Mineral Phase Ratio of MEBIOS Sinter and its Reducibility under Blast Furnace Condition with High Hydrogen Concentration

Abstract

For decreasing carbon dioxide emission from iron- and steelmaking industry, hydrogen utilization in the blast furnace has been considered. Iron ore sinter is one of the major iron sources for the blast furnace. It is known that the effects of the hydrogen concentration in the reducing gas on the reducibility and reduction disintegration of the sinter are strong. In this study, the evaluation of the mineral phase of the Mosaic EmBedding Iron Ore Sintering (MEBIOS) sinter and reducibility using a laboratory scale furnace simulated the blast furnace condition with high hydrogen concentration and the experimental blast furnace were carried out.

MEBIOS sinter has higher ratio of the mineral phases, i.e., hematite and acicular calcium ferrite with primary hematite, showing higher reducibility compared to the conventional sinters. The value of JIS-RDI of the MEBIOS sinter decreases with an increase in that of JIS-RI, which is the reverse trend of the conventional sinter. Further, the basket charge test with the experimental blast furnace confirmed that the MEBIOS sinter shows high reducibility under the condition of high hydrogen concentration while the reduction disintegration behavior at low temperature is similar to conventional sinter.

1. Introduction

Drastic decrease in carbon dioxide emissions from iron- and steelmaking industry has been strongly required. Especially, the amount of CO₂ emissions from ironmaking process is large, and its countermeasure is an urgent issue. There are many reports on the decrease in CO₂ from blast furnace (BF). Lowering the thermal reserve zone temperature in BF by decreasing the distance between iron oxide and carbonaceous particles may be one of the effective methods. The mixed charging of iron ore with coke to BF¹⁾ and the utilization of highly reactive coke²⁾ and iron ore–carbon composites³⁾ have attracted significant attention. Several studies have been conducted on the reduction behavior of composites containing iron oxide and carbonaceous materials from the viewpoint of the reduction and melting behaviors under the BF conditions.^4,5) However, the reducible values of CO₂ emissions, less than 10%, have not been sufficient.

The utilization of large amounts of hydrogen as reducing gas is other promising ways to decrease the amount of carbon dioxide emissions. When the hydrogen concentration of the reducing gas increases, the reduction behaviors of iron burdens become different from that under conventional condition, because the reduction rate by hydrogen is faster than that of CO gas.⁶⁾ Furthermore, the temperature distribution in the blast furnace will drastically change because the reduction of wustite by hydrogen is endothermic reaction.⁷⁾ An increase in hydrogen concentration leads to a decrease in the shaft region temperature of BF and the broadening in the shaft region at lower temperature. Therefore, higher reducibility of iron ore burden at lower temperature is required under higher hydrogen condition.

Iron ore sinter, pellet, and lump ores are typical burdens for BF, and especially sinter is major one in East Asian countries. Major phases of sinter are primary and secondary hematite, magnetite, multicomponent calcium ferrite, and slag. Multicomponent calcium ferrite can be categorized into four types based on coexisting phases such as 1H-ACF, 2H-CF, M-FCF, and M-CCF.⁸⁾ 1H-ACF is acicular calcium ferrite with the size of less than 10 μm coexisting with primary hematite. 2H-CF is columnar calcium ferrite with secondary hematite. M-FCF and M-CCF are fine and course calcium ferrite phases coexisted with magnetite, respectively. Increasing hydrogen concentration leads to the reduction of hematite, 1H-ACF, and 2H-CF in the shaft region of BF.

On the other hand, the progress of reduction at lower temperature accelerates the degradation of sinter. Generally, it is well known that higher reducibility of sinter results in worse degradation property.⁹⁾ These are based on the results of CO reduction of sinter. Increasing hydrogen concentration in BF also results in the acceleration of sinter degradation because reduction of sinter accelerates.¹⁰⁾ The main factor of the degradation is volumetric expansion during the reduction from hematite to magnetite.¹¹⁾ Especially, secondary hematite has a strong effect on the disintegration¹²⁾ while secondary hematite has high reducibility. These results indicate that it is difficult to realize both higher reducibility and lower reduction disintegration property for sinter simultaneously under high hydrogen BF condition.

An increase in gangue contents such as Al₂O₃ and SiO₂ in Australian ores will be inevitable.¹³⁾ Hence, it is expected to increase the utilization amount of concentrate ores which is produced by the beneficiation of pulverized iron ores. An increase in the ratio of fine ore such as concentrate in raw materials of sinter results in the decreases in the permeability of the sintering bed and therefore the productivity of sinter.¹⁴⁾ To maintain the gas flow pass in the sinter bed, Mosaic EmBedding Iron Ore Sintering, MEBIOS process was proposed.¹⁵⁾ In this process, the pre-granulated green pellets (GP) having the size between 10 and 15 mm^φ distributed in the sintering bed. The surface of GPs acts as a wall and it promotes the gas flow through the bed.

In this study, MEBIOS sintering bed structure which composes of GP produced using fine ore and sinter feed¹⁶⁾ as shown in Fig. 1 is focused on. During sintering, the shape of the pellets can be maintained by lowering the CaO content in the GP. Therefore, it has the functions of the shrinkage prevention of the sintering bed in addition to maintaining the gas permeability. Surrounding raw materials can be melted by increasing CaO content. It leads that the GP is designated as the starting point of formation of acicular calcium ferrite (1H-ACF).¹⁷⁾ The size of GP and the amount of coke breeze may affect the microstructure and formed mineral phases of produced sinter. However, there has been no report on such points of the MEBIOS sinter.

Fig. 1. Image of structure change in the sinter bed of MEBIOS sinter. (Online version in color.)

There has been no report on the relationship between the mineral structure of MEBIOS sinters and their reducibility under the reducing gas containing a larger amount of hydrogen, although their reducibility under CO gas reduction has reported.¹⁸⁾ In this study, the effect of the GP size and coke breeze ratio on the microstructure of the MEBIOS sinter and reducibility for N₂–CO–CO₂–H₂–H₂O gas system were examined using laboratory scale furnace and experimental blast furnace.

2. Experimental

2.1. Production and Evaluation of Pot Sinters

The MEBIOS sinter samples were prepared using a sinter-pod. Two types of iron ores, burnt lime, limestone, silica sand, return fine, and coke breeze were used as raw materials and their chemical composition were listed in Table 1. The mixing ratio of raw materials was listed in Table 2, together with their practice sizes. The mixing ratio was determined that the chemical composition of produced sinter became to be 10 mass%CaO - 4.6 mass%SiO₂ -1.4 mass%Al₂O₃ - 0.5 mass%MgO. Coke breeze was added only to the surrounding raw materials. The admixing amount of coke breeze was set at 4.5, 4.7, and 5.0 mass%. The sizes of the GP were varied as 5–8, 8–10, and 10–15 mm. The basicity (CaO/SiO₂) of GP was 2.0. The ratio of GP to surrounding raw materials was set at 30%. For comparison, uniform granulation (UG) sample with the same composition as the MEBIOS sinter was also prepared. The coke breeze ratio of UG sample was 4.7 mass%. Furthermore, standard sinter (STD) sample which has typical chemical composition in Japanese steel mill was prepared, whose chemical composition was listed in Table 3.

Table 1. Chemical composition of raw materials.

	T–Fe	SiO2	Al2O3	CaO	MgO
	mass%
Ore A	57.02	5.64	1.58	0.08	0.23
Ore B	65.55	1.48	1.12	0.03	0.07
Burnt Lime	0.00	0.77	0.17	95.41	1.69
Limestone	0.09	0.26	0.09	54.85	0.64
Silica sand	1.30	92.22	2.58	0.10	0.23
Return fine	56.83	5.28	1.84	10.60	1.24
Ash in coke breeze	0.56	6.20	3.56	0.31	0.12

Table 2. Compositions of surrounding raw material and green pellet.

		Ore A	Ore B		Burnt lime	Limestone	Silica sand	Return fine	Coke breeze
Particle size (mm)		−8	−8	−0.15	−0.25	−5	−2.3	−4.75	0.25–1.0
MP (mass%)	Surrounding raw materials (SRM)	48.0	6.4	0.0	0.0	15.9	1.1	28.6	4.5/4.7/5.0
	Green pellet (GP)	0.0	0.0	97.0	3.0	0.0	0.0	0.0	0.0
Uniform granulation (UG) (mass%)		33.6	29.1	4.5	0.9	11.1	0.8	20.0	4.7

Table 3. Chemical composition of STD sinter sample.

T–Fe	FeO	SiO2	Al2O3	CaO	MgO	CaO/SiO2
58.0	10.1	5.4	1.9	9.7	1.0	1.8

Figure 2 shows the pre-granulation process flow of raw materials. GPs were produced using a pan pelletizer adding water of 6 mass% after mixing the raw materials using a stirring and mixing machine. Raw materials for surrounding materials were mixed using a concrete mixer for 180 s with 20 rpm, and then it was granulated using a drum mixer for 360 s with the rotation speed of 7.5 rpm. After that, GPs were added to the surrounding raw materials in a drum mixer with the wight mixing ratio of 3:7, and these were mixed for 30 s with 7.5 rpm. The mixture of GPs and surrounding raw materials was charged into sinter pot,¹⁹⁾ and sinter sample was prepared.

Fig. 2. Granulation process of raw materials for sinter pot test. (Online version in color.)

The values of JIS-RI, JIS-RDI, and JIS-TI of the sinter samples were evaluated based on JIS-8713, JIS-8720, and JIS-M8712, respectively. Mineral phase ratio of hematite and magnetite in the sinter was determined by the internal standard method of XRD using NaF regent as a standard material. The ratio of calcium ferrites and slag was made by the image analysis using more than twenty microstructure images of pellet and matrix areas in a few sinter particles obtained using objective lens of 50 times. Furthermore, specific surface area of sinter sample was measured by BET.

2.2. Reduction Test of Sinter Sample

Sinter sample of approximately 100 g with the particle size of 6.7–9.5 mm was charged in the crucible with an inner diameter of 53 mm having several holes at the bottom to pass through a reducing gas. Alumina balls with the diameter of 10 mm were packed at its bottom. Then, it was set in the furnace as shown in Fig. 3⁸⁾ and pre-heated up to 200°C under nitrogen gas flow of 13.1 NL/min. After that, the reducing gas was flowed to the chamber with the same flow rate, and the sample was heated up to 700°C with the rate of 10°C/min. Subsequently, it was heated up to 1000°C with the rate of 4°C/min. Two types of the reducing gas composition were used; Low-H₂ (N₂ - 48%(CO+CO₂) - 5.8%(H₂+H₂O)) and High-H₂ (N₂ - 48%(CO+CO₂) - 13%(H₂+H₂O)) conditions. The ratio of CO/CO₂ and H₂/H₂O was changed with increasing temperature as shown in Fig. 4,⁸⁾ simulating the BF reducing gas composition. The weight change of sinter sample was measured during reduction to calculate reduction degree which was calculated from net weight change of sinter sample during heating as shown in the equation.

  

$$\text{Reduction}\mathrm{\hspace{0.17em} }\text{degree}=\frac{\mathrm{\Delta }{\text{W}}_1+\mathrm{\Delta }{\text{W}}_{2+}}{{\text{W}}_{3+}}\times 100$$(1)

Here, ΔW₁ is net weight change, ΔW₂₊ is the weight change assuming that Fe²⁺ in original sinter is oxidized to Fe³⁺, and W₃₊ is oxygen weight in sinter caused by iron oxide when all iron oxide is Fe³⁺. Net weight change was calculated from the difference between that of sample and blank test. Blank test was carried out using MgO crucible with only alumina balls using same gas and heating conditions.

Fig. 3. Schematic diagram of experimental apparatus for reduction. (Online version in color.)

Fig. 4. Partial pressure of reducing gas on the phase diagram of Fe–C–O system and Fe–H–O system. (Online version in color.)

2.3. Basket Charge to Experimental Blast Furnace

The sample of STD and MP47M (4.7mass%coke, GP: 8–10 mm) with the size of 10–25 mm was set into the basket with 60 mm in diameter and 150 mm in length, and the baskets were charged into the experimental blast furnace between the sinter and coke layers. Two types of temperature sensor made of oxide (Referthermo L and L1) were set into the basket, and the maximum temperature was obtained to measure the size change of the sensor before and after the experiment. The measurable temperature ranges of L and L1 are 1050–1300°C and 800–1150°C, respectively. The basket charge was started from 4 h 30 min before BF shut down. After shut down of the BF, the furnace was cooled by the blowing nitrogen gas. The samples were recovered from the furnace.

The height of experimental blast furnace is 6.5 m from the tuyere, and the maximum and top diameters are 1.6 m and 1.0 m, respectively. The operating conditions when the basket charge was carried out were as follows; the volume of blast gas was 1013 Nm³/h, blast temperature was 1080°C, PCI is 150 kg/t-HM, coke ratio was 390 kg/t-HM, the volume of COG was 95 Nm³/t-HM, the oxygen enrichment was 15.6%, the production amount was 34 t/d, and hydrogen concentration in COG was 50–55%.

After collecting the basket, particle size distribution of the samples was measured. The degradation ratio of the sample with less than 2.8 mm was calculated using the size distribution. Reduction degree was evaluated from the content of total Fe and Fe³⁺ and microstructure was observed using optical microscope.

3. Results and Discussion

3.1. Microstructure and Strength of Sinter

The appearance of nine types of sinter samples was shown in Fig. 5. Increasing the green pellet size and decreasing the coke ratio result in the increase in the amount of the non-melted brown particles indicated by the arrows in the figure. Especially, many bare pellets are observed in the sample using the pellet size of 10–15 mm. It indicates that the connection between the pellet and surrounding materials does not proceed suggesting the lower strength of the sample. Figure 6 shows the TI value of the samples. The TI value of UG sample is 65.2%, which is indicated as a plane in the figure and their values are also listed in Table 4. TI value decreases in increasing the pellet size and this result agrees with the prediction from the appearance observation. There is the trend that the TI value increases with decreasing coke ratio. The value of the sample using the pellet size less than 10 mm and below 4.7% coke is higher than that of UG. On the other hand, that using large pellets (L) is lower than that of UG. In this study, there is no coke addition into the pellets. It means that the strength of the pellets is low because of insufficient sintering of the pellets.

Fig. 5. Appearances of sinter samples prepared using sinter pot. (Online version in color.)

Fig. 6. Effect of average pellet size and coke ratio on TI of sinter. (Online version in color.)

Table 4. Size of green pellet, total-Fe, FeO, JIS-RI, JIS-RDI, JIS-TI, and initial reduction degree before experiment of the sinter samples.

	Size of GP (mm)	T-Fe	FeO	JIS-RI	JIS-RDI	JIS-TI	Initial reduction degree
		(mass%)		(%)
UG	–	56.81	9.74	67.5	36.8	65.2	4.44
MP45S	5–8	58.02	7.90	68.8	39.6	67.6	3.53
MP47S		58.28	8.59	69.6	38.0	66.4	3.60
MP50S		57.74	7.68	68.0	39.0	63.9	3.45
MP45M	8–10	57.81	6.67	70.1	31.4	69.2	3.11
MP47M		58.43	5.76	72.3	33.1	65.7	3.35
MP50M		57.57	9.63	68.9	34.8	61.0	4.22
MP45L	10–15	58.15	7.44	71.0	35.1	60.4	3.32
MP47L		57.72	9.02	71.4	35.2	54.4	3.03
MP50L		58.17	6.94	67.7	35.5	53.3	3.09

Cross-sectional view of MP47M sample is shown in Fig. 7. Red circles in the figure are the pellet part and the nuclear particle is observed in the pellet. Other area is matrix made of the surrounding raw materials. The interface between the pellet and matrix is not clear because these materials react each other. Large pores with the size of a few millimeters are observed in the matrix. There is a possibility that these pores may be an origin point during crushing and there are few large pores in the pellet. Figure 8 shows the microstructures of ① and ② areas in this MP47M sample. The area ① is inside of the pellet, whose mineral phases are primary hematite and 1H-ACF. Micro pores with the size of a few micrometers are observed among 1H-ACF phase. The interface between the pellet and matrix in the area ② is not clear. The matrix is composed of pore, primary hematite and 1H-ACF. The size of the primary hematite in the area ② is smaller than that in the area ① and the number of micro pores is lower. The reason is that the area ② has higher CaO content than the area ① and the amount of formed melts is larger. Therefore, many micro pores and primary hematite disappear. Figure 9 shows the typical microstructures of matrix, and major mineral phase in Fig. 9(a) is secondary hematite, 2H-CF, and slag and that in Fig. 9(b) is magnetite, M-FCF, and slag. There are typical microstructures of conventional sinter.⁸⁾

Fig. 7. Macrostructure of sinter sample MP47M. ① inside the pellet, ② interface between pellet and matrix, and ③ in the matrix. (Online version in color.)

Fig. 8. Microstructure a) in the pellet (① in Fig. 6.) and b) at the interface between pellet and matrix where it is ② in Fig. 6. in sinter sample MP47M. (Online version in color.)

Fig. 9. Microstructure of matrix area in sinter sample MP47M, where it is ③ in Fig. 6. (Online version in color.)

Mineral phase ratios of the pellet and matrix in nine sinter samples are shown in Figs. 10(a) and 10(b), respectively. Mineral phase ratios were calculated by the method described in 2.1. In the pellet, hematite, 1H-ACF, 2H-CF, magnetite and the small amount of slag are detected. The ratio of hematite, 1H-ACF, and 2H-CF, which are the phase accelerating the reduction under high hydrogen condition (Reduction Acceleration Phase Ratio, RAPR),⁸⁾ is high. In the matrix, on the other hand, the ratio of magnetite, M-CCF, M-FCF, which is categorized as the phase that reduction is not accelerating, is high. The reason is that the coke breeze was added into the surrounding raw materials, and the maximum temperature in the surrounding materials becomes higher than that in the pellet. Mineral phase ratio of magnetite increases with increasing the amount of coke addition. The calculated results from the sum of the mineral phase ratios of the pellet and matrix at a ratio of 3:7 which is based on the mixing ratio of raw materials is shown in Fig. 10(c) together with that of UG. Mineral phase ratio of magnetite in the UG sample is higher than that of MP samples, of which calcium ferrite phases is lower. In MP samples, the ratios of magnetite and RAPR are lower and higher than that of UG, respectively. In case of 4.7%coke addition, especially, RAPR is highest.

Fig. 10. Effect of green pellet size and mixing ratio of coke breeze on the ratio of mineral phase promoting H₂ reduction. (Online version in color.)

3.2. Reducibility of Sinter

The change in JIS-RI value with the average pellet size and the coke ratio is shown in Fig. 11. The result of UG is indicated as a plane (67.5%) as same as Fig. 6. JIS-RI shows higher value in case of 4.7 mass% coke and pellet size M. That of UG has lowest value. It is well known that conventional sinter has 30 to 40 vol% magnetite and an increase in magnetite content results in lowering reducibility.²⁰⁾ UG has highest magnetite content in this study. Therefore, JIS-RI has lowest value.

Fig. 11. Effect of average pellet size and coke ratio on RI of sinter. (Online version in color.)

Figure 12 shows the change in JIS-RDI value with the average pellet size and the coke ratio. In the range of coke ratio in this study, the effect of the average pellet size on the JIS-RDI value has stronger than that of the coke ratio although the effect of the coke ratio depends on the average pellet size. The value of MP-S sample which was prepared using small size green pellet is higher than that of UG sample. MP-S sample has smaller amount of magnetite and larger amount of 2H-CF than UG sample. As a result, it is expected that the amount of secondary hematite is larger and it seems to be the reason why JIS-RDI value of MP-S has higher. In case of MP-M sample which was prepared using middle size green pellet, the JIS-RDI value increases with increasing the coke ratio. Mineral phase ratios of magnetite, M-FCF, and M-CCF of MP50M are higher than those of the other MP-M samples while these phases have inert effect on the reduction disintegration behavior. However, an increase in coke ratio leads to increasing sample temperature described above. It results in the increasing the amount of secondary hematite. This may be the reason of the effect of coke ratio on the RDI value of MP-M sample. There is no dependence of coke ratio on the RDI value of MP-L sample.

Fig. 12. Effect of average pellet size and coke ratio on RDI of sinter. (Online version in color.)

Figure 13 shows the relationship between the JIS-RDI and JIS-RI values of the samples in this study together with that of the sinters produced in the actual steel mills in Japan (actual sinter). There is a trend that the JIS-RDI value of the actual sinter increases with increasing the JIS-RI value. The value of UG is included in this area. In case of MP samples, on the other hand, the JIS-RDI value decreases with increasing the JIS-RI value. This is an irregular trend on the conventional sinter, and it may be caused by the pellet addition because it leads to changing the mineral phase ratio. It indicates that there is a possibility to produce the sinter with both higher reducibility and lower reduction disintegration by the optimization of the pellet layout in the sinter bed.

Fig. 13. Effect of average pellet size and coke ratio on RDI of sinter. (Online version in color.)

Specific surface area, SSA of various sinter samples such as UG, MP47M, and eight actual sinters is shown in Fig. 14. The measurement of SSA of MP47M sample was carried out after separating into the pellet and matrix by sorting visually after pulverizing to a few millimeters. SSA of UG and matrix in MP47M has similar values to that of actual sinter. However, that of the pellet in MP47M has twice higher than that of other samples since the pellet has many fine pores. It is expected that the amount of secondary hematite formed is smaller because major mineral phase in the pellet is hematite and 1H-ACF as described in Fig. 10. This is the point to realize both higher RI and lower RDI values.

Fig. 14. Specific surface area of various sinter samples such as UG, MP47M, and actual sinter.⁸⁾ (Online version in color.)

Change in reduction degree of MP47M and UG samples by high-H₂ condition with temperature is shown in Fig. 15. Reduction degree of MP47M at 1000°C was highest in the MP samples. The initial value of reduction degree of the samples has different because the value was calculated from the equation (1). Reduction degree of all samples increases with increasing the temperature. However, reduction rate of MP47M is higher than that of UG above 600°C and that at 1000°C is also higher. To discuss the difference in the reduction rate, reduction experiment of MP47M was carried out using not only MP47M but also the pellet and matrix parts only. The start value of the pellet in MP47M is lowest because the ratio of Fe²⁺ in the pellet is lower. However, the reduction rate of the pellet is largest. The reason is that major phase in the pellet is hematite and there is large SSA value in the pellet as shown in Fig. 14. Compared with the reduction degree of MP47M, the value of matrix in MP47M is lower at 800°C and 1000°C.

Fig. 15. Reduction curves of P only, MP47M, M only and UG under high hydrogen atmosphere. (Online version in color.)

Reduction degree was calculated from the sum of those values of the pellet and matrix at respective temperature at a ratio of 3:7. Figure 16 shows the change in calculated and measured reduction degrees of MP47M sample with the temperature. The calculated result is good agreement with that of measurement one. It means that the reduction behavior can be estimated by the mixing ratio of the pellet and matrix.

Fig. 16. Reduction curves of MP47M and curve of summing 30% pellet only and 70% matrix only. (Online version in color.)

Figure 17 shows the effect of the average pellet size and coke ratio on reduction degree at 1000°C under low-H₂ and high-H₂ conditions. In all samples, reduction degree under high-H₂ condition is higher than that under low-H₂ condition. Reduction degree of some MP samples shows lower value than that of UG sample. Within all samples, MP47L sample shows highest reduction degree. However, this sample was crushed to the size of 6.7–9.5 mm before the experiment while the average pellet size was 10–15 mm. Therefore, RAPR cropped on the surface of the sample. In case of other samples such as MP47M, the pellet surface is covered with matrix. Therefore, MP47L shows higher value. It can be concluded that MP47M is the best condition.

Fig. 17. Effect of a) green pellet size and b) mixing ratio of coke breeze on reduction degree of sample at 1000°C under High-H₂ and Low-H₂ conditions. (Online version in color.)

It is reported that the difference between reduction degrees under high-H₂ and low-H₂ conditions at 1000°C bears a proportionate relationship to RARP.⁸⁾ Figure 18 shows the relationship between this difference and RARP. The previous data⁸⁾ was also plotted on. The ratio of slag is included in the RAPR value of this study while that is not in the previous report. In case of UG and a few MP samples, the RAPR value is in the same trend line of the previous study. However, most of MP samples has different trend. The reason may be the existence of the pellet in the sinter because the pellet has a different pore structure compared with the sinter. The sinter has many pores of various sizes from a few millimeters to micrometer scale. On the other hand, the pellet has many pores of similar size. The pore affect to the reduction behavior strongly.²⁰⁾

Fig. 18. Relationship between the difference of reduction degree of samples reduced up to 1000°C under High-H₂ and Low-H₂ conditions, and ratio of the phases of which reduction are accelerated by an increase in the partial pressure of H₂. (Online version in color.)

3.3. Basket Charge into Experimental Blast Furnace

MP47M and STD samples were charged into the experimental blast furnace using the basket. Microstructures of the reduced MP47M and STD samples recovered from the same level of the blast furnace (2.8 m from the tuyere) are shown in Fig. 19. Here, Figs. 19(c) and 19(f) are the surface and center of STD sinter particles. Microstructures of both the pellet and matrix in MP47M are shown in Figs. 19(a)–19(c). Metallic iron forms at the surface of the pellet and matrix of MP47M and STD and their microstructures are almost similar. In case of the center of the matrix in the MP47M particles, however, there is no metallic iron formation while small amount of metallic iron forms in the pellet. The amount of metallic iron formed in the pellet is smaller than that of STD. Figure 20 shows the coordinates of the recovery position of MP47M and STD samples from the experimental blast furnace. The coordinate (0, 0) means the center of the blast furnace at each level. The arrows indicate the coordinates of the recovery position of the samples shown in Fig. 19. Most of MP47M samples are located at the outer of the furnace than STD. There is a possibility that the maximum temperature of MP47M sample is different from that of STD because the temperature of the outer side of the blast furnace is lower than that of the center. Figure 21 shows the change in reduction degree with maximum temperature measured using the sensors. The arrows in the figure have the same meaning as Fig. 20. Reduction degrees of both MP47M and STD increases with increasing the temperature. These values rapidly increase at a certain temperature, and then, the reduction reaction finishes. This temperature of MP47M is 50°C lower than that of STD. It indicates that MP47M sample has high reducibility.

Fig. 19. Microstructure of samples obtained from experimental blast furnace. a) and d) are pellet part in MP47M, b) and e) are matrix, and c) and f) are STD. (Online version in color.)

Fig. 20. XY coordinate data of recovery position of MP47M and STD samples from experimental blast furnace. (Online version in color.)

Fig. 21. Change in reduction degree of MP47M and STD samples with temperature measured using oxide sensors in the experimental blast furnace. (Online version in color.)

Figure 22 shows the relationship between degradation ratio of the sample below 2.8 mm in particle size and reduction degree of the recovered sample. There is no significant difference between two samples when reduction degree is below 60%. It means that reduction disintegration behavior is similar while increasing hydrogen concentration leads to accelerating the disintegration in the blast furnace because the operation condition of this experimental blast furnace is higher hydrogen concentration than that of conventional one. At higher reduction degree than 60%, degradation ratio increases with increasing reduction degree. At this reduction degree range, however, the reduced sinter is partially melted because the temperature becomes higher than 1200°C. Therefore, this increase in degradation ratio is not important.

Fig. 22. Relationship between degradation ratio below 2.8 mm and reduction degree of MP47M and STD samples. (Online version in color.)

4. Conclusions

In this study, the concept of MEBIOS sintering process was applied to prepare the sinter samples having high reducibility under the reducing gas condition with high hydrogen concentration. The properties of the sinter samples such as TI, RI, and RDI were examined. Furthermore, reducibility and degradation behaviors of the selected sinter sample were under the basket charge test into the experimental blast furnace.

(1) When the mixing ratio of coke breeze decreases in the preparation of MEBIOS sinter, strength of the sinter sample decreases due to insufficient sintering of the pellet. On the other hand, higher levels inhibit the formation of acicular calcium ferrite with primary hematite (1H-ACF) that promotes the reduction by the gas with high hydrogen concentration.

(2) The pellet in the MEBIOS sinter is composed of primary hematite and 1H-ACF which are the mineral phase accelerating reduction. Matrix is composed of calcium ferrite with secondary hematite, magnetite, and fine and columnar calcium ferrite with magnetite.

(3) The ratio of mineral phase accelerating the reduction (RAPR) of the MEBIOS sinter is larger than that of the sinter produced by uniform granulation because 30% of raw materials is the pellet in the MEBIOS sinter. RAPR increases with decreasing the green pellet size because of increasing the interface area between the pellet and matrix. From the viewpoint of this ratio, there is the optimum value of coke ratio.

(4) JIS-RDI of the MEBIOS sinter decreases with an increase in JIS-RI, which is an opposite trend compared with the conventional sinter. There is no proportional relationship between RAPR and the difference the reducibility of the MEBIOS sinter under high and low hydrogen concentrations in the reducing gas.

(5) The part of the pellet in the MEBIOS sinter has high reducibility. This is the reason why the MEBIOS sinter has high reducibility. By the basket charge into the experimental blast furnace, it was confirmed that MEBIOS sinter has higher reducibility than the conventional sinter and shows a similar disintegration behavior.

Declaration of Competing Interest

The authors declare that they have no known competing financial interest or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgment

The present work was carried out as a part of COURSE50 project. Financial support by the New Energy and Industrial Technology Development Organization (NEDO) is gratefully acknowledged.

References

• 1)   S.  Watakabe,  K.  Takeda,  H.  Nishimura,  S.  Goto,  N.  Nishimura,  T.  Uchida and  M.  Kiguchi: Tetsu-to-Hagané, 92 (2006), 901 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.92.12_901
• 2)   S.  Nomura,  H.  Terashima,  E.  Sato and  M.  Naito: Tetsu-to-Hagané, 92 (2006), 849 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.92.12_849
• 3)   A.  Kasai,  M.  Naito,  Y.  Matsui and  Y.  Yamagata: Tetsu-to-Hagané, 89 (2003), 1212 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.89.12_1212
• 4)   T. Y.  Huang,  D.  Maruoka,  T.  Murakami and  E.  Kasai: ISIJ Int., 59 (2019), 1982. https://doi.org/10.2355/isijinternational.ISIJINT-2019-178
• 5)   T.  Zhang,  Q.  Wue,  G.  Wang and  J.  Wang: ISIJ Int., 58 (2018), 2219. https://doi.org/10.2355/isijinternational.ISIJINT-2018-372
• 6)   D.  Spreitzer and  J.  Schenk: Steel Research, 90 (2019), 1900108. https://doi.org/10.1002/srin.201900108
• 7)   H.  Nogami,  Y.  Kashiwaya and  D.  Yamada: Tetsu-to-Hagané, 100 (2014), 251 (in Japanese). https://doi.org/10.2355/tetsutohagane.100.251
• 8)   T.  Murakami,  H.  Wakabayashi,  D.  Maruoka and  E.  Kasai: ISIJ Int., 60 (2020), 2678. https://doi.org/10.2355/isijinternational.ISIJINT-2020-180
• 9)   K.  Mochizuki,  T.  Murai,  Y.  Kawaguchi and  Y.  Iwanaga: Tetsu-to-Hagané, 72 (1986), 1855 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.72.14_1855
• 10)   N.  Takeuchi,  Y.  Iwami,  T.  Higuchi,  K.  Nushiro,  N.  Oyama and  M.  Sato: Tetsu-to-Hagané, 99 (2013), 448 (in Japanese). https://doi.org/10.2355/tetsutohagane.99.448
• 11)   N.  Sakamoto,  H.  Fukuyo,  Y.  Miyata and  T.  Miyashita: Tetsu-to-Hagané, 70 (1984), 512 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.70.6_512
• 12)   T.  Inazumi,  K.  Shinada and  M.  Kawabe: Tetsu-to-Hagané, 68 (1982), 2207 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.68.15_2207
• 13)   K.  Saito: Nippon Steel Technical Report, 413 (2019), 15.
• 14)   K.  Nushiro,  N.  Oyama,  K.  Igawa,  K.  Sorimachi and  T.  Uetake: Tetsu-to-Hagané, 83 (1997), 473 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.83.8_473
• 15)   E.  Kasai,  S.  Komarov,  K.  Nushiro and  M.  Nakano: ISIJ Int., 45 (2005), 538. https://doi.org/10.2355/isijinternational.45.538
• 16)   C.  Kamijo,  M.  Matsumura and  T.  Kawaguchi: ISIJ Int., 45 (2005), 544. https://doi.org/10.2355/isijinternational.45.544
• 17)   Y.  Hida,  J.  Okazaki,  K.  Ito and  M.  Sasaki: Tetsu-to-Hagané, 73 (1987), 1893 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.73.15_1893
• 18)   M.  Matsumura,  Y.  Yamaguchi,  K.  Higuchi,  T.  Murakami and  T.  Maeda: Tetsu-to-Hagané, 107 (2021), 471 (in Japanese). https://doi.org/10.2355/tetsutohagane.TETSU-2020-073
• 19)   K.  Iwase,  T.  Higuchi,  T.  Yamamoto and  T.  Murakami: Tetsu-to-Hagané, 107 (2021), 483 (in Japanese). https://doi.org/10.2355/tetsutohagane.TETSU-2020-080
• 20)   N.  Sakamoto,  H.  Fukuyo,  Y.  Miyata and  T.  Miyashita: Tetsu-to-Hagané, 70 (1984), 504 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.70.6_504