Laboratory study of an organic binder for palletization of a magnetite concentrate

Minerals & Metallurgical Processing, Aug 2010 by de Moraes, S L, Kawatra, S K

Abstract This study aimed to identify a way to reduce the use of bentonite in the pelletization of magnetite. With this goal, different combinations of binders were compared to bentonite by examining the quality of pellets obtained by balling drum agglomeration. Pellets were subjected to routine tests that included simple compression of the wet and dry 105oC pellets, dropping wet pellets to determine their ability to survive handling and submitting pellets to thermal shock to determine how they tolerate drying and preheating. The best results use sodium silicate (1.5%) as a binder and show that it is possible to pelletize iron ore without using bentonite. Key words: Pelletization, Magnetite, Iron/ iron ores, Binders, Bentonite Introduction Pelletization is a method of agglomeration used to turn fine fractions of iron ore into an adequate product (pellet) to be fed to blast furnaces and direct reduction reactors, where it will be reduced to pig iron or sponge iron. Binder must be added to the pelletization mixture to raise the liquid phase viscosity inside the capillary spaces, to maintain the cohesion of "green" pellets and to improve the compressive strength of the thermally treated pellets, by contributing to the formation of ceramic or iron ore oxide bridges or to scorify discrete points during thermal treatment. The classic binder for iron ore pelletization is bentonite, which is characterized by its high swelling power, great superficial area, cation exchange capacities around 60 to 170 meq/100g and property of thixotropy. In contact with water, the bentonite platelets separate, forming a colloidal gel. Bentonite enhances the strength of the iron ore pellets by the following mechanism: the presence of colloidal material shortens the distance between particles, raising the intensity of the Van der Waals forces . The disposition of bentonite plates in the point of contact between ore particles also enhances the pellet's compressive strength. The typical dosage of bentonite as a binder for iron ore pelletization is 0.5 to 0 .7% of the ore mass (dry basis) . A disadvantage of bentonite is that it adds the undesirable contaminants alumina and silica to the pellet.

Binders that would not leave a residue after thermal treatment would be extremely interesting. They would be advantageous for minimizing the variations in quality of the pellets, lowering the dosage to be used to around 0.05%. Materials and methods Sample ore and binders. Asample of magnetite concentrate from the Empire Mine (Palmer, MI) and received by Michigan Tech on 16/03/2009 was utilized. The average chemical composition of the sample is shown in Table 1. The work was developed using the following binders: bentonite , sodium hydroxide , sodium carbonate , sodium metasilicate , sodium tripolyphosphate and carboxymethyl cellulose. The typical chemical composition of bentonite is shown in Table 2 and the specification of the binders in Table 3. The choice of binders to be tested was based on the author's previous experience using these binders in the agglomeration of hematite concentrate (Moraes and Cassola, 2008). The sodium hydroxide and sodium silicate were used in the form of a 50% (w/w) solution in water. The moisture of bentonite was determined by drying it in an oven at a temperature of 105o C until constant weight. The moisture content of bentonite was 13% (dry basis). Homogenization and reduction of sample mass. The sample of iron ore concentrate was received in three containers containing approximately 15 kg (wet weight) each. It was dried at a temperature of 1000C for 24 hours. The dry sample was disaggregated using a 28-mesh (0.71 mm) sieve, blended and split by a rotatory splitter. 24 subsamples of approximately 1.7 kg each were produced. Figure 1 shows the process of homogenization and reduction of mass of the sample. Pelletization tests. For each test the ore , binder and moisture were added to a laboratory mixer (Fig . 2a) for two minutes . The tests were carried out using a laboratory-scale balling drum (Fig. 2b) with aliquots of 1.7 kg (dry basis) ore. The balling procedure consisted of: 1) Delump mixed binder and concentrate through an eightmesh screen; 2) Add a small amount of ore mixture to drum rotating at 25 rpm; moisten it with a water spray to produce pellet "seeds;" 3) Enlarge seeds by adding more concentrate while spraying with water, periodically removing and screening to keep uniform size. Pellets are finished when they are between 1 .6 and 12.7 mm (0.06 and 0.5 in.) in diameter.

The composition of each mixture is given in percentage by mass concentration based on dry components . Moisture content and binder dosages are shown in Table 4. Characterisation of the pellet properties. Two standard tests are used to measure the strength of pellets, whether the pellets are green pellets or fired pellets. These tests are the "drop" test and the "compression" test. The drop test requires dropping a random sample of pellets from a height of 46 cm (18 in.) onto a steel plate. Pellets are dropped repeatedly until the pellets crack. The number of drops needed to crack each pellet is recorded and averaged. Compression strength is measured by compressing or applying pressure to a random sampling of pellets until the pellets crumble. The pounds of force required to crush the pellets is recorded and averaged. These two tests are used to measure the strength of both wet and fired pellets . The drop and compressive test measurements are important because pellets, proceeding through the balling drum and subsequent conveyor belts , experience frequent drops as well as compressive forces from the weight of other pellets traveling on top of them. Thermal shock resistance is a factor which must be taken into consideration in any process for agglomerating mineral ore concentrate . Increases in a pellet's thermal shock resistance improve the pellet's ability to resist internal pressures created by the sudden evaporation of water when the pellet is heated in a kiln. If the pellet has numerous pores through which the water vapor can escape, thermal shock resistance is improved. If the surface of the pellet is smooth and continuous, without pores, the pellet has an increased tendency to shatter upon rapid heating. This causes a concurrent increase in the amount of "fines" or coarse particles in the pelletized mineral ore. A binder which increases the pores formed in a pellet improves the pellet's ability to resist thermal shock. The pellets obtained were subjected to the tests for characterization described as follows. Determination of moisture. After the pelletizing, about 300 g pellets were randomly sampled and dried in an oven at temperature of 1050C until constant weight. Drop number green pellet. From each batch of green pellets obtained, 20 pellets were randomly selected and dropped freely from a height of 46 cm (18 in.) onto a platform with a steel shank. Each pellet was dropped repeatedly until the first fractures appeared. The number of drops incurred by each pellet was recorded. The result is the average of values given by number of falls per pellet. Wet compression strength. From each batch of pellets obtained, 20 pellets were randomly subjected to automatic compression until breaking. At this time the load was recorded and the arithmetic mean of the values represents the test result in kgf orN/pellet.

Dry compression strength. This test is similar to that done with green pellets, differentiated an additional step: the 20 pellets were dried in an oven at 105o C for 24 hours . After cooling for 15 minutes, the pellets were subjected to a compression test and the result was expressed in kgf or N/pellet. Thermal shock resistance test. From the batch of green pellets, groups of 15 pellets were selected at random. Then each group was placed in a muffle furnace at temperatures of 300oC, 500oC, 700oC and 900oC for 10 minutes at each temperature. After this time, the groups were left to cool in the air and subjected to a test of compressive strength. This step also evaluates the percentage of pellets that crack and/or explode. Results and discussion Table 4 shows the results obtained in the pelletization tests with iron ore concentrate and different binders. For each value reported, the mean and standard deviation were determined for 20 pellets. The error bars shown on the graphs represent the 95% confidence intervals calculated using the t-distribuition. The variation of the moisture content in each test was from 8.5% to 9.8%. The biggest moisture value was observed in the test CMC NaOH and the smallest in the test with sodium carbonate. The wet drop results are shown in Fig. 3. The highest value observed was that of the tests bonded with bentonite, which survived 3.2 drops. This value is not the minimum industrially acceptable, but in this case will be the value of reference. The worst result was observed for pellets bonded with sodium silicate, which only lasted 1.9 drops. Pellets bonded with sodium carbonate, bentonite NaOH and CMC (sodium carbomethyl cellulose) TPP (sodium tripolyphosphate) show the same result of 2.9 drops each. Effect of bentonite. Figure 4 shows the compressive strength of pellets bonded with bentonite . Two tests were carried out with bentonite at the dose of 0.66%. They showed wet and dry compressive strength results of 19 N/pellet and 38.3 N/pellet, respectively, higher than the minimum industrially acceptable (9 N/pellet, wet and 22 N/pellet, dry). Pellets bonded with bentonite at 0.4% and NaOH at 0.02% showed less wet and dry compressive strength compared to plain bentonite at O.66%. However, the result of the wet compression (18.1 N/pellet) is still greater than the minimum industrially acceptable standard. The dry compressive strength result (20.7 N/pellet) is almost the minimum industrially acceptable that is 22 N/pellet. These results suggest that the dose of bentonite and NaOH can be adjusted to reach the target strength. The behavior of pellets bonded with bentonite (0.66%) and bentonite (0.4%) NaOH (0.02%) during thermal shock can be seen in Fig. 4. As shown in this figure there are no significant differences between the results at 30O0C and 50O0C. At 70O0C the performance of pellets with bentonite (0.4%) NaOH (0.02%) falls and at 90O0C the compressive strength improves again. These results show that is possible to reduce the amount of bentonite without reducing the quality of the pellets.

Pellets bonded with bentonite presented cracks in all temperatures during thermal shock resistance tests (Fig. 5). The cracks increase the formation of dust. There were no explosions during thermal shock resistance tests in pellets bonded with bentonite and bentonite NaOH. Effect of CMC. The use of CMC in iron ore pelletization is not new. Studies carried out by Lima (1991) and Lima and Chaves (1992) with hematite concentrate showed the results using CMC and CMC TPP as binder. The authors found that CMC alone as a binder is not effective , because the pellets do not resist heat. CMC and the additive TPP present the greatest potential for use as binders. However, it should be noted that TPP adds phosphorus to the pellets, which is an undesirable contaminant. In 1998, Cassola and Chaves found that the dispersion of limonite from the surface of particles of hematite due to the dispersant effect of TPP was the fundamental mechanism for the process. The aim of using CMC and CMC NaOH and CMC TPP in this work was to check if the same dispersant effect happens when pellets are made using a magnetite concentrate. Figure 6 shows the compressive strength of magnetite concentrate pellets bonded with CMC and the additives NaOH and TPP. Magnetite pellets made using 0 .04% CMC alone as a binder had insufficient dry compressive strength. The behavior of the pellets during the thermal shock resistance test did not provide enough compressive strength to survive the drying and firing steps. When NaOH was added together with CMC to the magnetite concentrate pellets, the dry compressive strength increased, but was still insufficient. Similar results were found with the thermal shock resistance test. Pellets bonded with CMC and TPP showed better results than pellets bonded with CMC and NaOH. The addition of 0.02% TPP improved the pellet dry compressive strengths to the minimum industrially acceptable (22 N/pellet). Similar results were found with the thermal shock resistance test. These results suggest that dispersants are beneficial in making pellets with magnetite concentrate. There were no cracks or explosions during thermal shock resistance during all tests. Effect of sodium carbonate and sodium silicate. The compressive strength results for pellets bonded with sodium carbonate and sodium silicate are shown in Fig. 7. The choice of these binders was made primarily to check the effect of the binder in the firing step. Specifically, it was determined whether it would be possible to decrease the temperature of sintering.

Pellets using 1.5% sodium carbonate as a binder showed overall better results than pellets bonded with 0.66% bentonite. However, 27% of these pellets exploded during the thermal shock resistance test at each temperature (30OoC, 500oC and 900oC) . The explosions may have occurred because the sodium carbonate should be acting as a dispersant over the fines particles contained in the magnetite concentrate. This dispersion is very intense and saturates the pores of the pellet with water and solids , preventing the passage of gases of the evaporating water during the drying step . However, the compressive strength of pellets bound with sodium carbonate is higher at 700oC and 900oC than that of pellets bonded with bentonite (0.66%). Pellets made with sodium carbonate as a binder can be seen after explosion during the thermal shock resistance test (Fig . 8) . The best results obtained were those of the pellets made with 1 .5% sodium silicate as a binder. The addition of 1.5% sodium silicate improved the compressive strength, to well above the minimum industrially acceptable values. This result demonstrates that the dose of sodium silicate can be reduced. The behavior of sodium-silicate-bonded pellets during the thermal shock experiments was excellent. There was no formation of cracks and dust and no explosion occurred. The compressive strength of the pellets made with sodium silicate as a binder was higher at 70OoC and 90OoC than that of the pellets bonded with bentonite (0.66%). The pellets bonded with sodium carbonate (0.75%) and sodium silicate (0.75%) were made in a effort to eliminate explosions during thermal shock resistance test. This test showed that by reducing the dosage and using a combination of sodium carbonate and sodium silicate, it is possible to get the same results as that of the pellets bonded with bentonite (0.66%). In this case there were no cracks or explosions during the thermal shock resistance test. Conclusion * The results of magnetite concentrate pellets bonded with bentonite (0.4%) NaOH (0.02%) compared with pellets bonded with bentonite (0.66%) showed that is possible to reduce the bentonite dosage without reducing the quality of the pellets. * Pellets bonded with bentonite presented cracks at all temperatures during thermal shock resistance tests. Cracks increase the formation of dust. There were no explosions during the thermal shock resistance tests of pellets bonded with bentonite and bentonite NaOH. * Magnetite pellets made using 0.04% CMC alone as a binder had insufficient dry compressive strength. The thermal shock resistance test gave insufficient compressive strength to survive the drying and firing steps. * When NaOH was added together with CMC, the dry compressive strength increased, butnotenough. Similar results were obtained in the thermal shock resistance test.

* The addition of 0.02% TPP improved the pellet dry compressive strength to the minimum industrially acceptable (22 N/pellet). Similar results were obtained in the thermal shock resistance test. These results suggest that dispersants are beneficial in making pellets with magnetite concentrate. * Pellets made with 1 .5 % sodium carbonate showed better results than pellets bonded with 0.66% bentonite. However, 27% pellets exploded during thermal shock resistance test at 300oC, 500oC and 900oC. * The best results were achieved with pellets made with 1.5% sodium silicate as a binder. The addition of 1 .5% sodium silicate improved the compressive strength to over the minimum industrially acceptable value . This result demonstrates that the dosage of sodium silicate can be reduced. * The compressive strength at 70O0C and 90O0C of pellets made with sodium carbonate and sodium silicate is higher than that of the pellets bonded with bentonite (0.66%). These preliminary results suggest that is possible to optimize the sintering temperature of the pellets by adjusting the amounts and types of binders. * The pellets bonded using sodium carbonate (0.75%) and sodium silicate (0.75%) showed that reducing the dose and using the combination of sodium carbonate and sodium silicate, it is possible to get the same results as that of pellets bonded with bentonite (0.66%). In this case there were no cracks or explosions during the thermal shock resistance test. References Lima, J. R. B., 1 991 , Estudo da carboxi-metil-ce/ulose como aglomerante para pelotiza o, PhD dissertation, Escola Polit cnica, Universidade Sao Paulo, Sao Paulo, 145 pp. Cassola, M. S., and Chaves, A. P., 1998, "Effect of the addition of organic binders on the behavior of iron ore pellets, " KONA: Powder and Panicle, Osaka, No. 16, pp. 136-142. Lima, J. R. B., and Chaves, A. P., 1992, "Estudo da carboxi-metil-celulose como aglomerante para pelotiza o," Sao Paulo: EPUSP, 1992. (Bo m T et cnico da Escola Polit cnica da USP. Departamento de Engenharia de Minas, BT/ PM I/0 14). Moraes, S. L., 2004, Comparacao de desempenho de aglomerante org nico em rela o bentonita na opera o de pelotiza o de concentrados de min de ferro brasileiros de rio diversas procedencias, 2004, Master's thesis, Escola Polit cnica, Universidade Sao Paulo, Sao Paulo, 80 pp. Moraes, S. L., and Cassola, M. S., 2008, "Microstructure of iron ore pellets - organic and inorganic binders," 3rd International Meeting on lronmaking and 2nd International Symposium on Iron Ore. September 22-26th, 2008, Sao Luis, MA, Brazil, pp. 464-471.

Ripke, S. J., 2002, Advances in iron ore pelletization by understanding bonding and strengthening mechanisms, PhD Dissertation, Michigan Technological University, Michigan, 186 pp. S. L. de Moraes Research engineer, Environmental and Energy Technologies Center, Institute for Technological Research, Sao Paulo, Sao Paulo, Brazil S. K. Kawatra Professor, Department of Chemical Engineering, Michigan Technological University, Houghton, Michigan Copyright Society for Mining, Metallurgy, and Exploration, Inc. Aug 2010 Provided by ProQuest Information and Learning Company. All rights Reserved de Moraes, S L "Laboratory study of an organic binder for palletization of a magnetite concentrate". Minerals & Metallurgical Processing. FindArticles.com. 28 Mar, 2011. http://findarticles.com/p/articles/mi_hb5851/is_201008/ai_n55069970/ Copyright Society for Mining, Metallurgy, and Exploration, Inc. Aug 2010 Provided by ProQuest Information and Learning Company. All rights Reserved