Characterization and comparison of new limestone additives for LKAB's pellets according to texture and disintegration properties

(1)

2010:035

M A S T E R ' S T H E S I S

Characterization and Comparison of New Limestone Additives for LKAB´s Pellets According to Texture and Disintegration

Properties

Charlotte Fiquet

Luleå University of Technology Master Thesis, Continuation Courses Exploration and Environmental Geosciences

(2)

Abstract

The Swedish mining company LKAB is using limestone as additives for the production of its iron ore pellets. The company is considering five new proposals of limestones since the Nordkalk Company is soon no longer able to provide limestone from the Storugns quarry which is the one that is used today.

The first purpose in this study was to characterize and compare these five new limestones in respect to their texture and their disintegration during the transport from the quarry to the final destination in Malmberget, considering the Storugns limestone as a reference material. The second aim was to focus on finding any link between texture and disintegration properties of the limestones. Another intention is that the methodology which is used in this study will help the company to consider new proposals of limestones in the future.

Six limestones from the Orsa, Storugns, Stucks, Vasalemma, Verdal and Võhmuta quarries were analyzed by an optical microscopy analysis of the textures. Two types of physical tests were also included in the study: a tumbler test and a breakage test.

Six different textures were identified by a qualitative thin section analysis which shows similarities between the limestones from different origins but also a high variability of texture within a same quarry. A more quantitative optical microscopy analysis led to the assumption of both the degree of lithification and the amount of initial micro-cracks for each sample. According to the physical test results, the limestones disintegrate as follows, from the less to the more disintegrated:

Stucks (7,6%), Storugns (9,4%), Verdal (10,3%), Võhmuta (11,1%), Vasalemma (11,8%) and Orsa (17,6%).

There is no evident textural parameter which is controlling directly the disintegration of limestone. However the samples with a rather high lithification and a rather low fracturing disintegrate less than samples with a rather low lithification and a rather high fracturing. It is assumed that the combination of degree of lithification together with the amount of initial micro- cracks is somehow controlling the disintegration of limestone.

(3)

(4)

1 Introduction

Luossavaara-Kirunavaara Aktiebolag (LKAB) is an international mineral company that produces iron ore products mostly for the steel industry. The major iron ore mines are located in Kiruna and Malmberget in northern Sweden. The company is particularly well known for the manufacturing of iron ore pellets which are used to produce crude iron in the blast furnace. LKAB uses limestone as additives together with olivine and quartzite in their pelletization process. Indeed the addition of limestone is improving the mechanical and metallurgical properties of the LKAB’s iron ore pellets (LKAB, 2009).

The Nordkalk Company has been supplying LKAB for a long time with the limestone from the Storugns quarry in Gotland. Unfortunately the Nordkalk Company will soon (at the end of the year 2010) no longer be able to provide LKAB with the present limestone considered of good quality. Before being able to produce limestone from a new quarry in the same area (Bunge, 9 km east from Storugns), the Nordkalk Company have suggested LKAB different alternatives from three quarry sites: Orsa (Sweden), Vasalemma (Estonia) and Verdal (Norway). Another potential supplier is Svenska Mineral AB (SMA) suggesting limestones from two quarries: Stucks (Gotland) and Võhmuta (Estonia) (LKAB, pers. com.).

To change the quarry site will mean for LKAB possible changes in the logistic chain of the limestone but also changes of the different properties of the limestone compared to the Storugns one.

The aim of this work is to characterize five new potential sources for limestone additives in respect to texture and physical properties and to compare them to each other considering the Storugns limestone as a reference. The purpose is to focus on finding links between texture and disintegration due to transport of limestone by optical microscopy analysis and two physical tests performed at the LKAB’s Research and Development station in Malmberget. The classification of the five limestones together with the results of the physical tests will help LKAB to determine which of these limestones suits the best for its pellets according to their disintegration properties along the logistic chain. Finally, if the company has to take into account a new proposal, the methodology used all along this work should also be estimated as an example of how to proceed to characterize and determine if the new limestone is appropriate or not for LKAB’s pellets according to its disintegration properties.

(7)

2 Background

2.1 Limestone in general

Limestone belongs to the sedimentary rocks and more precisely to the carbonate rocks. Literally the term limestone is applied to rocks composed of at least 80 percent of carbonates (calcite, aragonite or dolomite). But generally the word is used less strictly and it designates the rocks in which the carbonate fraction is higher than the non-carbonate fraction. Finally a carbonate rock is called limestone only if the major minerals in the carbonate fraction, are calcium carbonates (CaCO₃) as calcite or aragonite, whereas it is called dolomite if the main mineral is magnesium carbonate (CaMg(CO₃)₂) (Pettijohn, 1975).

Although the sedimentary rocks represent only a few percent of the earth crust’s volume, whereas the igneous and the metamorphic rocks represent the major part of it, they are really common on the earth’s surface (Pettijohn, 1975). More precisely, the limestones represent 10 to 15 percent of the sedimentary rocks (Nichols, 1999).

Limestones are dated of all ages from the Archaean (2 500 millions of years) to nowadays, even though they are much more common in recent rocks than in older rocks (Pettijohn, 1975).

There are two different origins for limestones: detrital or biochemical. If the limestone is from a detrital origin, it is called an “exogenetic” limestone since it has been mechanically transported and deposited. If it is from a chemical origin, it is then considered as an “endogenetic” one since it has been precipitated and formed in situ.

However a third type called “epigenetic” exists and occurs if the limestone undergoes physical and chemical changes after deposition, i.e.

during the process “diagenesis” (cf. Figure 1).

All these three types of limestones show different textures. Indeed the exogenetic limestone shows hydrodynamic fabrics and structures typical of all rocks transported by waves and currents, whereas the endogenetic type shows very large variation in textures, finally the epigenetic rocks have their own frameworks (Pettijohn, 1975).

Figure 1: Genetic classification of limestones (Pettijohn, 1975).

(8)

2.1.1 Texture of limestone

The texture is the description of the grain-to-grain relations. It consists in the study of the size, the shape and the arrangement of the different constituents of a rock. It has to be done at a thin section scale (Pettijohn, 1975).

The carbonate rocks comprise material formed mostly at or close to their basin of deposition.

The principal constituents of limestone are the matrix, the cement, the organized grains made up of calcium carbonate known as allochemical components and the porosity (cf. Figure 2).

porosity matrix

cement

Figure 2: Identification of the different components of limestone (modified from Boulvain, 2009).

The matrix is the mud that has penetrated between the grains during the deposition. It is made of micrite. The term micrite, which is an abbreviation for microcrystalline calcite, refers to crystals of less than 5 µm in diameter. Micrite is formed in the basin of deposition and can result directly from seawater precipitation or from the decomposition of fragments of organisms made of calcium carbonate. Micrite often appears medium to dark grey or brown under the microscope. Indeed the thickness of normal thin sections (30 µm as in this study) is much bigger than the size of micritic grains so the grains are all stacked on each other (cf. Figure 3). That is the reason why it is impossible to isolate one single crystal of micrite under the microscope (Adams and MacKenzie, 1994). Sometimes during diagenesis the matrix can be recrystallized in coarse grains. This process is called “neomorphism” and is responsible for the formation of “microspar” and “pseudospar”

which measure respectively 4 µm to 10 µm and 10 µm to 50 µm according to Tucker (1981) in his classification of the matrix. Some micrite grains can also be found on the outer parts of shell fragments of allochems because of the micritization process. This process results in some holes (10 µm in diameter) made by blue-green microscopical algae in the skeleton of other organisms that are filled with micrite after the death of the algae. Sometimes the micrite can replaced entirely the original shell and the micritized area is called a “micrite envelope” (Adams et al., 1984).

(9)

The cement is the calcite that precipitated between the grains after deposition. It is made of sparite. The word sparite, which is an abbreviation for sparry calcite, refers to crystals of 5µm or more in diameter. Often the grains of sparite are much coarser and the crystals are typically tens to hundreds of microns in size (Adams and MacKenzie, 1994). Sparite appears clear under the microscope (cf. Figure 3) and the cement that it forms can be of different types (cf. Figure 4).

Figure 3: Observation of micrite grains compared to sparry calcite grains under the microscope in plane polarized light (PPL) with a normal thin section (30 µm thick) (modified from Boulvain, 2009).

coverslip sample

stage

Figure 4: Different types of spar cements. A: equant spar, B: equant microspar (could be either cement or a matrix having undergone neomorphism), C: bladed spar in isopachous rim, D: fibrous spar in sopachous rim, E: drusy mosaic spar (the size of grains is increasing over time from the margins of pore toward its center) (modified from Boulvain, 2009). Another type of cement is called “syntaxial” and refers to a cement which was developed in optical continuity on particular allochems such as fragments of echinoderms (Adams et al., 1984).

!

(10)

The allochemical (allochems) components are the organized aggregates of carbonate formed within the basin of deposition. This study includes three different types of allochems which are the bioclasts, the peloids and the intraclasts. The bioclasts are the complete or broken skeletal pieces of carbonate-secreting organisms. The peloids are aggregates of micrite without internal structure and those of fecal origin are specially called “pellets”. The intraclasts are new grains composed of original sediment that was later eroded and reworked within the same basin of deposition (Adams and MacKenzie, 1994).

The porosity represents the void space that is left between the different limestone’s constituents.

At the origin it might be filled with water, gas or oil and was then replaced by air or an impregnating medium during the manufacturing of the thin section. The porosity can be of different types (cf.

Figure 5).

Figure 5: Basic porosity types in sediments. The pores are shaded in blue (modified from Choquette and Pray, 1970).

Interparticle Fenestral Fracture Breccia

Interparticle Shelter Channel Boring

Intercrystal Growth

framework Vug Burrow

Mouldic Cavern Shrinkage

Finally, many limestones are not pure carbonates and do often contain quartz or clay (illite or kaolinite for instance). Since clay is below 4 µm in size, it is impossible to determinate the percentage of clay when it is mixed with fine carbonate from a thin section (Adams and MacKenzie, 1994).

(11)

2.1.2 Modifications of the texture

Some limestones do not show original texture, they were modified by several process (recrystallization, replacement, dissolution) through diagenesis, lithification or metamorphism.

The texture can also be modified by the formation of micro-cracks.

According to Pettijohn (1975), the higher the diagenesis or the higher metamorphism grade a limestone underwent, the more lithified and the more a solid framework predominates in this rock.

2.1.2.1 Diagenesis

After initial deposition, the sediment undergoes some chemical and physical changes during the process called “diagenesis”. Diagenesis occurs during burial at low temperature and at low pressure, with the increase of load pressure and the circulation of fluids through the sediment.

On one hand, the mineralogy and the texture of the original rock can be affected during the diagenetic process. The initial minerals are indeed susceptible to undergo dissolution, cementation, recrystallization or replacement. Dissolution means that a mineral is going into solution and can become a cement if it is precipitated later in an empty pore space. The recrystallization implies a change of crystallinity of a preexisting mineral without any chemical modification. The replacement represents both the change of crystallinity but also the chemical modification of the preexisting phase. For instance the mineral aragonite, which is metastable under normal conditions prevailing in sediments, is often dissolved or recrystallized into calcite, whereas the mineral dolomite can replace the calcite or aragonite minerals during diagenesis.

On the other hand compaction due to mechanical means or to chemical processes such as pressure-solution phenomenon can also occur during diagenesis. In the mechanical compaction, the uncemented grains will reorganize themselves in order to fit more tightly together. The two processes of pressure-solution are the grain-to-grain pressure-solution and the suture-seam solution. So when the limestone undergoes an increase of load pressure during burial, it is subject to dissolution. Indeed the material is going into solution at areas under high stress and can precipitate again in open pore spaces. Since this dissolution is selective, it means that the CaCO₃is normally dissolved and any less soluble material (such as clay or quartz) is concentrated along seams (also called “stylolite seams”).

Finally the limestone can also undergo crystal deformation that results in the bending of the calcite crystal twin planes (Adams et al., 1984).

2.1.2.2 Lithification

It is the process in which the loose sediment becomes gradually a solid rock. It causes the reduction of the porosity through compaction, cementation and recrystallization.

2.1.2.3 Metamorphism

When a limestone undergoes a regional or possibly a contact metamorphism it becomes a

“marble”. The texture of marbles is called granoblastic which means that the grains are formed

(12)

2.1.2.4 Micro-cracks

The limestone can also present micro-cracks that have been formed during its geological past or during the blasting in the quarry. If present, these micro-cracks act as path of weakness in the limestone (Friis et al, 2009).

2.1.3 Classifications of limestone

Many detailed limestone classifications have been proposed (Folk 1959 and 1962, Dunham 1962, Embry and Klovan 1971, James 1984, Insalaco 1998 etc.). The Dunham’s classification (cf. Figure 6) is one of the most useful and was chosen for this study because it is based on the depositional texture of the rock. Particularly it shows the relative proportions of grains and matrix using the concept of “support” which assumes the continuity of either the matrix or the grains. For instance, a limestone with 55 to 60% of grains is called “grain-supported” which means that the grains are interconnected, whereas a limestone is “mud-supported” when the grains float into a continuous matrix (Adams and MacKenzie, 1994).

Figure 6: Dunham (1962) classification of limestone according to depositional texture. Boundstones are sediments in which the components are organically bound during deposition to form a rigid structure. The boundstones include much of the sediment making up reefs and are normally identified at a hand-specimen level rather than microscopically.

2.2 Limestone in the industry

Limestone is really interesting because it has a high economic value and is used in many different applications. It also represents an important storage reservoirs for oil and one third to one half of the world’s oil production is coming from limestone or dolomite reservoirs (Pettijohn, 1975). It is widely used in several other industrial fields for its own properties. Indeed according to Jessica Roberts (2009), the lime (CaO) produced from limestone, is certainly one of the first minerals to have been manufactured and has surely been used in the antic world.

(13)

The demand for limestone and its consumption is really high. It has a very large panel of industrial applications such as in steel production, construction, building, chemical industry, environmental protection, fillers production (for paper, paints, coatings and plastics manufacturing), sugar refining, agriculture, pharmacy, sculpture, etc. In 1994, around 4.5 billion tonnes of limestone were consumed in the whole world with 31.5 percent used only in the cement industry (Tegethoff, 2002). In Europe the annual production of lime for 2008 was estimated by industry sources to 37.3 million tons per annum with the Nordkalk and the Svenska Mineral companies respectively in the 7^th and 8^th position among the European leading lime producers (Roberts, 2009). According to the European Lime Association (Roberts, 2009), its different sectors of consumption are in order of importance: metallurgy (45%), construction (25%), environmental (20%) and chemical and other industrial sectors (10%) (cf. Figure 7).

In LKAB, the limestone is used in the pelletization process to improve both the mechanical and the metallurgical properties of the iron ore pellets. Effectively almost all the calcium from calcite is included in the silicate slag that binds and strengthens the pellet. Moreover in the blast furnace it improves the slag formation which removes impurities from crude iron. The presence of limestone within the pellet is generating a more homogeneous distribution of limestone in the blast furnace. Finally it is advantageous for the blast furnace process if the energy required for the calcination of limestone is added already in the pelletizing process (LKAB, pers. com.).

2.3 From the quarry site to the customer

The quality and physical properties of the limestone products are depending on the type of limestone deposit, the production logistics and the tranportation.

2.3.1 Geology of the different quarry sites

The limestone which is used today by LKAB comes from the Storugns Nordkalk’s quarry in Gotland (Sweden).The five other types of limestone that are considered in this study are produced either by Svenska Mineral or Nordkalk. They are from different quarries located in Sweden (Orsa, Stucks), Estonia (Vasalemma and Võhmuta) and Norway (Verdal) that belong to approximately the same Cambrian-Silurian stratigraphic age (600-400 Ma) than the Storugns reference material (cf. Figure 8). The following geological information about the different quarry sites are from two geologists working for the Nordkalk and the Svenska Mineral companies.

Figure 7: Main consuming sectors of the European lime (European Lime Association, 2009).

(14)

2.3.1.1 The Gotland deposits of Storugns and Stucks: Silurian (430 Ma)

The bedrock of Gotland is a sequence of sedimentary rocks. The limestone pile strikes to the North-East and dips slightly at 0.15-0.3° to the South-East, so the layering is quite horizontal in outcrop scale. The limestone in Storugns and Stucks belongs more specifically to the Slite sub- group, which is the most important source of limestone on Gotland.

In Storugns the general stratigraphy is quite regular, although some lateral variations occurring over rather short distances. The sequence consists of a ca. 6 m-thick layer of stromatoporoid limestone, underlain by a ca. 10 m-thick layer of crinoid limestone. Finally reef formations, with associated fragmentary limestone, occur at the depth of ca. 17 m and are underlain by marl. Thin clay layers usually separate the different layers of limestone (cf. Figure 9).

• Types of limestone: There are four different fossiliferous limestones (stromatoporoid-, crinoid-, reef-, and fragmentary limestone) and a marly limestone observed in the area.

• Impurities: There are some clay occurrences as horizontal interlayers.

• Particularities: Thin clay layers and stylolite seams usually separate the different limestone layers and are responsible for an easy breakage of limestone along them.

• Uses: Steel industry because of its low sulfur content.

Figure 8: Location of each quarry site (modified from Nordkalk, pers. com.).

VÕHMUTA

STUCKS

(15)

In Stucks, the limestone pile is constituted of two different main types of fossiliferous limestones: a light one and a dark one. The dark limestone which is rich in sulfur and impurities is the lowest layer in the quarry and is connected to the underlying marl. The impurities consist in silica, sulfur, iron and aluminum in varying contents. Some clay occurrences and stylolite seams separate the limestone interlayers.

2.3.1.2 The Orsa deposit: Silurian (435-425 Ma)

It consists of a reef limestone, with some horizontal clay interlayers, which is surrounded by dark schists (cf. Figure 10). The limestone and the surrounding schist layers dip outward at 30-50º. Indeed the reef body has a shape similar to a mushroom, with thin edges and a thicker central part. The quarry is nowadays an open pit of 40 m depth.

Figure 9: Outcrop of the Storugns quarry. The thin, dark and clay-rich layers can be observed in the top unit of the stromatoporoid limestone. No visible layering is visible in the reef limestone (Nordkalk, pers. com.).

Present proﬁle

Dark schists Reef limestone

Open pit

Figure 10: Hypothetical profile over the Orsa quarry showing the location of the reef body and the present limits of the quarry (modified from Nordkalk, pers. com.).

(16)

• Type of limestone: It is a fossiliferous limestone.

• Impurities: Some few impurities of silica, iron and aluminum are concentrated in the clay layers and some clay filled fractures. Moreover a selective mining is done in order to separate impure limestone from pure limestone.

• Particularities: The limestone is really brittle because of the numerous micro- and macro- scale fractures, which are sometimes filled with clay. This extreme fracturing is explained by the major meteorite impact responsible for the Siljan Lake formation that occurred in this area about 400 million years ago after the limestone lithification.

2.3.1.3 The Estonian deposits of Vasalemma and Võhmuta: Ordovician-Silurian (respectively 450-440 Ma and 445-435 Ma)

The limestones from both Vasalemma and Võhmuta are really close in age.

The Vasalemma limestone belongs to the Vasalemma sequence that is an East-West oriented horizon, 25 km long and ca. 3 km wide. The average thickness of the limestone pile is 10 m which overlies the Kahula-marl.

• Types of limestone: There are two different types of fossiliferous limestones observed in the quarry, differing a lot by their texture. The so called “Vasalemma marble” is of metamorphic origin and has a phaneritic texture. The other type called “blue stone” consists of reef-like bodies between the Vasalemma marble layers and has an aphanitic texture (cf.

Figure 11).

• Impurities: It occurs some impurities as sulfur and silica in the limestones, mostly in the blue stone because of its inclusions of clay and marl. Moreover a selective mining is done in order to separate the blue stone from the marble.

• Particularities: The surface layers of limestones are often marked by stylolites seams where clayey material is concentrated.

• Uses: On one hand the Vasalemma marble has been used as a beautiful dimension stone since the 13th century. It is also useful for high grade carbonate products. On the other hand the blue stone is used in road building and as filling material in construction.

In Võhmuta it is only the upper part of the sequence which is exploited. It is composed of pentamerid - limestone layers (4-7 m thick) sometimes overlayed with a thin (2-3 m thick) upper layer of stromatoporoid-pentamerid limestone. The unexploited lower part consists of limestone with marl interlayers (10 m thick) which overlies a fine-grained limestone (10 m thick).

• Types of limestone: The area shows two similar fossiliferous limestones, which are stromatoporoid- pentamerid limestone and pentamerid- limestone.

(17)

2.3.1.4 The Verdal deposit: Cambrian (540-500 Ma)

The limestone formation of the Verdal deposit occurs between greenstones on the hanging wall side and phyllites on the footwall side. The dip of the limestone layers in the open pit is approximately 70°.

• Type of limestone: The type of limestone in Verdal is metamorphic, it is indeed a marble.

• Impurities: This limestone has a high carbon content and presents iron minerals like hematite, limonite and goethite. There are also clay occurrences.

• Particularities: Some fracture zones with quartz and chlorite are intersecting the deposit (cf. Figure 12). Fractures can also be filled with clay.

Figure 11: Outcrop of the Vasalemma quarry. The bluish colored aphanitic and clayey limestone to the right of the picture is surrounded by the light colored and layered marble (Nordkalk, pers. com.).

Figure 12: View of the south wall of the Verdal quarry. The limestone is colored in grey. The wall shows a few fractures filled with chlorite or contaminated with iron which are brown-yellow in color (Nordkalk, pers.

(18)

2.3.2 The logistic chain

LKAB is buying crushed limestone in the specific fraction 40-70 mm which has been transported from the quarry site of Storugns in Gotland to Malmberget. The limestone is crushed and sieved to the specific fraction in the quarry and then it is shipped to Malmberget by various means of transport.

By studying the logistic chain from each quarry site to the end destination in Malmberget (cf. Figure 13), it appears that the route the limestone is being transported would be quite similar to the one followed by Storugns limestone. The limestone would follow this logistic chain: by truck to the harbor closest to the quarry site, then by ship to Luleå harbor in the Baltic Sea and finally by train to Malmberget.

Normally the material would not be sieved during the transport (Nordkalk and SMA, pers. com.).

The loading is made by a front loader in the harbor near the quarry site and the unloading and loading is done by a grabber in the Luleå harbor. In Malmberget the limestone is transported by an excavator and a

dumper to the storage tent (cf. Figure 14).

During the winter time the limestone is first sieved below 10 mm before being stored in tent. The sieving is designed to remove the snow and the small amount of fines which is generated during the transport from the limestone aggregates. These few percents of fines are first stored and then a part of it is used in the process during the summer time. The limestone in fraction 40-70 mm is finally mixed together with aggregates of both quartzite and olivine in the following proportions:

72% of olivine, 18% of limestone and 10% of quartzite (cf. Figure 15) (LKAB, pers. com.).

Figure 13: The logistic chain from the quarry to the customer (modified from Nordkalk, pers. com.).

!

Norwegian Sea

North Sea

Baltic Sea

Figure 14: The three storage tents for additives at LKAB Malmberget. From the left to the right are stored the limestone, the quartzite and the olivine.

(19)

Figure 15: Flowsheet for the additives in Malmberget (LKAB, pers. com.).

(20)

2.3.3 Eventual problems during transport

During the transport the limestone undergoes a new stress at each step of loading and unloading.

This stress is responsible for the disintegration of limestones that results in the generation of fines.

On one hand the grains mostly suffer from disintegration by impact when they fall from a certain height during the loading. On the other hand, they mostly suffer from disintegration by abrasion, when the stones collide against each other during the unloading.

The fines (<10mm) represent not only a loss of material during the transport but also a problem when the material is unloaded from the railcars and from the shaft at the entrance of the concentrator plant in Malmberget. The fines agglomerate to each other and also with other coarse grains. Consequently the material is sometimes stacked in the railcars or in the concentrator shaft, in which blasting is needed in order to get the material moving. The generation of fines is mostly problematic with olivine which produces a lot of fines and is not that much a problem with the limestone from Storugns. That is why one of the purpose of this study is to find which limestone would generate the lowest amount of fines during transport or would have at least the same disintegration properties as Storugns.

2.3.4 Testing methods to link texture and disintegration during transport

The tumbler test is used by LKAB to study the resistance to disintegration by abrasion and impact during transport of its pellets. A previous work performed by Määttä and Lantz (2008) about the disintegration of limestones used an adapted protocol of the tumbler test. The same adapted protocol is applied for this study.

Another test called breakage test is also performed in this study to analyze the break surfaces of limestones due to transport. The study of mineral breakage is also really important in the mineral industry as in the communition process for instance. The breakage can be of two different types and occurs according to the easiest way to break. It is an inter-crystalline breakage when the crack follows the crystal boundaries and an intra-crystalline breakage when it follows one of the three cleavage plans of calcite (Friis et al., 2009). According to Petruk (2000), the type of breakage during grinding depends on two main features: the strength of the mineral bond and the regularity of the grain boundaries. Indeed strongly bonded minerals (due to crystallization or recrystallization) have a random breakage, whereas weakly bonded grains have a preferential breakage along grain boundaries. Moreover sinuous grain boundaries, with strong intergrowth, represent strong bonds whereas straight grain boundaries, with no intergrowth represent weak bonds. By optical microscopy it is possible for instance to distinguish weakly bonded grains from strongly ones because they present fractures and holes along the grains boundaries. Finally the study of Friis et al. (2009) shows that degrees of lithification and porosity together with the grain size can help to predict the type of breakage (intercrystalline along the crystal boundaries or intracrystalline along the calcite cleavage plans) that a limestone will undergo during the grinding process.

(21)

3 Method for the textural analysis and the physical testing

The test material which is used in this study consists of one bucket of Storugns material from the storage tent at LKAB Malmberget and seven bigbags of test materials around 500 kg each with different fractions depending on the origin of the limestones (cf. Figure 16). Two different fractions were sent for both the Orsa and the Verdal materials. The bigbags were all transported on pallets by trucks from the quarry sites to the LKAB’s Research and Development station in Malmberget. So it is assumed that the material underwent the same amount of stress and that it is really negligible compared to the stress caused along the normal logistic chain.

Figure 16: Weight and origin of the test materials (NK = Nordlkalk, SMA = Svenska Mineral AB).

Quarry Origin Company Weight (kg) Fraction

ORSA Sweden NK 1000

(2 bigbags) 15-50mm 50-70mm STORUGNS Sweden

(Gotland) NK 20

(1 bucket) 40-70mm STUCKS Sweden

(Gotland) SMA 450

(1 bigbag) 20-50mm

VASALEMMA Estonia NK 200

(1 bigbag) 40-70mm

VERDAL Norway NK 1100

(2 bigbags) 16-35mm 35-70mm

VÖHMUTA Estonia SMA 450

(1 bigbag) 60-120mm

In order to characterize the limestone which comes from each quarry site, it is important to identify the different types of limestones present in the same quarry. In order to collect a representative sample from each quarry, ten crushed stones of limestone are sampled randomly from each of the seven bigbags and from the bucket. The macroscopic comparison between the samples from the same bag allows finding out and collecting the different limestones from each quarry.

After the sampling, the amount of thin sections and polished thin sections to order is determined according to the different types which are identified among each fraction that was received from each quarry site (cf. Figure 17). The total amount which is ordered for each quarry site is: 5 for Orsa (two fractions), 4 for Storugns, 2 for Stucks, 4 for Vasalemma, 3 for Verdal (two fractions) and 2 for Võhmuta. Sixteen samples are ordered as thin sections and four as polished thin sections.

Previous chemical analyses on Vasalemma samples revealed a high amount of sulfur. Moreover sulfides are observed with a magnifying glass on Vasalemma limestones during the sampling.

That is why four polished thin sections are ordered to examine the sulfides in reflected light on the Vasalemma samples.

(22)

3.1 Textural description by optical microscopy analysis of thin sections

The optical microscopy analysis consisted first of a qualitative characterization of the texture according to the Dunham’s classification which was then completed by a more quantitative analysis such as the determination of the proportions of both the different phases and the different grain sizes and finally the estimation of the amount of micro-cracks.

3.1.1 Equipment

Two optical microscopes were used to characterize the texture:

- The Leica DMRME polarization microscope equipped with the Leica DFC 490 camera at the Metallurgical laboratory, Malmberget.

- The Leica DMLP polarization microscope equipped with the Leica DC 300 camera at the Research and Development station, Malmberget.

3.1.2 General description of the texture according to the Dunham’s classification The classification of Dunham is used in this study in order to characterize the texture of each limestone within a same quarry and also between different quarries. Moreover the name of the dominant allochem type is added to this classification in order to precise each texture. The studied parameters are: the sorting, the dominant allochem types and the proportions of grains and matrix.

By the analysis of these parameters, it is possible to characterize each limestone according to its texture in the Dunham classification. Moreover this first microscopical description allows identifying if all the different limestone’s types described by the geologists at Nordkalk (NK) and Svenska Mineral (SMA) were sampled in the initial step of this study.

3.1.3 Determination of the phases and grain sizes proportions

The article from Friis et al. (2009) led to focus this study on the grain size and the degrees of lithification (the latter being related to porosity, cementation, recrystallization and compaction) in order to find if there is a link between the texture and the disintegration properties.

Figure 17: Detailed list of thin sections and polished thin sections ordered to Vancouver Petrographics.

Origin Fraction

(mm) Number of different types macroscopically identified

Number of thin sections (TS) or polished thin sections ordered (PTS)

Sample names

50-70 3 3 TS ORSA 1a, 2a, 3

Orsa

15-50 2 2 TS Orsa 1b, 2b

Storugns 40-70 4 4 TS Storugns 1, 2, 3, 4

Stucks 20-50 2 2 TS Stucks 1, 2

Vasalemma 40-70 4 4 PTS Vasalemma 1, 2, 3, 4

35-70 1 1 TS VERDAL 1a

Verdal

16-35 2 2 TS Verdal 1b, 2

Võhmuta 60-120 1 2 TS Võhmuta 1, 1bis

(23)

In order to analyze the repartition of different grain sizes in every limestones, the proportion of each phases constituting every thin section is firstly determined following a specific scheme (cf. Figure 18). Then the proportion of different grain sizes is filled for every phase. There are two major categories of grain sizes: fine (micrite) and coarse (sparite). Every column is divided in three columns depending on how fine or coarse the grains are in each category.

Figure 18: Scheme for the phases proportions determination and the particle size distribution analysis in percentage of area.

Micrite (%)

(X≤4µm) Sparry calcite (%) (X≥4µm) Phases

proportion (%) Fine Medium Coarse Fine (“micro-,

pseudospar”) (4µm<X≤50µm)

Medium

(50µm<X≤200µm) Coarse (200µm<X)

Matrix Cement Allochems

Other grains

Real porosity

(voids, cracks) Grains proportion

(%)

100%

The crystal size of micrite is much less than the thickness of the thin sections which is used in this study (30 µm). Thus it is neither possible to see the limits of its crystals or to determine their diameter under the microscope (cf. Figure 3). Here the classification of the different sizes of micrite grains (fine, medium, coarse) is done according to how dark the micrite grains appear under the microscope in plane polarized light (PPL). This is assumed by considering that the finer the grains are, the more grains are stacked over each other, and so the darker it appears under the microscope. The top size of micritic grains is settled to 4 µm (Tucker, 1981). In some thin sections, clay (also below 4 µm in size) is mixed with fine micrite. But it is not possible to estimate the proportion of each phase so the total amount is considered as micrite.

The crystal size of coarse grains is determined by the graticule placed in the ocular piece of the microscope. Three major sizes are settled (fine, medium, coarse) depending on the definition of micro/pseudospar (4-50 µm) (Tucker 1981) and according to the grain size of the two other main grain families observed in all thin sections (50-200 µm and above). The types of porosity and cement are also examined.

(24)

3.1.4 Micro-cracks description

The major parameter which is analyzed about micro-cracks is mostly the degree of fracturing of the whole surface of the thin section. Some other parameters are also examined at the same time as the type of fracture (intercrystalline/intracrystalline or interparticular/intraparticular), the size and the nature of the infill material. The size of the micro-cracks is measured by the Leica QWin Pro software using the “interactive measurement” modul.

3.2 Method for physical testing

To complete the optical microscopy analysis, physical tests are carried out in order to find links between the textural and the disintegration properties of the different limestones.

The tumbler test was performed to estimate the resistance to disintegration by both abrasion and impact whereas the breakage test was done to study the type of breakage surfaces of the different limestones.

3.2.1 Tumbler test with a single fraction (50-60mm)

The tumbler test is used to estimate the resistance to disintegration both from abrasion and impact. This ISO-3271 (2007-10-01) standard test is normally used by LKAB to measure the disintegration rate of its iron ore pellets. But it can be adapted also to work with other types of materials like additives (quartzite, olivine and limestone). In a previous study it has already been performed on single fractions of Storugns and Vasalemma limestone (Määttä and Lantz, 2008).

3.2.1.1 Equipment

The equipment consists of a drum, two sieving devices and a weighing machine.

• A drum (cf. Figure 19) spinning at a speed of 25 rotations per minute. It is divided in two identical parts, each being composed by a 5 mm-thick steel plate with an inner radius of 1000 mm and an inner length of 500 mm. Inside the drum there are two L-shaped pads, which force the material to reach the highest altitude before it falls down to the lowest part of the drum (cf. Figure 20). Only one part of the drum was used for the test.

• Two automated sieving devices. The sieving device which is used to select the initial fraction 50-60 mm has sieves with wood frames whereas the other one is a Gilson test master model whose sieves have metallic frames.

• A weighing machine of Mettler Toledo model.

(25)

Figure 19: Drum used for the tumbler test (ISO-3271 2007:10:01).

(26)

3.2.1.2 Experiment

The same protocol is followed on the limestone test material from each quarry.

Firstly the test material is taken as a bulk fraction and sieved into a single fraction of 50-60 mm. This specific fraction was chosen in order to get enough material from every quarry site.

Moreover it is the same fraction that was used in a previous study on Storugns and Vasalemma limestones (Määttä and Lantz, 2008).

Then 30 kg of the most angular stones are put in the drum and one stone is marked with a pen in order to study its shape evolution. The drum is running for 5 minutes according to the specific test time which was settled in the study of Määttä and Lantz (2008) to represent the disintegration during the whole logistic chain. That test time was determined by the analysis of the shape evolution of the Vasalemma stone with respect to the curvature of the surface of the Storugns stone which is found in the storage tent today. The latter was considered as having undergone a complete logistic chain.

After the test material is removed from the drum and sieved during 2 minutes using 3 different screens (4 mm, 10 mm and 40 mm). Finally every fraction is weighted.

3.2.2 Breakage test

During the tumbler test the stones break, unfortunately it is not possible to analyze their break surfaces because of the large amount of fines which are generated during the test and contaminate the breakage surfaces. That is why another physical test called a breakage test was performed in order to study the types of macro-cracks. This test, although similar to a fall test, is more a qualitative test

3.2.2.1 Equipment

The equipment consists in plastic bags and a special height.

• Two garbage bags (LKAB model) per sample (cf. Figure 21). Doubling the garbage bags is better in order not to lose stone fragments through the holes generated during the fall if only one bag is used.

• A height of 5.16 meters (cf. Figure 22).

(27)

3.2.2.2 Experiment

The same protocol is followed on the limestone test material from each quarry.

Firstly 5 kg of test material are collected as a bulk fraction, taking care to select the most angular stones and put in a double plastic bag.

Then the plastic bag is thrown from a height of 5.16 meters. Only one fall is performed in order not to generate fines that could contaminate the breakage surfaces of the stones.

Finally the stones are removed from the bags to examine their breakage surfaces with the naked eye and the help of a magnifying glass.

4 Results

4.1 General description of the texture

The table in Appendix I shows six different textures of limestones. The texture of each limestone is illustrated by a photograph in Figure 23 and the correspondence between these textures and the Dunham’s classification is represented in Figure 24.

Among one quarry site, the limestones can have different textures as it is the case in the Storugns quarry for instance; whereas the limestones from different quarries can have the same texture according to the classification of Dunham. According to this general description of the texture it is demonstrated that the fraction range does not influence the nature of the limestone among one quarry. Thus the test material received at LKAB in two different fractions from the quarries of Orsa and Verdal is the same no matter the fraction range of crushed stones it is. Moreover some limestones which seemed to be different during the sampling prove to be similar under the

5,16 m

Figure 21: Breakage test installation. Figure 22: LKAB plastic bag.

50 cm

(28)

Figure 23: Texture of each limestone (photographs made in PPL).

(29)

The table in Appendix I shows also that the sampling which was done at the beginning of the study is quite well representative. Indeed only two types of limestones (marly- and stromatoporoid- pentamerid limestone of respectively the Storugns and the Võhmuta quarries) were not collected during the initial sampling.

4.2 Detailed description of the texture

The detailed description of the texture consists in three phases. First the determination of the phases proportions which also gives information about the mineralogy. Second the determination of the grain size proportions and finally the study of the micro-cracks.

4.2.1 Phases proportions

The limestones coming from the same quarry are similar regardless the fraction range. That is why the following descriptions of the texture are done only on a reduced number of thin sections or polished thin sections (ie. Orsa 1a, 2b; Storugns 1, 2, 3, 4; Stucks 1,2;Vasalemma 1, 2, 3, 4;

Verdal 1a; Võhmuta 1).

Figure 24: Correspondence between the classification of Dunham and the limestones which are studied.

(30)

The different phases proportions are indicated in Figure 25. The diagram in Figure 26 is another way to present the data.

Figure 25: Different phases proportions for each sample in percentage of area.

Note that for the Storugns 4 and Verdal 1a limestones it is not a cement.

Sample name Phases proportions (%)

Matrix Cement Allochems Other grains Porosity Total

ORSA 1a 55 28 16 0 1 100

Orsa 2b 15 6 45 34 0 100

Storugns 1 50 10 35 2 3 100

Storugns 2 20 25 50 2 3 100

Storugns 3 15 60 20 5 0 100

Storugns 4 0 100 0 0 0 100

Stucks 1 12 30 47 3 8 100

Stucks 2 15 30 45 3 7 100

Vasalemma 1 50 10 25 10 5 100

Vasalemma 2 0 44 25 16 15 100

Vasalemma 3 0 55 25 10 10 100

Vasalemma 4 0 55 25 8 12 100

VERDAL 1a 0 90 0 10 0 100

Võhmuta 1 20 12 43 10 15 100

Figure 26: Diagram showing the different phases proportions for each sample.

The natures of the other grains together with the one of the porosity and of the cement are specified in Appendix II.

(31)

4.2.1.1 Mineral composition

These samples which are studied are almost all composed by calcite. In fact the major part of the samples is composed by matrix, cement and allochems which consist here of calcite only. Thus the amount of other grains in these limestones is really low except for the Orsa 2b sample which is very rich in quartz (cf. Figure 27 a). Apart from the calcite, the other samples are composed of only a few other minerals as for instance pyrite, dolomite, clay and apatite (cf. Figure 27 b, c and d) which are included in the category “other grains” in Figures 25 and 26. The nature of these other minerals is specified for each thin section in Appendix II.

Figure 27: Mineral composition of the samples Orsa 2b and Vasalemma 3 and 4 (photographs in PPL).

a) Orsa 2b showing a lot of coarse silt-sized quartz grains (ca. 40 µm) which are angular and have a low sphericity.

b) Vasalemma 4 showing an apatite grain also called collophane (Hurlbut, C. S., Klein, C., 1998).

c) and d) Vasalemma 3 showing pyrite crystals which fill the porosity of a former organism (respectively in transmitted and reflected light).

4.2.1.2 Degrees of cementation, porosity and lithification

According to Figure 25, the samples can be classified depending on their degree of cementation as in the scale below (cf. Figure 28). The degree of cementation is divided in three ranges depending on the percentage of cement in each sample: weakly cemented (0 - 30%), moderately cemented

a

c

b

d

(32)

According to Figure 25, the samples can be classified depending on their degree of porosity as in the scale below (cf. Figure 29). The degree of porosity is divided in three ranges depending on the percentage of porosity in each sample: low/no porosity (0-5%), intermediate porosity (5-15%), high porosity (≥15%).

Figure 28: Scale which classifies the samples according to their degree of cementation.

Figure 29: Scale which classifies the samples according to their degree of porosity.

According to both the degrees of porosity and of cementation it is possible to assume the degree of lithification for every limestone and therefore to classify them (cf. Figure 30). The degree of lithification is divided in four ranges as follows: weakly/not lithified, moderately lithified, well lithified, highly lithified.

Figure 30: Scale which classifies the samples according to their degree of lithification.

(33)

4.2.2 Grain sizes proportions

The proportion of grains depending on their size is shown in Figure 31. The diagrams in Figure 32 and 33 are another way to present the data.

Figure 31: Grain sizes proportions in percentage of area for each sample.

Name

% Different grain sizes

Porosity andother grains

Total including porosity andother grains

Micrite Sparry calcite

Total Fine Medium Coarse Total

micrite Fine

4<x≤50µm Medium

50<x≤200µm Coarse

200µm<x Total sparite

Storugns 1 35 25 0 60 0 13 22 35 95 5 100

Storugns 2 5 10 0 15 12 49 19 80 95 5 100

Storugns 3 6 12 0 18 10 14 53 77 95 5 100

Storugns 4 0 0 0 0 35 65 0 100 100 0 100

VERDAL 1a 0 0 0 0 0 50 40 90 90 10 100

Võhmuta 1 0 32 0 32 8 5 30 43 75 25 100

Vasalemma 1 40 5 0 45 5 14 21 40 85 15 100

Vasalemma 2 20 0 0 20 7 3 39 49 69 31 100

Vasalemma 3 13 5 0 18 3 20 39 62 80 20 100

Vasalemma 4 8 10 0 18 3 20 39 62 80 20 100

Stucks 1 27 17 0 44 10 10 25 45 89 11 100

Stucks 2 35 10 0 45 7 10 28 45 90 10 100

ORSA 1a 15 20 10 45 10 18 26 54 99 1 100

Orsa 2b 25 20 5 50 0 16 0 16 66 34 100

Figure 32: Diagram showing the grain sizes proportions for each sample.

(34)

4.2.3 Micro-cracks description

The description of the micro-cracks is done on every representative thin section or polished thin section from each quarry site (ie. Orsa 1a, 2a, 2b; Storugns 1, 2, 3, 4; Stucks 1, 2; Vasalemma 1, 2, 3, 4; Verdal 1a, 2; Võhmuta 1). The observations are presented in Appendix III. According to their degree of fracturing the samples can be classified as in the scale below (cf. Figure 34). The degree of fracturing is divided in three ranges depending on the percentage of fracturing in each sample: low/no fracturing (0-7%), intermediate fracturing (7-15%), high fracturing (≥15%).

Figure 33: Diagram showing the detailed grain sizes proportions for each sample.

Figure 34: Scale which classifies the samples according to their degree of fracturing.

(35)

4.3 Tumbler test with a single fraction (50-60mm)

4.3.1 Disintegration rate based on material less than 10 mm

The different amount of materials weighted after the tumbler test are presented in Figure 35.

The different mass values are converted in percent and presented in Figures 36 and 37.

Figure 35: Weight of each fraction after the tumbler test.

Name -4 mm (kg) 4-10 mm (kg) 10-40 mm (kg) +40 mm (kg) Total mass (kg)

Võhmuta 3.1 0.24 1.04 25.6 29.98

Stucks 2.16 0 6.8 19.3 28.26

Verdal 2.68 0.44 2.62 24.62 30.36

Orsa 3.6 1.58 8.42 15.9 29.5

The comparison with the results of a former study is presented in Figures 38 and 39.

Name -4 mm 4-10 mm 10-40 mm +40 mm

Võhmuta 10.3 0.8 3.5 85.4

Stucks 7.6 0 24.1 68.3

Verdal 8.8 1.5 8.6 81.1

Orsa 12.2 5.4 28.5 53.9

Figure 36: Proportions of different fractions after the tumbler test in percentage.

Figure 37 : Diagram showing the proportions of different fractions after the tumbler test in percentage.

(36)

4.3.2 Shape evolution of the stones

Moreover the evolution of the shape of the stones shows that the limestones in general are much rounder than before entering the drum because of the abrasion during the tumblertest (cf.

Figure 40).

Name -10 mm 10-40 mm +40 mm

Võhmuta 11.1 3.5 85.4

Stucks 7.6 24.1 68.3

Verdal 10.3 8.6 81.1

Orsa 17.6 28.5 53.9

Storugns 9.4 2.9 87.7

Vasalemma 11.8 7.6 80.6

Figure 38: Proportions of the different fractions after the tumbler test in percentage with the results of a former study on Storugns and Vasalemma limestones.

Figure 39: Diagram showing the fractions proportions after the tumbler test.

a a’

(37)

4.4 Breakage test

The breakage surfaces are either characterized as dull/coarse or shiny/smooth. Almost all the samples show a mix of dull/coarse and shiny/smooth breakage surfaces, except the Orsa sample which shows only dull/coarse breakage surfaces (cf. Figures 41 and 42). Moreover it is also possible to do a qualitative analysis of the test material after the breakage test. In comparison with all the test materials, the Orsa sample generates a much higher amount of small stones which were reduced during the fall (cf. Figure 43).

Figure 40 : Comparison of the shape of the stones before (on the left) and after (on the right) the tumbler test.

b

c

d

b’

c’

d’

a) and a’) Võhmuta sample.

b) and b’) Orsa sample.

c) and c’) Stucks sample.

d) and d’) Verdal sample.

(38)

Origin Correspondance with the types of limestones studied

Breakage surface aspect Type of breakage hypothesis (intercrystalline/

intracrystalline) ORSA 1a, 2a, 3lp,

Orsa 1b Dull and coarse only Intercrystalline Orsa

ORSA 3 up, Orsa 2b Dull and coarse only Intercrystalline Storugns 1 Dull and coarse with very

few shiny surfaces Both Storugns 2 No breakage surface None Storugns 3 Dull and coarse with some

shiny surfaces Both

Storugns

Storugns 4 Dull and coarse with a very

few small shiny surfaces Both Stucks 1 Dull and coarse with a very

few shiny surfaces Both Stucks

Stucks 2 Dull and coarse with a very

few shiny surfaces Both Vasalemma 1 No breakage surface None Vasalemma 2 Shiny and smooth with

some dull and coarse areas Both Vasalemma 3 Shiny and smooth with

some dull and coarse areas Both Vasalemma

Vasalemma 4 Shiny and smooth with

some dull and coarse areas Both VERDAL VERDAL 1a, Verdal

1b, 2 Shiny and smooth with

some dull and coarse areas Both Võhmuta Võhmuta 1, 1 bis Dull and coarse with some

shiny surfaces Both

Figure 41: Macro-cracks description for each limestone.

Figure 42: Comparison between the breakage surfaces of Stucks (a, mix) and Orsa (b, dull/coarse) samples.

Characterization and comparison of new limestone additives for LKAB's pellets according to texture and disintegration properties

M A S T E R ' S T H E S I S