Agglomeration of Iron Ores D. F. BALL University of Salford J. DARTNELL British Steel Corporation J. DAVISON BISRA, The Corporate Laboratories of the British Stee! Corporation A. GRIEVE (assisted by A. F. Bradley and A. F. Ogg) ‘Simon Carves Limited R. WILD Consultant to BISRA, The Corporate Laboratories of the British Steel Corporation FOREWORD BY E. W. VOICE, &ritish Stee/ Corporation 74 228 LC, Frau Boke Gentasoi Heinemann Educational Books Limited 3] LondonHeinemann Educational Books LONDON EDINBURGH MELBOURNE AUCKLAND TORONTO HONG KONG SINGAPORE KUALA LUMPUR IBADAN NAIROBI] JOHANNESBURG NEW DELHI ISBN 0.435 720104 © D.F. Ball, J. Dartnell, J. Davison, A. Grieve, and R. Wild 1973 First published 1973 Published by Heinemann Educational Books Limited 48 Charles Street, London W1X 8AH Text set in 11/13 pt. IBM Press Roman, printed by photolithography, and bound in Great Britain at The Pitman Press, Bath.Foreword ‘One hundred years ago the average make of a blast furnace in the United Kingdom was 8000 tons per year. The Japanese have now made this quantity in a day and others are actively making plans to do so. Admittedly furnaces are now larger and there have been many improvements but it is fair to state that these outputs would not be possible without the economic and technical advantages of using agglomerates. Sintering started in America around sixty years ago in order to process arising fines, flue dust,and mill scale into a lumpy form for charging to the blast furnace. This ability to use fines opened the door to the use of crushers to size ore to a more appropriate size range than the run of mine ore. By the thirties and forties sintering was widely established for this purpose but the quality was such that while it could be used there was little enthusiasm to use more than was necessary. ‘After the war a few forward-looking blast furnace operators encouraged scientists and plant makers to study the process of sintering. Within ten years progress was considerable and fluxed sinter became generally accepted as a desirable blast furnace feed. Furnaces in some countries, notably Sweden, the U.K., and Russia, were reporting excellent results when using almost 100 per cent sinter. Simultaneously, others were enriching low-grade ores by fine grinding and mineral-dressing techniques. In most cases the upgraded concentrate was too fine for easy sintering and this encouraged the Swedes and Americans to find a more appropriate method of agglomerating, Pelletizing was found to be the right answer. The 1950s were a period of tremendous progress in the laboratory, in pilot plants, and in application to iron works throughout the world. Operators enthused and iron makes passed 2000 tons per day. Russia led with outputs of 2300 tons per day and confidently planned for five furnace ironworks to make five million tons per yea ‘The industry was well served by technical papers, national and intemational conferences etc. on agglomeration and ironmaking, but this period of intense development was in- appropriate for the production of an authoritative text. During the last ten years, progress has continued unabated and has been rewarded by some furnaces doubling their iron make in this period, and the Japanese reporting the phenomenal average at Fukuyama of 5940 tons of iron per day during the months of October and November 1968 from their No, 2 furnace (67 per cent fluxed sinter and the rest screened ore). Burden preparation and agglomeration have shown the advantages of getting the physics right and reducing the chemical load in the blast furnace. This has encouraged research into the manufacture of reduced agglomerate in order to reduce the thermal requirements to the minimum, Results obtained are exciting but local fuel costs and economics generally do not suggest that cheaper iron can always be made by this method. Instead, a good case can often be made for using highly reduced material direct in an electric furnace, and its use will certainly increase in the next decade. Agglomeration has thus transformed the blast furnace into the high productivity unit it is today, but unless the blast furnace continues to present a moving target to its competitors, it is possible that pre-reduced agglomerates will contribute to its decline. For many years operators have found it difficult to combine strength with high reducibility in sinter and, to a lesser extent, in pellets. There have been many arguments as to which is the more important and therefore which can be relaxed. Research in the laboratory and on the plant must be pursued vigorously in order to achieve both, because it is now wellvi Foreword established that both are vital if the blast furnace is to progress and survive. Fortunately encouraging results are already being achieved in the laboratory, but much greater under- standing of the fundamentals must be acquired as soon as possible in order to improve industrial performance. Much, also, must be done to improve the engineering of plants, to reduce down time, and to obtain greater process control and operational consistency in order to raise quality and reduce cost. What of the future? Liquid metal has the great advantage of separating gangue in the ore from iron. Classical methods of producing liquid metal involve the blast furnace or the electric furnace and these require agglomerates. New methods are being researched of producing liquid metal from fines and if successful would remove the need for agglomeration. Alternatively, steel products may be produced from iron powder and this could be produced by chemical routes or by using super concentrates and a direct reduction process. Again there would be no need for agglomeration. Clearly there are challenges ahead and the best way to meet them is to improve the quality of agglomerates and to reduce their cost. 1am confident that this book will help to encourage progress and I sincerely congratulate Dr D. F. Ball and his panel of authors for the considerable effort and time they have devoted to its preparation. E. W. Voice British Steel CorporationPreface ‘The agglomeration of iron ores not only allows fine materials to be used as a blast furnace charge but also brings about substantial increases in output and consequently lowers the cost of iron making, Realization of these benefits has resulted in the world-wide use of agglomera- tion with massive investment in both plant and research. ‘The agglomeration processes themselves involve an understanding of heat and mass transfer, fluid dynamics, and basic physical chemistry, whilst a proper appreciation of the properties of the product requires a knowledge of the techniques of mineralogy. The design, operation, and control of the very large plants used makes considerable demands on a variety of engineer- ing skills. This is, then, a multi-disciplinary subject involving both scientists and engineers. Although there has been a small number of symposia devoted exclusively to this subject, the present sources of information are exceedingly diverse. It is believed that there is now a need for a single text drawing together the various published data and opinions and augmented by the practical operational details which are not normally found in journals, ‘This book has three objectives: first, to enable scientists and engineers engaged in research on agglomeration to gain a detailed knowledge of the whole field and to place their own specific problems in true perspective; second, to help those responsible for the management and operation of agglomeration plants to utilize directly the enormous volume of data and theory that has been accumulated; and finally, to assist the designer and engineer to improve existing processes and to develop new ones on a more theoretical basis than hitherto. ‘A first text is faced with a number of choices. From these we have aimed to write a short, rather than encyclopaedic book. This necessitates selection and this has been based on the following considerations. To facilitate the use of the book by readers from widely different backgrounds, a comprehensive introduction has been included. The early part of this covers iron-ore treat- ment and smelting at a fairly elementary level before discussing the historical development of the agglomeration processes. The introduction then covers the methods used for testing agglomerates and for raw materials handling, aspects of which are common to both sintering and pelletizing. The two central sections of the book deal with the theory and practice of both sintering and pelletizing. In discussing theory, the emphasis has been to try to lay a good basic foundation incorporating that which has stood the test of time. The practice of both sintering and pelletizing is similar in that both are predominantly exercises in the complex handling of materials. To avoid undue length, we have chosen to describe in detail the plant and engineering requirements associated with sintering only, where the travelling grate is virtually unique. Much the same considerations apply to a good deal of pelletizing practice, and for this reason pelletizing plants are described in less detail. The advantages to be gained by using agglomerate in the blast furnace are described in a final section. ‘The book sets out to cover the agglomeration of iron ores for use in the blast furnace. Although it is possible to agglomerate and reduce iron ore simultaneously to produce pre- reduced pellets, these are not likely to have such widespread application as oxidized pellets, and for this reason the discussion of pre-reduced agglomerate has been deliberately restricted. Although in the initial stages of preparation each of the five authors was responsible for a major section of the book, the final result represents their ideas and views as a whole, In writing this book all the authors have been fortunate in the cooperation they have received from the organizations where they are working, and this is gratefully acknowledged. However, the statements and opinions expressed are their collective responsibility and are not necessarily endorsed by these organizations. 1973 D.F.B.Acknowledgements The authors are indebted to the following for permission to reproduce material from their publications: The Institution of Chemical Engineers The Iron and Stee! Institute ‘The Institution of Mining and Metallurey United States Bureau of Mines Centre National de Recherches Metallurgiques, Belgium ‘The Metallurgical Society of AIME American Society for Testing and Materials American Chemical Society Verlag Stahleisen MBH BISRA, The Corporate Laboratories of the British Steel Corporation. ‘They are also indebted to the publishers for permission to reproduce material from Mining Magazine and to Simon Engineering Ltd., Head Wrightson and Co. Ltd., Lurgi (U.K.) Ltd., Allis Chalmers Great Britain Ltd. and L.K.A.B. for the supply of photographs for use in the book, Mr A. A. Dor of the Hanna Mining Co. and Dr P. A. Ilmoni of L.K.A.B. kindly agreed to the substantial use of the papers which they read to the 9th International Congress on Mineral Processing in Prague in 1970, and this is gratefully acknowledged. Dr C. E. Capes, Professor P. V. Danckwerts, Sir Frederick Lea and Professor P. A. Young gave permission for the use of material from certain of their publications and this also is gratefully acknowledged. ‘Thanks are due to a great number of colleagues and others for assistance and advice and in particular to: Dr B.C. Blakey Dr P. R. Dawson DrJ.T. Fitton Dr D. C. Goldring Mr A. Khilkoff-Choubersky Mr G. Punch Mr G. Robinson Mr G. C. Carter of Davy Ashmore Ltd. assisted in the early stages of preparation of this book and his knowledge and advice were both valuable and much appreciated.Contents FOREWORD PREFACE ACKNOWLEDGEMENTS LIST OF PLATES UNITS AND NOMENCLATURE SECTION | Introductory CHAPTER Iron Ore 1 2 Iron Ore Treatment 3 Iron Ore Smelting 4 Historical Review of Agglomeration 5 Modern Agglomeration Practice 6 Trends in Agglomeration 7 Evaluation of Properties of Burden Materials 8 Pre-preparation of Materials and their Storage 9 Extraction, Proportioning, and Conveying SECTION 2 Sintering CHAPTER 10 Process Variables 11 Factors Affecting the Rate of Sintering 12. Sinter Quality 13. Preparation and Fek 14 — The Sinter Machine 15. Sinter Bed Ignition and Fuel 16 Product Handling 17 Main Waste-gas System 18 Sinter Plant Electrics 19 Process Control and Automation 20 Plant Start-up, Shut-down, and Emergencies 21 Plant Maintenance and Hygiene 22 The Economics of Sintering 23. Assessment Techniques and Results fing of Raw Mix SECTION3 _ Pelletizing CHAPTER 24 Theory of Pelletizing 25 Balling Process 26 Theoretical Aspects of Feed Preparation for the Balling Process 27 Mechanism of Ball Formation 28 Green Ball Production 29 The Use of Additives in the Pelletizing Proc: page vil vill xi xii 10 31 40 47 52 14 83 95 19 141 149 159 170 178 191 204 208 222 227 236 244 253. 259 266 276 283 303x — Contents 30 31 32 33 34 35 SECTION 4 CHAPTER 36 INDEX ‘The Drying of Green Balls Heat Transfer in Pellet Firing Pre-heating and Firing Cycles ‘The Firing of Individual Materials ‘The Behaviour of Pellets During Reduction, Pellet Firing Machines Conclusion ‘The Effect of Agglomeration on Blast-furnace Operation 308 314 320 324 332 343 365 385List of Plates PLATE PLATE PLATE PLATE PLATE PLATE PLatE PLATE PLATE 8 9 Puate 10 PLATE I PLATE 12 PLate 13 PLATE 14 facing Ore bedding and reclaiming installation 180 Feeder tables discharging from a line of mix component bins 180 Zone structure of the bed developed during sintering of a lean ore mix 181 Zone structure of the bed developed during sintering of a Labrador ore mix 212 12 ft wide sintering strand erected and trial run 212 Sinter strand 213 Straight line cooler with axial fans 213 70 ft diameter deep trough annular cooler 244 Balling discs during erection of a pelletizing plant 244 Balling circuit installed at a pelletizing plant 245 Balling drum 245 Feeder depositing green balls in a rill and valley pattern in the shaft furnace of a pelletizing plant 276 First down draught drying hood and direct recuperation header at an iron pellet plant 276 Scale model of part of a grate kiln installation 277Units and Nomenclature In the United Kingdom the preset time is very much one of transition from the Imperial system (F.P.S.) to the metric system. The metric system to be adopted is the Systeme Internationale d’Unités (SI). In the U.S.A. it would appear that a change to the metric system will certainly not take place as quickly as in the U.K., if at all. For these reasons it is, necessary that the book is acceptable to readers who work in either system of units. Accordingly, in the main the Imperial system has been used throughout, in the text, in all tables and on all diagrams, whilst in the text itself these units are followed by the appropriate SI units in brackets. There are two exceptions to this. The first relates to those sections dealing with theory where quantities are almost never expressed in units other than metric, and here SI units are used alone. An example of this is the surface area of a powder which is normally expressed in units such as m?/g rather than ft?/Ib. The second exception refers to the relatively small difference between the ton (2240 Ib) and the metric ton (tonne). This difference is about one and a half per cent, so that in circumstances where an approximate mass is given this is in tons only. However, where an accurate value is quoted in tons this is followed by the SI equivalent in brackets. The standard conditions of gas volume used are a temperature of 32°F (0° C) and a pressure of | atmosphere (101.325 kN/m*). To represent such standard conditions the prefix normal is used e.g. 23 700 normal ft?/ton (661 m?/t), In the text costs are given in £ sterling units followed in brackets by the cost in American dollars. The conversion factor used is 2.50 dollars = £1. In view of the change to decimal coinage in the U.K., all costs are expressed as decimal fractions of the pound sterling.Section 1. IntroductoryL Iron Ore Nature of iron ore Iron is only found in the metallic state in certain types of meteorites but, ina combined form, it is one of the most common elements, comprising some 5 per cent of the earth’s crust. It is most usually found as an oxide, sometimes a hydrated oxide, but it also occurs as carbonate and sulphide and as a component of a wide range of complex minerals, ‘An iron-bearing mineral can only be considered to be an iron ore if the total cost of extracting iron from it is comparable with the cost of extracting iron from other ores. This will be governed by many factors, of which the iron content of the mineral, the nature of the impurities and the location of the deposit are of particular importance [I]. ‘The more economically important iron-bearing minerals are the following: Hematite Fe,03 Magnetite Fe;O« Limonite, goethite and hydrogoethite Hydrated hematites Siderite FeCO, Chamosite 3FeO. ALO; 2Si0,.3H,0 In addition, iron is recovered from a number of minerals from which other saleable products are also extracted: for example, from pyrites (FeS.) and pyrrhotite (FeS), in addition to sulphur; from ilmenite (FeTiOs), in addition to titania, and from complex ores, in addition to such metals as nickel, copper, cobalt and vanadium. Hematite Hematite is widely distributed and is the most important source of iron. When pure, it contains 70 per cent of iron, Much of the hematite mined is high-grade, with 64-68 per cent iron and only small quantities of impurity, mainly silica and alumina. The sulphur and phosphorus contents are normally very low. There are also very large deposits of low-grade hematite containing only 20-40 per cent iron with high silica contents, some of which are now being mined. Much of the silica is removed by mineral dressing and a product is obtained which contains 60-70 per vent of iron, Magnetite Deposits of high-grade magnetites occur in a number of places, the best known being in Sweden. When pure, magnetite contains 72.4 per cent iron, the high-grade ore con- tains more than 60 per cent iron with phosphorus-containing minerals, e.g. apatite, as a common impurity. Low-grade magnetite deposits are also worked in many places, and a product is obtained after mineral dressing which has an iron content in excess of 60 per cent. Limonite, goethite, hydrogoethite These are all hydrated iron oxides containing, when pure, 60-63 per cent iron. On heating, the water molecules are removed. They can occur as primary minerals or may4 Agglomeration of Iron Ores be produced relatively near the surface as a result of weathering of the exposed ore. In low-grade silicate deposits, weathered outcrops of hydrated ore occur in which the iron content is greatly enriched. Such outcrops provided ore for the first steel industries. Siderite This mineral constitutes only a small proportion of the total world iron-ore reserves. When pure, it contains 48 per cent iron, but it is easily decomposed by heat (calcined) to hematite with 70 per cent iron. Siderite is still a commercially important source of ore in some countries, e.g. Canada. Chamosite This mineral occurs, together with limonite and siderite, in the relatively low-grade “Minette’-type deposits, such as those of France (Lorraine) and the United Kingdom. In these countries it is an important source of iron. These ores usually contain some sulphur and phosphorus and such minerals as quartz and calcite. Impurities All iron ores contain impurities, which are collectively known as gangue. The complex mineral chamosite can be regarded as a mixture of iron oxide (FeO) and gangue (SiO, and Al,0s). These impurities may be divided into categories. Slag forming These are, in the main, four oxides namely silica (SiOz) and alumina (A103) which are acidic and lime (CaO) and magnesia (MgO) which are basic. Of these oxides, only SiO, can be reduced during ironmaking and usually only to a very limited extent. In conventional ironmaking by the blast furnace process, a liquid slag is formed in which the ratio by weight of bases (CaO + MgO) to acids (SiO, + Al,Os), called the basicity, is normally about one. Most ores have an excess of SiO, and Al,O; and the ash of the coke used for fuel is mainly composed of these oxides, so a basic flux, e.g, limestone, must be added. In certain regions, in particular where Minette-type ores are used, it is possible to mix ores with a predominantly acid gangue with those whose gangue is basic to obtain a ‘self-fluxing’ ore mix, and so to avoid the use of limestone. Metallic oxides which are largely reduced to metal during ironmaking Manganese oxides are the most common, other oxides are those of chromium and nickel. Some metal oxides, e.g. nickel, are more easily reduced than iron oxide, while others, such as manganese and chromium, are less easily reduced so that only a pro- portion of them goes to the metal, the remainder going, as oxide, to the slag. A ‘manganese content of around | per cent in the metal is advantageous but small proportions of other metals, e.g. chromium, are undesirable. The large deposits of lateritic ores, which consist predominantly of hydrated oxides, contain appreciable quantities of chromium, nickel, and cobalt. In some areas, e.g. West Africa, they have been extracted and sold for their iron content, whilst in others they are worked for their non-ferrous metal content, e.g. nickel in New Caledonia.tron Ore 5 Deleterious impurities Both sulphur and phosphorus impart undesirable properties to steel and must be kept below certain maximum levels. Sulphur can be reduced to the desired level during iron making but an increased sulphur intake increases the cost of ironmaking. Normally ‘most of the sulphur enters the ironmaking process from the coke used as fuel, but some ores have a relatively high sulphur content. These can be given some pre- treatment to reduce the sulphur level; when iron is extracted from pyrite or pyrrhotite the sulphur is first removed by roasting, the ore being converted to an oxide. Phosphorus cannot be removed during ironmaking but this can, if necessary, be done when the iron is refined to steel. However, the most economic modern steelmaking process, basic oxygen steelmaking, can only be used if the phosphorus content of the iron is low. This means that low phosphorus ores, with a phosphorus/iron weight ratio of 0.002 or less, are preferred. Geographical distribution Iron-bearing minerals are very widely distributed over the earth’s surface. Figure 1.1 indicates the main known deposits of iron ore, with an indication of their size and the iron content of the ore. The largest deposits at present being worked are in the U.SS.R., U.S.A., Canada, Brazil, Australia, and West Africa (Liberia ~ Guinea). Many of the deposits of high-grade ore in industrially developed areas have been seriously depleted or even exhausted. This has meant that countries which were once self-supporting are now forced to import large tonnages of ore. An example of this is the U.S.A. where, until the end of the Second World War, the bulk of the ore utilized came from the high-grade deposits of the Lake Superior region. By this time they were seriously depleted, and increasingly the U.S.A. ore requirements are being met by imported high-grade ore together with concentrates from domestic low-grade deposits. The deposits of low-grade complex ores, largely of the Minette type, which pro- vided the raw materials for the steel industries of many industrial countries, are becoming of decreasing importance. These ores cannot be concentrated economically to yield a high-grade product. The development of new, rich ore deposits and the use of larger ore carriers, with a consequent lowering of freight costs, are making it more economic to import rich ores.than to use this type of low-grade domestic ore. For example, in West Germany the production of iron ore fell from 14.2 million tons in 1960 to 7.95 million tons in 1966 [2] On the other hand, the mining of low-grade ores which can readily be concentrated has increased markedly during recent years. In the U.S.A., for example, the tonnage of low-grade ore for concentration followed by pelletizing increased from 3.1 million tons in 1954 to 66.0 million tons in 1963 [3] The total world production of iron ore was 467 million tons in 1960, 514 million tons in 1963, and 598 million tons in 1966. It has been estimated that production will rise to about 700 million tons by 1970, 820 million tons by 1975, and 930 million tons by 1980. The iron ore production of the major producing countries is shown in Table 1.1 for the years 1960, 1963, and 1965 [4]. This table shows the production of ‘usable’ iron ore, i.e. iron ore dispatched to the consumers after, in some cases, the removal of gangue by concentration. As an example, in 1960 the United States mined about 157 million tons of crude ore to produce the 89.04 million tons of usable ore shown in the table [5]Agglomeration of Iron Ores 6 syisodap 10 ust 1ueodutt a10w ayp Jo UoNeIO7 "| a4nBLy Ty EST ey A weoutes seas A et e 00} wont (0001 PU OOF aamaeg S20s2804 0)Iron Ore Table 1.1 Iron ore production 1960-1965 Country Production {millions of tons) 1960 19631965 Europe Austria 3.54 3.93 3.54 Czechoslovakia 312 341 2.57 France 66.91 57.89 59.24 West Germany 14.18 9.51 7.95 Luxembourg, 6.98 6.99 6.32 Poland 241 2.61 2.86 Spain 5.49 5.28 5.19 Sweden 169 23.64 29.49 United Kingdom 1736-1515 15.66 USSR. 105.60 137.48 153.00 Asia China 55* s0* sot India 16.54 20.50 23.39 Japan 1.85 2.43 2.52 North Korea 3. 3.86 5.00 Malaysia 5.3 7.38 6.98 Africa Algeria 3.44 1.98 314 Liberia 3.05 656 15.07 Mauretania = 1.72 6.04 South Africa 3.07 4.46 5.82 North America Canada 19.55 27.34 34.21 United States 39.04 73.41 89.18 South America Brazil 935 11.22 15.98 Chile 6.04 851 12.13, Peru 5.13 6.45 6.25 Venezuela 2010 1186 17.65 Australasia Australia 4.40 5.60 681 “approximate of iron ores ‘There are differences of opinion concerning the precise mechanisms whereby iron- bearing materials became concentrated in the present deposits. However, there is general agreement on the broad principles [6]. Two basic processes are involved: 1. Chemical and physical processes In the process of crystallization from a silicate solution, the heavier iron-bearing minerals tend to segregate due to gravity or to more involved geological processes. Thus the final rock mass contains areas which are richer in iron than others. It is presumed that the Kiruna deposits of northern Sweden were formed in this way. When surface water descends or when under- ground water ascends through rock strata, iron-bearing minerals may be deposited or gangue minerals may be leached out, in particular silica. In either case, iron enrichment takes place. Iron-bearing minerals can also be concentrated by precipi- tation from lakes, rivers, and seas due to chemical or bacteriological action. The iron minerals may be accompanied by other mineral depositions and then iron-rich sand- stones, limestones, and other rocks are formed.8 Agglomeration of Iron Ores Mechanical action Weathering may break down the rocks into smalt particles, some of which are rich in iron. These particles are carried away by the water of rivers or seas and later redeposited, The higher specific gravity of the iron-bearing minerais causes them to be deposited sooner than other lighter material. This method of formation was responsible for the laying down of certain sedimentary hematite deposits and the magnetite beach sands. The iron ore deposits are associated with most of the main geological ages, as is, indicated by Table 1,2 (7]. It is clear, however, that the more important deposits are associated with rocks of the oldest period, the pre-Cambrian. Table 1.2 Geological period Deposit Location Tertiary Pliocene Kerch oolitic-Himonite Crimea, Russia Mesozoic Cretaceous Salagitter limonite and Germany hematite Jurassic Minette ores Western Europe Paleozoic Carboniferous Black band ironstones United Kingdom Devonian Siegerland siderite Germany Siturian Clinton hematite Alabama, U.S.A. Ordovician Wabana ores Newfoundland Pre-Cambrian Brazilian hematites Brazil Krivoi Rog Russia Bihar and Orissa India Lake Superior ores USA Cerro Bolivar hematites Venezuela Kiruna magnetites Sweden Classification of iron ores Iron ores ma conveniently be classified according to their mode of origin. This enables ores to be listed according to typel8] Lake Superior type This is the largest and most important group of ores which all resemble the ores of the Lake Superior region (U.S.A.)}. The original, sedimentary mother-rock typically con- sists of alternating bands of iron oxide and silica and is known by different names in different parts of the world — banded hematite quartzite (B.H.Q.) in India, calico rock in South Africa, itabirite in Brazil, and Taconite or Jaspilite in the U.S.A. It normally contains some 30 to 40 per cent iron. In certain areas the silica has been leached out and sometimes replaced by the deposition of secondary iron oxides. In these cases the iron ore is massive and often very pure. Where no secondary deposition has taken place the ore is porous and may be very friable. In certain regions, e.g. India and Brazil, avery pure, powdery iron oxide (iron content greater than 68 per cent) has been left known as ‘blue dust’.Iron Ore 9 Minette type ‘These ores were precipitated from shallow waters and the iron salts crystallized around granules, often of foreign matter, to produce the egg-shaped ‘ooliths’ of which they are composed. Not only are the Minette ores of Western Europe of this type but also the sedimentary ores of the United Kingdom, the Alabama (Clinton) ores, and the Wabana ore of Newfoundland, amongst others. ‘There are certain other sedimentary ores which do not fit into these two types; the most important, historically, is the British black-band ironstone which in past years was commercially important. Laterite type ‘These were formed by the decomposition of basic igneous rocks under tropical condi- tions. With alternating wet and dry seasons silica and alkalis were leached out, leaving mainly iron oxide and alumina. Any chromium, nickel and cobalt in the parent rock also remains. Kiruna type This type of ore is derived from igneous material of volcanic origin and consists mainly of magnetite. Magnitnaya type This, like the Kiruna type, was derived from igneous material, but gases and hydro- thermal solutions may have played an important role in modifying the nature of the ore deposit. There ores are mainly magnetites. Bilbao type ‘These ores are replacements of limestone. Circulating water has removed limestone, usually depositing iron carbonate in its place. Subsequent surface weathering has oxidized part of the iron carbonate to oxide or hydrated oxide. In addition to these ore types, there is a group of titaniferous magnetite ores which are not likely to be of commercial importance for ironmaking in the near future. References 1. Economic Aspects of Iron Ore Preparation, United Nations, 1966, p. 7. 2. The World Market for Iron Ores, United Nations, 1968, p. 20. 3. Reference, 1, p. 256. 4, Reference 2, pp. 20-21 5. Reference 1, p. 253. 6. The Making, Shaping and Treating of Steel, U.S. Steel Corporation, 1957, pp. 141-3. 7. Reference 6, p. 142. 8, FG. PERCIVAL: unpublished report.2. Iron Ore Treatment Iron ore mining Prospecting Indications that an area contains iron bearing minerals may come from many different sources, as, for example, the geological formations found during general exploration. Subsequent surveys, including aerial photography and detailed geophysical and geological surveys, enable more definite conclusions to be drawn of the possibilities so that a decision may be made as to whether a particular area is worth further study. When such a study has been decided upon, a diamond-drilling and tunnelling programme is carried out to determine the extent and quality of the ore deposit. From the results of the underground exploration programme, and a consideration of all the problems involved in mining the ore, processing it and transporting it to the consumers, the economic attractiveness of the deposit can be evaluated. In the late 1940s there was a general fear that a serious iron ore shortage might develop. This caused a large increase in prospecting which resulted in the discovery of large new ore fields — in Canada, Venezuela, Mauretania, and Liberia in particular. Ten years later, very large deposits were found in north-west Australia, whilst the reserves of the U.S.S.R. were increased from 5000 million tons in 1945 to over 100.000 million tons in 1965. World reserves were estimated at 85 000 million tons in 1955. By 1965 this estimate had been expanded to 205 000 million tons, with a further 250 000 million tons ‘probable’. Thus the known deposits contain sufficient ore to meet the steel industries’ needs for many decades to come. Most of the newly found deposits are in remote and under-developed areas. Thus it has been necessary to carry out a great deal of development work in addition to that specifically pertaining to the mining operation, such as the construction of railways to the coast, the construction of new port facilities, or the enlargement of existing ones, and, frequently, the building of mine and port townships. General considerations Normally, mining commences as an open-cast operation which may eventually change to underground mining if the ore body dips steeply. Most of the newer mines, for both high- and low-grade ore, are open-cast and are likely to remain so as very large reserves are available close to the surface. Large tonnages of ore are, however, mined under- ground, for example in Sweden and in Russia where, at Krivoi Rog, some 90 per cent of the ore, which totalled 48 million tons in 1963, came from underground workings. Underground mining is normally more expensive than open-cast working, as the development costs are higher and the labour productivity is lower. However, if the scale of the operation is large and other circumstances are favourable, ore mined underground can be competitive with ore mined by open-cast working. Iron-ore mining is carried out on a very large scale. A modern mine such as those in ‘Sweden and Brazil can produce ten million tons or more of ore per year. Extensive mechanization is employed and the labour force kept to a minimum. Open-cast mining Open-cast mining involves first the removal of the ‘overburden’, rock which has a low iron content, and then the establishment of a network of roadways which will enableIron Ore Treatment the ore to be transported from the faces. The process of mining is relatively simple; vertical holes are drilled behind the face, which is normally 30-50 ft (9-15 m) high, and the ore is then shattered by explosive charges set in the holes. The broken ore is loaded into dumper trucks which can carry up to 100 tons — or, in some older mines, railway cars — by power shovels of up to 12 cubic yards (9 m*) capacity and transported to the crushing and screening station. Many mines are on hillsides and it is possible to arrange for the ore to move continually downhill, Productivity in large modern mines is counted in hundreds of tons per man shift. In the United Kingdom open-cast mining often involves the removal of a thick layer of overburden — sometimes in excess of 100 ft (30 m) — followed by the excavation of the iron ore using walking drag-lines. Most of the ore is found in agricultural regions and, after the ore has been removed, the overburden is replaced and the land restored to cultivation. Underground mining Iron ore is mined underground by methods si generally on a far larger scale. In vertical ore bodies, access is achieved by sinking shafts or inclined roadways in the adjacent barren rock. These roadways are connected to the ore body by a series of cross- cuts carrying longitudinal drifts. This network serves to collect the ore and deliver it to the hoisting shaft. Between each drift, sub-levels are driven, spaced usually around 30 ft (9 m). These sub-levels carry the ‘stopes’ from which the ore is won. The precise method of extracting the ore depends on the angle of dip of the ore body, its width, the strength of the surrounding rocks, and the regularity of the ore. These mining methods are all highly efficient, but they require careful control if the ore is not to be diluted with gangue and the mixing of different types of ore is to be kept to a minimum. ‘The operations are highly mechanized; the ore is recovered by mechanical methods and dropped by chutes to the main haulage level, where it is collected into railway cars or trucks and dropped down into bins situated at the bottom haulage level. These bins usually feed a series of underground crushers and the ore is reduced in size to 4—6 in. (100-150 mm) before being hoisted up to the surface. In flat-lying, relatively narrow, deposits, access is made when possible by tunnels from hillsides, or by shafts, and the ‘pillar. and stall’ method of mining is employed. In this, the flat deposit is cut into a series of rectangles or parallelograms by longitudinal and transverse drifts. These drifts are progressively widened to form large chambers or ‘stalls’, and the rectangles are reduced in size to ‘pillars’ just sufficient to support the roof of the workings. This method is also highly mechanized, but has the disadvantage that up to 50 per cent of the ore may have to be abandoned as supporting pillars, according to ground conditions. ilar to those used for other metals, but Ore processing Ore loaded at the mine face is in the form of lumps whose maximum size can be up to 3-4 ft (1-14 m). With high-grade ores it is only necessary for the ore to be reduced in size and graded before it is used in the steel plant. At the mine the ore is crushed to a maximum size in the region of 48 in. (100-200 mm); further crushing and screening to give appropriately sized ore is carried out either at the mine, the export terminal, on receipt at the import terminal, or at the steel works. Before shipment to the steel works, the fine material may be further treated to give a sized product, i.e. pellets. Figure 2.1 shows the alternative treatment routes and indicates typical size ranges ofCOARSE CRUSHING fe é screen Crus Screen Gina 7 vetce 7 relia | z i 7 z 7 g E E 3 2 ° 8 Q| y 7 g Zi s| 3 S ‘< gE il g g s g| a 3 3 4 ¢ i é 3 é : : 8 8 ° g g 2 ‘crash and screen g A 2 at stecvorse i é + E z 7 : y 2/8 gE ale ie siost_ —Snter—=Sinter. Blast = Sinter_ Blast. fiimoce Bane pnt furnace Pane farmace Figure 2.1. Alternative treatment routes for iron ore the various products [1]. Increased treatment before shipment will inevitably result in increased ore prices. For example, the negotiated prices of Hamersley ore from north- ‘west Australia, for shipment to Japan, were quoted in 1967, as follows [2] Price £s f.0.b.* Product Size Range pate Lump ore 4-4 in. (6-100 mm) 4.027 Lump ore $-14in. (6-31 mm) 4.133 Fines Sin. (~6 mm) 3.200 Pellets 4-4in. (6-19 mm) 4.856 “free on board, ie. including all costs before the ore is in the hold of the ship at the exporting port. At the same time the cost of ore treatment at the steel works will diminish. Unfortunately degradation occurs during the handling and transport of ores and thus, for optimum blast-furnace performance, the screened ore and pellets need to be rescreened on receipt.fron Ore Treatment — 13 Consideration is now being given to extending the idea of pre-treatment before shipping to the steel works to include the removal of the major part of the oxygen from the ore. This is attractive if cheap fuel is available at the mine or the export terminal, since oxygen amounts to around 28 per cent of the weight of the iron ore. A plant is being constructed in Venezuela to produce briquettes of partially reduced iron ore (86 per cent iron) (3] for export and serious consideration is being given to this idea in north-west Australia (4]. The product shipped to the consumer should have as uniform a composition as possible. At times it may come not only from different faces of the same mine but from different mines in the same locality. In order to minimize variations in composition, the ore may be blended. With low-grade ores, the iron content must, where possible, be up-graded by gangue removal. This necessitates grinding to such a size that the iron- bearing mineral and the gangue are present in separate grains. Grinding is followed by mineral-dressing processes which separate the iron-rich from the iron-deficient grains. Certain low-grade complex ores are not amenable to concentration by mineral dressing. These are normally crushed to —§ or —} in. (~9 or ~6 mm) and then sintered. ‘This effects a degree of up-grading by removing the volatiles (HzO and CO,). For example, the ores used in the English Midlands have a mean iron content as mined of about 33 per cent, whereas the iron content of the sinter made from these ores is about 42 per cent. Crushing The 3—4 ft (1—1} m) lumps produced at the mine must be reduced to a maximum size of the order of 1—14 in. (25~37 mm) for rich ores and to a size which may be as small as 350 mesh (44 um) for certain low-grade ores, to enable a high-grade concentrate to be obtained. Since, normally, a single crusher cannot reduce the linear dimensions of a lump by more than a factor of about eight, multi-stage crushing is usually necessary. For the first stage of crushing, either jaw crushers or gyratory crushers are used, except for sticky, clayey ores. For these latter, double roll-crushers are frequently used. This first stage of crushing produces a material which is predominantly in the size range 6—10 in. (150-250 mm). 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 is to produce material in the size range 1—3 in, (27-55 mm). Figure 2.2 shows a diagram of a two-stage crusher station to produce material in this size range [5]. When smaller Jump sizes are required it is usual to reduce the lump size in stages, using closed-circuit crushing and screening to ensure that the material is allwithin the desired size range. Figure 2.3 shows the flow sheet of the No. | ore-sizing plant at the Fukuyama Ironworks, Japan, for the production of ore sized between # and 1 in. (10-25 mm) [6]. For complex low-grade ores designed for sintering, a third stage of crushing is normally required, which usually employs hammer mills. These reduce the ore to a maximum size of about 44 in. (6-10 mm). Material which has been crushed down to §—1 in. (10-25 mm) may be further crushed by the use of rod or ball mills or a combination of both. Rod mills grind the larger particles preferentially and so produce a smaller proportion of very fine particles. Frequently grinding is carried out in stages, i. rod milling first, followed by an initial concentration, after which further grinding in ball mills is carried out. It is usual to use rod and ball mills in closed circuit, that is to say that after passage through the mill the material is classified according to size, the undersize going forward and the oversize being returned.Approx SM Figure 2.2. Crusher station. LL, sour een 35mm} (primary , kas me Figure 2.3. Flow sheet of No. 1 sizing plant, Fukuyama Works,Iron Ore Treatment — 15 ‘As an alternative to this classical route, autogenous grinding may be used. Lumps of about 6 in. (150 mm) may be fed directly to the grinding mill, which may be operated wet or dry. These lumps act as grinding media, though in some instances some large balls are added to aid the grinding operation. This method also can be carried out én stages with an intermediate concentration, or it can be used as the relatively coarse srinding stage, the final fine grinding being affected by ball milling, Screening Screening is effected by passing the material over a grid or grids which will allow the fine material to pass through whilst the coarse material travels over. The nature of the grid, or screen, will vary, depending on the type of material to be screened, from a series of suitably spaced, fixed, parallel rods to welded or woven rod or wire mats. For some purposes a simple static grid is adequate, but more often it is necessary to impart a suitable motion to the screen. The most widely used screens are those which vibrate with a small amplitude at a comparatively high frequency. The movement of the screen may be gyratory, giving either a forward or a backward throw, or rectilinear. The amplitude of vibration and direction of throw affect the performance of a screen, as does the angle at which the screen is set. For effective screening it is necessary to select the most suitable screens and to ensure that a sufficient area of screen surface is avail- able. If wet, sticky ore is to be screened, particular care must be taken if blinding, i.e. clogging, of the screens is to be avoided. The screen vibration must be vigorous, and in certain instances the screens must be heated. For some ores, partial drying must be carried out before effective screening can be practised. In modern steel works, the crushing and screening facilities are very extensive in order to ensure that the final product is precisely sized. When ore has to be separated into different size-fractions at very small sizes, one of a number of methods can be used according to the conditions. Relatively coarse material can be separated by wet screening. Finer wet material can be separated using classifiers or hydrocyclones. In classifiers, the feed passes through a tank in which the coarse material falls to the bottom and is removed by mechanical rakes or a spiral, and the finer material is held in suspension and passes out with the overflow. In recent years hydrocyclones have been increasingly used. The coarse material is thrown to the walls and removed from the bottom, whilst the finer material is carried out by the overflow. Dry material can also be separated in cyclones. When dry autogenous grinding is used, the mills can be swept with heated air which, by carrying out the finer material, effects a measure of size separation Concentration After a low-grade ore has been crushed to a size small enough to liberate the iron oxide from the gangue, separation must be effected in the most appropriate manner. The following methods are used or are under consideration (7, 8]: Magnetic separation at either high or low intensity. Gravity separation using either heavy media, jigging, tabling or spirals. Flotation. Electrostatic separation, Magnetic roasting followed by a low intensity magnetic separation, In addition to these methods, some degree of concentration can be achieved by washing and by calcining (roasting). This last process, which removes the volatiles, was more16 Agglomeration of Iron Ores extensively practised in the past. Ores which were calcined before charging to the blast furnace are now usually sintered. Low-intensity magnetic separation When magnetite (Fes ) grains are to be separated from gangue, low-intensity (500-1200 oersteds) magnetic separation is normally practised. If the particles are of @ comparatively large size, greater than 4 in. (6 mm) dry magnetic separation is used. Figure 2.4. Principle of magnetic concentration Figure 2.5. Two-stage grinding in closed circuit with three-stage magnetic separation ore fea If sent Non-magnetic (angue) Magnetic material (concentrate) Crude ore Pinay wearate was | — Bat mill, ee ‘Secondary magnetic separators | ‘ory magnetic separators Fina’ conaontvate TolingsIron Ore Treatment 17 When the particles are less than 150 mesh (100 um) wet magnetic separation is used. If the size of the ore is intermediate, it is possible to use either method. Usually the ore is fed on to a rotating drum, part of whose surface is magnetized. The non-magnetic tailings fall from the drum before the concentrates, which are carried further under the influence of the magnetic field (Figure 2.4). It is usual for a series of magnetic separations to be carried out in combination with successive grinding operations. The precise sequence of operations and the number of stages will depend on the nature of the ore and the qualities required in the final product. Figure 2.5 shows a circuit in which two-stage grinding is practised in closed circuit with three-stage magnetic separation [9]. Occasionally, other concentration processes are used in conjunction with magnetic separation to improve the quality of the final product. For example, in one installation the final magnetic concentrates are recleaned by flotation to produce a high-grade concentrate with more than 67 per cent iron. High-intensity magnetic separation High-intensity (12 000-22 000 oersteds) separators can be used to separate weakly magnetic materials, such as hematite and hydrated hematite, from gangue materials. At present, this process is only suitable for use with dry ores. Very powerful electro- magnets are needed, and the production rate for each separator is relatively low, not greater than 6 tons/hour. Gravity methods 1. Heavy Media This method is normally used for separation at a relatively large particle size, e.g. }~1+in. (6-40 mm). The ore is fed into a container, such as an oscillating trough, containing the heavy media, normally a suspension of ferrosilicon or magnetite in water with a density between that of the iron oxide and the gangue. The concentrate is recovered from the underflow and the tailings from the overflow. The concentrate tailings are separated from the media using screens; the media is then thickened and recirculated. Frequently thé tailings are subjected to further treatment to recover the contained iron. It is possible to treat finer materials (0.01—0.2 in.: 0.2—S mm) in cyclones. Because of the centrifugal forces, the effective separation gravity is higher and it is usually possible to use magnetite as the separation medium. 2. Jigging In this method the feed is kept in motion by water pulsing vertically through it. The heavier grains move downwards to the bottom of the bed and are removed. Since the weight of the grain is governed by its size as well as its specific gravity, it is necessary for the feed to be sized within close limits. Many kinds of jig are in existence according to the size and the nature of the ore. The output of an individual jig is low. 3. Tabling The feed moves over an inclined, shaking table washed with a cross-stream of water. This method is now little used for iron-ore concentration. The capital and operating costs are high and the output of individual tables is small (1—2 tons/hour). 4. Spirals This method, which is best suited to material in the size range 10-150 mesh (1.5—0.1 mm), is becoming widely used in the treatment of hematite ores. The ore is washed down a spiral launder with a curved bottom. The heavier concentrate moves to the bottom of the curved track whilst the lighter tailings climb towards the outer rim. The investment costs are high because the output of each individual spiral is low and concentration can only be achieved by the use of a series of spirals. For example,‘Agglomeration of Iron Ores at the Carol concentration plant of the Iron Ore Co. of Canada, 3456 spirals are used to treat 17 million tons of ore per annum producing 7 million tons per annum of concentrate. However, the operating costs are low because no power is required, and the maintenance costs are also low. Flotation ‘This method of concentration, extensively used for other metal ores, has recently come to be used for non-magnetic iron ores, either as the sole concentration process or as a final concentration stage to ensure that a high-grade product is obtained. Figure 2.6 shows a simplified flow diagram of the concentration plant at the Benson Mine (New York State, U.S.A.) where flotation is used in addition to spirals [10] Rod mil Distributor tank Figure 2.6. Simplified flow diagram, Benson mines 1. screens 17 mim screen oth Feed tank Scavanger _ocuit >-—{_Rovaher spirals ns Feed tak Caner wile Cee ecaeaner rai Crees J 70% sols Cocoon Constr Oezwatrig com Potato ea Fier Fier Tailingstron Ore Treatment — 19 In the flotation process, air is bubbled through a suspension of fine iron ore in water to which a small quantity of flotation reagent has been added. This reagent modifies the surface of either the iron oxide or the principal gangue component, normally silica, so that these particles attach themselves to the air bubbles and are carried to the surface, where they are removed as a froth. Flotation can be divided into two categories, anionic and cationic, In the former the active ion, ie. the one which modifies the surface properties of the mineral particle, causing it to attach itself and the air bubble, is negatively charged. Petroleum. sulphonates or fatty acids, such as tall oil, are the most commonly used anionic reagents, the iron oxide being floated. Conversely, it is possible to float the silica if the basicity is kept high (11~12) and other reagents, such as starch or gum, are added to prevent the iron oxide from floating. In cationic flotation the active ion is positively charged. Normally silica is floated using fatty amine salts as flotation reagents. ‘Although only relatively small quantities of flotation reagent are needed, the cost of the reagent per ton of concentrate produced can be quite high. Much effort has been devoted to developing flotation processes for treating iron ores, particularly fine-grained hematites, but up to the present relatively few commercial plants use this process. Electrostatic separation ‘This process makes use of the differences in electrical conductivity between iron oxides and gangue materials. Its main use is for the final clearing of concentrates to produce a very pure material. When electrostatic separation is used in conjunction with magnetic separation, not only is the purity of the final concentrate improved, but the efficiency of the process is also improved as the quantity of iron rejected in the tailings is diminished. Magnetic roasting When a hematite ore is heated in a controlled, slightly reducing atmosphere, the hematite is reduced to magnetite. The roasting process can then be followed by a low- intensity magnetic separation. The ore has to be heated to 1100—1500° F (600-800°C) and thus there is a high heat requirement, 1}-2 million B.t.u. per ton of concentrate (1.62.1 GJ). This normally makes the process uneconomic. Magnetic roasting, using rotary kilns for the roasting stage, is used in Russia, and in Italy pyrites residues are roasted in a fluosolids reactor, Washing Some ores can be enriched by the simple process of washing, for example in trommel screens operated with water sprays. This removes any soft sandy or clayey material. In wet-treatment processes it is frequently advantageous to remove the finest materials. This process, de-sliming, is carried out either by wet screening or by the use of cyclones. Roasting (calcining) Carbon dioxide and combined water are driven off the ore by heating. Roasting is carried out either in heaps (clamps) or in shaft kilns. In the former a mixture of ore and coal slack is laid on a bed of lump coal and the latter is ignited. The whole pile, which may cover some 5000 square yards (4000 m?) and contain as much as 50 000 tons