Improvement of Sinter Strength and Reducibility through Promotion of Magnetite Ore Oxidation by Use of Separate Granulating Method

Abstract

In general, Fe content in iron ore is gradually decreasing. This fact affects worse performance of BF operation, for example, increase of RAR and Slag ratio. Depletion of high grade iron ore deposits is moving us to use concentrates in sintering process.

Through magnetite concentration deteriorates reducibility because of high FeO content in sinter product. Such situation makes it to promote oxidation of magnetite iron ore during sintering process for improving sinter reducibility. In addition, promoting oxidation of magnetite possibly increases sinter strength with using oxidation heat.

ISIJ sinter research group for utilization of magnetite concentration suggests that restricting melt formation is critical for promoting oxidation of magnetite concentration.

In this paper, It is confirmed that “Separate Granulation” has been examined to apply their suggestion by sinter pot test.

The main results obtained are described as follows:

(1) “Separate Granulation” in case that magnetite is pre-granulated with high Al₂O₃ iron ore without limestone and coke breeze resulted in decrease of FeO in sinter and improvement of both sinter reducibility and sinter strength.

(2) Sinter micro structure featured restriction of pore, low circle factor and small mineral texture, which supported that melting restriction worked during sintering.

(3) Magnetite decreased and hematite increased as sinter mineral, which corresponded with decrease of FeO content.

(4) These facts shown (1) to (3) concluded that “Separate Granulation” is effective to improve both sinter reducibility and sinter strength due to restriction of melting during sinter reaction.

1. Introduction

Iron ore sinter is the most widely used as raw materials for blast furnace in the world. And iron ore sinter production exceeds 80 million tons per year, and CO₂ emission from sintering process accounts for about 3% of the total domestic emissions. Here, iron ore quality used as raw material is getting gradually worse. And so, iron ore concentrated preliminarily in iron ore mining works will increase in the future. Among various beneficiation methods, magnetic beneficiation is an easier method for concentrating magnetite iron ore than other methods, for example, flotation method. Since magnetic iron ore is generally magnetite (Fe₃O₄), we have focused on it. Generally, in case of high blending ratio of magnetite ore in sinter mixture, Fe²⁺ in the iron ore sinter increases, which results in lower reducibility.¹⁾ Therefore, in the sintering process, promoting magnetite oxidation to hematite is important for high reducibility. In addition to the above-mentioned effects, oxidation promotion means increase of heat generation, which can cause high sinter strength.

Therefore, promoting oxidation of magnetite ore can achieve both high reducibility and high strength, however in the past, these two factors were regarded as opposite factors each other based on operational experience in case varying a sintering heat source ratio in sinter mixture.

Based on the above background, University research members in ISIJ Study Group have studied and found out useful technical information for promoting magnetite oxidation.

1) To promote oxidation of magnetite ore, increase the oxygen potential around the ore and suppress liquid phase formation.²⁾ And so, magnetite is located remotely from flux (CaO source) and carbonaceous coagulant (coke fines etc).

2) In the air atmosphere, compared with the reagent hematite, reagent magnetite has a low liquid phase temperature reacting with CaO.³⁾ Therefore, regarding suggestion described in 1), in order to suppress the liquid phase formation, magnetite is located remotely from the flux (CaO source).

3) In the FeOx saturation region of the FeOx–CaO–SiO₂ system, addition of Al₂O₃ reduces the liquid phase ratio.⁴⁾ Therefore, the Al₂O₃ concentration rise is effective for suppressing the liquid phase formation.

Locating remotely as mentioned above is possible effectively by use of several granulating method. One is separate granulation⁵⁾ which can make up two types of granules by granulating in two lines. Another is pre-granulating method⁶⁾ which can make up granules of two-layer structure by joining granules from part of raw materials to the other raw materials before main granulation. In this study, separate granulation⁵⁾ which can be more remotely arranged compared to separating granulating method⁶⁾ was adopted, and improvement of reducibility and cold strength through oxidation promotion of magnetite ore was evaluated by a sintering pot test of 60 kg scale.

2. Basic Knowledge of the ISIJ Research Group on Promotion of Magnetite Oxidation

First of all, basic information for influence of liquid phase formation on the oxidation behavior of magnetite by the ISIJ research group mentioned in the introduction is summarized in this chapter.

In promoting oxidation of magnetite, evaluating relationship with liquid phase formation is important for the sintering process. Here, the relationship means optimization of the chronological order of magnetite oxidation and liquid phase formation. If the liquid wets the surface on magnetite ore and interferes with the contact with oxygen, it is necessary to oxidize to hematite at the temperature rising stage before the liquid phase formation. On the other hand, if oxidation is carried out by sequentially exposing the unoxidized region inside the particle by liquid phase formation, it is necessary to simultaneously carry out the liquid phase formation and the oxidation reaction. Furthermore, thirdly, if it is desired to actively react magnetite and flux to oxidize at the solid phase precipitation stage in the liquid phase stage or the cooling stage after generating the liquid phase, it is aggressively promoted before the magnetite oxidation It should be formed in liquid phase.

Fujino et al.²⁾ placed green balls (1–2 mm in diameter) in the alumina sphere packed bed and investigated the oxidation reaction behavior under air flow. As a result, as shown in Fig. 1, the oxidation rate of the green ball added with 20 mass% of CaO together with the magnetite fine powder in the air flow of 900°C was lower than that of the green ball of only magnetite fine powder.

Fig. 1.

Change in reaction ratio of green ball of magnetite concentrate with and without CaO.²⁾

Based on the observation of the structure after cooling of the CaO-added green ball, unoxidized magnetite existed inside the green ball and calcium ferrite formed at surface area. and Forming calcium ferrite suggests existence the liquid phase during high temperature sintering. Therefore, it seems that the produced liquid phase blocks air and suppresses magnetite oxidation inside the green ball. Also, when mixing the magnetite fine powder green ball and the coke particles, the oxidation rate became lower than the green ball of only the magnetite fine powder. It is considered that the formation of hematite is suppressed by the low oxygen potential due to coke combustion gas or by the solid phase reaction with coke. Furthermore, the change in weight when the mixture of magnetite fine powder and fine coke was heated in a magnetite stable atmosphere was measured. As a result, as shown in Fig. 2, sample weight increased over 400 degree C, but decreased over 700 degree C. It is thought that weight increase over 400 degree C is oxidation reaction of magnetite, weight loss over 700 degree C is oxidation (combustion) of coke fine and reduction reaction of some magnetite fine powder ore. Therefore, magnetite can be oxidized at 400°C. or higher, it is important to promote oxidation as much as possible before coke burning. By remotely arranging with powdered coke, suppression of magnetite oxidation by low oxygen potential gas and direct reaction can be avoided.

Fig. 2.

Normalized weight change of magnetite concentrate and coke breeze with temperature.²⁾

Ohgi et al.³⁾ compared magnetite reagent and hematite reagent from the viewpoint of chemical reaction with CaO.

As shown in Fig. 3, in the air flowing, the minerals formed between magnetite and CaO are 4CaO·FeO·8Fe₂O₃ and 3CaO·FeO·7Fe₂O₃, both melting at 1200°C or less. Therefore, as compared with hematite, in case of magnetite, a liquid phase formed at a lower temperature by reaction with CaO.

Fig. 3.

Mineral phase of quenched sample and melting temperature during increasing temperature under reaction with CaO component.³⁾

Hayashi⁴⁾ analyzed the composition on the liquidus in the FeOx content saturation region in the FeOx–CaO–SiO₂–Al₂O₃ phase diagram. As a result, as shown in Fig. 4, FeOx concentration in liquidas composition decreased with increasing Al₂O₃ concentration. FeOx concentration decrease is also shown as increase in the distance from the FeOx vertex to the liquidus line in the ternary phase diagram. Increase of the distance means that the liquid phase region shrinks in the FeOx content saturation region, that is, the liquid phase fraction decreases at the same temperature. Industrially, it is suggested that the high Al₂O₃ ore existing near the magnetite ore is important in the suppression of magnetite melting. Regarding the influence of the Al₂O₃ component on the liquid phase formation in the sintering process, there are many reports^7,8,9,10) of melt suppression, and the results from this study⁴⁾ have basically been consistent with these previous studies.

Fig. 4.

Influence of Al₂O₃ content on equivalent FeOx content on liquidas composition at FeOx–SiO₂–CaO–Al₂O₃ Phase diagram.⁴⁾

Based on the basic knowledge^2,3,4) above, as a guideline for sintering operation, to promote oxidation of magnetite up to melting in the sintering material layer, magnetite is not placed close to CaO under high oxygen potential gas flow. Therefore, it is also important not to bring it close to coke fine. Furthermore, in order to suppress the liquid phase formation, it was concluded that it is effective that the iron ore with a high Al₂O₃ content is placed close to the magnetite ore.

For the actual sintering process, “Separate granulation method”⁵⁾ in which two kinds of granulates generated in each granulating line is effective to design sinter material placing as for close or remoting. Describing details, sintering raw material is divided into two groups, and one group mainly composed of magnetite ore and the other group mainly composed of flux and coke fine. Then these two group are granulated in each granulating system and after granulated materials are mixed and sintered. Effect of this method will be evaluated after next chapters.

3. Experimental Method

3.1. Pot Testing Method

Table 1 shows sintering raw material blending conditions and components of the sinter raw material bland. Regarding iron ore, four major iron ore used in Japan were used in addition to magnetite Conc. Coke fine blend ratio was changed from 4.5 mass% to 6.0 mass%. For cases 0A and 0B, magnetite blending ratios were 15 mass% and 0 mass%, respectively. For both cases, one line granulation method was adopted. On the other hand, Case 1 and Case 2 adopt separate granulation in which raw material is divided into two lines and granulated separately. And in both cases, blending ratio of magnetite ore was 15 mass%. In Case 1, hematite D ore and the minimum amount of quick lime required for granulation was blended in the separate granulation group (Sep. Notation shown in Table 1) containing magnetite ore. The other raw materials containing most CaO sources (quicklime, limestone) and coke fines was blended in the main granulating group (Main. Notation shown in Table 1). In case 2, high Al₂O₃ goethite B was compounded in place of hematite D ore blended in separate granulation group (Sep. In Table 1 notation). That is, in case 1, magnetite ore was located remotely from CaO source and coke fine, and in case 2, magnetite ore and high Al₂O₃ ore was located remotely from CaO source and coke fine and were located close to each other.

Table 1. Blending condition and calculated chemical composition.

Case Brand		0A	0B	1		2
				Main	Sep.	Main	Sep.
		(%)	(%)	(%)	(%)	(%)	(%)
Blending Ratio	Hematite_A	14.7	19.6	14.7	0.0	14.7	0.0
	Geothite B	19.6	24.5	19.6	0.0	5.9	13.7
	Geothite C	19.6	24.5	19.6	0.0	19.6	0.0
	Hematite_D	13.7	13.7	0.0	13.7	13.7	0.0
	Magnetite Conc.	14.7	0.0	0.0	14.7	0.0	14.7
	flux (MgO–SiO2)	2.0	2.0	2.0	0.0	2.0	0.0
	Quick lime	2.9	2.9	2.4	0.5	2.4	0.5
	Lime stone	12.7	12.7	12.7	0.0	12.7	0.0
	Total	100.0	100.0	71.1	28.9	71.1	28.9
	Coke breeze	4.5~6.0	4.5~6.0	5.0~5.5	0.0	5.0~5.5	0.0
Remarks	Moisture content	7.4	7.0	6.1	11.0	6.1	11.0

Case	0A	0B	1		2
			Main	Sep.	Main	Sep.
	(mass%)	(mass%)	(mass%)	(mass%)	(mass%)	(mass%)
t-Fe	57.40	57.16	53.02	66.71	54.19	64.21
SiO2	5.29	5.30	6.34	3.07	5.50	4.85
CaO	9.75	9.78	13.53	1.71	13.48	1.82
MgO	1.14	1.17	1.56	0.24	1.53	0.31
Al2O3	1.70	1.00	2.18	0.69	1.82	1.45
P2O5	0.04	0.05	0.06	0.02	0.05	0.03
CaO/SiO2	1.84	1.85	2.14	0.56	2.45	0.37

Regarding the granulation method, with respect to the one line granulation method and main granulation (Main Notation shown in Table 1), after mixing for 4 minutes with a drum mixer (diameter 600 mm, length 800 mm), moisture was added and further mixing with the drum mixer for 4 minutes. On the other hand, with respect to separate granulation (Sep. In Table 1), after mixing for 1 minute with a high speed agitating mixer, moisture was added and granulating with a pan pelletizer (pan diameter 800 mm, depth 200 mm) for 5 minutes.

Sinter pot with 300 mm in diameter and 500 mm in height was set on a wind box of the same diameter. After charging the sinter mixture, the thermocouple was set at the center of the horizontal section at the height of 240 mm and 430 mm from the top of the pot. In addition, a thermocouple was also set in the center of the wind box.

Ignition time was 1 minute with a burner using LPG as fuel while sucking at a pressure in the windbox −5.2 kPa. After the ignition, the pressure inside the windbox was −10.3 kPa constant. The pressure in the wind box was adjusted by the opening degree of the damper installed in the duct between the wind box and the blower. The exhaust gas temperature was measured with a thermocouple installed in the central part of the wind box, and the sintering time was defined as the time to the peak time of the exhaust gas temperature. Part of the exhaust gas was continuously collected from the end of the ignition to the end of the sintering, and CO, CO₂ and O₂ were analyzed after dehumidifying and dust removal treatment. CO and CO₂ were infrared absorption method, O₂ is magnetization method, portable analyzer was used. Air permeability during firing was determined as the flow rate of gas flowing through the sintering raw material layer by continuously measuring the air flow rate from the orifice plate installed in the exhaust gas duct.

The sintered cake discharged from the sintering pot was dropped from 2.0 m height four times. After dropping, the sinter sieved over 5 mm was defined as sinter product.

3.2. Sinter Quality

After product was sieved, the large sinter was crushed and sieved again, 15 kg sinter with 10–40 mm diameter was treated in rotating drum for 200 revolutions. Then cold strength (Tumble Index) was defined as +6 mm ratio. Meanwhile, an RI test based on the old JIS method was conducted on 500 g of a 19–21 mm sinter.

3.3. Pore Evaluation by Optical Microscope Cross-section Observation and Image Analysis

Sinter 15–19 mm was embedded in and cut into resin to observe the pores in the two-dimensional cross section. In the image analysis method,¹¹⁾ the number of pores in the cross section was counted.

The pore area was measured for each pore and the circle equivalent diameter of the same area was calculated. Pore miniaturization was evaluated by the frequency distribution of circle equivalent diameter. On the other hand, the perimeter was measured for each pore and defined as the perimeter length “A”, and the circle equivalent diameter of the equivalent area was defined as the perimeter length “B”, and the circularity degree = perimeter length B/perimeter length A was obtained. The degree of circularity is an index of the degree of irregularity, and under the formation of a liquid phase, since the pores are rounded, the circularity increases. On the other hand, in the case where no liquid phase is formed, the shape of the pores before firing is maintained, so generally the circularity becomes small.

3.4. Identification of Minerals by Powder X-ray Diffraction

Sinter 15–19 mm was pulverized to make the particle size under 10⁻⁵ m order. This pulverized samples were analyzed by a X-ray diffractometer method (X-ray source: CuKα, detector: scintillation counter). The focusing method was used for the X-ray optical system, and the measurement method was adopted by 2θ-θ scan. In order to perform Rietveld analysis, high S/N ratio is necessary in the 2θ range as wide as possible. Therefore, the measurement was carried out by a step scan method in which the range of 2θ was moved at intervals of a fixed interval (Δ2θ) and the diffraction peak at a specified time was detected under the condition in which 2θ was 10 to 120° and Δ2θ was 0.04° with exposure time of 20 s.

In this paper detailed explanation about the Rietveld method¹²⁾ is omitted. In shortly, this analysis method has superiority for optimizing the diffraction pattern. The optimization means matching the approximate structural model as closely as possible with the actually measured pattern. Here, the approximate structural model is calculated by use of the information contained in the whole powder diffraction pattern to the utmost.

4. Experimental Results

4.1. Sinter Reducibility and Sinter Strength

Figure 5 shows the influence of coke fine ratio and granulation method on FeO concentration. In both the ordinary granulation case (0A, 0B), FeO concentration increases with the increase in coke fine ratio in sinter mixture. Compared with Case 0B, FeO content was higher in Case 0A, where the magnetite blending ratio was higher, which means that blended magnetite ore is not perfectly oxidized to hematite. Compared with Case 0A FeO content was higher in case 1 and case 2, even though magnetite blending ratio was equal. Therefore, separate granulation was effective for lowering FeO content. Under the condition of coke fine blending ratio 5.5 mass%, case 1 and case 2 had FeO contents equivalent to case 0B of magnetite blending ratio 0 mass%. In other words, when separate granulation under blending conditions based on the technical idea gotten from the research group was adopted, oxidation of magnetite was promoted.

Fig. 5.

Influence of coke breeze ratio and granulation method on FeO in sinter.

Figures 6 and 7 show influence of coke fine blend ratio on reducibility (RI) and sinter strength (TI). Case 0A, which is ordinary granulation case with blending magnetite ore, indicate tendency that reducible property decreases, and the sinter strength rose as the blended coke ratio increases.

Fig. 6.

Influence of coke breeze ratio and granulation method on reducibility.

Fig. 7.

Influence of coke breeze ratio and granulation method on sinter strength.

Further, case 0B, which is ordinary granulation case without blending magnetite ore, indicated high reducibility and equal sinter strength compared with case 0A. This means that FeO content increased due to blending magnetite ore deteriorates reducibility, but generation heat value due to oxidation of magnetite ore was also small, so that the sinter strength was not improved.

Both reducibility and sinter strength were improved when the separate granulation method was adopted under the blending of magnetite except for Case 1 with a coke fine blending ratio of 5.5 mass%. This improvement is due to promotion of magnetite oxidation, FeO content in sinter decreases so that reducibility, and sinter strength improved through heat generation increase due to oxidation promotion.

Figure 8 shows the heat patterns at 100 mm position and 250 mm position from the upper surface of the packed bed in case of 0A, 1, and 2, which was the magnetite blending condition at the coke fine blending ratio of 5.0 mass%. In Fig. 8, the upper figures show data for 100 mm position and the lower figures shows data for 250 mm position. And the right-side figures focus on temperature range over 1000°C. Compared with ordinary granulation, the case of separate granulation showed small difference in maximum temperature, but the high temperature holding time over 1200°C extended.

Fig. 8.

Increasing high temperature (>1200°C) holding time by use of separating granulation method.

4.2. Pore Evaluation by Optical Microscope Cross-section Observation and Image Analysis

Figure 9 shows sinter structure of separate granulation (Case 2) and ordinary granulation (Case 0A) under the blending condition of 15 mass% magnetite ore at a coke fine blending of 5.0 mass%. First of all, in the case of separate granulation, there were many pores under 100 μm diameter but in the case of ordinary granulation, there were many pores around 50 μm diameter and pores with over 100 μm diameter. Furthermore, in the case of ordinary granulation, since the pores were also rounded, it is presumed that the integration of the pores was promoted through melting. Next, in terms of minerals, many spotted hematite was present in the case of separate granulation, but in the case of ordinary granulation, many of the skeletal hematite was found. In general, skeletal hematite precipitates from the liquid phase, and spotted hematite precipitates from the solid phase. It is considered that spotted hematite is derived from fine magnetite and from high Al₂O₃ goethite ore remotely located from limestone which is a liquid phase source. That is, it is considered that the magnetite was oxidized and changed to spotted hematite. From the characteristics of pores and minerals, it is considered that in the case of ordinary granulation, a large amount of liquid phase was generated in the sintering process as compared with the case of separate granulation. From the characteristics of pores and minerals, it is considered that in the case of ordinary granulation, a large amount of liquid phase was generated in the sintering process as compared with the case of separate granulation.

Fig. 9.

Comparison of microscope image between at ordinary granulation and at separate granulation under blending magnetite fine.

Figure 10 shows the circle equivalent diameter distribution of separate granulation (Case 2) and ordinary granulation (Case 0A) at a blending condition of 15 mass% magnetite ore at a coke fine blending ratio of 5.0 mass%. Two case samples were provided for each. Case 2 adopting separate granulation and case 0A adopting ordinary granulation have peaks of circle equivalent diameter of 40–70 μm and 100–250 μm respectively, and the equivalent circle diameter was smaller in the case of separate granulation. Also, the integrated value at the circle equivalent diameter means the pore volume, but in the case of the separate granulation, the pore volume was large.

Fig. 10.

Increasing number of pore and decreasing pore size by use of separate granulation method.

Figure 11 shows circularity distribution. In the case of separate granulation, there were many pores with circularity degree of 0.5 to 0.8. On the other hand, in the case of ordinary granulation, few pores with circularity under 0.8 and mostly over 0.8 were observed.

Fig. 11.

Decreasing pore circle factor by use of separate granulation method.

Thus, Figs. 9, 10, 11, compared with the case of ordinary granulation, it seems that the separate granulation case leads to reduction of skeletal hematite, inhibition of pore integration and decrease of pore circularity due to melt suppression. Here, suppression of pore integration and reduction in pore circularity are advantageous for acceleration of the reduction reaction.

4.3. Identification of Minerals by Powder X-ray Diffraction

Figure 12 shows the results of mineral quantification obtained based on Rietveld analysis. In the case of separate granulation, hematite increased, and magnetite decreased as compared with the case of ordinary granulation. That is, oxidation from Fe²⁺ to Fe³⁺ by separate granulation was promoted. For SFCA (multicomponent calcium ferrite) and silicate type minerals, the difference due to granulation method was small.

Fig. 12.

Influence of granulation method on mineral constitution.

5. Consideration

5.1. Concept of Heat Consumption

In this study, it was concluded that heat generation such as heat of magnetite oxidation affected both sinter cold strength and reducibility. Here, heat value of the sintered layer was quantitatively evaluated for each layer height as the amount of heat consumed in the sintered layer considering the heat generation amount of the sintered layer and the sensible heat of the gas flowing through the sintered layer. Figure 13 shows concept of heat generation in sintering bed. The sintered layer was divided into three layers, and heat value consumed in the sintering layer was calculated by heat inlet from upper layer plus the generated heat minus heat outlet to lower layer. Here, the heat inlet and the heat outlet were multiplied by time integrated value for product of air flowing volume, gas specific heat, and temperature measured by the thermocouple installed on the upper and lower sintered layers respectively. The integration time was determined as the time in which the flame front was present in each of the three layers. The arrival time of the flame front was set to the time when the thermocouple temperature reached 70°C.

Fig. 13.

Concept of heat generation in sintering bed.

5.2. Heat Production Rate of Sintered Layer

Heat generation in the sintered layer is coke combustion in consideration of combustion efficiency and oxidation from magnetite to hematite. The combustion efficiency was calculated based on the air volume and the exhaust gas composition (CO, CO₂, O₂). Figure 14 shows volume balance between inlet and outlet of the sintered layer. Regarding the gas composition, balance calculation was made with four components CO, CO₂, O₂, and N₂ ignoring a small amount of SOx and NOx as several hundred ppm. The inlet side gas composition was O₂ concentration 21%, N₂ concentration 79%. The N₂ gas concentration on the outlet side is obtained by subtracting the sum of CO, CO₂, and O₂ concentration from 100%. From N₂ concentration in outlet gas, the quantitative ratio between the inlet side gas and the outlet side gas is determined.

Fig. 14.

Volume balance between inlet and outlet gas.

And, the outlet gas amount is larger than the inlet gas amount, and the difference is the CO₂ amount produced from the carbonate such as limestone and the half amount of the CO generated by the coke combustion. The former is a desorbed gas from a solid which does not consume oxygen gas, and the latter is because 2 mol of CO is produced from 1 mol of oxygen molecule. Then, if CO₂ generated from carbonate is determined, CO₂ generated by coke combustion is obtained. From the coke fine amount in sinter mixture and the fixed carbon concentration in the coke, the amount of fixed carbon in the raw material layer can be determined, and CO and CO₂ derived from coke combustion can also be determined, so that the amount of unburned carbon after sintering can be determined. In this experiment, since the outlet gas amount is continuously analyzed, the complete combustion (C+CO₂→CO₂), incomplete combustion (C+ ½O₂→ CO) and unburnt amount of the coke is obtained. Here, the heat value in complete combustion and incomplete combustion is 408 J/mol and 125 J/mol as shown in the following formula. In this paper, the coke combustion efficiency is the ratio to the heat quantity obtained when the whole fixed carbon is completely burned, the coke combustion heat amount is taken into consideration from coke combustion efficiency.   

$$\begin{array}{l}\mathrm{C}+{\mathrm{O}}_2=\mathrm{C}{\mathrm{O}}_2+408\mathrm{\hspace{0.17em} }\hspace{0.17em}\mathrm{k}\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l}:\mathrm{\hspace{0.17em} }\hspace{0.17em}\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{e}\mathrm{t}\mathrm{e}\mathrm{\hspace{0.17em} }\hspace{0.17em}\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{b}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\\ \mathrm{C}+1/2{\mathrm{O}}_2=\mathrm{C}\mathrm{O}+125\mathrm{\hspace{0.17em} }\hspace{0.17em}\mathrm{k}\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l}:\mathrm{\hspace{0.17em} }\hspace{0.17em}\mathrm{i}\mathrm{n}\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{e}\mathrm{t}\mathrm{e}\mathrm{\hspace{0.17em} }\hspace{0.17em}\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{b}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\end{array}$$

On the other hand, oxidation heat from magnetite to hematite was given at 115 kJ/mol per 1 mol of magnetite as shown in the following formula.   

$$\mathrm{F}{\mathrm{e}}_3{\mathrm{O}}_4+1/4{\mathrm{O}}_2=3/2\mathrm{\hspace{0.17em} }\hspace{0.17em}\mathrm{F}{\mathrm{e}}_2{\mathrm{O}}_3+115\mathrm{\hspace{0.17em} }\hspace{0.17em}\mathrm{k}\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l}$$

Regarding oxidation heat from magnetite to hematite, it is difficult to calculate the absolute value of heat quantity from the iron ore sinter FeO value. This is because oxidation-reduction reactions involving CO and CO₂ need to be taken into consideration in addition to the oxidation reaction of the magnetite with oxygen. Therefore, in this study, oxidation heat from magnetite to hematite is relatively evaluated as heat amount difference between the experimental cases. That is, it is assumed that sinter FeO concentration difference described as ΔFeO is the oxidation reaction difference due to oxygen from magnetite to hematite.

Table 2 shows heat of coke burning and of magnetite oxidation. Compared with ordinary granulation, the case with separate granulation showed higher coke combustion efficiency. Heat generation increase by 15 to 30 MJ per 1 t sinter mixture, but this value corresponds to 1.2 to 2.4% with respect to the total heat amount of coke combustion. Meanwhile, as sinter FeO decreased, the heat of oxidation from magnetite to hematite was increased. This heat increase was 17 to 30 MJ per 1 t sinter mixture. That is, by adopting the separate granulation, the heat increase per 1 t sinter mixture was evaluated as 45 to 47 MJ by both effects of improving coke combustion efficiency and accelerating magnetite oxidation. It is considered that these two effects made high temperature holding time increasing.

Table 2. Increasing heat generation of coke combustion and magnetite oxidation by use of separating granulation method.

Coke combustion (A)
	Constitue of exhaust gas				Combustion (MJ/t)
	CO	CO2	CO+CO2	Unburnt	[difference]
0A	10.1%	83.5%	93.6%	6.4%	1236.9	–
1	11.6%	85.2%	96.7%	3.3%	1266.7	29.9
2	10.7%	84.4%	95.1%	4.9%	1251.7	14.8

Magnetite oxidation (B)				Heat generation (A+B)
	FeO in sinter		Oxidation (MJ/t)	Heat generation (MJ/t)
		difference	difference		difference
0A	9.5	–	–	1236.9	–
1	8.3	−1.2	17.0	1283.7	46.9
2	7.4	−2.2	29.8	1281.5	44.7

5.3. Sintered Layer Circulating Gas Sensible Heat

Gas sensible heat is the product of the flowing gas amount, the gas density, the gas specific heat, and the gas temperature. Here, the amount of flowing gas was taken as the exhaust gas amount measured at the exhaust gas piping orifice set after the sinter pot. Here, it took 2 seconds for the gas flowing through the sintered layer to pass through the orifice, so this time difference was taken into account. Gas density was taken as density value at exhaust gas temperature. The specific heat of the gas depends on the gas composition and the temperature, and the gas temperature was measurement value from thermocouple inserted in the sintered layer. For calculating specific heat, gas temperature was taken as the difference with reference to the gas temperature on the upper inlet side of the sintered layer. Therefore, inlet gas sensible heat amount for the upper layer is zero. Table 3 shows the inlet and outlet gas sensible heat amount in each of the three case layers. The sensible heat of inlet gas has a value of about 600 MJ/t for middle and lower layer because high-temperature gas is supplied from upper-layer. On the other hand, the sensible heat of outlet gas is about 20–30 MJ/t because outlet gas temperature for each divided layer were at about 60–70°C during calculating period.

Table 3. Gas sensible heat at each vertical position.

	Gas sensible heat(MJ/t)
Integlation period	Flame front in upper zone		Flame front in middle zone		Flame front in bottom zone
Inlet and outlet position	inlet	outlet	inlet	outlet	inlet	outlet
	surface	(a) in Fig. 14	(a) in Fig. 14	(b) in Fig. 14	(b) in Fig. 14	(c) in Fig. 14
0A	0	24	642	93	617	23
1	0	20	680	32	690	24
2	0	23	698	30	577	26

5.4. Heat Consumption of Sintered Layer

Table 4 shows the calculation results of heat consumption. It was higher by about 600 MJ/t in the middle and lower layers compared with the upper layer. This is the difference in sensible heat of inlet gas. Next, in comparison between the cases, case 1 in which separate granulation was adopted, showed higher heat consumption in the upper and middle layers compared with case 0A of ordinary granulation. And case 2 in which magnetite fine ore and high Al₂O₃ hematite ore were treated in the same separate granulation system, showed higher heat consumption in any of the upper, middle and lower layers compared with case 0A of ordinary granulation. In the lower layer, the case 1 was higher for case 0A and case 2 was equal for case 0A. The increase amount of heat consumption for case 1 to case 0A is 150 MJ/t in the middle layer and 120 MJ/t in the bottom layer. These values were higher compared with upper layer of 50 MJ/t. These higher values is considered as effect of increasing the sensible gas heat supplied from the upper layer to the middle layer and from the middle layer to the lower layer.

Table 4. Heat consumption at each vertical position.

	Heat consumption in sintering bed (MJ/t)
	Upper	Middle	Bottom	Average
0A	1213	1786	1831	1610
1	1264	1932	1950	1715
2	1259	1950	1832	1680

Based on the results and discussion, the mechanism for both sinter strength and the reducibility increase are summarized in Fig. 15.

Fig. 15.

Mechanism of promoting magnetite ore oxidation and Improving sinter strength and reducibility.

Separate granulation method in which magnetite fine ore and high Al₂O₃ ore are granulated in the same separate system with minimum quick lime necessary for granulation, has purpose that magnetite ore is remotely located from limestone which is the CaO source and close to the high Al₂O₃ ore. This method has an effect on suppression of melting reaction, which has a role of accelerating magnetite oxidation. As a result of this oxidation acceleration, hematite increases, and magnetite decreases in sinter, which has effect on improving reducibility. On the other hand, the promotion of oxidation of magnetite ore raises the high temperature holding time of the sintered layer and leads to improving sinter strength. Here, melting suppression was demonstrated in sinter structure with suppression of pore integration, miniaturization of minerals, and low circularity.

6. Conclusion

In this research, in order to improve both the sinter reducibility and strength in the condition of blending magnetite fine ore, separate granulation method in which magnetite fine ore and high Al₂O₃ iron ore with minimum amount of quicklime necessary for granulation was suggested and evaluated by a sintering pot test. As the results, the following knowledge was obtained as follows.

(1) When separate granulation by the above blending design is adopted, FeO in the iron ore sinter decreases, and both sintering reducibility and cold strength were improved.

(2) From sinter structure, pore integration suppression, pore low circularity, mineral refinement was confirmed. These observation results are matched with melt suppression during sintering.

(3) Magnetite reduction and hematite rise in the sintered mineral were confirmed. This result corresponds to the FeO decrease in (1) above.

(4) From the above (1) to (3), it is thought that this separate granulation method promotes oxidation of magnetite ore by suppression of melting reaction, which resulting in improvement in sintering reducibility and sinter strength.

References

• 1)   T.  Maeda and  Y.  Ono: Tetsu-to-Hagané, 72 (1986), 775 (in Japanese).
• 2)   K.  Fujino,  T.  Murakami and  E.  Kasai: CAMP-ISIJ, 29 (2016), 7 (in Japanese).
• 3)   H.  Ohgi,  K.  Ohno,  T.  Maeda and  K.  Kunitomo: CAMP-ISIJ, 29 (2016), 5 (in Japanese).
• 4)   M.  Hayashi,  Y.  Katahira and  T.  Watanabe: CAMP-ISIJ, 29 (2016), 11 (in Japanese).
• 5)   T.  Kawaguchi,  K.  Kuriyama,  S.  Sato and  K.  Takata: Tetsu-to-Hagané, 76 (1990), 1642 (in Japanese).
• 6)   T.  Haga,  A.  Ohshio,  Y.  Hida,  H.  Fukuda and  N.  Ogata: Tetsu-to-Hagané, 83 (1997), 233 (in Japanese).
• 7)   Y.  Hosotani,  N.  Konno,  S.  Kabuto,  M.  Kitamura and  T.  Abe: Tetsu-to-Hagané, 83 (1997), 347 (in Japanese).
• 8)   T.  Kawaguchi,  S.  Kasama and  T.  Inazumi: Tetsu-to-Hagané, 78 (1992), 1053 (in Japanese).
• 9)   T.  Matsumura,  K.  Morioka,  M.  Shimizu,  T.  Noda and  T.  Matsuo: Tetsu-to-Hagané, 82 (1996), 23 (in Japanese).
• 10)   S.  Machida,  K.  Nushiro,  K.  Ichikawa,  H.  Noda and  H.  Sakai: Tetsu-to-Hagané, 92 (2006), 755 (in Japanese).
• 11)   M.  Hoshi,  M.  Matsumura and  T.  Kawaguchi: CAMP-ISIJ, 18 (2005), 22 (in Japanese).
• 12)   H.  Rietveld: Acta Crystallgr., 22 (1967), 151.