape, size distribution, etc. For
an iron ore of 10–30 mm size, with an average mean size of 18 mm, the
angle of repose is around 33–35 degrees. For coke of 27–75 mm size, with
an average size of 45 mm, the angle of repose is around 35–38 degrees.
Similarly, the angle of repose for sinter is in the range of 31–34 degrees, and
for pellets it is around 26–28 degrees. The higher is the angle of repose, the
more the material has the tendency of forming ridges when charging in a
blast furnace. The central V Contour is deeper in practice for coke than that
for an ore, and it will be smaller for pellets. The drier the ore and the more
it is free from fines, the less pronounced the angle of repose. Thus, it has
less of a tendency towards segregation. The clayey ores tend to form ridges
because of their high angle of repose. The effective way to reduce the angle
of repose of any iron ore is to eliminate the fines, dry the ore if wet, and to
wash off the clay, if any, adhering the ore.
More angular particles create less segregation whilst a similar effect is
pronounced by moist but coarser materials. This is because the particles
here do not roll down to any significant extent. The spherical particles have
a large tendency to roll down and give rise to segregation if fines are present
in the charge. The angle of repose of charge materials, the charge size, and
the sequence of charging must be assessed to understand their effect, as a
whole, on the formation of ridges and their consequent tendency towards
segregation inside the blast furnace.
As the packing density or bulk density increases, the permeability

decreases with increase in the range of sizes of particles in a bed. Hence,
even if the permeability of the burden inside the furnace is somehow
achieved as uniform all over the cross-section, the net permeability of the
bed will be decreased with the increase in the range of sizes of raw materials
and also with increasing the percentage fines in the charge. Lower permea-
bility all through the cross-section results in decreased gas flow through the
burden (i.e. lower driving rate at a given pressure of blowing). Alternatively,
higher blowing pressure will be required to force the same amount of gas
through the burden. In short, increased fines in the charge decreases the
burden permeability and consequently the pressure drop inside the furnace
increases, along with its attendant problems. The reduction rate is there-
fore drastically decreased with increasing percentage fines in the charge.
Since the wider range of sizes in the charge is more prone to segregation
and crest formation, it is common practice to screen off the fines, particularly
the less than 5 mm fraction, before charging the materials in the furnace. Even
the sinter and pellets are also screened to remove the less than 5 mm fraction.
Many operators prefer to screen-off the less than 10 mm fraction. The frac-
tion of iron bearing material below this limiting size is therefore termed as
fines by the blast furnace technologists. It is invariably eliminated by screen-
ing at every possible stage. From the point of view of reduction, the maximum
top size of an iron bearing material should be as low as possible, since the rate
of reduction decreases, perhaps exponentially, with increasing size. The size
range of materials charged in the blast furnace represents a compromise to
create both good stack permeability and adequate bulk reducibility.
The following size ranges, as charged in the furnace, are typical of mod-
ern practice:
More reducible ores + 10–37 mm
Less reducible ore + 10–25 mm
Sinter + 5–50 mm
Pellets + 9–16 mm
The +5 mm fraction in the burden composes a maximum of 5%. Control

of this one single factor has improved the burden distribution, and thereby
the production, much more than any other factor.
A few of the old furnaces still employ buckets, and a few of the modern
ones employ a belt conveyor system for hauling the charge to the furnace
top. Barring these exceptions, the raw materials are hauled to the top of the
furnace by a mutually compensated two skip charging system on an inclined
railway. This arrangement requires an oval shaped hopper for the small bell
and the material is thrown from the skip at an angle of 45 degrees to the
bell control line (vertical).
The actual distribution of the raw materials on the large bell, prior to
their being dropped inside, affects the distribution in the furnace consid-
erably. In order to obtain better distribution on the large bell, and conse-
quently inside the furnace, several devices have been introduced. The early
distribution using an eccentric chute in place of small bell (i.e. the Brown
distributor) or using a deflector just above the small bell (i.e. the Baker and
Newman distributor) could not find wide use. The McKee revolving distrib-
utor as shown in Figure 5.5 or its modifications, however, were most popular
and continue to be used even today on most, old furnaces. Until recently,
all the furnaces in India were provided with this top only. In this design, the
raw materials are emptied in an elliptical hopper and then into the revolving
circular hopper of the small bell. The small bell hopper is rotated in a pre-
determined sequence. When one full charge, which consists of one or more
skipful of each of the raw materials, is charged uniformly around the small
bell, revolving the hopper after every skip is unloaded, it is lowered to drop
the charge on the big bell. This arrangement gives a better distribution of
charge on the bell and consequently inside the furnace as well.
(a) Both bells closed skip discharging the charge in the small belt hopper.
(b) Big bell closed, small bell open to allow the charge to fall in the big bell hopper.
(c) Small bell closed, big bell opened to allow the charge to be dropped inside the furnace.
(d) Both bells closed.
FIGURE 5.5 McKee top operation for charging the furnace.
Since one or more skipfuls (trash cans) of the individual materials are
charged, without prior intermixing in the skip, the revolving McKee top
ensures good mixing of the materials before these finally settle at the stock-
line. The extent of mixing that is possible with this technique is shown with
an illustration from the blast furnace charging practice at the Rourkela
Steel Plant in India.
Note: C indicates a coke skip and O indicates an ore skip.

The rotation is through 60 degrees.
FIGURE 5.6 Distribution of the burden on the small bell and on the big bell as affected by the rotation
of the small bell hopper.
The charging is carried out in the sequence CCOO↓ (that is two skips
of coke followed by two skips of ore and then the small bell is lowered), and
the hopper is rotated through 60 degrees after every skip is dumped on the
small bell, which is lowered after every four skips to dump the charge on
the big bell. When eight such charges (i.e. 32 skips) are charged, the big bell
is lowered to allow the whole charge to fall in the furnace. The distribution
of charge consisting of 32 skips, as would prevail on the big bell, is shown in
Figure 5.6 by a horizontal extension of the vertical position of the skipful of
charges in the form of rings. It is clear that vertically there will be alternate
layers of coke and ore in the charge on the big bell, and the mixing would
improve further when the big bell drops inside the furnace.
The use of several types of deflectors and distributors in the form of

rings, located immediately below the large bell, has been suggested in an
effort to obtain better distribution in the furnace. The McDonald and the
slick distribution rings, the moving frustum developed by Diamond, the
compensated charging device developed by Scott-Maxwell, etc. are but few
of the typical examples of efforts carried out before and just after the World
War II. While such mechanical devices were being developed, particularly
after the World War II, the high-top pressure operation was also intro-
duced for improving the furnace operation. The popular McKee revolving
top proved to be too inadequate to maintain a leak proof top with a suf-
ficiently long life. The philosophy of using additional mechanical devices
should therefore be regarded from the point of view of maintaining a good
leak proof top as well. A new range of top-designs were developed in the
immediate post second-war period and in the 1960s and 1970s and put into
operation. Some of these are described in Chapter 17. Once the superior-
ity of the Paul-Wurth-Belless-Top (PW-BLT) was universally established,
it was used on almost all blast furnaces constructed in recent times. The
well-known McKee Top is only found on old furnaces.
Any change in the blast furnace top design, as a means of moderniza-
tion, has to be preceded by a good length of experimentation. This may
include model studies as well as studies on full scale models constructed
specifically for this purpose. Design changes are brought about only after
successful full scale model studies are completed.
Tata Steel decided to adopt the moving throat armor in the early 1980s
on their old furnaces. A full scale model of 18 m height with 5 floors was set
up for this purpose. Nine notches were provided for setting the MTA posi-
tions. Each trial required 6000 kg of total charge in the desired proportion.
Over 200 trials were carried out to finally arrive at the decision of installa-
tion of MTA on the D furnace. A charging sequence of C7–8/O5/C3–5/S3–5 was
adopted and resulted in an immediate decrease in coke rate by about 40 kg/
thm. Coke rate fell further down with passing years. The furnace campaign
time increased from six to nine years following the introduction of MTA (i.e.
corresponding to about one million tons of extra hot metal from the furnace).
Similar results were obtained on other furnaces: A, C, and F when the MTA
was installed in them. The G furnace at Tata Steel is fitted with the PW-BLT,
but model studies were carried out for this as well and finally these proved to
be far beneficial in improving the furnace performance. The importance of
model studies in burden distribution hardly needs any extra emphasis.
Order of Charging
Mozumdar emphasizes the role of order of charging in furnace distribution
as: change in the system of charging of raw materials influence the distribu-
tion and character of gas flow to an extent considerably more than changes
of level of charging and size of the charge. Therefore, all conditions similar,
only 3.5% of the cases go to the credit of the control of distribution of gas
flow by changing the size of the charge and 85% control due to the change
of the system of charging.
It has been observed that more uniform gas flow through the bed can
be obtained much more readily by proper sequencing of the charge, if coke
is charged on the bell first. There is a definite excess of coke at the wall
and at the center of the furnace. When ore is charged first, there is no
pronounced layer formation in the stack and there is no excess of ore at
the walls; there is also a narrow zone of excess coke in the center. Changes
in the charging sequence and variation of the ore to coke ratio have to be
based on the existing physical conditions of the column inside the furnace.
With the usual M-type contour at the stockline, a large volume of gases
pass along the peripheral ring, which is not wide but occupies a consider-
able space. Maximum loading of this peripheral space with iron ore not only
corresponds to better utilization of gases, but it also prevents excessive ero-
sion of the refractory lining by the high velocity gases. Any excessive load-
ing of the periphery with iron ore, however, leads to excessive resistance of
the peripheral ring to the ascending gases. The situation can be rectified
by shifting the excessive peripheral load to the center. The same reasoning
may be applied if the central trough is overloaded with the ore. In short,
the charging sequence, in relation to the level of charging and size of the
charge, has to be adjusted to have maximum utilization of gas, ensuring
smooth performance of the furnace.
As discussed previously in this section, the commonly adopted system
of charging is CnOm↓, where n is the number of skip of coke and m is the
number of skips of ore. The ↓ sign indicates the lowering of the small bell.
Smooth performance of the blast furnace with this system is possible only
if the burden is well prepared. With this system of charging, the periphery
gets loaded with coke. The OmCn↓ scheme of charging is just the oppo-
site of the CnOm↓ scheme of charging, and it has just the opposite effects.
Continuous work on OmCn↓ system brings about gradual overloading of the
center and periphery with ore. In the system Om↓ Cn↓, the ore and the coke
are charged separately. In comparison to OmCn↓ system, this decreases the
ratio of iron ore to coke in the peripheral zone and increases the same in the
center. The system Cn ↓ Om ↓ is opposite Om↓ Cn ↓ as OmCn↓ is to CnOm ↓
in its effects.
The COmCn–1↓ system is more flexible as it distributes coke over a wide
range. The amount of coke charged prior to ore gradually and smoothly
reduces the ratio of ore to coke at the periphery and increases at the center.
The system OOCC↓, CO↓ CCC↓, or OC↓OCC↓ tend to load the
periphery with ore but coke consumption goes up a. Conversely the system
COOC↓, CCC↓OO↓, or CCCOO↓ loosens the already loaded periphery
but at the cost of increased coke consumption.
Any uneven behavior of the furnace may be eliminated by application
of the CCO↓CCO↓ system subsequently changed to the COOOC↓ system
or CCO↓CCO↓ or CC↓OOO↓ systems. Charging of extra coke and loosen-
ing of the periphery, as mentioned above, are commonly adopted if uneven
movements of the burden are observed.
The observations about the effects of any system of charging as mentioned
above are of a general nature. Since these are a function of various variables,
including the design parameters of the charging system, ones own experi-
ence of sequence of charging a particular blast furnace should weigh more for
future guidance while controlling the distribution of burden in any furnace.
The study of the distribution of burden on large scale models is increas-
ingly being adopted all over the world. Details of investigations on a 1 : 10 scale
model of 1719 m2 furnace has been reported.1 Similarly, a 1 : 18 scale model has
been used at the Bethlehem Steel Corporation to evaluate a parameter called
a uniformity index of distribution. This index quantified the degree of unifor-
mity of ore and coke layer thickness in radial direction. This concept of char-
acterizing the existing and proposed alternative filling practices was utilized in
developing improved charging practices. By selecting charging practices with
uniformity indices greater than those of existing practices, the Bethlehem Steel
Corporation could improve the performance of their blast furnaces.
5.4 Limitations
With the conventional bell-top charging system, the flexibility of burden
distribution is limited. The control in this case is limited to the shifting
of the peaks which inevitably occur in the form of M and V contours.
These shifts are brought about by placing the desired material towards
the wall or the center of the furnace. The optimum distribution can be
found out by testing alternative charging sequences, perhaps in relation to
the cyclic alternations of the stockline level, between specified lower and
upper limits.

In blast furnaces with conventional bell tops, the non-uniformity or segre-
gation of raw materials sets in from the beginning of its discharge through
the skip, small bell hopper, big bell, and finally to the furnace top. It
therefore requires that the distribution should be studied at each of these
stages of discharge until a final distribution is achieved at the stockline.
The top charging system plays a very important role in controlling the
burden distribution, thereby the gas flow, and consequently the furnace
performance. In these old designs, the sequence of charging plays a very
dominant role, besides burden preparation, in improving the blast furnace
performance.
CHAPTER
6
BURDEN PREPARATION:
TREATMENT OF IRON ORES
In This Chapter
● Introduction
● Burden Qualities
● Crushing and Sizing
● Beneficiation of the Iron Ores
● Calcining/Roasting
● Blending
● Shipping
● Products of Ore Processing
● Treatment of Iron Ores (India)
● Agglomeration of Iron Ores
● World Resources (2017 USGS)
6.1 Introduction
The developments in the field of ironmaking have always been directed
towards increasing the productivity of the blast furnaces with minimum
fuel consumption. In recent years, the emphasis has also been to obtain
consistent and better quality pig iron, with low silicon and sulfur contents,
owing to the consequent economy in its conversion to steel. Effective
process control and thereby maximum productivity can be achieved if bet-

ter quality charge materials, with respect to their uniform size, adequate
strength and good reducibility, are consistently used to run the furnaces.
Until the time when furnaces were small and consequently the production
rate was low, the ore was mainly hand-mined and sorted. The use of such
ores did not pose any serious problem in running the furnaces smoothly.
With the exception of drying very wet ores and calcination of carbonate
or hydroxide ores, hardly any other processing of the ores was considered
worth the trouble.
Increasing demand of iron and steel, particularly during and after
the World War II, led to the design and construction of blast furnaces of
increasing sizes. The large amount of ore required to feed these larger
capacity furnaces led to the development of mechanized mining and han-
dling of ores. Mechanized mining produces 50–70% of the output as below
10 mm size. Transportation and handling also produce a further amount of
finer fractions. Screening of ores therefore, to remove the fines, became
unavoidable. Varied sizes of ore cannot be fed, these days, to run a BF. They
need to be screened to obtain the correct size fraction for running the BF
smoothly. This crushing and screening broadly leads to the generation of
three different ore-size fractions below:
Lumps:
1. size 10–40/50 mm for direct charging in the BF.
2. size 5–18/20 mm for DRI production in coal-based kilns
3. size 10–30/40 for DRI production in gas-based processes
Coarse Fines:
1. size −10 mm and +100 mesh is usually sintered.
Very Fine Fines:

1. size below 100 mesh which can only be pelletized.
The −100 mesh fraction may be amenable readily for beneficiation, and if
required, it is generally beneficiated to improve upon the iron content and
decrease the gangue contents before pelletization. Since the acceptable BF
fraction is hardly 20–30% of the total run-of-mine, when the mechanized
mining operation is adopted, it is difficult to afford wasteful rejection of
nearly 70–80% of the run-of-mine. The two remaining fractions, coarse and
Burden Preparation: Treatment of Iron Ores • 97
fine fines below 10 mm, are therefore artificially agglomerated, whereby

these fines are converted into a lumpy, acceptable size for the BF, with bet-
ter blast furnace properties.
Once agglomeration of ore fines became a commercial reality, benefici-
ation of lean ores into a good quality concentrate of iron, and consequently
as good quality blast furnace feed, developed rapidly. A blast furnace opera-
tor would now rarely argue with the statement that, all other things being
equal, the most important element in blast furnace production and perfor-
mance is the condition of the burden. In fact, the cost of burden prepara-
tion can be more than offset by the increased productivity of the furnace.
It is therefore a common practice to screen the iron ore again at the plant
site to eliminate the fine fraction and to charge only the required lumpy frac-
tion to the BF and the fine fraction to be sintered. A last minute screening is
also conducted to remove fines; all this helps to run the BF more smoothly.
The very fine fraction, however, is amenable ideally to pelletization only
and can be adopted if otherwise economical. Generally, it is sold off or
dumped or micro-pelletized and fed to the sinter plant.
6.2 Burden Qualities

In order to obtain a smooth and hard (rapid) driving operation, the burden
charged in the furnace should ideally possess the following properties:
Physical
1. A close size range with minimum fines.
2. An ability to withstand the physical stresses incurred on being trans-
ported to the furnace, charged to the hopper and the bells, and, finally
in the furnace.
3. Non-decrepitating nature.
4. An ability to withstand mildly reducing conditions at lower tempera-
ture without breaking.
5. A good bulk reducibility to obtain equilibrium conditions between
solid and gas phases in the stack.
6. Low swelling tendency during reduction.
7. A high softening temperature with a narrow temperature range of fusion.
Chemical
1. A high percentage of iron and low gangue contents.
2. A low percentage of silica, alumina, etc. and a low alumina/silica ratio.
3. Proper overall chemistry of the burden to ensure adequate desulfur-
ization of metal and absorption of coke ash in slag.
4. Proper overall chemistry to ensure clean slag and metal separation at
minimum temperature and free flow of both slag and metal from the
furnace on opening the corresponding outlets.
The quality of naturally occurring raw materials may not be as ideal as
detailed above. It can however be improved, as far as is economically fea-
sible, by pre-processing the natural raw materials. These pre-treatments or
processings are known as burden preparation which includes the treatment
of ores, coke, and limestone. Many times the sole attention is focused on
to the treatment of iron ores alone, because of their dominant importance.
This does not underrate the importance of the quality of coke in smooth
operation of a blast furnace. Burden preparation often means agglomera-
tion of ore fines with or without flux.
The main purpose of the preparation of an iron ore is the modification
of the chemical and physical characteristics of iron bearing minerals. In
their original condition , they are not all that suitable as a blast furnace
feed. Any additional operations of preparation of ore for blast furnace
involve additional cost. Until the advantages that would accrue from ore
preparation were properly understood, it was considered unnecessary. It
has now been conclusively proved that the cost of preparation of ores is
invariably more than compensated, by way of increased production rate
and efficient control of the modern large capacity furnaces. The prepara-
tion of ore produces pig iron of more consistent and better quality. Today,
the use of unprepared or even a partially prepared burden is rather an
exception dictated by financial and or historical circumstances. In such
exceptional cases, efforts are being made to put ore preparation on a
sounder footing.
The possible general characteristics of natural iron ore are shown in
Table 6.1. The probable ore treatment to prepare a better feed from these
ores is also shown. These treatments are discussed in this chapter and the
agglomeration of fines is addressed in the following chapters. An iron ore
may need one or more such treatments to render it acceptable as a blast
furnace feed. The ultimate choice of treatment has to be made in terms of

the economy (i.e. both additional capital and running costs and the conse-
quent returns that might accrue from this modification).
TABLE 6.1 The Characteristics of Natural Iron Ores and Requirements

for Suitable Blast Furnace Feed
Available ore Required quality Processing needed
1. The run-of-mine iron ore 1. Lumpy (natural). 1. Screening after crushing to

may vary from fine powder obtain closed sized fraction.
to several hundred cm size
boulders
2. Ores may be lean with high 2. Lumpy with uniform compo- 2. Fines to be agglomerated.
percentage of gangue and sition, porosity and chemical
low percentage iron. responses.
3. Ore may be wet. 3. Low swelling characteristics. 3. Entire ore is crushed
and ground and then
agglomerated.
4. Ore may be carbonate or 4. Sufficient strength (for 4. Beneficiation like washing,
hydroxide. handling as well as during magnetic concentration, jig-
reduction). ging, etc. to be carried out.
5. Moisture associated with the 5. Dry 5. Drying
iron oxide may be consid-
ered as gangue.
6. Ore may vary in composition. 6. Chemically preferably as 6. Calcination for carbonate
Fe2O3. ores
7. Pre-reduced or metallized 7. Pre-reduction of ore.
pellets are a useful feed.
8. Uniform composition. 8. Blending of suitably sized
fractions for preparation of
a blended lumpy ore feed or
for agglomeration.
6.3 Crushing and Sizing

The choice of the right size of an iron ore suitable for efficient blast furnace
operation is dictated by many considerations. The adverse effect of fines
below 5–10 mm size on the distribution of burden inside the furnace has
already been discussed in section 5.3. Besides this, fines lead to excessive
dust losses and increase the load on the dust catchers and the gas cleaning
plant. Most furnace operators, therefore, avoid material below 5–10 mm
size in the charge. As screening efficiencies are never perfect, a small
fraction of material below the limiting size does get into the furnace. This is
inevitable even in the best practice. Any excessive proportion of fines in the
charge can be tolerated only at the cost of loss in production.
The reduction of ore inside the furnace largely takes place by gas which
has to diffuse into the ore particles. Too large an ore size would result
in unreduced material coming down the stack into the tuyere area and
increase the total demand of heat. The maximum limiting size is based on
ore chemistry, its bulk reducibility and other characteristics. An appraisal of
the prospective ores is generally made at each plant before the installation
of a blast furnace to assess the maximum tolerable size for a certain furnace
design. Assessment of ores generally continues even after the installation of
the plant. This improves the performance of the furnace and also takes care
of the possible variations in the mine product as mining progresses.
The run-of-mine in the modern mechanized mining operation is usu-
ally in the form of boulders whose maximum size may be up to 1.0–1.25 m
(3.4 feet). If the ore is high grade, oxide-type, and dry, the only preparation
necessary is to reduce the size and grade it before it is charged in the furnace
as natural ore lumps. The run-of -mine is generally crushed at the mine site
to a maximum size in the range of 100–200 mm (4–8 in). Final crushing and
screening is either carried out at the mine or at the steel plant. The various
alternative routes of treatments in principle are shown in Figure 6.1. The
cost of the ore increases with the extent of treatment carried out on the run-
of-mine and the distance over which it needs to be transported for smelting.
Therefore, international prices have to be referred to the size and or the
form of the ore with respect to the location (i.e., f.o.b., f.o.r., etc.).
The crushing and sizing operation, whether carried out entirely at the
mine or partly at the mine and the remainder at the plant, should yield max-
imum amounts of 25–37 mm size fractions. The fraction below 5–10 mm
size needs to be agglomerated to make it useful as a blast furnace feed.
Since normally a single crusher cannot reduce the linear dimension of lump
by more than a factor of about 6–4, multistage crushing is invariably nec-
essary. For the first stage of crushing, either jaw or gyratory crushers are
used, except for sticky and clayey ores, for which double roll crushers are
commonly adopted. The first stage crushing produces a material which is
predominantly in the range of 150–250 mm (6–10 in). The second stage of
crushing, which is usually carried out using cone or gyratory crushers, may
be carried out in a single operation if the aim of the operation is to produce
material of 25–55 mm (1–2.5 in). When a smaller size is required, crushing
is carried out in stages using a closed circuit screening to ensure all the
material is in the desired size range.
FIGURE 6.1 Principal treatments of ores prior to their charging. Four alternative routes are shown. The
dotted lines show further alternative possibilities in these routes.
Screening is carried out by passing the material over a grid or grids

which allow the fines to pass through while the coarse material travels over.
The screens are made of various types and materials. Quite often the screens
vibrate with a small amplitude at a comparatively high frequency, since it
improves the screening efficiency considerably. Sometimes the screens are
heated. Screening of wet ores requires additional care; relatively coarse
material can be separated from fines by wet screening. In modern steel

works, the crushing and screening facilities are very extensive in order to
ensure a more closely graded product. The improvement in blast furnace
operation brought about by the use of sized charge is far greater than any
other single improvement.
The ore particles descend in the blast furnace stock in solid state; this
means that when the particles are nearly at the bottom of the stack they
have to withstand, in bulk form, nearly 15 m of overlying burden. The par-
ticles, therefore, should be strong enough to stand this load, and they also
experience abrasion from other particles during their descent.The ore frac-
tion that is obtained by crushing and sizing, as suitable for direct charging,
should possess adequate strength and abrasion resistance. A hard ore would
produce such a desired fraction whereas friable ore would not. The crushing
and sizing treatment is therefore feasible only if the ore is strong and hard.
6.4 Beneficiation of the Iron Ores

The evaluation of an iron ore deposit had already been discussed earlier in
section 3.1. If the ore body cannot be made suitable for the blast furnace
merely by crushing and screening, then further beneficiation of the ore
becomes necessary to obtain a proper quality of blast furnace feed.
For beneficiation to be efficient, the iron minerals must by physically
liberated from the other components of the ore so that processes that are
employed can segregate them into an economically useful concentrate. The
phenomenon of liberation is achieved by comminution (i.e., crushing and
grinding). Both of the these processes break down the ore into even smaller
fragments. The extent to which the comminution is carried out is dictated
by the relative cost of comminution and the percent liberation as deter-
mined by its possible recovery. The structure, characteristics, and proper-
ties of the mineral deposit can have an important bearing on the method
of beneficiation which is adopted and on the economy achieved. Properties
like magnetic susceptibility, grain size and shape, density, electrical con-
ductivity, etc. are important in this respect. Each ore must be judged not
only in terms of its inherent qualities but also in relation to local conditions
(e.g. availability of water, skilled labor, etc.) and the distance it will have to
be transported to reach the blast furnace. Beneficiation is determined step-
by-step in order to ascertain the economics involved. Further processing is
considered only if the prior treatment is economical.
In these treatments, the low grade ore has to be first crushed and
ground to liberate the iron mineral from the gangue at the optimum cost.
The separation of iron mineral from the gangue can then be achieved by
any one or more of the following methods:
1. Magnetic separation at either high or low intensity magnetic fields;
2. Gravity separation using either a heavy fluid medium or by jigging,
tabling or the use of spirals;
3. Flotation;
4. Electrostatic separation;
5. Magnetic roasting followed by a low intensity magnetic separation;
6. Although not the goal, some degree of concentration is achieved by
washing and calcination.
Magnetic Concentration
Being strongly magnetic, magnetite (Fe3O4) can be readily separated at
low intensity magnetic fields (500–200 oersteds). Weakly magnetic mate-
rials like hematite can be separated at high intensity (12000–22000 oer-
steds) magnetic fields only. The dry method is suitable if the particle size
is + 6 mm and the wet method is effective if the size is less than 150 mesh
(100 μm). For the intermediate size range either method is equally suited.
Usually the ore is fed on to a rotating drum with a partially magnetized
surface. The non-magnetic tailings fall from the drum before the concen-
trate which is carried further under the influence of the magnetic field.
A series of magnetic separators in combination with successive crushing
and grinding operations is usually adopted for effective beneficiation.
Occasionally, other concentration operations may be used in conjunction
with magnetic separation to improve the quality of the final product. The
actual sequence of operations depends on the quality of the feed and that
of the product required.
Gravity Methods
1. Heavy Media. This method is normally used for separation when par-
ticle size is around 6–40 mm (¼–1½ in). An artificial heavy medium is pre-
pared by having ferro-silicon, sand or magnetite in suspension in water so
that the density is between those of iron oxide and the gangue. The iron
oxide concentrate sinks and the gangue overflows as tailings. A variety of

container designs are used for this process.
2. Jigging. In this method, a bed of ore particles, placed on a perforated
bottom, is subjected to periodic impulses of water flow through the perfora-
tions. The heavier particles move downwards and are removed as the con-
centrate. Since the weight of a particle depends on its density, it is necessary
for the feed to be sized within close limits. The output of an individual jig
is rather low.
3. Tabling. This method is suitable for sandy feed. The feed moves over
an inclined shaking table and is washed with a cross-stream of water. The
capital and operating costs are high and the output of an individual table is
small (1–2 t/hr).
4. Spirals. This method, which is best suited for materials of 1.5–0.1 mm
(10–150 mesh), is increasingly being used for upgradation of hematite ores.
The ore is washed down a spiral launder with a curved bottom. The iron
mineral, being heavier, moves to the bottom of the curved track while the
higher tailings cling towards the outer rim. Since the output of an individual
spiral is very low, a series of spirals are necessary. At the Carol concentra-
tion plant of the Iron Ore company of Canada, 3456 spirals are used to
treat 17 million tons of ore producing 7 Mt of concentrate per year. The
operational and maintenance costs are quite low since no power is required
to run the spirals.
Flotation Method
This method is based on the fact that the surface property of some miner-
als can be temporarily altered to make them hydrophilic or hydro-phobic.
Hydrophobic minerals can be floated and separated as froth if air is bub-
bled through a suspension of such minerals in water. Various reagents are
added to obtain the desired surface characteristics of the metallic mineral
or the gangue. Petroleum sulfonates, fatty acides, tall oil, etc. are used to
float iron oxide. Starch, gum, etc. tend to float silica and depress iron oxide.
Even though the flotation reagents are required in small amounts, they
are costly chemicals. The overall cost of chemicals required in floatation is
high. Relatively more power is required to run the flotation cells. The feed
to flotation cell should be finely ground. All these factors in general make
flotation uneconomical for upgrading iron ores. Flotation has been used
in conjunction with other concentration operations in exceptional cases to
ensure high grade product.
Electrostatic Separation
The difference in electrical conductivity between iron minerals and the
gangue is used to separate the two. Electrostatic separation is mainly used
in the final stages of concentration in conjunction with some other upgrada-
tion method. In conjunction with magnetic separation, the use of electro-
static separation in the final stages improves the purity of the product and
overall efficiency of concentration.
Magnetic Roasting
It is very easy to reduce hematite to magnetite under slightly reducing con-
ditions at 600°–800° C temperature and the ore is there by rendered ame-
nable to magnetic separation at low intensity. Magnetic roasting in rotary
kilns was practiced in the U.S.S.R. and in fluo-solid reactors in Italy.
Washing
The terms washing means separation of particles of widely different sizes,
on the basis of their sizes, using a fluid current in which the finer particles
are readily suspended and washed away while the larger particles fail to
get suspended and are separated from the fines. Washing is essentially a
classification process and concentration that may occur in the process is
incidental. This term should be differentiated from the washing of coal
wherein upgradation is primarily sought along with freeing it from the
extraneous clayey matter. Washing was adopted for obtaining lumpy ore
fraction from a moist clayey ore which otherwise is not amenable to dry
screening. Washing improves screenability markedly by removing the
finer clayey particles which confer plasticity onto the ore. If these ores are
not washed, the ore handling difficulties during crushing and screening,
are aggravated.
Drying
Many blast furnace operators preferred to charge wet ores directly in the
furnace with the hope that the sensible heat of the outgoing gases would
be best utilized in drying the wet ores. Such a furnace practice proved
quite detrimental to the furnace productivity. Wet ores cannot be prop-
erly screened to eliminate fines which enter the furnace. Even if such
ore charge gets dried by the outgoing gases, the fines associated with the
ore charge have adverse effects on the furnace productivity. It has been
observed that increasing the proportion of dried ore in an otherwise wet
charge has resulted in decreasing the coke rate. The economy is chiefly
attributed to the better screening of dry ores and its consequent improved
distribution in the burden. Wet ores are sticky and interfere with the screen-
ing. They also cause trouble in transit by sticking in the hopper, chutes, etc.
Moreover, they stick to the charging bell and thereby interferes with distri-
bution. Bashforth has described ore drying practices in the U.K. The driers
employed are generally rotary type and are heated by burning blast furnace
gas. These kilns are generally inclined at about 5 degrees to the horizontal.
Over-drying of the ore has been found to result in increased disintegration
of the ore particles into fines and dust which has to be screened off. Since
an increasing proportion of ores are being beneficiated in wet conditions,
dewatering operation is included as a part of the beneficiation flowsheet.
Ore drying was by no means a universal practice. In fact, only some plants
adopted it.
6.6 Calcining/Roasting
Iron is ultimately produced by reducing iron oxide present in the ore.
If local considerations dictate the use of hydroxide, carbonate or sulfide
type of ores these can be used for ironmaking only after calcination/roast-
ing. Normally, a good amount of coal matter and or sulfur contained in
the ore also gets eliminated during calcining. Bashforth32 has described
the ore calcination practices in the U.K: at one plant a Davies Colby kiln
of 24 feet in diameter and 45 feet tall was employed to deal with 300 t of
carbonate ore per day. In the past calcining was fairly widely practiced
where carbonate and or hydroxide ores were the only source available
for ironmaking. It has been completely replaced by sintering or pellet-
izing since calcining can be achieved readily and more effectively in the
latter processes, and these can also produce a product directly useful
for charging.
6.7 Blending
Blending means mixing up of two or more types of materials from two or
more sources to even out variations in physical and or chemical qualities
to obtain a more uniform material of desired qualities over an extended
period. The iron ore blend may be made up of ores from different sources
or it may also include materials much as coke breeze, flue dust, limestone,
etc. In the past, materials were stored in separate bunkers and appropriate
quantities were withdrawn from each and mixed to give the desired mix-
ture. A more satisfactory method, which is now widely adopted, is bedding.
The materials in appropriate ratios are laid as layers, one over the other, on
long beds, which have a triangular or trapezoidal cross section, by means
of a traveling reversing boon stacker or gantry. The blend is claimed in a
direction perpendicular to the direction of laying down the bed. A mini-
mum of two beds are employed, one being laid while the other is being
reclaimed. In a modern plant, ore is being laid at the rate of about 1000 t/
hr and reclaimed at about 500 t/hr.
6.8 Shipping
The large scale rise in iron output in the industrially advanced countries,
particularly after the World War II, was made possible through the import
of a substantial amount of iron ore from other countries. Iron ores are
now transported from the main iron ore producing countries like Brazil,
Australia, India, Chile, etc. to countries like Japan, U.S.A., Germany, etc.
in vessels of up to 200,000 t dwt capacity. The ports need to possess cor-
responding capacities to admit such large vessels and efficiently handle
materials of this order of magnitude. Blending may be carried out at the
exporting or the importing ports. Ore can be laid down and withdrawn
from modern stockyards at the rate of several thousand tons per hour.
Loading and unloading is carried out by adjustable loading arms and con-
veyor belt-systems.
Automatic sampling is a common practice. A small proportion of the
ore is continuously removed, crushed, ground, and a sample is drawn from
this for chemical analysis. The price of the ore depends on its size, composi-
tion, and pre-treatment, if any, done before loading.
It has already been indicated earlier that the iron ore, metallurgical
coal and coke trade is fast expanding on the international level. It is not
uncommon for steel plants to buy or lease mines in other countries. The
transportation in is carried out only via water ways, because of their rela-
tive cheapness. International trade involves lot of problems in ensuring the
right quality supply as agreed upon. Several independent, third party agen-
cies have emerged to ensure this for both the seller and the buyer. Steel
plants now have their own jetty ports, ships or carriers, and attendant varied
services as well as elaborate material handling systems from the ports to the
plant stock yard.
6.9 Products of Ore Processing

The products after ore treatment may be of the following types:
(a) Sized lumps suitable for direct charging in the furnace;
(b) Ore fines and slimes after screening the lumps or produced deliber-
ately for subsequent agglomeration;
(c) Upgraded or concentrate, usually in the form of fines, for agglomera-
tion (as fine as 80%–350% mesh).
All other products of ore processing, except the sized lumps, are in the form
of fines which must be converted into an agglomerate before charging in a
blast furnace. This forms the subject matter of the next two chapters on the
topic of burden preparation.
6.10 Treatment of Iron Ores (India)

Until the beginning of the Second Five Year Plan (1955), in India, high
grade iron ores were selectively hand-mined and sorted for smelting as well
as export. With the increase in demand, dependent on the setting up of new
steel plants and allowing for expansion of the old ones, mechanized min-
ing on an increasing scale had to be adopted. As a result, the run-of-mine
became less uniform in its size and composition. The fraction of fines in the
run-of-mine increased substantially.
The Indian iron ores are generally rich (over 58% Fe) and high in insol-
ubles (silica and alumina contents). The occurrence of blue dust as assaying
over 68% Fe along with the hard ore is a common feature of the iron ore
mines in India. Because of the soft nature of iron ores, mechanized mining
and the presence of blue dust, the run-of-mine contains as high as 50–60%
fines (–10 mm). These fines, particularly those of lateritic nature, occurring
in the upper mining horizons, are of inferior quality, except in Goa where
the fines are of better grade than lumps. They are not chemically suitable,
apart from size, for smelting indigenously. Due to the association of iron-
bearing minerals, more with laterite and clayey gangue than with siliceous
minerals, the alumina to silica ratio is usually greater than one. These ores
are practically free from bases, i.e., CaO, MgO, etc.
Since the predominant gangue constituents associated with the Indian
ores are laterite and clay, these make the ores rather soft and sticky in nature
and are chiefly responsible for all the problems encountered in crushing,
screening, and handling of the ore at the mine site or at the ports. The situ-
ation is worse particularly in the wet months when the moisture level goes
up to 7–10%. For effective dry screening, the moisture level must be below
4%. The ores, therefore, need to be dried in wet months. Alternatively,
it has been shown that scrubbing followed by wet screening is absolutely
necessary to solve the ore screening and handling problems encountered in
almost all the iron ore mines in India employing mechanized mining.
Since the liberation of the bulk of the gangue from the iron-bearing
minerals take place at a relatively coarse size, and since these ores are rela-
tively rich in iron content, the ore treatment is rather simple and inexpen-
sive. It consists of the following steps:
1. Crushing the mine output of the required top-size;
2. If necessary, scrubbing the crushed ore to release the adhering clay;
3. Wet screening the crushed or the scrubbed ore to separate lumps from
objectionable fines;
4. Classification of the screen undersize to separate ore from slimes and
dewatering of the product;
5. Concentration of the resulting ore fines prior to agglomeration to
separate the iron bearing minerals from the gangue either by jigging,
heavy media separation, or by cyclone treatment.
Iron ore treatments carried out in India are shown in Table 6.2.
TABLE 6.2 Ore Processing Schemes of Some Major Iron Ores in India
Mine Process adopted
1. Bailadila - XIV Wet screening and classification

2. Bailadila - V –
3. Daitari –
4. Donimalai –
5. Kundremukh Magnetic separation concentration
6. Rajhara Dry screening only
7. Kiriburu Scrubbing and wet screening
8. Gua Dry screening only
9. Dalli Dry screening as well as scrubbing and wet screening
10. Barsua Dry screening as well as washing and gravity concentration
Mine Process adopted
11. Bolani Dry screening only

12. Meghataburu Dry screening as well as Wet screening and classification
13. Noamundi Dry screening as well as wet scrubbing, screening and
classification
Ore Treatment at Tata Steel

Tata Steel has done some pioneering work in this direction. After elaborate
trials, they have set up the ore treatment plant at their Noamundi mines
with a throughput capacity of nearly 1200 t/hr in two circuits. The hard, mas-
sive ore is subject to scrubbing and wet screening whereas the friable ore
and the blue dust are treated dry, hence the name dry-circuit. The flowsheet
of the wet circuit is shown in Figure 6.2. It is designed to produce a washed
ore of 2.5 million t per year in the form of 1.17 million t as - 50 +10 mm (–2
+ 3/8 in) and 1.33 million t of fines of – 10 + 100 mesh (– 3/8 + 100 mesh)
size. The facilities are comprised of primary crushing, secondary crushing,
scrubbing, water clarification and disposal, overland conveyors, storage and
reclamation systems, rail loading, weighing, make-up water systems, etc.
FIGURE 6.2. Flow-sheet of ore washing facilities at Tata Steel’s Noamundi iron ore mine.
Under Indian conditions, the washing of ores generally leads to lumpy

ore fractions of 61.5–68.5% Fe with a 60–90% yield. The upgraded fines-
fraction-feed from slimes analyses 63.5–66.0% Fe. The rejected tailings
amount to about 4.11% by weight and which carry with it 12–60% silica,
11–30% alumina and 2–10% iron are present in the ore. The ratio of alu-
mina to silica in lumps increases from 1.08–1.66 to 1.46 –1.84 after washing.
This only means that more silica is removed during washing. The actual
contents of both silica and alumina are decreased. The ratio produced is
adverse to blast furnace operation, but it can always be adjusted by blend-
ing it with a suitable silica rich ore.
The lumpy ore obtained after washing is ideal for charging in the blast fur-
nace or for export whereas the upgraded fines constitute an excellent material
for sintering. The iron content of the sinter is higher by about 4% than that
made from fines from dry screening. This treatment leads to improvement in
the overall economics of mining and conservation of Indian resources.
Dry screening was only recently adopted in all iron ore mines in India.
The difficulties encountered in dry screening have now been fully appreci-
ated by all the iron ore producers. The washing of ores, more or less along
the same lines as developed by Tata Steel, is increasingly being adopted.
Scrubbing and wet screening facilities can be easily incorporated in the
existing crushing and dry screening circuits at the mines without much of
additional cost. In fact, in the Kiriburu and Barsua iron ore mines such
modifications have already been added.
Considering the relevant factors, the maximum top sizes of the ore for
Indian blast furnaces is 50–55 mm and the lowest is 10 mm. The size should
be as uniform as is economically feasible. The major portion is expected to
be in the range of 25–50 mm. Sieve analysis of the ores in some of the steel
plants in India are shown in Table 6.3.
TABLE 6.3 Sieve Analyses of Iron Ores in India Steel Plants

Sieve analysis %
Plant + 80 –80 –60 –40 –25 –12 –5 all mm
+ 60 + 40 + 25 + 12 +5
Bokaro − − − 21 71 8 −
Bhilai 9 22 23 11 17 10 7
Rourkela 8 8 16 18 40 5 5
Barsua
Kalta − 2 31 30 31 3 3
Alumina Problem and Beneficiation

Alumina is a refractory oxide, and in a blast furnace it increases the liquid tem-
perature of the slag. Consequently, the furnace operating temperature has to
be increased to run the furnace smoothly. Coke rate in this case goes up. If
alumina is less (i.e. less than 2%) many of these problems are eliminated.
Indian iron ores are high in alumina and silica and have adverse ratio
of alumina to silica. Fortunately, crushing, screening, and washing leads to
obtaining lumps of iron ores that are acceptable as direct blast furnace feed.
The –10 mm fines are inferior because of the increased alumina content.
This happens because the alumina is associated more with the laterite and
friable portion of the ore. These fines, if sintered, then make the sinter is
inferior. Against the world’s average of less than 1.0% alumina, the Indian
sinter would contain 3.5–4.5% alumina. This will be clear from the data
given in Table 6.4. These ore fines are therefore beneficiated using gravity
separation like jigging.
It has been shown that a concentrate of ore fines with less than 2% alu-
mina at a recovery of nearly 80% is possible.
TABLE 6.4 The Variation of Insolubles as Influenced by Processing of Iron

Ores at Noamundi Mines
Yield Fe Percentage Al2O3
SiO2
Run-of-mine ore 100 64–65 1.2–1.4 3.0–3.2
Crushing
Screening
Wt scrubbing in sized ore (−40 + 53.4 66.5–67.0 0.7– 0.9 1.7–1.9
washing drums 10 mm)
Classifier Rake up fines 32.9 63.5–64.5 1.0–1.5 3.3–4.0
Fines −10 + 100 mesh
Slimes (–100 mesh) 13.5 56–59 3.75–550 6–9
Treatment of Goa Ores

The iron ores of Goa are essentially fine or friable in nature. It can not be
directly fed to blast furnaces. The ores, although rich in iron content, are
still processed to upgrade them by washing and cyclone treatment. The ore
fines are then pelletized, after required grinding, and are exported mainly
to Japan. M/s. Choughule and Co. erected one pelletizing plant to produce
0.5 million t of pellets per year in 1967 and another pelletizing plant was
added to produce an additional 6 million t of pellets per year. This also
includes beneficiation of the ore.
Kundremukh Iron Ore Project

The largest material handling and beneficiation plant in India was created to
treat the magnetite deposit of Kundremukh in Karnataka. The process of ben-
eficiation essentially consists of crushing, to the liberation size of –150 microns,
followed by spiral classification, and magnetic concentration. It has 728 rougher,
336 cleaner, and 336 recleaner spirals and 39 magnetic separators. The slurry
of concentrate containing 60% solid and 40% water is transported over a dis-
tance of 66 km through a pipeline. It is filtered and the cake thus produced is
loaded at the rate of up to 600 t/hr in vessels of up to 60,000 t dwt capacity at
the Manglore port. The simplified flow-sheet is shown in Figure 6.3.
It is designed to treat 22.6 million t of the run-of-mine per year to yield
7.5 million t of the concentrate and 13 million t of the tailings to be disposed
of per year. The product was to be sold, under a 20 year contact started in
1980, to the National Iranian Steel Industries Co. to make up a total order
for the supply of 150 million t of the concentrate. The product was to be
entirely below 100 mesh with at least 60% below 325 mesh. It would analyze
at 66.5% Fe, silica and alumina together nearly 4%, sulfur less than 0.05%,
and phosphorus below 0.03%, with a maximum of 9% moisture. The entire
project was to cost nearly 800 crores of rupees in the 1980s. The concentrate
was to be sold at Rs. 198 per ton as against rupees 75–80 per ton of the ore.
FIGURE 6.3 Simplified flow-sheet of beneficiation plant at Kundremukh for the treatment of magnetite
deposit.
6.11 Agglomeration of Iron Ores

Iron ore fines obtained by screening of good grade ores, or after beneficia-
tion of lean ores, or collection of flue dust or pyrite residue from sulfur pro-
duction units and the like, need to be agglomerated into lumps of suitable
physical characteristics for rendering them useful as a blast furnace feed.
The processes that have been used can be classified as follows: briquetting,
nodulizing (rotary kiln sintering), vacuum extrusion, sintering (on grate),
and pelletizing.
The development of these processes, to the extent of their adoption in
commercial practice, became possible only during the World War II, when
a general shortage of ore made it imperative to use all available raw mate-
rials. Out of the processes listed above, sintering on grate and pelletizing
are widely used in preference to others because of their technical and eco-
nomic suitability for agglomeration of iron ore fines on a large scale. These
two processes are therefore discussed in detail while the other processes
are only mentioned briefly.
Briquetting
Briquetting essentially consists of pressing the ore fines, with or without a
binder, into a block or briquette of some suitable size and shape, and then
subjecting it to a hardening process. A wide range of organic and inorganic
binders like tar, pitch, cereal products, sodium silicate, ferrous sulfate,
magnesium chloride, limestone, cement, bentonite, etc. have been tried
with varying success. Although briquetting was used commercially during
and after the World War II, it has practically been abandoned since 1960.
However, more recently, there has been a revival of interest in briquet-
ting. On a small scale, briquetting does offer some advantages over other
processes.
In the early version of the process, fine ore mixed with water was
pressed into oblong blocks which were hardened in a tunnel kiln heated to
about 1350º C.
Cement briquettes with 1–5% cement as binder were produced in
the U.S.A. during the World War II on a large scale. Individual plant pro-
duction was of the order of 1000 t/day. Recently briquettes of adequate
strength have been produced in cold conditions (without firing) by adding
10% cement and aging them for several days like cement concrete.
Briquettes were produced in Germany during the World War II using

nearly 3% iron borings, 8–10% coke breeze and 3–5% magnesium chloride
as additives. These were fired after at least two days of aging. Recently, hot
briquetting has also been developed in the U.S.A. by pressing ore fines at
800–1050º C under normal or slightly reducing conditions. In the U.S.S.R.,
hardened briquettes were produced by using hydrated lime and molasses as
binders and hardening was carried out at about 300º C in an atmosphere of
carbon dioxide. Similar processes have also been developed in the U.S.A.,
but are yet to be adopted in practice.
Briquetting has been quite popular for producing hot-briquetted-iron
or HBI. It is also used commercially for producing agglomerate from DRI
fines generated during handling so that these could be used as solid charge
in electric steelmaking.
Nodulizing
In the nodulizing process, flue dust, pyrite residue, or fine ore concentrate
along with some carbonaceous material like tar are passed through a rotary
kiln heated by gas or oil. The feed travels countercurrent to the gases. The
temperature inside the kiln is just sufficient to soften the ore but not high
enough to fuse the ore. The kiln is inclined slightly to the horizontal and
rotates at 1–2 rpm and is nearly 30–60 m in length. The diameter at the
feed end is about 2 m whereas that at the sinter zone is about double that. It
takes about 1.5–2.0 hours for the charge to travel through the kiln. The size
of the nodules produced vary considerably depending upon the tar content
and the temperature. The nodules varied considerably in composition and
were too dense or slaggy. They were not profitable and this process could
not find great favor. It has been largely replaced by either sintering or pel-
letizing. Nodulizing kilns were first used in 1914 and continued in operation
until the 1960s.
Vacuum Extrusion
The process of vacuum extrusion otherwise used in the ceramic industry,
was adopted for producing iron ore compacts in the 1950s on a small
scale. Moist ore, with or without bentonite as a binder, is fed in the de-
airing chamber and extruded into a cylindrical product which is cut into
desired small sizes. The product is dried and fired before use in a blast
furnace.
Sintering
Sintering is essentially a process of heating a mass of fine particles to the
stage of incipient fusion (temperature a little below the melting or soft-
ening point) for the purpose of agglomerating them into lumps. Metal
powder compacts are sintered to obtain the desired solid shapes. In iron
ore sintering, the aim is to produce a strong, but porous, agglomerate
from a sandy, uncompacted mass. Compacting of ore fines prior to sin-
tering, as is done for powder metallurgical articles, is therefore unneces-
sary. Similarly, heating the entire bed of fine particles to the temperature
required for sintering, as is done in powder metallurgy, is neither practi-
cal nor economical. Sintering is carried out layer by layer in a mass of
charge. The process of sintering will be discussed at length in the next
chapter.
Pelletizing
The process of pelletization consists of rolling moist iron ore fines of less
than 100 mesh size, with or without a binder, into balls of usually 7–20 mm
in size. Since the moist balls lack the necessary strength, these are dried,
preheated and fired, to produce hardened balls which are an excellent
feed for a blast furnace. This process will also be discussed at length in
Chapter 8.
6.12 World Resources (2017 USGS):

U.S. resources are estimated to be 110 billion tons of iron ore containing
about 27 billion tons of iron. U.S.
resources are mainly low-grade taconite-type ores from the Lake Superior
district that require beneficiation and agglomeration prior to commercial
use. World resources are estimated to be greater than 800 billion tons of
crude ore containing more than 230 billion tons of iron (see Table 6.5:
Source: U.S. Census Bureau).
Substitutes
The only source of primary iron is iron ore, used directly as direct-shipping
ore or converted to briquettes, concentrates, DRI, iron nuggets, pellets,
or sinter. DRI, iron nuggets, and scrap are extensively used for steelmak-
ing in electric arc furnaces and in iron and steel foundries. Technological
advancements have been made, which allow hematite to be recovered from
tailings basins and pelletized.
TABLE 6.5 U.S. Exports of Iron Ore by country and type1, 2.

(Exclusive of ore containing 20% or more manganese)
2016 2017
March January–March
Country and 3
January–March Quantity Value Quantity Value3 Value3
type of product Quantity (thousand (thousand (thousand (thousand (dollars
(thousand metric tons) dollars) metric tons) dollars) per ton)
metric tons)
Canada 1,030 214 14,000 1,020 67,700 66.59
China – 30 2,450 30 2,450 81.57
Japan – 144 8,930 144 8,930 61.99
Other 1 (4) 65 62 5,530 88.61
Total 1,030 388 25,500 1,250 84,600 67.51
Concentrates (4) (4) 10 4 539 134.75
Fine ores (4) (4) 22 (4) 41 10.25
Pellets 1,030 357 22,900 1,180 79,000 66.80
Other (4) 30 2,480 68 5,090 74.78
Total 1,030 388 25,500 1,250 84,600 67.53
– Zero.
1
Data are rounded to no more than three significant digits; may not add to totals shown.
2
All countries with less than 5,000 metric tons of exports per month included in “Other”.
3
Free alongside ship (FAS) value.
4
Less than ½ Unit.
Source: U.S. Census Bureau.

An efficient blast furnace operation requires that the A12O3/SiO2 ratio be
kept at a low value. For achieving optimum levels of productivity, the iron
content of acidic ores should preferably be above 60%, the A12O3/Fe ratio
less than 0.05, and A12O2/SiO2 ratio less than 1. Merely washing the ores, as
is currently being done, is not sufficient. The use of polymeric dispersant
additives as developed by RRL Jorhat, for effective elimination of alumina
and silica, has been recommended to improve the Indian ores. The fines,
the concentrates, etc. of iron ore can be agglomerated into lumps, on a
large scale, by either sintering or pelletization.
CHAPTER
7
BURDEN PREPARATION:
AGGLOMERATION AND
SINTERING
In This Chapter
● Principle of Sintering
● Process Variables
● Sinter Quality
● Mechanism of Sintering
● Raw Material Requirements
● Recent Trends in Sintering Practice
● Alumina Problem in Sintering
7.1 Principle of Sintering

Iron ore sintering, in essence, is carried out by putting a mixture of iron-bear-
ing fines mixed with solid fuel on a permeable grate. Since coke breeze is nor-
mally available as a waste product in an integrated iron and steel plant (from
coke ovens), it is universally incorporated as a solid fuel in the sinter mix.
Coke breeze is the best choice among the available solid fuels. The top layer
of this sinter bed is heated to the sintering temperature (1200–1300 degrees
C) by a gas or oil burner. Air is drawn downwards, through the grate, with the
help of exhaust blowers connected, from underneath, to the grate.
FIGURE 7.1 A sinter bed a few minutes after the ignition of the top layer.
The narrow combustion zone developed initially at the top layer travels
through the bed raising the temperature of the bed, layer by layer, to the
sintering level. An illustration of a sinter bed a few minutes after the igni-
tion of the top layer is shown in Figure 7.1. The cold blast drawn through
the bed cools the already sintered layer and becomes heated. The heat
contained in the blast is utilized in drying and preheating the lower layers
in the bed. Before combustion, each layer is dried and pre-heated by the
heat transferred from the upper combustion zones. Much of the heat in the
gases is absorbed by the lower portion of the bed.
In the combustion zone, bonding takes place between the grains and a
strong and porous aggregate is formed as shown in Figure 7.2. The process
is completed when the combustion zone has reached the lowest layer of the
bed. The sinter cake is then tipped from the grate in hot condition or after
partial cooling. It is broken, screened, and cooled to produce desired frac-
tion. The undersize is recycled and the oversize is fully cooled and sent to
the blast furnace.
FIGURE 7.2 Sinter cake.

Burden Preparation: Agglomeration and Sintering • 121
The process described above is known as down-draught sintering since

the air blast is drawn through the sinter bed downwards. In contrast to this
process, some non-ferrous ores are sintered by up-draught sintering. In up-
draught sintering, the initial small layer of the charge on the grate is ignited
before making up the remaining bed thickness. Iron ores are, however, sin-
tered by down-draught sintering alone.
The earliest form of sintering, the Huntington-Heberlein pot process,
was developed in the 1890s for non-ferrous ores; a later modification was
the Greenwalt sintering machine. In the first decade of the present century,
Dwight and Lloyd in Mexico developed the continuous sintering process
for non-ferrous ores. Soon it was adopted for iron ore sintering. Today the
Dwight-Lloyd sintering machine is the only large scale sintering process in
use both for ferrous and non-ferrous ores.
7.2 Process Variables

During sintering, the aggregate of particles constituting the sinter charge is
heated layer by layer to the sintering temperature and also similarly cooled
within a matter of minutes (10–15). The temperature cycle which any
layer of the charge has to pass through is qualitatively and ideally shown in
Figure 7.3. The particles reach the peak temperature of incipient fusion,
so that, at the point of contact of particles, with each other, fusion bond-
ing takes place. This results in sufficiently stronger bonding between the
particles, and as a result, the fine particle-aggregate becomes a somewhat
porous cake or agglomerate. Bonding of grains is, therefore, a high tem-
perature phenomenon. It is a function of temperature and time for which
the mass is maintained at the high temperature level. In other words, the
area under the time-temperature curves as shown in Figure 7.3. essentially
determines the nature and strength of the bonds developed during sinter-
ing of a given mix. For a given mix, it is most unlikely that the bonds of
sufficient strength will be formed below a certain temperature level within
a reasonably short time. Hence, the area under the curve above a certain
temperature, which may be around 1000º C for iron ores, is the effective
factor in deciding the extent of sintering, rather than the whole area under
the curve from room temperature to the combustion temperature level.
The nature of the time-temperature graph will depend upon the rate of
heating and cooling of a given mix. The nature of this graph is of paramount
importance in assessing the sintering response. The factors that affect this
curve are then the variables of the process and should be adjusted properly
for effective sintering.
FIGURE 7.3 Temperature cycle during sintering of iron ore fines.
The variables of the sintering process are as follows:

1. Bed permeability as decided by the particle size and shapes of the mix.
2. Thickness of the bed.
3. Total volume of air blast drawn through the bed for its sintering.
4. Rate of blast drawn through the bed during sintering.
5. Amount and quality of solid fuel incorporated in the charge.
6. Amount and type of carbonates present in the charge.
7. Amount of moisture in the charge.
8. Nature of ore fines, (e.g. its chemical composition).
9. Any non-uniformity in the bed composition or in the process of sintering.
During sintering, heat exchange takes place between the solid charge and
the air blast drawn through the bed. At any time during sintering, the air
blast initially gets heated, that is, it cools the combustion zone, and in turn
heats the lower layers of the bed. It is, therefore, essentially a phenomenon
of gas-solid heat exchange. The process is carried out at such a fast rate that
the system is far from steady state conditions of gas-; solid heat exchange.
In order to carry both the heating and cooling functions of the gas phase
effectively (i.e. to obtain faster rate of heat exchange), the blast drawn
through the bed should be at maximum heat capacity. This means that
the volume of air drawn through the bed during sintering should be maxi-
mum. The more permeable is the bed, the more the blast will be drawn
through it. A more permeable bed, however, leads to loss of strength in the
resulting sinter. These two opposing factors should be adjusted to produce
optimum results.
It is usual practice to draw air blasts of 7003/t (25000 Nft3/t) for non-
volatile ores and up to 1100 m3/t (40000 Nft3) for volatile low grade ores
for sintering them effectively. This means that blast of total heat capacity is
almost equal to that of the solid charge is necessary to transfer the hot zone
through the bed and an equivalent blower size will have to be provided for
effective sintering. The amount of blast drawn per unit weight of the charge
(usually one ton) is often referred to as specific volume of the blast in sin-
tering. Heat flow studies during sintering have shown that the product of
specific volume and heat capacity of gaseous phase drawn through the bed
is normally constant.
Since sintering takes place layer by layer, it is obvious that every layer
will have to be heated to the sintering temperature level. This is possible
only if solid fuel is incorporated in the charge and that it burns layer by
layer. If no fuel is incorporated in the charge, and even if the top layer is
initially heated to the sintering temperature level, the peak temperature
attained by the lower layers will go on decreasing and sintering will not take
place uniformly and effectively.
Sintering is commenced by igniting the solid fuel in the top layer. The
combustion of solid fuel raises the temperature of the top layer to the
required level. The air blast drawn thereafter through the bed not only
cools the top layer but preheats the solid fuel in the next lower layer and
the combustion zone shifts downwards. A high sintering temperature in
the lower layers can be achieved by preheating the fuel to high tempera-
ture, by achieving combustion in a gas stream at high temperature and by
obtaining maximum amount of carbon to burn to carbon dioxide. The rate
of fuel combustion (i.e. the rate of travel of flame front) depends on the
reactivity of the fuel and the oxygen content of the gas. The heat generated
by combustion in the upper layers must travel down the bed at the same
rate as the combustion zone and be available at the right time for raising
the temperature of the lower layers. The heat wave travels down in the bed
with approximately a constant velocity given by an expression:
HgW
Hs(1 − F)
where Hg is the heat capacity of gas per unit volume, and Hs is the heat capacity
of solid per unit volume.
W is the normal volume of fluid per unit cross section of the bed
per minute,and F is the void fraction per unit volume of bed.
The time required for the waste gases to reach peak temperature
should, therefore, be inversely proportional to the product of apparent gas
velocity and heat capacity of the gas phase.
For any fuel type, there is a certain partial pressure of oxygen which will
give a combustion rate such that the flame front travels with the heat front.
This is often referred to as matching and it gives maximum bed tempera-
ture. Any other condition is called mismatching. It gives broad temperature
distribution not conducive to effective sintering. For more reactive fuels,
matching occurs at low oxygen partial pressures in the gas and vice versa.
It is a common thermodynamic phenomenon that an increase in partial
pressure of oxygen gives rise to an increase in combustion temperature as
is found during ignition and during the matched conditions of sintering. It
may not be so when the conditions are mismatched.
In spite of the fact that excess oxygen is always present in the sintering
atmosphere, all the carbon in the fuel does not burn to completion (to car-
bon dioxide) and some carbon monoxide is always present in the gas phase.
It is related to the poor reactivity of the fuel and the mismatching that exists
between flame front and the heat front. Fuels containing low temperature
volatiles are not suitable for use in sinter mix, not only because of their high
reactivities, but also because these volatiles distill out of the bed in advance
of the flame front, and interfere with the gas-solid heat exchange. For effec-
tive sintering, once the top layer is ignited, adequately matching conditions
should exist. The permeability of the bed, as decided by the particle size of
the bed, amount of solid fuel, moisture and carbonates incorporated in the
charge, and the suction applied will have to be interrelated to obtain the
matching conditions.
If suction is increased to draw more blast, transfer of heat between
the gas and the solid may become less efficient. On the other hand, if suc-
tion is less, the flame front will not move down the bed properly. In either
case mismatching occurs. Increase in the average size of the particles of a
sinter mix increase permeability, and the amount of air drawn through the
bed consequently increases. This leads to inefficient heat transfer and the
attendant troubles.
There seems to be no apparent relationship between specific volume and
the solid fuel content of the sinter mix. The specific volume is always dictated
by heat transfer considerations rather than by fuel combustion. With the cor-
rect fuel content in the bed, increasing the amount of moisture or carbon-
ates in the bed increases the specific air volume. The effect of water is more
predominant. This is one of the reasons for addition of a critical amount of
moisture in the sinter mix. A significant amount of sulfur can be removed
by sintering, and this has the advantage of reducing the sulfur load in the
blast furnace. It has been observed that the optimum coke content for sul-
fur removal is much greater at the bottom of the bed, in order to achieve
the correct temperature for sulfur removal. The considerations for obtaining
adequate strength of sinter are different from those necessary for reasonable
removal of sulfur.
The sintering of iron ore fines is now universally carried out on travel-
ling grate machines forming a closed loop. The design of an endless band
of pallets moving over static wind boxes was originally set out by Dwight
and Lloyd in the beginning of the 1900s in Mexico. Although the design
has undergone improvements, essential features of the original design still
continue to be adopted even on the latest of machines. In 1958, the largest
machine in operation was 3.7 m in width and 223 m2 in grate area with a
production capacity of some 8000 t/day. The largest machine now in use is
in Japan and is nearly 8 m in width with nearly 500 m2 as useful grate area
and a production capacity of nearly 24,000 t/day.
The number of sinter strands provided in Indian steel plants are shown
in Table 7.1. Except for IISCO, all other steel plants have their own sin-
tering machines and consume together around 12 million tons of iron ore
fines. The total capacity of sintering is nearly 20 million tons.
TABLE 7.1 Details of Sinter Strands Provided in Early Years in Indian

Steel Plants.
Bhilai Durgapur Rourkela TISCO Bokaro
No. of Strands 4 2 2 2 2(1.7M)
2(in second
stage)
Width, m 2 2.5 × 30 m 2.5 2 4
length
Working length, m 25 57 50 30 63
2 50 142.5 125 60 252
Working area, m
Annual prod. capacity, Mt 2 2.1 1.2 1.26 4.2
Depth of bed mm 300 300 300 300 350
Area of cooling section m2 – – – – 60
The Dwight-Lloyd sintering machine is essentially an endless band of

pallets moving over rails, stretched across and over two huge pulleys, one
of which is driven by a motor through a reduction gear system. The raw
materials are loaded at one end of the machine and the top layer is ignited
as it immediately comes under a fixed ignition-hood. As the pallets move,
the ignited portion comes over a series of stationary wind boxes connected
to an exhaust blower. Sintering of the charge is completed by the time the
pallet travels over nearly the whole, useful length of the machine.
FIGURE 7.4 Schematic arrangement in a sinter plant.
The sintered cake drops out at the other end when the pallets turn
upside down. The cake is broken, screened, and the oversize is cooled.
The undersize usually −9 mm, is returned to the machine for re-sinter-
ing whereas the oversize, after rescreening, goes to the blast furnace as
charge. From the discharge end, the pallets return to the feed end in
inverted overhung fashion on rails. The exhaust gases from the wind boxes
are let off into the atmosphere through a chimney after dust and possible
heat extraction.
The important parts of the machine and its accessories that make up
the complete sinter plant are as follows:
1. Storage bins, mixer(s), feeder(s), etc.
2. Charge leveler
3. Ignition hood
4. Band of pallets and rails for its movement
5. Drive mechanism
6. Sinter breaker, screen, cooler, etc.
7. Spillage collector
8. Wind boxes, dust extractor, exhaust fan, chimney, etc.
The general arrangement in the plant is shown schematically in Figure 7.4.
Raw materials like iron ore fines, coke breeze, reclaimed flue dust,
etc. are generally used for sintering. These may be stored in separate over-
head bins. The weighed amount of these materials are delivered to the
mixer to prepare the feed mix. Alternatively, the raw materials may be laid
in layers to prepare a huge pile of blend which is reclaimed and delivered
to the mixer.
The proportioned raw materials are moistened and thoroughly mixed
in the mixer to prepare a desired homogeneous feed for sintering. The
mixer is a trommel or a drum fitted with paddles. After initial mixing, the
moist charge is rolled in a drum to nodulize the fines and decrease the size
range of the particles. This improves the permeability of the sinter bed.
In the latest design of the sinter strand at the Bokaro Steel Plant (India),
a two stage mixing and two-layer charging design is employed as shown
in Figure 7.5. This ready mix is laid evenly across the whole width of the
pallet with minimum compacting. It is necessary to have a continuous
uninterrupted flow of charge on the pallets when the machine runs. A
swinging spout of a roll feeder or a vibrating tray feeder is generally used
to achieve this objective.
FIGURE 7.5 Two-stage mixing and double layer charging design as adopted for sintering in the Bokaro
Steel Plant (India).
For a smooth operation, it is necessary to have uniform thickness of the

bed on the pallets. A flat plate is fitted immediately after the feeder to level
off the charge. The cut off plate is generally adjusted to give a fixed thick-
ness of the bed in the range of 300–450 mm. The importance of this charge
leveler, an apparently simple device, should not be underestimated.
The pallet has to support the mix during its travel from the feed end
to the discharge end. It must also withstand high temperatures developed
during sintering without bending or getting damaged while maintaining
an efficient seal with the stationary wind boxes. The grate bars laid at the
bottom of the pallet should provide maximum openings without allowing
excessive fine materials to be pulled through into the wind boxes during
suction. The grate bars are generally made of high chrome cast iron, and in
order to avoid distortion, three or four rows of bars per pallet are provided.
The aperture in the grate can be a maximum of 6 mm to avoid excessive
dust losses. The maximum total free area is about 20% of the total area in
order to minimize the pressure drop across the grate. The sidewalls of the
pallet have a maximum depth of about 600 mm. They taper outwards to
ensure free fall of the sinter cake. The flanged rollers are attached on each
side of the pallet to make them run smoothly on the rails. This is shown
in Figure 7.6. Since the pallets move over static wind boxes, the pallets
must always ensure leak proof seals with the wind boxes in the longitudinal
direction. When the pallet moves over the wind box and when it crosses
past the last wind box, sealing in the transverse direction is also equally
essential. In the early designs, the pallets moved on flat face on the upper
track. A lubricant was force fed in the area of contact to maintain the seal.
Such arrangements worked well on small machines.
FIGURE 7.6 Pallet roller design.
Bigger modern machines, however, need different kinds of seals. A vari-

ety of mechanized seals have been developed and used. The use of plastic
low-friction seal or armored hose have also been proposed and used. The
sintering strand in the Bokaro Steel Plant with 252 m2 area of sintering is
one of the big plants and uses an armored hose sealing arrangement.
The air leakage at the time of the pallet entering and leaving the wind
boxes, in transverse direction, is minimized by suitable mechanical arrange-
ment. There is, however, much scope for improvement in this direction.
For a good leak-proof design, the leakage of air into a new machine
should not be more than 10–20% of the total volume drawn by the blower.
During the entire life of the plant, the leakage should not exceed 25%.
The pallets move from the feed end to the discharge end on the upper
rail track. After discharging the sinter cake, they return back to the feed
end in overhung fashion on the lower rail track. The drive mechanism is
expected to ensure smooth changeover of their movements from lower to
upper rails and vice versa at the ends.
The main drive normally operates through a spur wheel and pinion
arrangement to achieve the desired speed reduction. The speed of the
early sintering strands, as available at the Bhilai Steel Plant is 1.4–4.2 m/min
for a machine of 50 m2 working area. In the latest design as the Bokaro

Steel Plant, the speed is 1.5–7.5 m/min for a 252 m2 working area. The
high torques involved make it essential to have a safety interlock to operate
and stop the machine if the pallets are not engaged properly in the driving
mechanism.
The most critical part of the machine is the discharge end, and the
major problem in its design is how to accommodate the thermal expansion
of the pallets as they move from the feed to the discharge end. Two differ-
ent solutions are in use and can be described by illustrating the designs of
the early Bhilai strand and the latest in Bokaro. These arrangements are
shown in Figure 7.7. In Bhilai, the pallets pass around the track connect-
ing the upper with the lower rails solely under the action of gravity. A gap
of nearly 60–75 mm is left in the complete set of pallets even when these
are heated to the maximum attainable temperature. This hot-gap is located
after the last suction box as shown in Figure 7.7 (a). The track at this gap is
slightly inclined for easy movement. The design, however, is subject to the
impact of the sliding pallet on the pallet which just slid before. The wear
of the curved rails is also a problem. Excessive wear of pallet ends leads to
leakage when the pallets are on the top track.
FIGURE 7.7 Facilities at the discharge end of the sintering machine.
The Bokaro design which is presently widely adopted is relatively free

of these defects. The pallets are driven by power driven sprockets. By virtue
of meshing of the pallets with the teeth of the sprockets, the ends of the
pallets do not touch each other, but a guaranteed clearance is maintained
as long as the pallets are along the curved rails. The wear of pallet ends is
practically eliminated. The teeth of the sprocket engage either with the
pallet roller or on special haunches on the pallet ends. A gap is left for
accommodating expansion of the pallets at the bottom track as shown in
Figure 7.7 (b).
If proper ignition of the top layer is not achieved, both output and qual-
ity of the sinter will suffer adversely. For each mix, proper ignition condi-
tions are established by experiments and the hood and burner arrangements
are designed accordingly. Two main types of ignition hoods as shown in
Figure 7.8 are in use:
FIGURE 7.8 Ignition hood designs. (a) Combustion-chamber type: Two burners are located on the
discharge half side of the hood and hence heating takes place partly directly and partly by the radiations
from the hot roof. (b) Nozzle or torch-type: One or more rows of burners are fitted into the hood over
the entire width of the pallet to obtain ignition time of nearly a minute, the flame directly impinges on
the charge and radiation plays a relatively minor role.
Proper ignition depends on the nature, amount and distribution of solid

fuel, in the mix. Since the cost of solid fuel can amount to nearly 50% of
the total cost of sinter production, attempts are always made to decrease
the coke breeze consumption. In a country like India, where coke is scare
and the coke breeze analysis is 25–30% ash, alternative fuels should, as far
as possible, be employed to reduce the coke breeze consumption. Sintering
has been successfully carried out using gaseous fuels in place of coke. This,
however, is yet to come to the stage of its commercial applicability. An alter-
native to this is to use an xtended hood. The main part of the hood ignites
the top layer as usual, and a series of gas fired burners are fitted in the hood
extending over a substantial length of the stand to provide additional heat.
A modified version of this arrangement is used in the Bokaro Steel Plant in
the form of a double hood. The main ignition hood maintains a temperature
of 1250º C by burning a coke oven and blast furnace gas mixture. In the
second extended chamber, a 950–1100 degree C temperature is maintained
by burning coke oven gas with excess air.
Mixed firing system not only reduces coke percentage in the mix but
improves quality and yield of the sinter. It eliminates quenching of the
ignited layer by cold air, and the disproportionately high percentage of fines
produced from the top 25 mm layer of the bed is considerably reduced. The
degree of reduction of iron oxide in the sinter cake is mainly a function of
the solid fuel content of the bed. Adoption of gaseous sintering, partly or
wholly, may reduce the production rate of the strand, but it may give rise
to a product of such chemistry that eventually requires less fuel in the blast
furnace. The development of this type of sintering is of particular interest to
India where coke, being scarce, needs to be consumed sparingly.
Because of the straight line movements of the pallets from the ignition
to the discharge end, the sinter cake is practically one whole piece. At the
discharge end, when the pallets change the direction of their movement,
one by one, the sinter cake breaks into slabs of nearly the same dimensions
as that of the pallet. These big pieces slide away from the pallet on the crash
deck and the sinter breaker. The sinter breaker has a series of prongs or
claws mounted on a shaft, and these pass between fixed breaker bars. The
lumps are broken by bending rather than by shearing. The tip of the prongs
are hardened or have hard replaceable bits bolted to it.
It is difficult and inefficient to cool the sinter before breaking. The
sinter is broken, crushed, and screened. The oversize is cooled fully so
that it can be carried to the blast furnace even on rubber belt conveyors.
The fines are cooled and can be suitably recirculated. Cooling of the over-
size can be achieved by various means (e.g. cooling on dwarfs, cooling on
sintering machine (as practiced in Bokaro Steel Plant), shallow bed cool-
ers, deep bed coolers, etc.) The screens have been specially designed for
screening hot sinter. Although breaking and screening of sinter poses no
unsurmountable technical problem, the abrasive nature and the hot con-
dition of the product creates problems in the maintenance of the system.
This makes the operation more expensive and time consuming. Better
construction materials can reduce the down time and make the operation
more economical.
The feed is expected to be laid on the pallets, and the sinter cake
is expected to slide off the pallets on the crash deck. Materials do spill
over and these must be collected and reused again. The machine is cov-
ered on the sides and a longitudinal bunker is installed below the bottom
track such that the spillage from the top track and sinter fines or lumps
sticking to the grate, even in its inverted position, can be directed to the
apron conveyor which conveys all these to the sinter screen for further
processing. In the modern strand at the Bokoro Steel Plant, an additional
mechanical device has been provided to clean up the material sticking to
the pallets while they are moving in the overhung position on the bottom
rail track.
Depending upon the thickness of the bed and quality of the feed,
ignition condition, amount and quality of fuel incorporated in the mix,
etc. an exhaust blower of suitable size is provided with the machine. On
a modern, large unit, the fan suction capacity may be of the order of
45–500 m3/sec at 120–200 cm water gauge requiring about 10,000 h.p.
motor to generate this. The waste gas system mainly includes the down-
comers, the dust hoppers, the waste gas main, the dust extractor, the
exhaust blower, and finally the chimney. The entire design should ensure
as far as possible:
1. Uniform suction across the width of the machine;
2. Collection of dust and charge particles falling through the gate in dust
hopper itself;
3. Low velocity of waste gases to keep down abrasion due to dust laden
gases;
4. Control on suction of individual wind box;

5. Minimum distortion due to thermal expansion when the machine
comes to its peak performance, by providing expansion joints (e.g. of
the bellow type).
The waste gas main may be designed either for low velocity (10 m/sec)
by making the cross-section large or for high velocity by making it small.
The latter is relatively cheaper but the abrasion problem can be serious.
It is customary to have the direction of gas exhaust from discharge end
to feed end, as shown in Figure 7.4, to ensure adequate mixing of the hotter
gases from the latter stage of sintering (discharge end) with the cooler gases
from early stage of sintering (feed end). The gas main is, therefore, tapered
in size progressively increasing from the discharge end to the feed end.
The modern, big units are provided with two fans and two exhaust
mains. Each wind box is connected to both of these mains. A symmetrical
design ensures uniform gas flow and balanced fan operation.
All the dust collecting hoppers are fitted to the gas main over the entire
length. The hoppers are fitted to the dust extracting valves which are capable
of standing suction up to 200 cm water gauge and abrasive action of the dust.
As per the pollution legislation requirements, the exhaust gases may
be cleaned before letting them off into the atmosphere through a chimney.
The gases do contain sulfur and this leads to corrosion of the parts which
come in contact with the gases. If the sulfur content is beyond the permis-
sible level, it will have to be neutralized or eliminated before passing the
gases into open atmosphere.
The efficiency of a sintering machine can be assessed in terms of the
following parameters:
1. Productivity in tons of useful sinter per square meter of the working
area per hour. The larger the figure, the more efficient the unit.
2. Effective suction as determined by the effectiveness of leak proof seals
measured in terms of length of seals per square meter working area.
The smaller is the length, the better the efficiency.
3. The quantity of air drawn through the bed per unit time. The higher
the value, the better the efficiency.
The data from Indian plants may be important to be examined in this
regard. These are given in Table 7.2.
TABLE 7.2 Indian Sintering Plants and their Performances.

No. Steel Plant Rated Sintering area Suction w.g. Bed height Sinter prod. % sinter
capacity m2 × No. of mm mm t/m2/hr in B.F.
Mt/year strands burden
1. Bokaro 4.94 252/312 × 3 1350 350 1.30 70
2. Bhilai 4.18 75 × 4 1100 300 1.32 60
3. Rourkela 1.80 125 × 2 900 527 1.20 45
4. Durgapur 1.50 140 × 2 and 945 400 1.00 35
180 × 1
5. Tata Steel 2.54 75 × 2 and 1000 and 340 1.45 65
192 × 1 1328
600 1.50
6. VSP 2.45 312 1250 400 1.35 70
7.3 Sinter Quality

The process of sintering was originally developed to merely agglomerate the
otherwise wasteful iron ore fines into a useful lumpy feed for blast furnaces.
Once the beneficial effects of sinter as a blast furnace feed were realized,
the physical properties and chemical constitution of sinter were examined
more closely from the point of view of achieving better productivity. This
led to the understanding of ideal properties of the burden inside the blast
furnace. The understanding of the ideal properties of the burden and the
possibilities of achieving these in the sinter developed hand-in-hand. The
properties of interest in sinter are:
(a) strength as measured by the tumbler index (TI), based on the frac-
tion retained on +6.3 mm screen;
(b) fines content measured as – 5 mm fraction;
(c) mean size, taking into account the fractions of all plus sizes;
(d) granulometry as measured by +10 mm fraction;
(e) reduction degradation index, (RDI) as measured by the −3.2 mm
fraction.
The details of testing of raw materials are given in Chapter 9. The objec-
tives of sintering were therefore widened and are listed below:
(i) increase the size of iron ore fines to a level acceptable to the BF;
(ii) form as strong a sinter as possible coupled with high bulk density and
maximum porosity;
(iii) obtain good reduction degradation index, (close to 25);

(iv) remove volatiles like CO2 from carbonates, H2O from hydroxides or
sulfhur from sulfhide type of ore fines along with agglomeration;
(v) incorporate flux in the sinter as a substitute to its direct addition in BF;
(vi) obtain good granulometry and less fines;
(vii) obtain narrow softening-melting temperature range.
Besides these, the quality of sinter is determined by the appropriate
ratio of alumina to silica contents, FeO content, and ferrites formed. The
first objective of sintering is the most important. The extent to which
the remaining objectives are met shall very much depend on the pro-
cess control and nature of ore fines. The removal of volatiles is only inci-
dental. The vertical sintering rate, that is the rate of travel of heat and
flame, across the depth of the sinter bed influences the sinter quality
greatly. The quality of sinter produced is judged from the properties like
RDI, granulometry, tumbler index, fines generated, etc. As the bed depth
increases, the granulometry improves. The RDI increases with increas-
ing amount of FeO, (i.e. degree of reduction during sintering). Sinter
strength is very sensitive to the alumina content of the ore fines; it should
be as low as possible.
Flux addition is beneficial from the point of view of bond formation
between particles. It, however, decreases the reducibility. Attention is
therefore focused on the extent to which flux can be incorporated in the
sinter mix without, in any way, jeopardizing other properties of the resul-
tant sinter. Incorporation of flux in the burden in this way, rather than its
addition as a separate charge material, greatly improves the blast furnace
performance. The formation of slag then becomes relatively easy, and it
forms at the correct level in the blast furnace.
Three different types of sinters are being produced depending upon
the extent of addition of flux in the mix. These are:
1. Acid Sinter. The sinter mix does not contain flux at all. Flux is added
in the furnace separately. The production of this type is fast declining
and most modern practices produce self-fluxing or super-fluxed sinter.
2. Fluxed Sinter. The amount of flux added in the mix in such that
the basicity of the mix is equal to that of the slag to be produced
in the furnace. Separate addition of flux would be required only in
proportion to the amount of natural lumpy ore charged in the fur-

nace. This type of sinter may be produced when cent per cent sinter
charge is planned. This type is also fast declining in preference to
super-fluxed sinter.
3. Super-fluxed Sinter. The entire amount of flux required to be oth-
erwise charged in the furnace, when run on 100% natural ore charge
is added in the mix. The basicity of the mix would be more than that
of the slag in the furnace, since a part of the charge would either be
natural lumpy ore or sinter of lower basicity than that of the normal
blast furnace slag. This is the aim of most of the modern practices
wherein a high percentage of sinter (50–80%) charge in the burden is
contemplated. A small proportion of natural, good graded and sized
ore charge would also be necessary. Practically no separate flux addi-
tion may be required if super-fluxed sinter is used.
It has been conclusively proven that high basicity sinters are not
desirable for good blast furnace operation. These sinters leave a refrac-
tory residue on the coke surface thereby rendering the slag more vis-
cous. It affects the permeability of the burden in the lower portion of
the blast furnace. Any reduction in basicity of the sinter should be done
without affecting other quality parameters of the sinter. Addition of
dunite, a magnesia silicate mineral, has been found at the Tata Steel to
reduce the basicity without in any way adversely affecting the furnace
performance. Later, dunite was replaced by a cheaper mineral called
pyroxenite. This way, the sinter basicity was brought down from 3.0–3.5
to 1.8–2.3. Dunite was richer in chromium oxide, and as a result, the
chromium content of the metal was high. But the use of pyroxenite39
decreased the chromium level from 0.055 to 0.046% only on partial
replacement of dunite by pyroxenite. This level will go down further on
full replacement.
7.4 Mechanism of Sintering

The addition of flux in the sinter-mix to produce fluxed sinter is benefi-
cial because the limestone addition in the BF is reduced, and the heat
required for its calcinations inside the BF is also reduced proportionately.
The same thermal energy can be better utilized inside the BF for iron
oxide reduction for more productivity. The addition of flux in sinter-mix is,
therefore, a universally adopted practice and for some practices, no flux

will be added directly to the BF at all. This means that the actual basicity
in the sinter will be much more than what would be maintained inside
without losing desired reducibility. The strength of the sinter goes up with
increase in the flux addition, but it simultaneously decreases the reducibil-
ity of the sinter because of the formation of a slag layer on the surface of
the particles or the pore surface. Therefore, many prefer only a minimal
addition of super-fluxed sinters. The various zones that are formed dur-
ing sintering on a static bed have already been shown in Figure 7.1. Each
layer below the ignited top layer undergoes changes in the order: wet ore,
drying, calcining, preheating, combustion, and cooling. The same order of
changes takes place on a moving bed. The downward movement of each
zone and the forward movement of the pallet during sintering results in
the structure as shown in Figure 7.9. This is a vertical section through the
bed over the entire effective length of the strand. Each zone is extended
diagonally over the length. The nature of zones at any intermediate stage
is shown in the form of a transverse section in Figure 7.1 which is the
condition on a static bed. It also indicates the situation at any transverse
section of the bed.
FIGURE 7.9 Vertical section through the sinter bed over the whole length showing the various zones.
The chemical composition changes gradually across the various zones.

The proportion of ferrous iron is more in the zone of calcination and com-
bustion but it decreases on cooling. Two types of bonds may be formed
during sintering:
1. Diffusion or Recrystallization or Solid State Bond. It is formed as
a result of recrystallization of the parent phase at the point of contact
of two particles in solid state and hence the name.
2. Slag or Glass Bond. It is formed as a result of formation of low melt-

ing slag or glass at the point of contact of two particles, depending
upon the mineral constitution, flux addition, etc.
As a result, the sinter can have three different types of constituents:
1. Original mineral which has not undergone any chemical or physical
change during sintering, more at the core of individual particles.
2. Original mineral constituents which have undergone changes in their
physical structure without any change in their chemistry. Recrystalliza-
tion is the only change at some of the particle surfaces.
3. Secondary constituents formed due to dissolution or reactions be-
tween two or more of the original constituents (e.g. formation of fer-
rites, spinels, silicates, etc.) These new constituents either remain as
solution and consolidate as a slag (glass) or recrystallize out of solution.
4. Partially reduced iron oxides like magnetite, maghemite and wustite
formed as a result of reduction of original iron oxide by C or by CO at
the sintering temperature.
The proportion of the physical and chemical changes during sintering depends
upon the time-temperature cycle of the process. The higher the temperature,
the faster the proportion of new constituents will form in solution. The lower
the temperature, the longer the process of recrystallization in the solid state
will take. The higher the slag bonding, the stronger the sinter is but with less
reducibility; the higher the diffusion bonding, the higher the reducibility but
strength of the sinter is less. Since ores are fairly impure, slag bond predomi-
nates. The best practice is to sinter at lower temperatures and at such a rapid
rate as to form enough slag bond but not much recrystallization.
The amount and quality of solid fuel incorporated in the mix has a great
bearing on the nature and extent of chemical changes occurring during sin-
tering. Ferric oxide content decreases with increase of carbon content and
is nearly absent if carbon content is above 7%. Ferrous silicates are formed
if carbon is above 6%. The amount of magnetite increases with carbon up to
6–7%. It then decreases if the carbon content is increased further. In general,
fluxed sinter has been found to be more reducible than acid sinters. Strong
sinters can be produced at all basicities except at around 1.5. This basicity is
avoided perhaps because of the sinter’s open tubular structure with a thin
frame-work which is rather weak. The use of a sinter of basicity 1.5 can be
avoided by making two sinters. One relatively acid and the other highly basic
and using a mixture of these two in proper proportion in the blast furnace.
7.5 Raw Material Requirements

The efficiency of producing good quality sinter depends on many charac-
teristics of the charge mix. These qualities of the charge mix should be
adjusted at the optimum levels to obtain best possible quality of sinter with-
out affecting the production rate adversely.
Size
If all other things are similar, the strength of the sinter is directly related
to the size distribution of the charge mix. These factors dictate the contact
area of particles which in turn dictates the strength of the sinter. If the size
is large, the contact area will be less and the strength of the sinter will be
low. Conversely, if size is small, the contact area of particles will be large
and the strength will be high. For sintering, therefore, it is undesirable to
have large lumps of ore in the charge. The top size of more than 10 mm
is rarely preferred. Similarly, an excessive proportion of −100 mesh would
affect the sinter bed permeability adversely. It would require pretreatment
like nodulizing before actual sintering. The size distribution of ore fines
has a pronounced effect on sinter strength. The aim is to have a maximum
contact area with the particles without impairing the bed permeability. It is
desirable to have a large proportion of fines below 3 mm and a small frac-
tion of over 6 mm size. In general, the 100–150 mesh fraction is screened
off from the 10 mm fraction before sending it to the sinter plant bins. The
100–150 mesh fraction is ideal for pelletization. The two processes, namely
the sintering and the pelletizing, are not competitors but are complimen-
tary to each other. Coke breeze of −3 mm size or fine anthracite coal are
good solid fuels for sintering. The above considerations have to be viewed
in the light of the size of the solid fuel as well as that of the flux when arriv-
ing at the optimum size of the mix. Limestone or dolomite of −3 mm size is
the best. Very fine coal and coke dust is not useful because the majority of
heat released from their burning is not available for sintering.
Fuel Content
The proportion of fuel in the charge has a bearing on the constitution of the
sinter, as was discussed in the previous section. Fuel content is usually in
the range of 6–8%.
FIGURE 7.10 Heat balance in sintering.
With the progress of sintering, the air gets preheated to an increasingly

higher level so that the fuel required to achieve the same sintering tem-
perature decreases in the lower regions of the bed. The general practice is
to have a uniform fuel content in the bed. This leads to either friable sinter
at the top or over-burnt sinter at the bottom. At least a two-layer charging
system with relatively less fuel in the lower layer has been found to give bet-
ter results. A typical heat balance during sintering is shown in Figure 7.10
as a guide to help work out the thermal efficiency of the process.
The cost of solid fuel can be nearly 50% of the total cost of the sinter
operation. In countries like India, where coke is scarce and of very poor
quality (ash 25–30%), a double layer charging system is ideal. A fuel sav-
ing of 7–8% can thus be effected. The Bokaro Steel Plant sinter strand has
been equipped with a double layer charging arrangement.
The use of extended hood, gas sintering or double layer sintering have
been developed to decrease the solid fuel requirements in sintering. The
method of calculating the practical minimum of the theoretical fuel for
given conditions of sintering has been given by Wegman.
Moisture
The presence of moisture in the sinter mix has several advantages. It main-
tains proper permeability in the bed during sintering. The rate of rise of
temperature at higher temperature levels is more rapid in wet than in dry
mix, and this is what is desired for better sintering. The presence of mois-
ture increases the specific volume of air required for sintering. This is ben-
eficial from the point for view of heat transfer during sintering. The level of
moisture in the sinter mix is generally around 3.6%.
Circulating Load
FIGURE 7.11 Mass balance in sintering.
The appropriate crushing and screening of sinter cake leads to produc-

tion of required size fraction for the blast furnace. The fines left are to be
recirculated in the charge preparation circuit. A fraction of sizes 3–6 mm

is used to lay the hearth layer on the pallets. This is essential to obtaining
good permeability of the bed for effective down draft. The −3 mm frac-
tion is mixed in the charge-mix. It also improves the permeability of the
charge bed.
For higher output of the sinter strand, the circulating load should be
low. A low circulating load, however, reduces the permeability of the bed
and consequently the input rate to the strand has to be reduced. These two
factors oppose each other, and an optimum circulating load is established
for maximum output of the acceptable sinter to the blast furnace. A typical
mass balance during sintering is shown in Figure 7.11 as a guide to assess
production efficiency of the process.
The top layer is generally more friable and the return fines arise from
this portion. The circulating load can be reduced by decreasing the pro-
portion of this top layer in the overall thickness by just increasing the bed
thickness. This, however, reduces the air flow and a compromise is again
necessary. The circulating load, the permeability (as decided by the size
and the moisture contents), and the bed thickness should be related to the
fan size, amount of leakage, etc. before the optimum values are established
for efficient operation. Many arbitrary assumptions are inevitably made in
finalizing the design. The precise role of return fines in the sintering pro-
cess is still largely a matter of guessing. In the case of high-volatile ore, fines
the return fines decrease the heat required for calcination, whereas, in the
case of low-volatile rich ores, it makes no such difference. A double-layer
sintering process has been developed for high volatile ores. In this system,
the lower layer consists of only ore fines and the top layer is only calcined by
the heat in the upper sintered layers. The lower calcined layer is returned
to become the top layer in the next pass.
Performance of Sintering Machine

The performance of the sintering machine is assessed in terms of the fines,
usually −10 mm (can be set at any other screen size depending upon the
BF acceptability of sinter) produced and the Tumbler index of the required
size-product achieved. Machine performance also goes down if the machine
is turned on and off often. This is because it takes a while to stabilize the
machine operation after it is started. The dead weight of the machine is very
large and it takes a while to achieve thermal equilibrium of the machine
during the operation.
The operation of a sintering machine can be controlled by proper adjust-

ments of the following operational variables:
1. Fuel content for heat input;
2. Ignition intensity;
3. Moisture content of mix to control its permeability;
4. Machine speed to obtain complete burn through;
5. Percent returns;
6. Bed height.
For an ideal operation, these parameters are fixed and the operator must, as
far as is possible, ensure maximum consistency ( i .e. minimum of departure
from the standard conditions) in plant operation so that sinter of desired
properties is obtained.
Economics
Typical figures indicating the capital cost of setting up of a sinter plant are
shown below:
% of total cost
Civil work
Foundation 10
Buildings 23
Electricals 15
Sinter machine (including controls) 16
Sinter cooler 7
Blower, apron etc. 7
Raw material handling equipment 10
Gas main, bunkers, etc. (plate work) 7
Miscellaneous 5
Total 100
The operating cost-breakdown is typically as follows:
Wages 30%
Repairs maintainence, 50%
supplies, utilities, etc. Transportations and general 5%
services
Fixed expenses 15%
Total 100
7.6 Recent Trends in Sintering Practice

A couple of decades ago, when furnace oil was cheaper than coke, it was
extensively used to replace coke. Emphasis was then to produce sinter with
maximum cold strength as measured by the shatter and tumbler test indi-
ces. Now the scenario has changed, and coke and coal are being used in
blast furnaces. Now the aim to make the sinter more reducible. Sintering
with low heat input can lead to these objectives:
1. Better reducibility;
2. Less slag volume finally in the B.F.;
3. Good high temperature properties like softening and melting char-
acteristics;
4. Optimum strength and RDI.
These have resulted in changes in sintering practices as follows:

1. Use of magnesium silicate mineral as flux in sinter-mix.
2. Use of quick-lime as flux in sinter-mix to improve the bonds.
3. Increase in depth of sinter bed from the usual 350–400 mm to 600–
650 mm. This has resulted in much higher productivity and a de-
creased coke rate of about 5–10 kg/t sinter. This was actually achieved
at Tata Steel.
4. A deeper bed logically leads to the adoption of double layer sintering,
(i.e. different coke percentages in the two layers higher in the top and
lower in the bottom.) This improves the coke breeze rate by about 4
kg/t sinter and reduces the blast rate by about 0.5 Nm3/t sinter.
5. Oxygen enrichment of the igniting fuel gas and extending the ignition
area by about 10% more by extending the ignition hood length. This
improves productivity and shatter strength.
6. Nearly 50% of the heat required in sintering is discharged in the open
atmosphere as waste heat. The sintering process consumes nearly 10%
of the total energy required in an integrated steel plant. Heat recovery
systems have been developed by the Kokura Steel Works of Sumitomo
Metal Industries, Japan.
7. For low production of sinter, China has development an alternative to
the standard sintering machine. It is known as Step-by-Step Box Sin-
tering Machines. Its capital cost is half that of the standard machine.
7.7 Alumina Problem in Sintering

The cost of hot metal is one of the key factors that influences the economy of
steel production in a hot metal- based integrated steel plant. The cost of hot
metal is influenced by the quality of coke and the quality of iron ore in the form
of lumps and sinter. The ways and means of improving the quality of coke has
been discussed in Chapter 4. The quality of iron ore lumps can be improved by
crushing, screening, and washing of run-of mine ore as discussed in Chapter 6.
Indian Scenario
The alumina content of Indian iron ores in general is very high, and it
reduces the strength of sinter under Indian conditions by producing more
fines and less lumpy fraction. The alumina content of Indian ores is rarely
below 2%, and it is often in the range of 2.5–4.5%. Any excess of alumina
eventually increases the slag volume inside the blast furnace, which is
somewhat restricted by adopting lower basicity, and which is only margin-
ally effective. Alumina makes the slag more refractory, and this problem has
to be tackled by increasing the basicity and or addition of MgO as flux. This,
again, increases the slag volume and the attendant problems like decreased
productivity, increased coke rate, and high operating temperatures leading
to high silicon content in the metal which have to be tolerated. The adverse
role of alumina in the burden needs no extra emphasis. Besides the iron
content and strength, the alumina content of the ore decides the quality
of ore. Fortunately, with proper washing treatment, it is possible to reduce
the alumina content of the ore lumps to below 2% as against the maximum
1% all over the world. This is the best that can be done under Indian condi-
tions, as far as the lumpy fraction of the ore is concerned.
The adverse role of alumina in the sinter on its strength and reduction-
degradation properties (RDI) has now been conclusively proved as shown
in Figures 7.12, 7.13, and 7.14. Any increase in alumina content of the sin-
ter beyond 2% alumina decreases the sinter strength as determined by the
tumbler test. Similarly, it increases the reduction-degradation index, and as a
result, the coke rate goes up. For maintaining the same RDI, the basicity of
the sinter has to be increased. For better blast furnace performance, the stack
zone should be as extended as possible with the softening and melting zone
confined to as narrow a zone as possible in the lower part. This is possible only
if the RDI is low (i.e. alumina content is low), particularly at low basicities.
This has resulted in better quality sinter with a RDI of 28 and a TI of
the order of 72. Because of this design, it has been possible to incorporate
more than 70% sinter in their blast furnaces. This has greatly improved
their BF productivity.
FIGURE 7.12 Effect of alumina content on sinter strength.40
FIGURE 7.13 Effect of alumina content on RDI of sinter.40
FIGURE 7.14 Effect of alumina content on sinter RDI at various CaO contents at Tata Steel.
It should be noted that the productivity of a sinter plant drops with increasing
alumina content. This will be obvious from Figure 7.15 where the sinter plant pro-
ductivity of Indian plants are compared with more productive plants in the world
with respect to the alumina content of the sinter-mix.37b Hence all the efforts to
reduce the alumina contents in the ore before it enters the BF as a charge.
FIGURE 7.15 The effect of alumina content of the charge on BF productivity.37b

This type of sintering process is one of the most efficient and compact
designs for heating iron ore charge on such a large scale. In fact, such a
system of direct heating should be used for other useful applications in
metallurgical industries such as the calcinations of limestone.
The ability of sintering to convert a wide variety of sandy fines of iron
ores into a most useful lump, in fact better than natural ore lumps, at an
acceptable cost has now been established beyond doubt. Understanding of
the underlying principles of sintering process has led to the production of
good quality artificial burdens and charges for the blast furnaces. The suc-
cess of this process is seen in some of the most modern and large capacity
blast furnaces, where the cent per cent burden is artificially prepared to
obtain the desired properties. This has more than paid back the cost so
involved by way of increased productivity, lower coke rate, and far better
control of the operation. The major proportion of prepared burden in such
cases consists of sinter. In some cases, it is sinter and sinter alone.
It is not possible to sinter iron ore fines, if the particle size falls below
100–150 mesh. This, however, is an ideal size for its conversion into pellets,
which is also an equally good artificial feed for a blast furnace. In fact, these
two processes of agglomerations, sintering and pelletization, are totally
complimentary to each other in that there is no competition regarding their
applicability, since the charge size required is altogether different in these
two processes. By adopting these two processes, the entire lot of ore fines of
below 10 mm can be usefully consumed for iron production. Even the ore
fines that accumulated while supplying the lumps to the BF over the years,
can now be transformed into a useful agglomerate, in a way a better quality
burden. Agglomeration of iron ore fines by the process of pelletization is
the subject matter of the next chapter.
CHAPTER
8
BURDEN PREPARATION:
AGGLOMERATION AND
PELLETIZATION
In This Chapter
● Introduction
● The Pelletization Process
● Theory of Bonding
● Mechanism of Ball Formation
● Feed Size, Shape, and Porosity
● Green Ball Production
● Additives
● Induration of Pellets
● Comparison of the Pellet Firing Machines
● Indian Pelletization Plants
● U.S. Production (Pelletization)
8.1 Introduction
Although the process of rolling moist material into balls was patented in
1913, nearly 40 years elapsed before it became a practical and commercially
attractive method of agglomeration of ore fines or concentrates. It came to
be known as a process of pelletization and the small balls as its product as
pellets. For the first time, it was commercially adopted to agglomerate the
ultra-fine iron ore concentrates obtained by upgrading the taconite deposits
of the Mesabi range in U.S.A. Such a fine material could not be easily and
economically sintered on a grate to convert it into a useful blast furnace
feed. The process, however, has now been developed in its own right as a
method of preparing a high quality feed material for blast furnaces. This
method of agglomeration is ideally suited for fine iron ore concentrates
obtained after upgradation of lean ores. It is equally suited for agglomera-
tion of dust recovered from exhaust gases in iron and steel making units,
as well as pyrite residues. The size range ideally suited for pelletization is
−100 mesh. The process of sintering and pelletization are not competitors.
They are complimentary to each other in converting the fine iron bearing
material into a useful blast furnace feed.
With a very rapid increase in demand for metallics all over the world,
for increasing steel production, the iron ore is not only becoming more
scarce but more costly as well. Attempts are underway to utilize cent per
cent iron ore that is mined and supplied or purchased. This includes use
of all the size fractions of iron ore: the use of +10−40 mm as natural lump
feed directly to the BF, +100 mesh and −10 mm to be sintered at the BF
site, and all −100 mesh to be micro-pelletized and added to sinter-mix or
regularly pelletized thereby ensuring cent per cent utilization of iron ore.
For this, it is not necessary to carry out sintering and pelletization in the
same plant where lumps are consumed as direct feed. Usually, sintering is
carried out as an adjoining activity with the BF. At the mine head, screening
is employed to separate +150 mesh −40 mm for dispatch to the BF and the
−150 mesh is sold for pelletization.
The pellet production is, therefore, rising continuously, even though it
is costlier than sinter. At times, the −150 mesh size is micro-pelletized and
the product is mixed with sinter feed. The use of pellets as BF feed is rather
uncommon. However, it is more commonly used for DRI making.
8.2 The Pelletization Process

Pelletization essentially consists of formation of small green balls by roll-
ing a fine iron bearing material with a critical amount of water; an external
binder or any other additive may be added if required. These green balls
of nearly 5–20 mm size are then dried, pre-heated, and fired, all under
oxidizing conditions, to a temperature of around 1250–1350º C. Bonds of
Burden Preparation: Agglomeration and Pelletization • 153
good strength are developed between the particles at such high tempera-
tures. The sensible heat of the exhaust gases is recovered and is fed back
in the induration operation. The process, therefore, produces pellets in a
highly oxidized state, in contrast to the sintering process where solid fuel
is incorporated in the charge and results in a 5–20% ferrous iron.
The pelletization process consists of the following steps:
1. Feed preparation.
2. Green ball production and sizing.
3. Green ball induration:
(a) Drying
(b) Pre-heating
(c) Firing
4. Cooling of hardened pellets.
8.3 Theory of Bonding

The production of hardened pellets first starts with the production of green
balls. It is necessary that the green balls possess adequate strength to with-
stand bumps, overlying load, and abrasion during screening and handling
until finally they are hardened by induration. The nature of bonds and
how these are formed need to be properly understood in order to properly
design the equipment and the process of pelletization.
The observations on ball formation that eventually led to the develop-
ment of the theory of balling are as follows:
1. Dry material does not pelletize, and the presence of moisture is essen-
tial to rolling the powder into balls. Excessive water is also detrimental.
2. Surface tension of water in contact with the particles plays a dominant
role in binding the particles together in green condition.
3. Rolling of moist material leads to the formation of balls of very high
densities. This is otherwise attainable only by compacting powder un-
der the application of a very high pressure.
4. The ease with which material can be rolled into balls is almost directly
proportional to the surface area of particles, (i.e. its fineness).
From fundamental studies, it has been concluded that there are three dif-
ferent water-particle systems:
1. The pendular state, when water is present just at the point of con-
tact of the particles and surface tension holds the particles together as
shown in Figure 8.1 (a).
2. The funicular state, when some pores are fully occupied by water in an
aggregate system as shown in Figure 8.1 (b).
3. The capillary state, when all the pores are filled with water but there is
no coherent film covering the entire surface of the particles as shown
in Figure 8.1 (c).
The bond strength under these three conditions have been mathemati-
cally evaluated, and based on these formulas, equations to calculate the ball
strength have been developed by some workers.
FIGURE 8.1 Water-particle systems.
Besides the bonds formed from surface tension, the mechanical inter-
locking of particles also plays a significant role in developing the ball strength.
The maximum strength of a green ball produced from a given mate-
rial will be obtained by compacting the material to the minimum porosity
with just sufficient water to saturate the voids. The rolling action during
pelletization is beneficial in reducing the internal pore space by effecting
compaction and mechanical interlocking of the particles. The water content
is, therefore, quite critical to obtain good quality balls. The importance of
the critical amount of water in developing the strength can also be appreci-
ated from the following common experience. At the river-side or beach, the
water content increases from the dry sand of the shore to a very wet sand at
the point of contact with the water. There is a region in between where the
sand has maximum strength, it decreases on either side of this region. The
strongest region is that region where the water content is just sufficient to
fill all the pores without forming a coherent water film.
The water content of the pelletization feed is, therefore, quite critical
for obtaining strong green balls. The critical moisture can be assessed by
determining the porosity of an aggregate. The value does not vary with the
ability of the material to absorb water. The factors that alter the porosity also
affect its critical moisture level, and an efficient balling process is required
to produce balls of uniform properties in spite of these non-uniformities.
8.4 Mechanism of Ball Formation

The ball formation is a two stage process, (i.e. nucleation or seed formation
and their growth). The formation of balls on a pelletizer depends primar-
ily on the moisture content. If moisture is less than the critical amount,
its distribution tends to be non-uniform, a major amount being present in
the granulated material leaving the non-granulated material relatively dry.
If the moisture level is more than the critical value, growth rate is more
but the balls produced are liable to deformation because of their plasticity.
Seeds are formed only if a critical moisture level is maintained, and without
which the process cannot proceed properly. The seeds generally have slight
excess of water. This imparts a certain degree of plasticity which is neces-
sary for its growth. Growth takes place by either layering or assimilation.
FIGURE 8.2 Ball size as a function of extent of rolling (i.e. number of revolutions of the pelletizer).
It has been observed that the size of the balls produced in a pelletizer from
a charge containing the right amount of moisture depends on the time and
speed of the pelletizer, (i.e. number of revolution). The general nature of the
variation is shown in Figure 8.2 in which three regions can be clearly observed:
1. Nucleii formation region;
2. Transition region;
3. Ball growth region.
Nucleii Growth Region

When a wet particle comes in contact with another wet or dry particle, a
bond is immediately formed between the two. Similarly, several such par-
ticles initially join during rolling to form a highly porous loosely held aggre-
gate and crumbs. These undergo re-arrangement and partial packing in
short duration to form small, spherical, stable nucleii. This is the nucleation
period, a pre-requisite for ball formation, since these very nucleii later grow
into balls. The surface tension of water alone is the binding force in these
nucleii. The growth rate of nucleii appears to be related to their specific
surface and not to the specific surface area of the feed material. The size of
the nucleii is quite small.
Transition Period
After nucleii are formed, they pass through a transition period in which the
plastic nucleii further re-arrange and are compacted to eliminate the air
voids present in them. The system moves from a pendular state through a
funicular state to the capillary state of bonding. Rolling action causes the
granules to densify further. The granules are still plastic with a water film
on the surface, and they are capable of coalescing with other granules. The
size range of granules in this region is fairly wide.
Ball Growth Region

The plastic and relatively wet granules grow if they are favorably oriented.
In this process, some granules may even break because of impact, abrasion,
etc. Growth takes place by two alternative modes:
(i) growth by assimilation is possible when balling proceeds without the
addition of fresh feed material.
(ii) growth by layering is possible when balling proceeds with the addi-
tion of fresh feed material.
Growth by Assimilation
If no fresh feed material is added for balling, the rolling action may break
some of the granules, particularly the small ones, and the material coalesces
with those which grow. The bigger the ball, the larger it will grow under
these conditions. Since smaller granules are weaker, they are the first victims
and the growth of the bigger balls takes place at their expense. Growth con-
tinues until such a size when the torque tending to separate the contacting
balls exceeds the bond strength of coalescence. Balls with a higher percent-
age of moisture, therefore, grow to a larger size, but the strength of the
resulting balls is low. It has also been shown, that after the initial growth by
coalescence further growth may occur by assimilation, (i.e. joining of two
granules without prior fragmentation).
Growth by Layering
Growth of the seeds takes place by layering when the balls pick up material
while rolling on a layer of fresh feed. The amount of material picked up by
the balls is directly proportional to its exposed surface, (i.e. the increase in
the size of the balls is independent of their actual size). In practice, fresh
material is always supplied to the pelletizer continuously while withdraw-
ing the equivalent amount of balls from the circuit. The conditions are,
therefore, more suitable for growth to take place by layering. However, it
has been observed that if growth by layering is predominant, the size range
of the balls becomes narrower than the size range produced by assimilation
mechanism. Growth by layering is more predominant in the disc pellet-
izers, and growth by assimilation is more predominant in drum pelletizers,
at least beyond the feed zone.
8.5 Feed Size, Shape, and Porosity

The hardened pellets should be strong and adequately porous to possess
good reducibility. To achieve this, the green balls must also possess good
strength and porosity. In general, the strength and porosity vary inversely
and they have to be adjusted at their optimum values. It has also been
broadly accepted that the feed should be 100 mesh (200 μm) in size, with
more than 60% as 350 mesh (≈ 44 μm). A right combination of sizes and
shape of feed particles can lead to maximum mechanical interlocking and
surface tension and capillary force to obtain strong green pellets.
If the material does not have the correct size, further grinding is nec-
essary. The finer the size, the larger is the specific surface area. Broad
indications are that a surface area of 0.2 m2/g (Blaine) is likely to be a mini-
mum value for successful balling, and the grinding should be directed to
this end. Over grinding is wasteful; beyond a certain fineness there is no
marked improvement in the quality of the balls produced. Overgrinding is
detrimental since it reduces the porosity of the final product. The extent
of grinding that is necessary should be decided in relation to the desired
quality of the product. This, in turn, will very much depend on the fracture
characteristics of the ore minerals which determine the shape of the parti-
cles. Although very fine grinding overcomes the problem of the shape factor
of the material in balling, there exists a danger of overgrinding. The shape
factor, however, should be taken into consideration in evaluating the com-
paction, density, and strength of balls without overgrinding the material.
The surface roughness increases the surface area for the same size and
may improve green ball strength.
The size distribution of the material to be balled has a marked effect
on the porosity of the product. The smaller particles fill up the pores left
in between the bigger particles and reduce the porosity. The reduction
in porosity is most marked if the size of the smaller size fraction is one
fourth that of the bigger particles and 25% of the smaller component is
added. Since the actual size range is fairly wide, the actual porosity should
be experimentally established. Porosity value of about 20–30% in green
balls is considered to be adequate since such a product responds well to the
induration process.
8.6 Green Ball Production

Green ball agglomerates are economically produced in balling drums or
discs and rarely in other specialized equipment. The rolling action of the
equipment is chiefly responsible in compacting the fine particles into a
near spherical shape; the presence of a critical amount of water develops
the necessary strength in the green balls. The balling equipment should,
therefore, produce a maximum of rolling motion of particles, particularly as
they grow. The balling equipment would require a rotating surface where
the pellet mix is fed. Facilities for addition of moisture, removal of balls of
desired size, and recirculation of the undersize and oversize should be pro-
vided with the pelletizer. The feed mix may contain a binder and/or a flux
to obtain better pellets or fluxed pellets. No fuel is incorporated in the mix.
The pelletizer acts as a mixer in the mechanical sense, and no pre-mixing
is usually necessary. Water is generally added into the pelletizer itself. The
feed is generally dry but partially moist feed may also be used.
The formation of balls on a rotating surface is a two stage process. The
nuclei or seeds are first formed and their growth then takes place later. It has
been observed that these two stages of ball formation take place in two dif-
ferent zones of the pelletizer, (i.e. the nucleating zone and the growth zone).
The region where water is added on the pelletizer is the zone of nucleation.

Growth takes place over practically the rest of the effective area of the pelletizer.
Disc Pelletizer
It is essentially a disc, with an outwardly sloping peripheral wall, which
is rotated, around its own center, in an inclined position to horizontal as
shown in Figure 8.3. These discs resemble flying saucers and are normally
3.6–5.6 m in diameter and are inclined at about 45 degrees to the horizon-
tal. The material to be pelletized is generally fed directly onto the disc, and
the moisture level is maintained with the help of water sprays. A scraper is
provided to prevent buildup of moist material on the disc. It can also con-
trol the material flow pattern on the disc.
In the region where water is added, seeds are easily formed. With the
growth of these seeds, their frictional drag against the disc decreases and
the centrifugal force acquired increases. Consequently, they move out of the
nucleation zone. They also tend to rise on the inclined surface of the pellet-
izer in the direction of rotation and fall down against the toe section of the
disc. The height and width of trajectory of the ball movements increase with
the size of the ball until the balls are deflected downwards by the scraper.
During this movement, the balls encounter fresh feed and growth takes place
more by layering while compaction-assimilation plays a relatively minor role.
FIGURE 8.3 The disc pelletizer.

Pelletization can be carried out on a disc with reasonable flexibility. In

a continuous operation, the discharge of desired size balls is balanced by an
equivalent addition of feed. The concept of formation of balls in two stages
implies that there is a specific residence time for any given mix to produce
balls of the required size. For steady-state-conditions, the residence time
required for balling may be ascertained by dividing the weight of material
being actively balled on the disc by the feed rate.
The rate of production of balls on a disc is a function of the following
variables:
1. Diameter of the disc.
2. Height of the peripheral wall.
3. Angle of inclination of disc with horizontal.
4. Speed of rotation.
5. Place on the disc where mix is fed.
6. Place where water is sprayed on the disc.
7. Rate of feed.
8. Rate of moisture addition.
9. Rate of withdrawal of the product (i.e. amount of material on the disc).
10. Nature and size of feed.
11. Desired size range of pellets and percentage recycled load.
12. Any other additions like binder, flux, etc.
All of these factors affect, in the final analysis, the rates of nucleation and
growth. For optimum utilization of a given disc, the residence time of the
charge on the disc must be the same as the balling time. In other words,
the rate of feed should be equal to the rate of discharge. For maximum
production, optimum utilization of the disc has to be achieved. This occurs
when the un-balled material begins to move down the inclined plane after
having reached the highest point on the disc. Under these conditions, it has
been estimated that nearly 70% of the disc surface is being actively used
for balling.
As the angle is increased, compaction is increased and pellets of better

strength are obtained. As the speed is increased, the rolling improves and
consequently the rate of growth increases. Excessive speed, however, may
result in breaking the pellets by impact. This can be obviated by reducing
the slope. The limiting speed is that at which the entire material is cen-
trifuged without rolling taking place. The higher the speed, the more the
process becomes sensitive to even slight excess of moisture. For each mix,
the angle and speed are increased until maximum utilization of the disc is
achieved.
For a given slope and feed rate, the residence time can be adjusted
to some extent by varying the height of the peripheral wall. It has been
estimated that loading of a disc beyond 0.20 t/m2 area hampers the balling
operation as well as the input. The rate of the nucleation and growth can
be relatively adjusted, besides any other factors, by adjusting the points at
which charge and water are fed. If the feed point is such that on addition of
feed it immediately comes in contact with the rolling balls, then the growth
rate is more dominant than the rate of seed formation. The product in this
case would be more closely sized. The farther the feed point is from the
rolling trajectories of the balls, the higher the rate of seed formation and
consequently the wider the size range of the product. The ball size is also
determined by the residence time, (i.e. by the rate of feed and the rate of
discharge).
It has also been observed that the disc produces a self-preserving size
distribution, the range of which is a function of the nature of the mix and its
resistance time in the ball growth region. The size range becomes narrower,
as growth tends to attain steady-state conditions, particularly if it occurs
by layering rather than by assimilation. The production of balls of desired
size range from a continuously operating disc requires careful selection of
points to feed the mix and the water. The variation in these should be read-
ily feasible in the pelletizer design.
On a disc, the recycle proportion is nearly 15% which either is a perma-
nent feature or is meant to cure the non-uniform performance. Although
growth rate is more for increased moisture, it tends to decrease the strength.
The finer the feed is, the better the quality and the higher the growth
rate of the pellets. The disc pelletizing circuit may have one or more discs,
and their operation has to be synchronized to obtain a uniform rate of pro-
duction, so that induration unit works smoothly.
Drum Pelletizer
The drum pelletizer is a simple steel drum, open at both ends with a
length-to-diameter ratio of 2.5–3.5, rotating around its own axis in a slightly
inclined position to the horizontal. The drums are 2–3 m in diameter and
6–9 m in length and rotate at 10–15 rpm depending upon the diameter. The
angle of inclination in working position is about 2–10 degrees. The charge is
fed through the end which is at a higher level; water sprays are also located
there. The material rolls over the surface of the rotating drum and slides
downwards due to the inclination of the drum. This motion of the balls is
called cascading which takes place until the balls emerge out of the lower
end.
There are some fundamental differences between the behavior of a
drum and a disc pelletizer. The drum does not act as a classifier as the disc
does, and the time of ball growth is not equal to the residence time of the
feed. The size range of the output is, therefore, large and it must be oper-
ated in closed-circuit with a screen. The ball growth is usually achieved in
more than one pass through the drum. Under a continuous steady-state
operation, balls of the size 10–16 mm are screened out, and the undersize
and the over-size is shredded before being recirculated.
The rate of production of desired size balls is a function of the following
variables:
1. Speed of rotation, (i.e. peripheral speed).
2. Angle of inclination.
3. Diameter of drum.
4. Rate of feed.
5. Depth of material in residence (loading).
6. Moisture content.
7. Nature and size of feed.
8. Any other additions such as binder, flux, etc.
All these factors affect the rates of nucleation and growth and, therefore,
the production rate. Factors such as variations in the place where water
and feed are added, rate of withdrawal, etc. do not play a significant role
as they do in a disc pelletizer. In a continuously operated drum, the factors
numbered above as 4., 5., and 9. are automatically fixed in relation to each
other once one is fixed. The effect of speed of rotation is similar to that in a
disc pelletizer. The correct speed of rotation is 25–35% of the critical speed,
(i.e. the speed at which material will centrifuge).
The optimum loading of the drum is equivalent to about 5% of the
drum volume. The residence time or balling time is determined by the
slope, water content, speed of rotation, and the feed rate. Optimum loading
and correct speed of rotation leads to correct rolling and tumbling action
eventually resulting in a continuous steady-state operation. Even in a con-
tinuous operation, anything between 50–100% of the feed is the recycle
load which affects the ratio of growth by layering and by assimilation. A
high recycle rate favors growth by layering and produces a product of close
size-range. For a continuous, stabilized steady-state operation, the net rate
of production of seeds must equal the rate of production of green balls,
and the rate of incremental growth during the passage through the drum
should give rise to the correct size of the balls. Incremental growth can be
increased by decreasing the slope of the drum, thereby increasing the roll-
ing time per pass and vice versa. In order to achieve a constant production
rate, the incremental growth rate must exceed a certain specific value. In
practice, the drum pelletizers rarely operate under steady-state conditions
because of the difficulty in matching the variables. The production rate,
however, should not vary more than +8% in the case of a single unit for
smooth induration operation. In a multi-unit plant, variations up to +12%
are tolerated. The output, in practice, can be increased by increasing the
size of the drum, but in a given design the speed of rotation and feed rate
are adjusted to obtain maximum output from a given mix.
TABLE 8.1 Comparison of Disc and Drum Pelletizers

S. No. Item Disc Drum
1. Extent of mixing Acts nearly as a perfect Not to the same extent of
mixer. disc.
2. Classification of product Acts as a classifier also. Not like disc.
3. Nucleation In the wet zone. In the wet zone.
4. Growth Mainly by layering. More by assimilation.
2
5. Specific production rate 20 t/m of the pelletizer 7–12 t/m2 of the pelletizer
area. area.
6. Space requirements and 30% less than that required More than that for equiva-
capital cost for equivalent capacity lent capacity disc.
drum and proportionately
less capital cost.
S. No. Item Disc Drum

7. Circulation load Less; less power is required Both are more, 50–100% of
to run the disc (≈15% of the the feed rate.
feed rate)
8. Flexibility of operation More flexible with respect Less flexible.
to rate of nucleation and
growth.
9. Product quality Close sized product. Wide range of product size.
10. Screening of green balls Adopted only if very nar- A must.

row size range product is
required.
11. Size For the same output, the –
size of disc is much larger
than drum.
12. Supervision. More. Less.
8.7 Additives
Many times, the iron bearing material by itself may not roll into balls or may
not produce green balls of the required strength just by the addition of the
right amount of water. Under such circumstances, the addition of a small
amount of certain binders have been successful in improving the balling
operation as well as the green ball strength.
Use of organic substances like starch, dextrine, alginate products, etc.
has been beneficial in the preparation of green balls. But their use, many
times, does pose insurmountable problems during induration of these balls.
Hence, in general, these are not preferred. Inorganic binders like ferrous
sulfate, alkali chlorides, alkali and alkaline earth carbonates, etc. have been
successfully used. Bentonite has found the widest application since its use
improves induration and results in a better quality of fired ball with less
tendency to decrepitation and cracking.
0.5–1.0% bentonite is generally added in the feed during its prepara-
tion. Excess of addition can lead to undesirable properties. The choice of
binder additive and its optimum content in the feed should be experimen-
tally established for a given raw material and given process-plant design.
Use of lime or dolomite as an additive is beneficial since it ultimately pro-
duces slag bonds during induration, and this is exhibited as better strength
of the pellets.
8.8 Induration of Pellets

These green balls from the pelletization plant are sent to the induration
plant where these are subjected to drying, pre-heating, firing, and cooling.
For this purpose, the pellets are subjected to a continuous heating and cool-
ing cycle depending upon the variables of balling process, the mineralogical
nature of the raw materials, and the type of bonds to be developed. The
balls attain a maximum temperature during the firing operation. The time
for which the balls are subjected to this depends mainly on the quality of
the pellets ultimately required. The heat recovered during cooling of the
pellets is invariably utilized in drying the pellets.
Three different types of heat hardening equipment are in commercial
use at the present time: the vertical shaft furnace, the traveling grate, and
the grate kiln. In general, green pellets are fed at one end of these mod-
ern machines, and hardened and cooled pellets are withdrawn from the
other. In other words, these are continuous induration units. However, in
some earlier shaft furnaces, drying units were independent of the main fir-
ing furnace.
In some others, the drying and or cooling are carried out in indepen-
dent units.
Drying
The green balls are dried by passing hot air through the bed of the balls laid
on a grate. The temperature of the hot blast and volume required for drying
are largely a function of the moisture content of green balls and the nature
of the ore from which the balls are made. Specular hematites decrepitate
at temperatures in excess of nearly 300º C. Hence, these should be dried at
well below this temperature. On the contrary, natural magnetites may safely
be dried at even 475º C. The use of binders increases the highest safe per-
missible temperature for drying. The depth of bed has to be suitably fixed
to avoid condensation of moisture from gases in the cold zone.
The strength of dried balls may be due to the mechanical interlock-
ing obtained earlier during the rolling operation. Adequate dry strength
of the balls is necessary to stand the overlying burden during drying and
also to withstand the pre-heating and handling operation. The compres-
sive strength of pellets varies with its moisture content during drying as
shown in Figure 8.4. The use of binders is beneficial in obtaining better dry
strength.
FIGURE 8.4 Variation in compression strength of pellets as a function of moisture content during the
drying cycle.
Drying may be carried out by using either a down-draught or an up-

draught. Down-draught drying is believed to be safe for green balls contain-
ing less than 9% moisture. Up-draught drying followed by down draught
drying should be employed, using drying gas, at high temperature and low
flow-rates. Prior experimental work is essential to assess the best gas tem-
perature and flow-rates to obtain efficient drying with maximum thermal
efficiency.
Preheating and Firing

The hardened pellets should not only possess the optimum mechanical
properties in the oxidized state, but more so while these are being reduced
in the blast furnace. The firing cycle should aim for this when developing
the required nature and strength of the bonds. The role of additives has
to be regarded from the point of view of developing optimum mechanical
properties. As in sintering, the solid state bond (crystal growth) and the slag
bond (due to melt formation) may be formed during firing.
In the preheating stage, partial oxidation of magnetite and sulfides, if
present, occurs along with the calcination of carbonates and hydroxides.
The solid bonding and grain growth commences in the preheating period.
These are completed in the firing stage. Liquid formation by chemical com-
bination of various acid and basic constituents, and consequently, the devel-
opment of slag bonds occur predominantly during firing.
The strength of the pellet is developed initially at the shell and pro-
gresses towards the center with time at the firing temperature. The bond
strength is essentially developed at the firing temperatures, but the heating
cycle should be decided in relation to the chemistry of the pellets. This
develops an optimal quality of hardened pellets. The heat evolved during
the oxidation of magnetite or sulfides is by far the largest source of energy.
This heat must be carefully utilized in the promotion of grain growth
and liquid bond formation. The calcination of minerals, on the contrary,
requires heat and this must be accomplished without disintegration of the
pellets. The calcination of minerals should also be completed ahead of the
time when maximum firing temperature is reached. Optimum fuel econ-
omy should be achieved by carefully planning the firing cycle. It should be
clear that heat requirements for hardening hematite pellets are more than
those for magnetite. Similarly, those for goethite, limonite, etc. are more
than those for hematite.
In general, firing temperatures in the range of 1250–1350º C, are suf-
ficient to produce pellets of the required qualities. Higher temperatures
lead to excessive slag formation and less porous pellets, a quality detrimen-
tal to the blast furnace operation. The actual firing temperature is decided
in terms of the basic feed material, the additives, the holding time at the
maximum temperature, the size of the pellets, and the ultimate quality of
pellets required.
Firing Units
Three main types of pellet firing machines that are in commercial use are
described below:
The Shaft Kiln. It was developed in the 1950s and is shown in
Figure 8.5. The green or dried pellets are fed vertically downwards in a
central shaft of a rectangular cross section. Fuel is burned in two fire cham-
bers, one on each long side of the shaft, and the hot gases are allowed to
enter the main shaft through multiple flues. It is, in a way, a counter-current
gas solid heater. The fired pellets are cooled in the lower portion, and if any
chunks are formed these are broken by the chunk breakers and cooled pel-
lets are discharged from the bottom. Cooling air, introduced from below,
gets pre-heated and is either taken out to burn the fuel in the chamber or
to make available preheated gas for completing the combustion in the firing
zone of the shaft.
FIGURE 8.5 The shaft kiln for induration of pellets.
A typical furnace of this type has a shaft of around 4 × 2 m cross-

section, and it is nearly 20 m in height from the discharged point to the
stock level. The feed and the discharge rates are adjusted to maintain a
nearly constant stockline level. Nearly a third of the total air required for
combustion is admitted in the combustion chambers for partial burning
of the fuel; the remainder is introduced in the shaft for cooling from
below, which is then available as preheated secondary air for completing
the fuel combustion inside the shaft. The temperature in the combustion
chamber is around 1300º C. The furnace contains nearly 200 t of pellets
and the production rate of hardened pellets is around 1000–1200 t per
day. Some recent designs have incorporated external coolers or internal
coolers and heat exchangers to improve upon the thermal efficiency of
the process.
The Grate Machine. These are essentially the modified adoptions of a
Dwight-Lloyd sintering machine in which the green balls are fed at one end
on the continuous traveling grate, and the hardened and cooled pellets are
discharged from the other end. The length of the grate is divided into four
different zones: drying, pre-heating, firing, and a cooling zone of suitable
relative lengths. The hot air from the cooling zone is circulated in a complex
manner to carry out drying, dehydration, pre-heating, etc. Down-draught,
up-draught, or a combination of the two is suitably employed in the design
to carry out the preliminary operations before the actual firing. The combi-
nation varies with the mineralogical constitution of the base feed, and as a
result, the designs differ considerably from place to place.
FIGURE 8.6 The straight grate for induration of pellets.

FIGURE 8.7 The grate kiln for induration of pellets.
The Dravo-Lurgi straight grate systems are fairly popular. The same
has been adopted in the latest pelletization plant being set up by the M/s
Chowgule Mining Co. in Goa to produce 6 Mt of pellets per year for export.
The general scheme is shown in Figure 8.6. Oil or gas burners are used
in the firing zone to generate the necessary temperature. No solid fuel is
incorporated along with the pellets.
A peculiar feature of such a system is the necessity of a protective layer
of previously fired pellets on the bottom as well as on the sides of the bed.
The temperature, therefore, need not be raised to the firing temperature
level near the bottom or the walls, and the danger of overheating the metal-
lic pellets is thereby eliminated. If such a precaution is not taken, the pellets
near the bottom or the walls will remain underfired in trying to protect the
metallic grate and the walls.
The latest modified version of the grate machine is the use of a circular
grate in place of a straight grate. It eliminates the operational difficulties that
are encountered in the straight grate. One such unit has been recently com-
missioned in Mexico, and its performance is being watched with interest.
The Grate Kiln. It is a combination of a grate and a kiln. The dry-
ing and pre-heating is carried out in this machine on the straight traveling
grate, as in the grate system, and the firing is completed in a long rotating
kiln wherein the fuel is burned to generate the necessary temperature. Hot
gases from the kiln are carried over the grate, where drying and pre-heat-
ing are carried out under down-draught conditions or in combination with
up-draught movement as well. A typical scheme is shown in Figure 8.7.
The traveling grate, in this case, is not subject to very high temperatures,
and the life of the grate is very high. It is this single factor that makes the
machine a costly operation.
8.9 Comparison of the Pellet Firing Machines

A number of attempts, in detail, have been made to assess the relative cost
of various operations in the above mentioned three different pellet firing
machines. Some of the broad conclusion of these studies are as follows:
1. The induration time for continuous grate is very short. The grate kiln and
the shaft furnace require too long a induration as shown in Figure 8.8.
2. Fuel consumption of a shaft furnace is only slightly less than that of the
other two. Its electrical power requirement is, however, much higher
than that of a continuous grate and which itself is higher than that of
the grate kiln.
3. Capital outlay of the grate kiln is much more than the other two.
4. Maintenance cost of the traveling grate is more than that of a grate
kiln; shaft furnace requires minimum.
5. Overall production cost is highest for continuous grate, lowest for shaft
furnace, and intermediate for grate kiln.
FIGURE 8.8 Firing cycle of pelletizing in three types of induration units.

8.10 Indian Pelletization Plants

Indian iron ores are characterized by:
1. Presence of blue dust patches, rich in iron contents;
2. Generation of a considerable amount of iron ore fines during benefi-
ciation which are rich or can be enriched by concentration techniques;
3. Magnetic iron ores which can be beneficiated to give fine iron rich
concentrate.
All these have led to the installation of pelletization plants in India and
which are given in the table below:
TABLE: 8.2 Indian Pelletization Plants

S. No. Name of the plant Rated capacity Remarks
Mt/year
1. Kundremukh Iron Ore Co. Karnataka 3.0 60% Feed
–325 mesh
2. Chowgule and Co. Lale, 0.55
Goa
3. Mandovi Pallets Ltd. Mandovi, 1.8
Goa
4. Noamundi (Tata Steel) 0.8 Now closed
Indian steel plants do not use pellets as burden feed. All the pellets are
exported, except that from Tata Steel, but it has been closed. The Noamundi
pellets were once used in Tata Steel’s blast furnaces for some time.
Essar Steel is planning to install a pelletizing plant at Vishakhapatanam.
Jindal Ispat is contemplating settin up a 3.0 Mt plant in Karnataka for their
own steel production. SAIL is planning to set up a 400 tpd cold bonded
pellet plant at Gua in Bihar, based on indigenously developed technology.
In India, pellets are not used as a blast furnace burden. However, pel-
lets are prepared at Mandovi, Goa and Kundremukh, Karnataka which are
entirely exported: 1.8 Mt from Mandovi and 3.0 Mt from Kundremukh.
8.11 U.S. Production (Pelletization)

Domestic Production and Use: In 2014, mines in Michigan and Minnesota
shipped 93% of the usable iron ore products in the United States, with
an estimated value of $5.1 billion. Twelve iron ore mines (9 open pits and
3 reclamation operations), 9 concentration plants, 10 pelletizing plants,

2 direct-reduced iron (DRI) plants, and 1 iron nugget plant operated dur-
ing the year. Almost all ore was concentrated before shipment. Eight of the
mines, operated by three companies, accounted for the majority of produc-
tion. The United States was estimated to have produced and consumed 2%
of the world’s iron ore output (See Table 8.3).
TABLE: 8.3 U.S. Pelletization Plants

(Data in million metric tons gross weight unless otherwise noted)
Salient Statistics–United States: 2010 2011 2012 2013 2014
2 49.9 54.7 54.0 53.0 57.5
Production
Shipments 50.6 55.6 52.9 52.7 54.2
Imports for consumption 6.4 5.3 5.2 3.2 5.5
Exports 10.0 11.1 11.2 11.0 13.0
Consumption:
Reported (ore and total agglomerate)3 42.3 46.3 46.9 48.8 49.5
4 47.9 49.1 48.1 47.1 47.8
Apparent
Price,5 U.S. dollars per metric ton 98.79 99.45 98.16 104.90 101.00
Stocks, mine, dock, and consuming plant,
yearend, excluding byproduct ore 3.47 3.26 3.11 2.29 4.50
Employment, mine, concentrating and
pelletizing plant, number 4,780 5,270 5,420 5,644 5,750
Net import reliance6 as a percentage of
apparent consumption (iron in ore) E E E E E
SOURCE:USGS
2
Includes agglomerates, concentrates, DRI, direct-shipping ore, iron nuggets, pellets, and
byproduct ore for consumption.
3
Includes weight of lime, flue dust, and other additives in sinter and pellets for blast furnaces.
4
Defined as production + imports – exports + adjustments for industry stock changes.
5
Estimated from reported value of ore at mines.
6
Defined as imports – exports + adjustments for Government and industry stock changes.
U.S. iron ore production was expected to increase in 2014 and 2015 from
that of 2013 owing to new production that began in late 2013 and 2014.
In December 2013, one company in Louisiana began producing DRI pel-
lets from imported iron ore concentrates. The facility was the largest of its
kind in the world, with a 2.5 million-ton-per-year capacity, although equip-
ment failure, upgrades, and repairs temporarily idled the plant periodically
during the year. In September, Reynolds Pellet Plant began operations in
Indiana using iron ore concentrates from Minnesota reclamation opera-

tions. The plant was designed to produce 3 million tons per year of high-
quality flux pellets to feed blast furnaces in Ohio and Kentucky. Mesabi
Chief Plant Four, a 2-million-ton-per-year iron ore reclamation plant, was
set to begin operations in the first quarter of 2015.
In February, it was announced that the Empire Mine would remain
open through January 2017 following an extension of supply and joint-
operating agreements. Production rates reached 2 million tons per year for
the Comstock Mountain Lion Mine in Utah, which produces concentrates
for export. One company’s project in Minnesota to construct a 7-million-
ton-per-year open-pit iron ore mine, concentrator, pelletizing plant, and
DRI plant was expected to begin in the second half of 2015 after receiving
financing needed to complete the project. Construction began on a 2-mil-
lion-ton-per-year DRI plant in Texas, expected to be operational by yearend
2015. Weather-related delays on the Great Lakes reduced shipments from
January through April; however, record-high shipments of iron ore were
recorded during the summer months.

The process of pelletization is not a competitor of sintering. On the contrary,
these are complimentary to each other. Pelletization has to be adopted only
in the case of agglomeration of very fine ores, concentrates, etc. since sin-
tering cannot accept such a fine charge unless prior nodulization is carried
out. Transportation of pellets is easier and customary when compared to
sinter. Sinter is generally produced at the blast furnace site to avoid deg-
radation during transportation. Pelletization can be carried out on a small
scale to utilize small reserves of ore fines or concentrates. It is then trans-
ported over any distance to the blast furnace.
CHAPTER
9
CHARACTERIZATION
OF BF CHARGE MATERIALS
In This Chapter
● Introduction
● Room Temperature Physical Properties
● Reducibility
● Physical Behavior of Materials During Reduction at High Temperatures
9.1 Introduction
A single important factor that affects the output and efficiency of a blast
furnace operation is the quality of the charge materials fed into the furnace.
Over the past few decades, researchers have endeavored to develop a bet-
ter understanding of the physical chemistry of the blast furnace process,
and there-from improve the most beneficial characteristics of the burden
materials. It has been discovered that it is virtually impossible to simulate
conditions inside the blast furnace in these tests. Laboratory tests, however,
have been devised to investigate, identify, and quantify, on a relative basis,
an increasing number of burden properties to assess the extent of suitability
of a given material as a burden. The establishment of these tests also helps
to identify those materials which would jeopardize efficient and smooth fur-
nace operation. Consequently, attempts could be made to either improve
the properties of such detrimental materials or to eliminate these from the

burden, as is otherwise economically permissible.
These tests, in general, are empirical in nature and the results of two
different investigations do not readily agree unless extensive standardiza-
tion is sought. The International Standards Organization has been working
in the direction of adoption of rigorously and internationally standardized
tests for the different burden materials.
The properties of burden materials of interest are those properties
which have bearing on its performance during handling, until it is charged
in the furnace, and subsequently on its behavior inside the furnace. These
properties are described here as:
1. Room temperature physical properties;
2. Reducibility;
3. Physical behavior during reduction at high temperatures.
Size of material is an important physical property, and it has already been
described in Section 5.3.
9.2 Room Temperature Physical Properties

These include the physical properties such as the cold strength, resistance
to impact and abrasion, and porosity of the burden materials. The shatter
and the tumbler tests have mainly been devised for this purpose, besides
direct compression strength measurements.
Shatter Tests
The shatter test essentially consists of dropping a certain amount of mate-
rial from a standard height for a certain number of drops. The amount of
the material, retained on or passed through certain sieves, expressed as per-
centage of the original weight, is indicated as the shatter index.
The shatter test that was developed for coke in the U.K. is employed to
determine the strength of sinter and ores as well. A 20 kg (44 1b) sample
of +10 mm size is dropped four times from a height of 2 m. The material is
then screened and the shatter index is expressed as the percentage of mate-
rial greater than 10 mm surviving. A shatter index of less than 1% of −5 mm
is considered to be an indication of strong sinter under Indian conditions.
Characterization of BF Charge Materials • 177
An illustration of the details of the shatter test for sinter produced in the
Bokaro Steel Plant is given below:
Weight of sample dropped = 10 kg
Size of the particles = +25, −40 mm
Height of drop = 2 mm
Shatter index expressed = % of −5 mm produced
Typical screen analysis after the test is as follows:
+10 mm 82–93%
−10 + 5 mm 2–9%
−5 mm 1–2%
Shatter index 0.7–1.0%
The shatter index is related to the tumbling index as:
+10 mm shatter index = 39∙082 + 0∙570 (A.S.T.M. + ¼ inch index).
Tumbling and Abrasion Test

This is a more popular test to assess the strength of sinter (including its abra-
sion resistance) and has largely superseded the shatter test. The A.S.T.M.
test is followed in the U.K. and U.S., and the Micum test is followed on the
continent.
The tumbling test essentially consists of tumbling a standard weight of
material of a certain size in a standard drum; tumbling is carried out at a
standard speed for a fixed number of revolutions. The percentage of mate-
rial passing through or retained on a certain sieve is the index. The standard
conditions under which these tests are carried out are shown in Table 9.1.
TABLE 9.1 Tumbling and Abrasion Test

Standard A.S.T.M. Micum Half Micum Proposed I.S.O
Sample size: Sinter range −2 + 3/8 −40 + 10 mm −40 + 10 mm −40 + 10 mm
Pellets −1½–1/4 in −25 + 10 mm −25 + 10 mm −40 + 6.3 mm
Sample weight 25 lb 50 kg 25 kg 15 kg
Drum dia × length 36 ×18 in 1×1m 1 × 0.5 m 1 × 0.5 m
Number and size of lifters two, 2 in four, 100 mm four, 100 mm two, 50 mm
Speed (rpm) 24–1 25 25 25 ± 1
Standard A.S.T.M. Micum Half Micum Proposed I.S.O

No. of revolutions 200 100 100 200
Screen analysis after test 3/8 in 10 mm 10 mm 6.3 mm
1/4 in 5 mm 5 mm −
30 mesh 2.5 mm 2.5 mm 28 mesh
(595 μm) (500 μm)
Indices % retained on:
abrasion +1/4 in +10 m +10 mm +6.3 mm
dust −30 mesh −2.5 mm −2.5 mm −28 mesh
These tests have shown that the pellets are a high quality product and
sinter a poor quality product with natural ore, in general, falling in between
the two with respect to physical strength.
Degradation of materials and consequent generation of deleterious
fines (−5 mm) is an inherent feature of the behavior of materials during their
handling. The extent to which materials break down during handling has an
important influence on the quality of fines arriving at the blast furnace. If
these fines find their way into the furnace, this will have adverse effects on
the furnace operation. If the fines are screened out, this will involve addi-
tional costs in handling and re-agglomeration. It is also equally necessary
that the degradation of materials during their drop from hopper, via bells, on
the stockline is kept to a bare minimum. No further rectification in the qual-
ity of material is possible at this point. Similarly, the fines generated during
shipping also have an important bearing on the ultimate cost.
It is in these respects that it should be possible to evaluate the behavior
of materials during handling with a suitable test. It is equally important
to lay down proper specifications to ensure minimum degradation during
handling. A drum-tumbling test can be devised to match the breakdown in
the drum with that occurring in a given handling system to ascertain the
behavior of the material in advance. A suitable tumbler index can likewise
be specified to ensure tolerable minimum degradation of a given material
by a given mode of handling.
Although it is not possible to alter the quality of natural ores, the pro-
cess parameters can be adjusted to obtain sinter and pellets of required
handling properties. In general, the higher the tumbling index, the better
the material will withstand the handling process.
The cost of natural, lumpy ore can also be assessed by knowing its rel-
evant tumbler index for the given mode of transportation. Degradation
during handling should be minimized by designing the handling system

such that the drops and movements are as gentle as possible.
Compression Test
The direct measurement of compression strength of iron bearing material
is difficult because of the uncertain geometry of the specimens. The com-
pression strength of pellets, being more regular in shape, can be measured
with better accuracy than those of ores and sinters. A large number of tests
should be carried out, and the maximum, minimum, and mean strength
along with the standard deviation should be indicated.
The L.K.A.B. Laboratories in Sweden have developed a method to mea-
sure the compression strength of any material. A 2 kg material of 10–15 mm
size is dried and placed in cylinder of 200 mm diameter and a load of 100 t
is applied on it via piston. The +5 mm material surviving the test is the com-
pression strength index. A value of 75% is an indication of good strength
and 20% is an indication of friable nature of the material.
Porosity
It is the most difficult property to measure and is expressed as the volume
of pores as a percentage of the total volume of the material tested. Two
types of porosities, open and closed, are recognized. Open porosity is acces-
sible to fluids whereas the closed porosity is not. It is usually measured by
simple method based on the Archimedes principle.
9.3 Reducibility
The reducibility test essentially aims at measuring the rate of reduction of
iron bearing material under blast furnace conditions. Since the conditions
in a blast furnace vary from top to the bottom of the stack, a standardized
apparatus has to be designed to carry out this test under standardized con-
ditions. It is in this respect that the test is most difficult to simulate. The use
of reduction temperature in the range of 900–1000º C and a reducing gas
of CO or CO + N2 of constant composition is increasingly being used. The
sample is a known weight of a certain fixed size and is placed in the furnace
in the form of a static bed of fixed dimensions during the reduction stud-
ies. Several tests have been devised and are followed in various countries.
These are the:
1. Gakushin test;
2. Verein Deutscher Eisenhuttenleute (V.D.E.) method;
3. Centre National de Recherches Métallurgiques (C.N.R.M.) method;

4. Non-isothermal test;
5. Chiba test.
The standardized conditions in all these tests are shown in the form of a
comparative statement in Table 9.2.
TABLE 9.2 Details of Various Reducibility Tests

Specifications Gakushin Chiba V.D.E. C.N.R.M. Non-isothermal
Sample weight g 300–600 500 1000 450 Sufficient
Reducing gas 30% CO 30% CO 40% CO 40% CO 35% CO
70% N2 67% N2 60% N2 60% N2 65% N2
Bed dia mm 60–100 68 – 58 10–15
Gas flow rate 15, 1/min 20, 1/min 3 1000, 1/hr –
5 Nm /hr
Reaction temp C 900 900 900 1000 20–900
Test duration 3 hr 100 min – 1hr –
Index Total time Loss in % loss/min total % Oxygen loss
required for weight until 60% reduction during the test
complete (% oxygen oxida- (loss in (% reduction)
reduction lost) tion state weight)
(until con- survives.
stant weight)
9.4 Physical Behavior of Materials During Reduction at High

Temperatures
The behavior of materials from the time they are charged on the stock-
line until their descent into the bottom of the stock is of direct interest in
assessing the blast furnace operation. During this period the material may
disintegrate by any one or more of the following reasons:
1. Decrepitation;
2. Low-temperature breakdown under reducing conditions;
3. Failure under overlying load at high temperatures;
4. Swelling at high temperatures under reducing conditions;
5. Premature softening of materials.
The permeability of the stack may decrease and affect the furnace opera-
tion adversely. Tests have, therefore, been developed, under as closely sim-
ulated conditions as possible, to measure the tendency of burden materials
to degradation and these are described below:
Decrepitation
When iron bearing materials are suddenly exposed to the exhaust gas
temperature at the stock level on charging, breakdown may occur due
to thermal shock. This is known as decrepitation. Experimentally, it is
measured by dropping a known weight of material in a furnace previously
heated to a temperature level of 400–600º C, under normal atmosphere,
inert atmosphere or under mildly reducing conditions. After the charge
attains the temperature, it is removed, cooled, and sieved to measure
the breakdown.
In a typical test 500 g of 20-; 40 mm size undried ore is dropped in a
furnace previously heated to a temperature level of 400º C and retained
there for 30 min under a flow rate of 5000 liters of nitrogen per hour. The
sample is then removed, cooled, and the percentage of 0.5 mm and −5.6
+ 0.5 mm material in the product is determined by sieving. A typical set of
results are shown below:
Sample % −0.5 mm % −5.6 + 0.5 mm

Hematite (a) 1.1 8.9
Hematite (b) 0.04 0.07
It is believed that ores with more than 10% porosity will not decrepi-
tate. Similarly, a small amount of moisture may also be beneficial. Since the
natural ores do contain varying amounts of moisture the tests should take
into account this variation before arriving at the test procedure.
Low-Temperature Breakdown Test (L.T.B.T.)

It has been observed in the experimental blast furnace that the iron bear-
ing materials do disintegrate at low temperatures under mildly reducing
conditions. That is, in the upper part of the stack, affecting the furnace
permeability and consequently the output adversely. It is believed that
deposition of carbon in this region of the stack is also a contributing factor,
although with sinters, the breakdown has been associated with the presence
of micro-cracks.
In essence, the test consists of subjecting the charge to static bed reduc-
tion at low temperatures in a rotating furnace for a fixed duration. The per-
centage of fines generated is quoted as the L.T.B.T. index. The standardized
conditions in three different tests are comparatively shown in Table 9.3.
TABLE 9.3 Details of Low Temperature Breakdown Tests

Specifications Chiba Jumbo-Linder BISRA
(Japan) (Germany) (U.K.)
Sample weight g 500 Tube full 500
Tube dimensions dia. 20–25 140 140
Length (mm) – 650 650
Lifters – four, 20 mm four, 20 mm
Temperatureº C) of reduction 600 500 600

Tube speed, (rpm) 20 9.3 10
Reducing phase % CO 33 24 24
%CO2 – 16 16
%N2 67 60 60
Flow rate, Liters/min. 20 15 15
Test duration, hr 1 1 1
Index % material of % material of % material of
−1 mm −0.5 mm + 6.35 mm
and % −30 mesh
(0.5 mm)
Based on the amount of fines generated due to the L.T.B.T. index and
the amount of fines accompanying the charge, BISRA has developed the
concept of the bulk fines index to assess the suitability of either the ore,
sinter, and pellets or their combination as a burden in a given blast furnace
practice.
The L.T.B.T. properties are inherent in the materials and hence a
judicious use of sinter or pellets resistant to such breakdown in the stack
can minimize the effect, to a limited proportion, of ore with bad L.T.B.T.
indices.
Reduction Degradation Index Test (R.D.I)

500 gms of dry iron bearing material like ore and or sinter in the size range
of 10–15 mm is charged in a reacting vessel (770 mm L × 75 mm diameter)
made of Inconnel-600 grade steel. The vessel is pushed in an electric muffle
furnace maintained at 550º C. Pure N2 gas is passed at 5 liters/min until the

sample stabilizes at 550º C. Then, a mixed gas of N2 and CO in the ratio of
70 : 30 is passed for 30 minutes at 15 liters/min. Subsequently, the sample
is cooled down to room temperature by passing N2 gas at 5 liters/min. The
reacted sample is rotated in a small tumbling drum (200 mm L × 130 mm
diameter) at 30 rpm for 30 minutes. The sample is screened on 6.3 mm and
3.15 mm screens. The percentage of 3.15 mm fraction is reported as RDI
of iron bearing material.
Reducibility Test (R.I.) on Iron Bearing Materials

All the above methods are followed, but the temperature of the sample is
raised to 900º C and mixed gas (N2: CO in the ratio of 70: 30) at 15 liters/
min is passed for 3 hrs. The cooled sample is weighed to find out the %
weight loss against 500 grams (initial sample). RI is calculated as follows:
RI (Reducibility Index) = % Wt. loss × 100

(Fe(t) − FeO) × 48/112 + FeO × 16/56
Coke Reactivity Test (C.R.I.) and Coke Strength after Reaction (C.S.R.)
200 grams of dry coke of 19 mm to 21 mm size is placed in the reacting ves-
sel (450 mm L ×; 75 mm diameter) made of Inconnel-600 grade steel. The
vessel is pushed in an electric muffle furnace kept at 1100º C temperature.
Pure N2 is passed at 5 liters/min until the sample temperature stabilizes at
1100º C. Then pure CO gas is passed for 2 hrs at 5 liters/min. Subsequently,
the sample temperature is brought down to room temperature by pass-
ing N2 gas at 5 liters/min. Cold reacted coke is weighed and the % weight
loss against initial weight of 200 grams is reported as the CRI (i.e. Coke
Reactivity Index). Then, the reacted coke is rotated in an I/drum (770 mm
L × 130 mm diameter) at 20 rpm for 30 minutes. Tumbled coke is screened
on a 10 mm round screen. The percentage of +10 mm fraction is reported
as CSR (i.e. Coke Strength after Reaction).
Hot Compression Strength

The compression strength of burden materials falls rapidly with increasing
temperatures. A cold compression strength of 150–300 kg for certain pel-
lets falls to as low as 25 kg at 1000º C. This means that practically half-way
down the furnace stack, the pellets lose much of their strength. Materials
with low hot compression strength would breakdown under load in the
lower part of the furnace stack and affect the stack permeability adversely.
Knowledge of this property is, therefore, beneficial. The hot compression

strength can be measured directly in a compression testing machine pro-
vided with such facilities. A large number of tests should be performed to
obtain the average correct value.
Swelling
Some materials, in particular the pellets, show a maximum in their vol-
ume change with increasing degree of reduction. This is due to swelling of
the pellets. If it is excessive, it can give rise to serious troubles in the fur-
nace operation. Swelling can be measured for individual particles, but bulk
swelling data are of practical use.
In the C.N.R.M or the BISRA tests for swelling, four samples of 60 ± 2 g
each are used. Their apparent volume is measured by immersion in mer-
cury. The samples are then reduced to different degrees in a static bed
under the following conditions:
Temperatures 1000º C
Reducing gas 40% CO and 60% N2
Gas flow 1000 liters/min
Pre-heating under nitrogen atmosphere
The percentage reduction is ascertained from the percentage weight
loss or chemical analysis. The apparent volume is again measured, and the
swelling is correlated to the degree of reduction in a graphical form.
Softening of Materials in the Blast Furnace

During its descent in the furnace stack, the burden comes to a temperature
level where it softens. The charge becomes sticky and bed permeability
decreases resulting in a considerable pressure drop in this zone. The wider
the sticky zone, the greater the pressure drop is in the furnace, and the
more the furnace operation gets affected adversely. The earlier the soft-
ening commences, the wider the softening zone is, and consequently, the
wider the sticky zone is in the furnace. The softening behavior also varies
with the degree of reduction. It is, therefore, necessary to measure the soft-
ening temperatures of burden materials under reducing conditions in order
to assess their suitability as burden materials.
The C.N.R.N. Laboratories have developed a test for this purpose.
Materials at known states of reduction are heated in a neutral atmosphere
under a carefully programmed temperature cycle while being subjected to a

static load of 2 kg/cm2. The temperature at which softening commences and
ends were conveniently defined as those corresponding to shrinkages of 3%
and 25% of the initial bed height as measured by a dilatometer. The soften-
ing start and finish temperatures are plotted against degree of reduction.
Materials with less than 1200º C as the starting of softening were observed
to be far more detrimental to furnace operation than those with more than
1200º C as the softening-start temperature. Softening temperatures vary
with the nature of the iron bearing material, its iron content, initial state of
oxidation, basicity ratio, etc. In particular, the alkali content could have a
marked effect on the softening temperatures.
High Temperature Permeability Under Reducing Conditions and

Applied Load
The swelling, softening, disintegration, or decrepitation of the burden
materials may occur to a lesser or higher extent individually, but all put
together, they affect the stack permeability adversely. It would be worth-
while to devise a single test to assess the effect of all those characteristics
on the ultimate permeability of the stackt. The apparatus used for measur-
ing the reducibility has been recommended with several modifications for
this test as well. The test includes the measurement of the reduction, the
pressure drop across the bed, and the volume shrinkage simultaneously.
The materials can thereby be assessed on a relative basis regarding their
suitability as a blast furnace burden.

The testing of materials is not only useful to assess the suitability of a given
material as a burden, but it can lead to several suggestions regarding the
proportion of various constituents, like natural ore, sinter, and pellets to be
incorporated in preparing the best burden under given conditions. Tests
can also indicate the directions in which the properties of a given material
need to be improved for making it a more suitable burden material.
PHYSICAL-THERMAL-
CHAPTER
10
CHEMICAL PROCESSES
IN A BLAST FURNACE
In This Chapter
● Introduction
● The Blast Furnace Profiles
● Thermodynamic Equilibria
● Physical Chemistry of Blast Furnace Process
● Alumina Problem
● Kinetics of Reactions in Bosh and Hearth
● Efficiency of the Blast Furnace Process
● Quality of Hot Metal
● TiO2 Addition in Burden
● U.S. and World Events, Trends, and Issues. Source: USGS 2017
10.1 Introduction
Although iron production in shaft furnaces of various designs has been
occurring for at least several centuries, it was more as an art rather than a
practice based on the scientific understanding of the process. The practice
was improved upon over the years, but the understanding of the BF process
is only of recent origin. This understanding of the fundamental nature of
the BF process is of paramount importance in order to develop the practice
to its maximum efficiency and economy. Much effort has, therefore, been
directed towards establishing the various physical, chemical, and thermal
processes occurring in the blast furnace. A lot of laboratory experimenta-
tion to determine the basic data and its correlation with the practical obser-
vation, has taken place in the recent past. This research helped to improve
upon the practice with respect to material preparation, operational changes,
and modifications in blast furnace design and practice.
Small experimental furnaces have been installed at various places to
study the process directly and to establish the effects of practical variables
on the quality and efficiency of iron production. Similarly, full scale com-
mercial blast furnaces have been subjected to experimentation, particularly
towards the end of their useful campaign. One such furnace was quenched
at the end of its useful campaign by pouring water over it. Researchers
were able to study the physical structure of the charge inside the operating
furnace.
All this research has led to a fairly accurate insight into the physical-
thermal-chemical processes occurring in a blast furnace. In this chapter, the
fundamental nature of the blast furnace process, as is generally accepted, is
discussed in its simplified form.
10.2 The Blast Furnace Profiles

Thermal, Physical, and Chemical Profiles
The main chemical reactions that occur in a blast furnace have already been
indicated in Chapter 2 in the form of Equations (2.1) to (2.6). The same
may be recapitulated here since their understanding would help the reader
to understand the following. The reduction of iron oxide is a high tem-
perature process and the thermal requirements of the process are met by
burning coke, injections if any, and the enthalpy of the hot blast. The maxi-
mum temperature is, therefore, generated in front of the tuyeres where
hot blast burns preheated coke. This temperature may be as high as 2000
degrees C. The temperature inside the furnace decreases on either side of
the tuyere level. The exhaust gas temperature is as low as 200 degrees C
and the temperature in the hearth may be around 1300–1500 degrees C
depending upon the chemistry of the slag; a typical temperature is 1350
degrees C. The temperature variations inside the furnace along the vertical
axis are shown in Figure 10.1. The pressure and gas composition profiles
Physical-Thermal-Chemical Processes in a Blast Furnace • 189
are shown along with the temperature profiles. The horizontal variations
in temperature are shown on the vertical profile of the blast furnace. The
gas composition and temperature vary smoothly over the vertical as well
as the horizontal cross section of the furnace. The chemical processes are
therefore continuous from bottom to top and throughout the volume of
the furnace.
FIGURE 10.1 Temperature, pressure, and gas composition profiles in a blast furnace.
(a) Shows the profiles of temperatures in degrees C in a vertical section, (b) and (c) show the profiles
along the vertical axis of the furnace.
Physical Structure
It is now universally agreed that the blast furnace, in its working state, con-
sists of broadly six zones as shown in Figure 10.2. These are the:
(a) Granular zone where the entire charge is solid, (i.e. the stack).
(b) Cohesive zone where, except coke, the rest of the charge is semi-
solid, or is softened and at the lower end the slag is molten. This slag
is rich in FeO and melts at much lower temperature.
(c) Active coke zone where coke actively takes part in the reactions
involving direct reduction of iron oxide from the trickling slag and
its own interaction with carbon dioxide (i.e. mainly the bosh region).
(d) Tuyere zone is where coke burns in front of the tuyere by combin-
ing with the hot blast (oxygen) effectively to CO.
(e) Stagnant coke zone where the coke is still solid and that supports
the overlying entire burden and through which the slag and the met-
al trickles and metal gets saturated with carbon.
(f) Hearth zone where the slag and the metal clarify into two layers, and
they interact with each other with respect to the chemistry to move, as
far as possible, towards equilibrium.
The hot blast enters the furnace through the tuyeres at a velocity of
around 200–300 m/sec and at a pressure of 2–4 atmos. Such a high pres-
sure is necessary to push the reducing gases through the solid burden and
to overcome the top pressure. These conditions create a flame raceway
of 1–2 m in front of each tuyere which can be easily penetrated by a bar
pushed through a tuyere. Each raceway, at its sides and below, is bound by
rather firm regions of lump coke which has bypassed oxidation during its
descent through the furnace. Studies of quenched blast furnaces indicate
that this coke extends downwards into the iron pool and perhaps even to
the hearth bricks. This is what is shown in Figure 10.2. as stagnant coke
percolator.
The top of raceway region is bound by loosely packed lump coke since
gas has to flow upwards through it. The bottom most pieces in this region
periodically fall into the raceway and are consumed by the hot blast. The
coke bed, by and large, slowly moves down, gets consumed and is replen-
ished from top. The above description indicates that almost all the solid
material in the hearth and bosh is coke alone. Liquid iron and slag per-
colate through this and collect in the hearth. The hot gases rising from
the raceways heat up the descending coke, liquid slag, and metal. During
percolation, reduction of iron oxide is completed and it gets saturated with
carbon. Partial reduction of other oxides, like those of phosphorus, silicon,
manganese, etc., also occur and make the metal impure.
FIGURE 10.2 Idealized internal structure of a blast furnace in operation. The most notable feature of
this structure is the fusion zone with its alternate layers of 0.5–1.0 m thick of (i) coke and (ii) fused slag
plus iron. The moving active zone of coke through which the gases ascend, and the stagnant coke bed
through which liquid slag and metal percolate, are also shown.
Studies of commercial quenched blast furnaces have shown that the

region of loose packed coke above the raceways is bounded on top by
inverted U-shaped fusion zone. This consists of alternate layers of coke and
other charge materials as developed at the stockline level by way of charg-
ing sequence. The structure is shown in Figure 10.2. The inverted U-shape
is characteristically an important feature. It supports the furnace burden,
aided by the gas pressure from below. Until all the charge is solid, the entire
burden is permeable to ascending gases. In the fusion zone, the ore, flux,
etc. melt, and the permeability of the bed is only due to the existence of
alternate coke slits as shown in Figure 10.2. The gas flow, therefore, tends
to become horizontal and the coke quality must ensure this permeability
during the operation. As the charge melts in the fusion zone, it trickles
down through the coke bed and the coke slits slowly become a part of the
moving active zone. The slag, while percolating through the moving active
zone of coke, absorbs the ash released from the coke on burning. Any iron
oxide contained in the slag is reduced fully during percolation through the
stagnant coke percolator. The structure of the burden above the fusion zone
is practically the same as it is obtained at the stockline level by way of charg-
ing sequence. The reduction of iron oxide there is mainly by gas (i.e. indi-
rect reduction).
The charge, almost through half of the total furnace height, remains
fully solid and, therefore, maintains the bed permeability for the hot reduc-
ing gases to pass through. This is the granular zone. This is not in the form of
a horizontal slice of the furnace, but it has a shape as shown in Figure 10.2.
The structure of the bed, as created through charging at the top, remains
more or less in tact until this stage. Below this level is the semi-solid cohe-
sive state below which the bed permeability is maintained by coke alone,
since it descends in solid state right into the hearth. The properties of coke,
particularly the CSR and the RI, are of paramount importance to run the
furnace efficiently. Any premature breakdown of coke therefore affects the
operation adversely to that extent.
As the charge melts in the cohesive zone, it trickles down through the
coke bed and the coke slits slowly become part of the moving active zone.
The slag, while percolating through the moving active zone, absorbs the
ash released from the coke on its burning. The initial slag formed in the
cohesive zone is rich in FeO, and as it percolates through the coke slits,
it is directly reduced to iron. The iron also picks up carbon from coke and
its melting point is reduced considerably. The interaction of iron with CO
results in dissolution of carbon in iron. The slag must form within a very
narrow range of temperature. If it forms over a wide temperature range, it
interferes with the bed permeability and that is not beneficial for operation.
Use of fluxed sinters and formation of slag in sintering itself helps in this
respect considerably. The reduction of charge in the solid state is purely a
gas-solid interfacial phenomenon. It is an indirect reduction by gases like
CO or perhaps coupled with hydrogen, if available.
10.3 Thermodynamic Equilibria

Carbon-Oxygen Equilibria
This is of interest since the carbon of the coke is burned by oxygen of the air
blast inside the furnace to generate heat. The free energy data of relevant
reactions are as follows:
< C >g + {O2} = {O2} (10.1)
ΔGº(10.1) = −94,200 −0.2T cal/mole
and {CO2} + < C >g = 2 {CO} (10.2)
where the suffix g indicates graphite as its standard state. The forward reac-
tion in Eq. (10.2) is referred to as a solution loss reaction, that in backward
direction as Naumann reversion and the equilibrium as a whole is called
Boudouard equilibrium by the blast furnace technologists.
Carbon can also burn at high temperatures as
2< C >g + {O2} = 2{CO} (10.3)
ΔGº = −53,400 − 41.9T
and subtraction of Eqs. (10.1) and (10.3) gives Eq. (10.2)
where ΔºG(10.2) = +40,800 − 41.7 T
−ΔGº 8910
therefore log K (10.2) = =− + 9.11
4.575T T
The equilibrium CO/CO2 ratio in contact with graphite at various tem-
peratures has been computed from the above data as follows:
Temp degrees C 1900 1460 1180 980 820 670 530 410 300
5 4 3 2 −2
CO/CO2 10 /1 10 /1 10 /1 10 /1 10/1 1/1 1/10 1/10 1/10−2
It is, therefore, clear that at temperatures above 1000 degrees C, the

proportion of CO2 in the gas phase in equilibrium with graphite is negligi-
bly small, and similarly, below 400 degrees C the concentration of CO is
negligibly small. Between these temperature limits, the proportion of both
are significant. The oxygen potential of the gas phase can be conveniently
represented by CO/CO2 ratio in the gas phase. If a gas phase containing
CO and CO2 is in contact with iron oxide, the possibility and extent of its
reduction could be ascertained from the knowledge of its equilibrium CO/

CO2 ratio and the temperature.
Equilibria in Fe-C-O and Fe-O-H Systems

The equilibria between the various oxides of iron and metallic iron are
shown in Figure 10.3 in the form of the relevant Oxide Ellingham diagram.
Since the reducing power of the gas in a blast furnace is generally repre-
sented by its CO and CO2 contents, the same equilibria can be conveniently
represented in the form of reactions as:
3< Fe > + {CO2} = < Fe3O3 > + {CO} (10.4)
3< FeO > + {C02} = < Fe3O4 > + {CO} (10.5)
< Fe > + {C02} = < FeO > + {CO} (10.6)
The equilibrium of Fe2O3/Fe3O4 is not included here because actual data
are not available; only the hypothetical data are available and Fe2O3 does
not actually change to Fe3O4 but a phase of Fe3O4 containing Fe2O3. In
practice, Fe2O3 is readily reduced to Fe2O4 by a gas containing a very small
amount of CO at relatively low temperature.
FIGURE 10.3 Oxide Ellingham diagram for oxides of iron and carbon.
The standard free energy changes of reactions as in Eq. (10.4) to (10.6)

can be computed from the Ellingham diagram by algebraic summation of
Δ G° values of the relevant simpler reactions.
For example 2< Fe > + {O2} = 2 < FeO > (10.7)

and for which ΔG°(10.7) = −125,700 + 30.69 T cal/mole.
Hence ΔG°(10.6) = 1 ΔG°(10.3) + 1 ΔG°(10.7) − ΔG°(10.1)
2 2
= 4650 + 5.40T cal/mole
The equilibrium CO/CO2 ratio for reaction in Eq. (10.6) at any temperature
can be computed as:
a < FeO >·Pco
K(10.6) =
a < FeO >·Pco2
Pco
= for solid Fe and FeO
Pco2
ΔG°(10.6)
also K(10.6) log =
4.575T
1018
i.e. Pco/Pco2 = − + 1.181 (10.8)
T
from whence the equilibrium CO/CO2 ratio can be calculated.
The data in the Ellingham diagram can thus be replotted as equilib-
rium CO/CO2 ratio vs. temperature for the Eq. (10.4) to (10.6) as shown in
Figure 10.4. The CO/CO2 ratio in equilibrium with solid carbon is also shown.
From the Ellingham diagram, the relative stability of the various oxides
of iron can be assessed and the removal of oxygen from iron oxide by CO
can be worked out from Figure 10.4.
FIGURE 10.4 Equilibrium CO/CO2 ratio in contact with carbon and oxides of iron at various temperatures.
The normal blast furnace gas composition is also superimposed for comparison.
Both these figures indicate that the reduction of iron oxide takes place
in the following sequence:
< Fe2O3 > – → < Fe3O4 > – → < Fe > below 570 degrees C
< Fe2O3 > – → < Fe3O4 > – → Í < FeO > – → < Fe > …above 570 degrees C
since wustite (i.e. FeO) is a stable phase only above nearly 570 degrees C.
If hydrogen is available as a reducing component of the gas phase, the
equilibrium of various oxides with a H2/H2O ratio follow a similar pattern to
that for CO and is shown in Figure 10.5.
FIGURE 10.5 Equilibrium H2/H2O ratio in contact with oxides of iron at various temperatures.
It should be clear from the two Figures 10.4 and 10.5 that, thermo-
dynamically, a lower value of CO/CO2 ratio than that of H2/H2O is ade-
quate to effect reduction at relatively lower temperatures while a higher
CO/CO2 ratio than H2/H2O ratio is required to effect reduction at higher
temperatures.
Let us now calculate the equilibrium CO/CO2 ratio in contact with FeO
and solid carbon at the same temperature (say 950 degrees C) for compari-
son. The equilibrium CO/CO2 ratio in contact with solid graphite can be cal-
culated from the free energy data for Eq. (10.2). It is in proportion to nearly
100% CO and negligible amounts of CO2 at 950 degrees C. Similarly, the
equilibrium CO/CO2 ratio in contact with solid wustite (approx. FeO) can
be calculated from the free energy data for Eq. (10.7) given in Eq. (10.8)
and is in proportion to 70% CO and 30% CO2 at 950 degrees C. These can
also be obtained from the Figure 10.4. The equilibrium CO/CO2 ratio in
contact with carbon and wustite are of direct interest in understanding the
reduction of iron oxide by the gas phase inside the blast furnace.
Elements of Stoichiometry
In a blast furnace, iron oxide plus coke are introduced from the top and oxy-
gen, in the form of blast, through the tuyeres. The oxygen part of this charge
goes out of the furnace only through the top via gas phase, and the iron
only leaves as liquid metal from the bottom. The carbon part of the charge
mainly leaves the furnace in the form of top gas. A small part, however, goes
with the metal as dissolved carbon. In essence, the amount of coke and blast
required to produce a ton of iron should, therefore, be known. Similarly,
the effect of alterations of any parameter(s) on these must also be known.
Mathematical equations have been developed to describe these. Under the
simplified conditions that iron joins the metal phase alone, and that other
side reactions do not occur to interfere with the mass balance in a blast fur-
nace under steady state conditions, the reaction can be described as:
Item No. of moles No. of moles Equations depicting balance:

entering furnace leaving furnace per mole of product Fe
Iron niFe noFe niFe = niFe = 1
Carbon n iC noC niC = noFe = No. of moles of C in gas phase since

No. of moles of C in gas phase plus No. of moles
of C in iron phase equals total No. of C entering
the furnace, therefore, ngC + nmFe = noC
No. of moles of oxygen entering the furnace is
in the form of oxygen in blast and in iron ore
together. Hence,
Oxygen niO noC niO = niB + niX(ore)
Since all oxygen leaves the furnace as gas as CO
and CO2 the ratio of C and O in gas as C/OG shall
be a molar ratio in gas phase. Hence,
niO = n0O = (O/C)G
From the above it should be obvious that:

niB + niX = ngC. (O/C)G (10.9)
It is convenient to divide the functions of the input carbon into the
active moles, nAC, which react with oxygen in the blast and iron oxide, and
the moles of inactive carbon simply dissolve in the iron, both per mole of
iron product. Also, all the active carbon ends up in gas phase as CO and
CO2. Therefore, moles of active C equals moles of C in gas phase, or
nAC = ngC
Eq. (10.19), therefore, becomes
niB + niX = nAC .(O/C)G (10.10)
FIGURE 10.6 The Rist Diagram showing the operating line for a simplified blast furnace operation
defined by Eq. (10.10).
It means:
No. of moles of oxygen in the blast No. of moles of active carbon
+ multiplied by
=
No. of mole: of oxygen in oxide +
(all per mole of product Fe). No. of moles of oxygen in gas
(all per mole of product Fe).
Equation (10.10) is the stoichiometric equation for the blast furnace in
its most direct and simplest form.
The same observations have been represented graphically in the form
of the Rist Diagram in Figure 10.6 and the slope of the straight line is
equivalent to:
(no)B +(O/Fe)X
(O/C)g
which is nothing but the value nAC, (i.e. active carbon required to produce
one mole of product Fe). In other words, it is the coke rate of the blast
furnace under the simplified ideal conditions of its operation. The details of
the additional reactions such as calcination, reduction of oxides like MnO,
SiO2, P2O5, etc. can suitably be added to elaborate Eq. (10.9) in detail to
describe real stoichiometry. The changes in the quality of the iron oxide
burden and the quality of coke can be suitably incorporated in arriving at
the appropriate stoichiometry.
10.4 Physical Chemistry of Blast Furnace Process

The physical chemistry of processes occurring inside the blast furnace at
various places can now be discussed in the light of the above thermody-
namic information. The kinetic factors affecting these will be discussed
later.
Reactions in the Tuyere Zone

Coke is the only constituent of the burden that descends down into the
tuyere zone in the form of a solid column. During its descent, it gets pre-
heated to nearly 1500 degrees C by the time it reaches the tuyere level.
Such a highly preheated coke burns in front of the tuyeres with great inten-
sity and turbulence, when it comes in contact with a preheated blast at
around 800–1300 degrees C, it generates temperatures of the order of

1900–2000 degrees C. The combustion zone is in the form of a pear shape,
called the raceway in which the hot gases rotate at high speeds carrying
a small amount of burning coke in suspension. The situation in a blast
furnace tuyere zone is practically an adiabatic system as defined in clas-
sical thermodynamics. The burning of coke in front of the tuyere zone is
almost instantaneous, and the theoretical adiabatic flame temperature in
the raceway can be calculated without any difficulty. This is known as the
raceway adiabatic flame temperature (popularly called RAFT in abridged
form). This is the highest temperature available inside the furnace. There is
a temperature gradient in the vertical direction on either side of this zone.
This temperature is critically related to the hearth temperature known as
the operating temperature of the furnace. It is equally related to the top
gas temperature such that the hot raceway gases have to impart their heat
to the descending burden to the extent expected and leave the furnace as
off-gases at the desired temperature.
The raceway is a vital part of the blast furnace since it is the heat
source in a gigantic reactor, and at the same time, a source of reducing gas.
The structure of the raceway largely depends upon the quality of coke. It
has been proven, through coke combustion tests, that coke containing a
large amount of fines tends to raise the pressure in the raceway because
of decreased permeability. This leads to unbalanced pressure near the
raceway and an unstable raceway shape, while the coke containing smaller
amount of fines exhibits excellent permeability, forming a stable raceway.
The volume and depth of the raceway, therefore, deteriorates when the
quality of coke deteriorates. For efficient operation, the raceway should
penetrate the charge to such an extent that finally, after the flow becomes
vertical, it should cover the entire cross-section of the furnace. Any lack of
penetration of the raceway might lead to decrease in temperature along the
vertical axis of the furnace, a situation detrimental to furnace operation. A
good permeability of charge, achieved by the use of coke of right size and
quality, can only ensure a proper raceway.
The main reactions in the tuyere area is the combustion of coke as per
Eq. (10.1), but in the presence of excess of coke, reactions as in Eq. (10.2)
take place thereby amounting to the overall process as shown in Eq. (10.3).
Although the coke rate varies from furnace to furnace, and it is as low
as 450 kg/t of hot metal produced in some of the very efficient furnaces, it is
generally around 600–750 kg/t of metal product. Major portions of this coke
burns at the tuyere level to provide a major portion of the heat, (enthalpy)
necessary for the process. A certain proportion of the coke is consumed in

the stack in a solution loss reaction as in Eq. (10.2). About 20% to as high as
40% of the heat requirement of the process is, however, met by the sensible
heat of the blast.
There is always some moisture in the blast; it is higher in the rainy
season and less in the summer. Moisture reacts with hot coke in the tuyere
area as:
{H2O} + <C> = {CO} + {H2} (10.11)
thereby generating additional reducing agents in the form of hydrogen. The

reaction in Eq. (10.11) is endothermic in nature. However, for every unit of
coke burned, it generates more units of reducing gas. If coke is burned by
air alone, it generates one unit of CO, but if it is burned with H2O, it gener-
ates one unit of CO and one unit of H2. If extra heat is available in the tuyere
area, for example, by way of the extra hot blast temperature, it can be made
advantageous by adding extra moisture in the blast. This is precisely the
principle of humidification of blast in modern blast furnace practice.
Coke is a costly as well as scarce fuel, and in many furnaces, a part of the
fuel requirement is met by the use of alternative cheaper fuels. Natural gas,
coal or coke slurry, oil, tar, etc. have been successfully used in this respect. The
fluid fuel is injected through the tuyeres along with the preheated blast. The
fuel injection is quite efficient if the blast temperature is above 900 degrees
C. Many of these fuels that are injected contain hydrocarbons. In addition to
providing heat, they also generate hydrogen as an additional reducing agent.
The hot reducing gases produced in the combustion zone contain pre-
dominantly nitrogen which is quite inert as far as the blast furnace chem-
istry in concerned. Nevertheless, the large volume of gases are not only
useful, but these are essential in transferring the heat of combustion to the
upper part of the furnace for preheating and finally melting the charge in
the bosh region.
Unfortunately, combustion also releases the sulfur present in the coke,
which is practically all absorbed by the charge in the stack. It is carried
down into the hearth where it gets distributed between the slag and the
metal phases. The sulfur content of coke may be as high as 1.5%, and it
usually accounts for nearly 90% of the total sulfur entering the furnace. The
sulfur of the coke or any other fuel, injected through the tuyeres, if any, is a
serious problem. High sulfur coke or other fuels cannot, therefore, be used
in blast furnaces.
Reactions in Stack
The functions of a blast furnace stack are two-fold. First, it should progres-
sively reduce the iron oxide to iron using carbon monoxide of the gas phase.
Secondly, it should extract sensible heat from the ascending gases and thereby
get itself heated. Ideally, the gases at the stockline should be at atmospheric
temperature and the CO/CO2 ratio should be in equilibrium with iron oxide as
3Fe2O3 + CO = 2Fe3O4 + CO2 (10.12)
It would ensure maximum thermal efficiency at minimum of coke (total
fuel for that matter) consumption.
The effluent gases at the stockline are usually at around 200 degrees C
and 1.3 atmospheric pressure with the CO/CO2 ratio usually in the range of
1.5–3.0. The equilibrium CO/CO2 ratio for Eq. (10.12) is extremely small.
Even if reaction as in Eq. (10.4) or (10.5) were to achieve equilibrium at
the stockline, the CO/CO2 ratio is not much. The reducing potential of the
gases is, therefore, not utilized fully in actual practice. The furnace is oper-
ated at a higher coke rate than is theoretically necessary for the reduction of
ore. It is true that the blast furnace gas, because of its CO content, is used
as a fuel outside the furnace, but it is hardly economical to produce a low
grade gaseous fuel from an expensive fuel like metallurgical coke. Hence,
every effort in design and operation of the furnace must be made to reduce
the CO/CO2 ratio in the exhaust gases.
The relevance of CO/CO2 ratio of 2.3 in equilibrium with Fe/FeO and
nearly cent-percent CO in equilibrium with solid carbon at 950 degrees C
will now be clear with reference to the blast furnace operation.
During the ascent, if a volume of reducing gas comes in contact with
iron oxide at 950 degrees C, it would tend to move towards a CO/CO2 ratio
of 2.3 as per Eq. (10.6). If the same gas later on strikes a piece of coke, its reduc-
ing power will be regenerated and the CO/CO2 ratio tends towards a value as
per Eq. (10.2) (i.e. practically all CO at 950 degrees C). The cycle of reduction
and regeneration can theoretically be repeated indefinitely. In blast furnace prac-
tice, the CO/CO2 ratio is also affected in the temperature range of 800–1100
degrees C by the calcination of limestone as in Eq. (2.5). If the CO2 content
thus increases beyond the equilibrium value as per Eq. (10.6), oxidation of
Fe to FeO may take place. This is an endothermic reaction and requires heat.
The situation in an actual operation can be understood clearly by follow-
ing the changes in the composition of a small volume of gas from the time of
its entering the furnace as air blast to the time when it is discharged from the
furnace top. This is illustrated in Figure 10.7. The air blast enters the furnace
at around 1000 degrees C, and the oxygen potential is normal atmospheric
(PO2 = 0.21 atmos) which is shown as the starting point of the curve in Figure
10.7. The oxygen then combines with coke to effectively form CO and devel-
ops a temperature of the order of 2000 degrees C. This change is shown by
the sharply falling curve. The lower most point on this curve will be close
to C/CO line at 2000 degrees C and is the lowest attainable potential under
the prevailing conditions. The gases pass through the furnace at tremendous
velocities and the retention time may be only few seconds at best. Therefore,
before equilibrium as per Eq. (10.3) is attained at 2000 degrees C, the gas
probably comes in contact with iron oxide which is reduced, thereby raising
the oxygen potential towards the appropriate iron oxide line. Since the bur-
den is already getting reduced during its descent, it is most likely that only
FeO and no other higher oxide of iron exists at the lower part of the bosh.
The upper limit to which the oxygen potential of this volume of gas is raised
is the Fe/FeO line. The raising and lowering of the oxygen potential of the
gas is determined by the alternate contacts with iron oxide and coke particles.
FIGURE 10.7 Composition changes in a volume of gas as it goes through the blast furnace as a func-
tion of temperature, as depicted in the form of oxygen potentials.
The mean blast furnace gas composition, as estimated from experimen-

tal furnaces, is such that it is capable of reducing iron oxide at all tem-
peratures. The corresponding range of reduction is therefore shown in
Figure 10.7 by the hatched region.
The magnitude of the composition fluctuations of the gas in the furnace
diminish progressively as the two limiting lines (i.e. Fe/FeO and C/CO)
converge with decreasing temperature, and below about 1100ºC, they lie
in the range shown by hatched band. Since there is insufficient time for the
reduction to take place at lower temperatures, the oxygen potentials are
lower than even the CO/CO2 line. Finally, the gas is discharged from the
furnace at a point close to below Fe/Fe3O4 line at 200°C.
In the former discussion, it has been assumed that the iron oxide is
present as a pure phase and is not chemically combined with other oxides
that may be present in the ore. If it is combined, for example, with an oxide
like iron silicate, its activity will be lowered considerably below unity and
the equilibrium CO/CO2 ratio would work out to be much higher. This
problem is serious, particularly if sintering or pelletizing is adopted, since
there is danger of silicate formation in these processes.
Below 800 degrees C, the CO content of the gases is more than the
value in equilibrium with iron oxide or carbon. The conditions are prone to
deposition of carbon or the Naumann reversion reaction (10.2) in reverse
can take place in the upper part of the stack. Fortunately, but for the slug-
gishness of this reaction, the stack would get chocked with soot deposition.
A certain amount of deposition does take place in the temperature range
of 300–600 degrees, and this soot returns to the higher temperature range
of active zones at lower levels of the furnace. The presence of iron and iron
oxide act as catalyst for the deposition reaction. This can be hazardous for
the fireclay refractories in the upper part of the furnace.
Besides of main reduction (i.e. reduction of iron oxide), higher oxides of
manganese like MnO2, Mn3O4, etc. are reduced to lower oxides like MnO in
the lower part of the stack according to
< MnO2 > + {CO} = < MnO > + {CO2} (10.13)
ΔH° = −35,350 cal/mole
< Mn3O4 > + {CO} = < MnO > + {CO2} (10.14)
ΔH° = −12,250 cal/mole

Further reduction of MnO to Mn is possible only by carbon at higher

temperature levels. The kinetics of iron oxide reduction in the stack is of
special significance and is discussed below.
Kinetics of Iron Oxide Reduction

Since the gases pass through the furnace at tremendous velocity, attainment
of equilibrium between the gas and the charge at any level in the furnace is
out of the question. According to some estimates, the residence time of gas
inside is of the order of only a second. The gases, therefore, must react with
the solid charge as rapidly as possible in order to fully utilize their reducing
ability.
The rate of reduction of iron oxide has to be regarded from the point
of view of an individual iron ore particle, and the reduction of particles in
the form of a permeable bed. The mechanism and kinetics of reduction of
iron oxide particle by hydrogen, CO or a mixture of the two has been inves-
tigated extensively. It is now universally accepted that the reduction takes
place as follows:
Fe2O3 → Fe3O4 → FeO → Fe above 570°C
which means that there are Fe2O3/CO, Fe3O4/CO, and FeO/CO as gas/solid
interfaces, at which the indirect reduction is taking place. If a spherical
specimen of oxide is reduced by any reducing gas then, after partial reduc-
tion, the particle may have concentric rings of Fe, FeO, Fe3O4 and core as
Fe2O3 quite visible even to the naked eye. On a macro level, these rings
are very clear, but on micro-level, the same rings are not all that sharply
differentiated from each other because of individual particles and grains
constituting the spherical specimen.
Every iron ore particle inside the blast furnace is also reduced in the
same way in the stack; the charge is solid and the ascending gases mainly
contain CO as reducing agent. Any hydrogen present in the gas will also
reduce the oxide in the same way. This is known as indirect reduction.
Carbon of coke does not reduce oxide in the stack part. This is because the
system then resembles a solid/solid interfacial reaction which is extremely
slow to contribute substantially in reducing the ore.
In the cohesive and the bosh regions, where the FeO rich slag trickles
through the coke bed, carbon reduces the FeO in the liquid slag directly,
although the reaction may takes place by coke reacting directly with the
FeO but more predominantly through
2{CO} → {CO2} + < C > as earlier (10.2)

and (FeO) + < C > → Fe + {CO} (10.15)
This precipitated carbon in fact contributes in direct reduction of oxide

in the cohesive and the upper bosh regions. The CO2 dissolves carbon to
recuperate to CO. The factors that affect the rate of reduction of iron oxide
need to be understood in order to devise ways and means for improving
them. These factors are broadly indicated below:
1. Physical nature of iron oxide and its previous history.
2. Nature and composition of gangue in the charge.
3. CO content of the gas phase at the level in question.
4. Temperature of the charge and the gas at the level in question.
5. Chemical nature of iron oxide.
6. Pressure at the top of the furnace.
7. The extent of solid-solid reaction.
Physical Nature of Ore and Its Previous History

Reduction of oxide of iron is broadly a diffusion controlled gas-solid reac-
tion. The rate of reduction is increased with decreasing size and increasing
porosity of individual particles. The shape of the particle is important only
when particles are smaller than about 0.25 mm in size. The burden perme-
ability decreases with increasing percentage of fines in the charge. Hence,
the lower the permeability, the lower its reduction is. The more uniform the
size and the higher its porosity, the more efficient is its reduction.
The ore particles should be reduced completely, as far as possible, in
the stack itself, (i.e. within the time required for its descent from the stock-
line to the bosh level) under the operating conditions of the furnace. The
maximum size that can be reduced before the ore reaches the bosh region
is determined largely by its chemical composition and porosity. Usually a
mix-size of 50 mm is considered satisfactory and the lowest size is 20 mm.
The natural ore often lacks necessary size range, strength, and porosity
as required for an efficient blast furnace operation. It is a common practice,
therefore, to agglomerate the ore fines and produce sinter and or pellets
with necessary physical properties at the optimum, to the extent permis-
sible by the economics involved. The burden in this case is made up of
lumpy ore plus sinter and or pellets. In many a places, particularly for larger
furnaces, the entire ore is crushed to produce sinter and or pellets which
constitute the burden. The additional cost of artificial agglomeration can be
more than offset by the increased efficiency of the operation. The fines are
screened off from the pellets and the sinter to obtain adequate permeability
in the furnace. Ideal sinter size is +10-50 mm and pellets of +6-30.
Nature and Composition of Gangue in the Charge

In order to improve blast furnace performance, it is necessary that the stack
zone of the blast furnace is as extended as possible and the softening and
melting zone confined to the lower part over a narrow region. The presence
of other oxides has to be judged from this point of view. This is quantita-
tively assessed in terms of the reduction-degradation index of the ferrous
material. The lower the index, the better it is. The combination of oxides
that tend to increase the temperature gap between the solids and the liquids
tend to form a viscous semi-molten mass at lower temperatures and extend
this zone into the stack. This adversely affects the reducibility of the ore and
delays the reduction of iron oxide in the stack. This, on the whole, reflects in
inefficient blast furnace operation. The gangue contents, in terms of quality
and quantity, have to be assessed with respect to the whole operation.
In general, the presence of usually associated gangue oxides is known
to reduce the rate of reduction of an iron oxide. Gangue that forms low
melting slag often tends to cover the ore particles with viscous slag before
the reduction is complete in the stack. The slag layer acts as a barrier for
the reducing gas to reach the oxide, and the rate of reduction slows down.
Silica is one such constituent which is invariably present in ores. It tends to
form a low melting iron-silicate slag. If a low melting slag layer is formed,
subsequent reduction, if any, may take place by its contact with solid coke
in the bosh region.
Temperature and Carbon Monoxide Content of Gases

For reduction to take place, CO content of the gases must be above the
equilibrium value as per Eqs. (10.4) to (10.6) at the designated temperature
and, thus, at the corresponding level of the blast furnace. The higher the
CO content of the gas and the lower its utilization in the stack, the higher
the CO content will be in the exit gases and the lower the efficiency of oper-
ation will be (i.e. higher coke rate). Also, the higher CO content in gases has
a decelerating effect on the direct reduction reaction which must occur to
some extent in order to improve the efficiency of the process.
Although it is very difficult to assess the gas composition inside the

material column, because of lack of access to the interior, the gas composi-
tion and temperature of only the effluent gases are determined to under-
stand the overall reduction rate. With other conditions remaining the same,
the lower the CO content and temperature of the exit gases, the lower will
the coke rate will be and the higher the efficiency will be.
Chemical Nature of Iron Oxide

It has been observed that the rate of reduction of hematite is faster than that
of magnetite, although its reduction involves the intermediate formation
and subsequent reduction of magnetite. A greater total quantity of oxygen
has to be removed in this process. The reason is that that Fe3O4 produced
by reduction of Fe2O3 has a porous structure because of the removal of oxy-
gen. These vacant sites allow oxygen from the core to diffuse to the reaction
interface readily. This is not so if Fe3O4 is the starting material. Even the
rate of reduction of the FeOt formed during reduction may be slow. It is
for this reason that the magnetite ores are given an oxidizing roast to form
hematite agglomerate before feeding these in the blast furnace, (e.g. the
taconite ore deposits of the U.S.A. are treated in this fashion).
Pressure
The increased pressure helps to increase the effective concentration of the
reducing agent, that is carbon monoxide, and the reduction rate of a pro-
cess, which is away from equilibrium, is also increased. The high top pres-
sure also helps to increase the contact time of gas with the charge. Hence,
the efficiency of reduction is improved.
Solid-Solid Reaction
Solid carbon is potentially a good reducing agent for iron oxides even at low
temperatures. The speed of reduction, however, is limited by the area of
contact between the solid coke and the ore. Once the oxygen at the contact-
ing areas are removed further reduction, being controlled by the diffusion
of reactants, oxygen from the core to the surface or carbon to the interior
across the reduced layer, slows down. At low temperatures, this is slow,
and consequently, the reduction is also slow. Above about 1000 degrees C,
the diffusion, and consequently, the rate of reduction is appreciable. The
formation of slag above about 1000 degrees C tends to decrease gaseous
reduction, and therefore, reduction by carbon may become dominant.
Below about 800 degrees C, the gas tends to deposit carbon according
to Eq. (10.2). Carbon is deposited in a finely divided form upon the surface
as well as in the pores and fissures of the oxide particles where the deposi-
tion reaction is catalyzed. This tends to increase the pores and fissures in
the oxide particles where the contacting area is larger, and consequently,
the reduction is improved. Even if all this carbon is not consumed in the
lower temperature region, on its descent, it would do so at higher tempera-
ture levels. The contribution of the carbon reduction reaction is substantial.
In most cases, the most efficient blast furnace operation is obtained when
about one-third of the reduction is achieved by solid carbon (i.e. direct
reduction) and the remaining two-third by gaseous reduction (i.e. indirect
reduction).
The deposited carbon, being very finely divided, is readily available
to replenish any CO consumed in reduction, by solution reaction as in
Eq. (10.2) in-situ.
Reactions in the Bosh

The reduction of iron oxide by the CO contained in the ascending gases
in the stack and directly by solid carbon of coke or that deposited by the
Naumann reversion reaction continues side-by-side as direct reduction of
iron oxide against the indirect reduction by the CO. Out of these, the gas-
solid reaction kinetics is faster, and the indirect reduction is dominant in the
stack. As the gangue, fluxed in sinter-form or in un-fluxed form in an ore, in
combination with FeO, formed during the indirect reduction in the stack,
forms highly oxidizing and low melting slag, it trickles through the coke bed
below the cohesive zone. It brings the FeO in the slag in intimate contact
with coke which directly reduces the iron oxide to iron. This direct reduc-
tion can be represented as:
(FeO) + < C > = < Fe > + {CO} (10.16)
The iron reduced is still in solid form because of its high melting point.
But it starts picking up carbon as it trickles down through the bosh region,
and its melting point thereby decreases, and it becomes molten. It eventu-
ally gets saturated with carbon.
The initial slag with high FeO content is acidic, but it then picks up the
CaO and achieves the desired basicity and viscosity as it passes through the
cohesive zone. Concurrently, all the P2O5, most of Cr2O3, if present, and part
of the SiO2 and MnO forming part of the slag are reduced in the bosh zone.
The iron is carburized, it melts and it picks up Si, S, Mn, Cr, and P as per
their extent of reductions. Phosphorus is nearly completely reduced and
joins iron. The others are distributed between slag and metal as per their
actual equilibria, under the prevailing conditions, between slag and metal
phases.
The processes occurring in the bosh zone have been studied in some of
the most interesting ways. In one, an experimental furnace midway through
the campaign was quenched with nitrogen and the solidified charges from
various levels of interest were analyzed for interpretation. In others, sam-
plers were introduced in the experimental furnaces to draw samples from
the various levels of the bosh and hearth during their normal operations. In
yet another instance at the time of blowing down, while in normal operating
condition the No. 5 blast furnace of NSC, Higashida, was quenched with
38000 t of water and the samples were taken out from the solidified charge,
from various levels, to predict bosh processes.
All these direct observations, coupled with laboratory investigations
have revealed that the carburization of reduced iron in solid state itself is
quite slow, but after melting in the lower portion of the bosh, it increases
rapidly. The silicon content of the metal phase is high even before melt-
ing. This suggests that the silica reduction is taking place rapidly even
when iron is not fully molten. It is inconceivable to assume that silica is
reduced because of the slag metal reaction in this region. Laboratory stud-
ies indicate that SiO gas is formed during the burning of coke. The SiO
gas formed due to the reduction of silica from the coke ash in the tuyere
zone could be as:
< SiO2 > + 2 < C > = {SiO} + 2 {CO} (10.17)
which, during its ascent, contribute to siliconization of the metal.
The possible reactions of absorption of Si by the metal could be as:
{SiO} + {CO} = [Si] + {CO2} (10.18)
{SiO} + < C > = [Si] + {CO} (10.19)
{SiO} + (FeO) = (SiO2) + < Fe > (10.20)
A part of silicon in the metal in the lower part of the bosh may come from:
(SiO2) + 2 < C > = [Si] + 2 {CO} (10.21)
Similarly, the other oxides are likewise reduced as:

(P2O5) + 5 < C > = 2[P] + 5{CO} (10.22)
(MnO) + < C > = [Mn] + {CO} (10.23)
(Cr2O3) + 3 < C > = 2[Cr] + 3{CO} (10.24)
The sulfur released from the fuel, on its burning, in the tuyere combus-
tion zone is all absorbed by the trickling metal droplets through the coke.
Whatever is not absorbed in the bosh, is all picked up by the iron in the
stack. It is believed that the dominant mechanism of sulfur transfer is
through:
< CaS > (coke ash) + {SiO} = {SiS} + < CaO > (10.25)
FeS (coke ash) + {SiO} + <C> = {SiS} + {CO} + Fe (10.26)
{SiS} = Si + S (10.27)
And partially, bosh slag also absorbs sulfur as
< CaS > (coke ash) = (CaS) in slag (10.28)
It has been proven that the SiS is the dominant gaseous species in the bosh
region. Hence, the possible sulfur transfer reactions are as shown above.
Metal and slag beads trickling through the bosh zone have been col-
lected from working furnaces. Their analyses reveal that the sulfur in the
metal decreases, and that, in slag, it increases during the passage through
the bosh zone up to the slag surface in the hearth.
Reactions in the Hearth

The hearth is the part of the furnace where the molten products of the blast
furnace process accumulate and stay there until they are tapped out peri-
odically. In fact, melting takes place in the bosh region itself because of the
prevailing high temperature. There are two immiscible liquid products of
the blast furnace smelting process. One is the metallic phase containing iron
and all those elements which are reduced and collected by the metal. The
other phase is the oxidized phase containing oxides like alumina, magnesia,
alkali oxides, alkaline earth oxides (unaffected oxides), silica, manganese
oxide (partially reduced oxides), and sulfur. These are generally known as
molten metal and slag. These trickle down through the bosh and the tuyere
zone through the coke layer and finally collect in the hearth, a receptacle
for both these phases. The specific gravity of slag is much lower than that of
the metal. Therefore, they stratify into two distinct layers of molten immis-
cible liquids. Both of these go on accumulating in the hearth up to a certain
level, and then they are tapped out separately. The slag is tapped every 2–3
hours, and the metal is tapped every 5–6 hours in a normal practice. The
desired product of the blast furnace smelting is the molten metal. Slag is
the inevitable by-product without which the metal cannot be obtained.
On economic grounds the process must operate at the minimum tem-
perature (in the hearth), commensurate with high productivity and efficient
operation. This will keep the coke rate down to a minimum. Any unnec-
essary increase in the operating temperature would require extra coke or
other fuel, and thus, unnecessary additional costs of operation. What is this
minimum operating temperature and how it is fixed?
Since the molten metal trickles down through coke bed in the furnace,
it is always saturated with carbon and it is molten even at around 1200
degrees C. The slag mainly consists of CaO, SiO2, Al2O3 constituting nearly
90% of the slag and the balance being minor oxides like MgO, alkalis, FeO,
etc. No mixture of these oxides is completely molten below 1200 degrees
C as is obvious from the ternary diagram shown in Figure 10.8. Thin liquid
oxide product of the general composition encountered in a blast furnace is
not readily possible below 1300–1400 degrees C. In order to obtain molten
metal, it is necessary that both the metal and the slag form thin free-flowing
liquids, so that these clarify into two clear liquid phases and are clearly
tapped separately from the furnace. For this to be achieved, the operating
temperature of a blast furnace will have to be more than the liquid tem-
perature of the slag formed (i.e. around 1300–1400 degrees C). If no flux
is added, and if silica and alumina are the only main gangue constituents
of the ore, the liquid temperature of this combination is far too high. Flux
like CaO is added to develop a composition such that the liquid is in the
range of 1300–1400 degrees C. There are some exceptional ores like the
Minnette ore deposit which are self-fluxing and need no external flux to
obtain a slag composition having the above range of liquid temperatures.
Most ore deposits contain silica and alumina as the main gangue constitu-
ents and need lime as flux to obtain proper slag composition. The problem
is serious if:
(a) the total gangue is high and
(b) alumina is high and ratio of alumina to silica is high.
Any excess gangue would need extra flux, and thereby the slag volume
(amount of slag per ton of metal made) would increase. Any such increase
means less effective utilization of the furnace volume for iron ore reduc-
tion and consequently lower productivity. Every attempt must therefore be
made to:
(a) use as high grade ore as is possible
(b) beneficiate the ore to reduce the gangue as a whole and
(c) beneficiate the ore to reduce selectively the alumina content of the
ore and ore fines.
A proper burden chemistry thus obtained can ensure proper slag chemistry
to obtain free flowing slag at the minimum operating temperatures of the
furnace. In fact, the burden chemistry decides the operating temperatures
of a blast furnace. One single constituent that matters more in this respect
is alumina. It comes into the burden as associated with iron ore lumps, iron
ore fines as sinter or pellets, as insoluble in the flux, as a part of coke-ash, as
part of manganese ore charge, etc.
Obviously, the liquid temperature of the slag is much higher than that
of the metal, and therefore, the operating temperature of a blast furnace is
always at least 50–100 degrees C above the liquid temperature of the slag
formed in the furnace. At this temperature, both the slag and the metal will
be thin, free-flowing liquids. In reality, the blast furnace operating tempera-
tures are in the range of 1300–1500 degrees C depending upon the total
gangue in the burden and the flux added. The gangue in the burden comes
from:
1. Iron ore lumps;
2. Iron ore fines charged as sinter or pellets;
3. Coke ash;
4. Coke breeze ash used as fuel in the sintering process;
5. Manganese ore gangue, if added in the burden;
6. Ash of coal injected through the tuyeres, if any;
7. Insolubles in the charged flux.
All these have to be scrutinized and processed to minimize the insoluble
contents. The efficiency of the furnace operation is directly related to the
quality and quantity of these gangue in the burden. These are minimized
by adopting the following:
1. Iron ore washing to obtain lumps of low gangue;
2. Iron ore fines beneficiated to obtain a concentrate with acceptable
gangue;
3. Manganese ore charge eliminated and sulfur problem tackled in an
alternative way like external desulfurization;
4. Good quality limestone preferred with less of insolubles;
5. Imported variety of coking coal with very low ash content (around 10–
12%) are used to replace a part of the local coal of high ash content for
making coke with ash content of around 20% as against the 28% ash
coke made from entirely local coals.
By adopting most of the above recommendations in their G furnace, Tata
Steel has been able to reduce slag volume to just 350 kg/thin.
Sulphur Reaction:
The slag composition has also to be regarded from the point of view of its
reactions with the metal, in particular its desulfurizing ability. This is the
main reaction that occurs in the hearth. Concurrently, and consequently,
some other reactions do occur in the hearth. Carbon dissolution in the
metal reaches near the saturation limit in the hearth.
The sulfur transfer across the slag-metal interface along with the associ-
ated reactions are often described in the molecular fashion as:
[FeS] + (CaO) = (CaS) + (FeO) (10.29)
(FeO) = [O] + [Fe] (10.30)
2[O] + [Si] = (SiO2) (10.31)
[C] + [O] = {CO} (10.32)
[Mn] + [O] = (MnO) (10.33)
and alternatively in ionic form as:
[S] + (O2−) = (S2−) + [O] (10.34)
[S] + 2e = (S2−) (10.35)
[Si] − 4e = [Si4+] (10.36)

(O2−) −2e = [O] (10.37)
[Fe] −2e = [Fe2+] (10.38)
which can be electrically balanced as:
2[S] + [Si] + 2(O2−) = (SiO2) + 2(S2−) (10.39)
[S] + [C] + (O2−) = {CO} + (S2−) (10.40)
Therefore, molecularly
a
(CaS).a(FeO)
K(10.29) = a
[FeS].a(CaO)
and assuming Henrian behavior
[%S].(%FeO)
Kv(10.29) = (10.41)
[FeS].(%CaO)
(%S) (%CaO)
i.e. ∝ (10.42a)
[%S] (%FeO)
The same is obtained in ionic form as:
a
(S2−). a[O]
K(10.34) = a
[S].a(O2−)
and assuming Henrian behavior
(%S).[%O]
Kv(10.34) =
[%S](O2−)
and therefore
(%S) (%O2−)
∝ (10.42b)
[%S] [%O]
The oxygen ion concentration in slag is represented by CaO or similar
basic oxides in the slag, and the oxygen potential of the system is repre-
sented by either (%FeO) or [% O]. The ratio (% S )/ [% S] is often referred
to as the desulfurizing index which varies with basicity and the oxygen
potential as shown in Figure 10.8.
The overall equilibrium between slag and metal in a working blast furnace
hearth is complicated because of the possibility of transfer of several elements
simultaneously across the interface. Individually, some elements may achieve

equilibrium, but the system, as a whole, is always away from the equilibrium.
It has been fashionable to correlate the partition of sulfur with that of man-
ganese as:
[Mn] + [S] = (Mn2+) + (S2−) (10.43)
(%MnO).(%S)
Kv(10.34) = (10.44)
[%Mn].[%S]
It has been found experimentally that
(%S) [%Mn]
= 28 (10.45)
[%S] (%Mn)
for conditions prevalent in the blast furnace. It has also been shown by
laboratory tests on the slag and metal samples, simultaneously drawn from
working furnace hearth, that although manganese-sulfur attains equilib-
rium inside the furnace, manganese and sulfur do not individually achieve
equilibrium with respect to the whole system.
Most of the pig iron that is produced is subsequently converted to steel,
and it must not contain more than 0.04% S, and more often not, more
than 0.03% S in order to economize the process of steelmaking. Adequate
desulfurization of metal must, therefore, occur inside the furnace by way
of slag-metal interaction. In the absence of equilibrium, the kinetics of
desulfurization assumes vital importance. The process of desulfurization in
a blast furnace is so complex that even whether it is diffusion controlled or
activation controlled, the process cannot be inferred with certainty.
In spite of this lack of equilibrium, the fact remains that presence of manga-
nese in metal helps removal of sulfur from metal to slag inside the furnace. Iron
ores generally do not contain manganese, and therefore, manganese bearing
material like Mn-ore or some manganese slag is added in the blast furnace bur-
den to ensure adequate manganese in the metal phase. Presence of manganese
does help remove sulfur from the slag to metal in the hearth of the blast furnace.
Historically, the blast furnace was seriously used as a means of removal
of sulfur in the burden to obtain desired sulfur levels in the hot metal and
thereby obtain the desired sulfur specifications in the steels produced. This
was economical because sulfur removal is far more efficient under the reduc-
ing conditions of the blast furnace than under the oxidizing conditions of
steelmaking. This approach was, on the whole, economical for overall iron
and steel production. But, as sulfur specifications of steel became more and
more stringent, the desired sulfur level in the steel could only be obtained
by first adopting external desulfurization of hot metal prior to steelmaking.
In that case, there was no need for the blast furnace operator to bother with
desulfurization. He was at liberty to produce high sulfur irons which were
later externally desulfurized to take care of the sulfur problem in steelmaking.
This situation then did not need any manganese to be present in the hot
metal. This means there was no need to put manganese bearing material
in the blast furnace burden. Now, many plants having external desulfuriza-
tion facilities do not add manganese in the blast furnace burden. This has
resulted in very low manganese content in hot metals. This situation has also
led to decreasing the operating temperatures of the furnace and decreasing
silica reduction in the bosh and hearth. Consequently, it produces much low
silicon iron which are far better suited for basic steelmaking.
The overall economy of iron and steelmaking has improved in spite of
adopting the external desulfurization facility in such cases.
If the total sulfur burden in the BF is low, particularly by using low-S
coke, then there is no need to put manganese bearing charge in the burden.
In that case, low-S and low-Si hot metal can be produced, even at lower
temperatures and at a lower coke rate.
FIGURE 10.8 Desulfurizing index as a function of basicity and reducing power of slag.
10.5 Alumina Problem

Iron ores invariably contain Al2O3 and SiO2 as main gangue constituents.
The pyro-metallurgical operation of smelting iron ore in a blast furnace is
supposed to produce molten iron on the one hand and molten slag, taking
these gangue constituents in solution, on the other, to clearly separate the

metal from the slag. The slag so produced at the working temperature of
the furnace (hearth temperature) must therefore be very fluid for its clean
separation from the metal phase. Since no liquid of reasonably low viscosity
can be produced merely by melting these gangue oxides alone, some flux
in the form of CaO (added as limestone) is added to render these gangue
oxides readily fusible. The slag produced in a blast furnace is, therefore,
essentially an oxide-melt of the ternary CaO-SiO2-Al2O3 type system, or if
sufficient MgO is also present in the burden along with lime, then of the
quaternary type CaO-SiO2-Al2O3-MgO system. Let us first consider the ter-
nary type slag. In order to appreciate the melting of the oxides from this
ternary system, the liquid temperatures as superimposed on this ternary
quilibrium phase diagram, are shown in Figure 10.9.
If no flux is added, then the slag will have a composition denoted by
a point on the SiO2-Al2O3 line of the triangle. The ratio of percentages of
these two oxides in the ore shall decide the point, since only these two shall
form the slag. Since this combination does not produce the required thin
liquid slag by themselves, CaO is added as flux and with its addition, the
composition tends to move towards the CaO corner from this point on the
SiO2-Al2O3 line depending upon the amount added. This is depicted in the
Figure 10.8 by three different lines originating from three different points,
as decided by their three different assumed concentrations in the iron ore.
These three lines are the lines joining A, B and C to the CaO corner.
FIGURE 10.9 The basic ternary CaO-SiO2-Al2O3 system of interest to slag composition in ironmaking.
Each of the points A, B and C denotes a specific ratio of Al2O3/SiO2 in three

different ores. These three points are chosen to depict the two extreme
cases and the optimally balanced proportion of alumina and silica in the ore.
The lowest liquid region in this ternary system is shown in the figure by a
circle. Blast furnaces must be operated to produce slag falling in this region
of low liquid temperatures. The addition of CaO as flux to the ore shall
change the slag composition on the corresponding line as shown, depending
upon the amount of CaO added. In order to obtain the desired slag depicted
by the circle, the original ratio of Al2O3/SiO2 in the ore will have to be within
a certain limit as shown by the hatched area. If the original ratio is outside
this area, the addition of CaO as flux is not going to result in slag of the low-
est liquid point. The slag, in that case, will have to be produced by operating
at a somewhat higher temperature, which is bound to increase the coke rate
unnecessarily. This is why, besides their actual percentages, the original ratio
of these two oxides in the ore and sinter, must be within a certain accept-
able range. If this ratio is not within that range, then the ratio will have to be
adjusted by blending of two different ores of suitable composition to obtain
the desired ratio. This often results in the final ore blend of high gangue
percentage. This is also equally detrimental to the blast furnace operation
because it unnecessarily increases the coke rate, increases the slag volume,
and affects productivity adversely. The best course left is to beneficiate the
ore, to decrease the gangue proportion, if otherwise economically feasible,
to finally obtain low gangue contents and the correct ratio of these two main
gangue constituents. All these alternatives are adopted in practice depend-
ing upon the local nature of the ore to be used for smelting.
Quite often, a high alumina ore is blended with siliceous ores to obtain
the correct alumina/silica ratio to ensure smooth blast furnace operation.
This will no doubt increase the slag volume, and consequently decrease
the productivity, which will have to be tolerated. This is also the reason
why quartzite or dunite (magnesium silicate) is added in the sinter mix to
produce sinter of acceptable chemistry to finally produce the acceptable
alumina/silica ratio in the burden.
In a ternary slag system, the alumina/silica ratio should be in the range
of 0.5-;1.5, more closely around 0.6–0.7.
The situation improves if the burden contains substantial amount MgO
as flux or as arising out of blending. The presence of MgO can help to obtain
slags of lower liquid temperatures, when in certain proportion, and which
can be appreciated from the quaternary phase diagram. These are usually
depicted as pseudo-ternary diagram in the form of ternary, equivalent to

only the total percentage of three major constituents, excluding the MgO.
The higher the alumina/silica ratio, and alumina being a more refractory
oxide, the liquid temperature goes up which necessitates higher operating
temperatures, and which in turn. means high coke rate and high silicon con-
tent of the molten iron. Higher operating temperatures, though useful for
effective desulfurization, are not preferred these days, to achieve overall
better economy. This problem is dealt with by adoption of external desulfur-
ization or by using a ladle furnace in steelmaking where it is removed to the
desired extent. If any of these alternatives are not possible, for any reason,
then adequate desulfurization will have to be achieved in the BF itself. Use
of low sulur coke helps to a large extent, in this respect. It is also possible to
work out slags of much lower liquid temperatures by decreasing the basicity
if desulfurization is not the requirement to be met by blast furnace operator
Indian Ores
Indian iron ores are notorious because of their high alumina/silica ratio and
high total percentage of alumina and silica together. The problem was com-
pounded by the use of high ash coke produced locally. This is the reason
why the slag volume in Indian BFs is around 400–500 kg/thm. In fact, it
should be lower than 350 kg/thm. The use of imported low ash coal or coke
has had a very beneficial effect on blast furnace performance in India after
the import was liberalized after 1991. Of late, many iron ores are benefici-
ated by merely washing to improve the total gangue content and the alu-
mina/silica ratio favorably. The fines are also beneficiated successfully, from
the same point of view, to improve the sinter quality.
All these factors put together have decreased the blast furnace oper-
ating temperatures in India from 1500–1550 degrees C to 1450–1500
degrees C. The slag volume has also come down to below 350–450 kg/thm.
In general, the CaO, SiO2 and Al2O3 contents in the BF slag are more or
less equal with the V ratio; close to one is generally maintained. In that case,
the slag composition is roughly as follows: CaO = 32–35%, SiO2 = 28–32%
and Al2O3 = 18–25%, the balance being minor oxides. However, MgO is
invariably present in the blast furnace burden and the slag in that case is as
follows: CaO = 30–32, SiO2 = 28–32%, Al2O3 = 18–25% and MgO = 4–15%
with the remaining being minor oxides like MnO, Cr2O3, TiO2, etc.
The primary slag formed in the cohesive region is always rich in FeO, and
consequently, has very low liquid temperatures. It is, therefore, amenable
to reduction by carbon readily. But, as the FeO is directly reduced in the

bosh region, the slag composition changes because of the loss of FeO. It
also continuously changes with dissolution of silica and alumina, as ash con-
stituents, released by the burning of coke in the bosh and tuyere regions.
The normal expected slag composition, in reality, is achieved below the
bosh level.
Earlier, the basicity value, depicted as the V-ratio, was maintained at
around 1.2 to effect sufficient desulfurization of metal inside the furnace.
But now, the V-ratio is invariably maintained at around only 1.0. It is also
true that low sulfur coke is increasingly being used to keep the sulfur level
in the metal on the lower side.
The effect of alumina in sinter on BF performance has already been
shown in Figure 7.15. It was to emphasize that because of this one single
unfavorable factor, the productivity values in Indian BFs were low as com-
pared to many others in the world where low alumina-burden is maintained.
10.6 Kinetics of Reactions in Bosh and Hearth

The melting of the solid charge, except coke, and the direct reduction of
wustite by carbon of the coke as
< FeOt > + < C > = < Fe > + {CO} (10.46)
are the main processes occurring in the bosh region. The rate of these two
processes is important from the point of view of achieving a maximum pro-
duction rate, since the production rate is directly proportional to the melting
rate of a blast furnace. The slag that is formed in the bosh region is high in
basicity because it has yet to take coke-ash in solution, which is mainly silica
and alumina, to attain the working basicity in the hearth. The rate of dis-
solution of ash in high limey slags under reducing conditions is relevant to
understand the bosh processes. Fortunately, the bosh temperature is much
higher than the slag temperature in the hearth. The dissolution of ash in
such limey slags therefore does not pose any difficulty. It has, however, been
shown that if the basicity of the bosh slags is reduced, by not incorporating
lime in the charge, and then by injecting the lime where it is to flux the coke
(through the tuyeres), it is possible to decrease the viscosity of the bosh
slags and hasten the process of melting of bosh slags. This is the principle of
lime injection through the tuyeres. The rate of reduction of wustite by solid
carbon does depend on the structure of the wustite formed in the previous
stages of reduction. This is a highly endothermic reaction, and any excessive

direct reduction in the bosh zone calls for extra heat demand which may be
detrimental to the overall operation. If the indirect reduction of iron oxide
to the wustite stage in the stack is carried out appropriately, the burden of
direct reduction in the bosh zone is within the limits and the bosh processes
would proceed at proper pace keeping the coke rate to a minimum.
The composition of the pig iron is finally decided by the processes occur-
ring in the hearth, particularly with respect to its silicon, manganese, and sul-
fur contents. The process of desulfurization in a blast furnace is so complex
that even whether it is a diffusion controlled or an activation controlled pro-
cess, it can not be inferred with certainty. Nevertheless, laboratory tests have
indicated that the rate of desulfurization of metal increases, besides stirring,
with temperature and basicity of the slag. Higher temperature reduces the
viscosity of slag and promotes mass transport. For effective desulfurization,
therefore, temperature and burden chemistry (basicity) should be adjusted to
obtain efficient desulfurization and free-flowing slag at the optimum economy.
The effect of temperature on the hearth reactions need to be under-
stood in detail because it affects the chemical composition of the pig iron
to a large extent. This is why the working temperature of a blast furnace is
supposed to be the hearth temperature. The hearth temperature is one of
the parameters to be adjusted to obtain the desired grade of pig iron. The
effect of increased temperature is the result of increased silicon content
and decreased sulfur content; the sulfur and silicon contents of iron vary
inversely with temperature within certain limits. The reduction of silica by
carbon as per Eq. (10.19) is an endothermic process. Hence, a rise in tem-
perature helps silica reduction. On the contrary, a rise in temperature pro-
motes mass transport and thereby helps desulfurization of the metal phase.
A significant amount of absorption of carbon, silicon, and manganese by
the metal and almost entire desulfurization of the metal occurs below the
tuyere level and above the lower surface of the bulk slag layer. This zone is
referred to as the reacting-hearth. The nearly no reaction zone below it is
the decanting-hearth. The reacting-hearth is just the extension of the work-
ing zone of the furnace. The variations in composition of the metal from
the beginning of the tap to the end of the tap have been attributed to the
reactions in the reacting-hearth. During the tapping operation, the bur-
den descends rapidly into the hearth because of the room generated in the
hearth. A relatively cooler charge, therefore, descends into the reacting-
hearth during tapping and thereby alters the thermal profile in the furnace.
Consequently, the reduction of silica and manganese oxide and desulfuriza-

tion of the metal are delayed. This produces metal with higher contents of
sulfur and lower contents of silicon and manganese as observed in practice.
It emphasizes that the interface reactions between the bulk slag and
metal phases play a negligible part in the overall hearth reactions. No won-
der that in spite of long duration of contact of slag and metal in the hearth,
between two taps, equilibrium is not achieved between slag and metal
phases in working furnaces.
10.7 Efficiency of the Blast Furnace Process

In the early open-top shaft-type blast furnaces, the effluent gases contain-
ing the unused carbon monoxide used to burn at the furnace top on their
combining with atmospheric oxygen. This was quite erroneously attributed
to the lack of adequate height of the furnace shaft for effective utilization of
the reducing gases. Increasing the height of the shaft, however, failed to solve
the problem. The lack of understanding of the chemico-thermal processes,
and the furnace efficiency prevented this problem from being appreciated.
The first serious attempt to assess the thermal efficiency of a blast furnace
was made by Gruner. He correlated the top gas CO/CO2 ratio with furnace
efficiency and proposed that maximum thermal efficiency (i.e. minimum of
coke rate) will be obtained if all the iron oxide is reduced by CO above the
tuyere level to form CO2 (i.e. indirect reduction), and that none of this CO2
should consume coke and get reduced back to CO (i.e. direct reduction). The
concepts of direct and indirect reduction as stipulated by Gruner are no longer
acceptable. The chemico-thermal processes occurring in a blast furnace are too
complex to assess the efficiency in terms of just the top gas CO/CO2 ratio alone.
The fundamental nature of the mass balance, heat balance, heat exchange
and reduction in a counter-current gas-solid reactor have, therefore, been
studied in detail. The theoretical heat and mass balance is best summa-
rized in the form of Rist diagrams, one of which has already been shown in
Figure 10.6 dealing with the mass balance in its most simplified manner.
Elements of Enthalpy Balance

Reduction of iron oxide is a high temperature process, and the enthalpy
requirement of the process must be met at the minimum expenditure of
energy in the form of carbon (in the form of coke and or other injected
fuels). In a most simplified form, the enthalpy input and output per mole
product of liquid iron should balance. The enthalpy of n moles of Fe2O3

(required to produce 1 mole of product Fe and in this case it is n = ½)
should be equal to the summation of enthalpy of one mole of molten Fe and
that of equivalent moles of CO and CO2 generated as gas.
The furnace can operate between two extreme top gas compositions of
either 100% CO or 100% CO2, (i.e. on the Rist Diagram O/C = 1 or O/C = 2).
This means that when Fe2O3 is reduced entirely, either directly by carbon or
indirectly by CO. If all the Fe2O3 is reduced indirectly by CO, the oxygen input
as blast nOB would be equal to one to first produce CO at the tuyeres. The
operating line in that case would be that joining points O/C = 2 on the x-axis
and nOB = 1 on the y-axis as shown in Figure 10.10. The other extreme case is
when top gas is all CO (no CO2) (i.e. all Fe2O3 is reduced directly by carbon).
FIGURE 10.10 The extreme conditions of operating a blast furnace are shown by two operating lines.
The areas in between are type typical wedges of operation. The actual operation can be depicted by any
straight line falling in these wedges. The mid-point M depicts particular conditions of operation as
decided by the charge composition and enthalpy equipment of the process.
In this operation, carbon will have to be burnt at the tuyeres to produce

the necessary heat for the process. This will have to be assessed and the
equivalent amount of blast oxygen will have to be supplied for its combus-
tion. The line in this case joining the points O/C = 1 and nOB = appropriate
value, as shown in Figure 10.10, indicates the operating line for the furnace.
The blast furnace operates between these two extreme lines as indicated in
Figure 10.10. Either of the extreme cases is too impractical to be adopted
in practice. Since the slope of this line gives the coke rate of that operation,
the coke rate of a blast furnace operation shall always lie between these two
extreme slopes. The area between these two extreme lines, because of its
shape, is often referred to as the wedge of permissible operating lines in the
field of blast furnace technology.
In actual practice, the operation is so balanced that major part of the
reduction of oxides takes place indirectly by CO generated at the tuyere
level, while simultaneously generating the required enthalpy for the pro-
cess, and a small amount of oxide is reduced directly by solid carbon. The
proportion of this indirect and direct reduction is to be adjusted with
respect to the furnace design and quality of the charge material under the
available operating conditions like hot blast temperature, humidity and oxy-
gen contents of the blast, etc.
Process Zones in a Blast Furnace

Studies of blast furnace processes, in laboratories and in commercial prac-
tices, have led to the understanding of the various process zones in a blast
furnace as shown in Figure 10.11. This figure sums up the chemico-thermal
processes occurring in a blast furnace.
In this, the blast furnace is divided into two main zones. The condi-
tioning zone is the upper nearly half of the furnace, where preheating and
partial reduction of the burden by the gas takes place. The lower nearly
half of the furnace is the processing zone where reduction of iron oxide is
completed and fusion of the reduced product is effected to finally obtain
slag and pig iron in molten condition.
The essential feature of this diagram is the existence of two heat
exchange zones separated by an isothermal zone occupying the greater part
of the stack. The thermal capacities of the gas and the solid phases can be
correlated in these three zones as follows:
If Tg and Ts represent the thermal capacities or water equivalents of
gas and the solid phases flowing through a given section of the furnace
per unit time, the heat exchange in the top heat exchange zone can be
described as:
Tg > T8
It means that the gas contains more heat than can be absorbed by the bur-
den. The excess heat is, therefore, carried away with the gas which is dis-
charged at a higher temperature than that of the solid.
FIGURE 10.11 Chemical-thermal processes occurring in a blast furnace.
The situation in the lower heat exchange zone is just the reverse (i.e.
Tg < Ts ), and very efficient heat exchange takes place in the lower heat
exchange zone. It also means that the charge can not be heated to the tem-
perature of the gas rising from the tuyeres. On the whole, the heat exchange
is maximum at the tuyere level and minimum at the throat.
In the isothermal zone that separates the two heat exchange zones,
the Tg = Ts. This is why it is the isothermal zone and is referred to as the
thermal reserve zone. The oxides are reduced at around 950 degrees C
in solid state in this isothermal reactor. The oxides remain in this zone
for a sufficient length of time for chemical equilibrium between wustite
(FeOt) and gas to be virtually reached. This region consumes the greatest
quantity of heat and predominantly decides the consumption of fuel and
the blast, as well as production of furnace gas. It is, in a way, a recupera-
tor for sensible heat and reducing the power of the gas, since the ther-
mal and the chemical reserve zones overlap to a significant extent. The
solution loss reaction recuperates the reducing power of the gas phase in
this zone. The heat, due to the exothermic reactions like gaseous reduc-
tion of oxide, is fully utilized in carrying out the endothermic processes
like calcination of flux, solution loss reaction, etc. in the reserve zone.
Hence, the isothermal nature is established. The thermal operation of
each stage is independent, except that the temperature of the gas reach-
ing the upper stage is dependent on the temperature of the gas leaving
the lower stage.
Thermal Efficiency
The maximum gas-burden temperature difference is 400–500 degrees C
at the tuyere level, and this gap is soon closed because of the efficient
heat exchange in the lower zone. In the top heat exchange zone, the gas-
burden temperature difference is of the order of maximum 200 degrees C
just at the throat. The existence of an isothermal zone over an appreciable
extent of the shaft, particularly after the initial heat transfer zone, should
be accepted as an indication of remarkably high thermal efficiency of the
reactor.
The top gas temperature is a function of the quantity of heat escaping,
and which is of the order of 5–6% of the total heat released by coke con-
sumption in a modern blast furnace. The variation of gas temperature will
not, therefore, alter the thermal efficiency significantly, since the top gas
temperature cannot be decreased substantially below about 200 degrees
C. Only a very marginal improvement may be possible by increasing the
height of the furnace. Any increase in the height of the furnace demands
more stringent, and therefore costly, requirements of coke. What may be
gained by increasing the height of the furnace may have to be paid back,
perhaps more, by way of cost of coke. Increasing the height of the furnace,
therefore, offers no solution for improving the thermal efficiency beyond a
small range.
The gas-burden temperature difference of almost zero or a reason-
ably low value may be obtainable after an extremely long time of contact
between gases and solids, and not by increasing the shaft height. The
time of contact of gas and the burden can be increased by high top
pressure application alone, and the same has been found to improve
the thermal efficiency. It has also been observed that in the upper heat-
ing zone and in the reserve zone, chemical processes have little effect
on the variation in the top gas temperature. A modern furnace has a
considerable reserve height. The reduction and heat transfer processes
slow down in the thermal reserve zone. An increase in the height of
the furnace can not be justified from the point of view of heat transfer.
Neither the ratio of height to diameter of the furnace, nor the residence
time of the burden in the furnace has any effect on the development of
the blast furnace processes.
The top gas temperature does not increase with high driving rates (and
sometimes not even when the rate is doubled). This is because, as the heat
load increases, the zones of heat transfer are enlarged at the expense of
the reserve zone. If the reserve zone is t reduced to zero, then only the top
gas temperature will rise on increasing the driving rate, which is practically
impossible in actual practice.
The modern furnace can therefore be driven hard without any dan-
ger of decrease in its efficiency. Improving prepared burden can only allow
harder blowing. The efficiency could further be increased readily by the
application of high top pressure alone. This will be discussed further in
Chapter 17.
10.8 Quality of Hot Metal

The useful product of the BF process is molten iron generally meant to
be refined further to produce steel. Modern steelmaking practice imposes
stringent quality restrictions on hot metal. The quality of molten iron means
its chemistry, its temperature, and particularly the sulfur and the silicon
contents. This is decided by the steelmaker so that his operation is more
economical. In fact, the ironmaking and steelmaking processes are viewed
in an integral-way to finally obtain steel at the best possible economy.
The steelmaker demands molten iron of certain chemistry and at cer-
tain temperature. When supplied, it should be free from BF-slag contam-
ination. Attempts are, therefore, made to tap metal as free of slag as is
otherwise possible. The higher the temperature of the metal, the better it
is for the steelmaker. But this taxes the iron producer in terms of coke rate,
and the hot metal is always supplied by the iron producer at the lowest pos-
sible temperature.
Silicon Content of Hot Metal

For steelmaking, the lower the S and Si contents, the better the pro-
duction of quality steels which ensures a higher metallic yield as well.
Although sulfur and silicon contents can both be controlled by external
treatments, it is always attempted to get the desired quality from the BF
operation without losing productivity and economy. It is therefore desir-
able to achieve the desired silicon and sulfur contents in hot metal in the
BF itself.
The problem of sulfur control has already been discussed. There is a
limit to which sulfur can be brought down in molten metal. Use of low-
sulfur bearing iron ore and coke can give low-end-sulfur iron. This is, how-
ever, a costly affair. Some sulfur can be removed by using high slag basicity
and high hearth temperature. This, again, adds to the cost of ironmaking.
External desulfurization has been found to be the solution.
However, it is rather difficult to explain and control the silicon con-
tents. Several models have been developed to predict hot metal silicon con-
tent. Traditionally silicon was considered to be picked up by metal via the
reaction:
(SiO2) + 2 < C > = Si + 2{CO} (10.47)

Now, it is believed that SiO gas is first generated by reduction of silica in
coke ash with solid carbon of coke, in situ in the tuyere region according to:
SiO2 (coke ash) + C(coke) = SiO(gas) + CO(gas) (10.48)
And then SiO(gas) + C(metal) = Si(metal) + CO(gas) (10.49)

These reactions are favored by high temperature. The higher the temper-
ature of the furnace, in particular in the raceway, the higher the silicon
content will be of the hot metal. For effective desulfurization, higher tem-
perature is preferred, and eliminating sulfur silicon content is likely to go
up.
Lower silicon is achieved by adopting the following strategy:
1. Suppression of SiO formation in the raceway, which will be helped by
lower ash in coke and lower coke rate.
2. Decreasing the activity of silica by having higher basicity. This is rather

difficult to adopt, but lime injection may be helpful.
3. Increasing the blast pressure to suppress formation of SiO gas.
4. Optimal control of heat input in BF, lowest possible RAFT.
5. Keeping as far as possible stable hearth conditions.

6. Maintaining proper viscosity stable hearth conditions.
7. Effective burden distribution to obtain effective raceway.
8. Minimizing irregularities in BF operation.

BF process control strategies are now available to effectively control sil-
icon at low levels at around 0.5% Si. This strategy involves the use of
low basicity (close to one), lower hearth temperature, and overlooking
desulfurization. The silicon and sulfur contents normally very inversely,
and this fact is also taken into consideration while decreasing silicon con-
tent of the metal. Ideally, a silicon content of about 0.5% is what the
steelmaker asks for to minimize flux consumption in steelmaking, thereby
minimizing the slag volume and minimizing iron oxide loss in the refining
slag in steelmaking.
10.9 TiO2 Addition in Burden

Being associated with iron ores, titanium is usually present in the blast
furnace burden as ilmenite (FeO.TiO2) or rutile (TiO2). The titanium
oxide is far more stable than silica, and is reduced to a very minor extent
in the blast furnace. Titanium, however, forms very stable carbides,
nitrides, and carbo-nitrides when in contact with carbon and or nitrogen.
These are refractory compounds and have low solubility in liquid iron,
even in the presence of carbon. Inside the blast furnace, particularly in
bosh and hearth regions, these precipitate out and form encrustations
on the BF lining in the bosh, tuyere, and hearth regions. It is believed
that being refractory in nature, this layer of titanium carbides, nitrides,
or carbo-nitrides provides good resistance to attacks from metal and slag
on the lining.
If the ore does not contain enough TiO2, then titano-magnetites or

ilmenite containing ore is deliberately added in the burden, as a matter of
routine practice, to obtain this safety cover on the lining. This is a common
practice in most of the plants.
10.10 U.S. and World Events, Trends, and Issues.

Source: USGS 2017
Domestic Production and Use (Data in million metric tons unless oth-
erwise noted)
Iron and steel (ferrous) slags are co-products of iron and steel manu-
facturing. After cooling and processing, ferrous slags are sold primarily
to the construction industry. Data are unavailable on actual U.S. ferrous
slag production, but it is estimated to have been in the range of 15 to 20
million tons in 2016. Domestic slag sales1in 2016 amounted to an esti-
mated 18 million tons, valued at about $350 million (ex-plant). Iron (blast
furnace) slag accounted for about 47% of the tonnage sold and had a
value of about $300 million; about 85% of this value was from sales of
granulated slag. Steel slag produced from basic oxygen and electric arc
furnaces accounted for the remainder.2Slag was processed by about 29
companies servicing active iron and steel facilities or reprocessing old slag
piles at about 135 processing plants in 31 States; included in this tally are
some facilities that grind and sell ground granulated blast furnace slag
(GGBFS) based on imported unground feed. Prices listed in the table
below are weighted averages (rounded) or iron and steel slags sold for a
variety of applications. Actual prices per ton ranged widely in 2016, from
a few cents for some steel slags at a few locations to about $110 for some
GGBFS. Air-cooled iron slag and steel slag are used primarily as aggre-
gates in concrete (air-cooled iron slag only), asphaltic paving, fill, and road
bases; both slag types also can be used as a feed for cement kilns. Almost
all GGBFS is used as a partial substitute for portland cement in con-
crete mixes or in blended cements. Pelletized slag is generally used for
lightweight aggregate but can be ground into material similar to GGBFS.
Owing to low unit values, most slag types can be shipped only short dis-
tances by truck, but rail and waterborne transportation allow for greater
distances. Because of much higher unit values, GGBFS can be shipped
longer distances, including from overseas.
Salient Statistics – United States: 2012 2013 2014 2015 2016e

1, 3
Production (Sales) 16.0 15.5 16.6 17.7 18.0
Imports for consumption4 1.2 1.7 1.8 1.4 2.0
5 5 5
Exports () () 0.1 () (5)
4, 6
Consumption, apparent 16.0 15.5 16.5 17.7 18.0
Price average value, dollars per ton,
17.00 17.50 19.00 19.50 19.50
f.o.b. plant7
Employment, numbere 1,800 1,700 1,700 1,700 1,600
Net import reliance8 as a percentage
7 11 10 8 11
of apparent consumption
e
Estimate
1
Data are from an annual survey of slag processors and pertain to the quantities of processed
slag sold rather than that processed or produced during the year. The data exclude any
entrained metal that may be recovered during slag processing and then sold separately or
returned to iron and, especially, steel furnaces. The data are incomplete regarding slag returns
to the furnaces.
2
There were very minor sales of open hearth furnace steel slag from stockpiles but no domestic
production of this slag type in 2012–16.
3
Data include sales of imported granulated blast furnace slag, either after domestic grinding or
still unground, and exclude sales of pelletized slag (proprietary but very small). Overall, actual
production of blast furnace slag may be estimated as equivalent to 25% to 30% of crude (pig)
iron production and steel furnace slag as about 10% to 15% of crude steel output.
4
Based on official (U.S. Census Bureau) data. In some years, the official data appear to have under-
stated the true imports; the apparent discrepancy was small for 2012, but may have been nearly 0.4
million tons in 2013 and 2014, depending on whether imports from Italy were mischaracterized as
being from Spain or not. The U.S. Geological Survey canvass captures only part of the imported slag.
5
Less than 0.05 million tons.
6
Although definable as total sales of slag (including those from imported feed) − exports, appar-
ent consumption of slag does not significantly differ from total sales owing to the very small
export tonnages.
7
Rounded to the nearest $0.50 per ton.
8
Defined as imports minus exports.
Source: U.S. Geological Survey, Mineral Commodity Summaries, January 2017
Recycling
Following removal of metal, slag can be returned to the blast and steel
furnaces as ferrous and flux feed, but data on these returns are incomplete.
Entrained metal, particularly in steel slag, is routinely recovered during slag
processing for return to the furnaces, and is an important revenue source

for the slag processors, but data on metal returns are unavailable.
Import Sources (2012–15)

The dominant imported ferrous slag type is granulated blast furnace slag
(mostly unground), but official import data in recent years have included sig-
nificant tonnages of nonslag materials (such as cenospheres, fly ash, and silica
fume) and slags or other residues of various metallurgical industries (such
as copper slag) whose unit values are outside the range expected for granu-
lated slag. The official data appear to have underreported the granulated slag
imports in some recent years, but likely not in 2011–12. Based on official
data, the principal country sources for 2012–15 were Japan, 33%; Canada,
31%; Spain, 16%; Germany, 5%; and other, 15%; however, much of the ton-
nage from Spain in 2013–14 may in fact have been from Italy, and slag from
the Netherlands and Switzerland in 2015 may have been from Germany.
Trends and Issues

The supply of blast furnace slag continues to be problematic in the
United States because of the closure and (or) continued idling of a num-
ber of active U.S. blast furnaces in recent years, including one in 2015,
the lack of construction of new furnaces, and the depletion of old slag
piles. Only a limited quantity of locally produced granulated blast furnace
slag was available. At yearend 2016, granulation cooling was available at
only two active U.S. blast furnaces, down from three in 2014. Installation
of granulation cooling was being evaluated at a few blast furnaces, but it
was unclear if this would be economic. Pelletized blast furnace slag was in
very limited supply (one site only), and it was uncertain if any additional
pelletizing capacity was planned. Basic oxygen furnace steel slag from
domestic furnaces has become less available recently because of the clo-
sure of several integrated iron and steel complexes; thus, the long-term
supply of steel slag will be increasingly reliant on electric arc furnaces,
which now contribute the majority of U.S. steel production. Where not
in short supply, slag (as aggregate) sales to the construction sector tend
to fluctuate less than those of natural aggregates. Domestic- and import-
supply constraints appear to have limited domestic demand for GGBFS
in recent years. Although prices have increased, sales volumes have failed
to match the relative increases that have characterized the overall U.S.
cement market since 2010. Long-term demand for GGBFS likely will
increase because its use in concrete yields a superior product in many
applications and reduces the unit carbon dioxide (CO2) emissions footprint
of the concrete related to the Portland cement (clinker) content. Recent
regulations to restrict emissions of CO2 and mercury by coal-fired power
plants, together with some power plant closures or conversion of others to
natural gas, have led to a reduction in the supply of fly ash in some areas,
including that of material for use as cementitious additive for concrete.
This has the potential to increase future demand for GGBFS, but the
availability of material to supply this demand will increasingly depend on
imports, either of ground or unground material. Imported slag availability
may be constrained by increasing international demand for the same mate-
rial and because not all granulated slag produced overseas is of high qual-
ity. New restrictions on mercury emissions by cement plants may reduce
demand for fly ash as a raw material for clinker manufacture, and this could
lead to use of air-cooled and steel slags as replacement raw materials.
World Mine Production and Reserves

Slag is not a mined material and thus the concept of reserves does not apply
to this mineral commodity. Slag production data for the world are unavail-
able, but it is estimated that global iron slag output in 2016 was on the order
of 300 to 360 million tons, and steel slag about 160 to 240 million tons,
based on typical ratios of slag to crude iron and steel output.

The study of chemico-thermal-physical processes, in commercial blast
furnaces and under simulated conditions in laboratories, has led to the
development of reliable theories of ironmaking in a blast furnace. It has
also established the limits to which the efficiency of a given furnace and
operation can be improved in practice. All these attempts have resulted
in decreased specific consumption of coke and maximum productivity to
improve upon the overall economy of ironmaking. It is also essential to
understand the ternary and quaternary oxide phase diagrams to appreciate
the nature of slag that should be produced to obtain the optimal slag metal
conditions inside the furnace.
CONSTRUCTION OF A
CHAPTER
11
BLAST FURNACE AND
ITS ACCESSORIES
In This Chapter
● Blast Furnace Refractories

● Blast Furnace Structure
● Blast Furnace Cooling Arrangements
● Tap Holes
● Cast House
● Tuyere Assembly
● Raw Material Section
● Charge Hoisting Appliances
● Top Charging System
● Blowers, Boilers, Pumps, etc.
● Slag-granulation
● Instrumentation and Control
T
he general description of a blast furnace and its accessories
have already been covered in Chapter 2. The present chapter
deals with some more details of its construction and those of
its accessories.
11.1 Blast Furnace Refractories

For many years, blast furnaces were lined throughout with high duty fire-
bricks, and despite the development of a special type of refractories, the
use of high duty firebricks still predominates. The life of the lining, under
the conditions prevailing inside the furnace, decides the furnace campaign
which should not be less than a few years. Although the thickness of lining
depends on the furnace size and on the location in the furnace, it may by
around one meter or so in thickness in most modern furnaces.
The chief causes of failure of the lining are:
1. Carbon monoxide attack.
2. Action of alkali vapors.
3. Action of limy and alkaline slags.
4. Action of other volatile matters.
5. Abrasion by solids, liquids, and gases.
6. Temperature.
7. Action of molten metal.
8. Conditions of operation and design.
9. Blowing-in procedure.
All these factors may not be operative at all the areas in a furnace. One or
a few factors, at best, may be dominant at any one area in the furnace. For
example, in the stack, the lining has to withstand predominantly abrasion
by a solid burden and attack of carbon monoxide, while in the bosh region,
the lining has to withstand high temperature, erosion by ascending gas, and
attack of molten lime and alkali slags. Similarly, the hearth has to withstand
the action of molten slag and metal without breakouts.
The present day high duty firebrick for blast furnace lining are made
of different grades to suit the requirements at various levels: the stack, the
hearth, the hearth walls, the bosh, etc.
Stack Lining
The lining here should have a very good abrasion resistance and resistance
to carbon monoxide attack; refractoriness is relatively of less significance.
Construction of a Blast Furnace and its Accessories • 237
A good, dense refractory is ideal for this purpose. It is a common practice to

use armor plates at the throat to withstand the abrading action of the falling
burden. Immediately below this, over a length of nearly 2–3 m, high-fired,
super duty firebricks are used. Trials to replace even the armor plates with
high alumina bricks have been successful. The entire stack below the top
few meters of height is lined with high duty firebricks. A 35–40% Al2O3
firebrick with a close texture is usually preferred for the stack. A 60% Al2O3
brick has been recommended for the lower parts of the stack. These bricks
are made by machine molding under high pressures and are de-aired to
improve their density. The use of standard type bricks or even large blocks
is currently in vogue.
Hearth Lining
The lining in the hearth should primarily prevent breakouts. The use of
bricks of high alumina to silica ratio, lower permeability and porosity, with
well laid joints can minimize breakouts. In spite of this, the earlier firebrick
hearth still suffered from frequent breakouts and the attendant troubles.
The occurrence of breakout was believed to be due to the oxidation of iron
by gases that penetrated the firebricks and the consequent effect of iron
oxide in lowering the melting point of the refractory, ultimately resulting
in its failure. Ramming of the carbon plus tar mixture at such vulnerable
areas exhibited excellent resistance to such breakouts. Finally, this lead to
the development and use of carbon lined hearths. High refractoriness, high
thermal conductivity, high abrasion resistance, high bulk density coupled
with low porosity, good crushing strength, almost complete inertness to
carbon saturated iron and slag, and such other properties make carbon an
almost ideal material for hearth construction provided bricks or blocks of
these are keyed into position with the thinnest possible joints. It has been
observed that not only the carbon hearth contour is better maintained dur-
ing the campaign, but the problem of breakouts is virtually eliminated.
In the early adaptation, carbon blocks were used only as the facing lin-
ing with high duty firebrick for backing. However, all carbon hearth, (i.e.
the whole wall thickness and a considerable bottom thickness), has almost
universally been accepted as a standard method of preparing the hearth;
the remaining bottom thickness is made up of high duty firebricks. The
shape and size of the carbon blocks used for making the bottom vary con-
siderably, but they all aim at achieving the thinnest possible keyed joints,
preferably without the use of external jointing material.
(a) details of hearth lining,

(b) method of laying the hearth layers and
(c) individual carbon block
FIGURE 11.1 Carbon lining of a blast furnace.
The modern, all carbon hearth construction uses large blocks of a few
square meters in size; lengths of approximately half the hearth diameter are
increasingly being adopted. These are laid horizontally and keyed together.
Each block is anchored firmly at the hearth wall to prevent it from floating
by molten iron. The shape of the individual carbon block and the method of
laying as adopted at the Bhilai Steel Plant in India are shown in Figure 11.1.
Some carbon hearths are air cooled or even water cooled at the bottom
by pipe or ducts inserted underneath the lowest course of carbon blocks.
The bottom of the 1719 m3 volume furnace at Bhilai is air cooled in
this fashion.
Careful control of the manufacture of hearth blocks and construction of
the hearth is of fundamental importance in achieving the desired campaign
life of the hearth.
Hearth Walls
Either the entire carbon lining or carbon block facing with high duty fire-
brick backing is employed for the walls right up to the central line of the
tuyeres or up to the top of the tuyeres in modern furnaces. The recent
trend is to adopt all carbon hearth up to the tuyere level or an even higher
level. Effective cooling of the hearth walls is improved by employing graph-
ite as the backing layer with carbon blocks facing to make up the lining.
This design improves thermal conductivity and the cooling efficiency of
the lining, particularly when spray cooling or water jacket cooling methods
are employed.
Bosh Lining
Considering the severity of temperature and chemical attack in this region,
the lining should possess good refractoriness. The refractoriness under the
load should be low after expansion and resistant to the action of molten
limy and alkali slags. The majority of bosh linings are of high duty or super
duty firebricks with 45–65% Al2O3, laid in the conventional banded bosh
construction with copper cooling plates. The extremely successful use of
carbon blocks for lining the hearth and its walls led to its adoption, even in
the bosh region. The carbon refractory possesses better properties, espe-
cially high thermal conductivity, than those of the conventional high duty
firebricks. Carbon lined walls can be cooled by either spray coolers or water
jackets. The changeover from firebrick lining to carbon lining, therefore,
eliminates the corrugated pattern of construction produced by rows of
coolers inserted in the lining and permits simpler construction which also
results in smoother and uniform wear. Carbonaceous materials like high-
conductivity graphite, electrographite or plumbago-clay graphite, etc. have
been used for preparing the bosh lining. The effective chill through the
high conductivity lining in the bosh generally results in early deposition of
essentially a mixture of slag and graphite on the face which in fact becomes
the effective lining.
The development and use of graphite-silicon carbide-brick in Japan has

given excellent performance both for the bosh and hearth and might find
wider use in near future. There still remains some doubt about the future of
carbon lined bosh, particularly when fuel injection is adopted.
Carbon-lined Furnaces
The successful use of an all carbon lined hearth has led to the development
of an all carbon lined furnace. The adoption of carbon even up to the bosh
level simplifies the cooling arrangement required in laying the lining. With
a carbon lined bosh and hearth, the conventional cooling plates are gener-
ally replaced by the external cooling of the shell. A carbon hearth, in par-
ticular, must be protected from oxidation by a thin layer of fireclay or some
other material during the blowing-in period. The advantages of carbon lin-
ing can be summarized as follows:
1. Increase in overall campaign life of the furnace.
2. Minimum breakouts and reduced scaffolds.
3. Cooling design becomes more simple.
4. More uniform wear of the lining.
5. Stack cooling may become unnecessary.
6. Clean surface in contact with slag and metal both.
7. A relatively thinner lining is adequate and more furnace volume is
effectively available for the same size shell. This would result in im-
proved productivity.
All these advantages compensate for more than the additional cost of lin-
ing, and in fact, its wider use may reduce the cost of the carbon refractories
in the near future. In spite of the advantages, the use of carbon lining can
create serious problems in practice if careful operation and maintenance of
the hearth, especially in the slag and metal tap hole regions, is not scrupu-
lously met. Gases and steam can damage the carbon lining. Faulty coolers
can lead to the formation of steam, when the leaking water comes in contact
with the red hot lining. A short tap hole aggravates the problems. Armoring
bars are generally embedded in the lining at the throat to withstand the
falling burden and safeguard the lining. This has been extremely success-
ful. It should, however, not continue up to the zone of carbon deposition by
the Naumann reversion reaction Eq. (10.2), since iron acts as a catalyst for
this reaction. Carbon deposition has been observed to corrode the furnace
lining in practice.
11.2 Blast Furnace Structure

The structural design of the blast furnace has undergone radical changes
with the progress in its overall design, and particularly with its increasing
size. In the early designs, the stack as well as the entire superstructure of
the furnace including charging arrangement, gas uptakes, etc. were sup-
ported by the solid iron pillars from the mantle ring.
FIGURE 11.2 Supporting structure of a modern blast furnace in which the pillars have been replaced
by the box-girders.
For larger furnaces, this is not a stable structure and the modern
trend is to have a box girder jacket surrounding the furnace to support
the furnace in place of the pillars. The furnace is generally provided
with two box girder jackets, one above the other; one is for the stack and
the other is for the super-structure with an expansion joint in between.

The simplified scheme is shown in Figure 11.2, in which the pillars are
eliminated. In this way, the furnace shell, which is subjected to heat and
distortion, is relieved of the loads imposed on it by the superstructure.
It also facilitates replacement of the furnace shell more readily than in
any other design. It allows more clear space for working around the fur-
nace bottom, particularly for tuyere repairs and replacements, iron and
slag hole opening and closing, etc. It also allows a maximum number
of tuyeres to be fitted onto the furnace without any other restriction,
with all its attendant advantages. In this structural design, there need not
be any interrelationships between the number and location of tuyeres,
iron hole(s), slag holes and the supporting columns, since the latter are
non-existent. This design facilitates unobstructed access to the bosh and
hearth to carry out relining more easily. Consequently, this saves on time
and cost.
The box type structure also poses some difficulties related to accessing
the furnace parts for repairs and installing fuel injection systems. Therefore,
BFs are now being designed and constructed as freely standing structures
with free access from all around to any part of the furnace. It also facilitates
tuyere maneuvers without any difficulties. The newly constructed modern
G furnace of Tata Steel is of this type.
11.3 Blast Furnace Cooling Arrangements

The provision of good quality refractory lining is by itself no guarantee that
the furnace life will be adequate. For a longer life and normal functioning
of the blast furnace, cooling of the lining, particularly in the hearth and
the bosh regions, is essential. It keeps the shell temperature within limits
and thereby prevents expansion of the furnace during normal working con-
ditions. Interruption of cooling, even for a short duration may, therefore,
cause serious consequences. An effective cooling arrangement can bring
about a proper temperature gradient across the lining thickness. It can be
advantageously used in reducing the effective lining thickness required for
the furnace without in any way sacrificing its campaign life and obtain more
useful volume for the same size of the shell along with all its consequent
beneficial effects.
The type of cooler designs used vary with the size of the furnace and
also with reference to the location on the furnace where it is to be installed.
Four different modes of cooling arrangements are in use, and these are as
follows:
1. The box cooler to be inserted in the lining and affixed to the shell of
the furnace.
2. The spray coolers used externally on the shell.
3. The water jacketed shell.
4. The cooling pipes for air or water circulation, particularly for cooling
the bottom.
5. The evaporation cooling.
Three different types of box coolers are currently in use, namely the
cantilever-type, the L-type, and the plate type. For the same furnace, the
designs of box coolers for shaft, bosh, and hearth are different. In general,
cast iron boxes, with steel tubes embedded inside, are used for shaft cool-
ing, whereas copper boxes with copper tubes embedded inside are used
for belly, bosh, and hearth, since cooling has to be more effective at these
areas than in the shaft. The modern trend is to use copper coolers even for
the shaft. To make the coolers the pipe through which water circulates, it
is first bent in the required form and embedded during founding in cast
iron or copper casting as per the requirements. The water inlet and outlet
project are outside the casting. Holes are also provided for the insertion of
the thermocouple to measure the temperature. The coolers have a flange
which, after insertion through the furnace shell, can be bolted onto the
flange provided on the shell. The coolers are provided all around the shell
at any desired level and individually each one is to be properly fixed with
respect to the other facilities like tuyere holes, slag and iron notches, etc.
Shaft Coolers
For smaller furnaces, cantilever type coolers are used whereas for larger
furnaces L-type coolers are used. The cantilever-type coolers as laid in
the 1719 m3 volume furnace of Bhilai are shown in Figure 11.3. There
are twelve rows of each 24 coolers, and these are staggered in the checker
fashion. The L-type coolers are used in a 2000 m3 furnace of Bokaro are
shown in Figure 11.4. There are seven rows with forty coolers in each
row. The L-type coolers are cast with firebrick on one face for clean con-
tact with the lining. A recent trend is to use external water sprays to cool
the shell. This is more effective if carbon lining is used. In the Rourkela
Steel plant, all the furnace shells are cooled by water sprays half-way
down the stack to the bottom. The upper half of the stack is cooled by the
cantilever type coolers.
FIGURE 11.3 Cantilever-type blast furnace coolers (only a few are shown)
FIGURE 11.4 L-type coolers (only a few have are shown)
On the contrary, the furnaces in Bhilai Steel Plant are cooled by coolers
inserted in the lining, in spite of the furnace being bigger than those of the
Rourkela Plant.
Hearth and Bosh Coolers

Plate-type flat copper coolers are used to cool the lining in the bosh and
hearth walls of the furnace. Maximum shell surface area is covered by these
coolers. The number of coolers in the bosh ring are almost fifty per cent
more than those in the lower portion because of the presence of several slag
and metal holes in the lower region and the relatively smaller diameter of
the shell in the hearth.
In addition to the inserted coolers, external water cooling of almost
the entire shell has been adopted since carbon is invariably used for lining
the hearth or the hearth and the bosh. Thermal conductivity of graphite is
nearly 18 times superior to ordinary carbon blocks. Use of backing layers of
graphite with carbon blocks on the face have, therefore, been particularly
recommended for effective cooling of the hearth.
External spray cooling of the hearth is rather difficult because of the
oblique walls. Use of water jackets for cooling the bosh and the belly is also
recommended. For this purpose, cooling pockets are welded onto the shell
jacket and the water passes through in an upward direction. This has been
claimed to give nearly perfect cooling, and the lining thickness could be
reduced to only 600–800 mm.
In the normal water cooling arrangement, heat is extracted by the
water. In this process, the water may get heated to a certain safe limit
like 40 degrees C or so. This means that each unit of water extracts a heat
equivalent to that required in raising its temperature from the normal room
temperature to 40 degrees C, that is, by about 20 degrees C. The amount is
fairly small, and the amount of water required to be circulated for effective
cooling is, therefore, enormous. For effective cooling, large pumping facili-
ties have, therefore, to be provided.
The amount of heat extracted by water in getting itself converted to steam
is more than 20 times that extracted in heating the water by one degree cen-
tigrade. Cooling designs, using water at 100 degrees C, have been developed
and commercially adopted in Germany. The heat is extracted in the cooler
by converting the water to steam at the same temperature. This is known as
evaporative system of cooling which extracts heat equivalent to the amount
of steam formed during cooling at the boiling point of water. The amount of
water to be circulated for effective cooling is, drastically reduced as compared
to traditional water cooling. Corrosion and other problems are increased, but
these have to be solved satisfactorily. The system is shown in Figure 11.5.
FIGURE 11.5 Evaporative cooling arrangement.
The furnace shell is a double plate shell, and boiling water is circulated
in the space in between these in the upward direction. This may prove use-
ful particularly useful for the very large capacity furnaces.
Cooling of Hearth Bottom

When the hearth diameter is small, the thermal conductivity of carbon
blocks is adequate to remove sufficient heat from the center of the hearth
by the hearth cooling staves. The increasing size of the furnace, however,
increases the specific thermal load on the hearth bottom and makes it more
vulnerable. Bottom cooling has been recommended and used to minimize
this vulnerability. In a simple cooling arrangement, cast iron plates are laid
to cover the entire bottom. Steel pipes are inserted through these plates
to ensure air circulation for cooling. The gap between the steel pipes and
the cast iron plates is packed with iron putty consisting of cast iron chips,
ammonium chloride, sulfur and sulfuric acid.
In another but more effective design, air was supplied through a centrally
located vertical pipe and carried outwards through a radial pipe. This arrange-
ment, as shown in Figure 11.5, has the added advantage in that the cold air
is supplied at the hottest spot of the bottom. It is theoretically estimated that
this cooling arrangement may reduce the bottom thickness by 600–800 mm.
The load on the bottom cooling can be reduced by effecting larger dissipation
of heat through the hearth walls with its improved design and improved cool-
ing arrangement. The bigger 1719 m3 furnaces of Bhilai are bottom cooled by
air whereas the Rourkela furnaces are not.
On the very big modern furnaces, bottom cooling by water has been
adopted as shown in Figure 11.6.
FIGURE 11.6 Bottom cooling arrangement.
11.4 Tap Holes

Molten iron accumulates in the hearth and is periodically removed from
the furnace through a hole located a little above the topmost brick layer of
the bottom. This opening is known as tap hole or iron notch. An opening
at the appropriate level is left in the furnace shell, and a water cooled cast
steel frame is fitted in this opening to make up the tap hole as shown in
Figure 11.7. The steel frame is lined all around from inside with fire-
bricks, and a considerable space is left which is later packed with fireclay.
The actual tap hole is much smaller as compared with the area patched with
fireclay. The tapping hole is normally closed with the help of an electric
clay gun which swings-in in front of the hole. For tapping metal, the hole is
opened by drilling through the patched fireclay with the help of an electric
tap hole drilling machine. The hole is closed again after the tapping is over
with the help of the electric clay gun.
FIGURE 11.7 Tap hole for metal.
The slag or cinder, as it is often called, is removed from the furnace

through a hole located at a level above the iron notch, but the two are
spread out through an angle of nearly 45 degrees. The slag hole is made by
fixing a water cooled copper ring in the furnace shell.
FIGURE 11.8 Slag-notch.
It is commonly known as the monkey on the shop floor. The slag-notch

assembly consists of three concentric, water cooled frustums of cones of cast
copper as shown in Figure 11.8. The outer-cone is the largest and is called
the monkey cooler, and the intermediate and the innermost segments are
known as the monkey itself. These are tapered and are designed to support
each other so that the replacement of the faulty cooler is easily accomplished.
Water is supplied constantly to these coolers since they are always exposed
to a high temperature. The life of the monkey is as low as a few days, and at
best, it may usually be a month. The life can be improved few fold by giving
a calorizing treatment to the monkey proper. The central cooler is packed
with fireclay. Since the slag hole is opened nearly every 2–3 hours, it is closed
with a readily removable slag-notch-stopper which carries a head to fit tightly
in the monkey. The stopper rod is also water cooled and is held in position
with a pneumatically compressed piston. Most furnaces provide an additional
cinder notch, but it is staggered further through nearly 45 degrees from the
main cinder notch.
For furnaces producing up to about 1500 t/day, provision of one iron notch
and two slag notches are adequate. For larger sizes, additional metal and slag
holes become necessary for removing these molten products without prolong-
ing the tapping duration. With the progress of furnace campaign, the bottom is
eroded and a large mass of metal known as the salamander (or bear) gets accu-
mulated at the bottom of the hearth. The metal accumulates at a lower level
than the iron notch and does not get tapped during normal tapping operation.
Additional iron notches may be provided to remove the salamander at the time
of blowing-out before shutting off the furnace and wrecking the old lining.
11.5 Cast House

The area around the blast furnace, in the form of a raised concrete platform in
which channels are provided for metal and slag to flow to their receptacles, is
known as a cast-house. It is a huge, massive concrete platform resting on col-
umns. The cast-house is provided with an electric drill to open the iron notch,
an electric clay gun to close the iron notch, and a slag-notch-stopper actuat-
ing mechanism. Tuyeres and hearth coolers are often serviced from the cast-
house with the help of chord gantry cranes. An E.O.T. crane is provided in
the cast house for erection and maintenance of equipment in the cast-house.
Until recently, the cast-house used to have a sand bed in which chan-
nels were cut for the flow of pig iron and slag. Pig iron was allowed to
solidify in these channels. The resemblance of this pattern of solidification
with a litter of pigs sucking the mother led to referring to the solid product
as pig iron. In spite of the fact that pig irons are no longer produced by this
old practice in the cast-house, the name still continues.
In a present day cast-house, the concrete platform is covered with a layer

of firebricks. Steel troughs lined with firebricks are laid on the platform in
front of the holes to lead the molten metal and slag to their ladles. The chan-
nels are provided with gates to control the rate of flow and its direction to any
one of the two or more spouts constructed at right angles to the channel. The
spouts are spaced in such a way that each ladle comes below one spout when
the ladle car is brought to the cast-house to receive the metal and the slag. The
general scheme of arrangement of stationary runners is shown in Figure 11.9.
This arrangement varies with the size of the furnace and its design. A certain
amount of slag always flows out of the iron notch, and it is detrimental to allow
this slag to go to the metal ladle. The metal runner is, therefore, connected to
the slag runner in such a way and the gate is so opened that the metal passes
out through the gate opening, and the slag flows back into the slag runner.
For a large size furnace producing more than about 2000 t/day, the
number of spouts and the network of runners required, involve a volumi-
nous cast-house preparation for tapping. Swiveling types of runners have,
therefore, been developed, and they can do the cast-house job with mini-
mum preparation. This type of runner has been utilized on the bigger fur-
naces of Bhilai and Bokaro Steel Plants.
FIGURE 11.9 Stationary and swivel runner arrangement in a cast-house.
The general arrangements are shown in Figure 11.9. In this arrangement,

the ladles are brought on two parallel rail tracks. The ladles on the main track
are filled through the stationary runners, while those on the secondary track
are filled using the swiveling runner. The ladles on the main and the second-
ary tracks are filled alternately. The swiveling runner swings into action every
time after filling one ladle on the main track by the stationary spout.
11.6 Tuyere Assembly

The hot blast from the stove is introduced in the furnace through a bustle
pipe and a tuyere assembly. The tuyeres are located uniformly all over the
furnace periphery, just above the hearth level. Hence, the bustle pipe sup-
plying the hot blast circumscribes the furnace at the bosh level. The bustle
pipe is a large steel pipe with asbestos and firebrick lining from inside to
reduce heat losses. The function of the bustle pipe is to supply hot blast at
uniform pressure to all the tuyeres so that a uniform combustion zone is
obtained inside the furnace.
The tuyere assembly is schematically shown in Figure 11.10. It essen-
tially consists of a bend pipe, expansion joint, elbow, and a nozzle. The water
cooled tuyere breast assembly is bolted or welded to the furnace shell. The
water cooled tuyere nozzle fits tightly in the breast assembly. The elbow is
connected to the connecting rod assembly which keeps the elbow pressed
against the nozzle body even after thermal expansion. A glass window is
provided in the elbow to observe the combustion zone inside the furnace.
FIGURE 11.10 The tuyere assembly. The modified design with expansion joints is shown in (a).
The tuyere assembly is made up of several parts in order to help replace

the faulty part easily and speedily. The tuyere nozzle and its body (blow
pipe), and the breast assembly are all made of copper either as cast or rolled
shape. The rest is as hematite or steel castings.
The original bellow-type expansion joints connecting the blowpipe with
goose neck are not satisfactory for higher temperature and pressure. Lately,
they have been replaced by two universal-type expansion joints. The elbow
is directly connected to the bustle pipe while eliminating the goose neck.
This is shown in Figure 11.10 (a). Its performance for hot blast at as high as
1350 degrees C temperature has been successful. In the improved design,
the heat loss for a blast volume of 3300 Nm3/min at a blast temperature of
1100 degrees C in the design was 7400 kcal/min. It changed to less than half
this value with the improved design.
11.7 Raw Material Section

At least a few blast furnaces are generally located in a row in a steel plant.
The raw material section runs parallel to the furnace line on one side, while
the gas cleaning system is located on the other side. It has a large ore yard
with ore blending and handling facilities, the trestle, two rows of bins, scale
car, ore transfer car, etc. The design of the ore yard depends upon the
sources and mode of transport of iron ore to the plant. The plants, which
run on imported ore, are generally situated on the waterfront and have a
more elaborate arrangement for unloading and transporting the ore from
the ship to the ore yard. Whether the ore is brought in ore carrier ships or
in rail wagons from mines, the ore is finally dumped in the ore-yard. The
ore-yard is a large space running parallel to the furnace line where ore
blending is carried out to prepare the charge of more uniform composition.
Ore bridge cranes with grab buckets are provided in the yard for bedding
and reclaiming.
The trestle consists of a reinforced concrete wall on the stockyard side
and steel columns on the furnace side, between which a cross-work of
transverse and longitudinal girders are laid. Rail tracks are laid on these
girders and bins are provided underneath. The rail track nearest to the ore-
yard carries a side dump bin or transfer car to transfer the blended ore to
the bins of all the furnaces. The rail track next to this carries limestone,
dolomite, sinter, manganese ore, etc. The remaining one or two tracks are
meant to supply coke to the bins from the coke ovens. The bins are laid in
two rows beneath the trestle; one row is generally kept for coke while the
other is kept for all the other types of burden materials. The materials are
dumped on the trestle, through which these fall in the appropriate bins, by
manual operation of the hopper gates of the cars or the wagons.
Coke of the right size is supplied to the bins from the coke ovens by
belt conveyors, and the coke yard becomes unnecessary. The same is true
with sinter or pellets since these are supplied from the sinter or the pel-
let plant located either inside or outside the plant. If pellets are received
from a distance, the appropriate storage facilities like that for ore will have
to be provided in the blast furnace plant. Limestone, manganese ore, and
such other raw materials are dropped directly in the bins from rail wagons
through the trestle.
It is positively beneficial to screen ore, pellets, sinter, and coke just before
putting them in the skip. Many plants provide facilities for this last minute
screening. The fines of iron bearing materials thus screened off are sent for
further agglomeration and those of coke for secondary uses. Coke is screened
at the furnace in India in the Bhilai and Bokaro plants only. This facility does
not exist in other plants. Other raw materials are not screened at the furnace
in any of the plants in India. One need not be surprised to see a large propor-
tion of ore fines, associated with mine dispatch and those generated in the
ore-yard, because of handling, finds its way ultimately in the furnace. The fine-
fraction of the natural ore charge under Indian practices is, therefore, as high
as 18–30%. It definitely affects the furnace operation adversely. (cf. Table 16.2)
Automatically controlled systems for sampling, weighing, and trans-
porting the raw materials have now come into vogue, particularly in large
furnaces. Beneath each bin, weighing mechanisms are fitted and belt con-
veyors are used to deliver the material to the skip pit. Computer controlled
programmers are assigned to control the operation effectively.
11.8 Charge Hoisting Appliances

The solid charge materials are hoisted up to the furnace top by either of the
following three methods:
1. Bucket hoist;
2. Skip hoist;
3. Conveyor belts.
Although bucket hoist was adopted widely on the continent, it has

almost completely been replaced now by the skip hoist. The conveyor belt
system has also proved to be quite satisfactory and has been adopted even
in some of the recently constructed blast furnace plants.
In the skip hoist, two rail tracks are laid over an incline extending from
the furnace top to the bottom of the stock house. Two skips, one on each
track, move up and down over these two tracks. The skips are hauled up and
down by a steel rope wound on a steel drum driven by an electric motor.
The skips are so attached to the two ends of the rope that when one skip is at
the bottom, the other is at the top. In other words, when the bottom skip is
being filled with raw materials, the other at the top is discharging the mate-
rials in the top charging device. The drum, the motor, their control system,
etc. are housed in the hoisting cabin located above the skip pit as shown in
Figure 11.11. The raising and lowering of the skips and the operations of the
furnace top charging device are all controlled from the hoisting cabin.
FIGURE 11.11 Simplified design showing the details of the skip-pit.
In the belt conveyor system, the materials are raised to the furnace
top by a system of belts. Since the maximum angle through which belt can
raise the materials is fairly shallow (maximum angle decided by the angle of
repose of the material), the bins have to be located further away from the
furnace than that of the skip hoisting system. An independent belt conveyor
system for each furnace, therefore, becomes unwieldy and uneconomical.
A common belt conveyor system for a group of furnaces has been found to
be convenient and economical. This is the reason why it has been preferred
in a few of the recently constructed blast furnace plants. An interesting belt
conveyor system has been described by Hookham.
Four old blast furnaces were replaced by the three modern, large capac-
ity furnaces in one of the plants in Belgium. Since the space available did
not permit installation of independent skip-hoist-systems, a belt conveyor
system was successfully adopted for all three furnaces. It has been working
to the designed degree of satisfaction.
11.9 Top Charging System

The furnace top is a complicated design since it includes the charging sys-
tem and the gas outlet assembly along with their necessary safety devices.
Until recently, the double bell arrangement was universally adopted as the
furnace top charging system. It prevented the gas from escaping out while
charging the furnace. The details of this design have already been covered
in Sections 5.1 and 5.3. The raising and lowering of the bells, the operation
of the charge level indicators, etc. are controlled from the furnace control
cabin to ensure proper burdening of the furnace.
The modern designs of the furnace tops, particularly suited to adopt
high top pressure, will be described in Chapter 17, while describing the
high top pressure operation. The details of uptakes, downtakes, and bleed-
ers will be described in the next chapter.
11.10 Blowers, Boilers, Pumps, etc.

One of the most important raw materials in blast furnace practice is the air.
It is almost universal that steam driven turbo-blowers are used to supply
air to the furnace via stoves. The modern practice is to blow the furnace
at a constant volume of air, the pressure being regulated by the size of the
tuyeres. The blower is automatically controlled to regulate constant volume
of blast. Boilers fired by blast furnace gas are provided to generate steam
for running these blowers.
Since enormous amounts of water are required to cool the furnace and a
few of its accessories, large pumps are provided to maintain a constant supply
of water to the furnace without interruption even for a short duration.
11.11 Slag-granulation
Until recently BF slag was collected in slag thimbles and was dumped in
molten condition in the dumping yard. In this process, it got contaminated
with external muck of unknown chemistry. It took longer to cool in these
dumps. But this was immaterial, as it used to lie there thereafter almost
for ever. Now-a-days, BF slag is entirely used as a raw material for cement
making. But this requires the slag to be granulated for easy transport, uni-
form chemical composition, and the right glassy structure of solidified slag
without external contamination. The molten slag is carried, through the
channels, into stirred water so that it is quenched and converted into gran-
ules. This arrangement has to be provided anew, on old furnaces, for the
end-utilization of the slag. Various designs of slag granulation plants are
available. The amount of water to be circulated must be such that its tem-
perature should not rise during this process beyond the designed level.
11.12 Instrumentation and Control

Each furnace operation is controlled from the furnace control cabin which
is provided with several measuring, recording, and controlling devices.
The optimum operating conditions of the furnace are known, and every
attempt is made to run the furnace under those conditions. Instruments are
provided to measure and adjust the variables affecting the operation. The
instruments that may be provided are as follows:
1. Blast temperature and pressure recorders and indicators.
2. Blast volume recording indicator.
3. Stockline recorders and indicators (usually four in number).
4. Top gas temperature and pressure indicating recorders.
5. Automatic hot blast temperature controller.
6. Stove dome temperature recorder.
7. Automatic combustion control device for stoves.
8. Stove stock temperature recorder.

9. Sequence recorders of large bell movements and revolving distributor.
10. Alarm warning for low gas pressure, high stove dome temperature,
high stack temperature, low blast pressure.
11. Furnace wall temperature indicators.
The furnace operation is essentially controlled by maintaining the blast
temperature, blast pressure, and blast volume within the accepted limits
and by proper sequencing of the charge materials.

On the whole, a number of technological features, as indicated below are
provided in modern blast furnaces:
1. Free standing structures
2. Bell-less top: either Paul-Worth or Gimble design
3. Belt conveyor charging in place of double-skip hoists
4. Fully automatic conveyorized stock house with electronic weighing
system
5. Hot blast stove with external or top combustion chamber
6. Intense furnace cooling with soft water in closed circuit or evapora-
tion cooling
7. Modern refractory lining design
8. Cast house with more than one tap holes and slag notches
9. Slag granulation facilities
10. Expert system for complete or partial automated process control
11. Dust extraction system at the stock house
12. Fume extraction system at the cast house
13. Fuel injection system
All of these may be incorporated in the modernization of the existing fur-
nace, to the extent permissible and possible. The large capacity modern
furnaces must have all these modern design features to a certain extent.
BLAST FURNACE
CHAPTER
12
PRODUCTS: BLAST
FURNACE GAS CLEANING
AND UTILIZATION
In This Chapter
● Uptake, Downcomer, and Bleeder

● Dust Catcher
● Semi-Fine or Primary Cleaning
● Fine or Secondary Cleaning
● Utilization of Blast Furnace Gas
● Hot Blast Stove
I
t has been made clear that even the most efficient modern blast fur-
nace would produce an effluent gas containing a significant proportion
of carbon monoxide which cannot be used for iron oxide reduction. The
actual CO content may vary around 20–30% by volume. This has a calorific
value of nearly 900 kcal/m3. The quantity of gas produced depends upon the
amount of fuel burnt. For one ton of coke burnt, nearly 4000 m3 of effluent
gas may be produced. A blast furnace requiring 1000 t of coke per day would
generate nearly 4 ×106 m3 of gas with a total energy content of 3600 × 106
kcal which is nearly equivalent to 500 t of coke. Although various uses of the
blast furnace gas, including recovery of by-products, have been suggested, it
has not been possible to use the gas for any other purpose other than just a
lean fuel.
The effluent gas from the furnace cannot directly be used as a fuel, since
a substantial quantity of dust from the burden is also discharged along with.
This may lead to accumulation of dust and wear in the equipment using
the gas. The gas is, therefore, cleaned before its use, and in so doing, the
sensible heat of the gas is invariably lost. The chemical heat of the cleaned
gas is what is utilized.
Neglecting the large lumps blown out occasionally when the furnace
slips, the dust particles carried in the gas may vary in size from nearly 6 mm
down to few microns. The average dust content may vary in the range of
7–30 g/m3. In general, cleaning is carried out in three stages: coarse, semi-
fine, and fine cleaning. A typical scheme is shown in Figure 12.1. The coarse
cleaning is done in dust catchers and cyclones in dry condition. The dust
content of the coarse cleaned gas is nearly 5–10 g/m3. The semi-fine clean-
ing is carried out in scrubbers, venturi washers, cyclone separators, centrif-
ugal disintegrators, feld washers, or even in electrostatic precipitators. The
dust content is thereby reduced to 0.5-; 1.5g/m3. Fine cleaning is carried
out mainly by electrostatic precipitators, or at times, by high speed rotary
disintegrators. The dust content is thereby reduced down to 0.01 g/m3.
FIGURE 12.1 Scheme of blast furnace gas cleaning system.
The semi-fine and fine cleaning is carried out either in wet or dry
conditions. Wet methods are generally preferred to dry methods for their
better efficiency and smooth operation.
12.1 Uptake, Downcomer, and Bleeder

The effluent gases from the furnace are carried out of the furnace by large
vertical pipes called uptakes, normally four in number.
Blast Furnace Products: Blast Furnace Gas Cleaning and Utilization • 261
FIGURE 12.2 Scheme of uptakes and downcomer arrangement two different designs are shown in (a)
and (c) while (b)shows the other view of the design in (a).
Two adjacent uptakes are joined together to form one single duct, and
the two such ducts, thus formed, are connected to form only one duct which
carries the gas downwards into the dust catcher. The downcoming pipe or
duct is called the downcomer. The general scheme of uptakes and down-
comer are shown in Figure 12.2. An open-in or open-out bleeder valve is
provided at the top of each joined pair of uptakes.
A bleeder valve is a safety device, which opens automatically or is
opened, to release extra pressure developed inside the furnace and thereby
eliminate the danger of explosion. Bleeder valves are either counter-
weighted or operated by air cylinders. Some furnaces have separate bleed-
ers for each uptake. If the bleeders are situated on the joint-pipe, a third
bleeder is provided at the top of the downcomer. Bleeders are located at
the highest levels of the furnace structure.
The uptakes and the downcomers are steel pipes and are lined from
inside with firebricks or a monolithic castable cement applied over a wire
mesh that is anchored from inside to the shell. The sizes of the uptakes and
downcomers and the angle of their joints are such that gas flows smoothly
out of the furnace. The uptakes should be located on the furnace-top
periphery at those points which are not directly vertical above the iron-
notch, slag notch, blast main entrance to the bustle pipe, etc. These are
active points of the furnace, and if the uptakes are located right above these
points, it may cause uneven distribution of the gas through the burden. The
entire design should also ensure that a minimum amount of dust is carried
from the furnace with the gases.
12.2 Dust Catcher

The blast furnace gas is led by the downcomer directly into a dust catcher,
which as the name implies, removes as much of the flue dust as possible.
The dust catcher operates on the principle that the dust is contained in the
gas by virtue of its velocity. Hence, if the velocity is zero, the dust particles
will be eliminated from the gas. The state of zero velocity is achieved in the
dust catcher by simply reversing the direction of the gas flow, from verti-
cally downward to the vertically upward direction.
It is a cylindrical steel structure with a conical top and bottom section as
shown in Figure 12.3, and it is lined from inside with firebricks.
FIGURE 12.3 The dust catcher.
The downcomer leads the gases into the dust catcher through a vertical
pipe that flares out and extends nearly to the bottom of the dust catcher.
The gas comes down vertically and suddenly changes its direction to verti-
cally upwards since the outlet is located nearly at the top of the dust catcher.
This brings the gas to a point of zero velocity in from of the tip of the vertical
inlet pipe, and coarse dust falls in the bottom cone of the dust catcher from
where it is periodically removed. The dust catcher has valves for gas inlet
and dust removal. A cut-off valve when closed separates the dust catcher
from the downcomer, if required.
The efficiency of the dust catcher is usually 60–75%. It is also a practice to
install two dust catchers in a series, on some furnaces, to effect proper coarse
cleaning of the gas and to reduce burden on the semi-fine gas cleaning devices.
12.3 Semi-Fine or Primary Cleaning

The gas from the dust catcher is led into the primary or semi-fine gas cleaning
system which consists of any one of the following units:
1. Scrubbers or spray towers (stationary);

2. Venturi washer;
3. Revolving spray-towers;
4. Feld washers;
5. Spray fans.
Out of these, the wet scrubbers are most popular, and lately spray towers,
in their modernized designs, are becoming popular. The venturi scrubber
is also being adopted.
Scrubbers or Spray Towers

Scrubber or spray towers consist of a tall cylinder with conical top and bot-
tom. The gas enters at a point near the bottom of the cylinder and leaves
through a centrally located outlet at the top as shown in Figure 12.4. Water
is fed from top in the form of, one or more rows of water sprays to cover the
entire cross-section of the cylindrical shell. Inside the cylindrical shell, several
layers of wooden packing with perforation are fitted to break up the gas flow
into small jets and bring intimate contact of gas and water. The dust contained
in the gas sticks to the wooden packings and is washed down with water.
The coarse, cleaned gas from the dust catcher enters the scrubber at an
angle of nearly 50 degrees. Its velocity is reduced considerably because of
sudden increase in volume, and as a result, some coarse dust settles at the
bottom as it becomes wet. Finer dust is washed down from the top.
FIGURE 12.4 The wet scrubber for primary cleaning of blast furnace gas. The details of the baffles are
shown in (a) and the details of the water spray arrangement are shown in (b).
The scrubber not only cleans the gas but cools it to 30–35°C. The slurry
that is formed collects at the bottom and is continuously removed from
there into a thickener for separating water from the solid for recirculation.
Nearly 80–90% of the dust contained in the incoming gas is removed by this
method with about 3.5–4.5 liters of water consumption per cubic meter of
gas. The dust content of the outgoing gas is 0.5–1.5 g/m3. A centrifugal type
water separator at the exit end of the gas separates any entrained water in
the outgoing gas.
Venturi Washer
It is a vertical unit with a construction where water is injected in the form of
sprays to wet the gas moving downwards as shown in Figure 12.5. Two sets of
water sprays are used, one at right angle to the direction of the gas flow and
the other at an obtuse angle like 110–120 degrees. The washer is lined from
the inside with abrasion resistant material to withstand the erosive dust-laden
gas. A water separator is located at the bottom of the venturi washer and a tall
cooling tower is located by the side to cool the gas. A moisture eliminator is
provided finally to remove the excess entrained water in the cleaned gas. The
dust content of the clean gas is around 0.05 g/m3.
FIGURE 12.5 Venturi washer for primary cleaning of blast furnace gas.
12.4 Fine or Secondary Cleaning

The fine or secondary cleaning is carried out both in wet and dry conditions.
Wet-type cleaners are either electrostatic precipitators (Cottrell design),
high speed disintegrators, or Theisen disintegrators.
The Theisen disintegrator is essentially a fan that brings water and
gas in intimate contact and separates them in the opposite direction. It is
already out of date, and hence, is not described here.
The high speed disintegrators essentially mix water and gas with a high
speed rotor. Blades are provided both on the rotor and the stator body from
inside to achieve this objective. Water is atomized to an almost misty condi-
tion. The centrifugal action of the rotor forces the wet dust to stator walls
where it forms part of the slurry and is removed. The cleaned gas from the
spiral space goes out into a water separator. The dust content of the cleaned
gas is 0.015–0.020 g/m3.
Electrostatic Precipitator (ESP)

It is essentially based on the principle that under an action of high applied
static voltage, the dust particles in a gas phase acquire electrostatic charge
and are attracted to an electrode of opposite polarity where they are col-
lected and washed out. The corona effect also helps this behavior. The
unit consists of either tubes or plates as one electrode and wires, centrally
located in each tube or parallel with the plates, as the other electrode as
FIGURE 12.6 Wet electrostatic precipitator (ESP) for secondary cleaning of blast furnace gas. ESP can
be used as an independent unit as in (a) or in combination with scrubber as in (b).
Rectified high voltage is applied across these two electrodes. The gas enters
from the bottom and rises upwards through a system of such cells, while a thin
film of water flows over the surface of the tube or plates from inside down-
wards. The dust particles on ionization get attracted to these wet surfaces and
are washed down. The slurry is led to a thickener. The dust particles are pre-
cipitated on the electrode. The precipitator is usually divided into two zones,
one of which is working while the other is under repair, if required. In some
designs, the precipitator is mounted directly above the primary washer.
The rotary disintegrator has a lower capital cost but higher running
cost as compared to the electrostatic precipitator. The cleaning efficiency is
however better in the case of electrostatic precipitator and it is practically
universally adopted. The ESP is also used widely for cleaning the dry efflu-
ent gases before they are let off into the atmosphere as per the pollution
control legislation. The wet cleaning unit is used for blast furnace gas clean-
ing due to its superior performance.
Although wet cleaning is better, it results in a wash water disposal prob-
lem which necessitates the use of an elaborate arrangement involving thick-
eners, filters, pumps, etc. The dust thus recovered consists mostly of iron
ore and is often circulated back into the agglomeration circuit. The clean
gas is stored in a huge gas holder and is distributed from there in the plant
wherever necessary.
12.5 Utilization of Blast Furnace Gas

Although the potential energy of the high top pressure furnace gas has been
successfully used by converting its mechanical energy into electrical energy,
through turbines, as the latest innovation, it has yet to be widely adopted in
practice. The use of blast furnace gas as a fuel is, however, a universal prac-
tice. In general, cleaned blast furnace gas may be used for the following:
1. Preheating the blast for the blast furnace through stoves.

2. Firing of reheating furnaces, in an integrated steel plant.
3. Firing of an open hearth furnace by mixing it with coke even or pro-
ducer gas.
4. Steam raising for varieties of uses including vacuum steelmaking.
Nearly 25–40% of the total gas generated is consumed in preheating the blast
in a modern blast furnace plant (i.e. for firing of hot blast stoves). It was first
adopted by Neilson in 1828, and since then, it has been increasingly used to
obtain a preheated blast of higher temperature. The cleaned gas is stored in

huge gas holder and is supplied therefrom wherever necessary in the plant.
The construction and management of the gas holder involves capital as well
as running expenditure. The Bhilai Steel Plant has dispensed with the holder
by balancing the amount of gas generated and used in the plant. In case any
excess is left, it is burnt. A pilot flame of coke gas is kept burning to light the
excess available gas. The consumption is kept slightly on the higher side so that
not much is available as excess gas. The maximum amount of thermal energy
contained in the BF off-gas is recirculated back into the BF via the pre-heated
blast. Over the years, the hot blast temperature has steadily increased to nearly
1200°C. The more the temperature increases, the more gas is used for its gen-
eration. Thereby, the more the thermal value of coke is fed back into the BF.
12.6 Hot Blast Stove

The blast required for running a blast furnace is first heated in what are com-
monly known as hot blast stoves. Although their designs vary, it essentially
consists of a tall cylindrical structure comprised of a combustion chamber
and heat regenerator unit of checker bricks. The clean blast furnace gas is
burnt in the combustion chamber and the hot products of combustion later
heat up the checker bricks. In this case, the stove is said to be on on gas
and is maintained on gas until the checker bricks are heated to a certain
temperature. Firing is stopped and the cold blast is passed through the
checkers which impart the heat stored in them and a produce preheated
blast. The stove is said to be on blast. It can continue heating the blast till
a certain minimum temperature of the blast is obtained. The stove is again
put on gas and the cycle is repeated.
The stove design and the number of stoves employed should ensure
a steady supply of preheated blast to the furnace. This duty demands that
the amount of heat generated by way of combustion of gas per unit time
should be adequate to heat up the required amount of blast to the required
temperature per unit time, taking into account the usual efficiency of heat
transfer via the checker system and the usual heat losses from the system.
The thermal efficiency of the stove varies between 70–90%. The checker
work cools more rapidly, whereas it takes a longer time to heat it up. In
practice, a stove may be on gas for 2–4 hours and on blast for 1–2 hours. For
an uninterrupted steady supply of blast at specified temperature, a battery
of at least three stoves is necessary. A two stove system is quite unsatisfac-
tory, and a three or four stove system is preferred.
In a three stove system, one is on blast and the other two are on gas at
any time. In a four stove system, two are on gas and on blast at any time.
If any stove needs repair, it can be cut off from the line and repaired while
the others can continue normal working as a battery of three stoves only.
A four stove system is more flexible than the three stove system in such an
eventuality. Historically, a four stove system was fairly popular. The modern
trend is to manage with a three stove system of efficient design to last prac-
tically the whole furnace campaign without any major intermediate repairs.
The interrelationships of number of stoves employed, hot blast tempera-
ture desired, calorific value of fuel gas used for heating, amount of blast to
be heated, and gas to be burnt per unit time can be established.
Construction of the Stove and Its Accessories

The stove is a tall cylindrical steel shell with heights ranging between
20–36 m and a diameter around 6–8 m. It has a dome-shaped top. In the
commonly adopted design, the products of combustion rise upwards in the
combustion chamber, and then flow downwards through the checker only
once. This is called the two-pass design. There are other designs, in one of
which the checkers are divided into two chambers and the gases have to
pass once downwards and then upwards before leaving the stove. Such a
design is known as three-pass design.
(a) side-combustion type.

(b) central-combustion-type.
FIGURE 12.7 Hot blast stove design with respect to location of the combustion chamber.
Similarly, a four-pass design is also in use. The efficiency of the stove

depends upon the exit gas temperature, and in order to reduce this temper-
ature, to increase the efficiency, the multiple passes through the checkers
were designed. It is no longer necessary since, even with one pass through
the checkers, the exit gas temperature can be brought to a substantially low
value by using improved checker brickwork. That is why modern stoves are
invariably a two-pass design.
Stoves differ in yet another major feature (i.e. the location of the combus-
tion chamber) as shown in Figure 12.7. In one type, the combustion chamber
is located in the center and the checkers are all around it in an annular shape.
This is called the central-combustion-type. In the other, the combustion
chamber is on one side and oval in cross-section. This latter design is more
popular than the former and is known as the side-combustion-type.
(a) Vertical section.

(b) Horizontal section showing the details of column support for the checker-work.
(c) Horizontal section through the hot blast main and
(d) Section showing the wall lining and the checker-work area.
FIGURE 12.8 Modern hot blast stove:
The stove is a steel shell lined thoroughly inside with good insulation and
firebricks. The wall separating the combustion chamber from the checkers is
also of heavy duty firebrick construction. The dome is subject to maximum tem-
perature, and hence, needs to be properly insulated and lined with heavy duty
firebricks. The checkers are a refractory construction of special brick shapes.
The checker-work has to absorb maximum heat at a faster rate while heating
and should desorb heat equally rapidly to the incoming cold blast. The larger
the weight of the bricks, the larger their heat storing capacity will be. The larger
the surface area exposed as flues, the faster the heat exchange is with the gas.
The bricks should have maximum weight with maximum surface area of flues
(i.e. maximum openings to allow free passage of gases). It has been found that
a ratio of weight of bricks in kilograms to a heating surface in square meters of
about 5–6 in minimum is optimal. Below this, structural difficulties may arise.
The checker bricks are supported on steel grids which, in turn, are
supported by cast iron or steel columns as shown in Figure 12.8. Since
the maximum temperature during combustion is generated near the
dome, and since the top portion of checker bricks have to stand higher
temperatures, with progressively decreasing value downwards, the quality
of checker bricks used also varies accordingly. Heavy duty firebricks are
essential for dome construction. The top 3–6 m height of the checkers is
made up of higher alumina bricks or semi-silica bricks while the remainder
is made up of good quality firebricks. The checker bricks are invariably
of special shapes with many flues of several different shapes and sizes as
FIGURE 12.9 A few of the checker brick designs.

These shapes facilitate laying and handling. The brick lining including
the checkers are expected to last 15–20 years before replacement becomes
necessary because of spalling, deformation, clogging, etc. However, 3–6 m
of the top checker bricks, are usually replaced after each furnace campaign
(i.e. after 4–7 years).The general design of the stove is shown in Figure 12.8
and some more details are shown in Figures 12.10 and 12.11.
FIGURE 12.10 Some details of hot blast stove construction.
FIGURE 12.11 Some details of hot blast stove design.

Besides the checkers, the chamber stop valves to close and open the
cold blast, gas and hot blast mains, as desired, are also at convenient loca-
tions. An air stove also has several other attachments. A proper burner pre-
mixes air and gas and burns it in the combustion relief valve which is used to
release the extra pressure developed inside while changing the stove from
on-blast to on-gas. Since the hot blast temperature varies, within certain
limits, as the checkers cool down, and since it is desirable to supply the
hot blast at a fixed temperature to the bustle main, a mixer valve by-passes
a certain amount of cold blast into the heated blast and by its regulation
maintains a constant hot blast temperature. A pyrometric control circuit is
used to regulate the mixer valve openings.

In spite of all improvements in blast furnace processing, the effluent gas
from the furnace does contain carbon monoxide that was not utilized for
reduction purposes. A fairly efficient gas cleaning system has been devel-
oped and used in practice to obtain a clean fuel gas for reuse in the plant. A
part of the clean gas is used for generating the hot blast for the blast furnace
itself, and the remainder is used elsewhere in the plant as a lean fuel gas.
CHAPTER
13
BLAST FURNACE OPERATION
In This Chapter
● Blowing-in
● Banking
● Blowing-out
● Tapping
● Fanning
● Back Drafting
● Conclusions
13.1 Blowing-in
The process of starting a newly lined furnace is called blowing-in. In gen-
eral, the operation involves four main steps: drying, filling, lighting, and
operation until normal production is established. There is no standard
practice of blowing-in, and the details of each of the above mentioned
steps adopted depend on local conditions and customs.
Drying
The new lining of a furnace contains a significant proportion of moisture
which must be slowly and completely removed before the temperature of
the furnace is raised. This operation is known as drying, in which the furnace
is slowly heated. The stoves that are newly lined are dried by introducing a
lighted gas pipe in the combustion chamber, slowly increasing its intensity
in about 10–15 days to the level when firing under standard conditions can
begin. Repaired stoves can be brought to working temperatures in a day or
two. Blast furnace drying can be accomplished by any one of the three dif-
ferent techniques:
1. Supply of hot blast from the stove;

2. Use of a Dutch oven as an auxiliary furnace to generate and supply
hot gases;
3. Use of wood or coke fire on the hearth.
Any amount of time and trouble taken in ensuring careful and gradual dry-
ing of the furnace is more than repaid in the furnace’s subsequent operation.
Filling
At the end of drying, depending upon the method used, the furnace is
cleared off. The coolers are turned on, and once the inside temperature is
tolerable, furnace personnel can enter and prepare for filling the furnace.
The inspection of coolers at this stage is necessary, since the rectification
of faulty coolers is readily possible at this stage. In fact, a check list is pre-
pared and each item is checked off as reports are received of its satisfactory
performance. Simultaneously, the equipment for weighing, charge hoisting,
material disposal, etc. are also prepared to be put into action at the appro-
priate time. The tap holes are prepared, and the coolers, which were earlier
removed to have access inside, are packed in position.
Filling of the furnace usually means filling the hearth with light kindling
wood and shavings saturated with oil up to the tuyere level; old timber slip-
pers are laid over this scaffold. Coke is charged above the timber scaffold
from the top up to the bosh level. A quantity of limestone sufficient to flux
the ash in the charged coke, is also charged along with the coke after the
initial coke blanks. A small amount of old blast furnace slag is also incor-
porated with coke after the coke level rises beyond the mantle level. The
early slag volume is deliberately maintained at a high level to heat up the
hearth and prepare it to receive the iron. A light burden charge of ore,
stone, and coke (i.e. the ratio of iron ore to coke is low, about 0.5–0.6.) are
laid on the coke blanks. The burden weight that is iron ore is increased
every 8–10 charges, in increments of approximately 0.03–0.05 ratio of iron
ore to coke. The furnace is now ready to be lighted or ignited.
Blast Furnace Operation • 277
Lighting and Operating Until Routine Practice is Established

After filling the furnace for blowing-in, the bells are opened and the dust
catcher dump valve is closed. The air in the dust catcher is purged out
with a full head of steam. The furnace is lighted either by inserting red-
hot bars through the tuyeres or the slag hole and iron notch. Alternatively,
the gas torch may also be used. Generally, highly combustible material
is kept in front of the tuyeres during filling to light the furnace readily.
Burning is allowed with natural draught alone for the first 24–36 hours;
a light blast is put on thereafter. As soon as good amount of gas emerges
from the furnace top, the bells are closed and the dust catcher dump
valve is slowly opened to conduct exit gases through the gas cleaning
system. The blast volume is fairly rapidly increased to the normal volume
of blowing. Usually more than three quarters of the standard volume of
blast is blown by the end of fourth day, and the full blast is on by the end
of a week after the furnace is ignited.
Tap holes are kept open for hot gases to escape out during the early
period. Once coke burning and slag formation starts, furnace crews are vigi-
lant in observing the tap holes. The first indication of a sudden decrease
in the outgoing gas through the tap hole is an indication of the beginning
of slag accumulation in the hearth, and the tap hole is immediately closed
thereafter. Nearly six to eight hours may elapse after this, before sufficient
slag has accumulated to warrant flushing.
After nearly two days, as the ratio of iron ore increases in the burden,
the first cast may be due. The amount of slag and metal flowing out of the
furnace is correlated to the charge schedule and proportion, in order to
assess the progress of the blowing-in operation. After the first cast is over,
charging and tapping schedules are established and are strictly adhered to
until routine production is established. Even if changes in these procedures
become necessary, alternatives are kept ready for immediate execution
without any delay. The changes become necessary due to fluctuations in
temperature, composition of metal, and in the relative amounts of slag and
metal tapped.
The maximum blast temperature and volume rate are achieved rather
quickly to have a proper blowing-in.
It generally takes nearly a week to establish normal routine practice
so that the metal and slag of desired composition will be tapped out of the
furnace at the desired intervals.
13.2 Banking
After blowing-in, a blast furnace is expected to run continuously for
5–7 years. Some exceptions to this are due to infrequent shutdowns for
repair of the coolers and the like or a major shutdown of 4–5 days to
repair or replace the bells. A modern blast furnace, therefore, may pro-
duce 4–6 million tons of pig iron during its full campaign before it is
shut down for relining.
A furnace, however, at times may have to be shut down for an extended
period during the campaign for variety of reasons like: labor troubles, short-
ages of raw materials, serious breakdowns, etc. Since the furnace can not be
turned on or off readily, the furnace needs to be prepared for such a tem-
porary shutdown, so that when production is resumed, it can be done with
minimum effort. The temporary shutdown of a blast furnace is known as
banking, which is more akin to banking a fire. This means reduction in com-
bustion rate, achieved not by extinguishing the fire but by taking the blast
off, covering the fire with excess coke, luting up all air inlets and smooth-
ing the stock with fine material. The fire is thereby maintained inside for
future use. In general, a blank charge of coke is used to bank a furnace. The
amount of blank coke charge depends upon the length of the anticipated
shutdown. The blank coke charge is followed by light ore charges, 50–60%
of the normal burden. The longer the anticipated shutdown is, the lighter
the burden. A banking burden for a shutdown of an indefinite length of
time is very similar to blowing-in burden. As soon as the coke blank reaches
the top of the bosh the blast is taken off, all the metal and slag accumulated
in the hearth its tapped, leaving nothing to solidify and the stock is covered
with fine ore to prevent drafts. The tuyeres and notches are luted with clay
to prevent air infiltration. The bells are closed and the dust catcher dump
value is also closed to isolate the furnace from the gas cleaning system.
The stock level is inspected every day when the furnace is banked. A
slow downward movement of the stock indicates slow consumption of coke
by air that is infiltrating. A rapid drop in stock level can lead to a defi-
ciency of coke at the time of restarting the furnace. An excessive drop may
even necessitate spraying to seal the bosh from the inside. The furnace is
restarted by replacing the tuyeres and notches, and charging sufficient coke
to fill up the space created by the sinking of the stock. The rest of the prac-
tice normally resembles the blowing-in practice. It may not take more than
a week to bring the furnace operation back to normal.
Blast Furnace Operation • 279
13.3 Blowing-out
The process of stopping the furnace operation at the end of its campaign
(due to the lining being worn out) is known as blowing-out. It is accom-
plished in two ways:
1. Charging is stopped and the stock is allowed to descend until a minimum
of it remains inside. As the stock sinks, the blast is reduced and the top
of the stock is cooled by water sprays inserted through the top. To be-
gin with, the burden quality is altered to produce highly siliceous slag.
The furnace is isolated from the gas cleaning system midway through the
blow-out. The last cast is carried out as completely as is feasible. Towards
the end, the remaining stock is quenched with water, and the furnace be-
comes cold in about 24 hours so the crew can start the raking operation.
2. In the alternative method, the blowing-out starts with a blank coke charge
followed by a charge of clean silica gravels of +25 mm and −50 mm size.
The stockline is kept at the normal level in the beginning, but later on, it
is allowed to sink. Water is used sparingly to control the stockline temper-
ature. The blast volume is not reduced to the extent as done in the previ-
ous method. The leftover gravels are washed out with water in the end.
The time required for blow-out is very short, generally about 6–8 hours.
13.4 Tapping
The metal and slag are removed from the furnace periodically, when the
furnace is in operation. The time interval between two taps depends upon
the production rate and size of the furnace. On an average, 20–22% of the
ore weight charged in the furnace is in the form of oxygen that combines
with carbon to get eliminated as gas. Similarly, 70% or so of the coke is
gasified, and almost 50% of the flux (as stone) is gasified during calcination.
This leaves approximately 78% of ore, 25–30% of coke (as ash) and 50% of
stone to finally come out as condensed phases (i.e. metal and slag). From
the usual amounts of charge materials put in per ton of pig iron produced,
nearly 300–700 kg of slag will also be produced. The density of the slag is
one third metal, and the volume of the slag is three times the metal of the
same weight. Hence, the slag volume will be nearly twice that of the metal
in a commercial blast furnace. This would necessitate removal of the slag at
least twice as frequently as the metal. In general, the metal is tapped every
4–6 hours and the slag is removed every 2 hours or so.
The tapping of the slag and metal needs prior cast-house preparation,
subsequent cleaning of the cast-house, and closing and opening of the tap
holes. It is a fairly laborious job performed in trying conditions. The ladles
must be ordered in advance and kept waiting to receive the slag and the
metal. Since the furnace is under pressure, the metal and the slag flow out
of the tap holes with force on their opening.
13.5 Fanning
If the full production capacity of the furnace is not required for any reason,
it is decreased by reducing the blast input. If the blast is reduced to a value
less than 25% of the normal volume, the technique is known is fanning.
This is far better than totally shutting off the blast, because normal produc-
tion can thereby be resumed without any difficulty.
13.6 Back Drafting

Whenever the furnace is required to be put off the blast for a short duration
of an hour or two, to repair the tuyeres or the skips, or the coolers, etc. it
is back-drafted. The blast is put off and the bustle pipe is put under nega-
tive pressure to force the furnace gases to flow in a reverse direction into
the stove where it is burned. Air may be drawn downwards in through the
bleeders to achieve this. A special back-draft chimney may be used instead
of allowing the gases to go through the stoves.
13.7 Conclusions
It is neither easy nor simple to put the blast furnace in operation, and tem-
porarily or permanently, put it out of operation. The complexity of the blast
furnace operation can be observed from this single fact.
CHAPTER
14
IRREGULARITIES IN BLAST
FURNACE OPERATION
AND THEIR REMEDIES
In This Chapter
● Hanging
● Scaffolding
● Slip
● Chilled Hearth
● Pillaring
● Breakout
● Choking of Gas Off take
● Flooding and Coke Ejection Through the Tap Hole
● Leaking Tuyeres, Tap Holes, and Coolers
● Channeling
● Salamander Formation
I
n spite of the developments in the understanding of the blast furnace
process and construction of better designed and controlled furnaces,
the blast furnace operation in practice is still challenging. The furnace
is still capable of acting in unpredictable ways, and hence quite aptly it
has been always referred to by the pronoun she. The irregularities in blast
furnace operation may arise due to faulty mechanical devices like coolers,
valves, etc. or due to faulty mechanical operations like tapping, charging,

etc. or due to abnormal processes occurring inside the furnace with respect
to physico-chemical changes in the charge. Proper mechanical and elec-
trical maintenance of all the devices, their normal change-overs, replace-
ments, etc. can minimize these problems to a large extent. Innovation in
design and better construction materials for these devices can also prove
to be equally useful. Similarly, possible faulty mechanical operations can
be anticipated by experienced operators with forethought and good judg-
ment, and they be minimized with prompt action. The irregularities related
to processes inside the furnace are often due to variations in the operating
variables of the process, and these do vary in spite of the control, that is
exercised.
In general, immediate detection and correction of any irregularity is
essential to avoid serious trouble and loss of output. It should be noted that
the burden takes 6–10 hours to travel through the furnace. If any corrective
action is required, the action is likely to bear the result only after several
hours. It is, therefore, essential to anticipate the trouble and initiate the
corrective action without delay. Some of the common irregularities in blast
furnace operation are discussed below.
14.1 Hanging
If uniform descent of the burden in the furnace is interrupted either by
wedging or bridging of the stock or by scaffolding, it is known as hanging.
Hanging can be caused by five different reasons:
1. Solidification of previously fused slag, particularly in sinter, into a large
impervious mass.
2. Bridging of ore particles in the vicinity of fine coke particles, which
instead of separating the ore particles, flow in the interstitial positions
of the ore particles.
3. The deposition of large amounts of carbon due to Naumann reversion
(as in Eq. (10.2)) in the voids and consequent decrease in permeability
of the burden.
4. Condensation of alkali vapors in the upper part of the stack, thereby
cementing the charge into a impervious mass.
5. Excessive blast pressure resulting in counteracting the downward move-
ment of the stock.
Irregularities in Blast Furnace Operation and their Remedies • 283
Insufficient batter in the stack walls may accentuate hanging. Hanging has
been classified broadly in three different types: top hanging, bottom hang-
ing, and general hanging. Top hanging occurs in the stack and mainly arises
due to the carbon deposition reaction. Alkali vapor condensation can also
be responsible for this type of hanging.
Bottom hanging occurs anywhere below the tuyeres due to low voidage
in the stack. General hanging occurs anywhere above the tuyeres due to low
voidage in the stack.
Remedies
1. Add limestone of large size which on dissociation generates CO2, which
in turn forces the solution loss reaction to take place and improves the
permeability of the bed.
2. Hanging furnace responds better to a colder blast.
3. Reduction in blast pressure distributes the gas flow, though slowly, but
more uniformly.
4. A persistent hanging can be cured by blowing down the furnace to
bosh level and filling it with coke blank.
14.2 Scaffolding
A scaffold is a large mass of material that gets stuck to the furnace wall, usu-
ally as a single block, particularly in the top portion of the bosh. It results
in reduction in the cross section of the furnace and all the attendant effects
like uneven movement of the stock, rise in blast pressure, increase in dust
losses, decrease in output, increase in fuel consumption, decrease in the
amount of gas produced, etc.
Formation of a scaffold is mainly attributed to the presence of alkali
oxides and not chlorides or other compounds of alkali metals in the bur-
den. The alkali vapors condense on the brick lining and form low melting
alkali-alumina-silicates, to which the ore particles subsequently stick, pro-
gressively resulting in the formation of a massive block of charge materials
stuck to the walls. The formation of a scaffold is influenced by factors like
improper fluxing, heavy-burden, bad furnace design, etc.
Scaffolds are also believed to be formed due to a sudden lowering of
the fusion zone adjacent to the walls; this mass progressively builds up to
ultimately form a scaffold. Formation of a scaffold is a progressive process
of adhesion of a mass of material at a definite zone in the furnace. A scaf-

fold generally has a interstratified, laminated texture with thin sheets of
decarburized iron and associated alkalis. The concentration of alkali oxides
occurs at the scaffold-brickwork interface.
Scaffold formation may therefore be minimized by reducing the follow-
ing: alkali content of the burden, refractoriness of the slag, erratic working
of furnace, decrease in fuel rate, and shut-offs of blast for any reason.
The trouble can be corrected by charging a few blanks of coke and
reducing the blast pressure on the affected side; the burden may be made
more siliceous on that side. Scaffolds have been successfully removed by
jumping the furnace, that is, the blast is put off suddenly to relieve pressure
under the scaffold. Rigid scaffolds have been ultimately removed by the use
of dynamite when every other remedy failed.
14.3 Slip
Slipping is defined as the sudden sinking of the stock caused by a collapse
of the hanging, wedging, scaffold, etc. In short, it is the aftermath of these
irregularities. In severe cases, slipping causes a chilling of the hearth or
even explosion. The best remedy is to allow the furnace to slip on its own,
by adjustment of the blast temperature and pressure. A poor bosh design
may cause slipping, and the bosh behavior of burden materials needs to be
assessed before finalizing the bosh design.
14.4 Chilled Hearth

This is a very serious irregularity since it affects tapping adversely. It may
result because of low fuel input, excessive moisture in the blast, water leaking
in from the tuyeres, etc. If it is due to these reasons, chilling is gradual and can
be rectified before it becomes serious. Slipping also can cause chilling, but it is
sudden. Depending upon the extent of chilling caused by slipping, gas torches
are used to open the tap holes. The metal may even be tapped through the
cinder notch, if required and if possible, under such circumstances.
14.5 Pillaring
If the blast is unable to penetrate right up to the center of the furnace, it may
lead to the formation of a cold, central column of the stock with an annu-
lar hot zone all around it. This is known as pillaring. A bar inserted right
through the furnace through the tuyere hole would show a red-hot portion
at both ends and a cold middle portion, if pillaring exists in the furnace. The
extent of a cold middle portion of this bar indicates the extent of pillaring
in the furnace. Pillaring can be eliminated by increasing the blast pressure
which will penetrate the cold, central column and heat up the pillar.
14.6 Breakout
Generally, breakouts are caused by the failure of the bosh walls, the walls of
the hearth, or the hearth bottom. As a result, the liquid iron or slag or both
flow out of the furnace in an uncontrolled fashion. This can be disastrous
and may even cause an explosion if it breaks coolers. Although a break-
out can occur anywhere, they are serious if they occur in the hearth. Slag
breakouts are not as serious as metal breakouts, since slag breakouts can
be readily repaired. Wherever a breakout occurs, it is essential, to remove
maximum amount of liquid products from the furnace through the tap
holes. The furnace is also taken off the blast. The breakout area is fully
repaired, and the furnace operation is resumed only thereafter.
A properly designed hearth, with a minimum of refractory joints and
absence of continuous straight joints up to the shell, can minimize break-
outs. The hearth refractory also needs to be of proper grade. The earlier
firebrick hearths were more prone to breakouts because of the leakage
of atmospheric air in the hearth. Its consequent combination with iron to
form iron oxide and fluxing action of the fire-back, finally resulted in the
refractory giving way. The problem, to a large extent, has been satisfactorily
solved by adopting a carbon hearth.
14.7 Choking of Gas Off take

Furnace operation has to be suspended if dust accumulates in the uptakes
and downcomer, and it can be resumed only after clean-up. This happens
because of a faulty gas offtake design, particularly the inadequate cross sec-
tion and improper joints.
14.8 Flooding and Coke Ejection Through the Tap Hole

In the bosh, liquid slag and metal trickle through the permeable coke bed
against the upward thrust of the ascending gases. An increase in gas or
liquid flow can prevent the liquid metal and slag from flowing downwards,
causing it to accumulate in the coke interstices until the weight of the liquid
overcomes the upward thrust of the gases and descends suddenly into the
hearth. This phenomenon is known as flooding, which can be minimized
by having a high voidage (i.e. by using a higher mean size of the coke).
Better quality coke is also beneficial since the degradation inside the fur-
nace would be reduced. Consequently, the permeability in the bosh region
would be improved.
Anything that imposes a constraint on the tuyere raceway volume causes
a hold up and a subsequent tendency to flooding. The interruption in uni-
form blowing rate causes the raceway to collapse, and when it is resumed
again the small particles of coke can not re-enter the raceway. Consequently,
they descend into the hearth, instead of burning in the tuyere area, result-
ing in a choked hearth thereby causing the well-known phenomenon of
coke ejection from the slag and the iron notches during tapping. This often
leads to an unjustified criticism of coke quality. Uniform blowing of the
furnace is the best remedy to avoid flooding.
14.9 Leaking Tuyeres, Tap Holes, and Coolers

In spite of proper design, the water cooled parts of the furnace may give
way and these should be immediately replaced or repaired. The monkey
is the troublesome part and often needs frequent replacements. If it is not
possible to repair or replace the faulty cooler, it must be cut off from the
water mains and put out of use.
Leaking tuyeres or coolers in the lower part of the furnace can have
disastrous effects if these are not rectified in time. The leaking tap hole
coolers lead to steam generation, which on coming in contact with the car-
bon hearth, erodes the hearth lining and the campaign has to be stopped for
capital repairs. The maintenance instructions for tap holes must be scrupu-
lously followed to minimize these troubles.
14.10 Channeling
Preferential flow of ascending gases through certain areas of the burden,
because of their relative permeability, is known as channeling, since these
passages appear as channels. This arises due improper distribution and a
wide range of the charge in the furnace. Use of a more uniformly sized
burden and proper distribution can minimize this. Channeling otherwise
reduces the effective cross section of the furnace for gas-solid interaction,
and thereby decreases the productivity directly.
14.11 Salamander Formation

Presently, the BF operates for 5–10 years before it is due for relining. During
this period, the molten metal slowly erodes the hearth in general and the
bottom in particular. This results in lowering the level of the hearth-bottom
below the tap hole. As a result, the amount of metal accumulated never gets
a chance to drain out. When the furnace is blown out for major repairs, this
poses a major problem of draining out the accumulated hot metal. This is
called salamander formation. In order to understand the salamander forma-
tion, several thermocouples are embedded in the hearth bottom and walls,
and the temperatures are recorded to decipher the size, shape, area of loca-
tion, and the rate of its formation. This helps to determine its size, shape,
and location. Then, blast furnace operators can develop ways and means to
tackle this problem while blowing-out the furnace for capital repairs.

By and large, proper burden distribution allows gas to pass through the
central, middle, and the peripherals zones as per the requirements of effi-
cient operation. The alkali control in the iron bearing charge materials is
another component that ensures minimum wall accretions; this is the prime
approach of avoiding irregularities in a BF operation. Proper burden distri-
bution coupled with controlling the other parameters is the only guarantee
of a more regular BF operation and extended campaign. But in spite of
good process control, irregularities do occur in a BF operation, and these
have to be remedied as per the exigencies of the situation. The better the
quality of the charge materials and the more effective the process control,
the better the regular BF operation will be.
BLAST FURNACE
CHAPTER
15
PRODUCTS II –
SLAG AND METAL
In This Chapter
● Introduction
● Composition of Pig Iron
● Slag Composition
● Production of Ferro-Manganese
● The Sulfur Problem
● Basic Burdening of Blast Furnace and External Desiliconization
● Acid Burdening of Blast Furnace and External Desulfurization
● Disposal of Metal
15.1 Introduction
Blast furnace produces three different products viz. the gas, the metal, and
the slag. The quality and the amount of off-gas produced by the blast fur-
nace and the extent of cleaning required to make it as a useful fuel and,
where all it is utilized in the plant has already been described in Chapter 12.
The present chapter deals with the quantity and quality of slag and metal
generated and their utilization. Emphasis has been laid on the operation of
the blast vis-a-vis their quality and quantity produced.
A blast furnace is used to produce molten iron to finally convert it into
steel. Sometimes it is cast and sold as pigs (solid pig iron). In the latter case
the iron that is produced is usually foundry-grade iron. The amount of iron so
produced is usually from small-capacity blast furnaces (MBF) meant to pro-

duce merchant iron (i.e., pigs). The major amount of iron produced from blast
furnaces is invariably meant for its immediate conversion to steel without
allowing it to solidify. Such furnaces are a part of an integrated iron and steel
plant wherein iron ore is the main input material, ironmaking is just the first
step, and the molten iron is converted to steel, which is cast and rolled into
flat or long semi-finished marketable steel products. This iron and steelmak-
ing and production of such steel products in an integrated steel plant are an
integrated activity. The way all these processes are carried out and the quality
of product at every step so produced has to be acceptable as a raw material
for the next processing. The economics of every step are viewed in the form of
integrated economics of production of finished steel products from iron ore.
The product specifications are therefore adjusted as per the require-
ments of the next step of processing so that the overall plant operation is at
its best economy and productivity, of desired steel products, from the avail-
able ore. The chemistry of all such intermediate products, in particular that
of iron, is very critical for its efficient conversion to steel. In other words,
the specifications of products, at each stage of production, are laid down
and correct processing steps are chosen to ultimately produce economically
the required steel products from the given iron.
The quality of slag produced depends mainly on the burden chemistry and
temperature at which the furnace is operated. Since the temperature in a BF
varies from top to bottom, a question might arise as to the temperature at what
zones then considered as the operating temperature of the furnace. Since the
interest is in the molten products, more specifically the metal, whose chemistry
is ultimately decided by the hearth temperature, it is considered to be the oper-
ating temperature of a furnace, unless anything contrary is specifically stated.
Slag is an unavoidable by-product along with the useful metal. The chem-
istry and the volume of slag thus produced depends upon the burden chem-
istry, in particular the gangue constituents and their actual proportion in the
ore and flux, the chemistry, and the amount of coke ash. These dictate the
type and amount of flux to be added to finally obtain thin and free-flowing
slag at the working hearth temperature. The more the gangue contents of the
ore/sinter and the more the ash of coke, the higher will be the slag volume
indicated, which is often expressed as kg/thm.
The best practices and high productivity blast furnaces produce less
than 300 kg of slag per thm produced. It can be as bad as 700–800 kg/thm
because of very poor quality of burden materials. Major practices produce
Blast Furnace Products II – Slag and Metal • 291
400–500 kg of slag/thm produced. The old Indian practices using inferior

quality of Indian cokes, slag volume was approximately 500–600 kg/thm.
With the use of imported low-ash, good-quality coal/coke, the slag volume
has come down to nearly 400–450 kg/thm, depending upon the proportion
of imported coal/coke incorporated in the charge.
15.2 Composition of Pig Iron

The metallic product from a blast furnace is variously known as pig iron,
molten metal, et cetera. It contains carbon, silicon, phosphorus, manga-
nese, and sulfur as the chief impurities totaling anything up to 8% or so by
weight. Since pig iron trickles through coke lumps in the bosh region, it
gets saturated with carbon, which may vary around 3.5–4.2%. Whatever is
the phosphorus load, practically the whole of it in the burden is reduced,
and it finally joins the metal. There is, therefore, practically no control on
the contents of carbon and phosphorus in the blast furnace practice per
se. Although the presence of carbon in pig iron is beneficial during its con-
version to steel, the presence of phosphorus is positively detrimental and
it adds to the cost of steelmaking. It should therefore be the minimum
amount in the hot metal. It means that it should be the minimum in the
burden itself, to whatever extent it is otherwise feasible.
The contents of silicon and sulfur depend upon the volume of slag, its
basicity and the operating temperature of the furnace. These can therefore be
varied within certain limits by adjusting the burden chemistry, fuel rate, and
those conditions that affect the temperature of the blast furnace. Similarly,
the manganese content of iron is a function of manganese oxide content of
the burden, the slag basicity, and the operating temperature of the furnace.
Because of these compulsions, the molten iron from a blast furnace
is quite impure. The same impurities are subsequently eliminated during
steelmaking, as inevitable stages, in the production of steel from iron ore.
The composition of pig iron varies as follows:
Carbon ... 3.5%–4.2%

Silicon ... 0.5%–2.5%
Manganese ... 0.1%–1.5%
Phosphorus ... 0.04%–2.0%
Sulfur ... 0.04%–0.15%
If the phosphorus and sulfur contents are each below 0.04%, the pig iron
is known as Swedish or hematite or acid iron. The acid process of steelmaking
is adopted for its conversion to steel. If the phosphorus content is more than
0.04%, it is known as basic pig iron since the basic process of steelmaking has
to be adopted for its conversion to steel. In the basic variety, if the phospho-
rus content is more than 0.5%, special basic steelmaking processes have to
be adopted for their conversion to steel. The iron containing more than 0.5%
phosphorus is traditionally known as Thomas iron or basic Bessemer iron.
By and large, much of the world’s pig iron contains 0.15 to 0.35% phos-
phorus. Thomas iron is produced particularly on the continent in countries
including France, Belgium, Luxemburg, and Germany, and also in South
Africa and some other places. The production of Swedish iron is rather
exceptional in that it is produced only at a few places in the UK and Sweden.
On the whole, the composition of pig iron is subject to a variety of condi-
tions under which the furnace is operated and the type of charge materials
used. The temperature of the hearth also varies considerably from plant to
plant. It is as low as 1280° C to 1350° C in India to as high as 1500° C to
1550° C in Japan. The hot metal composition as obtained in different coun-
tries is shown in Table 15.1.
Table 15.1 Typical hot metal compositors produced for steelmaking in

different countries. The variation is due to the variation in raw material
qualities and individual practices
Japan Europe USA Sweden India

Carbon everywhere saturated with carbon in the range of 4.0%–4.5%
Silicon 0.30–0.45 0.30–0.45 0.40–0.70 0.5–2.5 0.5–1.5
Manganese 0.20–0.40 0.20–0.7 0.50–0.75 0.50–1.0 0.04–0.10

Phosphorus 0.09–0.10 0.06–1.5 0.04–0.08 0.04–0.05 1.10–0.35
Sulfur 0.02–0.025 0.02–0.04 0.025–0.065 0.02–0.03 0.05–0.08
The presence of silicon in hot metal is useful in steelmaking, as it acts

as a heat producer. Any excessive amount, however, leads to increased flux
consumption and to consequently a large slag bulk and all related problems,
which add to the cost of steelmaking. Silicon content of 0.5% is considered
ideal for a basic process of steelmaking.
The sulfur content specified for hot metal varies considerably, and
the actual value depends upon the steelmaking process adopted, the steel
product quality required, and so on. In general sulfur content of hot metal
should not be more than 0.06% for basic oxygen steelmaking processes
or preferably below 0.04% to meet the end product quality requirements
readily. Since the silicon and the sulfur contents of pig iron vary inversely
with temperature, suitable optimum values of silicon and sulfur contents
are arrived at by adjusting the ironmaking practice. Temperature adjust-
ment is effective in this respect.
From the point of view of designing the steelmaking technology that is
to be subsequently adopted, the hot metal compositions are broadly classi-
fied as basic, Thomas, and Swedish types, as shown in Table 15.2.
Table 15.2 Typical hot metal composition of different steelmaking grades

Impurity % Swedish Iron Basic Iron Thomas Iron Indian Iron
(Typical)
Carbon 3.5–4.2 3.5–4.2 3.5–4.2 3.5–4.2
Manganese 1.0 0.7–1.0 0.7–1.0 0.7–1.2
Silicon 2.5 0.5–1.0 0.5–1.0 1.2–1.8

Phosphorus 0.04 0.15–0.35 0.5–2.5 0.25–0.35
Sulfur 0.04 0.04–0.06 0.04–0.06 0.04–0.06
The above-mentioned compositions are produced to suit steelmaking

requirements. This is not suitable for foundry purposes. Foundry iron is
produced with nearly 2.5%–3.5% silicon and 1%–2% manganese to have
reasonable fluidity. The silicon content of this order can be produced by
raising the operating temperature of the blast furnace so as to expedite
silica reduction and obtain a high silicon iron.
Much of the iron in an integrated steel plant is produced for its subse-
quent conversion to steel. A small amount is, however, produced as foundry
grade iron, to be sold as foundry pigs. Another small fraction of the molten
iron is required for home foundry, particularly for the production of ingot
molds in an integrated steel plant.
15.3 Slag Composition

The BF slag cannot be of any specific chemistry. Its composition depends on
the chemistry of the burden, coke rate, fuel injection, and temperature of fur-
nace operation. The charge contains alumina, silica, magnesia, and lime as the
main constituents. The minor oxides may be Cr2O3, TiO2, FeO, MnO, etc.
Out of these, nearly 15% of silica gets reduced, and Si joins the metal
phase; balance 85% joins the slag. Temperature has a decisive role in decid-
ing this proportion. Nearly 60% of silica in slag comes from iron-bearing
materials, and the remaining 40% comes from coke ash.
These days, the MnO content of slag is negligible because Mn-bearing
ore/slag is not added in the charge.
Practically all the phosphorus in the charge, except that blown out as
dust, is reduced and joins the metal phase. Nearly 60%–70% of P in the metal
comes from iron-bearing charge materials, and the remaining 30%–40%
comes from coke.
Nearly 90% of input manganese find its way into the metal, only 10%
may be present in the slag.
Anything up to 90% of input sulfur can be made to join slag. But these
days, for high productivity, the practice is such that only 40% may join the
slag, and balance may join the metal.
Slag Volume
The exact slag volume produced is a function of the following variables:
1. Actual gangue amount in iron bearing charge

2. Actual ash of coke and chemistry of ash
3. Ash contributed by fuel injection through tuyeres
4. Basicity of slag aimed at
5. Amount of flux added and its chemistry
6. Amount of these blown out as dust in the off-gas
How to Calculate the Slag Volume?

Given:
(a) Ore contains 2.0% Al2O3, 2.5% SiO2, and balance 95.5% Fe2O3
(b) 1.495 t of ore is required to produce 1 t of metal
(c) Coke rate is 500 kg/thm containing 20% ash and which contains 40%
Al2O3 and 60% SiO2
(d) Basicity is to be maintained at 1.0
(e) Limestone is assumed to be 90% CaCO3 and 10% MgCO3
Let us see what is the slag volume.

For one metric ton of metal produced the amount in kgs of
SiO2 in the burden → 37.5(ore) +60(coke ash) = 97.5 kg
Al2O3 in burden → 30(ore) +40(coke ash) = 70 kg
CaO added as flux → 97.5 kg for basicity one

Limestone required is 174.1 kg to give finally basicity of 1.0.
This much limestone will give 17.41 kg of MgCO3 (i.e., 7.43 kg of MgO)
The slag will contain CaO, MgO, Al2O3, and SiO2 as assumed situation.
Hence the slag weight will be the total weights of these oxides.
And it is: 97.5 + 70 + 97.5 + 7.43 = 272.43 kg
Hence the slag rate is 272.43 kg in this case. In actual practice the ore,
coke, and limestone shall contain several minor oxides that will end up in
the slag. The slag volume therefore will be more than this amount.
Disposal of Slag
Earlier when slag was considered as typically a waste product inevitably
produced to obtain useful molten pig iron, it was transported from the BF
to slag dumping yard in slag ladles. The slag was tipped on the heaps in
molten condition and was allowed to be air cooled. This air-cooled slag was
used to make mineral wool, glass sand, and ceramic ware, or a proper size
fraction was used as filler material on railways.
Blast furnace slag essentially contains CaO, SiO2, MgO, and Al2O3 as the
major constituents and that on solidification it contains mainly silicates and
aluminates of calcium and to some extent of magnesium, Portland cement
is produced by semi-fusing CaO (obtained from limestone), Al2O3, and SiO2
to produce non-crystalline silicates and aluminates of calcium. The hydra-
tion of these silicates and aluminates develops the strength for which it is
used as a useful jointing construction material. BF slag as such does not
develop that good strength after hydraulic bonding because of lack of ade-
quate lime.
Although, qualitatively, blast furnace slag does contain the same silicates
and aluminates, the actual chemistry is not the same as that of Portland
cement. The reason is that the CaO content is much less in BF slag than in
Portland cement. Hence higher order aluminates and silicates do not form
in BF slag, as they form in a cementmaking kiln. The BF slag, however, can

be used as a cheaper substitute to the usual raw materials for the manu-
facture of Portland cement, replacing part of the virgin raw materials. This
replacement to the tune of 50% has now been achieved. Hence the value
for BF slag.
This is now a universal practice to use BF slag as a raw material for the
manufacture of Portland cement. For this the slag needs to be granulated.
BF slag as it flows out from the furnace, through channels in the cast house,
is allowed to fall in stirred water held in a steel ladle, whereby it gets granu-
lated. In this form it is easy to transport it to cement manufacturing sites
and finally to the cement making kiln.
This has already been adopted as a universal BF practice. This has obvi-
ated the need to dump it within the plant itself and a considerably costly
area of the plant for its dumping.
The pollution requirements ideally aim at not producing any waste
product in any manufacturing process. Use of blast furnace slag in this way
has contributed in a big way towards this end. The extent to which the pol-
lution problem arising out of this slag has thus been solved can be guessed
from the relevant data. Usually blast furnace practice produces 300–350 kg
of slag/thm. It means that for every million ton of hot metal produced, 3–3.5
lac metric tons of slag is generated and was stored as dumps inside the plant
until recently.
In the ultimate production of steel, nearly 60% of hot metal is con-
sumed. Therefore, a 5-Mt steel plant may produce 1 Mt of BF slag. This is
an enormous quantity, earlier stored within the plant itself, but now used
as a by-product. In this way now all the BF slag is disposed of from the
ironmaking plant.
Expanded or lightweight slag is a foamed slag product produced by
quenching it in a very controlled amount of water and air/steam. The prod-
uct should be fairly dry and lumpy, which is useful for making tiles, bricks,
building blocks, etc. The amount of slag that can be consumed by this route
is far too small as compared to its generation.
The steelmaking slag, however, is still dumped within the plant itself
in liquid or solidified condition. Generally, the volume of steelmaking slag
produced is 15% of the steel made. For a 5-Mt plant, nearly 7.5 lac t of
steelmaking slag is produced and stored as dumps inside the plant. This
causes many problems of storage space and pollution. It is also necessary
to evolve suitable process for its utilization. The high iron oxide content of
steelmaking slag is the main culprit. If it is possible to win iron from this
slag, it could also be used for cement making. In fact, this slag would then
be far superior to BF slag because of its high basicity.
15.4 Production of Ferro-Manganese

Until the advent of electro-thermal smelting soon after the World War II,
ferro-manganese was regularly produced in iron blast furnaces. The man-
ganese content of the product used to be lower than what is now obtained
by electro-thermal smelting. The use of oxygen-enriched blast minimized
many of the difficulties by raising the temperature of the furnace. In spite
of all these innovations, blast furnace cannot compete anymore with the
electro-thermal smelting processes for the production of ferro-manganese.
Ferro-manganese is therefore almost universally produced by submerged
electric arc furnaces.
Towards the end of the normal campaign of an iron blast furnace, how-
ever, it is switched over to ferro-manganese production at some plants. A
substantial proportion of Fe-Mn produced in countries including the USA,
Russia, France, and Germany comes from iron blast furnaces in the above
manner[54]. The usual high-carbon grades of Fe-Mn are thus produced.
15.5 The Sulfur Problem

A simple multiple regression analysis of practical data, particularly from
BOFs, have indicated that the process of steelmaking finds it difficult to
control effectively the sulfur specification of steel products, and the guar-
antee of achieving the steel specification can best be realized by reduced
sulfur input in the process via the hot metal. The problem is expected to
aggravate further in the future because of the general trend of laying down
more and more stringent sulfur specifications for more and more variety of
steel products. The significance of hot metal sulfur analysis will be increas-
ingly seen to determine the overall success in meeting the required quality
standards of final products.
It has been observed that the widely adopted faster BOF processes
of steelmaking are capable of removing only up to 20%–30% of the sulfur
present in the hot metal charge. It means that in order to meet the gener-
ally specified 0.025%–0.040% S in the final product, that hot metal charge
should not contain more than 0.04% S. It should preferably be approxi-

mately 0.02% S to accommodate erratic analysis of the blast furnace from
tap to tap. This is possible in a blast furnace practice using coke contain-
ing approximately 0.6% S maximum. If the sulfur in coke is approximately
1%, as is now commonly the case, it will be difficult to meet the hot metal
and subsequently the final steel specifications with respect to sulfur con-
tent. The sulfur problem is more acute when continuous casting of steel is
adopted. The finished molten steel should have no more than 0.025% S for
continuous casting to operate smoothly.
As with coke, the auxiliary fuels used for injection in the blast furnace
should also be low in sulfur. The cost of fuel oil is in inverse proportion to
its sulfur content and hence oils with approximately 2% S may have to be
tolerated for injection. In other words, with respect to sulfur content, sub-
stitution of coke by oil has practically no advantage.
For economical reasons, therefore, every effort must be made to eliminate
sulfur and produce the required quantity of hot metal before putting it in
the steelmaking furnace. The possible areas of desulfurization can be sum-
marized as follows:
1. Decrease of sulfur load to the blast furnace: Although it may be ideal,

it is impractical to put it into practice anymore.
2. Reduction of coke rate (coke plus fuel oil rate): It can be done by
employing rich ores, producing less slag volume, etc. The Steel Co. of
Canada operated a blast furnace with rich ores (98% Fe2O3) as charge
producing only 150 kg/t of slag without any difficulty. The average
hot metal sulfur was 0.021%. This cannot be universally applied since
poorer ores have to be used.
3. Desulfurization of coal or coke. It is yet to be developed to the stage of
its commercial exploitation.
Sulfur control in the blast furnace:
Desulfurization inside the blast furnace can be effected in three
different ways:
(i) By increasing the slag volume
(ii) By having higher basicity
(iii) By raising the hearth temperature
The more is the slag volume, the higher is the basicity and the higher
the hearth temperature, the better will be the desulfurization of metal
inside the blast furnace. All three factors in general require additional
heat, which demands an increased coke rate, which in effect adds to the
sulfur load in the furnace. It has been estimated that an average increase
of 10 kg of slag volume requires an increase in the coke rate by approxi-
mately 1.6 kg.
In essence, the forgoing indicates that it will be difficult for the blast
furnace, under the prevailing circumstances of operation and conditions of
raw materials, to meet the requirement of hot metal sulfur specifications
without impairing the production and economy.
The alternative is therefore to interpose one more processing step
between the ironmaking and steelmaking to desulfurize the hot metal to
achieve the required specifications and thereby meet the steelmakers’
requirements of hot metal. This is known as external desulfurization.
Any additional operation such as these adds to the cost of steelmaking.
The external treatment, however, must be viewed in terms of the overall
economy of iron and steelmaking production. Wahlster and other[59] have
shown that, on the whole, it is economical to adopt external desulfurization
along with certain other necessary changes in the blast furnace practice.
Every effort must be made to obtain sulfur and silicon contents in the
pig iron, as desired by the steelmaker, in the blast furnace itself. If this is
not possible, for whatever reasons, either of the following two alternatives
may be adopted.
15.6 Basic Burdening of Blast Furnace and External Desiliconization

If the alumina content of the burden is high, the furnace hearth tempera-
ture has to be on the high side to keep the slag thin and free flowing,
since alumina raises the softening temperature of the CaO-SiO2-Al2O3-
type slags. The basicity has also to be raised to obtain proper slag in this
case. These inevitable conditions are more conducive to effective desulfur-
ization of metal inside the furnace. This can be fully exploited in obtain-
ing pig iron containing less than 0.04% or preferably 0.02% sulfur, by
maintaining the highest possible basicity at optimum coke rate, provided
that the load of sulfur in the burden is within limits. This often leads to
high silicon content of pig iron, since silica reduction is favored at high
temperatures. The excess silicon necessitates adoption of the steelmaking

processes duplex in principle or else the pig iron is treated externally to
remove silicon and produce hot metal of the specification required
for steelmaking. This external treatment for elimination of silicon from
molten iron is known as external desiliconization.
The blast furnace operator in this case aims to produce low sulfur iron,
leaving silicon content to be worried about by the steelmaker.
This treatment is adopted at the Durgapur Steel Plant in India and many
a plants in the world. The efficacy of this depends on the total sulfur load in
the blast furnace burden. It worked well until recently. The sulfur load in
the furnace is steadily increasing, and the sulfur specification is progressively
deceasing all over the world. Basic burdening of a blast furnace is therefore
gradually losing ground, and indications are that external desulfurization of
pig iron will increasingly be adopted all over the world.
15.7 Acid Burdening of Blast Furnace and External Desulfurization

Once it is assessed that it is not possible to achieve the required desulfuriza-
tion of metal inside the blast furnace to meet the hot metal sulfur specifica-
tions of the steelmaker, by manipulating the blast furnace burden or the
practice, economically, external desulfurization of the hot metal outside the
blast furnace is resorted to in order to achieve the desired end. The blast
furnace is then no longer operated to achieve desulfurization as such inside.
Since one more additional step is to be adopted, the blast furnace operation
is adjusted to obtain maximum productivity and minimum of coke rate by
reducing the hearth temperature. The slag basicity is also reduced, since
high basicity is no longer necessary. The furnace is often operated with
basicity less than 1.0 (i.e., the slag produced is acid and for which the bur-
den chemistry is adjusted with more acid and less of basic oxides. Hence
the name acid burdening has come into vogue. The net result is that the
silicon content of the metal is also low, which is advantageous for the subse-
quent basic process of steelmaking. This is being practiced almost all over
the world now.
The acid burdening of blast furnace and external desulfurization
can be very readily practiced if the total gangue in the ore and coke are
more siliceous in nature. If the gangue is more aluminous in nature, the
hearth temperature required for free running of slag is on the high side.
Here also the slag composition is adjusted to obtain free-flowing slag at
the lowest possible temperature. The basicity may be little more than
1.0. Desulfurization is later on achieved externally.
In this practice, the blast furnace operator worries about production
rate and fuel economy and does not bother about the sulfur content of iron,
which is the worry of the steelmaker.
Indian Practice of Acid Burdening

The alumina content of Indian iron ores is rather on the high side but
adoption of washing-wet-screening practice of ore beneficiation at
almost all the major iron ore mines has led to iron ore lumps of less than
2% alumina. The fines are also being classified and upgraded to reduce the
alumina content of sinter subsequently produced from these fines. The net
result is that the operating temperatures of blast furnaces need not neces-
sarily be as high as was earlier necessary. Taking advantage of this and adop-
tion of external desulfurization, it is now possible to follow “acid” burdening
of blast furnaces. Many steel plants in India have adopted external desul-
furization. Consequently the slag basicity in blast furnaces is being brought
down to less than 1.0 from the earlier 1.2–1.3. Previously, MgO was added
in the burden in the form of part replacement of limestone by dolomite
to obtain fluid slag with high alumina content. Mn-ore/slag was added to
obtain required Mn content of the iron.
Now for 0.95–1.0 basicity, it is necessary to add some silica in the burden.
This is being done by direct addition of quartzite in the burden or in sinter-
mix. In high alumina ores, assimilation of silica requires a high heat input,
which defeats the very purpose. A suitable flux material must have MgO and
SiO2, and when introduced in the sinter-mix it can solve the problem of assim-
ilation without any difficulty. Olivine, a silicate of magnesium, can be a useful
material in this respect. Dunite, a rock form of olivine, was tried in Tata Steel
and proved successful. Now a cheaper substitute in the form of pyroxenite has
been found and is currently being used to finally obtain the required basic-
ity of 0.95–1.0. These necessarily must be added in the sinter-mix and not
directly in the furnace. Addition of olivine in some form in place of dolomite
in the sinter-mix, the limestone addition goes up by about 60%–70% as com-
pared to when dolomite is added. But this is better in performance, finally.
There is a limit to which super-fluxing of sinter can be carried out.
The higher the super-fluxing, the lower shall be the proportion of sinter in
the burden. The above practice allows much high proportion of sinter to be
included in burden.
15.8 Disposal of Metal

Iron is produced in two different types of plants if viewed in respect to dis-
posal of molten iron. In merchant blast furnaces iron is cast invariably in pig
casting machines to produce solid pigs for their sale to foundries. The prob-
lem of disposal of molten iron in such cases is a simple affair. But if iron is
produced in an integrated steel plant essentially for its conversion to steel,
the iron disposal is quite a complicated affair. An integrated steel plant is
a material-movement-intensive industry and it can be run efficiently only
by managing the logistics of the entire plant, including that of iron disposal
itself. There is no standard solution to this problem because every plant has
its own peculiarities with respect to the following:
1. Number of blast furnaces provided
2. Their individual capacities
3. Their tapping schedules
4. Number of steelmaking shops provided
5. Capacities of individual SMSs and their iron requirements
6. Types and capacities of ladles provided for transport of iron
7. In-house foundries and their iron requirements
8. Mixer capacities provided
9. Initial design and inefficiency generated with use or modifications
adopted at various stages of SMS operations
10. The transport system itself, which may include bogies, locomotives,
ladle relining facilities, etc.
11. Maintenance and pre-heating of metal ladles
12. Variations in tap-to-tap time in each shop and its variations with
shutdowns and breakdowns
This is a highly complicated logistics and in order to obtain the highest
possible efficiency of hot metal supply, to finally obtain maximum steel
production, the logistics need to be controlled by an expert electronic
system alone.
Because of the above considerations, the logistics of iron flow need

to be reviewed again and again to minimize delays and loss of production
down the flow lines.
Foundry iron and any excess basic iron are normally sent to pig casting
machine for producing pigs, which are generally of handlable size, such as
25 kg or so.
Data in Tables 15.3 and 15.4 indicate the complexities of production
and distribution of iron by way of an example of Tata Steel.
Table 15.3 Details of blast furnaces

Blast Furnaces Production No of Minimum Maximum
(t/day) Castaways Cast-to-cast Time Cast-to-cast Time
A 1300 4 50(2) 150(4)
B 700 3 65(2) 180(3)
C 800 3 60(2) 180(3)
D 1300 4 50(2) 150(4)
E 900 3 50(2) 180(3)
F 1600 2 50(2) 140(4)
G (new) 3600 2 30/1/ 100/4/
Where the figure in the brackets ( ) indicate open top ladles produced
and figures in // indicate the number of torpedo ladles produced.
Table 15.4 Details of the hot metal – consuming shops

Consuming Requirements Vessel Cycle No. o f Open-top Ladles No. o f Torpedo Ladles
Place (t / d a y ) Time (min) Required per Day Required per Day
LD shop 1 4480 45 70 cannot take
LD shop 2 4060 40 – 20
SMS–3 650 – 10 cannot take
Foundries 130 – 2 cannot take
The pig casting machine is an endless belt of grey iron molds of nearly
50 cm in length, 20 cm in width, and 10 cm in depth, with adequate taper for
easy stripping. The molds travel at a linear velocity of approximately 10–25 cm
per second. The average rate of pouring may be more than one metric ton per
strand per minute. The molds are protected by a tar or lime coat after every
pour to make it ready for the next pour.
The orthodox system of casting of pig iron in a sand bed in the cast-
house does not exist anymore since pig casting machines, which can do the
casting job easily and readily, are now universally adopted.
The possibility of granulation of pig iron, like slag granulation in place
of casting pigs in pig casting machine, is currently being investigated.
Preliminary trials and assessment indicate that it might work out to be more
economical than the pig iron casting machine. If it comes out to be true on
a commercial level, metal granulation may come up in the near future to
substitute for pig casting machines.
Pig iron for steelmaking is transferred to the mixer, which is usually
located near the steelmaking shop, in simple or specially designed torpedo
ladles. The torpedo ladles are more complex in construction and costly.
These reduce the heat losses during transit and minimize the blast furnace
slag, entrained with the metal, being carried into the mixer and, from there
to the steelmaking furnace. Molten metal for steelmaking is stored in a
holding vessel known as a mixer. Molten iron is therefore available for steel-
making, as and when required, rather than waiting for the blast furnace tap.
The blast furnace slag that goes with the hot metal into the mixer may
ultimately find its way into the steelmaking furnace. Blast furnace slag is
acidic and is therefore detrimental to steelmaking, which is generally a
basic process. The blast furnace slag must therefore be skimmed off as far
as possible from the metal while being tapped. Additional efforts to remove
slag from the top of the metal, for example, by way of water quenching, are
also beneficial.

The main product of BF smelting is molten iron. Slag production cannot
be avoided, but its volume is kept to a minimum. Fortunately, slag is no
longer a waste product. The entire slag can be used as a cement-making raw
material. The chemistry of metal is very important from the standpoint of
steelmaking. Every effort is made to obtain the chemistry of metal such that
the process from ore to steel, as a whole, is more economical.
IMPACT OF BURDEN
CHAPTER
16
PREPARATION AND
DISTRIBUTION ON BLAST
FURNACE PERFORMANCE
In This Chapter
● Introduction
● Burden Distribution
● Burden Trajectory and Profiles
● Burden Distribution Patterns
● Blast Furnace Performance
● Effect of Sizing
● Use of Agglomerate as Burden
● Sinter vs. Pellets
● Agglomerate vs. Natural Lumpy Ore
● Sizing of Coke and Its Quality
● Flux Quality
16.1 Introduction
The remarkable furnace productivities and low coke rates that have been
obtained in the post–World War II era have exceeded most of the earlier
expectations. This has been possible mainly due to better distribution of
more carefully prepared burden, better furnace design in the light of raw
material properties, and introduction of a few additional innovative tech-
niques in running the blast furnace.
The problem of burden distribution inside the blast furnace has already
been covered in Chapter 5 with respect to use of a traditional double-bell-
top charging and distribution system. As the ideal qualities of burden mate-
rials were increasingly revealed, more and more attempts were made to
obtain those properties even by additional processing. The quality require-
ments of the raw materials and the ways of improving the same in the natu-
rally available raw materials have been dealt with in Chapters 6, 7, and 8.
The rate of production in a blast furnace essentially depends on the rate of
air blast passing through the furnace. The higher it is, the more is the produc-
tivity. The blast furnace, being a packed bed of particles, provides the neces-
sary voids or gross permeability, in the charge inside the furnace, for gases to
pass through in the upward direction. If the burden consisted of particles of
uniform shape and size, then the permeability is fixed and thereby the limiting
rate of air blown through it is also automatically fixed. The limiting rate would
be reached when the bed gets almost fluidized due to the pressure of the gas.
However, the burden is far from such an ideal situation. The charge is neither
uniform in shape and size nor is it possible to obtain uniform gross permeabil-
ity all through the horizontal section of the furnace.
This non-uniform size develops variable permeability in the bed, depend-
ing upon the non-uniformity and the way it is charged inside the furnace. It
is reported that a 10% decrease in this permeability increases the resistance
to gas flow by nearly 300%. The adverse influence of decreased permeability
is so aggravated that every attempt is made to improve the permeability. The
furnace productivity can therefore be increased by preparation of the bur-
den to minimize the non-uniformity in size and shape. It is generally done
by sintering and/or pelletization of ore fines. It is further aided by adopting
appropriate charging methodology with respect to iron-bearing material and
coke to obtain uniform permeability all through the furnace cross section. The
agglomeration of ore fines have been dealt with in Chapters 7 and 8. The tra-
ditional distribution of charge has been dealt with in Chapter 5.
Impact of Burden Preparation and Distribution... • 307
Whatever is the pattern of charge distribution obtained at the throat

where the charge is put in the furnace, it continues below until the cohe-
sive zone. If any non-uniformity is generated at the top, then nothing can
be done to alter it. This non-uniform radial distribution of charge affects
gas flow adversely. Hence all possible attention is paid to obtain optimal
distribution of charge at the top. Several devices have been developed to
achieve this optimal distribution. Once the limitations of the double-bell
Mckee Top were understood, a bell-less top was designed by Paul-Wurth,
which has been very popular until recently. It is now being challenged by
the Gimble-Top. Hence it’s a two pronged attack, viz. the burden prepara-
tion and its charging.
Burden preparation essentially includes the following:
1. Sizing to as close a range as is economically feasible
2. Agglomeration of fines into useful-size agglomerate
3. Incorporation of flux in the agglomerate
4. Improvements in physical and chemical properties of coke
5. Last-minute screening just before charging to remove fines generated
in transit
Burden preparation is resorted to in order to obtain properties as close to
ideal properties as is economically feasible. The effect of such prepared
burden on the performance of a blast furnace is discussed in the present
chapter along with the distribution of charge at the stock-line using modern
charging equipment.
16.2 Burden Distribution

The charge consists of various materials of varying sizes and physical prop-
erties. It is difficult to distribute them at the top such that it offers mini-
mum resistance to the gas flow in horizontal as well as vertical sections.
Distribution is therefore assessed in terms of its three zones of the charge
inside the furnace:
1. Peripheral
2. Intermediate
3. Central
The intermediate zone consists of nearly 50% of the total cross-sec-

tional area, and nearly 30%–35% of the gas passes through it. Each of the
central and the peripheral zones consists of 25% each, and 65%–70% of the
gas passes through them. However, the majority of the gas passes through
the central and the peripheral zones. More uniform distribution of charge
in these regions is of paramount importance.
It is possible to achieve this by burden distribution at the stock-line. It
should be understood that gases pass readily through coke regions because of
better permeability. The more the gas passes, the more will be the heating in
that region. Hence more coke must be present in the central region so that
the central zone will be at higher temperature. The temperature at the periph-
ery will be relatively lower, and the wall refractory life will consequently be
improved. Coke is therefore charged in the form of rings towards the center,
and iron charge is put into rings more towards the periphery, resulting in more
pronounced central working of the furnace. This, together with properly sized
ore and agglomerated burden, is bound to give better performance.
16.3 Burden Trajectory and Profiles

The distribution pattern has been discussed in Chapter 5. Two different
profiles of the burden emerge, depending upon whether the particles hit
the wall directly or the stock surface. If it hits walls directly, then the fines
get accumulated more at the periphery, and stock takes the “V” profile.
If it hits the stock surface a little distance away from the walls, then “M”
profile of stock is formed, wherein the fines get accumulated at the inter-
mediate zone. Major gas thus passes through the central and peripheral
zones in M-shape. In V-shape, coke and coarse material gets accumulated
in the center. Fines accumulate at the peripheral zone. In M-shape, profile
fines get accumulated in the intermediate zone, and the central and the
peripheral zones are relatively more permeable and hence allow more gas
to pass through. It is necessary to understand that if peripheral permeabil-
ity is more it will get unnecessarily overheated and may cause damage to
lining. The radial distribution by a bell-less top, by adjusting the ore/coke
ratio readily, helps to obtain the distribution much better. The charging is
therefore resorted to in order to obtain V and M profiles to finally get the
desired distribution as mentioned above.
The permeability of the burden gives rise to temperature profiles
accordingly. More permeability means more gas flow and therefore higher
temperature of the burden and vice versa. Broadly, six different tempera-
ture profiles have been identified depending upon the charge distribution,
and are shown in Figure 16.1.
FIGURE 16.1 Six classes of temperature profile in the burden as a function of the burden distribution
pattern.
16.4 Burden Distribution Patterns

The patterns can be described as below:
Flat profile is as shown in (1)
Inverse V or U is as shown in (2)
-shape with some peripheral gas flow is shown in (3)
W-shape with considerable peripheral flow is shown in (4)
Inverse V-shape with some peripheral gas flow is shown in (5)
Inverse V-shape with little peripheral gas flow is shown in (6)
Profile (1) is abnormal and may be found when marked decrease in gas
flow is seen, perhaps due to changing or such irregularities.
Profile (2) is again abnormal and is seen if the deadman is inactive. It

may also occur if excessive coal injection is carried out and it is not burned
completely.
Profile (3) represents a commonly adopted balanced distribution in
which small peripheral gas flow is taking place and which prevents wall
accretions and other irregularities.
Profile (4) means excessive peripheral gas flow and is adopted only
when cleaning up of the wall build-up is resorted to.
Profile (5) is more similar to profile (3), but has a more pronounced
central gas flow. This may be observed when the fuel rate is very low.
Profile (6) is preferred when there exists no danger of wall accretion because
of low alkali contents in the charge and because high PCI is resorted to.
Circumferential balancing is adopted through charging and ore/coke
ratio to obtain uniform gas flow all through the cross section and along the
height. This also minimizes the peripheral gas flow to safeguard the walls
from getting heated and becoming prone to accretion.
The required distribution is readily obtainable using a bell-less top with
a chute distributor. The mechanical details of this along with its brief his-
tory and other designs, that preceded this top design, will be described
in the following chapter. In each furnace, with the given type of charge
materials and coke rate, the charging practice is developed to obtain the
optimum burden distribution. If anything goes wrong the ways and means
of off-setting such a situation are also evolved. Computerized expert-sys-
tem-control helps a lot in this respect. A simple precaution is often taken
to see that the stock-line does not drop below the vertical cross section of
the throat. Whenever it drops, then the distribution gets affected adversely.
The importance of maintaining the stock-line need not be over-empha-
sized. Charging sequences should be such that the material does not strike
the wall and help create segregations of sizes.
16.5 Blast Furnace Performance

The performance of a blast furnace is assessed in terms of:
(a) Productivity as shown earlier and usually as thm/m3/d (Section 2.8)
(b) Coke rate per metric ton, since it is a major cost item
(c) Campaign life of the furnace
(d) Working temperature of the furnace, that is, hearth temperature

(e) Regular shutdowns required to maintain the furnace campaign
(f) Chemical quality of iron ore, in particular the alkali content
(g) General irregularities in operation
(h) Furnace design efficiency
(i) Hot blast temperature
(j) Quality of staff manning the furnace operation
(k) Chemistry of the hot metal, particularly with respect to silicon and
sulfur contents
(l) Slag rate, that is, amount of slag produced per metric ton of hot metal
(m) General furnace operational control
(n) Guniting frequencies before campaign ends
Out of these, except for the burden quality and the coke rate, the rest are
assumed to be at their possible best status. Hence in actual practice, and
unless some irregularity with respect to these variables crops up, much of
the attention is paid to improve the burden quality and the coke rate. The
coke rate does not depend only on the burden quality, but it also depends
on how the furnace is operated.
For a given furnace design and operation, the performance of a blast
furnace shall be essentially decided by the blast rate and the coke rate. The
higher the blast blown and the lower the coke rate, the better the furnace
performance. The role of burden preparation in order to ultimately obtain
better distribution inside the furnace need not be over-emphasized. The
charge quality that results in higher gross permeability of the burden has
to be worked out. Similarly, that design of charging equipment that gives
better gross permeability is to be adopted. Burden preparation means as
uniform a size and shape as are economically possible. Burden is prepared
in practice by the following means:
(a) Sizing and elimination of fines

(b) Agglomeration of fines by any economically viable process
The effect of each one of these factors on the productivity is assessed in the
present chapter.
16.6 Effect of Sizing

It would be an exception now that the charge is put in the furnace without
screening. Screening is often effected only at the mine site for the ore and
at the coke oven plant for coke. The effects of sizing of ore on productivity
and coke rate, as observed in various practices, have been summarized in
Table 16.1.[62, 63] The ill-effects of fines present in the burden on blast
furnace performance have been summarized in Table 16.2.[62, 64]
TABLE 16.1 Effect of sizing of ore on production and coke rate

Plants Ore Size Increase in Decrease
Production in Coke
(%) Rate (%)
from mm to
mm
Columbia Steel Co. Irontown plant crude ore −50 20 18
Kaiser Steel Corporation Fontana plant −38 −25 14 14
Nippon Kokan, Kawasaki Iron Works (20%–25%) +40 −40 N.A. 5.3
Yawata, Japan Higashida 40−10 30−8 10 3
Fuji Steel, Japan Fuji plant 50−13 25−10 3 8
Yawata, Tabata Works 40−8 30−8 3.6 5.5
TABLE 16.2 Effect of removal of fines from the burden
Plant Ore Burden Reduction in Fines Increase in

Production (%)
from (%) to
Edgar Thomas Works sinter 60−6 mm 38−6 mm 21

U.S. Steel Corp.
Armco, Middleton pellets 46−6 mm 26−6 mm 17
Appleby sinter 35%−10 mm 5%–10 mm 30
Frondingham, U.K.
Australian Iron and Steel Co., sinter 35%−10 mm 11%−10 mm 12
Port Kembla
Cherepovets works Russia sinter 9.2%−5 mm 3.1%–5 mm 7
Fines are detrimental to furnace operation no matter where these are

generated. Even after early screening, the material undergoes degradation
during its transit from mine to the furnace. The example of Bhilai plant in
India can be quoted as an illustration. Although screened iron ore lumps
are dispatched from the mines, the natural ore charged in their furnaces
may contain as high as 30% fines. Degradation occurs due to physical
bumps and drops as the ore is in the process of charging. The ill-effects of
transportation can be minimized by the last-minute screening of the charge
just before it is loaded in the skip. This is fast becoming a common practice
on modern large-capacity furnace since it has further improved the furnace
performance. Even the existing old plants, wherever possible, have provided
this facility. For example, coke rescreened in Bhilai plant at the 1719 m3 fur-
naces just before loading the skip.
The actual size of the ore used as a burden depends on its crushing
strength, reducibility, etc. In deciding the size of any charge material, the
relative sizes of ore, agglomerate, coke, flux, et cetera, have to be taken into
account to obtain the best performance of the furnace.
16.7 Use of Agglomerate as Burden

The remarkable improvement in furnace performance achieved by using
sized lumpy ore as charge led to the development of agglomeration of ore
fines as a routine method of burden preparation. Of all the agglomeration
methods, sintering and pelletizing of iron bearing fines alone have been
universally adopted for their overall economy.
Sinter
The beneficial influence of addition of sinter in place of ore lumps has to be
assessed in terms of the following parameters, which are nothing but those
factors that affect reducibility, degradation, slag formation at desired level,
coke rate, and so on:
1. Maintenance of or improvement in permeability of the burden
2. Reducibility of sinter vis-a-vis ore lumps
3. Slag formation at the correct level inside the furnace
4. Coke rate
On all these accounts it has been observed that sinter in the burden improves
the situation as a whole. The strength factor and reducibility as assessed in
terms of RDI index of sinter is better than that of ore. The temperature
zone of slag formation and softening is better controlled by use of sinter
because of prior adjustment of sinter and flux combinations in the form of
fluxed or super-fluxed sinters. Because of these improvements, the coke
rate comes down as a consequence.
Experience with sinter has indicated a rise of blast furnace output by
3.5% and reduction in the coke rate of 3% for every 10% increase in sinter
proportion of the burden. This trend continues up to 50%–75% sinter in
the burden beyond which the advantages gained decreased.[71] Coke can
be decreased further by using fluxed sinter, depending upon the propor-
tion of total flux requirement incorporated in the sinter mix. The use of
superfluxed sinter is now a common practice, and the reasons are given
below:
1. The limestone is calcined during sintering, and this large heat-con-
suming reaction is eliminated from the blast furnace process.
2. The limestone in the sinter mix combines with the gangue of the ore
forming the slag forming constituents, which facilitate slag formation
in the furnace.
3. Increase in basicity of sinter beyond 1.5–1.6 increases its strength
and hence, degradation of sinter during handling and charging will
be relatively less in high-basicity sinters. It would improve perme-
ability in the furnace and the attendant advantages would accrue
from it.
Blast furnace operations show improvements in productivity and smooth
the operation with increasing strength and basicity of sinter. According to
the Japanese practice, it is desirable to keep the shatter index of +10 mm
sinter above 84% so as to keep the operation smooth. Elimination of fines
and use of a close-size-range of sinter is highly beneficial.
Pellets
Iron-bearing fines agglomerated into pellets also forms a good blast fur-
nace burden material since it possesses better blast furnace properties.
The nearly spherical shape, close size, and desired porosity, coupled with
good strength, lead to high permeability in the charge-containing pellets.
All these factors have beneficial influence on the gas utilization, as can
be seen from the top gas analysis. The usual CO/CO2 ratio of two was
found to decrease to one by using pellets in the charge at the Steel Co
of Canada.
The pioneering work in the field of use of pellet burden was carried
out at the Armco Steel, USA. The Armco furnace, when operating on
lump ore and sinter, had output of 1460 short tons per day with a coke
rate of 1623 lbs. and wind rate of 83,000 ft3/min. With 90% pellets, the
output increased to 2804 short tons per day with a coke rate of 1276 lbs.
and wind rate of 109,000 ft3/min. The high permeability of the pellet
burden allowed higher blowing rates. The production rate later stabi-
lized at 3200 short tons per day. Similar data from other plants have been
compiled in Table 16.3 to indicate the beneficial effects of pellet burden
on furnace performance.
The swelling property of pellets as a result of transformation of
hematite to magnetic during reduction is an important parameter that
decides the extent to which pellet burden affects the furnace perfor-
mance. Excessive swelling decreases the crushing strength and conse-
quent danger of disintegration of the pellets and loss of permeability of
the burden.
At an experimental blast furnace running on 65% pellets and 35% coarse
ore charge, the operation was smooth if the degree of swelling was less than
20%. When the swelling rose to 40%, the furnace could be operated only
with difficulty even at high blowing rates. It has therefore been concluded
that[65]
1. Pellets with swelling less than 20% are safe for use in the furnace to
any extent
2. Pellets with 20%–40% swelling should not occupy more than 65% of
the total burden
3. Pellets with more than 40% swelling are positively detrimental and
should not be used in the burden as far as possible
It has been suggested that the degree of swelling reduces as the crushing
strength improves. Use of bentonite as a binder has therefore been recom-
mended to produce stronger pellets.
TABLE 16.3 Effect of pellet burden on furnace performance
Plant Agglomerate in Burden % Increase in % Decrease in

Production Coke Rate
from to
Pellet Sinter Pellet Sinter

(%) (%) (%) (%)
Armco Steel, USA 10.9 nil 77.7 nil 83.5 3.0
Ford motor, USA 5.0 45 60.0 40.0 21.4 3.8
KNHS, Netherlands nil 90.0 50.0 40.0 20.0 5.8
Tsurmi No. 1 B.F. nil 58.2 65.0 11.2 6.5 3.8

NKK Japan
Tsurmi No. 2 B.F. nil 42.0 30.0 38.2 2.8 4.3

NKK Japan
It has been suggested that the size, chemistry, mineralogy, and physical
and reducibility properties required of the pellets should be specified for
a given practice. The Germans insist on crushing strength of minimum
200 kg/pellet for 10–15 mm size. The size distribution should be 80%
between 10–15 mm, 85% between 9–16 and 5% maximum of less than
5 mm size. The A.S.T.M. tumbling index of −5 mm should not exceed 5%
and −6 mm 6%. The B.S.C. Laboratories have also recommended more or
less the same specifications but have also specified other requirements, par-
ticularly swelling, reducibility, and L.T.B.T. values.[45] The L.T.B.T. values
should be a minimum 10% +6.3 mm−5.0 mm index of 7.5% maximum. The
swelling should not exceed 20%.
16.8 Sinter vs. Pellets

In general, natural lumpy ore or sinter or pellets or a suitable combina-
tion of two or more of these, form the burden. The relative proportion of
each in the burden is dictated by various economic considerations. The
modern large-capacity furnaces necessarily need fully prepared burden
to maintain their productivity, since the required blast furnace properties
cannot just be met by natural lumpy ore. The selection of the process of
agglomeration, whether sintering or pelletizing, will depend upon the
type of ore fines available, the location of the plant, and other related
economic factors involved.
Sintering is preferred if the ore size is −10 mm to +100 mesh and if it

is −100 mesh, pelletizing is generally adopted. Pelletizing in fact requires
ultrafines of more than 75% of −325 mesh. These processes are therefore
not competitive.
Production of super-fluxed sinter is now a common practice, whereas
production of fluxed pellets has yet to surmount some barriers on a com-
mercial level.
Porosity of sinter is 10%–18% and that of pellets is 20%–30%.
The shape of pellets is near spherical and hence bulk permeability of
the burden is much better than that obtained from sinter, which is non-
uniform in shape.
The installation cost of a pelletizing plant will be 30%–40% more than
that of a sintering plant of an equal size. The operating cost of sintering
is, however, slightly less than that of pelletizing. Almost 50% of the cost of
sinter comes from fuel used for sintering.
Both sinter fines and pellet fines are equally detrimental and should be
eliminated from the charge to ensure better performance of the furnace.
It is possible to produce pellets of required quality economically. The
use of pellet burden reduces incidence of hanging, which is otherwise
observed only if a high proportion of sinter is used. The periods of hanging
while using sinter are rather too long, which affect productivity and quality
of iron adversely.
The advocates of pellets often attribute to pellets more advantages than
are actually due. The higher driving rates claimed for pellet burden have
also been achieved by sinter burden. In one Japanese plant, two similar
furnaces were operated, one with 80% pellet burden and another with 80%
sinter, and both were observed to have similar performance. In fact, the one
on sinter had an edge over that on pellets. This only means that good sinter
is as acceptable to the furnace as good pellets.
It has been observed that burden with nearly 40%–60% pellets and
balance sinter, with at the most 10% ore was found to be a better burden
than 90% sinter burden. The difference in productivity was nearly 15%,
and coke saving was approximately 25 kg/t.
It should also be noted that degradation of sinter during its transit is
much more than that of pellets. The sinter, therefore, has to be produced
nearby the blast furnace plant, whereas pellets can be carried over a long
distance without appreciable degradation. If high rates of productivity
demand elimination of fines, and since sinter happens to contribute more to
the generation of fines than that of pellets, the latter will have to be chosen
as the burden in preference to sinter.
16.9 Agglomerate vs. Natural Lumpy Ore

All iron-bearing materials, on entering the blast furnace, are known to dis-
integrate to varying degree and produce fines in the stack region (i.e., due
to the breakdown at low temperatures under reducing conditions). The
main culprit of this degradation inside the furnace appears to be the volume
change accompanying the reduction of hematite to magnetite or wustite.
The markedly lower tendency of sinter to degrade is due to the resistance
of the slag-bond region to crack propagation, since the crack propagation
is arrested at these bonds during reduction of sinter. The total proportion
of CaO + SiO2 + FeO in sinter should therefore be greater than 25.0% to
eliminate disintegration of sinter inside the furnace.
An interesting study of blast furnace data for the 1965–78 period at
TISCO reveals that improved blast furnace performance when sinter and/
or pellet up to a total of 70%–80% is used in the burden, is more due to the
lesser tendency of agglomerate to degradation inside the furnace than that
of natural ore[68]. This is substantiated by the fact that dust losses in the
corresponding period are proportionately decreased while using agglom-
erate in the burden. Under identical conditions, smaller size (10–15 mm)
particles of ore degrade more than larger size (20–25 mm) particles.
Emphasis on last-minute screening of the charge can now be appreci-
ated since it will at least eliminate all those fines generated until then.
16.10 Sizing of Coke and Its Quality

It has been observed that by sizing the coke, blast furnace operation is
markedly improved. With a smaller size of coke the endothermic gasifica-
tion of coke (E.C.G.) at an equivalent coke rate is decreased whereby the
coke rate decreases and the production rate increases.
Coke is the backbone of the blast furnace process since the burden
permeability in the bosh regions, where everything else is molten, is
maintained by coke alone. The extent to which this is achieved by the coke
directly affects the furnace performance. Since coke also undergoes degra-
dation before reaching the bosh zone, the initial size has to be decided by
regression analysis knowing the size requirement in the bosh region. The
performance could be improved only after improvement in coke quality by
way of its size and strength.
High productivity is obtained if +40 mm fraction in coke charge is at
least 80% with a top size of 80–100 mm. This is the reason why last minute
screening, just before charging the skip, is resorted to for coke at many
plants as a routine practice.
16.11 Flux Quality

Since every attempt is being made to produce superfluxed sinter or pellets,
thereby avoiding charging of flux as a separate charge material, the impact
of preparation of flux as a separate charge material has more or less become
of minor relevance and hence it has not been discussed here. The relevant
information is not readily available either.

Preparation of burden, by way of obtaining close size range for ore, coke,
sinter, and pellet, has resulted in improving the blast furnace perfor-
mance remarkably. Agglomeration of fines into sinter or pellet of better
strength and other properties has almost become a “must” to run blast
furnaces efficiently.
CHAPTER
17
MODERN BLAST FURNACE
DESIGN AND PRACTICE
In This Chapter
● Introduction
● Large-Capacity Blast Furnaces
● Coke Quality
● Burden Preparation
● Improved Distribution of Charge
● High Top Pressure
● Charging Devices for High Top Pressure
● Higher Blast Temperature and Driving Rate
● Oxygen Enrichment of Blast
● Humidification of Blast
● Fuel Injection
● Lime Injection
● Pre-reduced Ore as Burden
● Injection of Reducing Gas at the Bottom of the Stack
● Utilization of Plant Iron-Bearing Wastes
17.1 Introduction
Considerable advances have been made in recent years in blast furnace
design and practice to improve upon its economy, both in terms of reduced
consumption of coke and increased productivity. The improvements in
the design of the blast furnace and its accessories have been mainly in
the direction of increasing the size and its capacity. This has necessitated
improvement in the blast furnace structure, cooling efficiency, gas clean-
ing system, hot blast generation, and so forth. The design of the lining and
the quality of lining material, coupled with efficient cooling, particularly
of bosh and hearth, have undergone improvements that have led to reduc-
tion in its thickness and attendant increased useful volume for the same
shell size. Efficient material handling, last-minute screening facilities just
before charging, better cast-house designs, efficient liquid product dis-
posal, et cetera, have been inevitable developments to run large-size fur-
naces efficiently. Proper burden preparation and distribution of charge
in the furnace are of vital importance in improving furnace productiv-
ity. Adoption of high top pressure has necessitated several drastic design
changes in the furnace top charging system. The understanding of the
physico-chemical nature of the blast furnace process has led to numer-
ous improvements in the practice. Even the existing furnace designs have
been modified, wherever feasible, to incorporate the necessary changes in
the design and consequently the practice to improve the overall produc-
tion economy.
The principal achievements are listed below and also are discussed later
in the present chapter.
1. Large-capacity blast furnaces
2. Better-prepared burden
3. Efficient burden distributor
4. Better coke quality
5. Auxiliary fuel injection through tuyeres
6. Humidification of blast
7. (Oxygen enrichment of blast
8. Higher blast temperature through better stoves
9. Blast furnace gas injection through the stack
10. Lime dust injection through tuyeres
Modern Blast Furnace Design and Practice • 323
With these improvements in design and practice, it has been possible

to achieve a production rate of more than 2.5 t/m3 useful volume or more
than 2.8 t/m2 of hearth area with fuel consumption of the order of 350–
450 kg/thm, which may include additionally 150 kg/thm PCI. Since coke
contributes nearly 70% of the cost of iron, any reduction in its rate by way
of its replacement by any other cheaper fuel is bound to contribute most
to the economy of production. Injection of pulverized or granular thermal
coals, natural gas, or any other gaseous or liquid fuel is aimed to reduce
coke rate and thereby the cost of iron production. Any productivity increase
by adoption of one or the other means also contributes to improving the
economy of production. The general picture of increasing the blast furnace
size and the related operational parameters are shown in Table 17.1.
The trends in improvements in design and operation of BFs should be
seen from two perspectives: one is to install a modern furnace with all these
facilities and the other is to improve the existing furnace, to the extent possi-
ble, by incorporating the new design features and operational practices which
can improve the economy of production. These improvements, along with
the underlying metallurgical principles, have been discussed in what follows.
Table 17.1 Operational impact of increasing size of blast furnaces

Blast Furnace Small Medium Large
Hearth diameter (m) 8.5 12 15.5
2 1550 3300 5700
Inner vol. (m )
No. of tuyeres 24 32 42
Production (tpd) 3358 7150 12350
3 2565 5500 9434
Blast vol. (Nm /min)
Blast vol./tuyere (Nm3/min) 106 172 232
Injection rate (kg/thm) 150 150 150
Injection rate/tuyere (tph) 0.9 1.4 1.8
17.2 Large-Capacity Blast Furnaces

Having realized that the capital and operational costs of a large blast fur-
nace are lower than those of a battery of furnaces of the same total capac-
ity, the size of the blast furnaces have been progressively increasing, and
blast furnaces of smaller capacities are no longer economically attractive.
The same is true with the auxiliary units such as coke ovens, agglomeration
plants, stoves, etc.
The size of a blast furnace can be compared either in terms of its hearth
diameter or the total useful volume. The hearth diameter of approximately
5.6 m prevalent in the 1925–1930 period increased to nearly 8.9 m by
1950–1955, nearly 12 m by 1970, and now it has reached nearly 15 m in the
largest blast furnace. The profile of one such modern furnace is shown in
Figure 17.1. The height has more or less remained constant and perhaps
would continue to be the same even in the near future. The useful volume
of furnaces has also gone up from the earlier 1000 m3 to nearly 4500 m3,
and a furnace of more than 5000 m3 is designed. The daily production of an
individual furnace increased from a mere 500 t/day in 1930 to 2000–3000 t/
day in the fifties, 6000 t/day by 1970, and presently nearly 12,000 t/day from
the largest furnace. Details and location of some of the very large-capacity
furnaces that are more than 13 m diameter are shown in Table 17.2. A
16,000 t/day furnace may be planned in the future.
FIGURE 17.1 Profile of a modern large-capacity blast furnace.
The specific production capacity in the last few decades has also shown
a tremendous improvement from nearly 0.8–1.0 t/m3/day to more than
2.5 t/m3/day from some of the most efficient furnaces of today.
In Japan, two furnaces in the 5000 m3 useful volume class are in opera-
tion. Construction of a furnace of this scale requires 6000 t of bricks and 4500
t of steel for the furnace body alone. It also requires an additional 4000 t or
more for the auxiliaries and piping. A blast furnace of this type in operation
holds approximately 10,000 t of iron ore, coke, and iron, bringing the total
weight of the furnace to approximately 24,000–25,000 t. to enable the blast
furnace, whose height reaches 120 m, to withstand earthquakes, its foun-
dation must be built strongly enough to resist lateral pressure equivalent
to 5000 t. Such a foundation requires some 300 pipe piles 1.2–1.3 m in

diameter and 10,000 m3 of concrete. All this is for the blast furnace alone,
and when all the other equipments in a blast furnace unit are included, the
foundation work for a single furnace requires 2000 piles weighing a total of
10,000 t, and steel bar of approximately 1800 t and 30,000 m3 concrete. The
thickness of the shell plate is 36 mm at the top, 60 mm in the middle, and
65 mm at the bottom. The shell thickness at the tap hole is 110 mm and
90 mm around the tuyeres. Carbon blocks are used for lining the hearth.
Bottom thickness is between 3–4 m and the wall thickness is 1–1.5 m. The
individual carbon block is 500 mm wide, 600 mm high, and 2–3 m in length
and weighs approximately 1.5 t. Space between the blocks is less than one
millimeter, requiring a high degree of precision in their installation. This
furnace is designed to run for 7–10 years and thereby produce 25–35 mil-
lion metric tons of pig iron during its campaign.
Table 17.2 Large blast furnaces in the world

Country Time of Firm Works B.F. Hearth Inner Volume
Blowing-in No. Diameter (m) (m3)
Japan April 71 Nippon Kokan Fukuyama 4 13.2 4197
Japan Sept. 71 Nippon Steel Kimitsu 3 13.4 4063
Japan April 72 Nippon Steel Oita 1 14.0 4158
Japan July 72 Nippon Steel Tobata 4 13.5 3799
Netherlands Nov. 72 Estel Ijmuiden 7 13.0 3667
Japan Jan 73 Kobe Steel Kakogawa 2 13.2 3850
Japan March 73 Sumitomo Metal Kashima 2 13.8 4080
Japan April 73 Kawasaki Steel Mizushima 4 14.4 4323
France May 73 Usinor Dunkirk 4 14.0 4526
Japan Nov. 73 Nippon Kokan Fukuyama 5 14.4 4617
Italy Nov. 74 Italsider Taranto 5 14.0 4128
U.S.S.R. Dec. 74 Krivoy Rog Krivoy Rog 9 14.7 5026
Japan March 75 Nippon Steel Tabata 1 13.4 4140
Japan Oct. 75 Nippon Steel Kimitsu 4 14.0 4930
W. Germany April 76 A.T.H. Schwergern 1 13.6 4689
Japan Sept. 76 Sumitomo Metal Kashima 3 15.0 5050
Japan Oct. 76 Nippon Steel Oita 2 14.8 5070
Japan Nov. 76 Nippon Kokan Ohgishima 1 13.5 4052
Japan June 77 Kawasaki Steel Chiba 6 14.1 4500
The general picture of increasing the blast furnace size and the related
operational parameters are shown in Table 17.2 and Table 17.3.
Table 17.3 Design and operational data of some of the largest blast
furnaces in the world
B.F. Inner Hearth Period Productivity Utilization
Volume diameter (%)
(M3) (m)
tpd/m3 tpd/m3
SMI Kahima BF-3 5050 15 2003 1.9 54.3 98.3
JEE Chiba BF-6 5153 15 2003 1.8 52.5 97.5
Thyssen Schwelgerm BF-2 5513 14 2005 2 .1 66.4 95.3
NSC Kimitsu BF-4 5555 15.2 2003 1.9 58.2 97
Severstal Cherepovets BF-5 5500 15.5 2002 2.0 59.1 98
NSC Oita BF-2 5775 15.6 2003 2.26 68 97.1
These very large-capacity furnaces cannot be run efficiently; that is,

the required permeability in the burden and driving rates cannot be main-
tained, unless entirely prepared burden of the right quality is fed to run
them. No natural lumpy ore is incorporated in their charges. Sintering
and/or pelletizing units of correspondingly large capacities are needed to
feed these furnaces. Such large agglomeration plants have therefore been
designed and installed to cater to these furnaces.
17.3 Coke Quality

These large-capacity furnaces require burden of a very consistent and
better quality so as to control the operation nearly fully automatically
in order to produce metal of equally consistent and better quality.
The quality of coke perhaps plays a more vital role in the running of
these large-capacity furnaces than even the role played by the other
burden materials. This is not to under-rate the importance of burden
quality in the smooth running of these large furnaces. In fact, the dam-
age caused by poor-quality coke may be more severe in large furnaces
than in smaller-size furnaces as shown in Figure 17.2. This is because of
increased degradation of coke close to the tuyeres in the large furnaces
than in the smaller furnaces.
FIGURE 17.2 Schematic diagram showing the influence of inferior quality of coke on the blast furnace
operation. [69]
A poor-quality coke, in its net effect, decreases penetration of the race-

way to the center and relatively increases the slag level in the hearth, both
of which affect the practice adversely.
Coke is the heart of the blast furnace on which the performance ines-
capably depends. Firstly, it is the costliest item amongst all the coke qual-
ity indices (and its behavior in the furnace), and thereby establish control
indices suitable for large furnaces. Alternatively, blast furnace technology
matched to the available coke quality will have to be developed by improv-
ing other factors that influence the production rate.
FIGURE 17.3(a) Interdependence of coke properties and blast furnace performance.
FIGURE 17.3(b) Functions of coke in a blast furnace.
It is the inter-relationships of these factors vis-a-vis their economics that

would dictate how these can be optimally used for better economics of pro-
duction. This has been discussed at length earlier while dealing with coke
preparation. It is, however, certain that the CSR must be, at a minimum, 65 for
better productivity, and for larger furnaces it should be closer to 80. Such good-
quality coke is used by British Steel, Sumitomo, UsinoSA, BHP Steel, and oth-
ers. Similarly, the maximum amount of coal/natural gas/tar may be introduced
through the tuyeres to partially replace the function of coke as heat producer, to
minimize coke consumption without affecting the productivity adversely.
FIGURE 17.4 The effect of coke CSR on BF productivity as seen at Tata Steel.
Ash content of coke has also great bearing on production economy. It

increases the slag volume by assimilating this extra ash and by the extra flux
charged. Ash also results in high temperature breakdown of coke since it
generates additional cracks by differential expansion of carbon and ash par-
ticles on heating. Many offset the lower CSR values by using low ash coke.
In India, coke made from entirely local coals contain as high as 25%–28%
ash. This is somewhat decreased by blending it with imported low-ash coals
to approximately 18%–20%. This is still higher than the 10%–12% ash con-
tent coke used by many well-known overseas steel plants. The productivity
as influenced by the CSR is shown in Figure 17.4. The influence of coke ash
on productivity is shown in Figure 17.5.
The increase in ash content is reflected in many ways. It results in
increased slag volume, which requires higher temperature, and conse-
quently the silicon content of the metal increases. Increased PCI demands
more stringent coke properties.
FIGURE 17.5 The effect of coke ash content on productivity.

It is obvious that in order to cater to the blast furnace, a closer coopera-

tion between coke producer and user is essential and is increasingly being
achieved.
17.4 Burden Preparation

This is the one single parameter in blast furnace practice in which improve-
ments have paid greater dividends than those in any other Burden preparation
is being adopted at every plant, almost without exception, on an ever-increas-
ing scale, the importance being coke manufacture and agglomeration.
Iron-bearing fines are increasingly being agglomerated by adopting
either sintering or pelletization. It was estimated that the production of
sinter was nearly 400 Mt and that of pellets 106 Mt in the year 1968. This
accounted for approximately 50% and 15%, respectively, of the total pig
iron production in the world. It was then estimated that the pellet produc-
tion would reach 200 Mt, which it did, with only a small increase in sinter
production by 1977.
The size of an individual agglomeration unit is progressively increasing
to make the operation more and more economical. Automation of the units
with sophisticated computer control is slowly becoming the order of the
day. The main direction of metallurgical developments in agglomeration
practice is the incorporation of flux on an ever-increasing scale in the sinter.
Production of fluxed pellets is the logical future development in the same
direction. Super-fluxed sinter with basicity of up to four is being contem-
plated to be produced commercially in the Indian plants. Sinter containing
MgO is being recommended to offset the ill-effects of high alumina under
Indian conditions.
Blast furnace operation with high MgO slags has been slowly adopted
over the last decade in the Indian plants to offset the ill effects of high alu-
mina content of the slag. It has posed no problems. The MgO is incorpo-
rated in the burden as dolomite and/or MgO-bearing superfluxed sinter. It
is desirable to operate the blast furnaces with high MgO through high MgO
sinters wherever possible.
Besides the developments in the shaft, straight grate, and the grate-
kiln system, two more new systems are likely to come up in the near future
for firing the pellets. One is the circular-grate system and the other is the
annular-kiln system. At least one circular-grate machine is in operation
at Mexico. However, only pilot plant units of the annular-kiln type are in
operation. It is claimed that there are no engineering difficulties in the con-

struction of such commercial units up to nearly 11 m in diameter.
Attempts to obtain adequate strength of pellets, as in thermal process-
ing, by chemical bonding at low temperatures have been made. The aim of
this work has been to produce an inexpensive plant, a simpler operation,
and lower fuel consumption. Low-sulfur portland cement, slaked lime, et
cetera, are used as binders to develop the strength. The Carbonate Bonding
(C–B) process, the Hydrothermal, and the COBO process are the typical
examples of this category. In the C–B process, slaked lime is used as binder
during balling, and the pellets are hardened in carbon dioxide atmosphere;
the calcium carbonate thus formed develops the strength. In the COBO
process, hardening is carried out at 200° C in an autoclave under 15 kg/cm2
pressure of carbon dioxide to expedite the process. It must be remembered
that feed preparation, balling, and pellet quality requirements are similar
for both the thermal as well as the cold processing systems.
The benefits that may accrue from cold bonding of pellets are tremendous,
and the developments are therefore being watched with greater interest. The
SAIL R and D claims the following benefits of this process of agglomeration:
1. Relatively coarser material can be used.
2. Metallurgical wastes such as blast furnace flue dust, B.O.F. flue dust,
or ultra-fine concentrates and slimes of adequate grade can be used.
3. Self-fluxing pellets can be readily made.
4. Coke breeze can be incorporated and fed to the furnace.
5. Cheaper, since the heat hardening process is eliminated.
Large-scale trials carried out at the Kalinga Iron Works, Barbil by SAIL R
and D are quite encouraging. Use of 360 kg of pellets per metric ton of
hot metal produced resulted in an increase in productivity by 22% and a
decrease in the coke rate by 12%.
17.5 Improved Distribution of Charge

The importance of burden distribution vis-a-vis the production rate has
already been discussed in Chapter 4. Elaborate burden preparation is no
substitute for better distribution of the charge inside the furnace. The
problem of distribution of charge acquires added significance, particularly
if sinter and/or pellets are incorporated in the burden.
Early efforts to improve distribution by using additional rings and

deflectors did not lead to any satisfactory and acceptable design other than
the standard McKee revolving top, which always results in the formation
of trough at the stock-line just below the large bell. The introduction of
high top pressure complicated the requirements a step further. The vari-
able diameter stock-line armor was fairly popular since the early 1960s in
various countries. However, designs in which the large bell is replaced by a
distributor or a chute have come into vogue. The stock-line armor is being
described in this section while the other modifications of the distribution
system will be dealt with while dealing with the high top pressure operation.
Stock-line Armor
This is an improved version of the earlier armor plate meant to protect the
lining exposed to the falling charge and to prevent its premature failure.
The earlier armor plates were firmly secured to the furnace wall, but in the
modified design these are suspended in a semi-rigid fashion from the top. It
consists of several cast steel plates suspended individually in such a way that
they form a cylindrical shape. These are basically of two different types:
1. Cylindrical. Wherein the diameter of the overall armor varies, always
keeping the shape cylindrical, as shown in Figure 17.6.
FIGURE 17.6 Variable stock-line armor—cylindrical design.

FIGURE 17.7 Variable stock-line armor—conical design.
2. Conical. Wherein the diameter of the overall armor varies by making

the shape conical, as shown in Figure 17.7.
In both of these designs, there is sufficient overlap of individual plates to
ensure the required cylindrical or conical shape at any of the expanded or
contracted positions.
In the cylindrical design, the diameter varies between a maximum limit
and a minimum limit. The design is such that the height of armor as a whole
also changes simultaneously, as shown in Figure 17.4. In one of the German
blast furnaces with hearth diameter of 8 m the armor has 18 plates of each
1250 × 2400 × 50 mm size. The settings vary between 5.68 m (18.5 ft) to 6.76 m
(22 ft), that is, the difference between minimum and maximum settings is 1.04
m (4 ft). It permits eleven different settings with an increment of 120 mm.
In the conical design, the angle of an individual plate is varied with
respect to the vertical plane. The range of adjustments is between −15 and
+15 degrees.
In either of these designs, the design of individual plate is such that it some-
what yields to the impact of falling charge and hence the charge practically
slides past the surface of the plates. Since the plates are more or less vertical
in the cylindrical design, the charge falls nearly vertically in the furnace and
the variable settings are responsible to alter the stock-line contour.
Any setting(s) can be selected or it can be synchronized in the automatic
charging control system to obtain the desired contour as well as distribution
in the furnace. This type of armor is being widely used on the continent,
and the advantages claimed are as follows:
1. The big bell diameter need not increase in proportion to the throat
diameter, as the latter increases with increasing capacity. It thereby
decreases many of the problems associated with big bell.
2. The system is reliable and easy to maintain and operate.
3. Since the plates absorb the kinetic energy of the falling particles, the
materials drop practically vertically in the furnace. The operator can
therefore distribute the charge as desired to obtain the required dis-
tribution.
4. The materials of its construction have been so developed that the life
of the armor can be the same as that of the furnace campaign and
hence no mid-campaign changeover or repairs are necessary.
17.6 High Top Pressure

Julian Avery, for the first time in 1934, put forth the idea that the gaseous
reduction in a blast furnace would increase if the static pressure in the fur-
nace is increased by the throttling back of the discharge gas pressure. This
increase in the furnace top gas pressure is known as high top pressure.
Although sound in principle, the idea was not actually put into practice until
1940 because of engineering difficulties. By the late 1940s a few furnaces
in the USA were working satisfactorily at top pressures of 0.7–0.85 kg/cm2
gauge. The application of this much high top pressure was claimed to have
improved the production rate by nearly 11% and that it resulted in coke
savings of approximately 35 kg/t of pig iron, with marked decrease in flue
dust losses. The use of high top pressure presented several engineering dif-
ficulties in the early days and the Russians were the first to overcome many
of these difficulties.
The adoption of high top pressure in a furnace provided with standard
McKee double bell top leads to a differential pressure across the big bell.
The pressure below the bell is high and that above the bell is practically
atmospheric pressure. As a result, the bell would get pressed against its
hopper seat and would not lower down. The big bell can be lowered only
after admitting gas under pressure in the chamber above the big bell (i.e.,
between the two bells) to develop zero pressure differential across the big
bell. Similarly, the small bell could be opened only after withdrawing the
gas pressure from below the small bell, thereby bringing it down to the
normal atmospheric pressure. The adoption of high top pressure therefore
requires a system of pressuring and de-pressuring the big bell hopper often
to maintain the high top pressure while working the furnace in a normal
way.
The scheme of high top pressure application is shown in Figure 17.8. It
includes a throttling device located after the second stage of cleaning
FIGURE 17.8 Scheme of high top pressure operation.
(i.e., after scrubbers). This led to an increase in pressure not only in the
furnace but also in the uptakes, downtakes, dust catcher, and scrubbers. If
an existing furnace is capable of withstanding the increased pressure, then
only the adoption of high top pressure is feasible, otherwise not. It may
be of interest to note here that the Bhilai blast furnaces in India could be
converted to high top pressure (1.0 kg/cm2) readily, whereas the German-
designed Rourkela furnaces were found to be incapable and hence these
are still being run without any high top pressure.
A septum valve is provided just after the wet scrubbers, and part of the
semi-cleaned gas is bypassed to the big bell hopper to obtain zero pressure
differential across the big bell. One of the problems of operating at very
high pressures has been the noise level emitted at the septum valve control
area. This problem has been overcome to some extent by adopting multiple
orifice-type scrubbers for gas cleaning. Their efficiency of cleaning is also

excellent.
High top pressure of the order of 1 kg/cm2 gauge is most commonly
adopted even on the existing old furnaces. The problems and difficulties
multiply with increasing top pressure. Several modified top charging sys-
tems have therefore been designed to obtain reliable sealing while charging
the furnace. The efficiency of sealing and the duration of its life, besides
other less important factors, dictate the value to which the top pressure
could be raised. It has been estimated that the limit of effective top pres-
sure is approximately 3 kg/cm2 gauge for obtaining increased production
rate and decreased coke consumption.
In general, maximum burden preparation is adopted along with high top
pressure application, and it is difficult to assess the benefits of high top
pressure in isolation. However, the benefits claimed for top pressure appli-
cation are as follows:
1. A major benefit that is so obvious is increased production rate because
of increased time of contact of gas and solid as a result of reduced ve-
locity of gases through the furnace. Increased pressure also increases
the reduction rate of oxide. The results are shown in Figure 17.7.
2. The potential improvement in efficiency of utilization of reducing
gas results in fuel savings. Although no quantitative theory to corre-
late the top pressure with fuel rate is known, a marginal decrease in
coke rate with increase in top pressure has generally been observed.
Above the optimum level of production, coke rate tends to increase
with increase in production rate. High top pressure delays this rise in
coke rate even beyond the normal optimum level.
3. More uniform operation with lower and more consistent hot metal
silicon content has been claimed to be the benefit of high top pres-
sure. The Bhilai and Bokaro blast furnaces with high top pressures
have shown a marked control as well as almost 80% decrease in
silicon content of the hot metal because of the application of high
top pressure. The Japanese, who operate their furnaces at very high
top pressures, have attained consistently very low silicon contents.
This is perhaps due to the flexibility of operation that accrues from
the use of high top pressure. Smoother operation has a beneficial
effect on lining life and hence high top pressure lengthens furnace
campaign.
FIGURE 17.9 Production rate and hot metal silicon content as a function of high top pressure. The
solid line indicates the variations in production rates, and the dotted line indicates the hot metal silicon
contents.
4. High top pressure may be applied at constant blowing rate to obtain

fuel saving with lesser increase in output. Alternatively, maximum
blowing rate may be adopted to maximize output at the expense of
fuel-saving.
5. High top pressure markedly decreases channeling and dust losses in
the effluent gases. The load on the gas-cleaning system is thereby de-
creased.
6. High top pressure allows faster driving rates at remarkably low dust
losses in the outgoing gas.
7. The solution loss reaction is stifled because of rise in pressure. In other
words, the Boudouard equilibrium moves to the left because of high
top pressure and may consequently have beneficial effect on coke
consumption.
17.7 Charging Devices for High Top Pressure

In spite of the developments in successful throttling and pressurizing the big
bell hopper, the conventional McKee-type two bell charging system poses
serious problems of gas leakage through the joints of bells and improper
sealing. Leakage here causes abrasive cutting and soon destroys the seal-
ing function of the big bell with its hopper. This not only restricts the top
pressure but may even, in due course of time, nullify the benefits gained
through the high top pressure operation. The blast furnace designers should
provide equipment to operate effectively at top pressure up to 3 kg/cm2
gauge with adequate life. Many devices, including hard facing of the bell
and its hopper set, have been mooted and tried in practice. However, the
problem of excessive wear on seats and rapid deterioration of effective seal
could not be satisfactorily solved. The developments have been rapid after
adopting the design philosophy of separating the two functions of charge
holding and distribution on the one hand and gas sealing on the other. In
essence it means that hard-surfaced seats are used for bells, gates, chutes,
and others, which handle flowing materials, whereas soft seats, kept out of
way of flowing materials, are used as seals. Based on the above philosophy,
a number of designs for top charging, to whatever extent feasible, have
been developed. The development of top charging devices, while catering
to high top pressure maintenance, have always kept the goal of achievement
of better distribution of charge in the furnace, as was its primary aim earlier,
as an equally important function. A few of these modern designs, especially
developed for high top pressures, have been described below.
Tops with Only Bells

In these designs, the functions of distribution of charge and assurance of
pressure-tight gas seals are both achieved by the use of only bells. The con-
ventional McKee top design is improved while retaining the concept of
a revolving distributor to distribute the charge more evenly. The typical
examples are the Paul-Wurth/CRM top and the NKK’s four bell top.
Paul-Wurth/CRM Top
A line sketch of this design is shown in Figure 17.10. The big bell arrange-
ment is the same as in the McKee top while an additional small bell, called
the seal bell, is attached just below the charge small bell. Since the sealing
bell is different, the small bell along with its hopper can rotate to receive
the charge at various points along its periphery. The sealing bell fits tightly
to the large bell chamber and hence the small charge bell is outside the
purview of high top pressure. In operating this device, the charge is initially
put in the small bell hopper, receiving the skips at various preset angles.
During charging, the seal bell is closed. After one charge is put in the small
bell hopper, the seal bell is lowered to break the seal. The charge small
bell is then lowered to drop the charge on the big bell. The distance to
which the burden small bell is lowered, from the seal bell, is adjusted such
that the sealing surface remains out of the trajectory of sliding material,
and it is thereby protected. After dropping the charge, the small bell is
raised to close its hopper and then the seal bell is raised to seal the big bell
chamber, the pressurization and de-pressurization of which is carried out
as described earlier.
FIGURE 17.10 Paul-Wurth/CRM top.
NKK’s Four Bell Top

A line sketch of this design is shown in Figure 17.11. In this, two additional
bells are interposed between the rotating small bell and the large bell. The
middle bell holds the charge and also seals the large bell hopper for high
top pressure. The small bell carries only the seal, and because of this, the
large bell chamber is always under pressure and hence its wear is relatively
much less. The CRM large bell, for a certain proportion of time, is under
full furnace pressure, which eventually can cause severe wear, but such is
not the case in the NKK four bell design.
Both the CRM and the NKK top designs are such that the charge par-
ticles may find a way into the sealing area, causing some gaps through which
high-pressure gas leaks, with resultant wear. These designs therefore cater
to equipments for rapid changing of the bells.
FIGURE 17.11 NKK’s four bell top.
Tops with Bells and Valve Seals

The IHI-top as shown in Figure 17.12, the Yawata-top as shown in
Figure 17.13, the McKee Head Wrights on top as shown in Figure 17.14,
and the Demag-top as shown in Figure 17.15 are the best known design in
this group. Valve seals are provided to obtain effective seals. These designs
can be adopted on even the existing furnaces with a minimum of disturbance
to the large bell, top structure, gas mains, and other existing parts of the fur-
nace top structure. The seals are so provided that the large bell chamber
can operate permanently pressurized to that of approximately the furnace
pressure, thereby minimizing the wear on the large bell seatings. The gas
seal valves swing down and out of the way while dropping the charge, and
then quickly return to close and form a gas-tight housing around the bell.
FIGURE 17.12 IHI-top.
FIGURE 17.13 Yawata top.

In the IHI top, two hoppers with seals are provided to admit the
charge. The revolving distributor is in the pressurized zone and distributes
the change on the small bell. Contrary to this design, charge is fed through
a rotating chute, outside the pressurized chamber, via four hoppers with
valve seals in the Yawata design.
In the McKee Head Wrightson design, a six-position chute feeds six
valve seal inlets to distribute the charge evenly on the small bell, which
does not rotate. In the Demag design, the small bell is replaced by a rotat-
ing chute under the lower seal. This chute distributes the material directly
on the large bell. The central rod for raising and lowering the large bell is
replaced by a yoke operated from outside the pressurized chamber. Two
valve seal hoppers receive the charge and feed it through another pair of
lower seals into the rotating chute.
All these designs, even though they improve the gas sealing, do not
eliminate the drawbacks of distribution arising out of the use of a large bell.
FIGURE 17.14 McKee Head Wrightson top.

FIGURE 17.15 Demag top.
Bell Construction and Operation

The top design, the materials of construction of particularly the bells, the
seals, etc. should be such that full furnace campaign is obtained without
major intermediate repairs and replacements. In particular, the large bell
and its seat should at least last for the whole campaign since with increasing
size of the furnace, the problems associated with its manufacture, instal-
lation, and maintenance are escalating rapidly. On a modern furnace with
hearth diameter of 14 m, the large bell diameter might be approximately
7.5 m and it would weigh nearly 200 t. Such a unit is lifted into position and
if intermediate repairs, during the furnace campaign, are required, special
hoists are required to lift such a bell at nearly 80 m above the ground level,
let alone the actual repair problems. Hence the design as a whole should be
sound enough to last as long as possible, the objective being full campaign
of the furnace. The latest trend therefore is to dispense with the bells alto-
gether or use smaller bells and additional mechanical devices to distribute
the charge uniformly in the furnace at high top pressure with a minimum of
maintenance problems. These are discussed in the following section.
FIGURE 17.16 Paul-Wurth bell-less top.
Top with Only Valve Seals (No Bells)

This is a unique design in which the large bell is replaced by a distributor
chute that has no sealing function whatsoever. The problems of distribution
associated with large bell are entirely eliminated. A rotating chute is provided
inside the furnace top cone. All the materials are charged via holding hoppers,
with seals at its top and bottom that are charged and discharged alternately,
while the third is acting as a spare. Regulating gates in each hopper are pro-
vided to control the rate of charging to facilitate uniform distribution on the
stock-line. A typical design, known as Paul-Wurth bell-less top, is shown in
Figure 17.16.
Since the chute can rotate in a circular or helical fashion and at a vari-
able angle and hence, not only are the fundamental characteristics of large
bell distribution retained but charging at a point (stationary), in sequence
and in spiral form, is possible. It means that virtually any type of distribu-
tion as desired can be obtained.
The advantages claimed are:

1. With three bunkers in the design, the performance of this top has
been at least equally satisfactory and comparable to any of the avail-
able alternatives.
2. There is greater charge distribution flexibility with a small amount of
mechanical equipment.
3. Access to any part of the system is far easier and hence one or two
parts can be changed even during normal shutdown of the furnace.
4. Wearing parts are rather few and inexpensive and hence these can be
regularly changed during routine maintenance.
5. The total overall height of the top can be much less.
6. There is substantial reduction in investment for top construction.
7. It gives more operational safety and easy control over varying charging
patterns.
Even if the whole bunker is to be replaced during the furnace campaign,
the total maximum lift is reduced to only 30 t, resulting in reduced require-
ments in terms of hoisting facilities.
Gimble Top Distributor

This is the latest innovation in charging devices employed for blast furnaces.
It is claimed to be superior to the Paul-Wurth Bell-less Top, which dominated
for nearly 2–3 decades, superseding the earlier McKee Two-Bell charging
device. It was installed in India for the first time on BF-C of Tata Steel while
renovating it, fairly drastically, in 2007. This will be the first-ever application
for the blast furnace. It is costlier than the bell-less top. The design is being
supplied by Seimens VAI. It has already proved its worth under the arduous
environment of Corex, Finex, and direct-reduction plants.
It has a conical charge distribution device supported by rings in a gimble
arrangement. It consists of several robust rings and a chute. This tilting
chute is driven by hydraulic cylinders, operating through shafts, connecting
rods, and universal joints in order to drive the gimble rings. This allows any
angle of the chute to distribute the charge as per any specific requirements
at any point in the stock-line surface. This gives very precise material distri-
bution and therefore provides almost innumerable charging patterns to be
achieved at varying speeds. In fact, the charge particles are literally placed
wherever required, on the stock-line, by this device.
17.8 Higher Blast Temperature and Driving Rate

Since part of the fuel energy in the outgoing gases is pumped back into
the furnace in the form of pre-heated blast, the thermal efficiency of the
process could be improved by increasing the rate of blowing of hot blast as
well as its temperature. The average hot blast temperature of 600°C pre-
vailing in the 1950s was therefore raised to approximately 900°C by the late
1960s, and now 1200°C–1300°C is the level at which hot blast is blown in
some of the most modern furnaces. This requirement led to improvements
in all aspects of stove design to store more energy at high temperatures.
In other words, it necessitated investigations into the heat transfer pro-
cesses vis-a-vis checker shapes and materials and as a result, high-efficiency
checker shapes have been developed. Since the dome is a vulnerable part of
the stove, its general design underwent several improvements. Some of the
typical improved designs are shown in Figure 17.17. In one, the combus-
tion chamber is located outside the main stove body so that the entire stove
volume is available to build checkers and thereby store more energy. In the
other, the combustion chamber is located in the dome itself by altering the
dome shape to hold the burner. The Germans have developed the vulkan
burner, which is far more efficient in utilizing the fuel and can be used
FIGURE 17.17 Improved hot blast stove designs.

in the existing designs or in the new stoves with external combustion cham-
ber or at the top of the stove as well. The burner does not get affected
by the radiant heat during the on-blast period since it is made of ceramic
materials.
All these have led to dome temperature reaching a value of
1500°C–1600°C, at its peak, when on-gas, and dropping down to 1450°C
at the end of the on-blast period. The required refractory material has also
been developed. All these have improved the stove efficiency to almost 90%.
A recuperator system on the stove has been used to pre-heat the gas
and combustion air to generate higher temperatures while side-by-side
improving its efficiency. The stove waste gases are used to pre-heat the gas
and combustion air to almost 450°C.
Higher hot blast temperature reduces proportionately the fuel con-
sumption in the furnace. A higher hot blast temperature is often used in
conjunction with humidified blast so that flame temperature in the com-
bustion zone of the blast furnace is still within proper limits. This will be
further discussed while dealing with the humidification of the blast.
The latest technology to come on the horizon of hot blast generation is
the Kalugin Stoveto be installed at Sunflag Iron and Steel Co, Bhandara.
This is a shaftless-combustion stove design, fired from the top as shown in
Figure 17.15 (c). It gives a blast temperature of 1250°C readily.
This much hot blast temperature is capable of generating 2400°C –2500°C
as RAFT temperature, which is too high to be used. It also means that it can
take up readily 150–200 kg/thm of PCI or 100–150 Nm3 of natural gas injec-
tion along with 3%–5% oxygen enrichment and 5%–10% humidification of
the blast. This combination of high blast temperature, oxygen enrichment,
and humidification again brings down the RAFT to a normal 1900°C–2000°C.
The same is achieved by using PCI, which decreases the actual coke rate. It
also allows use of oil and natural gas injection together in various proportions,
which can lead to a coke replacement ratio of as high as 1.6–1.7 kg/kg, which
otherwise cannot be achieved by any other single fuel injection.
17.9 Oxygen Enrichment of Blast

For every unit weight of coke burnt at the tuyere by the air blast, nearly
4–5 unit weights of nitrogen of the blast are also heated to nearly 2000°C.
Although large amount of furnace gases are beneficial for heat transfer in
the stack, the presence of 79% nitrogen by volume in the blast restricts
the temperature generated in the combustion zone. This temperature can
be increased by decreasing the nitrogen content of the blast, that is, by
oxygen enrichment of the blast. An enrichment of only 2% (by weight) oxy-
gen reduces the nitrogen burden by approximately 4 units per unit weight
of coke, and a higher temperature would be possible. There is, however,
a limit to which higher temperature in front of tuyeres is desirable since
any excess over that causes bridging and sticking of the stock and a higher
silicon content in the pig iron. Oxygen enrichment is therefore beneficial
for producing foundry iron or for smelting ferro-alloys such as ferro-man-
ganese. For basic pig iron production, excess silicon content is detrimental
and hence any excessive heat release at the tuyeres must be absorbed by
some other beneficial endothermic reaction. Use of oxygen enrichment up
to 25% oxygen in the blast has been found to be advantageous, if it is bal-
anced by adequate humidification. A higher level of oxygen enrichment
poses several other problems that weigh more than the advantages and
hence is not used in practice.
FIGURE 17.18 Effect of oxygen enrichment of blast on furnace productivity.
Combined use of oxygen enrichment and humidification of blast offers

a unique method of blast furnace process control. Control of the blast
furnace by coke rate is not effective before a few hours after the change,
whereas variation in oxygen enrichment and humidification can control the
temperature almost immediately; it can also be monitored continuously.
The cost of oxygen, however, should permit its use in a given setup, for it
may otherwise work out to be costly.
For every percent increase in oxygen content, an increase in production
rate of approximately 3%–4% can be obtained with a marginal saving in
coke rate. The saving in coke rate is due to the cracking of moisture giving
rise to hydrogen, which acts as a reducing gas up in the stack. The increase
in output with oxygen enrichment is shown in Figure 17.16. It also shows
that the percent returns decreases with increasing enrichment.
17.10 Humidification of Blast

A uniform and steady RAFT is one of the basic pre-requisites for smooth
blast furnace operation. RAFT is sensitive to the moisture content of the
blast, and moisture varies from season to season, being maximum in rainy
season and minimum in dry summer. Nearly five times by weight air is
required for burning coke efficiently. Pre-heating of blast not only adds
heat energy in the furnace but it increases the RAFT. There is a limit to
which RAFT can be allowed to increase, depending upon the burden char-
acteristics and furnace profile, and beyond which it is detrimental to fur-
nace operation. Blast temperature can be increased still further without
increase in RAFT if equivalent coolants are added along with the blast.
Steam is one such corrective addition because of its endothermic reaction
with carbon as:
H2O + C = CO + H2
H°(1200°C) = +2700 kcal/kg C
= +1800 kcal/kg H2O (17.1)
It means that, for other conditions remaining the same, the RAFT will
be much less in the rainy season as compared to that on a dry summer
day. As a consequence, the blast furnace operation will be subject to sea-
son. This is not acceptable in modern times, as it would affect the produc-
tivity adversely. A uniform RAFT in this case would be obtained only by
adjusting every now and then the blast pre-heat temperature. This is next
to impossible.
There is a brighter side of this phenomenon. Presence of moisture

in the blast generates double the volume of reducing gas per mole of
carbon burnt. As per Equation (17.1) for every carbon burned, one mole
of CO and an additional mole of hydrogen will be available as product of
burning of coke for reduction in bosh and stack. The more the moisture,
the more of this additional hydrogen will be available. Kinetically, hydro-
gen reduction of iron oxide is faster than that by CO because of its small
size. Presence of moisture helps to burn coke at a faster rate with its
attendant favorable effects. Some of the endothermic heat of moisture
disintegration is compensated by way of exothermic reduction of iron
oxide by hydrogen.
It has been estimated that for an increase of 20 g/Nm3 moisture in the
blast, the endothermicity can be compensated by a rise of 200°C in the blast
pre-heat. This ratio can be used as a rule of thumb for further moisture
addition. Initially, it was considered enough to maintain the moisture level
in the blast at the maximum that occurs in the atmospheric air any time
round the year, but it soon became clear that it can be increased still further
to take advantage of the beneficial effects of steam addition. By increasing
moisture and compensating for it by additional rise of pre-heat means that
cheaper heat energy can be used to feed the furnace and thereby decrease
the coke consumption and economize the operation.
Oxygen enrichment of the blast and moisture enrichment have quite
opposite thermal effects. The two can be saddled together to obtain better
inputs. Hot blast temperature, extent of oxygen enrichment, and humidi-
fication of blast have to be adjusted as interrelated parameters simultane-
ously to obtain optimum conditions of operation for maximum benefits
such as minimum coke rate, higher productivity, and so on.
For humidification of blast, steam is introduced in the cold blast before
it is pre-heated in the stoves. If it is introduced in the hot blast, since steam
temperature will be much less as compared to that of blast, it will have a
cooling effect, which is not desirable.
Steam is also also cheap and requires energy for its generation. It is
therefore worthwhile to introduce water in the blast in place of steam.
These trials were taken in Tata Steel in 1996. This is more endothermic
in nature than introducing steam because additional energy is required
to vaporize water to steam in this case. In their “C” furnace, water was
to be injected at the rate of 800–3600 liters/hr at a pressure of 2 kg/cm2.
With water injection in place of steam, the adjustments in blast tem-

perature, oxygen enrichment, and water will be more tricky for getting
optimum RAFT.
For humidification, steam is introduced in the cold blast before it is
pre-heated in the stoves. If it is introduced in the hot blast, since the steam
temperature will not be as close to that of the hot blast, it will have a cool-
ing effect, which is not desirable. Humidification eliminates the day-to-day
changes in humidity level and thereby eliminates one major variable that
affects blast furnace operation markedly.
17.11 Fuel Injection

The necessity to adopt fuel injection in a blast furnace arises from the fact
that coke is not only costly but it is becoming more and more scarce and
hence it should be replaced by other cheaper and readily available fuels,
as far as is feasible, to run blast furnaces without impairing their efficiency.
Also, all coke furnace operation is affected by a higher flame temperature
and formation of an inactive zone near the furnace walls. Auxiliary fuel
injection tends to set this right and leads to a more stable furnace operation
because of additional control parameters available to control characteristics
of the raceway. The heat-producing function of coke is partially replaced
by injecting auxiliary fuels in the tuyeres. They are therefore readily avail-
able for combustion in the combustion zone in front of the tuyeres. Either
solid, liquid, or gaseous fuels can be injected in the tuyeres. Solid fuels
are coal, either pulverized or in the form of slurry. Liquid fuels are either
light oils, naphtha, or more predominantly heavy oil. Natural gas or coke
oven are the gaseous fuels that could be readily injected in the tuyeres.
The choice of the type of fuel for injection almost entirely depends on its
availability, economy, and feasibility of injection. Oxygen enrichment of
the blast along with fuel injection is required to achieve full advantages of
fuel injection.[78]
It has been discussed already in Chapter 4 that coke has two major
functions inside the BF. One is as a fuel and the other is as a spacer to
maintain permeable charge, which allows the gases to pass through it
smoothly. The function of coke as a fuel is therefore minimized, to the
extent possible, by using substitute pulverized, liquid, or gaseous fuels.
Fuel is burned in front of the tuyeres to generate the required thermal
energy and to compensate for that lost in decreasing the coke charge.
The substitution of coke is also done at this level only by injecting pul-
verized/granular coals or gaseous/liquid fuels such as natural gas/heavy
oil through the tuyeres. This is one of the most potent ways of reduc-
ing coke and replacing it with cheaper and available fuel. The basic
requirement is that the injected fuel must burn completely and to that
extent only the replacement is possible. Any excess injection can cause
damage rather than benefits by disturbing the bosh reactions and melt-
ing rate.
All coke operation is affected by higher flame temperature and forma-
tion of an inactive zone near the furnace walls. Auxiliary fuel injection tends
to set this right and leads to a more stable furnace operation because of
available additional control parameters to control the raceway.
The choice of the type of fuel to be injected almost entirely depends
on its availability, economics of its use, and feasibility of injection. From
this point of view, only fluidized fuels can be injected. Fuel injection is fully
exploited when adopted along with oxygen enrichment and humidification
of the blast.
The actual amount of fuel injected in any furnace in any plant depends
on many factors. On some furnaces individually up to 100 kg oil or 200 kg
pulverized coal or 150 Nm3 natural gas have been successfully injected, and
the efficiency has come up to expectations. The replacement ratio, kg of
coke saved divided by kg of fuel injected, depends on the practice, quality
of fuel, and effectiveness of furnace control.
These injections affect the following furnace parameters:
1. Flame temperature
2. Bosh gas composition and volume
3. The top gas temperature and CO/CO2 ratio and hence the efficiency
4. Operational efficiency, that is, the proportion of direct and indirect
reduction
5. Slag volume and its basicity
6. Overall production efficiency
The influence of such injectants is shown in Table 17.4
Table 17.4 Effect of various injectants on BF conditions

Parameter Change Change in Flame Change in Coke Rate
Temp. (°C) (kg/thm)
Natural gas +100 kg/thm −513 −82.6
(+132 Nm3/thm)
Anthracite coal +100 kg/thm −162 −92.0
Heavy volatile coal +100 kg/thm −218 −76.1
Heavy fuel oil +100 kg/thm −321 −98.4
The demand on heat is a net result of heat generated by its burning and
consumed for ash dissolution in slag and for any moisture, taking part in
the water gas reaction. Similarly, the net resultant gases at the tuyere shall
take part in additional reactions as they rise. It is necessary to compensate
this loss in flame temperature by increasing the hot blast temperature, and
if that limit is already reached then, by other means such as oxygen enrich-
ment, blast temperature, and so on.
More efficient reducing gas is generated at the tuyere area by disinte-
grating moisture into hydrogen via a reaction of the type:
H2O + C = H2 + CO (17.2)
or by
CH4 + 1/202 = 2H2 + CO (17.3)
These are endothermic reactions and are possible only when surplus
heat is available at the tuyere area.
The replacement ratio for oil is of the order of 0.9–1.1 kg of coke per kg
of oil. For natural gas, this ratio is 0.75 kg per Nm3 of gas. These are for
blast temperature of 900°C–1300°C. Natural gas or oil generates hydro-
gen, as a product of burning, which is a better reducing agent in the stack
as compared to CO. The H2/H2O ratio in the top gas does not indicate its
utilization efficiency because only 4% H2 is used for reduction. It therefore
improves the productivity. Natural gas does not contribute to slag forma-
tion and the replacement ratio is also good and, hence is preferred. It has
resulted in 6%–10% saving in coke and increase in productivity by approxi-
mately 0.5%–2.0% in a specific case.
Pulverized or granular thermal coal has also been routinely injected
in the range of 50–200 kg/thm. The aim is to go up to 300 kg/thm in the
near future. The cost of coke can be brought down by 10%–15%. If pul-
verized, then 80% of coal is ground to less than 75 micron and dried to
a level of less than 1% moisture. The system of injection should ensure
that the desired amount of coal is injected, within expected accuracy. It
is claimed that the performance of injecting even granular coals (sandy
in nature) is adequate. It is preferred since the cost of pulverization is
far more than that of granulation. However, pulverized coal injection is
more popular.
The limit of injecting PC is dictated by a parameter called Coal
Combustion Potential (CCP), which is defined as
Oxygen supplied to raceway (Nm3/hr)

CCP =
Oxygen required to burn injected coal (kg/hr)
The ideal theoretical value of CCP is 1.0. This consumption of oxy-

gen takes into account the humidification of the blast as well as the oxygen
enrichment, since all these go hand-in-hand to obtain the required condi-
tions. The reduction of coke rate by various such parameters is assessed
beforehand and is shown in Table 17.5.
Table 17.5 Assumed norms for calculating the changes in coke rates
Parameters Change in Coke Rate (kg/thm)
1. Sinter +10% −2.2
2. Blast temp +1°C −0.14
3. 2 −11.55
High-Top Pressure +1kg/cm
4. Central Working Index +1 −54.34
5. 3 −0.88
Humidity +1 g/Nm
6. Coke ash +1% +10.00
7. Oxygen enrichment +1.0% −6.82
This is quite independent of the efforts to raise the hot blast tempera-
ture to its maximum.
The replacement ratio (RR) of the order of 0.8–0.9 has been achieved.
Ash content of the coal affects the RR adversely, as will be obvious from
Table 17.4. The efficiency of replacement is not just limited to its com-
plete burning. It should also thereby result in improvements in bosh and
stack reactions as well as heat transfer and rate of melting. One of the chief
problem to be faced in PCI is its effect on deadman. In case of incomplete
combustion, the PCI can give rise to suit or char formation and may
decrease the permeability of the deadman. This can have disastrous effects
on hearth behavior.
Pulverized Coal Injection

Injection of gaseous or liquid fuels through the blast furnace tuyeres for
part replacement of coke was relatively easier from the point of view of
engineering involved therein. But the success at these led to attempt injec-
tion of coals, as these are readily and cheaply available at most places. Coal
is injected through the tuyeres in cold condition (non pre-heated). It is
therefore bound to influence the flame condition in front of the tuyeres.
Almost all modern blast furnaces all over the world use coal injection as a
matter of routine practice. The best performance has been to inject 100–
200 kg/thm with resultant coke rate falling to as low as 300 kg/thm. Tata
Steel commissioned coal injection unit on their “F” furnace in March, 1991.
This system is capable of injecting 150 kg/thm.
The trials have indicated that the rate of coal injection is influenced by a
variety of parameters that have to be adjusted to maintain or improve blast
furnace performance.
Coal injection in the existing design increases the tuyere velocity
beyond the optimum 205 m/sec and as a result, the tuyere diameter had to
be increased from 125 mm diameter to 150 mm diameter to obtain opti-
mum velocity.
Injection rate was influenced by blast pressure and blast volume. For
pressure below 10 psi, coal could not be injected. For pressure in the range
of 10–15 psi, injection rate was halved, and for pressure above 15 psi, full
injection was possible. This practice eliminates the possibility of tuyere
blockage. Above 15 psi pressure change in blast volume in percentage was
matched by the same percentage change in the injection rate for better
performance. Necessary changes had to be incorporated to obtain uni-
form injection through all the tuyeres. The injecting lance angle had to be
adjusted to 11° C for efficient operation of the lance.
The coal quality has a dominant influence on the furnace performance.
Maximum combustion of coal in the raceway is desirable. The ash content
should be low so that no deposition takes place in the lance or the tuyere
region. The melting point of the ash should be as high as possible but not
that much so as to influence the melting characteristics of the slag inside the
furnace. Since coal is injected in cold condition, it has a cooling effect on

the raceway. Increasing the injecting rate has increasing cooling effect and
unless counteracted it can have an adverse effect on furnace performance.
Usually a high injection rate, in this respect, is compensated for by a propor-
tionate increase in oxygen enrichment of the blast to obtain optimum RAFT.
The size of coal injected does influence the furnace performance. This
can be studied by experimenting on coal combustion. It was found at Tata
Steel that with 80% coal of −200 mesh poses no problems with respect to
its size. Although a very wide range of volatile matter coal has been studied
and found to be giving good results, it is better to have V.M. in the range of
20%–30%.
The calorific values also have an influence on RAFT. Coals with C.V. in
the range of 6000–7000 Kcal/kg have been found to pose no specific prob-
lems. The burden charging sequence had to be adjusted suitably to suit coal
injection.
For normal practice, nearly 50 kg/thm coal could be injected with
replacement ratio of 0.9. With 1.4% oxygen enrichment, the same could
go up to 70 kg/thm with replacement ratio of 1.08. In 1994, coal injection
was fitted on Tata Steel‘s “D” furnace and injection rate of 100 kg/thm with
1.5% oxygen enrichment was possible for replacement ratio of 1.0. Their
“G” furnace is the latest and most modern. There, 130 kg/thm could be
injected with 3.0%–3.5% oxygen enrichment with a replacement ratio of
1.3. The coke rate came down to 450–500 kg/thm.
In all these the RAFT was maintained at approximately 1900°C. The
productivity remained unchanged. As regards quality of hot metal, sulfur
remained unaltered but silicon went up from 1.0 to 1.36, which had to be
brought to 1.0 by oxygen enrichment.
Tata Steel has provided coal injection facilities on their “D”, “F”, and
“G” furnaces. Whereas on “A” and “B”, tar injection facilities have been
provided.
17.12 Lime Injection

The origin of lime injection lies in controlling the quality of bosh slag. The
quality of slag produced in the bosh region has direct bearing on the productiv-
ity of the furnace. If this slag is more viscous, then it affects the acro-dynamics
adversely. As a result, the productivity falls. Increase of bosh-slag viscosity by

even 1 poise may decrease the productivity by approximately 0.5%.
Flux is added in the BF charge to flux the gangue of the iron ore/sin-
ter and the ash of the coke. The entire flux requirement is met by adding
limestone as a separate constituent in the charge or via sinter in part or
in full. Where coke ash is very high, such as that of in India, the bulk of
the slag is formed out of the ash of the coke. A major portion of this ash is
released only after the combustion of coke in the tuyere region. It means
that the lime proportion of the slag formed before the bosh region is far
too large than is required to maintain the desired basicity. In other words,
the basicity of the slag that is formed before the bosh region is very high
and consequently its viscosity will be very high. Such highly viscous slag
poses several problems. Firstly, it trickles slowly through the coke bed.
Secondly, it interferes with the aerodynamics adversely and decreases the
rate of upward gas flow and thereby reduces the rate of reduction of ore
and rate of heating of the charge in the stack. It means, directly, a decrease
in the productivity.
In fact, the lime required to flux the ash of coke is ideally required at
the tuyere level. If it can be provided at that region, then the slag formed
up in the stack and in bosh will be almost the same type of slag as aimed at
the end. This will not be more viscous than necessary and hence will readily
trickle through the coke bed.
High ash content affects the BF operation adversely, as can be seen in
Table 17.6.
Table 17.6 Effect of variation of ash content of coke

Coke % Fixed Ash Limestone Rate Coke Rate Slag Rate Production Rate
Carbon (kg/thm) (kg/thm) (kg/thm) (t/day)
76 22 416 874 510 1346
72–58 26 497 972 610 1276
68–58 30 600 974 737 1197
All such slag-related problems can be solved to a great extent if part

of the total lime requirement can be injected through the tuyeres in a way
similar to fuel injection. It will be available where it is needed most, that is,
to flux ash of coke. If this is resorted to, then part of the flux is added, more
often via sinter and the rest via injection through the tuyeres. In that case,
no free limestone may be required in the charge material. It is the pow-

dered lime that is injected through the tuyeres and not limestone powder.
It means the heat requirement for its calcinations is met outside the BF
using any alternative fuel rather than coke. The coke rate would thus also
decrease proportionally.
FIGURE 17.19 Blast furnace operation with auxiliary fuels injected at the tuyeres with fully oxidized
burden.
Full-scale industrial trials on a small blast furnace of the Kalinga Iron

Works, Barbil, under the direct supervision of SAIL, R and D Division,
were carried out and have borne very good results. The same has therefore
been recommended to be applied for normal blast furnaces. The advan-
tages claimed are:
1. It improved the bosh permeability due to improvement in mass flow
rate in the bosh region.
2. It decreased the viscosity of the slag in the bosh region with all the at-
tendant advantages.
3. It decreased the metallurgical coke requirement in the blast furnace

due to decrease in limestone-flux charged from the top.
4. It increased the productivity of the furnace.
5. It controlled the basicity of slag and thereby improved the quality of
hot metal.
6. The slag rate is marginally reduced.
7. Metal sulfur content was somewhat lower.
8. On the whole, the furnace operation was smoother with decreased
hanging and slipping.
It is, however, surprising that it was not adopted on large furnaces even in
India where the coke ash problem was very seriously affecting the produc-
tivity. In fact, it is now recommended for its adoption in MBFs to decrease
their coke requirements and improved productivities.
17.13 Pre-reduced Ore as Burden

The underlying principle is to decrease the requirement of indirect reduc-
tion of ore in the stack. It means decreasing the generation of reducing gas
at the tuyere area. The indirect reduction is to be carried out outside the
furnace using cheaper fuels such as thermal coals. In a way, it is a combina-
tion of DRI production with BF. Although it is not possible to completely
charge pre-reduced burden in the blast furnace and eliminate the indirect
reduction, partial replacement of ore by the pre-reduced feed has a distinct
advantage. It has been occasionally tried and has resulted in full benefits.
The economics are, however, precarious. This is the reason why it is still not
used as a regular practice.
Adoption of fuel injection, humidification, oxygen enrichment, higher
blast temperatures, high top pressure, and fullest preparation and proper
distribution of burden, et cetera, can be carried out to a certain extent
only and hence these, therefore, exhaust the possibilities of improvement
of blast furnace technology. Further improvement in its productivity can,
however, be achieved by the use of pre-reduced ore as the burden material.
The trials on experimental blast furnaces using pre-reduced iron ore have
borne spectacular results. These have also been computed and are shown
in Figure 17.20.
FIGURE 17.20 Blast furnace operation with pre-reduced ores as burden and auxiliary fuels injected at
the tuyeres.
Theoretical coke rate for 100% pre-reduced burden comes to as low

as 150 kg/t of pig iron. Such a low amount of coke is too inadequate to
ensure the minimum required permeability of the charge in the furnace.
For pre-reduced burden, the loss of coke through solution loss reaction is
minimum. It means that the coke arrives in the fusion zone almost intact.
Under these conditions, using 30%–50% pre-reduced burden coke rate
could be decreased to 250–300 kg/t of pig iron. This much coke rate is
adequate to maintain required permeability of the charge in the furnace.
Use of this much pre-reduced burden can therefore raise the production
rate by approximately 25% and reduce the coke rate as well, over and above
that obtainable by other innovations as mentioned above.
This has been tried in commercial blast furnaces with very encouraging
results in increasing the productivity. This technique may be adopted in an
emergency when more hot metal is required or if the economics otherwise
permits it.
FIGURE 17.21 Blast furnace operation with fully oxidized burden and reducing gas injected in the shaft.
17.14 Injection of Reducing Gas at the Bottom of the Stack

Since the furnace gases pass through the stack at tremendous speed, the
efficiency of indirect reduction in the stack is limited. This is improved
to an appreciable extent by adopting high top pressure. It could further
be improved by injecting reducing gas at the bottom of the stack (i.e., at
the bottom of the isothermal zone in the furnace). The proposed gas must
therefore contain carbon monoxide and hydrogen only, and it should be at
a temperature of approximately 1000°C. This will be in addition to the fuel
injection that may be adopted through the tuyeres. It has been estimated
that the coke rate can thus be brought down to 250 kg/t of pig iron using
fully oxidized burden. The problem of burden permeability now disappears
as the coke, in this case, is expected to descend into the tuyere zone nearly
intact. Since reducing gas is injected, the solution loss reaction is no longer
indispensable (to regenerate the reducing gas) and as a result, the solution
loss of coke may not be more than 25 kg/t of pig iron produced.
The blast furnace in this way becomes a two-fold counterflow reactor.
The theoretical performance of the furnace as computed is shown in Figure
17.19. Within this framework, the output could be increased by approxi-
mately 50% in comparison to conventional practice, and coke rate reduced
to approximately 250 kg/t of pig iron.
17.15 Utilization of Plant Iron-Bearing Wastes

In the good old days, the wastes generated by the plant at various places
were dumped as wastes. This was because good-quality lumpy ore was
available for charging into the BF. As the size of BF increased and as a
consequence, more and more iron ore fines were generated, and the need
to use them as BF feed became necessary. This led to the development of
sintering. and today a sinter plant is an inalienable part of ironmaking.
A steel plant also generates the following wastes bearing iron oxide:
1. Dust generated from ore handling or blending inside the plant
2. Dust blown out of BF with exhaust gases and collected through elabo-
rate gas cleaning system
3. Dust collected in dry condition or as slurry while fine cleaning the BF gas
4. Fines generated from last-minute screening of lumpy ore and sinter
at the BF
5. Dust collected in gas cleaning system of steelmaking unit (EAF/BOF)

6. Sponge iron sludge generated in DRI plant
7. Sponge iron dust spread around
8. Steelmaking slags containing high percentage of iron
9. Dumps created historically by dumping of wastes
The successful development of the process of sintering, as an adjunct of
the BF unit, has now led to utilization of all such iron-bearing wastes by
mixing them in appropriate proportion in the sinter charge-mix. Since this
definitely improves the economics of ironmaking, it is increasingly being
adopted to the extent possible without compromising on the economics of
sintering. A typical material flow in a sintering plant in an integrated steel
plant is shown in Figure 17.22 as an illustration.
FIGURE 17.22 A typical recirculation of iron-bearing waste materials in a steel plant.

The BF technology for ironmaking has undergone innumerable improve-
ments with respect to furnace design, charging arrangement, burden
preparation, utilization of plant wastes, off-gas cleaning, and its utilization,
by-product disposal, hot blast generation, higher hot blast temperature up
to 1250°C, auxiliary mechanical and electrical equipment, on-line process
control, slag and metal tap holes, the cast-house design, and almost anything
and everything related to BF design and operation to such an extent that it
has almost reached its pinnacle in relation to its efficiency and economics.
The majority of the burden is artificially agglomerated as fluxed or
superfluxed charge of the correct size and quality. The earlier McKee top
has almost been replaced by the bell-less top for better distribution of the
burden. It might also get replaced by the gimble top in the near future.
High top pressure is invariably used to improve gas solid reactions, in spite
of top-sealing problems. The earlier gas cleaning, using cyclone and dry gas
cleaning, was replaced by a dust-catcher and wet gas cleaning system. But
it is again being replaced by cyclone and dry gas cleaning, but of more effi-
cient designs. Slag is granulated and sold as raw material for cement mak-
ing. Oxygen enrichment and humidification is resorted to in order to make
use of hydrogen as more efficient reducing gas in the stack. Fuel injection
through tuyeres is becoming the order of the day. Instead of using a tradi-
tional two-skip hoist for lifting the charge materials to the furnace top, a
belt conveyor system is being preferred.
FIGURE 17.23 The way carbon rate decreased in the blast furnace practice over the years, as result of
modifications.[71]
Process models of various zones and activities have been developed

and the operation is, fully or partially, controlled using some kind of expert
system. The useful volume of the blast furnace has almost reached the max-
imum size of nearly 5000 m3. In general, furnaces with less than 1500 m3
useful volume will not be constructed hereafter for being unable to accrue
the advantages of scale. MBFs may still prove to be useful under certain
specific conditions for the manufacture of merchant iron.
Although the theoretical minimum coke rate required for smooth BF
operation may be approximately 250–300 kg/thm, it has not been possible
to reach this level as yet. This is the most important parameter affecting
the production economy. Because of these improvements it has continu-
ously decreased, as shown in Figure 17.23. The coke rate of 300–350 kg/
thm achieved in some of the very modern practices is therefore a great and
significant achievement.
These achievements will be obvious if the performance of today’s mod-
ern BF is compared with those of yesteryear.
CHAPTER
18
BLAST FURNACE
OPERATIONAL CONTROL
In This Chapter
● Introduction
● Process Variables
● Probes for Measuring the Process Variables
● BF Operational Models
● Process Control Strategies
● Central Working Index
● Advantages of Process Control
18.1 Introduction
From the beginning until very recently, ironmaking was an art. Gradually,
the process variables became more and more apparent. Initially it was only
the charge proportion and their individual weights, and the blast rate were
considered as process variables. This list of process variables continuously
increased, and now a number of such variables, which influence the BF
operation, are understood and measured.
Slowly the chemistry and thermodynamics of reduction of iron oxide,
its mechanism and kinetics, became evident. The irregularities in BF
operation, occasional low productivities or excessive coke consumption
or inability, at times, to push required blast through the tuyeres and such
odd situations compelled the operators to think about the operation and
device means to control them to ensure, as far as is possible, a smooth
furnace operation.
All such developments slowly evolved into the concept of blast furnace
process control, soon after the World War II. The advent of the computer,
especially around the same time, helped to develop mathematical models
to explain what was going on inside the blast furnace at various regions and
zones. A variety of such models is now available and is used to understand
one or the other operation/phenomena inside the furnace or the operation
as a whole.
The situation as it is and, along with the changed requirement of
more and more consistent quality of the hot metal, is desired for its
effective and economical conversion to steel. When the steelmaking pro-
cesses were slow, any quality of hot metal could be tolerated and adjusted
to produce finally the desired quality of steel. The advent of very fast
processes such as LD/OBM required more consistent chemistry and
temperature of the hot metal to ensure process productivity and econ-
omy of steelmaking. Any relaxation in hot metal quality and its consis-
tency finally affected the production rate of steelmaking adversely. Mere
manual control and charge balancing could not ensure such consistency.
Hence an on-line expert process control system became more and more
necessary, nay almost imperative.
A complex apparatus such as the modern blast furnace can produce
consistently the same quality of hot metal only if the operation is carried out
with minimum deviation in input variables and is controlled in its entirety.
In fact, the quality and consistency of hot metal is directly proportional to
the extent of control exercised on blast furnace operation.
On-line control was not possible without measuring process variables
and, consequently, the means to carry it out in an actual running blast fur-
nace was increasingly felt as a priority for efficient BF operation. General
developments in engineering came in handy to develop gadgets and devices
to measure these variables. Over the last few decades these gadgets became
more and more sophisticated and started measuring the variables of the
process digitally, to obtain more accurate and reliable instant-data, to be fed
to the expert system for it to respond properly.
Blast Furnace Operational Control • 369
18.2 Process Variables

In a modern blast furnace, the following variables may be measured for
effecting process control using one or more computerized mathematical
models:
1. Stockline level with respect to a certain fixed point
2. Charging chute position, angle, rotation, and material flow rate
3. Top off-gas temperature, pressure, flow rate, and composition
4. Temperature distribution over the circular cross section at the top
5. Charge material weights, rate of addition, moisture analysis, et cetera
6. Blast furnace pressure drop
7. Shaft gas temperature distribution
8. Stave and lining temperatures
9. Cooling water volume, its outlet temperature at all cooling plates
10. Hot blast temperature and pressure
11. Tuyere cooling water temperatures and water volume throughout
12. Blast volume, humidity, content and oxygen enrichment
13. Amount of injectants blown through tuyeres
14. Hot metal tap weight and rate, its analysis, particularly silicon content
15. Slag weight and rate
16. Hearth side wall and bottom temperatures
17. Temperatures at any point inside the furnace
Variety of Operational Variables

Let us first understand the variables in the BF operation. There are inde-
pendent as well as dependant variables. These are described below:
The gross independent variables of the process are:
1. Charge weight and their inter se proportion
2. Blast temperature, pressure, and rate of blowing
3. Blast moisture content

4. Chemistry of burden, including ash of coke
5. Reducibility of burden
6. Reactivity of coke
7. Size distribution of charge
8. Physical properties of burden materials
The gross dependent variables of the process are:
1. Top gas analysis
2. Top gas pressure
3. Blast pressure
4. Top gas temperature
5. Heat losses to cooling water
6. Rate of production of metal and slag
7. Chemistry of metal and slag
8. Temperature of slag and metal
9. Distribution of charge at the throat
Out of the above input variables there are some that cannot be varied
immediately at the asking. It takes a while, may even take more time, to
adjust them. But there are some variables that can be instantly varied to
control the process. Such instantly controllable variables are the blast
rate, pressure, temperature, oxygen enrichment, humidification, and so
on. The variables such as coke rate and distribution take a while to be
effective even if adjusted instantly. In fact, these are changed gradually
to have the desired effect. The chemistry of the burden, its reducibility,
size, and such variables takes a very long time to understand their effect
on a given blast furnace operation, and it takes equally long to correct
them, when required. These are therefore maintained the same as far
as possible to ensure more effective control or are adjusted with a very
long-term view.
18.3 Probes for Measuring the Process Variables

Process control is based on measurement of both, independent and depen-
dant variables of the gross system. It fundamentally assumes “what cannot
be measured cannot be controlled.”
In order to measure the above-mentioned process variables, it is neces-
sary to have the right kind of probes to measure them. The following probes
are generally used in a modern BF to measure the process variables men-
tioned against their names:
1. Profilometer — Charge layer thickness across the radius at the throat
— Radial ore to coke ratio
— Burden descent rate
2. Top probe — Gas flow distribution
3. Vertical probe — Location and shape of various isotherm

— Location of reserve zones
— Burden reduction rates and decrepitation
4. Upper shaft probe — Gas flow pattern

— Shape of cohesive zone
5. Middle shaft probe — Gas flow pattern

— Shape of cohesive zone
6. Stave thermometer — Scaffold development indication

— Location of root of cohesive zone
7. Belly probe — Layer thickness in the cohesive zone

— Status of burden reduction
— Early warning of change in furnace thermal level
8. Tuyere coke sampler — Coke condition at tuyere

— Raceway depth
9. Core probe — Deadman condition

— Coal combustion behavior
10. Hearth thermometer — Wear pattern in the hearth

— Fluid flow phenomena
The places and how the various probes are used are shown in Figure 18.1.
FIGURE 18.1 The position of various probes vis-a-vis the furnace zones for direct measurements[81].
18.4 BF Operational Models

Initially, static thermal and material balancing models were developed. The
dynamic nature of a blast furnace is not readily explicable by such models
except for some simple calculations of charge and fuel requirements. They
were alright until the advent of mathematical simulation of the complex
BF processes. This increasingly revealed the dynamic nature of its com-
plex operation. It was impossible to solve these equations manually. Use
of a computer for processing these complex equations was imperative.
The understanding of the complex BF operation thereby improved con-
siderably. A large number of such mathematically simulated models were
developed and are available to explain one or the other phenomena occur-
ring inside the blast furnace. These are:
—Burden distribution model
—Wall temperature distribution model
—Top gas distribution model
—Hearth erosion model —Deadman status model
—Silicon prediction model —Accretion status model
—Shaft simulation model —Tuyere state model
All these are taken together to work out the overall model.
These models assume a certain set of idealized conditions to exist inside
the furnace. BF, however, seldom operates under such idealized conditions. It
is subject to many disturbances that cannot be predicted beforehand. In fact,
this unpredictable operation of blast furnace was reflected by the operators
in calling it by the pronoun “she.” The relevant probes indicate the deviation
and its extent. The term process control means the ability to adjust the mea-
surable operating variables, when they go off the reference standard state.
The ability to sense and adjust the operating variables therefore decides the
effectiveness of the probes as well as the process control. All these measure-
ments are obtained as digital signals and therefore are amenable to be pro-
cessed by a computer network that works on the mathematically simulated
software at the back-end. It gives the corrective action needed to bring back
the system as close to the assumed idealized state as is possible.
However effective the process control system is, it cannot take any
changes, particularly those that are not exactly measurable, into account for
control adjustments. It is therefore very necessary to minimize the variable
nature of the inputs, (e.g., quality of ore, sinter and coke, blast temperature,
blast volume, charging sequence, ore/coke ratio, etc.). The more these are
controlled, within the prescribed limits, the better and easier it is for pro-
cess control.
18.5 Process Control Strategies

It is necessary to produce iron at the best possible economy by running the
BF optimally. The production costs of a given plant operating under certain
prescribed operational strategy are determined by:
—Productivity of the furnace defined as thm produced per day

—Hot metal quality in the light of steelmaking requirements
—Fuel consumption (coke rate and rate of PCI, if any)
—Raw material selection
These factors are such that improvement in one may mean an unfavorable
situation for the other. The overall view therefore has to be taken to work
out the economics of production and marketing. Mathematical models
with an expert computer control system are used to adjust these to obtain
optimum economy. The system is in closed-loop with the input variables.
By sensing the deviation, it indicates the corrective action needed and its
extent in terms of the following input variables:
—Coke rate
—Burden basicity
—Humidification
—Oxygen enrichment
—Fuel injection
—Fine-tuning of burden distribution
The system is an expert system containing the knowledge-base represent-
ing the know-how of the BF operation gathered from accumulated expe-
rience the world over. The system either operates automatically, wherein
the corrective action is automatically conveyed to the inputs for immediate
implementation, or it gives the indication of necessary input changes, which
may be considered and then implemented as per the judgment of the expe-
rienced operator.
Overall Control Strategy

The automated system aims at minimizing judgments of the operators
for adjusting the furnace inputs. The system has interfaces with various
measuring devices and the analytical tools to accept the status of the
process in terms of output quality. Various furnace zones/activities are
controlled using their individual models, and these are all integrated in
the overall expert system, if required. The expert system structure is
FIGURE 18.2 Structure of the expert system.
The system is expected to counteract the fluctuations in the process by rec-

ommending appropriate changes in the inputs.
The overall aims of the strategy are:
1. Reduction in fuel consumption
2. Avoidance of critical process situations such as hanging, scaffolding,
slipping, et cetera
3. Stabilizing the product quality, particularly with respect to silicon con-
tent and temperature
4. Burden distribution control as per the needs
5. Avoidance of heavy control actions
6. Equalization of operational decisions over all the shifts
Control exercised via control of combustion in the raceway, by adjusting
the blast rate and pressure, instantly controls the quality of reducing gases,
temperature, and volume, and thereby affects the rate of indirect reduc-
tion almost instantly. These are very effective control tools. As compared
to this, the control exercised from top by way of burden distribution and
ore/coke ratio takes hours before it is effective. Although both control tools
are useful, the combustion control is more often resorted to, for immediate
rectification. In fact, the actual control is most often the thermal control of
the process only. The level of oxygen enrichment and humidification, hot
blast temperature and rate of blowing, fuel injection rate, and so on are,
in general, only thermal control parameters. Out of them, humidification
control is instantly effective and easily possible because the thermal effects
are immediately apparent. On the whole, the more the number of variables
measured and controlled, the more effective is the process control.
The process control system assesses the furnace condition at a fixed
interval of time, which can be increased or decreased suitably. The correc-
tive signals are available at the same frequency. Whether or not the correc-
tive action is to be taken depends upon the operator. No instrument can
entirely substitute for human judgment. The system might not have taken
into account or there may be some unknown parameters yet to be taken into
account in the system. Hence, the operator’s judgment still reigns supreme.
18.6 Central Working Index

The control of the BF operation starts with obtaining the physical condition
of the burden. It means actual stock-line level, rate, and uniformity of its
descent. The gross physical condition of the burden has a direct bearing on
its rate of heating and reduction. This is assessed in terms of temperature
and composition of the outgoing gas. Ideally the temperature of the off-gas
should be approximately 200°C and the CO2 content in the range of 7%–12%
by volume. This should be uniform all over the stock-line (top surface area of
the burden). These could be measured continuously or periodically to inter-
pret the gross physical condition of the burden inside the furnace.
The stock-line is measured by lowering stock level indicators. They
need to be protected against the falling charge. Although several designs
are available, no design has yet proved entirely satisfactory. The application
of high top pressure has only aggravated the problem further.
Horizontal probes are inserted to obtain a gas sample from any point in
the stack in running conditions. Similarly, temperature is also measured at
any spot in the stock-line. Probes are also available that can be pushed into
the stack at any point horizontally whenever needed. These probes either
penetrate up to the center or right across to the opposite wall. It extracts a
small gas sample for analysis, and a thermoelectric element provided
therein measures the temperature at any point in the stack in a running
furnace. It is therefore possible to know the gas composition and tempera-

ture at any point in the stack.
Standard charts of gas composi-
tion and temperatures are prepared
along four mutually perpendicular
radii of the throat, corresponding to
various characteristics of the opera-
tion. This gives the normal gas and
temperature curves, and any devia-
tion from this has to be corrected by
a suitable corrective action.
The temperature inside the stack
will be as shown in thermal profiles
in Chapter 10. The temperature at
any horizontal plane shall be some- FIGURE 18.3 Schematic variation in
what like that shown in Figure 18.3. temperature at any horizontal.
Based on these measurements the
central working index (CWI) is evaluated as follows:
Sum of any two temperatures at the central line
CWI =
2 × average temperature of all the over burden temperature
This index is correlated with process variables such as coke rate, %Si in
the hot metal, productivity, efficiency of gas utilization, and so on. In order
to obtain the expected performance, the CWI is adjusted at the required
value by effecting the proper burden quality and charging sequence.
18.7 Advantages of Process Control

The following benefits have been derived from process control:
1. It has been possible to reduce the variance in silicon content by the
process control; the extent of achievement depends on the effective-
ness of the control system.
2. The understanding of the dynamic nature of the blast furnace process
has been greatly achieved through such control systems.
3. The thermal evolution of the lower zone of BF has been greatly
improved.
4. Several sensing devices have been developed and put in commercial

furnaces.
5. Continuous control often leads to advance knowledge of possible faults
or irregularities in furnace working, and advance corrective action can
be initiated to off-set these.
6. It has also greatly improved the general understanding of the process
whereby performance of the furnace, in terms of economy and pro-
ductivity, has improved considerably.

The data collected from various probes and measurements, after required
processing through the expert system, give out corrective signals, which can
be initiated manually, semi-automatically, or fully automatically, depending
upon the facilities provided therefore, to ensure smooth furnace operation.
The key word in process control of blast furnace is consistency of inputs.
It means the charge must be the same in terms of chemical and physical
properties to the extent possible so that variation in the charge does not
affect/deflect the operation from its standard state. In this respect, since
natural ore cannot be so consistent, more stress is laid on using artificially
prepared burden. Similarly, the rate of blast, its pressure, and its tempera-
ture are also fixed within the limits possible. Burden distribution is fully
exploited to obtain commensurate distribution at any moment to suit the
smooth furnace operation.
The furnace operator is provided with the knowledge of the status of
the charge column inside the furnace, assessed by the various probes. Any
irregularity, depending upon its extent, can be corrected by taking suitable
corrective action, particularly by changing the sequence of charging and
distributing the burden suitably and raceway control.
The chemistry of hot metal, particularly its silicon content, is known
only after its analysis. If it goes astray then, even corrective actions will be
to correct it only slowly.
The measurements that take a long time are not suitable to be incorpo-
rated in the control scheme. The probes must sense quickly and convey the
signal about the status to the expert system.
CHAPTER
19
ALTERNATIVES TO BF I –
LOW-SHAFT AND
CHARCOAL FURNACE
In This Chapter
● Introduction
● The Low-Shaft Furnace
● Mini-Blast Furnaces (MBFs)
19.1 Introduction
The blast furnace design and practice that has evolved over the last hun-
dred years or so can be efficiently and economically run only on very good-
quality coke of a certain size fraction. It cannot be run without coke or using
a smaller-size fraction of coke. With increasing production of steel, the
requirements for coke have risen so much that it has now become a scarce
commodity. Good-quality metallurgical coals are available only in some
countries, and they have developed a sort of monopoly for this material
in the international trade. The price has continuously risen in the past and
is still on the rise because of this monopoly. The continuously rising steel
output on a global level has only aggravated the situation. This challenge of
keeping the global production of iron and steel at such high levels is there-
fore being tackled by a multi-pronged approach. On the one hand, the effi-
ciency of BF has been continuously improved to decrease the coke rate and
substituting as much as possible the fuel requirements using ordinary coals
using injection techniques. On the other hand, several alternative routes

of iron production have been developed to produce iron without using the
coke that is required to be used in a BF. These alternative routes employed
either that fine-size-fraction of coke, not used in BF, or non-metallurgical
thermal coals, charcoals, natural gas, propane, hydrogen, et cetera, as fuel
and as reductant in the production of iron. One of the alternative used is
electricity as a thermal energy source and coal/coke as a reductant. These
alternative routes can broadly be classified as follows:
1. Use of low-shaft blast furnace, using a nut-size fraction of coke re-
jected from the normal blast furnace operation. This is the nut coke of
+5 and −30 mm size.
2. Use of charcoal as a fuel, entirely replacing the coke altogether.
3. Production of sponge iron (solid state reduction of iron ore) using
cheaper solid, liquid, or gaseous fuels.
4. Production of ferro-coke or carbonized iron using cheaper and abun-
dantly available non-coking coals.
5. Use of a submerged electric arc furnace using poorer variety of fine
sized coke as reductant and electrical energy to supply thermal re-
quirements of the process.
6. Smelting reduction or SR processes, which operate in two stages (i.e.,
first pre-reduction followed by reduction smelting to produce molten
iron). Thermal requirements to be met by burning ordinary coals with
enriched air or pure oxygen.
The importance of adopting alternative routes of iron production using
cheaper thermal coals as fuel and as reductant hardly needs extra emphasis.
Out of these the low-shaft furnaces, including the mini-blast furnaces and
the charcoal furnaces, are described in the present chapter. The rest of the
processes are described one-by-one in the following chapters.
19.2 The Low-Shaft Furnace

The extremely high thermal efficiency of a conventional blast furnace is
essentially due to the tall shaft, where processing of the ore and pre-heating
and reduction take place before it descends into the melting zone. The
effective height of the shaft required to carry out the processing of the
ore depends on the reducing power and the temperature of the furnace
Alternatives to BF I – Low-Shaft and Charcoal Furnace • 381
gas generated at the tuyere level. In a normal blast furnace, the gas phase
generated at the tuyere level contains nearly 60% nitrogen, which is inert
and acts only as a heat transfer agent. It therefore needs a tall shaft, which
in turn necessitates the use of good-quality coke as the fuel.
The question that arises, therefore, is whether it is feasible to reduce
the shaft height without impairing the thermal efficiency of the process. It
means that the gases from the low-shaft furnace should go out of the furnace
at a temperature and carbon monoxide concentration that are comparable
with the normal blast furnace. If this is feasible, then poorer-quality coke
or even non-metallurgical coal could be used as a fuel to run such furnaces.
The shaft height can be reduced without impairing the thermal effi-
ciency if the heating and smelting zones in the furnace are compressed
by accelerating the heat exchange and chemical reactions, that is, more
sharper thermal and reducing potential gradients are obtained. The imme-
diate suggestion to achieve this is the use of oxygen-enriched blast, which
would produce higher temperature and furnace gas containing lower nitro-
gen content and higher reducing potential. Consequently, the necessary
heat transfer and reduction would be achieved over a shorter shaft.
Based on the above reasoning, the idea of a low-shaft furnace was con-
ceived by a group of seven countries, viz. France, Italy, Belgium, Luxemburg,
Holland, Austria, and Greece who jointly set up the first ever experimental
low-shaft furnace at Ougree, near Liege in Belgium (C.N.R.M.) in 1952–
1955. The original design underwent some changes after preliminary trials
and extended trials using a variety of charges available in these countries
were carried out to establish the optimum conditions of their efficient
smelting. The raw materials that were tried included low-grade lump ores,
concentrates obtained by upgrading the lean ores, rich ore fines, poor-qual-
ity coke, lignite, et cetera. The charge was fed either in the form of separate
lumpy material or in the form of briquettes made from a suitable mixture
of these charge materials. The difficulties that were anticipated, such as
high thermal losses, low indirect reduction, insufficient conditioning of the
charge before reaching the tuyere zone, incapability of high top pressure
application, etc. either did not exist or these could be encountered success-
fully. As a result of these trials a few commercial low-shaft furnace plants
were set up in these countries in the 1950s.
The low-shaft furnace plants at Colbe, Germany may be cited as an illus-
tration. This plant had ten furnaces, each producing 80–100 tpd. and produc-
ing 250,000 t of foundry-grade iron per annum. The furnace had an 11 m2
rectangular cross section at the tuyere level and an effective height of 4.8 m.
The blast was pre-heated to 780° C–800° C. It has a total of eight tuyeres, three
each on the two long sides and one each on the two short sides. The furnace
operated on a bedded charge of small lumps of iron ore, lime stone, and lignite
coke briquettes produced by low-temperature carbonization. The iron ore ana-
lyzed was 21% –27% Fe and 12% –40% SiO2. The consumption of lignite coke
is approximately 2000 kg per metric ton of pig iron.
Low-shaft furnaces have been successfully run on two types of charges.
One of the types, as mentioned above, consists of lumps of individual charge
materials. The other type consists of briquettes made from suitable mixture
of fines of ore, stone, and lignite, with tar as binder.
If the briquettes mixture is used, then the furnace should be run in such
a way as to prevent condensation of tar at the top, as this is evolved when bri-
quettes get pre-heated at the furnace top. The top-gas temperature is, there-
fore, kept rather high, that is, 350° C–400° C. The briquettes also should be
made such that they do not burst when exposed to this temperature on charg-
ing. The use of briquettes not only assures uniform distribution of the charge
but guarantees sufficient permeability for efficient heating and reduction.
FIGURE 19.1 The low-shaft furnace.

In the low-shaft furnace very little reduction occurs in the upper part
of the furnace but rapid reduction occurs a short distance above the tuy-
eres. The reduction of the ore takes place predominantly directly by carbon
rather than by the gas because of the short shaft, which is unlike the blast
furnace. The top-gas from a low-shaft furnace therefore is not only at higher
temperature, but it contains more carbon monoxide as compared to a blast
furnace. The CO/CO2 ratio in the top-gas is approximately six. Average
analysis of top-gas is approximately 30%–35% CO, 5%–6% CO2, 0.5% CH4,
1%–2% H2, and the balance, N2. Its calorific value averages 4400–5000 MJ/
Nm3 (110–150 B.Th. N/cft.) The volume ratio of slag depends very much
on the type of raw materials used for smelting.
The low-shaft furnace is shown in Figure 19.1 in a simplified form. In
general, the low-shaft furnaces seldom have effective height more than 5 m
(16 ft.). Although circular cross-section furnaces are in operation, rectangu-
lar cross section is very popular. The hearth area is generally approximately
8–10 m2. Nearly 500–700 Nm3 of blast is blown to generate 8000–11,000 Nm3
of top-gas. The average consumption of lignite is more than 2000 kg/t of
iron. This value, although high in comparison with the coke consumption in
a modern blast furnace, is not disadvantageous when the fuel requirement
of sintering and coke ovens, which are additionally required for the blast
furnace smelting, are considered together. It could be reduced by increasing
the pre-heat of the blast. Many of the recent innovations in normal blast fur-
nace practice, such as high top pressure agglomerated burden, and others,
can be adopted in a low-shaft furnace practice also.
The advantages of a low-shaft furnace can be summarized as follows.
1. It can take fine, friable, low-grade and such inferior type of ores in the
charge.
2. Fuels of inferior grade, such as lignite, can be successfully used. It can
therefore be adopted for making iron where metallurgical coal is not
available but non-metallurgical coal is abundant.
3. The furnace top-gas, a by-product, is a rich fuel that can be economi-
cally used elsewhere in the plant or partly for pre-heating the blast.
4. The blast pressure necessary to run the low-shaft furnace is of the
order of 2 kg/cm2, which is much lower than that for a normal blast fur-
nace, and hence costly high-capacity turbo-blowers are not necessary.
5. Pre-heating of the blast can be readily accomplished in a recuperator,
and hence costly stoves are not necessary.
6. Prior carbonization of fuel and agglomeration of ore is unnecessary if

briquettes are used. It therefore eliminates the coking and agglomera-
tion plants and thereby reduces the capital investment.
7. The furnace can be readily started and stopped without any expensive
and prolonged procedure involved.
8. Control of smelting operation is easy and is exercised by control of
blast temperature, pressure, and volume.
9. Silicon content of the pig iron would be low in the case of basic irons.
10. The production rate is approximately 2.5 t/m3 of useful volume of the
furnace, which is much more than that of a blast furnace.
11. The through-put time of the charge is only 1–2 hours, as compared to
8–10 hours in a blast furnace.
The furnace, however, suffers from the following disadvantages as well:
1. Briquetting of the charge mix may be necessary for efficient produc-
tion, and it adds to the cost of raw material preparation.
2. The physical and chemical preparation of the charge is desirable, if not
necessary, for batter efficiency.
3. The daily production of an individual unit is low, of the order of 100–
200 tpd, and hence the battery of furnaces are required to obtain even
a sizable production rate.
The successful adoption of a low-shaft furnace on the continent for smelting
iron ore with lignite as the fuel led to investigate the possibilities of adop-
tion of the same technology in India for smelting Salem magnetite or other
ores with Neyveli lignites. The quality of Neyveli lignites is comparable to
that of the lignite used on the continent successfully. Consequently, a low-
shaft furnace pilot plant of 15 tpd capacity, along with all other facilities,
was set up at the NML, Jamshedpur in the late 1950s. Several diverse trials
using a variety of charge materials were carried out over a sufficient length
of time. Contrary to expectations and in spite of the apparent similarity
of our raw materials with those being successfully smelted in commercial
low-shaft furnaces on the continent, it has not been possible to successfully
produce iron, from our diverse raw materials, to the extent that it could
be adopted for its commercial application. The whole project was pursued
with vigor in the beginning, but it was later abandoned.
The idea of assessing the possibilities of adoption of low-shaft furnace

for Indian conditions still needs to be pursued to its logical end by carrying
out extensive trials on the same pilot plant, supported by laboratory investi-
gations of parameters of related importance.
19.3 Mini-Blast Furnaces (MBFs)

In spite of developing several sponge ironmaking technologies on a com-
mercial level, production of liquid iron has its own merit and economical
advantages. It is essential in foundries. In trying to find out the alternatives
to a normal blast furnace, the possibilities of designing a small shaft blast
furnace capable of using fuel other than the one required for a normal
blast furnace becomes an attractive proposition. Initially, iron was pro-
duced using charcoal alone. As a historical legacy, use of wood charcoal
continued, wherever possible, even after the advent of coke, to produce
iron. The small shaft commensurate with the low strength of charcoal was
then necessary.
Lower-size fraction coke, otherwise not acceptable to a normal blast
furnace and therefore screened off as the nut-size fraction from normal
coke, was found to be useful as a fuel in such small shaft blast furnaces.
These led to the development of a variety of designs popularly known as
mini-blast furnaces, or MBFs for short. These have been widely and suc-
cessfully adopted in Germany, Bulgaria, Vietnam, Russia, Brazil, China,
and India. Several charcoal blast furnaces were switched over to the low-
shaft operation with suitable modifications. Their share in iron production
is approximately 4% in world iron production.
Special Features of MBF

It is not only a miniature version of a blast furnace but has several additional
characteristics in terms of simplicity of design and economy of its operation.
1. The working volume of such furnaces is normally in the range of
60–500 m3, corresponding to production capacities in the range of
30,000 to 200,000 thm/annum.
2. The top pressure is just sufficient enough to be able to clean the off-
gases using a double venturi scrubber (i.e., 0.3 kg/cm2 pressure level for
gas cleaning is sufficient and there is no need to use ESP). Even this
much cleaned gas can be used for blast heating and power generation.
3. No elaborate turbo-blowers are necessary since blast can be generated

by a set of fans, which are simple to operate and maintain.
4. Hot blast temperature of the order of 750° C–850° C is obtainable and
is good enough.
5. Hot blast is not generated by stoves but by metallic recuperators,
which are simple to operate.
6. The lining is of heavy-duty medium alumina bricks. It is not necessary
to have water-cooled plates to cool the lining and instead, the outer
steel shell itself is cooled by water spray. It still gives a campaign life
of 3–5 years.
7. It can take Jumpy iron ore or sinter as charge without any difficulty.
However, use of sinter does not add very specific advantages, as is be-
lieved.
8. The instrumentation and control is much simpler as compared to nor-
mal BF.
9. The coke rate is usually 550–650 kg/thm.
10. The specific productivity is approximately 1.5–2.0 t/m3/d.
11. The surplus off-gas is used to generate more than required electric
power for self-use.
Modern MBFs
The old charcoal blast furnaces were suitably modified into MBFs to take
nut-size coke as fuel and spacer inside these small blast furnaces. The early
trials were therefore carried out on these furnaces to evolve the MBF
design.
Although the Chinese small furnaces attracted the attention of India’s
steel producers in the 1970s, no further action was taken pertaining to
their adoption in India. It was only after the economic liberalization and
de-control of the iron and steel industry that a number of MBFs were set
up for either merchant iron production, especially for foundries, or as a part
of integrated steel plants of smaller sizes.
Almost 85% of all MBFs installed in India adopted the Tata-Korf design
of MBF. These are up to 250 m3 volume. The productivities reported are in
the range of 1.5–2.0 t/m3/d, with a coke rate of 650–900 kg/thm. All types of
cokes have been successfully used in MBFs.
Chinese MBF
Once, approximately one-fifth of iron production in China came from
MBFs (a total of 55 mtpa). They innovated this technology considerably.
Eighty-five to one-hundred percent sinter was incorporated in the charge.
Coke rate, therefore, varied by approximately 550–630 kg/thm. Pulverized
coal up to 60 kg/thm was successfully injected to bring down the coke rate
by 40–50 kg/thm. Hot blast temperature up to 1200° C was developed and
used. For this purpose, dry cleaning of gas was adopted effectively. All these
culminated in getting specific productivity up to 3.0 t/m3/d. Chinese ores,
though, contain only 40%–45% Fe but are very low in alumina (max 2%).
Chinese coking coals are very good quality (ash only 10%–12%). These bet-
ter input parameters are certainly responsible in obtaining good productiv-
ity values. The production details are given below:
150 m3 MBF 300 m3 MBF

Hot blast temperature, °C 900–1000 900–1000
Production, tpd 330 660
3 2.2 2.5
Sp. productivity, t/m /d
Coke rate, kg/thm 560–570 550
3 3.56 3.56
Thermal value of off-gas, MJ/Nm
Slag rate, kg/thm 500 500
Pig Iron (%) Slag (%) Coke (%)
C–4.0 CaO–38–39 F.C. –85
Si–1.4 (special) MgO–12 V.M. –1.8
0.7 (normal) Al2O3–13 Ash–13
Mn–0.12 SiO2–35 S–0.7
P–0.07 FeO2–0.5 Size–25–70 mm
S–0.03 S–0.7
First MBF in India

Kalinga Iron Works of Orissa were the first to install a MBF in India way
back in the early 1960s. It was designed by Dr. C. Otta and Co. of Germany
and supplied and erected by M/s Krupp, A.G. Eissen, Germany. There are
in all three furnaces, one with working volume of 37.19 m3 and two each
with a volume of 61.21 m3. They have an overall height of 14–15 m, with
stack height of 5.8 m. They have only one tap hole and 100-mm-diameter
tuyeres. It produces 100 thm/d.
Initially, the nut-coke size of 20–40 mm as screened off from usual BFs was
used as a fuel. The furnace was operated more or less in the same way as the
normal BF, maintaining the slag basicity approximately 0.9. The productivity
was 1.8–2.0 t/m3/d. When even the nut coke was not available, it was operated
with Beehive coke of high ash content of the size of 10–30 mm. Slag basicity
was therefore raised. As a result, the productivity decreased to 1.5–1.6 thm/d.
The practice was therefore again changed to use rejected coke of
10–25 mm size with some nut coke and hard coke. Basicity was reduced
to 0.7–0.8 and MgO was raised to 8%. Alumina was maintained at 22%–
26%. This improved the productivity to 2.4 t/m3/d. However, the coke rate
increased to 800 kg/thm due to a high slag volume of 500 kg/thm.
Small MBFs suffer from low heat transfer and low indirect reduction effi-
ciency because of the short stack height. The off-gas is richer in CO to the extent
of 28%–30% with a CO/CO2 ratio of 3, against the 1.5 obtained in usual BFs.
The history of this plant and experiences in running the furnaces against
so many heavy odds is an interesting technological research into the use of
MBF carried out much before these furnaces became very popular in India
in the 1990s.
KIW experimented on the use of low-temperature carbonized coke,
formed coke, sponge, iron, and coal injection through the tuyeres, et cetera,
to improve upon the productivity. They also tried external desulfurization
using soda ash to improve sulfur content of the hot metal from approxi-
mately 0.15% down to 0.07%. The first Tata-Korf MBF of 175 m3 size was
commissioned by SESAGOA of Goa in 1992 and since then this technology
is being increasingly adopted by many as shown before. Some of the fea-
tures of this Goa plant are given below:
Installation cost – Nearly Rs. 6000 per annual t.

Blast pressure – 0.7 kg/cm2
Top pressure – 0.3 kg/cm2
Stove – Metallic recuperator system, two in parallel
Hot blast temperature – 900° C.
Top-gas temperature – 250° C
Gas cleaning – Wet system, dust conent, 5 mg/Nm3
Sp. power consumption – 165 kwh/thm
Gas holder – 5000 m3N
Man-power – 175
Slag rate – 275–300 kg/thm
Coke rate – 600 kg/thm

Sp. productivity – 1.5 t/m3/d
Pig casting machine – 5 kg pigs
Metal quality –Temp–1450° C, C–4.0%, Si–2.0%, Mn–0.6%, S–0.03%.
P–0.06%.
Metal temperature – 1450° C
Slag basicity – 0.7–0.8
Slag chemistry – CaO-30%–32%, SiO2-35%–38%, MgO-8–10%, Al2O3-22–26%
Iron ore charge – +64% Fe of + 10–40 mm size
Charcoal Blast Furnace

Purposeful smelting of iron ores started off using charcoal as a fuel as well
as a reductant. It continued during the entire evolutionary period of the
blast furnace until the design and operation were properly evolved, pro-
ducing ultimately molten iron as the product. Although inferior in quality,
it became a universal fuel for smelting of iron ores until almost the eigh-
teenth century. The use of charcoal had to be abandoned, if not totally, at
least substantially, due to the restrictive legislation forbidding cutting forest
trees and burning the wood for making charcoal, since the forests were
getting wiped out at a faster rate than their natural growth. Although coke
was introduced as an alternative fuel in 1619, use of charcoal continued,
with decreasing rate. The charcoal blast furnaces continue to be operated
wherever the problem of charcoal is not so acute. Brazil, Sweden, China,
and some continental countries are such examples of where this takes place.
The use of charcoal in Bhadravati Iron and Steel Works in Karnataka, now
known as Visvesraya Iron and Steel Ltd., was also abandoned for the same
reasons in favor of electric smelting in the 1950s.
Most of these furnaces were less than 100 m3 useful volume, although
even 800 m3 furnaces were not uncommon. Once they produced nearly
30% (nearly 7 Mt) of the total iron production. The use of charcoal in fact
withheld progress of the blast furnace, since height and size could not be
increased on account of low strength of charcoal. Once this was replaced by
coke, the size increased rapidly.
Very low bulk density of charcoal resulted in its occupying substantial
useful volume of the furnace. As a result, the productivity of charcoal fur-
naces was relatively poor. The slag volume, however, was much less, that
is, 120 kg/thm because of the very low ash content of charcoal. The off-gas
was much richer in CO. It has a calorific value of 4.0–4.2 MJ/Nm3 against
3.1 MJ/Nm3 for coke furnaces.
Brazil still produces 6–7 Mt of iron through such blast furnaces. Each
such unit, on an average, produces approximately 100,000 t of iron/annum.
Charcoal is a very costly fuel. It is screened at the plant to remove −10 mm
fraction, which is sold at a throw-away price. With the advent of pulverized
coal injection in normal BFs, this −10 mm charcoal is also being used for
injection through tuyeres to the tune of up to 120 kg/thm in these furnaces.
It does require some further grinding before its injection.

Small-size merchant iron production plants or integrated iron and steel
plants, using any one of these alternatives, has a distinct advantage in a
country such as India, to be able to produce iron and steel at a large number
of places spread all over India to cater to the local needs rather than pro-
duce them at a few places and then transport them to all corners of India,
straining the already strained transport system.
Out of the alternatives mentioned herein, the low-shaft furnace was
tested for Indian conditions, but it failed to lead to any success. The rea-
sons are not all clear. The question of using charcoal does not arise in India
because of dangerously depleted forests, about which everybody is worried.
Stricter legislations do not permit production of charcoal on such a large
scale, either. The only alternative left is the MBF. In fact, it is not a real
alternative to the blast furnace.
As compared to standard BF, the installation of MBF may be eco-
nomical in terms of capital investment and gestation period involved. A
large number of entrepreneurs can take up this challenge, and inter se
competition can give a better product at competitive prices to the cus-
tomer. The future of MBF is therefore promising in India at present. This
is the reason a large number of such units have already come up and are
operating fairly well.
It is, however, necessary to develop some alternative fuel for these if
they are to be really competitive to standard BF. The use of waste plastic to
make coke may be seen with interest in this regard.
Investors will have to very carefully assess the competing economics of
MBF and DRI routes for ultimately making steel at a given place before
making their final choice. But their role for producing foundry-grade pigs
will still continue without much challenge.
CHAPTER
20
ALTERNATIVE ROUTES OF
IRON PRODUCTION II –
ELECTRO-THERMAL
PROCESSES
In This Chapter
● General
● Historical
● The Submerged Arc Furnace Plant
● Operation
● Smelting Practice
● Modern Trends
● Special Features
● Irregularities in Operations
20.1 General
The heat-producing function of coke, that is, the burning of its carbon by
hot blast in a normal small blast furnace, is replaced by an electric arc heat
input in electro-thermal smelting processes. Since no fuel is to be burned
for heat production, air blast is unnecessary. No furnace gas is therefore
generated, as in the blast furnace, and consequently pre-heating and reduc-

tion are not possible in the shaft. The electric smelting furnaces have there-
fore low shafts and no significant indirect reduction of iron oxide by carbon
monoxide is possible. Coke or other forms of carbon are still necessary
because the reduction of iron oxide takes place predominantly by the clas-
sical endothermic reaction:
FeXOY + YC = XFe + yCO (20.1)
The electro-thermal pig ironmaking process therefore essentially consists
of reduction of iron oxide by coke or an inferior quality of coal at high tem-
peratures when the thermal requirement is met by an electric arc; the other
requirements of smelting are met in the usual manner to finally produce
molten iron and slag separately.
20.2 Historical
Electric smelting of pig iron developed in the beginning of the present cen-
tury, particularly in Italy and Sweden because of the pioneering efforts of
Heroult and Strassano. Some 12 furnaces of maximum individual capacity
of 10 MVA were in operation in Italy and Sweden in the early 1920s. These
designs had to be abandoned later because of several defects. However,
inspired by these commercial installations and understanding the reasons
for their failures, Tysland and Hole in Norway developed the classical
Tysland-Hole low-shaft submerged electric arc furnace of approximately
5–6 MVA capacity in the pre- World War II years. The electrodes in this
furnace were in a row, which proved to be defective in larger-capacity fur-
naces. In the Imatra design of 12 MVA size, therefore, the electrodes were
arranged in a standard triangular formation, which continues even today.
Smelting of pig iron in an electric arc furnace on an extensive scale was
possible only in those countries where electrical power was cheap and metal-
lurgical coal was not available. The Scandinavian countries are typically the
best-suited places for such a process for ironmaking. They have therefore
adopted this on an extensive scale and have also contributed much in the
development of this technology as an alternative to the blast furnace. Electric
smelting of pig iron made rapid strides in the period 1950–1970. The present
total output of pig iron by this technique exceeds 4 million t/annum. The larg-
est individual unit of 60 MVA has been built, and this seems to be the limit of
electrical capacity with a three-electrode furnace. Higher-capacity furnaces
with six electrodes may be designed to assess their commercial exploitation.
Alternative Routes of Iron Production II – Electro-Thermal Processes • 393
India
The Visvasvaraya Iron and Steel Ltd. (formerly Mysore Iron and Steel Co.)
Bhadrawati were the pioneers to install two such furnaces each of 100 tpd
capacity in 1952 in India. Recently, two more furnaces each of 200 tpd
capacity have been added to increase the pig iron production. One furnace
of 100 tpd capacity is operating in the private sector in the name of Sandur
Manganese and Iron Ores Ltd at Sandur, district Bellari, Karanataka. The
Maharashtra Electrosmelt Ltd. Chandrapur was a mini steel plant with
electric pig iron and LD steelmaking facilities. The furnace is the biggest in
India and is of a 33 MVA rating. Since the steelmaking facilities were being
built, the furnace was temporarily switched to ferro-manganese produc-
tion, which was later changed over to iron production. Now there are two
33-MVA furnaces, but both produce Fe-Mn only. The steelmaking facilities
have been dismantled.
20.3 The Submerged Arc Furnace Plant

Electro-thermal production of pig iron is carried out commercially in a
three-electrode submerged arc furnace of the fixed or the non-tilting type.
The general scheme of arrangement of such a plant is shown in Figure 20.1.
The plant consists of the following:
1. The furnace proper, along with the power transformer, secondary bus-
bars, flexible cables, electrodes, electrode holders and their contacts,
shell rotation mechanism, et cetera
2. The charge storage bins, charge weighing, mixing, and charging sys-
tems, among others
3. Furnace gas disposal system
4. Cast house liquid product disposal system
5. Raw material preparation plant, if any, such as bedding, sintering, etc.
6. Furnace operation control.
The transformer capacity is an indication of the size of the furnace. Primary
voltage of the order of 33 KV or more, and low secondary adjustable voltage
with high amperage, such as 150 KA, are employed. A step-down trans-
former, circuit breaker, flexible cables, et cetera, are therefore needed.
Provision for star-delta switching on the primary side is a convenient way of
increasing the secondary voltage range, thereby increasing the flexibility of
the transformer, or the power can be fed from either a three-phase trans-
former or three independent single-phase transformers. The latter arrange-
ment is cheaper and is more effective since in the event of breakdown, the
furnace can be run with the remaining one or two phases.
FIGURE 20.1 General scheme of submerged electric arc plant for smelting iron ores.
The furnace is a steel shell lined from inside with basic refractories
such as magnesite or chrome-magnesite. Carbon lining of the lower por-
tion has also been adopted without any problems, as in a normal blast
furnace. The furnace is closed from the top by an inverted bowl-like
roof, lined from inside with ordinary grade firebricks, since it does not
have to stand high temperatures. The lining in general does not have to
perform any arduous duty because the arc heat zone is surrounded on
all sides by the charge.
The roof is provided with openings for insertion of three electrodes
vertically downwards. It is also provided with one or more holes for
charging and one or two holes for gases generated during smelting to
go out of the furnace. Modern furnaces are provided with self-baking
Sederberg-type electrodes since these are cheaper and they can be of
any desired dimension to carry the desired current. The electrode is
made by joining piece-by-piece steel sheet cylinders with welded ribs
from inside to support the electrode paste while baking. The electrode
paste is made from coke or calcined anthracite and bituminous binder.
As the electrode is consumed and is allowed to slip in, fresh steel cyl-
inder is affixed to it from top and is held in position, and the electrode
paste is filled in this. The paste gets slowly baked in situ to form a hard
electrode before it is consumed. The steel frame provides a suspension
link for the electrode grip-and-slip system.
Most submerged arc furnaces are fed with premixed charges that have
to be proportioned accurately be weighing each raw material individually.
The dried raw materials are stored in bins equipped with constant-weight
feeders, belt weighers, or a scale car. Pre-mixed and weighed charge is
raised to the furnace top either by skips or belt conveyors and is fed to
the furnace through pipe openings provided in the furnace roof. Weighed
quantity of raw materials can be layered on a conveyor belt, and proper
mixing can thereby be achieved. If a scale car is used, its contents are tipped
into a skip hoist to raise it to the charging level where these are mixed in the
bins and fed to the furnace under gravity through one, two, or three charg-
ing pipes or chutes.
Tipping of charge twice or thrice during this process is adequate to give
reasonably homogeneous mixing.
The furnace may be open or closed from top. In open-top furnaces,
the charging pipes have two gates or two valves that work in much the
same way as the two bell system of a normal blast furnace. In the open-
top furnaces, the furnace gas, which is a rich fuel, is burned at the top
of the furnace and is wasted. A fume-removal hood is provided in the
open-top furnace and is put under suction by connecting it to a stack.
This system gives out fumes through the stack. On the contrary, the
closed-top furnaces are choke-fed (i.e., the charging pipes are kept full
of charge to provide gas seals). The gases are led to cleaning through
openings provided in the roof or in the top portion of the furnace body.
The cleaned gas is used elsewhere as a fuel.
Modern anti-pollution restrictions necessitate gas cleaning before let-
ting them off in the atmosphere. Since the actual amount of gas let off
through a stack in an open-top furnace is almost 40–100 times more in
volume than the actual gases generated by the furnace reactions, the load
on the gas cleaning system unnecessarily increases, which make the gas
cleaning equipment prohibitive in terms of costs. The closed-top furnace
design is therefore generally preferred. The gas recovered from the closed-
top furnace can be used as a fuel for secondary uses.
One or two tap holes may be provided, depending upon the size of the
furnace, to tap the liquid iron and slag. The metal and slag flow out from the
same hole and are separated afterwards outside the furnace. A simple slag
dam can divert slag from the launder into the slag line while the metal flows
through underneath the dam. Molten iron may be sent to the steelmaking
section or may be covered into pigs in a pig-casting machine.
20.4 Operation
The reductant in electro-thermal smelting does not have to support a
tall burden as is necessary in a normal blast furnace. It should therefore
theoretically be possible to use any carbonaceous material as a reduc-
tant. The reductant still has to perform a dual function of being the
reducing agent and also the resistance for the generation of heat within
the furnace.
The reactivity of the reductant is important since reduction of iron
oxide has to take place directly by carbon in relatively shorter duration.
The electrical conductivity of the reductant is also equally important
since during smelting, a bed of reductant is formed under the electrode
and its resistance to passage of current, in fact, generates most of the heat
produced in the furnace. The size and proportion of reductant influences

the smelting temperature, and these are adjusted to obtain the desired
hearth temperature.
The reactivity and the electrical conductivity of the reducatant are
therefore important parameters, and the choice of the reductant has to
be made in the light of its above properties. The temperatures estimated
by Anderson [86] are shown in Figure 20.2. It is seen that most the of the
heat is developed within a meter from the electrode tip. The size of the
reductant bed must be under the electrode, and its electrical conductiv-
ity thus decides the temperature in the furnace. The proportion, quality,
and the size of reductant mixed with the charge must therefore be closely
controlled to obtain the desired optimum thermal profile in the furnace,
which in turn will decide the extent of furnace reactions and finally the
composition of pig iron produced.
FIGURE 20.2 Estimated temperature distribution in an electric arc furnace for smelting of iron ores.
In Figure 20.2, the charge next to the walls of the furnace is a solidified
layer of slag, which must be formed by even adjusting the practice in order
to protect the lining.
20.5 Smelting Practice

The charge consists of ore, coke (or cheaper fuel), and limestone. The
chemistry of the charges is adjusted to finally produce a thin, free-flowing
slag at the operating temperature of the furnace (i.e., at the bottom). The
general considerations in adjusting the chemistry of a blast furnace charge

hold true here also. Coke or fuel rate is almost half that of the normal blast
furnace. The flux requirement is therefore reduced in proportion to the
decrease in the actual reduction of ash in the charge.
Higher basicity causes hanging in a normal blast furnace, whereas no
such problem is posed in electric smelting, and hence higher basicity up
to 1.7–1.8 can also be employed, depending upon the sulfur load in the
charge. Considerations of charge for producing basic iron and foundry
iron differ. For producing foundry iron, higher basicity, nearly 1.6–1.8,
is preferred, with higher proportion of reductant as compared to those
for basic iron.
The charge need not be as lumpy as is necessary for a blast furnace, but
the proportion of −3 mm fines do affect the operation adversely, and hence
it should be eliminated by screening. The ore fines can then be agglomer-
ated using coke fines in the usual manner.
The use of self-fluxing sinter has been very encouraging. It decreased
power consumption by 10%–15%. Pre-heating the charge by using furnace
gases has also proved to be beneficial in reducing the power consumption
by up to 25%.
Reference to Figure 20.2 reveals that a small zone just above and around
the electrode tip gets pre-reduced by the CO produced by direct reduction
of iron oxide. The indirect reduction can contribute up to 20% reduction,
which is reflected in the furnace gas composition. If all iron oxide is reduced
by carbon the CO content of furnace gas can be above 90%–95%. But as
the proportion of indirect reduction increases, more CO gets converted to
CO2. The furnace gas normally analyzes:
CO.......... 72%–75%
CO2.......... 13%–15%
H2.......... 7%–10%
N2.......... 1%–3%
Hydrocarbons.......... 2%–3%
If highly reducible self-fluxing-type sinter is used, the CO propor-

tion can fall to 50%–52% and CO2 can rise to 40%. The CO2 content is
a true reflection of the proportion of indirect reduction taking place.
Pre-heating of charge also helps indirect reduction. The gas normally

generated in the furnace with 72%–75% CO has a calorific value of
approximately 2500 kcal/Nm3, and nearly 600–700 m3 of gas is gener-
ated per metric ton of pig iron.
The electric power consumption is generally 2200–2500 KWH/t for
basic iron, and it increases up to nearly 3000 KWH/t for foundry-grade iron.
The analysis of slag very much depends on the burden chemistry,
that is, the composition of ore, limestone, and ash of the reductant used.
It therefore varies from practice to practice and plant to plant. As an
illustration, the general analysis can be given as follows:
CaO.......... 40%–45%
SiO2.......... 25%–30%
MgO.......... 2%–10%
A12O3.......... 5%–25%
The analysis of iron varies considerably. It is saturated with carbon, and

the silicon varies with operating temperatures. Entire phosphorus in the
burden is reduced to join the metal (except the loss as flue dust), and much
of the sulfur will be eliminated by maintaining high basicity; it may be as
low as 0.05%.
The furnace is generally tapped every 1.5–4 hours, depending upon
the practice.
The operation and control of an electric furnace is rather tricky. This is
because the furnace is operated to produce a slag with nearly 1.5 V-ratio. If
Figure 10.8 is recollected, then the way the ratio of alumina to silica of the
ore shall move, to form the slag, due to flux addition, from the point on the
line joining the two corners towards the lime corner shall be obvious. The
area through which this line must pass is the lower liquidus temperature
area of the ternary system CaO-SiO2-Al2O3. The charge composition and
the electrical conductivity of the charge materials have to match such that
the slag composition falls within the low liquidus temperature region. In
this respect, the electric furnace offers a very narrow area of slag on the
basic ternary diagram. That makes the smelting operation and therefore its
control rather a tricky issue. It needs a great deal of experience to run the
ironmaking electric furnace.
20.6 Modern Trends

In electro-thermal smelting, the modern trend is to use closed-top fur-
naces and provide facilities for cleaning the furnace gas to recover the
unused CO for its subsequent use as a fuel. One of the important devel-
opments in electro-thermal smelting has been the use of furnace gas for
pre-heating or pre-reduction of the iron ore charge in a rotating kiln
located just above the furnace or beside the furnace. The pre-heated
or pre-reduced charge is fed in hot condition, out of contact with air,
directly into the furnace. The degree of pre-reduction is aimed at 50%
removal of total oxygen content of the ore. The Skopje Steelworks in
Yugoslavia and Highveld Steel and Vanadium Corporation at Witback,
South Africa already have a total of nine furnaces with a 30–45 MVA rat-
ing with such pre-treatment kilns.
20.7 Special Features

A sizable proportion of iron occurs in nature as titano-magnetites, which
cannot be smelted in a normal blast furnace because of the tendency of
formation of titanium carbides, nitrides, and other infusible compounds,
which tend to accumulate in the bosh and can eventually block the fur-
nace. The nitrogen atmosphere facilitates smelting of titaniferous materi-
als without any smelting problems. Most of the titaniferous iron minerals
contain vandium or other valuable constituents whose recovery make the
smelting process more economical. In fact, the Quebec Iron and Titanium
Corporation at Sorel, Quebec smelts titaniferous magnetites for the pro-
duction of vanadium-bearing slag as their main product, while pig iron
forms the by-product. In this process, titano-magnetites are smelted with
less of reductant to even allow a few percent of iron to go to slag. The FeO
content of slag, up to a few percent, then ensures free-flowing slag contain-
ing large amount of TiO2. Most of the vandium is reduced during smelting
to join the pig iron. The pig iron is then blown with oxygen in shaking ladles
when the first slag that is formed contains most of the vandium. This slag is
separated for further processing and recovery of vanadium while the iron is
blown to steel in the usual manner.
The titano-magnetites of Masankere in Karnataka have been success-
fully smelted for recovery of vanadium and production of steel in one of the
electric furnace of VISL, Bhadrawati in trials over extended duration, and
this is going to be adopted as a routine practice.
20.8 Irregularities in Operations

An excellent account of irregularities in an electric furnace smelting, observed
at the VISL plant, has been given by Kumarswami. He quotes slag boiling,
non-slipping of electrode, breakage of electrode, explosions, tap-hole leakage,
et cetera, as the important irregularities in electric smelting of pig iron.
Normally, a small coke bed formed at the bottom of each electrode is
responsible for the maintenance of a steady current in the electrode. If this
bed is disturbed because of inadequate addition of coke or addition of fine
coke, then the electrode tries to achieve stable condition by its up and down
movements. In doing so it may touch the slag and cause spurting, eventually
resulting in hanging of the overlaying burden and overheating of the slag in
the hearth. This is known as slag boiling. It can be rectified in the early stage
itself by suitable extra coke addition, or it can be remedied in the later stages
by judicious adjustments of the burden. In the worst circumstances, poking
may have to be resorted to by suspending the operation for several hours.
The electrode consumption in a 100 t/day furnace is nearly 30 cm per
day, which is made up by allowing slipping of the electrode in stages of 10
cm each time. Excessive quantities of reductant in the charge may not need
so much of slipping, but it tends to make the electrode work at a higher-
than-normal position. The heating of the charge at the bottom is affected
adversely and in the worst circumstances, the hearth may even solidify and
pose all the attendant problems. Correct charge proportioning is therefore
necessary for smooth running of the furnace.
The breakage of the electrode can be due to a variety of reasons. Since it
can cause considerable disruption in production, it must be minimized by care-
ful control of the electrode paste and following the details of electrode making.
Very fine size reductant is believed to cause explosions, which can be
disastrous. Careful control of the quality of reductant is therefore necessary.
Tap holes are generally made from carbon blocks that may wear out
from inside because of the action of molten metal and slag. A worn-out tap
hole starts leaking and causes difficult working conditions. A regular repair
and periodic replacement of the tap hole block can get rid of these troubles
and ensure smooth tappings.
Besides these, several other irregularities may occur in running an elec-
tric furnace. Careful control of the operation and rigorous routine mainte-
nance can minimize these and make the operation more smooth.

Electro-thermal smelting of iron ores is now an established commercial
practice for the production of iron where coke is scarce and use of electric-
ity is economically viable. The cost of pig iron made by this is still high,
and the process still needs moderation to make it a competitor to the blast
furnace. On the whole, electro-thermal smelting is more difficult and tricky,
in the sense that the fluid slag region is very narrow and restricted, and
furnace control has to be very effective to operate the furnace efficiently.
Few electric installations came up in the past several couple decades. And
now, mini-blast-furnaces are more economical than electro-thermal smelt-
ing, and hence the latter’s wider popularity, as is obvious from the previous
chapter.
CHAPTER
21
ALTERNATIVE ROUTES
OF IRON PRODUCTION III –
SPONGE IRON PRODUCTION
In This Chapter
● Introduction
● Physical Chemistry of DR Processes
● Sponge Ironmaking Processes
● Midrex Process
● Fluidized Bed Processes
● Coal-based DR Processes
● Coal-based Shaft Furnaces
● Gasified Coal-based Processes
● Sponge Iron Production in India
● Metallurgical Properties of Sponge Iron
● Use of Sponge Iron
21.1 Introduction
The production of iron is only an intermediate inevitable step in the ulti-
mate production of steel from a natural source such as iron ore. Where
coking coal is not available but other gaseous (natural gas) or solid fuels
such as non-coking coals are readily and abundantly available, an alterna-

tive method of iron production using other fuel is feasible. Production of
sponge iron is yet another alternative route of iron production besides those
described in Chapters 19 and 20.
Sponge iron is a term given to the product that is obtained by reduc-
tion of iron oxide ores in solid state. Because separation of iron pro-
duced from the gangue oxides is not readily possible, solid state sponge
iron is a mixture of reduced metallic iron, unreduced iron oxide, and the
associated gangue constituents of the original ore. It is variously known
as sponge iron, directly reduced (DR for short) iron, metalized ore, etc.
Directly reduced iron ore is neither iron nor ore, as either are normally
considered. Much of the oxygen associated with the iron oxide is elimi-
nated, thereby raising the content of metallic iron. Oxygen iron is fairly
big and hence its removal from the solid iron oxide lattice leaves the
remaining mass, after the reduction, more porous, and hence the term
sponge iron. The term directly reduced iron originates from the fact
that the iron oxide is directly reduced by the reductants (often solid).
The term metallized ore is derived from the fact that the metallic iron
content goes up considerably in these processes.
The percentage oxygen, associated with iron as oxide, removed
from the ore particle is often referred to as percentage reduction,
whereas the percentage of iron as part of the whole iron, existing as
metallic iron is called degree of metallization in the sponge ironmak-
ing technology.
The idea of production of iron by reduction of ore in solid state is at
least more than one hundred years old. In fact, the first patent stands in the
name of Henry Bessemer, the pioneer of modern steelmaking.
Interest did not arise seriously in this field until after the end of the
World War II. Thereafter a host of processes for direct reduction were sug-
gested but only a few could go beyond the laboratory level investigations.
Over these nearly 40 years, the underlying principles have not changed
greatly. However, the technical, economical, and social parameters that
decide ultimately the viability of the process on a commercial scale have
certainly changed considerably. Some of the processes that appeared prom-
ising for commercial applications after the war seem to have been aban-
doned in favor of a few others.
Alternative Routes of Iron Production III –Sponge Iron Production • 405
21.2 Physical Chemistry of DR Processes

The physico-chemical principles of reduction of iron oxide in solid state
are complex because, although the reduction occurs, the exact role played
by the gaseous and solid reducing agents in the reduction of iron oxide is
uncertain. Much of the mechanism and kinetics of iron oxide reduction
in the blast furnace stack hold true for direct reduction processes also.
In fact, the efficient nature of reduction of iron oxide in the blast furnace
stack, where a gaseous and/or solid reductants reduce iron oxide in solid
state, led to the development of many a direct reduction process carry-
ing out only the role of the blast furnace stack and the removal a partially
reduced product without fusion.
The reducing agents used in direct reduction are either gaseous and/or
solid carbonaceous materials. The main reduction reactions are:
Fe2O3 + 3CO = 2Fe + 3CO2 (21.1)
Fe2O3 + 3H2 = 2Fe + 3H2O (21.2)
Since hydrogen is too costly to be used for such commercial applica-
tions natural gas, wherever available and in reformed condition, is used as a
substitute for CO or H2. Reforming carried out as:
CH4 + H2O = CO + 3H2 (21.3)
This mixture of CO and H2 is used in some of the commercial direct
reduction processes. Other hydrocarbons such as methane naphtha, petro-
leum, et cetera, are used for reforming.
Solid reductants, such as non-metallurgical coals, when used, may carry
the reduction as:
Fe2O2 + 3C = 2Fe + 3CO (21.4)
Fe2O3 + 3/2C = 2Fe + 3/2CO2 (21.5)
All the above reactions (21.1) to (21.5) are endothermic and hence heat
must be provided for these reactions to take place. The heat transfer con-
siderations therefore play a very dominant role in the sponge iron produc-
tion processes. The effluent gases from these processes contain CO and
CO2 in a proportion commensurate with the actual reactions occurring in
the reactor and hence vary from practice to practice considerably.
21.3 Sponge Ironmaking Processes

The sponge ironmaking or DR processes can be conveniently classified into a
few categories, depending upon the type of reductant used, gaseous or solid
or both, the design of the reactor, and so on. These are summarized below.
Reductant Used Reactor Design Process Charge Type
Static bed HyL Sized lump/pellets
Reducing gas produced from Continuous retort Midrex –do–

natural gas or naphtha
Purofer –do–
Armco Pellets
Fluidized bed Hi iron Sized ore
Esso-Fior Briquettes
Nu-iron –do–
HIB, Novalfer etc. –do–
Solid and/or liquid or gaseous Rotary kiln SL/RN, Lurgi Sized lumps or
Krupp-Renn pellets
Retort Echevarria Pellets
Crucibles Hoganas Sized lumps/pellets
Out of all these, only some have found popularity in commercial appli-
cations. The rest either were confined to pilot plant scales or were commis-
sioned on a commercial level, but the very first plant could not continue
operation for long or, even if continued, the performance has not been sat-
isfactory for its further applications elsewhere.
These processes are described below under basically two categories viz.
the coal-based direct reduction processes and the gas-based direct reduc-
tion processes. Within themselves, further classification is possible with
respect to the type of reactor/furnace used.
Similarly, classification on the basis of reductant used is also possible, viz.:
—Coal based rotary kilns
—Coal based rotary hearth, retorts, etc.
—Batch type gas-based retort processes
—Continuous gas-based shaft-furnace processes

—Gas-based fluidized bed processes.
The popular processes (i.e., those proven as commercially viable) are HyL
(including their version III), Midrex and may be Wiberg in the gas-based cat-
egory. Whereas in the coal-based category the kiln processes are most popular
and in that also the SL/RN, CODIR, ACCAR, and DRC are more popular.
The SL/RN is the topmost choice for DRI coal-based reduction.
These are described below in more detail.
HyL Process
After several years of development work, the HyL process was patented by the
Hojalata y Lamina, S.A. and Pullman Incorporated and hence the name HyL
for this process. In this process sized iron ore or pellets are reduced by passing
reformed natural gas, downwards through a static bed of sized ore or pellets.
[105] The natural gas is desulfurized and reformed by mixing it with steam
and passing it over nickel as catalyst when reaction as in Eq. (21.3) occurs.
The reformed gas has approximately 75% H2, 14% CO, 3% CH4, and 8% CO2.
Reduction is carried out at 870° C–1050° C and reactions as in Eqs. (21.1) and
(21.2) are expected to occur. The charge needs to be pre-heated for efficient
reduction. The overall efficiency of the operation is improved considerably by
carrying out the process in four stages. These steps are:
1. Pre-heating the charge and partial reduction by hot gases coming from
the primary reduction step. This is also called secondary reduction step.
2. Primary reduction, in which the partially reduced charge (from the
secondary step), is further reduced by strong reducing gas coming
from the cooling stage.
3. Cooling of the reduced charge by fresh gas which itself gets pre-heat-
ed and thereafter will be available for primary reduction. This step
completes the reduction and allows controlled deposition of carbon,
which can be varied between 1.5%–2.25%.
4. Discharging of the sponge and recharging of the reactor with fresh
charge to start the process all over again.
Four reactors are therefore employed and in each of the above mentioned
operations are carried out in the sequence mentioned above.
The individual reactor operations are so staggered that, at any instant,

each of the four reactors is in one of the four stages of the cycle. The opera-
tion at any one instant is shown in Figure 21.1. A minimum of one group
of four reactors is therefore essential to carry out the process. The capacity
can be increased either by increasing the individual size of the reactor or by
multiplying the units into a battery of units.
FIGURE 21.1 Scheme of the HyL process.

The HyL process has been proved to be commercially a viable proposi-

tion for the production of sponge iron. Nearly 5.6 Mt of sponge produced
by this process has already been converted to high quality steels. The first
plant of its type was commissioned in 1957 in Monterrey, Mexico. The fol-
lowing is the description of the commercial HyL plants that were set up in
earlier years[103].
Owner Location Actual Rated Capacity Year of

t/annum Commissioning
Hojalata y Lamina, S.A. Monterrey 105,000 1957
Mexico
–do– –do– 220,000 1960
Tubos de Acero de Vera Cruz 175,000 1967
Mexico, S.A. Mexico
Hojalata y Lamina, S.A. Pueblo 245,000 1969
Mexico
U.S.I.B.A. Bahia 245,000 1972
Brazil
Govt. of Morocco Morocco 245,000 1973
This process is highly flexible. It can be operated at the optimum cycle of

12 hrs. and 90% metallization to produce 1 Mt of sponge iron per annum. The
same plant can be operated to produce 870,000 t/annum at 95% metallization
by increasing the cycle to 14 hrs. and 1.165 Mt/annum with 70% metallization
on an 8-hr cycle without any design modification in the plant. The earlier gas
consumption of 25,000 cu ft. has been lowered to 20,300 cu ft. per metric ton
of iron produced.
The HyL plant has several unique features in its design. The rich fresh
gas is used to reduce the already pre-reduced charge, and the cooling is
accomplished by passing fresh reducing gas through the charge so that the
reduction is achieved to the desired extent with minimum of gas consump-
tion and without the danger of reoxidation of the sponge that is produced.
The cooling stage acts as a recuperator to pre-heat the reducing gas and
thereby maximizes the thermal efficiency. Even after the use of gas for sec-
ondary reduction stage it still contains appreciable quantities of CO and H2
and hence it is used as a by-product fuel in the plant. It can also be used in
the process itself for heating by mixing it with additional natural gas. The
reaction temperature may be varied between sized natural ore since sticki-
ness is appreciably reduced. The HyL process product, because it is cooled
in a stream of fresh reducing gas, is not found to be pyrophoric and hence
can be transported over long distances without any danger of reoxidation of
iron. Reformed naphtha can be used if natural gas is not available.
After successful installation of several plants within a short period after
the first successful demonstration of the commercial plant in Mexico, the
originators were led to improve upon the process by continuous commer-
cial research. All this culminated in the development of the HyL II as the
modified version of the first design, and later in 1979, the HyL III was put
into commercial practice.
The HyL design and process were further improved into the HyL IV
M process wherein the separate gas reformer is eliminated. Gas reforming
is carried out inside the shaft reactor. The presence of very porous metallic
sponge iron as a reduced product in the reactor acts as an efficient catalyst.
Here both the reforming and the reduction reactions take place inside the
same reactor. The recycled gas is mixed with natural gas and the mixture
is pre-heated to 900° C, and on its way to the reactor a controlled amount
of oxygen is introduced in it to burn some gas and produce required heat
to raise the reducing gas temperature to 1000° C–1050° C. This also helps
to carburize the iron to the extent desired. Like the HyL III, the HyL IV
also uses humidifier so that hydrogen is produced during reforming and is
available for the reduction reaction. The product of hydrogen reduction is
steam, which is condensed from the exit gas. The carbon dioxide is removed
by compressing the gas. In case the product is to be cooled, then part of the
natural gas for reduction is injected cold from the bottom discharge end to
cool the sponge iron and thereby get itself pre-heated.
This is characteristic and quantum improvement in sponge ironmaking.
This has motivated other gas-based sponge-ironmaking processes to adopt
self-reforming of gas in place of using an external reformer, as has been the
practice in older units.
In the HyL III, the four fixed-bed reactors are modified into one single
moving bed reactor using the same gas reforming plant, auxiliary equip-
ment, etc. The material moves down in a vertical reactor. The mechanism
of the process is just the same as in batch reactors. The zones are separated
by using a isobaric zone in the middle. This modified version requires less
energy, can accept much wider grades of raw materials, lower size gas
reforming unit is required, higher sulfur input is tolerable, and so on. Steam
is used for reforming the gas and it results in higher hydrogen content of
the reducing phase and as a result the reduction rate is faster. In 1991, the
HyL III plants produced 5.4 Mt DRI of steelmaking grade.
The process efficiency has been improved so much that the HyL III
requires only 10% –15% more energy than theoretically required value.
It is necessary to have carbon associated with sponge iron as a product.
It helps its melting during steelmaking. Special efforts and additional gas
expenditure are required to get carbon containing sponge iron. The actual
carbon content varies, but it is generally in the range as above. Carbon is
made to react with iron to form iron carbide at a temperature where this
can take place efficiently. In gas-based plants there are reducing gases avail-
able but not carbon as such to form iron carbide. Carbon is therefore delib-
erately generated through soot formation as per the reaction:
2CO = CO2 + C
or through interaction of methane with hot iron to form iron carbides.
This CO/CO2 reaction is possible approximately 600° C temperature.
In gas-based processes such as Midrex, the reducing gas such as naphtha
is additionally introduced at this temperature level in the shaft. This is an
extra cost item and the cost goes up with increasing carbon content require-
ment in the sponge.
21.4 Midrex Process

The process essentially consists of reducing the ore pellets by reformed natu-
ral gas in a vertical shaft furnace. It is a continuous countercurrent gas-solid
reactor resembling more or less the stack of a normal blast furnace. The indu-
rated pellets are led in the shaft from the top in lots and the sponge iron, in
cooled condition, is removed from the bottom. The reducing gas composition
is similar to that used in the HyL process. An excellent system for off-gas
recovery and circulation, as shown in Figure 21.2, have been developed.
In Midrex process the reformed gases enter the reduction furnace at
approximately 800° C–870° C and generally contain 55% H2 and 33% CO.
These gases reduce the iron oxide to 92%–93% metallization. The spent
gas is cooled scrubbed and nearly 60% is used for reforming, whereas the
remaining 40% is available as a fuel for use elsewhere in the plant. It is
often used for reheating the reformer tubes. The carbon content of the
product is readily controlled within 1.0%–2.0% to best suit the steelmaker.
The product of Midrex is strong enough not to degrade during reduc-
tion and handling. However 5%–8% fines of −3 mm are generated and are
screened off. These are either hot briquetted, cold briquetted or directly
melted in captive electric furnace. The briquetted can be sold as product.
The fines are generally better metalized.
One of the requirement for smooth operation of Midrex is that the
charge descends smoothly and does not hang up due to fusion of particles
on the surface resulting in diffusion bonding and sintering into large lumps.
Pellet feed is therefore preferred because of their minimum area of contact
of particles. However, many Midrex plants run on lumpy ore charge alone.
In that case the size is usually +6 and −40 mm with major fraction of 10–25
mm size. A mixture of lumpy ore and pellets is also common. In that case
30%–50% of lumps is preferred.
Specific consumption figures are as follows: (per t of DRI)
Pellets/lumps of ore (min 66% Fe) 1.41–1.45 t
Natural gas equivalent of 9.5–0.5 GJ
Electrical power 100 Kwh
Water 1.0–1.5 m3
This process can accept high sulfur natural gas and can still produce
good quality DRI and run the process efficiently. Several chemical treat-
ments are available to make the sponge inactive for transportation over long
distances without much reoxidation.
Midrex is the most widely adopted process of making DRI all over the
world. In 1990, 61% of the total world production of DRI came from Midrex
process alone, purely because of the very high capacity individual units.
The unit-size of Midrex plants are far more than coal-based vertical or hori-
zontal kilns. The productivity of the Midrex process is therefore far more than
any other process. This is due to the gas-solid reduction kinetics of Midrex being
much faster than the solid-solid reduction kinetics of coal-based processes.
Other Gaseous Reduction Processes

Besides HyL and Midrex, several other processes were developed for
producing DRI using gaseous reduction. These are Wiberg, Plasmared,
ARMCO, Purofer, NSC, etc.
They are essentially similar in principle in their design and operation.
All them use natural gas, plant off gases such as coke oven gas, cracked fuel
oil, and the like as reducing agent. Generally approximately 60% of the off
gas from reduction zone is reformed using nickel catalyst with steam and
fuel supply gas. Their individual designs differ with respect to cycle times,
% gas recycled, charge mix (pellets or lumps or their mixture), quality of
charge, sulfur load and ways of eliminating it for avoiding contamination of
the catalyst and so on. They also differ in temperatures at which reduction
is carried out. The mode of heating of the charge may also differ, either by
the fuel reducing gas itself or by electrical energy.
Out of these, Wiberg did not flourish beyond its first installation, Plasmared
could not become a workable independent unit and it had to be used as an
adjunct facility with other process, ARMCO abandoned its efforts at pilot plant
stage itself, Purofer discontinued its interest after failure in its first installation,
NSC failed to show commercial viability of their plant in Malaya and so on.
A modification of Midrex process in the form of “AREX” has been set
up at SIDOR in Venezuela. In this, the reforming of gas is carried out using
hot DRI itself as catalyst. In other words, the reduction and reforming take
place in a single reactor. The hot DRI is subsequently used in a steelmak-
ing process using a smelting furnace (to be described in next chapter). This
coupling has been found to be more economical in terms of energy con-
sumption productivity, operational costs, carbon control, and so on.
Gas-based Processes
All gas-based DR processes work on counter current principle wherein the
gases pass through either a static permeable column of solid ore or the solid
ore and gas move counter to each other. The charge permeability is main-
tained properly by control on size of the charge. Pellets are therefore pre-
ferred because this charge can maintain burden permeability more readily.
This design exploits the advantages of high efficiency of countercurrent gas-
solid reactions. The small column height, or bed height, in the reduction zone
is effective in giving faster reduction rates at more uniform temperatures.
FIGURE 21.2 Scheme of the Midrex process.
The energy requirement at the Portland plant (USA) is nearly 3.5 Mk

cal per metric ton of iron production and approximately 90% of this is met
by using natural gas. This efficient use of energy inputs may partially explain
the unusually rapid acceptance of the Midrex process on a commercial level.
Midrex direct reduction plants are owned and operated by Midland-
Ross Corporation. These plants are provided with facilities for the manu-
facture of indurated oxide pellets by any one of the conventional procedure.
Reduction may, however, be carried out on purchased pellets as well.
Midrex plants are at the following places.
Portland, U.S.A........... 350,000 t/annum
Georgetown, U.S.A........... –do–
Hamburg, F.D.R. Germany.......... 350,000 t/annum
Contrecoeur, Canada.......... 400,000 t/annum
Louisiana, U.S.A............ 800,000 t/annum
Osaka, Japan.......... –do–
ISPAT, India.......... 100,0000 t/

annum
The average normal size of these plants is 1200 t/day and in the latest
plants the same has been nearly doubled.
Processes such as Wiberg, Purofer, Novalfer, Armco, ICEM, among

others, also work on the same principles as that of the Midrex. The Midrex
process, however, has achieved an impressive degree of popularity in a rela-
tively short-time.
It produces 92%–95% metallized pellets from good-grade ore.
21.5 Fluidized Bed Processes

In this category fall the processes known as FIOR (fluidized iron ore reduc-
tion), HIB or Orinoco, Novalfer, et cetera. These processes reduce iron
oxide fines by a reformed gas in a fluidized bed.
In FIOR, the reduction is carried out in series in four fluidized bed
reactors. The first reactor pre-heats the ore. Reduction is started in the
second and is continued in the third and completed in the fourth reac-
tor. These reactors operate at approximately 10 atmospheres, pressure.
The reducing gas is rich in hydrogen (hydrogen reduction of iron oxide
is faster than that by CO) and is generated by steam reforming of natural
gas. The reduction takes place at approximately 800° C. The fines are
discharged at approximately 650° C and are hot compacted into pil-
low-like briquettes. The first and the only plant was commissioned at
Matanzas in Venezuela in 1976. After several problems it is reported to
be working successfully.
HIB (high iron briquettes)—the original name being Nu-Iron pro-
cess—is fairly similar to FIOR in its design and operation. Reduction is
started at approximately 700° C and is completed at 870° C in the fourth
stage. The product is hot briquetted. Although iron content is high, its met-
allization is low. Only and the first plant of this type is in operation at Puerto
Ordaz in Venezuela. The product, being low metallized, is used as a blast
furnace burden.
Although advantageous for recycling the ore fines, plant wastes and
so on these processes have not been able to come up in competition with
other DR processes. One of the key parameter for improving DR pro-
cess economy is to reduce energy consumption in these processes. The
fluidized processes require relatively more energy for their operation.
Future of fluidized processes is not all that bright.
21.6 Coal-Based DR Processes

Alternative to gas-based reduction reactors are the coal-based reduction
reactors. This is the only solid fuel commercially available for such pur-
poses. Non-metallurgical cheaper coals are readily available, and processes
have been developed wherein carbon of such coals is used to provide heat
as well as the reducing gas. Iron oxide in contact with carbon when heated
results in the reduction of oxide to metallic iron. This reduction is more due
to the burning of carbon first followed by reduction by the carbon monox-
ide gas generated from burning. This is the reason why these processes are
termed as coal-based processes. These are more popularly carried out in
horizontal rotary kilns. But vertical shaft furnaces have also been used as
reactors for this process in some rare cases.
This is also a countercurrent process to the extent that solid oxide trav-
els downwards and the gases formed on burning of coal ascend with respect
to the inclination of the kiln or the shaft reactor.
A long kiln, easily 70–80 m in length and few meters in diameter,
slightly inclined to the horizontal and rotating slowly around its own axis is
employed as a reactor in these processes. The individual designs do differ
in details. The charge is fed from that end which is at a higher level. The
charge travels under gravity aided by the rotation motion, through several
heating zones and the reduced iron oxide product comes out of the other
end of the kiln. The throughput rate of solid iron oxide, its reducibility,
its size, gangue contents, the rotation of kiln, and other related factors
are adjusted such that the oxide is reduced to the extent of 93%–95% by
the time it travels the whole length of the kiln. This reduction is slow and
continuous and takes place over the length of the kiln. In this process, the
iron oxide gets heated to a maximum of 1050° C. The temperature profile
in the kiln over its entire length is adjusted and controlled using second-
ary blowers at appropriate places. The product comes out of the kiln at
high temperature and, being non-equilibrium material, is very prone to
reoxidation. It is therefore cooled in rotary coolers to almost room tem-
perature before it comes in contact with atmospheric oxygen. The scheme
is shown in Figure 21.3.
The most critical factor to be controlled in this process is the burning
of coal and its conversion to CO. This is achieved by controlling the total
air blown in the kiln and its division at various places of blowing. Primary
air is supplied through the discharge end and the secondary pipes at vari-
ous points across the whole length while the kiln is rotating. The charge
gets pre-heated to more than approximately 30% of its length from the
charge-end (i.e., by that time it attends nearly 800° C). Coal is burned
after this zone.
FIGURE 21.3 Scheme of rotary kiln plant.
In general, all the coal is introduced at the charge-end, but under cer-
tain situation a part is introduced with the charge and the remainder at
appropriate place where temperature is above 800° C.
Much of the volatiles in coal leave the kiln without taking part in the
reduction process. Hence, high volatile coals are not suitable for coal-based
kilns. It is the fixed carbon that is useful in this process and it should be high
in the coal used for coal-based reduction processes.
General Design Features

There are all types of designs available for coal-based horizontal rotating
kilns. However, to get some idea of these features, the details of a 100 t/d
output kiln are given below:
Length ~ 40–45 m
Air blown ~ 2000 m3/hr
Charge thickness ~ 600 mm
Charging rate ~ 200 t/d
Coal required ~ 1.2 p-1.5 t/t sponge

iron (with minimum
40% FC, <25% ash and
approximately 30% VM)
Ash fusion temp. ~ above 1200° C
Output ~ 100 t/d
Coal feed ~ 5–6 t/hr or 130–140 t/d
Residence time of ore (inside the kiln) ~ 7–8 hrs
Reduction time ~ 3–4 hrs
Pre-heating time ~ 3–4 hrs
Coal required for pre-heating ~ 60%
Coal through burner for heat ~ 40%
and CO for further reduction
Such a plant shall produce approximately 30000 t of sponge per annum.

This is a typical plant operating under Indian conditions today.
Raw Materials
The quality of raw material used decides the quality of DRI produced
by this process. This material is used for steelmaking and hence what-
ever factors related to ore that give difficulties in steelmaking have
to be minimized while selecting the raw material. For example, the
higher the gangue content and the more the alumina and silica, the
more will be the slag generated in steelmaking and hence it must be
low. In other words, iron ore with higher iron content, preferably more
than 65%, is preferred.
Phosphorus content of the ore has to be considered such that finally

there is no problem is getting the correct phosphorus specification in final
steel produced by melting this DRI. So is the case with sulfur and sulfur
content of the coal.
The optimum size of the ore fed is 8–20 mm with mean size approxi-
mately 12 mm. Although pellets of 6–22 mm are preferred, the majority are
in the range of 9 to 16 mm. The ore size should be such that it should not
lead to more fines in the product.
The reducibility of ore or the pellets should be adequate. Similarly,
degradation during and after reduction should be low. If this is not low,
the fines content of the product is high. DRI of less than 3 mm size are
not preferred, and these have to be compacted into briquettes before use.
Generally 80%–85% of the product must be above 3 mm in size. The charge
quality, its decrepitation, reduction degradation, abrasion resistance, and so
on, have to be adequate to be able to produce good quality DRI.
The coal is assessed in terms of the following:
1. Proximate analysis
2. Grain size and moisture content
3. Reactivity
4. Ash softening temperature
5. Sulfur content
6. Coking and swelling indices
7. Calorific value
Because of their high reactivity, lignites are found better as reductant in
these processes.
Limestone or dolomite of 1–4 mm size is used as a flux to fix up the
sulfur load of the charge. The smaller size is more effective in sulfur fixation
because of the gas-solid kinetics involved in this reaction.
As has already been discussed with reference to the Midrex process,
the carbon content of sponge iron is always specifically asked to help melt it
efficiently in the electric furnace for steelmaking. It is also beneficial to pre-
vent its oxidation during transportation to steelmaking plant. Although in
coal-based processes carbon is in direct contact with the sponge iron at high
temperatures, deliberate attempts have to be made to make up the carbon

specifications. This sluggishness in dissolving carbon in sponge iron is due
to the inherent slow process of solid carbon dissolving in solid iron, primar-
ily due to lack of adequate solid-solid contact. It also adds up to the cost of
sponge iron as a product. The more carbon, the higher the price will be.
Product
The product, even after cooling, is a mixture of DRI solid flux, ash, and
char of coal. DRI is magnetic and hence is separated from the rest by mag-
netic separator. Screening removes the fines, which can be sold as it is or
in briquette form. DRI contains or is allowed to pick up carbon so that it is
stabilized, and the carbon is useful subsequently in steelmaking.
The kiln rotation and fall of hot particles due to the cascading effect
generally results in rounding off the hot ore particles and hence the product
is smoother than the ore fed. This partially seals off the surface porosity and
protects the DRI from quicker reoxidation.
Ring Formation in a Rotary Kiln

At some stage, the temperature is such that some constituents of the charge
soften. For example, in the wustite stable region, fayalite is the low-softening
compound that readily makes the particles stick to each other as well as to
the lining. There are several such compounds that can possibly form and cre-
ate conditions such that the material sticks to the kiln wall and slowly a ring
is formed as a result of this accretion. This is the ring in such kilns. This ring
reduces the effective diameter of the kiln and interferes with the reduction
process. Productivity is decreased and when the ring is too large the opera-
tion has to be stopped, the ring is removed, and then only it is possible to
continue using it. This results in more down time and loss of production.
This has to be assessed in terms of softening behavior of the ore, qual-
ity of gangue, quality of ash, quality of flux, and so on. The working of the
kiln has to be controlled to minimize ring formation. Any excess tempera-
ture inside the kiln can hasten ring formation. The inclination, speed of
rotation, charging rate, rate of reduction, and so on, are all related to ring
formation phenomenon.
Coal-based Process Designs

The most popular design is the SL/RN process developed in 1964. In 1991,
out of the total coal-based DRI produced in the world, the share of SL/RN
was 56%. Nearly 0.83 million metric tons were produced. The product is
cooled by external water cooling.
The CODIR (coal-ore-direct-reduction) is another design. In this, the
product is cooled by water mist injected directly on to the DRI under very
controlled conditions so as not to allow reoxidation of DRI.
The rotary kiln of Sunflag Iron and Steel Co was installed at Bhandara,
Maharashtra. It is 80 m long kiln and 5 m in diameter and was commis-
sioned in 1989. It has completed continuous production of 138 days. Coarse
coal is injected from the discharge end with sizes of 5–25 mm. This may
amount to 60%–80% of the total requirements.
The ACCAR (Allis–Chalmers Controlled Atmospheric Reduction) pro-
cess was developed in 1972 in which natural gas or oil can be used singly or
together to supplement coal as fuel and reductant. Coal is still above 80% of
the total requirements. Fluid fuel is introduced at the discharge end. This
gives higher carbon content of the product—as is what a steelmaker desires.
The DRC (Direct Reduction Corporation) process was developed as a
modification of the kiln earlier used for removing iron from titanium ores
by pre-reduction. It commissioned its first plant in 1978 in the USA. In this
also a part of coal is injected from discharge end.
In India, the TDR (Tata Direct Reduction) process was developed and
put into practice by the Tata’s under the name IPITATA.
Several modified versions were also developed by several players in this
field to take care of their own specific requirements.
Process Control
A huge amount of information is available about this technology with respect
to material movements inside the kiln, reduction rates, rate of accretion,
kiln geometry, speed of rotation, charging rates, temperature profiles, and
so on.
The following variables are controlled in controlling the overall process:
1. Charge mix and rate of charging
2. Temperature profile inside the kiln
3. Rate and place of secondary air supply
4. Gas pressure inside the kiln
5. Kiln rotation speed

6. Exit gas temperature and DRI temperature at exit
7. Waste gas temperature and composition
8. Inclination of kiln and retention time of charge inside
9. Coal proportion introduced from each end
The charging rate, quality of charge, temperature profile in the kiln, exit
gas temperature, and discharge rate are all integrated in such a way as to
obtain the correct working conditions and thereby producing right quality
of product and at the required rate of production.
21.7 Coal-Based Shaft Furnaces

Vertical shaft furnaces also have been used as a reactor using coal as reduc-
tant. Two main processes have been commissioned on a commercial level.
The KM (Kinglor-Metor SpA) process was developed in 1971.
It consists of three twin vertical self-supporting retorts, housed in a fur-
nace. The charge travels down under gravity through pre-reduction, reduction,
and cooling zones. The total residence time of the charge inside the furnace is
15–18 hours. Coal in the charge burns, generates heat, and reduces gas.
Two such plants, one in Italy and the other in Burma, were set up but
had to be discontinued finally.
A modified Midrex using coal as reductant is the Midrex Electrothermal
Direct Reduction Process. It is a vertical shaft furnace divided into three
zones to carry out three different function viz. pre-heating, reduction, and
cooling. Heat required for the process is generated by passing electricity
through the charge. It has not yet been tried on a commercial scale.
21.8 Gasified Coal-Based Processes

Since iron oxide reduction requires reducing gas and since gas-reduction is
faster and cleaner, some techniques have been developed to produce gas from
coal and then using this gas, iron oxide is reduced to produce sponge iron.
These are exclusively of auto-thermal type (i.e., coal gassified by its interaction
with steam or oxygen). The use of steam in this case results in generation of
hydrogen in turn decides carbon monoxide. Hydrogen is faster in reducing
iron oxide. Although it is advisable to assess every possible technical alter-

native finally, the alternative has to be economical. Use of a coal gassifier is
uneconomical because coal directly can be used for reduction. At present, coal
gassification for DRI production is unlikely to be adopted.
The smelting-reduction processes, however, have used this concept in
developing processes wherein molten iron is produced as a product.
21.9 Sponge Iron Production in India

India is blessed with large reserves of good-grade iron ores but is unfortunate
with respect to good-quality coking coals. Fortunately, large reserves of non-
metallurgical coals are available in India. Considering these aspects, atten-
tion was focused on to the coal-based direct reduction processes to augment
iron production in the 1960s. Accordingly, a kiln was set up at Kothagudam in
Andhra Pradesh to produce sponge of 120 t/d with the help of UNIDO, in 1980.
Another plant using ACCAR technology was set up soon by Orissa Cement Ltd.
Since rotary kiln processes are highly sensitive to type of raw materials
NML and Tata set up their experimental kilns to study this process under
Indian conditions. Tata’s study culminated in setting up their commercial
plant as IPITATA. Their process is known as TDR.
Meanwhile, natural gas from the west coast became available and was
used to put up gas-based plants by Essar, Ispat India, and Vikram Ispat.
After liberalization in 1991 and because of a surge in demand for steel,
nearly 300 coal-based rotary kilns have been installed for sponge iron pro-
duction in India at places distributed all over India in the iron ore belts.
21.10 Metallurgical Properties of Sponge Iron

Being primarily a solid state reduction process, the original shape of the
iron ore particle is more or less retained in the final product with approxi-
mately 30% reduction in its weight due to the loss of oxygen from iron oxide
lattice. The density changes from approximately 4.4 g/cc to 7.8 g/cc. This
amounts to nearly 54% reduction in volume but this is manifested in the
formation of pores and not in outer gross shape. This is the chief reason for
calling it sponge iron. It is very prone to reoxidation because of its spongy
nature (i.e., large surface area/unit weight). It is absolutely necessary to
have certain amount of carbon in the sponge iron to
FIGURE 21.4 Theoretical iron content in DRI for given % metallization.
prevent its reoxidation on storage and to efficiently melt it in electric fur-

naces. Usually the carbon content is maintained in the proportion of 0.023%
C for every 1.0% of iron oxide in the sponge. This ensures adequate car-
bon-boil during its melting, which prevents iron oxidation to a large extent.
The carbon varies from 0.1–0.15 in coal-based DRI to as high as 2.5% in
gas-based DRI. The metallization achieved depends upon the quality of
ore and the processing parameters. Usually +65% Fe ore is preferred to
produce sponge iron. However, lower grade ores are often used. The lower
the grade, the lower will be the % iron in the sponge iron.
The iron content of the ore charge decides the Fe2O3 (assuming it to be
hematite) content and the balance shall be the gangue (assuming it to be dry
and volatile matter–free). For a given iron content, the amount of gangue
is thus fixed. The lower the iron content, the more shall be the gangue con-
tent. Depending upon the metallization achievable, the final composition
of the sponge iron shall be automatically fixed. The sponge iron contains
metallic iron, oxygen in the form of some iron oxide, gangue, and carbon
deliberately introduced in it. The iron content shall vary depending upon
the carbon introduced in the sponge. The theoretical metallic iron content
for a given % metallization is depicted in Figure 20.4.
Typical sponge iron chemistry as obtained in coal-based DRI plants is

shown in Table 21.1.
Table 21.1 Typical sponge iron chemistry as obtained in coal- and gas-
based plants
Ore C S P SiO2 Al2O3 CaO + MgO %Metallization
Coal-based
A 0.10 0.020 0.010 2.0 0.60 0.20 88
B 0.30 0.060 0.040 4.0 3.00 3.0 90
Gas-based
C 1.20 0.010 0.010 2.0 0.60 0.20 92
D 2.40 0.015 0.040 4.0 3.0 3.00 96
Usually the size of sponge iron is decided by the size of the ore charge
and is usually in the range of 6–35, of which 85% must be in the range of
10–35 mm and the −5 mm fraction should not be more than 5% of the total
charge. This is ideal from the point of view of subsequent steelmaking oper-
ation. Any larger size poses problems in charging, and the rate of its melting
in steelmaking is adversely affected. Similarly, any extra-fine material below
5-mm size is likely to result in increased dust losses or reoxidation, causing
problems in steelmaking.
Apparent density in the range of 4–6 g/cc is considered good. In spite
of the carbon contained in sponge iron, any exposure to moisture and/or
oxygen does result in some reoxidation of the metallic iron. The extent of
this shall depend on the extent and severity of such exposure. Because of its
spongy structure, it has a low thermal conductivity, which can create prob-
lems if the charging rate is more than commensurate to the melting rate, as
defined by the power input, in an electric furnace.
21.11 Use of Sponge Iron

An excellent account of preparation, properties and utilisation of pre-
reduced pellets is available.
The possibility of making sponge iron opens up an alternative route to
steelmaking. The conventional route is the blast furnace—open hearth, or
blast furnace—B.O.F. combination. The alternative route is the sponge iron –

electric arc furnace combination. This can work out economically viable, under
certain circumstances, for otherwise noneconomic areas. The Dwight-Lloyd-
McWane (D.L.M.) process, the Strategic – Udy process, the Elektrokemisk
process, and the like, are typical examples of combinations of direct reduction
coupled with electric arc furnace for the manufacture of steel.
Sponge iron can also be fed in an open hearth, L.D., et cetera, as scrap
for its conversion to steel. Similarly, 30%–50% of the charge can be sponge
in an otherwise scrap-based electric arc furnace process for making steel.
[108] A blast furnace can also readily take up to 50% of the ore burden as
sponge to improve upon its productivity.
Unlike scrap, sponge iron does not contain any residual elements such as
Cu, Sn, Ni, etc. and hence its use allows more stringent specifications of steels
to be met by an electric arc furnace. Sponge iron, being a product, unlike the
scrap, which is a by-product, allows better control during its subsequent pro-
cessing to arrive at correct end conditions. Steelmaking trials using sponge iron
have yielded satisfactory quality and recovery of metal values.
The possibility of using sponge in a cupola for making foundry iron has
also been shown to be commercially viable.
All these possibilities are bound to lead to the setting up of a large
number of mini-steel plants, based on this technology to meet the social
aspirations of people at a large number of places rather than setting us a big
integrated steel plant at one place only.

In the 1970s the DR technology as a whole came to be considered as a near-
perfect alternative to the BF route and was expected to revolutionize the
ironmaking. But these hopes, however, did not come true because of the
following reasons:
1. Very low productivity of an individual unit as compared to BF and
hence cannot avail the economic benefits of scale of production (sp.
investment for DRI is approximately Rs.10000/— per t capacity as
compared to Rs.5000/— for BF)
2. The individual BF size continuously increased to 10000 t/dhm
3. Transportation problems of DRI as it is prone to re-oxide
4. Relatively high cost of production

5. Non-availability of liquid iron
6. Electric power needed to melt and make steel out of it
7. Costly to get carbon in DRI product
The situation, however, changed because of heavy rise in demand for metal-
lics for steelmaking after globalization and liberalization, particularly in
India and China. Now there are nearly 300 such coal-based plants operat-
ing in India, and they are doing well to meet local steel demand of commer-
cial steels. The operating 222 DRI units produced 12.64 Mt in 2005–2006
(i.e., 79% capacity utilization). The gas based three units produced 6 Mt out
of this total production.
The Indian scenario is very specifically suited for coal-based DRI pro-
duction, and that is the reason for its growth in India and its occupying the
leadership role in DRI production on global level.
On the whole, the DRI production is expected to triple in the near
future. It will still be a negligible fraction of the total production of metal-
lics the world over.
CHAPTER
22
ALTERNATIVE TO BLAST
FURNACE IV –
SMELTING REDUCTION (SR)
PROCESSES
In This Chapter
● Introduction
● Classification of SR Processes
● Thermo-chemical Model
● Vertical Shaft Furnace Processes
● Electrical Processes
● Converter Processes
● Bath Smelting Processes
● Rotary Hearth Furnace Processes (RHF)
● TUPKARY-SR Process
22.1 Introduction
The SR process principle owes its origin to early attempts at melting iron
ore and coal mixture in an open hearth furnace. It revealed that reduction
of iron oxide in liquid state indeed occurred in that way.
The achievements with respect to developments in the form of MBFs,
sponge iron production, and electro-thermal smelting of iron ores, no doubt
contributed to the production of iron by alternative routes to blast furnace.
Out of these, the applicability of electro-thermal smelting depended on cheap

electric power. Similarly, the MBFs in one or the other forms still rely on some
form of metallurgical coke only. The low shaft furnaces also required some
form of coke and were not fuel efficient. The sponge iron production was
truly the alternative process with respect to use of cheaper and abundantly
available fuel such as ordinary coal. This, however, produced solid iron mixed
with gangue to be melted only in an electric furnace or used as coolant in oxy-
gen steelmaking. It means again the use of scarce and costly electric power,
which to some extent at least, defeats the main aim behind developing these
alternative routes of iron production. The DR processes also suffered from
inherent drawbacks such as low productivity, high cost of production, and
problems in handling sponge iron. The quest to produce liquid iron by some
alternative route, bypassing the use of coke, therefore remained a challenge,
in spite of the development of these alternative routes of iron production.
Any pyro-metallurgical smelting process in general, and more so the
blast furnace for smelting iron, essentially depends on melting the iron ore
along with its reduction to finally produce metallic iron and slag formed out
of the gangue associated with the ore. How the reduction and melting are
to be brought about efficiently depends upon the ingenuity of designing the
furnace so that the fuel and the reductant are utilized to their maximum
extent, thereby economizing the process. The blast furnace is a typical
example of a very efficient countercurrent reduction reactor and a melter.
The only problem is that the efficiency of BF depends on the use of costly
metallurgical coke as a fuel.
The fundamental processes occurring in the blast furnace are all neces-
sary to be carried out to produce iron from its ores by any process. Their
proportion inter se may vary, depending upon the requirements. In sponge
iron production the stack part of the blast furnace is in a way reproduced
and hence the iron is not separated from the gangue associated with the
ore. In an electric furnace the bosh-reactions are so dominantly carried out
that the stack-part processes play a very minor role. In the MBF it is strictly
a miniature copy of the blast furnace and nothing else.
A real alternative to blast furnace is not thus available in any of the
above-mentioned processes. Any true alternative process to BF should
produce molten iron as the product cleanly separated from the associated
gangue in the form of slag, while not using metallurgical coke as fuel in
any form and the fuel and the reductant requirements are to be met by
any of the cheaper solid, liquid, or gaseous fuels. This quest finally led to
the development of several processes known as smelting-reduction or
Alternative to Blast Furnace IV – Smelting Reduction (SR) Processes • 431
SR processes for short. Their difference inter se lies in how and in what
relative proportion all the blast furnace processes are carried out, starting
with solid state reduction of iron oxide to produce molten iron as the final
product, obviously separated from the molten slag.
In this respect the electric smelting can also be classified as a single-
stage SR process, when most of the SR processes are two or, exceptionally,
three-stage operation.
In this period of a few decades, when several SR technologies were
under development, as a side effect of these, it was realized that any modi-
fications even in BF itself to reduce the coke rate are worthy of exploitation.
The plasmasmelt process, though classified as SR technology, is a typical
example of such approaches.
In all these processes some of the following activities are carried out in such
a way as to obtain optimally iron with a minimum of energy input:
1. Iron oxide is pre-reduced in solid state at relatively lower temperatures.
This pre-reduction normally does not proceed beyond 70% as against
more than 90% obtainable in DRIs. This is invariably the first step.
2. In the second stage this pre-reduced iron oxide is melted and reduc-
tion is completed in molten state to obtain liquid iron with required
carburization.
3. The reduction and melting (i.e., smelting) essentially involves chemi-
cal reaction such as:
FeO + C = Fe + CO
and 2C + O2 = 2CO
The burning of coal generates the necessary heat for the process. This
generates CO, gas which is available for subsequent reduction and as
an export fuel.
4. The carbon monoxide gas, generally along with H2 generated by simul-
taneous reaction of coal with moisture, is used for pre-reduction in the
pre-reducing reactor.
5. Electrical energy is also used for melting using slag as the resistor, and
in such a case coal is still injected for final reduction of iron oxide. The
off-gas in this case is again CO, which is used for pre-reduction of iron
oxide as well as in-house electricity generation.
6. Fine iron ore and coal mixture is used as charge along with lumpy iron
oxide or pellets.
7. Rather than using one single continuous reactor, the above operations
are carried out in separate reactors to obtain optimum efficiency and
economy.
8. Reducing gas regeneration is also carried out if required.
9. Excess off-gas consisting of CO and H2 may be used elsewhere as rich fuel.
FIGURE 22.1 Schematic principle of SR processes based on use of coal and oxygen
The schematic principle of SR processes using coal and oxygen is shown

in Figure 22.1. The way the various operations involved in these processes,
mentioned in this figure, are actually carried out differentiates one SR pro-
cess from the other.
22.2 Classification of SR Processes

All ironmaking processes basically differ with respect to the extent of ore
fines used proportion of coal used as reductant/fuel and the amount of oxy-
gen gas used for burning the fuel. At one end is the BF wherein no fines
are tolerated, only 10%–15% of the fuel may be in the form of coal, and
a small amount of oxygen gas is used to burn the fuel at the tuyere. The
other extreme is fluidized bed reduction processes wherein only ore fines
are used, cent-percent coal is used, as fuel as well as reductant and, oxygen
gas is fully used for burning the fuel as well as for post-combustion. The
DRI processes broadly need lumpy ore with solid coal or gaseous hydro-
carbon as fuel and as reductant. No oxygen is required; air is good enough
for burning fuel.
The SR processes can work on ore fines, coal as fuel, and reductant and
oxygen is fully used for burning the fuel as well as for post-combustion. The
processes can be classified scientifically on this basis. These can also be clas-
sified on another basis as well, such as:
1. In terms of stages as single- or two-stage operation.
FIGURE 22.2a and 22b Schematic difference between one and two stage SR processes using electricity
and coal as source of thermal energy.
2. In terms of types of furnace used, such as shaft or converter or sub-

merged arc or hearth or fluidized bed, et cetera.
3. In terms of thermal energy source and reductant, such as carbon and/
or electricity used for heating and melting.
4. In terms of how and to what extent post combustion is utilized.
Figure 22.2 shows the schematic difference between the one- and the
two-stage processes in which thermal energy is supplied by electricity and
reduction by coal.
The classification may be based on any one of the above criteria. Any
process may fall in one or the other category, depending upon the criteria
used for classification. However, many may not belong to any specific class
because of their characteristic features.
The attempts to enhance extra scrap melting in a basic oxygen furnace
during steelmaking led to incorporation of coal in the charge along with
the scrap. The usual top-blown oxygen burnt this coal and produced the
extra heat needed for melting the scrap. This concept was further extended
to evolve SR technology. Such SR processes, wherein coal in the molten
oxide charge is burned by oxygen jet, fall in the BOF-based categories.
Processes such as Corex, Finex, HI smelt, AISI direct steelmaking, DIOS,
Combismelt, COIN, Hoogovens, BSC continuous melter, IDI, CIG, and
Kobe-Process fall in this category.
In another group of processes, coke bed is used in a vertical shaft fur-
nace to reduce iron oxide in molten state. Coal and oxygen are injected
to aid melting of the pre-reduced ore made in another reactor. They par-
tially replace the use of coke, that is, its function as in a blast furnace. The
processes such as Sumitomo SC, Kawasaki SR, Plasmasmelt, Technored,
IT-Mark 3, among others, belong to this category. The Sumitomo SC
requires 30%–40% of the coke required in BF under similar conditions.
The Corex may be included in this category, except that the pre-reduction
is carried out to a very high degree (i.e., almost 90%), unlike in the other
processes in this group.
Some use a rectangular open hearth type reactor for smelting the ore
with coal. The Romelt (earlier FLPR) and the Ausmelt processes fall in this
category.
Those that use electrical energy for heating and melting, and coal as
essentially a reductant, and generating required electrical energy using the
off-gases of the system itself, fall in the category typified by the INRED and
the ELRED.
The process such as Plasmasmelt relies on use of coal and oxygen in the
BF itself to reduce coke consumption to its bare minimum. Coal may also
be used as a reductant in this process.
The rate of production and the energy efficiency of these processes is a
critical parameter in their final commercial viability. The reduction of iron
oxide can be carried out at different places in these processes. It may be
in the slag, in the melt, at the lower or higher levels in the emulsion or in
lower or the upper reactor. All these differences are characteristics of the
individual processes.
Similarly, post-combustion of CO generated from iron oxide reduc-
tion by coal, at least to some extent, in a suitable manner and at a suitable
place, is a must to improve the energy efficiency. How, where, and to
what extent this is brought about decides the overall thermal efficiency
of the process.
22.3 Thermo-chemical Model

This model takes into account the heating and the reduction to be carried
out in two distinctly different stages to achieve the highest possible thermal
efficiency and the fastest possible reduction and melting. It assumes:
1. Two different stages of reduction (i.e., pre-reduction and smelting-
reduction) to be carried out in two different stages.
2. The pre-reduction stage consists of the upper, middle, lower, and car-
burizing zones.
3. Iron ore is partially reduced in pre-reduction using the off-gases of the
smelting zone.
4. The smelting zone partially avails heat of post-combustion zone either
below or above the tuyere level.
5. Gases generated from the reduction and combustion reactions in
the smelting zone are partially burned to get heat for the smelting
reactions and to reduce the CO and H2 potential of the off-gases

before these enter the pre-reduction stage.
6. The hot and partially reduced charge from the pre-reduction stage en-
ters the smelting zone above the tuyere area, then passes below it, and
finally joins the melt-pool to produce liquid iron and slag as products.
7. The pre-reduction stage is similar to the standard BF stack opera-
tion, with the difference that the degree of reduction may vary, de-
pending upon the overall process design and the way the operation
is carried out.
8. Similarly, the second-stage smelting below the pre-reduction stage is
conceptually very much akin to the bosh and hearth of the BF with a
very compacted area of operation.
9. The process is optimized by using preferably raw ore and coal as
charge without and pre-processing and making full use of the scale of
operation.
10. The volatile matter content of the coal has to be used for the process
very judiciously.
11. In the smelting process, the reductant must be used efficiently such
that minimum carbon is required. The theoretical carbon requirement
is shown below:
(a) For direct reduction of FeO by C at 900° C is 322 kg of C/t Fe.
(b) For indirect reduction of FeO by CO 704 kg of C or 1314 Nm3 of
CO will be required/t Fe.
(c) By a combination of these two processes, reduction can be carried
out using only 172 kg of C/t Fe, wherein the CO gas utilization
is 88%.
(d) If energy is supplied from an external source, then carbon
is used as reductant only, and the balanced process is shown in
Figure 22.3 (a) and (b).
12. It is therefore necessary to determine the optimum proportion of pre-

reduction and the proportion of reduction smelting in the second stage
to achieve the best possible fuel efficiency, making use of carbon both
as the source of heat and as reductant, taking into account the post-
combustion.
FIGURE 22.3 Simplified flowsheet of a balanced two-stage SR process using external electrical energy
is shown in (a), and the (b) shows the process using coal as a source of heat as well.
The two-stage design of SR is more popular since the efficiency is high.

Some of the common features are:
1. Coal and iron ore are charged into the reactor.
2. The coal is first devolatilized and produces char.
3. The iron ore dissolves in the slag and is reduced by carbon present in
the char and by carbon dissolved in iron.
4. Oxygen/air is injected through the lances/tuyeres.
5. The process energy requirements are met by combustion of carbon to CO
and post-combustion of CO and H2 to CO2 and moisture, respectively.
6. The ore is pre-reduced in the pre-reducer by the gases generated in
the smelting zone.
7. The off-gases, even after pre-reduction, are still rich enough in fuel
value and hence are used to generate electric power, which is often
more than enough to run the plant.
Some of the important SR processes have been described below.
22.4 Vertical Shaft Furnace Processes

This includes the Corex, Finex, Technored, Sumitomo SC, Kawasaki SR,
and such other processes. These are described below.
Corex Process
This essentially eliminates the coke making step and uses coal directly as
the fuel as well as the reducing agent. Corex evolved from the original KR
process of Voest-Alpine AG and KORF Engineering GmbH. The first com-
mercial plant was installed in 1988 at ISCOR’s Pretoria Works in South
Africa. This process produces molten iron in a two-step reduction melting
operation. One reactor is melter-gasifier and the other is pre-reducer. In
the pre-reducer, iron oxide is reduced in counter-flow principle. The hot
sponge is discharged by screw conveyors into the melting reactor. Coal is
introduced in the melting-gasifying zone along with oxygen gas at the rate
of 500° C–600 Nm3/thm. The flow velocity is chosen such that temperature
in the range of 1500–1800° C is maintained. The reducing gas containing
nearly 85% CO is hot de-dusted and cooled to 800–900° C before leading
it into the pre-reducer. The process is designed to operate at up to 5 bars.
Fluxes such as limestone, dolomite, silica sand, etc. are added along
with the ore to finally make up the right quality of slag. Approximately 500–
700 Kg of fixed carbon is required to make a metric ton of hot metal, along
with 300 Nm3/thm oxygen gas.
The export gas generated in the melting zone is rich and may have
calorific value of approximately 7500 kJ/Nm3 and can be used in the plant
usefully.
A variety of coals are usable in this process. Excessive volatile matter
is not desirable. Indian coals do suffer from this drawback and have to be
blended with appropriate imported coals. The Corex process accepts a wide
range of raw materials but better quality of ore, flux, and coal can give bet-
ter performance. The general considerations as applicable to blast furnace
practice should be kept in mind here as well. Typical hot metal analysis is
C = 4%–4.5%, S = 0.02%–0.09% and Si = 0.3%–0.8% or higher if required.
FIGURE 22.4 Basic material flow-sheet and equipment of the Corex process
The characteristics of this process are:

1. Unlike BF, high alkali content of the charge does not affect the opera-
tion adversely.
2. The specific melting rate is nearly twice as that of BF (i.e., 3.0–3.5 t/m3/d).
3. Although it is designed to take up coal as fuel, but 10%–15% coke,
amounting to 120–150 kg coke/thm, is required along with balance
nearly 800 kg coal/thm.
4. Although it can take any lumpy charge, but a minimum of 50%–60%
pellets are required.
5. A 2000–3000 thm/d plant is economically and thermally a minimum
viable size of the plant.
Looking at the performance of JSW Steel Plant, Essar purchased a Corex

unit from Korea. Although JSW has established international benchmarks
in Corex, the economics of this process still remain precarious. It can be
seen from the fact that while augmenting the hot metal production, JSW
decided to install a normal BF rather than adding an additional unit of
Corex to add to the existing Corex line. It only means that all is not well
with Corex, after all.
Sumitomo SC Process
The S of this process means Sumitomo, shaft, and smelting and C means
coal, coke, and cupola. The melter in this case resembles a cupola in which
a coke bed is maintained. Pulverized coal and oxygen are fed into this on
hot coke to produce heat to melt the charge and to generate hot reducing
gas. The pre-reducer is a separate reactor in which ore lumps are charged
and melter gas is used for its reduction. This arrangement results in 48%
coke and 52% coal consumption at the combustion rate of 1.2 kg of coal/
Nm3 of oxygen. Fuel rates of 840 kg/thm and a productivity of 3.34 t/m3/d
have been attained. The main advantage of this process is that it reduces
the coke consumption in iron production considerably. The scheme of the
process has been shown in Figure 22.5.
FIGURE 22.5 The scheme of Sumitomo SC smelting reduction process.
As it is a shaft furnace, it requires agglomerated charge and not ore

fines or alternatively sized lumps.
FIGURE 22.6 Scheme of Kawasaki smelting reduction process, which is carried out by either of the two
alternative (a) and (b) routes.
Kawasaki Process (SR)

This process essentially uses ore fines and low-grade coal or coke for its
reduction. In a shaft furnace such as a cupola, a coke bed is maintained. An
ore and coal/coke mixture is fed through tuyeres on this coke bed. It melts
and trickles through the coke bed where it gets reduced to iron saturated
with carbon. The process of melting is controlled using two layers of tuyeres
rather than only one. It can be operated with a pre-reduction reactor or

without it. The pre-reduction, if used, is carried out using the melter gas.
Pre-reduction is carried out in a fluidized bed, which has a faster reduction
rate and can use unprepared ore fines. This has been found more suitable
for ferro-alloy production. Under certain well-defined conditions, this pro-
cess can be economically operated.
The experimental set-up has shown productivity of 2–3 t/m3/d. The
scheme is shown in Figure 22.6.
Finex Process
It is developed by the POSCO, South Korea to use large amount of fines
accumulated and generated in routine ironmaking by BF. It is essentially
a combination of a Corex smelter unit coupled with fluidized bed pre-
reduction carried out in four different shaft-like reactors (such as the HyL
process) wherein the pre-heating, initial reduction and final pre-reduction
up to 90%, is achieved. The pre-reduced fines are hot compacted to pro-
duce HBI to suit the shaft of Corex for maintaining charge permeability.
By adopting the compacting step, the limitations of Corex in not accept-
ing the ore fines and coal fines is indirectly circumvented. This has also
decoupled the necessity of matching the rates of production of pre-reduced
material and rates of melting-reduction, as in the Corex operation. No coke
is required to run this process, and 900–1000 kg coal/thm is required. The
hot metal composition is C 4.3%–4.55%, Si 0.02% and S 0.02%. It is sche-
matically shown in Figure 22.7.
FIGURE 22.7 Schematic representation of the Finex process.
POSCO is going to adopt this technology to smelt iron ore at their steel
plant soon to be installed in Orissa.
Technored Process
This process was designed and developed specifically to use iron-bearing
fine-waste material generated in a number of ways within the plant. These
fines could not be used in BF in any way, and alternative route had to be
devised for their smelting. It used cold-bonded, self-fluxing, carbon and
coke containing and therefore self-reducing pellets as the charge material.
Hydrated quicklime was used as the bonding and fluxing material. A vertical
short shaft-like furnace was designed wherein coal/coke was also fed as fuel,
like the way it is injected through BF tuyeres. The fast kinetics are due to
the intimate contact of iron oxide with carbon. Part of the heat generated at
bottom by burning fuel melted the charge and also provided the hot reduc-
ing gases to pre-heat and pre-reduce the charge above. Hot blast was used
as in a normal BF. Cold blast was blown in upper part for post-combustion.
All this resulted in consumption of 250 kg of coal and 310 kg of coke
fines to produce a metric ton of hot metal containing 3%–4% C and 0.2%–
1.0% Si at a temperature of approximately 1400° C.
It is a promising process for using a variety of waste fines of iron and
fuel. This was developed in Brazil in 1998 and is proposed to be used at the
mine site to use ore fines and purchased fuel fines. It is also claimed that it
is also suited for producing ferro-alloys such as Fe-Mn.
22.5 Electrical Processes

INRED Process
FIGURE 22.8 Scheme of INRED flash-smelting-reduction process in one single stage.

INRED means an intensive reduction process. It was developed and

installed at MEFOS, Sweden in late 1970s. The chief advantage of this pro-
cess is that it can use good-grade ore fines or concentrate as such without
agglomeration.
It uses cheaper non-coking coals and has competitive cost with low
investment. It is environmentally acceptable. The process is shown sche-
matically in Figure 22.4.
The fine ore, coal, and oxygen are tangentially introduced into the reac-
tor through the roof. They create a vortex that is amplified by secondary and
tertiary jets of oxygen at additionally two levels. Flash smelting takes place
under these conditions maintaining temperature approximately 1900° C.
The lower portion of this reactor is an electric furnace in which the nearly
fully reduced material is melted to produce molten iron. The gangue forms
slag for which suitable flux has to be incorporated in the charge mix. The
product of flash smelting settles down into the reactor as sponge iron, coke,
burnt lime, and unreduced material. It cools down rapidly to 1450° C and
becomes pasty and porous. Final reduction and melting is effected at a
lower level around the electrodes. A small bed of coke is maintained under-
neath the slag layer.
It has been claimed that this plant can operate to obtain its own require-
ment of electricity from the hot gases available from the system.
The operating details are:
Iron source 65% Fe minimum
Coal moisture 7% maximum
Coal ash 15% desired
Net C.V. dry basis 28 MJ/kg
Coal consumption, dry 620 kg/thm
Oxygen, 95% pure, 5 bar gauge 680 m3/thm
Electrical energy for furnace 300 kwh/thm
Electrical energy for auxiliaries 100 kwh/thm
Electrical energy for oxygen plant 280 kwh/thm
Electrical energy produced 680 kwh/thm

Special features:
1. Ore/conc up to 100 mesh has been successfully used.
2. Coals with high sulfur can be used.
3. Coking grade coal not required.
4. No dust clogging problem experienced.
5. Once started, the process can be stopped and put on without much
time loss.
6. Carbon content is controlled by controlling the coke bed.
7. Normal silicon is 1% but controllable to some extent.
8. Nearly 70%–80% sulfur is removed and remaining gets distributed be-
tween slag and metal with distribution ratio of 25 when slag contains
less than 1% Fe.
9. Volatile impurities, including phosphorus, are eliminated in dust, and
effective de-dusting is possible in the last stage.
10. Low volume of gas generation and hence smaller gas-cleaning facilities
required.
11. Simple reactor design with less of moving parts.
In spite of all these tall claims, the only demonstration plant set up in the
late 1970s has been abandoned. The claims are therefore open to reason-
able doubt.
This was the pioneer of SR development and it set the pace for use of
fine iron ore, ordinary coal and oxygen, and very fast reduction kinetics for
molten iron production and to design processes that could start and stop at
very short notice. Similarly it established that such processes would nec-
essarily have fuel value–rich export gas and it could be mandatory to use
it for in-house power generation or any other important use. That is how
the power generation became an integral part of the SR process designs
thereafter.
ELRED Process
The ELRED process was developed in Sweden by ASEA and Stora
Kopparbergs and is designed to operate in two stages, first, pre-reduction
up to 60%–70% metallization using carbon as fuel and second, melting this
pre-reduced material in an DC arc furnace. Ore fines/concentrates can be

directly used but should have minimum 65% Fe. Plant wastes can be thus
recycled. The pre-reduction is carried out under pressure in fluidized bed
with faster reduction and non-sticking of particles because of extra carbon
in the charge. Coal powder and air are injected directly in the bed to gener-
ate temperature of approximately 1000° C, provide CO + H2 as a reducing
gas mixture and excess carbon required in the process. Hot cyclone is pro-
vided for de-dusting the reducing gases.
The pre-reduced material is continuously fed to the melting zone at
the bottom by adjusting the residence time and temperature in the reac-
tor. A hollow electrode is used to carry the hot pre-reduced material into
the melting zone along with suitable fluxes to produce molten iron and slag
of required compositions. The slag volume is much less. The hot reducing
gases can generate more than required electrical energy for the process.
FIGURE 22.9 Scheme of two-stage ELRED process of smelting reduction.
In spite of these tall claims, the development of this process has been
suspended.
The process scheme is shown in Figure 22.9.
COMBISMELT Process
This process was developed jointly by Lurgi and Mannesmann Demag of
Germany by combining coal-based sponge iron producing SL/RN process
with submerged arc furnace for open slag smelting. The electrical energy
requirement of this process is met to the extent of 70% from the hot gases
of the kiln. The continuous nature of the process makes this generation
little easy. The heat recovered through electricity generation very much
depends upon the type of coal used in the kiln.
The hot reduced ore is fed directly to the arc furnace. Slag depth is
maintained rather high so that electro-thermal energy generation is rel-
atively easier with a high-power factor. The resistance heating system is
more flexible with respect to variation in power availability. It also helps
to accept a varying degree of metallization of the reduction part. The pre-
reduced material is charged in batches and slag metal reactions are very fast
to reach near-equilibrium conditions.
This process can be extended to produce steel as the final product. The
carbon content of the pre-reduced charge is controllable and helps in the
steelmaking process.
The process is yet to develop to the stage of its commercial application.
Plasmasmelt Process
This was developed by SKF Steel of Sweden circa 1975. Herein the ore
fines are first pre-reduced in fluidized bed reactor using gases from the
shaft smelter. The shaft furnace is filled with coke and, pre-reduced ore,
flux, carbon, etc. are introduced in this. Clean off-gas is introduced through
tuyeres to generate plasma wherein the temperature is of the order of 3000°
C–5000° C. It rapidly drops to 1700° C–2000° C just outside the plasma.
The reduction of oxide is completed fairly rapidly, and the molten iron and
slag trickle down through the bed into the hearth and are tapped much like
that in blast furnace.
The predominantly CO-containing gases from the smelting zone are,
after cleaning and cooling to 850° C or so, fed into the pre-reduction stage.
The hot metal produced has low sulfur and normal silicon and manganese
contents, which can be adjusted to the desired values. It has been found to
be capable of smelting ferro-alloys.
This process uses almost half the fossil fuel required in a blast furnace.
On the whole, the cost is claimed to be 25% lower than that in a blast
furnace. This has a distinct disadvantage if electricity is costly. In such cases

coal plus oxygen can be used as a source of heat for melting.
With all these favorable factors, the process remains confined to bench
scale trials.
Fastmelt Process
The Fastmelt process uses a rotary hearth furnace to reduce the ore fines
to an approximately 90% reduction. This produces product such as the
DRI, which needs melting. The charge is therefore put in an electric
furnace to melt, to complete the reduction and produce molten iron.
The problems of handling iron oxide rich slag are yet to be solved to the
necessary satisfaction. It is true that the reduction kinetics in this process
are very fast because of intimate mixture of iron oxide and carbon in the
charge.
22.6 Converter Processes

Krupp–COIN Process
Krupp of Germany developed this process. The name COIN is made out
from coal (CO) and injection (IN). It uses shaft furnace for pre-reduction
and usual converter for smelting. The product is intermediate between
steel and molten iron, that is, carbon approximately 1%, sulfur 0.1%–0.2%,
and low silicon and manganese.
PCIG Process
The full form of the process is Pressurized Coal Iron Gassification. This
process is based on the principle that in molten iron, if coal and oxygen
are blown, carbon monoxide will be generated as the product gas along
with hydrogen to some extent. This gas is to be used for pre-reduction, and
the pre-reduced material is melted in the gassifier during the gassification
process itself. The process is very flexible with respect to the production of
hot metal and gas.
HI-Smelt Process
This is the process that aims at using ore fines or steel plant wastes. The
off-gas is used for pre-heating and pre-reduction of the charge in a vertical
shaft. Hot blast is also generated using off-gases to maximize productivity.
Pre-heated blast up to 1200° C, using oxygen enrichment, if needed, is

injected through water-cooled lances in the smelting vessel for post com-
bustion of carbon monoxide and hydrogen of natural gas injected as fuel
and generated from smelting reactions.
It’s a fast-responding system to changes in charge feed rate. The
world’s first plant has been operating since 2005 at Kwinana, Western
Australia. It promises bright prospects. The schematic system is shown in
Figure 22.10.
FIGURE 22.10 Basic flowsheet and equipments of HI-smelt process.[88]
Reactor Steelmaking Process

This RSM process can be used for making iron or steel of any desired car-
bon content. Carbon is burned using liquid iron as the medium to obtain
high combustion rates. The hot reducing gases from the smelting zone pass
through the stack of the furnace wherein pre-reduction is carried out. The
pre-reduced ore drops down into the melter-gassifier in hot condition. It is

melted by a coal-oxygen burner. A large amount of CO is combusted over
the molten slag as post-combustion to help improve the energy efficiency
of melting zone.
AISI-HyL Direct Steelmaking

Ore is pre-reduced and is melted in hot condition using coal and oxygen
jets directed to iron bath. Continuation of oxygen blowing alone (without
coal) can lead to steelmaking conditions and can produce steel as the end
product. Steel is thus produced at lower productivity but at much less cost.
It is still on an experimental level.
22.7 Bath Smelting Processes

Bath smelting processes include processes such as the HIS-melt, the
Ausmelt, the DIOS(NKK), the Romelt, and the like.
Romelt (Earlier FLPR) Process

This is a typical single-stage SR process wherein iron oxide is reduced
in a molten state by carbon already mixed with the iron ore charge.
Commercial purity oxygen is blown to partially burn coal to produce heat
and the rest of the carbon is used for oxide reduction. This was devel-
oped by the Moscow Institute of Steel and Alloys, Russia. A part of the
CO generated by reduction as well as by burning is post-combusted to
obtain heat for the process. It is essentially a hearth furnace process. The
furnace is rectangular in shape and oxygen tuyeres are fitted on the long
walls. The oxygen or enriched air blown through these tuyeres generates
emulsified slag above the tuyere level, while the bottom part remains
quite and helps to clarify as slag and metal layers. It can take all types
of iron oxide generated as waste in steel plants. The scheme is shown in
Figure 22.11.
The off-gas temperature is nearly 1600° C and is rich in CO with calorific
value of 7100 kJ/Nm3. The volume is also large (i.e., 2800 Nm3/thm). This rich
fuel needs to be used efficiently to improve economy of the process.
No commercial plant has yet been installed based on this technology.
This is also classified as the hearth process.
FIGURE 22.11 Scheme of Romelt process.
DIOS (Direct Iron Ore Smelting Process)

DIOS essentially includes fluidized bed pre-reduction followed by smelt-
ing. It also has a gas reforming unit. The pre-reduction and post-combus-
tion are adjusted to suit the thermal efficiency. This triple combination,
with nearly 40% –60% post-combustion, gives adequate energy efficiency.
It needs faster reduction smelting and for which nitrogen is blown through
the slag for stirring and oxygen is blown through side tuyeres.
The results of a pilot plant study have been quite encouraging.
22.8 Rotary Hearth Furnace Processes (RHF)

Typically these processes use rotary hearths for reduction.
IT mk3 Process
The name is coined out of the major historical generations through which
the ironmaking has progressed. The first is BF, which is named as IT mk1.
The next was the gas based reduction to produce DRI which is named as
IT mk2 and this is supposed to be third generation and hence the name as
IT mk3. This was developed by Kobe Steel of Japan.
FIGURE 22.12 Schematic of process flow sheet of ITmk3 process.
It also employs rotary hearth wherein the reduction of iron oxide is

carried out as in the Fastmelt process to produce iron-nuggets instead of
DRI as the product. This is possible because the temperature employed is
much higher, that is, of the order of 1350° C–1500° C. This is much higher
than that employed in any rotary hearth process. The charge is similar to
the Fastmelt process. It consists of fine iron-bearing material, reductant,
and binder and is pelletized before charging. The scheme of the process is
shown Figure 22.12.
The reduction is not only fast but because of high temperatures the
product is in the form of semi-solid nuggets. This also leads to get slag as
practically a separate phase as distinct from iron phase though sticking to
individual particles. It is thus easy to separate the iron from the slag read-
ily. The process produces nuggets of iron containing 2.5%–4.5% C, 0.2%
Si, 0.1% Mn, and less than 0.05% each of S and P. These nuggets can be
directly fed into the electric arc furnace as solid iron charge or in basic oxy-
gen processes as coolants.
This process appears promising because of the following characteristics:
— Ingenious method of removing gangue from low grade iron bearing
materials
— Uses composite pellets as charge
— One step process to produce straightaway the iron-nuggets
— Time required is just ten minutes or so for reduction, melting and slag
separation
— Any iron oxide is easily processed
— Claims to produces at least 20% less CO2 than emitted from BF
— In an integrated plant hot nuggets can be directly fed to EAF
— This is a distinct two-step process
Two commercial plants, one in Indiana and another in Minnesota in the
USA, each of 0.5 Mt annual capacities, are being built. These will prove its
applicability on a commercial level, beyond doubt.
22.9 TUPKARY-SR Process

Steel is produced from a mix of molten iron, pigs, steel scrap, and sponge
iron using basically either a Basic Oxygen Furnace process or an electric
furnace process. No matter how steel is produced by any of these pro-
cesses, it has to be produced using a refining slag, which contains a high
proportion of CaO, other oxides, and a considerable proportion of iron
oxide. The amount of slag produced is generally approximately 15% of
the total weight of the metallic charge in the steelmaking furnace. This
slag contains iron in both forms, as iron oxide and mechanically entrapped
iron. The amount of mechanically entrapped iron varies considerably and
depends on the dynamic nature of the slag metal system inside the furnace
during steel refining. In electric furnace steelmaking this proportion may
be quite high.
So far, this slag is dumped in some corner of the plant site itself since no
application worth the name has yet been found anywhere in the world. This
disposal system involves collection, transportation, and dumping of the slag.
It requires a great deal of costly land of the plant. It also causes pollution
problems, and after its accumulation over more than a century, now have
acquired serious proportions, besides the cost involved.
BF slag was also dumped like this once but now it is used as a substitute
raw material for Portland cementmaking. The values in BF slag are in the
form and decreasing order as CaO, Al2O3, and SiO2. Steelmaking slag also
contains the same constituents and in more valuable proportion than in BF
slag. The defect is that it contains too much of iron, in metallic as well as
oxide form. That reduces its value to practically zero.
Presently, the mechanically entrapped iron in this slag is won by crush-
ing and magnetic separation after solidification of this slag, as a standard
practice. But such a process does not collect all the iron contained in the
slag. The iron so obtained is contaminated by magnetic oxide, and therefore
by other oxide as well, produced during cooling of the slag.
The situation was almost beyond redemption until the advent of good
and efficient SR processes. In most of the SR processes as described
above, even lean iron ores or iron waste material from the plant, in fine
form, are smelted with ordinary coal as fuel and reductant. Steelmaking
slag may contain total iron as low as 15%–20% and as high as 30%–50%,
depending upon the steelmaking process. Besides, it will have to be fed
in the SR furnace in molten form. Its temperature is approximately 1550°
C–1600° C.
Any standard SR process takes solid iron-bearing charges at room tem-
perature. Hence, thermodynamically enthalpy is required to heat and melt
before the reduction of iron oxide can occur. This enthalpy requirement will
be almost one-third that required if molten charge was used. Hence, if mol-
ten slag from steelmaking is charged in the SR process, the total enthalpy
requirement for producing iron from this would be almost 30%–40% of that
required in the standard process. When iron is so reduced, the mechani-
cally entrapped iron will also be collected without much additional enthalpy
investment. The slag left over will be better than the BF-type slag because
it may contain a higher proportion of CaO besides other above-mentioned
main oxides. It may be more valuable than usual BF slag because of its
higher CaO content.
One of the possible problems would be the use of such high-temper-

ature slag. It could be fluxed with additional iron ore and then reduced to
finally get the correct smelting temperature.
The main drawback of this smelting operation will be that the metallic
iron so produced in this process would also contain practically all the phos-
phorus that was removed during steelmaking. It may find use as pigs rather
than for recirculation in the steelmaking circuit of the same plant.
All that would be needed would be to develop a means of collecting
molten steelmaking slag and transporting it to the SR unit located at a suit-
able place in the plant. It may not be an easy task, but it will not be an insur-
mountable task, either. The logistics will have to be developed for collection
and transportation of the molten steelmaking slag. The SR process will also
have to be modified to accept molten charge at such high temperatures.
If it works out, suitable legislation may force steel plants to adopt it for
decreasing pollution, in the first place. At least to some extent this will ease
the shortage of iron ore sources for producing iron. It will also augment
cement production and improve its economy of production.
The advantages claimed are:
1. Two more useful products will be produced in the form of molten iron
and slag.
2. Valuable plant land will be free for any other better use.
3. The crushing, grinding, and magnetic separation, etc. of solidified slag
will no longer be necessary, amounting to some savings.
4. The already-accumulated dumps of such slags can slowly be consumed
in the same process at some additional cost.
5. The limitation of allowing dynamism in the electric furnace steelmak-
ing would no longer be a restriction.
6. The yield in the steelmaking process would indirectly increase.

A large number of SR processes have been described in the preced-
ing sections. Many of them were tried on a pilot plant level and they had
to be abandoned at that level. These could not be adopted in practice.
Notwithstanding this, these processes have been described here to under-

stand their varieties and the reasons why they could not come to the stage
of commercial viability. Any future attempt would thus be benefited from
these abortive experiments as well.
One of the striking features of all these processes is the total cost of
auxiliaries such as oxygen plant, power plant, et cetera, needed to install
them commercially, and their costs are much more than the cost of the
main reduction and smelting equipment. This increases the installation cost
of the whole plant.
Out of all the SR processes, the Corex, Finex, ITmark, and HIsmelt
have been put to commercial practice. The prospects of Ausmelt, Romelt,
et cetera, appear little bright.
In India, Corex was used for the first time by the Jindal-SW plant at
Vijayanagaram in Karnataka. It is claimed that the cost of iron produced
there is little more than half of the cost if produced by a blast furnace
route. This is mainly due to the fact that nearly 17 GJ/thm is produced by
way of off-gases in this process against only 7 GJ/thm worth produced by
conventional blast furnace route. It reflects the fuel efficiencies obtained
in these two types of processes. It also indicates the value of export-gas
in the overall economics of Corex process. Out of the total investment of
the Corex plant, nearly 38% cost is of electrical power plant and 2% is for
related oxygen plant. The export gas energy is useful only after such heavy
investments.
The only problem in such a setup is that the power plant is also shut
down when the furnace is down. This creates difficulties, particularly when
power sale agreements have been made with outsiders to sell this power,
since power will not be available for sale during the furnace shut-down
period, which can be a real disturbance to outsiders.
The Ausmelt process has been well established for the non-ferrous
industry but it is yet to be commercially installed for iron production. It is
claimed that a minimum size of half a million metric tons per annum plant
has to be installed to make it commercially viable.
The Fastsmelt, the ITmk3 and the Finex are fast coming up to be
exploited on commercial level. Prospective users with one or more BFs are
watching these developments with interests because many would like to
augment their iron production hereafter by installing one such unit rather
than by adding additional BF.
In spite of several such attempts, no SR process has been able to operate

as fuel efficient as BF. This is simply because of employment of a very tall
stack in BF wherein the reducing and hot gases are fully utilized to reduce
and heat the charge. The off-gas is therefore very lean and at very low tem-
perature. Without exception, all SR processes use low-height shaft, and that
is why the fuel is not fully utilized. The stack cannot be increased using
coal as reductant and as fuel. Tall stack needs coke as a spacer. This is the
dilemma in which these processes are caught while trying to find an alterna-
tive route to BF. These processes are attractive from the point of view of uti-
lization of iron oxide waste-material collected within the plant, without their
preparation of any kind. Similarly, the coke-breeze, generated as unavoid-
able waste in any steel plant, may also be useful in such processes.
FIGURE 22.13 Flowsheet and material balance of combined reduction process with typical Indian
non-coking coals used as fuel as well as reductant.
The possibility of carrying out an SR process using typical Indian

raw materials, that is, non-coking coals as reductant and fuel is shown
in Figure 22.13 in the form of a material balance sheet[89], producing a
metric ton of hot metal at the end with usual composition.
The SR processes, however, have been claimed to be more useful for
producing ferro-alloys, and they may find commercial applications in this
field in the near future.
Out of the alternative routes, the MBF have been extensively adopted
in India and in that the Tata-Korf design has proved to be popular. There
are nearly 25 such merchant iron-producing units in India.
One of the pressing problem of steel industry is the accumulation of
steel furnace slag and its attendant pollution, along with cost of its storage.
The need is to have a zero-waste steel plant. Suitable SR technology on the
lines of the Tupkary SR Process has to come up to tackle this menace once
and for all.
APPENDIX A
Iron Ore In The U.S.
(March 2017 Report-USGS)1
U.S. mine production and shipments of iron ore in March 2017 were 3.76
million metric tons (Mt) and 1.34 Mt, respectively (table 1). Average daily
production of iron ore was 121,000 metric tons (t), slightly less than that
of February and 20% greater than that of March 2016. Average daily ship-
ments of iron ore were 43,200 t, nearly seven times those in February and
slightly greater than those in March 2016. Mine stocks at the end of March
2017 were 24% greater than those held at the end of February and 9% less
than those held at the end of March 2016.
Monthly iron ore data are reported as gross weight unless otherwise
noted and do not include iron metallics such as pigiron, direct-reduced
iron, hot-briquetted iron, or iron nuggets U.S. exports of iron ore totaled
388,000 t in March 2017, more than double those in February 2017 and
March 2016. U.S. imports of iron ore totaled 11,000 t in March 2017, less
than 10% of those in February and 3% of those in March 2016.
Seasonal variations of imports and exports during the winter months
are attributed to freezing of waterways and the closing of the Soo Locks,
which are used to transport iron ore in the Great Lakes region through the
St. Lawrence Seaway.
1 SOURCE: USGS http://minerals.usgs.gov/minerals/

TABLE 1
U.S. PRODUCTION, SHIPMENTS, AND STOCKS OF IRON ORE1, 2
(Exclusive of Ore containing 5% or more of manganese)
(Thousand metric tons, usable ore)3
Production Shipments3 Stocks
Period
Monthly Year to date Monthly Year to date End of Month
2016:
March 3,120 9,190 1,300 3,060 13,600
April 2,700 11,900 4,260 7,320 12,100
May 3,610 15,500 4,640 12,000 11,000
June 3,740 19,200 4,540 16,500 10,200
July 3,670 22,900 4,430 20,900 9,470
August 3,100 26,000 4,270 25,200 8,300
September 3,290 29,300 3,950 29,100 7,630
October 3,580 32,900 4,050 33,200 7,170
November 3,520 36,400 4,200 37,400 6,500
December 3,510 39,900 4,140 41,500 6,050
2017:
January 3,240 3,240 1,740 1,740 6,730
February 3,420 6,660 179 1,920 9,970
March 3,760 10,400 1,340 3,260 12,400
1 Data are rounded to no more than three significant digits.

2 Excludes byproduct ores and iron metallics.
3 Includes rail and vessel.
Appendix A • 461
TABLE 2
U.S. PRODUCTION OF PIG IRON AND RAW STEEL,
BY TYPE OF FURNACE1
(Thousand metric tons)
Pig iron production Raw steel production
Period Basic oxygen furnace Blast furnace Electric furnace
Monthly Year to date Monthly Year to date Monthly Year to date
2016:
March 2,380 6,840 2,040 5,960 4,390 12,800
April 2,150 8,980 1,860 7,820 4,460 17,300
May 1,910 10,900 2,210 10,000 4,770 22,000
June 2,280 13,200 1,950 12,000 4,550 26,600
July 2,220 15,400 1,890 13,900 4,480 31,100
August 1,860 17,200 2,160 16,000 4,490 35,600
September 1,770 19,000 2,070 18,100 4,120 39,700
October 1,610 20,600 1,940 20,000 4,290 44,000
November 1,660 22,300 1,920 22,000 4,260 48,200
December 1,810 24,100 2,090 24,100 4,370 52,600
2017:
January 1,940 1,940 2,240 2,240 4,740 4,740
February 1,930 3,870 2,140 4,380 4,280 9,020
March 1,960 5,830 2,250 6,630 4,640 13,700
Source: American Iron and Steel Institute.
1 Data are rounded to no more than three significant digits.

TABLE 3
U.S. EXPORTS OF IRON ORE, BY COUNTRY AND TYPE1, 2
(Exclusive of ore containing 20% or more manganese)34
2016 2017
January–March March January–March

Country and
type of product Quantity Quantity Value3 Quantity Value3 Value3
(thousand metric (thousand (thousand (thousand (thousand (dollars per
tons) metric tons) dollars) metric tons) dollars) ton)
Canada 1,030 214 14,000 1,020 67,700 66.59
China – 30 2,450 30 2,450 81.57
Japan – 144 8,930 144 8,930 61.99
Other 1 (4) 65 62 5,530 88.61
Total 1,030 388 25,500 1,250 84,600 67.51
Concentrates (4) (4) 10 4 539 134.75
Fine ores (4) (4) 22 (4) 41 10.25
Pellets 1,030 357 22,900 1,180 79,000 66.80
Other (4) 30 2,480 68 5,090 74.78
Total 1,030 388 25,500 1,250 84,600 67.53
– Zero.
1 Data are rounded to no more than three significant digits; may not add to totals shown.
2 All countries with less than 5,000 metric tons of exports per month included in “Other.”
3 Free alongside ship (FAS) value.
4 Less than ½ unit.
Appendix A • 463
TABLE 4
U.S. IMPORTS FOR CONSUMPTION OF IRON ORE,
BY COUNTRY AND TYPE1, 2
(Exclusive of ore containing 20% or more manganese)34
2016 2017
Country
January–March March January–March
of origin
and type of Quantity (thousand Quantity Value3 Quantity Value3 Value3
product metric tons) (thousand (thousand (thousand (thousand (dollars
metric tons) dollars) metric tons) dollars) per ton)
Peru – 11 1,290 20 9,910 495.30
Other 481 (4) 10 628 55,200 87.96
Total 481 11 1,300 648 65,100 100.53
Concentrates 92 – – 99 7,530 76.07
Fine Ores – 11 1,300 20 9,930 496.60
Pellets 339 – – 529 47,700 90.13
Other 50 – – – – –
Total 481 11 1,300 648 65,100 100.53
– Zero.
2 All countries with less than 5,000 metric tons of imports per month included in “Other.”
3 Customs value. Excludes international freight and insurance charges.
TABLE 5
U.S. IMPORTS FOR CONSUMPTION OF IRON ORE,
BY CUSTOMS DISTRICT1, 2
(Exclusive of ore containing 20% or more manganese)
(Thousand metric tons)3
Pellets Total, all products
Customs district (code no.) January–March March January–March March
2016 2017 2017 2016 2017 2017
Baltimore, MD (13) – – – – (3) –
Los Angeles, CA (27) – – – – 11 11
New Orleans, LA (20) 339 529 – 417 637 –
Other – – – 64 (3) –
Total 339 529 – 481 648 11
– Zero.
2 Customs Districts with less than 5,000 metric tons of imports per month included in “Other.”
APPENDIX B
Iron and Steel (U.S. and World)12
By Michael D. Fenton
The world steel industry continued to add more production capacity dur-
ing the period from 2009 to 2014 despite continuing weak demand as the
world economy recovered slowly from the economic downturn in 2008–9.
By 2014, global overcapacity was an estimated 330 million metric tons (Mt)
with an average utilization rate of less than 80%. North American steel
plants managed to limit overcapacity, whereas substantial overcapacity
developed in Central Asia, the European Union (EU), Russia, and South
America. Overcapacity in China, estimated at 200 Mt, has reached the point
where the Chinese Government has limited bank financing to add capacity
and has required the closure of small, unprofitable, and polluting plants.
Implementation by local governments that are concerned about economic
development and social stability continues to be an issue. Similar problems
of closures and planned closures took place in Belgium, France, Italy, and
Serbia. Apparent consumption in China, the world’s leading consumer of
steel, increased steadily from 101 Mt in 1995 to a high of 766 Mt in 2013
and then decreased to 740 Mt in 2014 (World Steel Association, 2015a).
Production
Raw steel production in the United States was about 88.2 Mt in 2014, up
slightly from that in 2013 (table 1). U.S. capacity utilization for raw steel
production, which had dipped to a low of 41% in April 2009, rose steadily
to a high of 80.2% in August 2014, ending at 74.6% in December 2014. The
American Iron and Steel Institute (AISI) estimated raw steel production
capacity in 2014 to be 114 Mt, up slightly from that in 2013. Production
1 Source: USGS
https://minerals.usgs.gov/minerals/pubs/commodity/iron_&_steel/
2 Domestic survey data and tables were prepared by Hoa P. Phamdang, statistical assistant, and
world production tables were prepared by Glenn J. Wallace, international data coordinator.
represented 77.5% of estimated capacity, up from 76.7% in 2013 (American

Iron and Steel Institute, 2015, p. 73).
Integrated steel producers smelted iron ore to make liquid iron in
blast furnaces and used basic oxygen furnaces (BOFs) to refine the liquid
iron, with some steel scrap, to produce raw liquid steel. The BOF pro-
cess was used to make 33.0 Mt of steel in the United States (American
Iron and Steel Institute, 2015, p. 70). The use of this process decreased
to 37.4% of total steel production in 2014 from 39.4% in 2013. Blast
furnaces in the United States were operated by three companies at 11
locations in 2014. AK Steel and Severstal North America closed one
blast furnace each, both in Dearborn, MI, during 2014 (Iron and Steel
Technology, 2015b, p. 294).
Minimills and specialty mills are nonintegrated steel producers that
use electric arc furnaces (EAFs) to melt low-cost raw materials (primarily
scrap). They also employ continuous casting machines and hot-rolling mills
that are often closely coupled to casting operations. Specialty mills include
producers of electrical alloys, stainless, and tool steel; high-temperature
alloys; forged ingots; and other low-volume steel products. About 58 com-
panies operated about 109 EAF facilities in the United States during 2014
(Iron and Steel Technology, 2015a, p. 158). These mills produced 55.2 Mt
of steel in 2014, 5% more than in 2013, and accounted for 62.6% of total
steelmaking (American Iron and Steel Institute, 2015, p. 70).
Raw liquid steel is mostly cast into semifinished products in continu-
ous casting machines. Only 1.5% of U.S. production was cast in ingot form
in 2014 and subsequently rolled into semifinished forms, a slightly higher
percentage than that of 2013. Continuous casting production was 86.8 Mt,
or 98.5% of total steel production, slightly higher than in 2013 (American
Iron and Steel Institute, 2015, p. 70).
Consumption and Shipments

Steel mill products are produced either by forging or by rolling into forms
normally delivered for fabrication or use. Some companies purchase semi-
finished steel mill products from other steel companies and use them to
produce finished steel products. The accumulated shipments of all compa-
nies less the shipments to other reporting companies are identified as net
shipments to avoid double counting.
U.S. apparent steel consumption, an indicator of economic growth,
which had decreased to a low of 63 Mt in 2009, after a high of 120 Mt in
Appendix B • 467
2006, increased to 108 Mt in 2014 from 99 Mt in 2013. Net shipments of

steel mill products by U.S. companies increased by 3% to 89.1 Mt com-
pared with those of 2013 (American Iron and Steel Institute, 2015, p. 23).
Compared with those in 2013, shipments of construction products, the
leading single end-use market for steel, increased by 10.6% in 2014; auto-
motive product shipments decreased slightly; shipments of agricultural
products increased by 15%; shipments of industrial machinery, equip-
ment, and tools increased by 9.8%; oil, gas, and petrochemical shipments
increased by 10.1%; steel service center shipments decreased slightly;
lumbering, mining, and quarrying industries shipments decreased by 17%;
shipments of appliances increased by 8.3%; and shipments of containers,
packaging, and shipping material decreased by 50% (American Iron and
Steel Institute, 2015, p. 27).
Prices
The U.S. Bureau of Labor Statistics (2015) producer price index for steel
mill products increased slightly to 200.2 in 2014 from 195.0 in 2013 (1982
base = 100) (table 1). The average monthly price of hot-rolled steel sheet
fluctuated within a narrow range from a high of $447 per metric ton in
January 2014 to a low of $599 per metric ton in December 2014 (American
Metal Market, 2014).
Foreign Trade
Export shipments of steel mill products by AISI reporting companies
decreased to 10.9 Mt from 11.5 Mt in 2013 (table 4). Canada received the
largest amount of U.S. exported steel products, 5.8 Mt, 5% less than that
in 2013 (table 4). Mexico was again the second-ranked export receiving
country, 3.8 Mt, 3% more than that in 2013. Domestic imports of steel
mill products increased by 37.9% to 40.2 Mt from 29.2 Mt in 2013 (table
4). Brazil, Canada, China, the EU, Japan, Mexico, the Republic of Korea,
Russia, and Turkey were major sources of steel mill product imports in
2014 (table 4).
Imports of semifinished steel (table 6) by steel companies are taken
into consideration in evaluating apparent consumption (supply) of steel
mill products in the United States and the share of the market represented
by imported steel. To avoid double counting the imported semifinished
steel and the products produced from it, the amount of semifinished
steel consumed by companies that also produced raw steel is subtracted
from domestic consumption. Between 1993 and 2014, semifinished steel
imports ranged between 2.5 and 9.6 million metric tons per year (Mt/yr).
Prior to 1993, the amount was less than 0.2 Mt/yr. Taking the imported
semifinished steel into consideration, the share of the U.S. steel market
represented by imported steel was an estimated 37% in 2014 compared
with 30% in 2013.
World Review
World production of pig iron totaled about 1.19 billion metric tons (Gt),
slightly more than that in 2013 (table 9). China continued to be the leading
producer of pig iron in the world, producing 712 Mt, slightly more than that
of 2013, followed by Japan (83.9 Mt), India (55.2 Mt), Russia (51.5 Mt), the
Republic of Korea (46.9 Mt), the United States (29.4 Mt), Germany (27.4
Mt), Brazil (27.0 Mt), and Ukraine (24.8 Mt). Russia and Ukraine were the
only major pig iron producers in the Commonwealth of Independent States
(CIS), where production in 2014 decreased by 5% from that in 2013. In
South America, the only major pig iron producer was Brazil. Germany was
the leading producer in the EU, producing about 27.4 Mt, slightly more
than that in 2013.
World capacity for direct-reduced iron (DRI) production in 2014 was
estimated to be about 108 Mt/yr (Midrex Technologies, Inc., 2015, p. 11).
DRI production worldwide increased by 6.6% to 71.3 Mt in 2014 from 66.9
Mt in 2013 (table 9). The leading producer of DRI was India (20.4 Mt),
followed by, in descending order of tonnage, Iran (14.6 Mt), Saudi Arabia
(6.46 Mt), Mexico (5.98 Mt), and Russia (5.20 Mt). In 2014, additional DRI
capacity of almost 16 Mt/yr was under construction in China, Egypt, India,
Russia, the United States, and Venezuela. The leading technology, accord-
ing to declining order of production, was the Midrex process, followed by
coal-based and HYL/Energiron.
World production of raw steel was 1.67 Gt, 3% more than that in
2013 (table 10). Steel production during 2014 increased in North America
(1.8%), decreased in the CIS (2.5%), and decreased in the EU (1.0%).
Positive growth also took place in India (6.5%) and China (5.6%). As in
previous years, production varied widely among major regions of the world.
China produced 49% of world total crude steel in 2014. Countries in Asia
produced about 68% of the world’s steel; the EU, 10%; North America, 7%;
and the CIS, 6%.
According to the World Steel Association (2015b), China was the
top steel producer in the world during 2014 (823 Mt), with its leading
Appendix B • 469
steelmaker, Hebei Iron & Steel Group Co. Ltd. (47.1 Mt), placing third
behind the world’s leading steelmakers–ArcelorMittal (98.1 Mt) and
Nippon Steel and Sumitomo Metal Corp. (49.3 Mt). Japan was the world’s
second-ranked steel producer (111 Mt), followed by the United States
(88.2 Mt), India (86.5 Mt), Russia (71.5 Mt), and the Republic of Korea
(71.0 Mt). These six countries accounted for 75% of world production.
Russia and Ukraine remained the leading producers in the CIS. China
ranked first in the world for total exports of steel during 2014 (92.9 Mt),
followed by Japan (41.3 Mt) and the EU (37.1 Mt). The United States
ranked first in the world for total imports of steel in 2014 (41.4 Mt), fol-
lowed by the EU (32.4 Mt). Increases in steel production during 2014 took
place in Africa (3.1%), Asia (3.9%), the CIS (2.7%), the EU (1.0%), and
North America (1.8%). Steel production decreases during 2014 took place
in South America (3.3%) and Africa (4.7%).
Outlook
The expansion or contraction of gross domestic product (GDP), the broad-
est measure of a nation’s economic activity, may be considered a predic-
tor of the health of the steelmaking and steel manufacturing industries,
worldwide and domestically. The World Bank’s forecast of global GDP
growth for 2015, 2016, and 2017 is 2.8%, 3.3%, and 3.2%, respectively,
after 2.6% in 2014 (World Bank, The, 2015). The International Monetary
Fund (2015) projection of GDP growth is 3.3% for 2015 and 3.8% for
2016. The U.S. Federal Reserve’s September 2015 projections for GDP
rate of growth for the United States are 2.1% for 2015, 2.3% for 2016, and
2.2% for 2017 (Board of Governors of the Federal Reserve System, 2015).
The 2014 rate of GDP growth for China was 7.4% and is projected to be
6.8% in 2015 and 6.3% in 2016. The rate of GDP growth for India was
7.3% in 2014 and is projected to be 7.5% in 2015 and 2016 (International
Monetary Fund, 2015).
MEPS (International) Ltd. forecast total world steel production in 2015
to be 1.62 Gt, down slightly from that in 2014 (World Steel News, 2015).
MEPS also forecast an increase to 1.64 Gt in 2016.
World apparent steel consumption (ASC) is forecast to be 1,513 Mt
and 1,523 Mt in 2015 and 2016, respectively, after consumption in 2014
was 1,540 Mt. (World Steel Association, 2015c). China’s ASC is expected
to decrease to 672 Mt in 2016 from 711 Mt in 2014. The ASC in India is
expected to increase to 87.6 Mt in 2016 from 75.9 Mt in 2014. The ASC in
the United States is expected to decrease to 105 Mt in 2016 from 107 Mt in

2014. The EU’s ASC is expected to increase to 153 Mt in 2016 from 148 Mt
in 2014. In Japan, the 2015 ASC is expected to decrease to 66 Mt in 2016
from 68 Mt in 2014. The ASC of the CIS is expected to decrease to 49.9 Mt
in 2016 from 56.1 Mt in 2014.
References Cited
American Iron and Steel Institute, 2015, Annual statistical report 2013:
Washington, DC, American Iron and Steel Institute, 120 p.
American Metal Market, 2014, Custom pricing reports: New York, NY,
American Metal Market. (Accessed August 4, 2014, via http://www.amm.com.)
Benteler Steel/Tube GmbH, 2013, Benteler Steel/Tube breaks ground at
$975M U.S. facility in Louisiana: Houston, TX, Benteler Steel/Tube GmbH
press release, September 16. (Accessed October 10, 2014, at http://www.
fabricatingandmetalworking.com/2013/10/benteler-steeltube-breaks-
ground-at-97SM-u-s-facility-in-louisiana/)
Board of Governors of the Federal Reserve System, 2014, Minutes of the
Federal Open Market Committee, Summary of economic projections,
September 16–17, 2014: Board of Governors of the Federal Reserve System.
(Accessed June 24, 2016, at http://www.federalreserve.gov/monetarypolicy/
fomcminutes20140917ep.htm.)
Board of Governors of the Federal Reserve System, 2015, Minutes of
the Federal Open Market Committee, Summary of economic projec-
tions, September 16–17, 2015: Board of Governors of the Federal Reserve
System. (Accessed January 27, 2016, at http://www.federalreserve.gov/
monetarypolicy/fomcminutes20150917ep.htm.)
Business Wire, 2013, Borusan Mannesmann pipe partners with Siemens
for its first U.S. manufacturing facility: New York, NY, Business Wire press
release, June 4. (Accessed April 23, 2014, at http://www.businesswire.com/
news/home/20130604006484/en/Borusan-Mannesmann-Pipe-Partners-
Siemens-U.S.-Manufacturing#.VFJrmPnIako.)
Ernst & Young, 2013, Global steel 2013—A new world, a new strat-
egy: Ernst & Young, 44 p. (Accessed September 5, 2014, at http://www.
ey.com/Publication/vwLUAssets/Global_steel_2013_-_A_new_world_-
_a_new_strategy/$FILE/Global_Steel_Report_India_conference_
Feb_2013.pdf.)
Appendix B • 471
Ernst & Young, 2014, Global steel 2014: Ernst & Young, 41 p. (Accessed
September 5, 2014, at http://www.ey.com/Publication/vwLUAssets/EY_-_
Global_steel_2014/$FILE/EY-Global-steel-2014.pdf.)
International Monetary Fund, 2015, Slower growth in emerging markets,
a gradual pickup in advanced economies: New York, NY, International
Monetary Fund press release, July 24. (Accessed January 27, 2016, at http://
www.imf.org/external/pubs/ft/weo/2015/update/02/pdf/0715.pdf.)
Iron and Steel Technology, 2015a, AIST2015 EAF roundup: Warrendale,
PA, Iron and Steel Technology, January, 280 p.
Iron and Steel Technology, 2015b, AIST2015 North American blast furnace
roundup: Warrendale, PA, Iron and Steel Technology, March, 352 p.
Midrex Technologies, Inc., 2015, 2014 world direct reduction statistics:
Charlotte, NC, Midrex Technologies, Inc., 14 p. (Accessed August 4, 2014,
via http://www.midrex.com/.)
Miller, L.B., 2014, Developments in the North American iron and steel
industry—2013: Warrendale, PA, Iron and Steel Technology, April, p.
192–212.
Nguyen, Lananh, 2012, U.S. oil output to overtake Saudi Arabia’s by 2020:
Bloomberg News, November 12. (Accessed October 10, 2014, at http://
www.bloomberg.com/news/2012-11-12/u-s-to-overtake-saudi-arabia-s-oil-
production-by-2020-iea-says.html.)
Smith, Grant, 2014, U.S. seen as biggest oil producer after overtaking Saudi
Arabia: Bloomberg News, July 4. (Accessed October 10, 2014, at http://
www.bloomberg.com/news/articles/2014-07-04/u-s-seen-as-biggest-oil-
producer-after-overtaking-saudi.)
U.S. Bureau of Labor Statistics, 2015, Producer price index highlights:
Washington, DC, U.S. Bureau of Labor Statistics. (Accessed January 6,
2016, at http://www.bls.gov/xg_shells/ro4xgppihi.htm.)
World Bank, The, 2015, Global economic prospects: Washington, DC, The
World Bank, June. (Accessed January 27, 2016, at http://www.worldbank.
org/en/publication/global-economic-prospects/summary-table.)
World Steel Association, 2015a, Steel statistical yearbook 2015: Brussels,
Belgium, World Steel Association.
World Steel Association, 2015b, World steel in figures 2015: Brussels,
Belgium, World Steel Association. (Accessed January 13, 2014, at http://
www.worldsteel.org/publications/bookshop/product-details.~World-Steel-
in-Figures-2015~PRODUCT~World-Steel-in-Figures-2015~.html.)
World Steel Association, 2015c, World steel short range outlook 2015–
2016: Brussels, Belgium, World Steel Association press release, April 20.
(Accessed January 27, 2016, at https://www.worldsteel.org/media-centre/
press-releases/2015/worldsteel-Short-Range-Outlook-2015–2016.html.)
World Steel News, 2015, MEPS expects a slight recovery in global steel pro-
duction in 2016: World Steel News, November 11. (Accessed January 27,
2016, at http://www.worldsteelnews.com/meps-expects-a-slight-recovery-in-
global-steel-production-in-2016/.)
General Sources of Information

U.S. Geological Survey Publications
Iron. Ch. in United States Mineral Resources, Professional Paper 820, 1973.
Iron and Steel. Ch. in Mineral Commodity Summaries, annual.
Iron and Steel (Fe). Ch. in Metal Prices in the United States Through 2010,
Scientific Investigations Report 2012–5188, 2013.
Iron and Steel Scrap. Mineral Industry Surveys, monthly.
Iron and Steel Scrap (Fe). Ch. in Metal Prices in the United States Through
2010, Scientific Investigations Report 2012–5188, 2013.
Iron and Steel Slag. Ch. in Mineral Commodity Summaries, annual.
Iron and Steel Slag. Ch. in Minerals Yearbook, annual.
Iron Ore. Ch. in Mineral Commodity Summaries, annual.
Iron Ore. Ch. in Minerals Yearbook, annual.
Iron Ore. Mineral Industry Surveys, monthly.
Other
American Metal Market, daily.
Annual Statistical Report, American Iron and Steel Institute,
Washington, DC.
Direct from Midrex quarterly, Midrex Direct Reduction Corporation.
Appendix B • 473
Directory of Iron and Steel Plants, Association of Iron and Steel Engineers,
Pittsburgh, PA.
HYL, the Iron & Steel Technology Division of HYLSA, S.A. de C.V.
HYLReport.
Iron and Steel. Ch. in Mineral Facts and Problems, U.S. Bureau of Mines
Bulletin 675, 1985.
Iron and Steelmaker, American Institute of Mining, Metallurgical, and
Petroleum Engineers—Iron and Steel Society, Warrenton, PA.
Making, Shaping, and Treating of Steel, Association of Iron and Steel
Engineers, Pittsburgh, PA.
Metal Bulletin, weekly.
Steel Manufacturers Association, Washington, DC.
Steel Statistical Yearbook, International Iron and Steel Institute, Brussels,
Belgium.
Steel Times International.
APPENDIX C
Iron and Steel Slag (January 2017)1
(Data in million metric tons unless otherwise noted)
Domestic Production and Use: Iron and steel (ferrous) slags are coproducts
of iron and steel manufacturing. After cooling and processing, ferrous slags are
sold primarily to the construction industry. Data are unavailable on actual U.S.
ferrous slag production, but it is estimated to have been in the range of 15 to 20
million tons in 2016. Domestic slag sales1 in 2016 amounted to an estimated
18 million tons, valued at about $350 million (ex-plant). Iron (blast furnace)
slag accounted for about 47% of the tonnage sold and had a value of about
$300 million; about 85% of this value was from sales of granulated slag. Steel
slag produced from basic oxygen and electric arc furnaces accounted for the
remainder.2 Slag was processed by about 29 companies servicing active iron
and steel facilities or reprocessing old slag piles at about 135 processing plants
in 31 States; included in this tally are some facilities that grind and sell ground
granulated blast furnace slag (GGBFS) based on imported unground feed.
Prices listed in the table below are weighted averages (rounded) for
iron and steel slags sold for a variety of applications. Actual prices per
ton ranged widely in 2016, from a few cents for some steel slags at a few
locations to about $110 for some GGBFS. Air-cooled iron slag and steel
slag are used primarily as aggregates in concrete (air-cooled iron slag
only), asphaltic paving, fill, and road bases; both slag types also can be
used as a feed for cement kilns. Almost all GGBFS is used as a partial
substitute for portland cement in concrete mixes or in blended cements.
Pelletized slag is generally used for lightweight aggregate but can
be ground into material similar to GGBFS. Owing to low unit values,
most slag types can be shipped only short distances by truck, but rail and
waterborne transportation allow for greater distances. Because of much
higher unit values, GGBFS can be shipped longer distances, including
from overseas.
1 Source: U.S. Geological Survey, Mineral Commodity Summaries, January 2017

https://minerals.usgs.gov/minerals/pubs/commodity/iron_&_steel_slag/mcs-2017-fesla.pdf
TABLE 1: STATISTICS2
Salient Statistics–United States: 2012 2013 2014 2015 2016e
Production (sales)1, 3 16.0 15.5 16.6 17.7 18.0
Imports for consumption4 1.2 1.7 1.8 1.4 2.0
Exports (5) (5) 0.1 (5) (5)
Consumption, apparent4, 6 16.0 15.5 16.5 17.7 18.0
Price average value, dollars per ton, 17.00 17.50 19.00 19.50 19.50
f.o.b. plant7
Employment, numbere 1,800 1,700 1,700 1,700 1,600
Net import reliance8 as a percentage

of apparent consumption 7 11 10 8 11
e Estimated.
1 Data are from an annual survey of slag processors and pertain to the quantities of processed slag
sold rather than that processed or produced during the year. The data exclude any entrained
metal that may be recovered during slag processing and then sold separately or returned to iron
and, especially, steel furnaces. The data are incomplete regarding slag returns to the furnaces.
2 There were very minor sales of open hearth furnace steel slag from stockpiles but no domestic
production of this slag type in 2012–16.
3 Data include sales of imported granulated blast furnace slag, either after domestic grinding or
still unground, and exclude sales of pelletized slag (proprietary but very small). Overall, actual
production of blast furnace slag may be estimated as equivalent to 25% to 30% of crude (pig)
iron production and steel furnace slag as about 10% to 15% of crude steel output.
4 Based on official (U.S. Census Bureau) data. In some years, the official data appear to have
understated the true imports; the apparent discrepancy was small for 2012, but may have been
nearly 0.4 million tons in 2013 and 2014, depending on whether imports from Italy were mis-
characterized as being from Spain or not. The U.S. Geological Survey canvass captures only
part of the imported slag.
5 Less than 0.05 million tons.
6 Although definable as total sales of slag (including those from imported feed) − exports, appar-
ent consumption of slag does not significantly differ from total sales owing to the very small
export tonnages.
7 Rounded to the nearest $0.50 per ton.
8 Defined as imports minus exports.
Appendix C • 477
Recycling: Following removal of metal, slag can be returned to the blast

and steel furnaces as ferrous and flux feed, but data on these returns are
incomplete. Entrained metal, particularly in steel slag, is routinely recov-
ered during slag processing for return to the furnaces, and is an impor-
tant revenue source for the slag processors, but data on metal returns are
unavailable.
Import Sources (2012–15): The dominant imported ferrous slag type is
granulated blast furnace slag (mostly unground), but official import data in
recent years have included significant tonnages of nonslag materials (such
as cenospheres, fly ash, and silica fume) and slags or other residues of vari-
ous metallurgical industries (such as copper slag) whose unit values are out-
side the range expected for granulated slag. The official data appear to have
underreported the granulated slag imports in some recent years, but likely
not in 2011–12. Based on official data, the principal country sources for
2012–15 were Japan, 33%; Canada, 31%; Spain, 16%; Germany, 5%; and
other, 15%; however, much of the tonnage from Spain in 2013–14 may in
fact have been from Italy, and slag from the Netherlands and Switzerland
in 2015 may have been from Germany.
Tariff : Item Number Normal Trade Relations 12–31–16
Granulated slag 2618.00.0000 Free.
Slag, dross, scale, from manufacture
of iron and steel 2619.00.0000 Free.
Depletion Allowance: Not applicable.

Government Stockpile: None.
(Prepared by Hendrik G. van Oss [(703) 648–7712, [email protected]])
Iron and Steel Slag

Events, Trends, and Issues: The supply of blast furnace slag continues to
be problematic in the United States because of the closure and (or) contin-
ued idling of a number of active U.S. blast furnaces in recent years, includ-
ing one in 2015, the lack of construction of new furnaces, and the depletion
of old slag piles. Only a limited quantity of locally produced granulated
blast furnace slag was available. At yearend 2016, granulation cooling was
available at only two active U.S. blast furnaces, down from three in 2014.
Installation of granulation cooling was being evaluated at a few blast fur-
naces, but it was unclear if this would be economic. Pelletized blast furnace
slag was in very limited supply (one site only), and it was uncertain if any
additional pelletizing capacity was planned.
Basic oxygen furnace steel slag from domestic furnaces has become
less available recently because of the closure of several integrated iron and
steel complexes; thus, the long-term supply of steel slag will be increasingly
reliant on electric arc furnaces, which now contribute the majority of U.S.
steel production. Where not in short supply, slag (as aggregate) sales to the
construction sector tend to fluctuate less than those of natural aggregates.
Domestic- and import-supply constraints appear to have limited domestic
demand for GGBFS in recent years. Although prices have increased, sales
volumes have failed to match the relative increases that have characterized
the overall U.S. cement market since 2010. Long-term demand for GGBFS
likely will increase because its use in concrete yields a superior product in
many applications and reduces the unit carbon dioxide (CO2) emissions
footprint of the concrete related to the portland cement (clinker) content.
Recent regulations to restrict emissions of CO2 and mercury by coal-
fired powerplants, together with some powerplant closures or conversion of
others to natural gas, have led to a reduction in the supply of fly ash in some
areas, including that of material for use as cementitious additive for con-
crete. This has the potential to increase future demand for GGBFS, but the
availability of material to supply this demand will increasingly depend on
imports, either of ground or unground material. Imported slag availability
may be constrained by increasing international demand for the same mate-
rial and because not all granulated slag produced overseas is of high qual-
ity. New restrictions on mercury emissions by cement plants may reduce
demand for fly ash as a raw material for clinker manufacture, and this could
lead to use of air-cooled and steel slags as replacement raw materials.
World Mine Production and Reserves: Slag is not a mined material
and thus the concept of reserves does not apply to this mineral commodity.
Slag production data for the world are unavailable, but it is estimated that
global iron slag output in 2016 was on the order of 300 to 360 million tons,
and steel slag about 160 to 240 million tons, based on typical ratios of slag
to crude iron and steel output.
World Resources: Not applicable.
Substitutes: In the construction sector, ferrous slags compete with
crushed stone and sand and gravel as aggregates, but are far less widely
available than the natural materials. As a cementitious additive in blended
Appendix C • 479
cements and concrete, GGBFS mainly competes with fly ash, metakaolin,
and volcanic ash pozzolans, and to a lesser degree with silica fume. In this
respect, GGBFS also competes with portland cement itself. Slags (espe-
cially steel slag) can be used as a partial substitute for limestone and some
other natural raw materials for clinker (cement) manufacture. Some other
metallurgical slags, such as copper slag, can compete with ferrous slags in
some specialty markets but are generally in much more restricted supply
than ferrous slags.
APPENDIX D
Bibliography
1. Rao K.N.P., Souvenir, Annual Tech Meeting, IIM, Hydrabad, 1979,
p. 1.
2. Sriniwasan, S and Ranganathan, S, India’s Legendary Wootz Steel—
An Advanced Material of the Ancient World, Tata Steel, 1996.
3. Ranganathan, S., Iron Heritage in India, IIM and Tata Steel, Jamshedpur,
1997.
4. Biswas, A.K., Biswas, Sulekha, Minerals and Metals I Ancient India,
Vol. 1 and 2, D.K. Printworld, New Delhi.
5. Krishan, R.M. et al., Proc. Symp Iron and Steel Industry in India,
NML, Jamshedpur, 1959, p. 342.
6. Rao, K.N.P., Metals News, 1991, Vol. 13, No. 5, p. 9.
7. Balasubramaniam, R., Iron Pillar—A metallurgical Marvel, Founda-
tion, 2005.
8. Kumar P.P., Gupta, D., Naha T.K., and Gupta S.S., Ironmaking and
Steelmaking, 2006, Vol. 33, No. 4, p. 293.
9. Survey of Electric Furnace Industry in India, Report of Joint Plant
Committee, 2004, Sept., Vol. 1 and 2.
10. Chatterjee, A., Talk at Bhubaneshwar, 28.9.2007.
11. Annamalai, G., et al., Tata Search, 1994, p. 29.
12. United Nations Report, “Economic Aspects of Iron Ore Preparation”
Geneva, 1996.
13. U.S. Geological Survey, Jan. 2007.
14. Iron Ores—A Market Survey, IBM, GOI, Oct. 2007.
15. Laskar, B., Commerce, Annual Number, 1976, Vol. 133, p. 145.
16. Basu T.N. and Ghosh, S.K., Proc. Semi, Energy for Iron and Steel
Industry, IIM, Rourkela, 1977, p. 1-1-1.
17. Chatterjee, A., Trans IIM, 1975, Vol. 28 (1), p. 44.
18. Krishna M.G., Ref. No. 16, p. 1-3-1.
19. Dasgupta, N.N., et al., J Mines Metals and Fuels, 1958, 1, p. 53.
20. Lahiri, A and raja, K., J Inst. Engr. (I), Chem. Engg., 1971, 52 (2), p. 3.
21. Singh B.N. and Das A., Metals new Bulletin, 1993, Vol. 15 (2), p. 8.
22. Rao, T.T.V. and Chatterjee, A., Tata Search, 1994, p. 14.
23. Das, A.K. et al., Metals News, 1992, Vol. 14 (5), p. 1.
24. Sandhu H.S., et al. Tata Search, 1999, p. 21.
25. Pandey B.D., Tata Search, 1999, p. 25.
26. Singh B.K. et al., Tata Search, 1998, p. 13.
27. Roy T.K., IIM, Metals News, 2007, Vol. 10 (5), p. 14.
28. (a) Raja K., et al., Trans IIM, 1973, Vol. 26 (2), p. 39.
(b) Paretkar, R.K. and Pathak, S.U., Proc. All India Sem on Recent
trends in ferroalloys Tech., Inst. of Engrs. (I), Nagpur Centre,
1978, p. 221.
29. Zherebine, B.N. and Mojumdar, B.N., Trans IIM, 1968, Vol. 21 (2),
p. 103.
30. Suzuki, G., et al., Silver Jub Symp., IIM, New Delhi, Iron and Steel
Section, 1972, p. 191.
31. Ojha, S.K., et al., All India Sem on Extraction of Iron, IIM, Bhilai,
1984, p. II-1.
32. Poveremo, J.J., and Hlinka, J.W., Third Proc. Tech. Conf. Pittsburg,
1982.
33. Davison, J., Ref. No. 30, p. 1.
34. Bashforth, G.R., Manuf of Iron and Steel, Vol. 1, Iron Production,
ASIA, 1959.
Appendix D • 483
35. (a) Mahakud, B.N. and Sharma, K.K., J Inst. Engrs. (I), MM, 1974,
Vol. 55, p. 18.
(b) Murthy V.G.K., et al. Tata Search, 1994, p. 7.
36. Wild, R., Symp Advances in Extractive Met, Inst. MM, London, 1967,
paper 4.
37. (a) Rao, J.A. and Sooch, J.S., Trans IIM, 1972, Vol. 25 (4), p. 74.
(b) Monograph on sintering at Tata Steel, Ed. Chatterjee, A., et al., p. 45.
38. Mukherjee, T., et al. Tata Search, 1996, p. 16.
39. Lonial, S.K. and Verma, A.K., Tata Search, 1997, p. 1.
39. (a) Wegman, E.F., Trans IIM, (1973), 26(3), p. 61.
40. Uday Kumar, C., et al. Tata Search, 1995, p. 20.
41. Nishida, R. et al., Proc. Int. Conf. Science and Tech. of Iron and Steel,
part I, ISI Japan, 1970, p. 70.
42. Mishra, V.N., Trans IIM, 1973, Vol. 26 (4), p. 24.
43. Singhai, A.K., Trans IIM, 1973, Vol. 26 (5), p. 33.
44. Traice B.P. and Lappin R.L., Developments in Ironmaking Practice,
Iron and Steel Inst., London 1973, p. 17.
45. Carter, G.C. and Write, F., Ref. No. 36, paper 6.
46. Vidal, S. et al. Trans IIM, 1973, Vol. 26 (3), p. 14.
47. Kodama, K and Hashimoto, S, Ref. No. 41, p. 112.
48. Yad U.S., et al. Tata Search, 1997, p. 10.
49. Hatch, G.G. and Chipman, J, Trans AIME, 1949, Vol. 167, p. 134.
50. Singh U., et al. Tata Search, 2007, p. 183 S.
51. Brandi, H.T., et al., Ref. No. 41, p. 176.
52. Mojumdar, B.N., Trans IIM, 1968, Vol. 21 (4), p. 41.
53. Mojumdar, B.N., Trans IIM, Vol. 27 (1), p. 56.
54. Jakkiwar, M.S. and Tupkary, R.H. Ref. No. 28 (b), p. 101.
55. Didkovskii, V.K., Stahl in English, 1967, Vol. 12, p. 988.

56. Brugner, R., Trans IIM, 1972, Vol. 25 (4), p. 74.
57. Ramasubbu, N.C., Proc. Modernisation, Steel Plants, IIM, Bhilai,
1978, p.c.4.1.
58. Mcky, J.C. and Peart, A., Proc. Ironmaking, AIME, 1963, Vol. 22, p. 87.
59. Wahlster, M, Stahl Eisen, 1961, Vol. 89, p. 9.
60. Bhasi, M. et al. Tata Search, 1998, p. 188.
61. Pandey, B.D., et al., Tata Search 2003, p. 95.
62. Trognitz, W.R., Blast Furnace Steel Plant, 1964, Vol. 52, p. 315.
63. Shriaishi, Y., et al., Testu-to-Hagane, 1964, Vol. 4, p. 393.
64. Leview, L.Y., et al. Stahl in English, 1962, p. 503.
65. Jayaram and Viswanatham, S., TISCO, 1966, Vol. 13 (4), p. 153.
66. Chan, M.C., Blast Furnace Steel Plant, 1967, Vol. 55 (11), p. 1019.
67. Viswanatham, S., TISCO, 1970, Vol. 17 (4), p. 111.
68. Mitra, A.N. et al., Steel Indian, 1980, Vol. 3 (2), p. 49.
69. Nakamura, N., Ref. No. 16, p. xi.
70. Singh, R.N. et al., Trans IIM, 1979, Vol. 32 (2), p. 90.
71. Roy, T.K., IIM Metal News, 2007, Vol. 10 (10), p. 14.
72. Pandey, B.D. and Jha, U.K., Iron and Steelmaking Symp. 1999, March,
p. 39.
73. Aswath, H.S., Ref. No. 41, p. 78.
74. Heynert, G. and Legile, E., Ref. No. 41, p. 88.
75. Jansen, H.W. and Hilger, H.T., Blast Furnace Steel Plant, 1970, Vol.
58 (3), p. 171.
76. Agayev, K., Iron & Steel Review, 2006, April, p. 56.
77. Jha, M.N. et al., Tata Search, 1997, p. 15.
78. Sanjaykumar, et al., Tata Search, 1994, p. 35.
Appendix D • 485
79. Chatterjee, A., Metal News, 1995, Vol. 17, p. 1.

80. Girard, M., Ref. No. 41, p. 176.
81. Chatterjee, A and Ghosh, A., Ironmaking and Steelmaking—Theory
and Practice PHI, 2008.
82. Malcore, H., Ref. No. 5 (2), p. 230.
83. Chatterjee, A.B., Ref. No. 5 (2), p. 244.
84. Chatterjee, A., Ironmaking and Steelmaking, 1996, Vol. 23 (4), p. 293.
85. Patel, N.K., Metal News, 1991, Vol. 13 (5), p. 2.
86. Anderson, H.C., Can Min Met Bulletin, 1963, July, p. 509.
87. Raja, B.V.R. and Pal, N. Metal News, 2006, Vol. 9 (2), p. 11.
88. Vaish, A.K., Gupta, R.C. and Mehrotra, S.P., Metal News, 2007,
Vol. 10 (2), p. 6.
89. Das, S.N. and Pal, P.K., Metal News, 1994, Vol. 16 (2), p. 7.
INDEX
A foundation, 20, 325

gas cleaning, 24, 260
Abrasion test
hearth, 20, 32, 215
for coke, 64
instrumentation, 386
for ore, 177
large capacity, 323
Acid burdening, 301
operation, 27–28
Agglomeration, 114
profiles, 188
briquetting, 114
raw materials, 35
of burden, 114, 116, 146
stoichiometry, 197
Alternative fuels, 76
thermal efficiency, 227
Alternative routes of iron
Blast furnace burden
production, 380, 391, 403
acid or basic, 292
coal based, 420
bulk reducibility, 100
gas based, 406
characteristics, 172
Alumina problem, 112, 146, 217
distribution, 63, 287, 331
Ancient ironmaking, 2
permeability, 64
Angle of repose, 86
preparation, 98, 114,
Ayas, 2
116, 329
qualities, 97
B testing, 175
Blast furnace charging, 7, 89
Back Drafting, 280
devices, 8, 81
Banking, 278
hoist, 253
Bell, 22
order, 89
Bell and hopper, 23–24
systems, 393
Big bell, 81, 335, 339
top designs, 340
Blast, 4
Blast furnace construction, 20, 231
humidified, 352, 354
bosh, 22, 29
oxygen enriched, 381
bustle pipe, 23
Blast furnace, 8, 17
charge distribution, 79
bosh, 22
hearth, 20, 32
charging devices, 8, 337
mantle, 22
control, 27–28, 352
stack, 22–23, 29–30, 222
efficiency, 32, 206
structure, 241
enthalpy balance, 223
tap holes, 247
Blast furnace cooling, 243 Blast furnace products, 259, 289

Blast furnace designs, 321 disposal, 25–26, 276, 302
chemical composition, 60 Fe-Mn, 297
cleaning, 260 gas, 256
downcomer, 260 metal, 289, 302
dust catcher, 7, 260 slag, 276
electrostatic precipitator, Blast furnace reactions, 28, 193
8, 265 in bosh, 205
function of, 59, 328 in hearth, 210
in India, 456 in stack, 202
preparation in India, 66 Blast furnace refractories, 236
primary cleaning, 263 Blast furnace top designs, 8, 90
reactivity, 60 with high top pressure, 332
scrubbers, 263 Bleeder, 260–261
secondary cleaning, 265 Bloom, 3
size, 63 Bloomery, 3
storage, 9, 253 Blowers, 255
thermal stability, 64 Blowing-in, 277
uptake, 260 Blowing-out, 279
utilization, 7, 256 Bosh, 22, 29
venturi washer, 265 reactions, 352
Blast furnace fuels, 7, 58 Bronze age, 2
Blast furnace gas, 7, 260 Burden testing, 175
Blast furnace operation, 27–28, 281 Bustle pipe, 23, 251
back drafting, 280
banking, 278
blowing-in, 277 C
blowing-out, 279 Carbon hearth, 237
breakouts, 285 Cast-house, 25, 249
charging sequence, 90 Catalan forge, 4
fanning, 280 Channeling, 286–287
filling, 276 Charcoal B.F., 389–390
irregularities in, 281 Charging appliances, 8, 353
lighting, 277 Chilled hearth, 284
sulphur problem, 214 Coal
tapping, 279 blend, 69
Blast furnace performance, 32, 305 dust injection, 322
coke rate, 202, 312 metallurgical, 52
effect of pellet charge, 314 non-metallurgical, 77
effect of sinter charge, 363 reserves, 53
effect of sizing, 312 reserves in India, 53
Blast furnace plant, 20 washing, 69
Index • 489
Coke, 8, 59 External desiliconization, 299–300

abrasion index, 64 External desulfurization, 299
breakdown, 66
chemical composition, 60
ferro, 77 F
formed, 76 Fanning, 280
function, 59, 328 Fe-Mn production, 297
IS-specifications, 65 Ferro-coke, 77
ovens, 63 Filling, 276
quality, 60, 329 Five year plan, 108
rate, 53, 350, 354 Flooding, 285–286
reactivity, 60 Flux, 53
shatter test, 64 evaluation of, 53
size, 62 quality, 319
strength, 64 Formed coke, 76
thermal stability, 64 Four bell top, 339–340
Coking coal blend, 68 Fuel injection, 317, 322, 351
Computer control, 374
Cowper stove, 7
Cup and cone, 7 G
Gangue, 29, 39
Gas cleaning, 7, 24, 260
D
Gimble top, 345
Decrepitation, 181
Desulfurization, 230, 299
Desulphurisation, 214 H
external, 299 Hanging, 282–283
Direct reduction, 28, 205, 398 Hearth, 20
Distribution of charge, 8, 80, 89, 331 breakouts, 285
Downcomer, 260–261 coolers, 243
DRI-iron, 11 decanting, 222
DR-iron, 404, 405 lining, 237
Driving rate, 228 reactions, 211
Dust catcher, 260 walls, 239
Sintering machine, 130 High top pressure,
Dwight-Lloyd, 126 332
charging system, 255
History of Ironmaking, 2
E
in India, 11, 388
Electric furnace, 12 Hot blast, 7
submerged arc, 393 storage, 296
Electrostatic precipitator, 8, 265 stove, 23, 280
Hot metal, 26, 228 K

Humidified blast, 352, 354
Krupp-Renn process, 406
HyL process, 11, 407
Kundremukh project, 113
I
Indian deposits, 46
preparation of, 102
L
reducibility, 179 Ladle, 26
structure, 39 Lime injection, 356–359
treatments in India, 101 Low shaft furnace, 11, 430
types, 41 L.T.B.T. test, 181–182
valuation, 49
washing, 112
world deposits, 45 M
Indirect reduction, 28, Mckee distributors, 88
205 Micum test, 64
Injections, 352 index, 65
coal dust, 322 Midrex process, 415, 419, 468
lime, 322 Mini Blast Furnaces, 380
moisture, 352, 353 Chinese, 386
oxygen, 352 Indian, 290
reducing gas, 353 Models, 369, 372
Iron age, 2 Modern Ironmaking, 9–10
Iron making, 2
history, 2
in India, 12, 456, 457 N
Iron mineral, 38 Naumann reversion, 193
characteristics, 38 Nodulizing, 115
classification, 39
Iron ores, 38
alumina problem, 114 O
beneficiation, 112 Oxygen enrichment, 381
classification, 39
export, 48
fines, 96, 112 P
gangue, 39 Pallet, 126
minerals, 38 Paul-Wurth Top, 344
Iron oxide reduction Pelletizer, 155
kinetics of, 205 disc, 157
solid-solid, 206 drum, 157
Isothermal zone, 225 Pelletizing, 114, 149
Index • 491
Pellets, 116, 149, 314 heat balance, 141

bonds, 152 in India, 141
growth, 155 mass balance, 142
Pig casting machine, 8, 26, 302, 303, 304 matching, 124
Pig iron, 5 process control, 136
composition, 291 use, 313
Pillaring, 284–285 Slag volume, 446
Pre-reduced iron, 431 disposal, 295
Probes, 371 Slip, 284
Process control, 368, 369 SL-RN process, 407, 420
Small shaft furnace, 430
Smelting reduction
R corex, 434
Raceway, 30 ELRED, 445
RAFT, 31, 200, 347 INRED, 435
Raw materials processes, 429
for ironmaking, 35 Sponge iron, 11, 403
for steelplant, 58 coal, 416
Reducibility test, 179–180 fluidised bed, 415
Rotary kiln, 416 gas based, 406
ring formation, 420 in India, 421
Runners, 250 processes, 447
SR processes, 429, 431
corex, 434
S finex, 434
Salamander, 287 Hismelt, 456
Scaffolding, 282 IT mk3, 452
Scrubbers, 263–264 Tupkary, 453
Shatter test Stack, 22–23, 29–30
for coke, 64 reactions, 202
for ore, 176 Stamp charging, 72–74
Sinter, 135, 316 Stockline
acid, 136 armour, 91, 332
bonds, 138 level, 93
breaker, 127 level indicator, 376
fluxed, 9, 136 Stove, 280
quality, 136 construction, 272
superfluxed, 301 Stukofen, 5
Sintering, 116, 135 Submerged arc furnace, 393–396
circulating load, 142 Sulphur problem, 217
economics, 144 Sulphur reactions, 214–217
on grate, 114 Swelling, 184
T V
Tap holes, 247–249 V-contour, 86, 92
Tapping, 250 Venturi washer, 265
Trompe, 4
Tuyeres, 22
assembly, 251 W
reactions, 201 Waste utilization, 362–363
zone, 31 Working index, 376–377
U
Uptake, 260–261

Iron Making by Tupkary

Uploaded by

Copyright:

Available Formats

Iron Making by Tupkary

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Iron Making by Tupkary

Uploaded by

Copyright:

Available Formats

MODERN

MLI Handbook Series

Mercury Learning and Information

Publisher: David Pallai

R. H. Tupkary and V. R. Tupkary. Modern Iron Making Handbook.

Library of Congress Control Number: 2017948166

171819321 Printed in the USA on acid-free paper.

2. MODERN BLAST FURNACE PLANT AND PROCESS ...... 19

2.3 Hot Blast Stove .................................................................................24

3. RAW MATERIALS FOR IRONMAKING .............................. 35

4. MAIN BLAST FURNACE FUEL: COKE .............................. 57

Stamp Charging of Coke Ovens .......................................................72

5. DISTRIBUTION OF TRADITIONAL BURDEN ................. 79

6. BURDEN PREPARATION: TREATMENT OF

6.6 Calcining/Roasting ..........................................................................106

7. BURDEN PREPARATION: AGGLOMERATION

8. BURDEN PREPARATION: AGGLOMERATION

9. CHARACTERIZATION OF BF CHARGE MATERIALS ... 175

Low-Temperature Breakdown Test (L.T.B.T.) ..............................181

10. PHYSICAL-THERMAL-CHEMICAL PROCESSES IN A

10.6 Kinetics of Reactions in Bosh and Hearth .....................................221

11. CONSTRUCTION OF A BLAST FURNACE AND

11.11 Slag-granulation ..............................................................................256

12. BLAST FURNACE PRODUCTS: BLAST FURNACE

13. BLAST FURNACE OPERATION ....................................... 275

14. IRREGULARITIES IN BLAST FURNACE

14.3 Slip ...................................................................................................284

15. BLAST FURNACE PRODUCTS-II SLAG AND METAL ... 289

16. IMPACT OF BURDEN PREPARATION AND

16.5 Blast Furnace Performance............................................................310

17. MODERN BLAST FURNACE DESIGN

17.15 Utilization of Plant Iron-Bearing Wastes .......................................362

18. BLAST FURNACE OPERATIONAL CONTROL ............... 367

19. ALTERNATIVES TO BF I – LOW-SHAFT AND

20. ALTERNATIVE ROUTES OF IRON PRODUCTION II –

20.7 Special Features..............................................................................400

21. ALTERNATIVE ROUTES OF IRON PRODUCTION III –

22. ALTERNATIVE TO BLAST FURNACE IV – SMELTING

Kawasaki Process (SR) ....................................................................441

APPENDIX A ............................................................................. 459

INDEX ....................................................................................... 487

1.1 Ancient Ironmaking

The modern blast furnace (B.F.) is a very expensive operation which

1.2 Historical Ironmaking Processes

FIGURE 1.1 Stukofen, the progenitor of the modern blast furnace.

Blooming hearth 60–70 450 20 0.03–0.15

Catalan Forge 140–160 360 31 0.5–0.7

Stukofen 300–900 250 40 1.0–3.0

1.3 Developments that Led to Modern Ironmaking

as a fuel although it suffered from some disadvantages. Because of its fri-

Hot Blast Generation

Blast Furnace Gas Cleaning

Pig Casting Machine

1.4 Modern Ironmaking

operations. These changes in quality resulted in a considerable increase in

1.5 Alternative Methods of Iron Production

because of matching improvements in quality of coking coal. Attempts

1. Use of electric-arc furnace to partly replace role of coke by electrical

1.7 Twenty-ﬁrst Century Trends (U.S. and the World)

U.S. Production of Selected Mineral Commodities in the First