1 Introduction

Luossavaara-Kiirunavaara AB (LKAB) is a state-owned Swedish iron ore producer in Northern Sweden with underground iron ore mines in Kiruna and in Malmberget as well as an open-pit operation in Svappavaara. At LKAB, the iron ore deposits and their amenability with respect to mineral processing have earlier been characterized and evaluated mainly based on the grade of important elements such as iron (Fe), phosphorus (P), vanadium (V) and silicon (Si). This kind of characterization and evaluation seemed obviously to be suitable when high grade ores were treated. However, it has drawbacks when the crude ore shows more complex mineralogy and where a specific element is distributed in several different minerals.

The first stage of the more complete characterization and evaluation of an iron ore deposit with respect to mineral processing is to predict the industrial scale plant performance from selected process mineralogical test work at the laboratory scale. In the case that test works in the laboratory scale have a positive outcome, they are followed by enlarged test work and by pilot scale test work as well as full scale tests at the beneficiation plant.

2 Background

In 2006, LKAB tackled a challenging expansion strategy that included studying known iron ore deposits to investigate the possibilities of increasing the ore resources and, in an extension, to have the opportunity to mine iron ore at several sites in the future. One of the iron ore deposits under investigation is the Mertainen iron ore deposit, which is located about 14 km NW of Svappavaara village and 29 km SE of Kiruna. The latest preparatory investigations were carried out in Mertainen deposit from 2007 to 2010.

There are several challenges in the characterization and evaluation of the Mertainen deposit for mineral processing, especially related to the gangue mineralogy. A high amount of silica (SiO2) including alkali (Na and K) bound to the silicates may end up in the final concentrate without successful beneficiation process and therefore impact the quality of the final products, blast furnace and direct reduction pellets. That might cause problems in the customers’ steel production processes. Another challenge related to the mineral processing of Mertainen magnetite ore might be the loss of fine magnetite material what can occur in the separation by flotation [1].

3 Mertainen Iron Ore Deposit

3.1 Geology and Mineralogy

The Mertainen iron ore deposit (IOA) is a Kiruna type Fe oxide apatite deposit (IOA), of which the world class Kiirunavaara deposit is largest and most well-known (e.g. [3,4,5]). The Mertainen iron ore deposit has the character of a large ore breccia containing also larger lenses or veins of massive magnetite ore situated in its central part (Fig. 1). These richer and massive part of the deposit occurs within a large positive magnetic and gravimetric anomaly suggesting the magnetite breccia is very extensive. The high-grade part in the deposit is surrounded by zones of ore breccia with successively lower magnetite content. Mineralization is known to a depth of at least 500 m [6, 7]. Estimated mineral resources (measured, indicated) and inferred of the Mertainen deposit are 229 Mt [8].

Fig. 1
figure 1

A simplified geological map of the Mertainen iron ore deposit after Lundberg and Smellie [6]

Full size image

Magnetite is the main and practically the only ore mineral of the Mertainen deposit of economic value. A restricted amount of martite (hematite) occurs also very local. The most important gangue mineral in the ore deposits is dark green actinolite, which occurs as dispersed crystals or as coarse aggregates forming bands and veins through the massive ore. Magnetite is also locally accompanied by some apatite and calcite. The average iron content is 35% Fe. The phosphorus content is very low, namely, 0.05% P. However, within certain restricted parts of the deposit the average P content may be as high as 0.2 to 0.9% P [4, 7].

3.2 Ore Types

Historically, mineralization has been divided into three ore types based on their Fe content: the massive magnetite ore (>52% Fe), high-grade magnetite breccia ore (36–52% Fe) and low-grade magnetite breccia ore (20–36% Fe). The Fe content in the host rock is normally under 20% Fe. The massive magnetite ore (or rich ore) is found mainly against the foot wall of the deposit while rich and poor magnetite breccias occur irregularly mixed in the deposits [6]. In a recent study the Mertainen deposit has been divided into seven different ore types based on their geology and mineralogy, on the type of Fe mineralization and alteration, not only on the Fe grade [9].

4 Magnetic Separation

4.1 Sample Preparation

For test work with magnetic separation, 33 samples from drill cores were selected, representing different parts and depths of the Mertainen deposit, at ALS Metallurgy in Australia. Six samples represented the host rock (<20% Fe), seventeen samples represented low grade magnetite breccia (20–36% Fe), six samples represented high grade magnetite breccia ore (36–52% Fe) and four samples represented massive magnetite ore (>52% Fe) [9].

The samples were crushed, homogenized and divided into two subsamples of 150 g. The subsamples were ground to a P80 = 45 µm. After grinding, the samples were screened wet and divided into two fractions: −45 µm and −100/+45 µm. After drying and homogenization, the material was weighed to several subsamples (á 20 g) for wet low intensity magnetic separation with Davis tube magnetic separation. The rest of the material was stored as a backup sample [9].

4.2 Davis Magnetic Tube Test

The magnetic separation with Davis magnetic tube test can be regarded as one of the most important of the mineral processing tests carried out at LKAB’s mineral processing laboratory, especially with a focus on liberation of magnetite and gangue mineral such as silicates and phosphates [5]. It should be noted that there is proof that Davis magnetic tube testing provides results that cannot be directly transferred to technical magnetic separators in industrial scale. For plant design and prediction of final behaviour of intergrowths pilot tests have to be done. In case of enlarged test work for wet low intensity magnetic separation a SALA wet magnetic laboratory separator can be used as well as when there is a need for more material, e.g. for flotation tests.

For this study, 38 iron ore samples (28 magnetic products and 10 non-magnetic products) from magnetic separation tests with Davis magnetic tube were submitted for chemical assays and mineralogical analysis with QEMSCAN (Quantitative Evaluation of Minerals by Scanning electron microscopy) [10].

5 Silicate Flotation Test Work in Laboratory Scale

The mineralogy of the ore to be beneficiated has a large impact on the chemical specification on the final product, i.e. the iron ore pellet or fines [11]. As reverse cationic flotation can be applied to reduce SiO2 content in magnetite concentrate (pellet concentrate) obtained by low intensity magnetic separation, a laboratory flotation test work was performed to find a starting point for the collector dosage in the pilot plant test [1, 2].

For the feed for the flotation test work in the laboratory scale the ore samples were obtained from massive magnetite ore and magnetite breccia with higher iron content, which was derived from test mining at Mertainen deposit. The ore samples were crushed to −20 mm and separated with dry magnetic separator. The concentrate was then crushed to −10 mm and beneficiated (comminution and wet low intensity magnetic separation) in the pilot plant in Malmberget. Table 1 presents the chemical composition, the specific surface and the particle size distribution of the produced feed for the flotation test work [1].

TABLE 1 Chemical and physical composition of feed concentrate for flotation test work [1]
Full size table

The flotation tests were carried out in a Svedala D12 laboratory flotation equipment with a cell size of 3 l. The flow sheet for the flotation test is shown in Fig. 2 [1].

Fig. 2
figure 2

Flow sheet for the laboratory flotation test work carried out at the LKAB’s process mineralogical laboratory in Kiruna [1]

Full size image

Two different amines were used as collectors (Table 2) and potato starch as depressant (Dextrin). Furthermore, synthetic process water corresponding to the process water in the process plant in Svappavaara site was used during one of the test runs. The test work parameters are presented in Table 2 [1].

TABLE 2 The parameters for laboratory test work for reverse silicate flotation [1]
Full size table

The laboratory flotation tests were performed as a stepwise flotation where the collector is added in three steps and the froth product from each step is collected and analyzed together with the final concentrate. The flotation time for each step was 3 min and pH and temperature were measured continuously, but not adjusted during flotation. Chemical analysis of the samples was carried out with XRF. Furthermore, particle size and specific surface were also measured [1].

6 Silicate Flotation Test Work in Pilot Plant

A process design was built up at the Geological Survey of Finland (GTK) corresponding to LKAB’s flotation process in the processing plant in Svappavaara site. The feed material was also obtained from massive magnetite ore and magnetite breccia with higher iron content, which was derived from test mining at Mertainen deposit and beneficiated in the pilot plant in Malmberget, consisting of comminution and wet low intensity magnetic separation [1].

The test work was divided into 3 different setups. In setup 1 and 2 the dosage of collector and functionality of flotation setup were evaluated. In the last stage, setup 3 presented in Fig. 3, the focus was to investigate how well the closed circuit worked, to confirm dosage selection and to produce magnetite concentrate (pellets concentrate) for further investigations on pelletizing (not presented in this article). The flow sheet for the third set up is presented in Fig. 3. It corresponds to the flotation flow sheet in the processing plant in Svappavaara except for the regrinding mill. The third set up was added as it was desirable to evaluate the effect of a regrinding step for the recycled material to try to recreate specific surface and increase the iron (Fe) recovery. Sampling was made from every stream, once a day (when possible) in one hour on every 10th minute [1].

Fig. 3
figure 3

Flow sheet for the pilot scale flotation test work carried out at the mineral processing laboratory of GTK (Geological Survey of Finland) in Finland [1]

Full size image

7 Results and Conclusions

7.1 Mineralogical Investigations and Liberation Analysis

Based on the mineralogical analysis carried out by ALS Metallurgy [10] magnetite is clearly dominating the iron oxide in the Mertainen deposit. Only minor to trace hematite (martite) is present in selected samples, associated with magnetite. Magnetite dominates also the magnetic fraction after magnetic separation with Davis magnetic tube making up between 94 and 99% in the selected samples. The most common gangue minerals in the samples representing magnetic fraction are albite, titanite and biotite. Small amounts of K‑feldspar, plagioclase, amphibole, chlorite and epidote as well as calcite and apatite are also present. Silicates (excluding titanite) typically make up between 0.5 wt% and 2.5 wt% of the sample mass in the magnetic fraction. In contrast, the amount of magnetite varies significantly between the non-magnetic fractions.

Furthermore, it can be concluded that magnetite is well liberated in nearly all samples representing magnetic fraction, with typically >90% classified as well liberated. In only two samples representing magnetic fraction, 74% and 83% of the magnetite is classified as well liberated. Unliberated magnetite in magnetic fraction is mainly locked with albite, titanite and biotite, but also with other silicates such as chlorite, K‑feldspars, epidote and amphibole. Particle image (QEMSCAN) in Fig. 4 show examples of unliberated magnetite grains mostly with intergrowth with silicates [10].

Fig. 4
figure 4

Typical intergrowths of magnetite and gangue minerals in a magnetic fraction (brown = magnetite, purple = albite, red = titanite, dark blue = amphibole, light blue = K‑Feldspar, turquoise=calcite) [10]

Full size image

Unliberated silicates are mainly locked with magnetite in high grade particles (with magnetite occupying >60% of the particle area) in the magnetic fraction. Titanite contributes typically <2 wt% of fraction mass. Furthermore, it is important to note that unliberated titanite is mainly locked with magnetite, either in well-liberated magnetite particles or in binary magnetite-titanite particles.

The intergrowths of magnetite and silicate minerals, mostly with albite but also with titanite and amphiboles is an important texture considering the wet low intensity magnetic separation, which is a crucial part of the beneficiation process in the processing plants in the LKAB sites in Kiruna, Malmberget and Svappavaara. The intergrowths of magnetite and silicate minerals can be expected to end up in the magnetite concentrate (feed for pellets plants) and to increase the SiO2 content based on the experimental studies on iron ore from Kiirunavaara deposit [5]. Silicates may also increase alkali (Na + K) and titanium (TiO2) content in the magnetite concentrate, because the silicates are the most important source of these elements in the Mertainen iron ore deposit.

7.2 Flotation Test Work

The aim of the flotation test work in the laboratory scale on magnetite ore from Mertainen deposit was to evaluate the effect of the silicate flotation on the quality of the concentrate. The most important task is to examine the reduction of SiO2 content in magnetite concentrate (SPC) after reverse silicate flotation. Furthermore, it can be noted that SiO2 content in SPC (Svappavaara pellet concentrate) should not exceed 0.5% in order to keep the alkali content (Na + K) and titanium (TiO2) content within the current product specification of LKAB [12].

A laboratory flotation test work carried out at the LKAB’s process mineralogical laboratory in Kiruna (Chap. 4) was performed to find a starting point for the collector dosage in the pilot plant test. The dosage of the collector (monoamine and diamines, see Table 2) as a function of iron recovery and SiO2 grade is presented in Fig. 5. As results of the test work it can be noted that monoamine requires a dosage of about 70 g/t to reach 0.5% SiO2 with an iron recovery of approximately 87%. The dosage required for the diamine is over 80 g/t with an iron recovery of about 90% in the laboratory scale [1].

Fig. 5
figure 5

The dosage of the collector (monoamine and diamine, Table 2) as a function of iron recovery and SiO2 grade [1]

Full size image

The dosage level of the collector for pilot scale test, adjusted to a required level with the monoamine, was about 50 g/t and with diamine approximately 100 g/t to reach SiO2 content of 0.5% The results from the flotation test work on iron ore from Mertainen deposit in pilot scale are shown in Table 3.

TABLE 3 The results from the flotation test work on iron ore from Mertainen deposit in pilot scale for set up 3 presented in Fig. 3 [1]
Full size table

The most important result from the flotation test work on iron ore from Mertainen deposit in pilot scale is that cationic silicate flotation can be regarded as a possible solution for reducing the silica (SiO2) content in the magnetite concentrate (feed for pellet production). All samples showed SiO2 content near the reference value of 0.50% (monoamine as collector) or clearly below the reference value (diamines as collector). A similar result was obtained in reverse cationic silicate flotation test work in laboratory scale regarding SiO2 content for SiO2-rich ore type B2 from the Kiirunavaara mine [2]. Furthermore, it can be noted that recovery of iron (Fe) is high in all the test runs and both sieve analysis and specific surface analyses shows a loss of fine material during flotation. The performance of monoamine and diamine seems to differ according to particle size. In order to further investigate the potential loss of specific surface, test runs with over dosage of amine were made. High dosage showed an increased loss of specific surface, lower iron (Fe) recovery and a lower SiO2 grade [1].

7.3 Conclusions

Although the geology and mineralogy of the Mertainen iron ore deposit differ from the other iron ore deposits operated by LKAB, i.e. Kiirunvaara and Malmberget underground mines and Leveäniemi open pit, it seems to be possible with cationic silicate flotation to reduce the silica (SiO2) content in the magnetite concentrate produced from this deposit.

The test work in the pilot scale showed that the required SiO2 content of 0.50% was achieved with an iron (Fe) recovery of over 95%, with both reagents (monoamine and diamines). For monoamine the froth was very stable, thereby accumulating in large volumes. This might cause some problems in full scale operation, because it may clog the froth drains. With diamine the froth was easier to handle and the selectivity was better. There was no notable difference between the use of synthetic process water and the original process water in content and recoveries. It can therefore be concluded that the origin of process water does not seem to have a large effect on the flotation conditions. It seems that diamines are more effective on coarser particles and monoamines on fine fractions.

If the excessive froth formation can be handled, it could be of interest to mix diamine and monoamines in order to enhance the selectivity in all size ranges. However, this needs further investigations and more test work both in the laboratory scale and pilot scale as well as full scale test at the processing plant.