ISIJ International
Online ISSN : 1347-5460
Print ISSN : 0915-1559
ISSN-L : 0915-1559
Regular Article
Low Temperature Reduction Disintegration Mechanism of Self-fluxing Pellet under High Hydrogen Condition of Blast Furnace at 500°C
Koki Momma , Daisuke Maruoka, Eiki Kasai, Taichi Murakami
Author information
JOURNAL OPEN ACCESS FULL-TEXT HTML

2025 Volume 65 Issue 6 Pages 719-727

Details
Abstract

Currently, the Japanese steel industry is developing the technologies to reduce 30% of domestic CO2 emissions from iron- and steelmaking industry using an innovative ironmaking process such as the COURSE50 project, which focuses on H2 reduction and CCUS technologies. On the other hand, it has been reported that an increasing H2 gas ratio in the blast furnace reducing gas promotes the low temperature reduction disintegration of the iron ore pellet. In this study, the low temperature reduction disintegration mechanism of self-fluxing pellet under higher hydrogen condition at 500°C was examined. During hydrogen reduction which proceeds uniformly, the fine cracks with few micrometers in the primary particles form. It leads to the formation of fine particles with the size of less than 0.1 mm on the surface of the pellet. Increase in the particles with few millimeters makes a significant impact on the permeability of the blast furnace. On the other hand, CO reduction without hydrogen gas proceeds topochemically, and volumetric fracture progresses with generation of the cracks with several millimeters in length due to the stress difference between near the surface and in the center of pellet. Under higher H2 conditions, reduction proceeds uniformly, so such cracks causing volume fractures are hard to form, but the finer cracks are easier formed than the case of CO reduction, and the amounts of finer particles increases.

1. Introduction

The demand for reducing CO2 emissions from iron- and steelmaking industry has increased. In Japan, the iron- and steelmaking industry accounts for approximately 14% of the country’s total CO2 emissions.1) In particular, CO2 emissions from the blast furnace ironmaking process account for more than half of the steel industry’s total emissions.2) Therefore, there is an urgent need to reduce CO2 emissions from blast furnaces ironmaking. ULCOS project in Europe and the introduction of the FINEX process in South Korea had been promoted so far.3) Currently, Japan has attempted to reduce 30% of domestic CO2 emissions through the development of an innovative ironmaking process (COURSE50), which focuses on H2 reduction and CO2 capture technology.4) This project includes the goal of reducing 10% of CO2 emissions from blast furnaces by replacing CO reduction with H2 reduction. Therefore, it will be more important to study the hydrogen reduction. According to the estimation of temperature distribution in blast furnace by Nogami et al.5) using mathematical model, it is reported that the low temperature region from 400°C to 600°C in blast furnace expands with increasing hydrogen concentration. Temperature distribution changes in the blast furnace may promote the low temperature reduction disintegration and the pulverization of iron ore burden such as sinter and pellets, leading to lowering the permeability inside the furnace. Gas permeability in the blast furnace is one of extremely important factor for stable operation.6) The gas permeability in the shaft region of the blast furnace deteriorates with an increase in the amount of powder with grain size of 5 mm or less. This phenomenon results in the abnormality of the burden descent.7) Therefore, the phenomenon of reduction disintegration of sinter is one of major issue.8,9,10,11,12) Based on previous reports, low temperature reduction disintegration of sinter cause to reduce the skeletal hematite (a kind of secondary hematite) to magnetite at the low-temperature region of 400–700°C, which results in the volumetric expansion of the sinter. The volumetric expansion leads to the formation of the cracks, and the sinter is disintegrated by impacts during burden descent.13) On the other hand, low-temperature reduction disintegration of pellets has not progressed as much as sinter under conventional blast furnace conditions.14) Pellets are mainly produced by solid-phase sintering, dissolution of primary hematite is unlikely to occur, and the skeletal hematite is almost nonexistent as a mineral phase. However, it was reported that the reduction of pellets by the gas including hydrogen leads to significant low temperature reduction disintegration. Wu et al. studied the reduction of pellets at the temperature from 450°C to 650°C and found that H2 gas promote the low temperature reduction disintegration of the pellet.15) Mizutani et al. carried out the in-situ measurement of the crack formation in the pellet during hydrogen reduction using the acoustic emission (AE) method and found that more fine cracks were generated during hydrogen reduction than in the case of no hydrogen addition.16) However, these studies do not examine the driving force of crack initiation in detail. Furthermore, these reports do not take into account water vapor in the reducing gas, while water vapor includes in the reducing gas of the blast furnace. In addition, although microstructural observation after reduction is generally used to identify the location and shape of crack initiation, the disintegration mechanism of pellet by hydrogen reduction is still unclear.

It is pointed out beneficiation processes of iron ore will increase due to lowering the grade9,10) in the future, and the production of concentrates which is the raw material for pellets, will also increase. It results in increase the supply capacity of the pellets. Considering the increase in H2 utilization as the blast furnace reductant for reducing CO2 emissions, it is necessary to control the reduction disintegration of pellets by the gas including hydrogen. It is essential to elucidate its mechanism. In this study, the reduction disintegration mechanism of pellets is examined by conducting reduction experiments and nondestructive microstructural observation of samples before and after reduction in the gas composition of high hydrogen condition of blast furnace with water vapor.

2. Experimental Procedure

The actual basic pellet was used in this study. Its chemical composition is listed in Table 1. The pellets were sieved to grain size of 9.5–11.1 mm. The value of Reduction Disintegration Index (JIS-RDI) was 2.6%.

Table 1. Chemical composition of the sample (mass%).

T.FeFeOCaOSiO2Al2O3MgO
66.900.271.641.710.340.50

Figure 1 shows the schematic diagram of the reduction experimental apparatus used in this study.6) The flow rates of N2, CO, CO2 and H2 were controlled by mass flow controllers. H2O gas was prepared by heating distilled water in a vaporizer maintained at 473 K. The mixed gas was preheated by ribbon heater set at 250°C to prevent the condensation of water vapor, and it was introduced into the reaction tube from the bottom of the furnace. The sample holder with the inner diameter of 53 mm and height of 10 mm has 34 holes with diameter of 4 mm at the bottom of the holder for the smooth gas flow in the furnace. Quartz tubes are installed inside the reaction tube to set the sample holder, and the change of the sample weight can be continuously measured by an electronic balance installed at the bottom of the apparatus. The sample temperature was measured using a K-type thermocouple inserted into the furnace from the top.

Fig. 1. Schematic diagram of experimental apparatus. (Online version in color.)

Alumina ball of 100±1.0 g with the particle size of 10 mm was inserted in the sample holder. Subsequently, the pellet samples with the total weight of 100 ± 1.0 g were randomly picked up and inserted in the sample holder. 30±5 number of pellets were used for each experiment. The sample was heated up to 500°C in a N2 gas flow with 13 NL/min and kept for 20 min. Subsequently, the gas was changed to a reducing gas with the same flow rate and the reduction experiment was carried out for selected time.

Table 2 shows the reducing gas composition. Two types of the reducing gas composition were used: N2-20%CO-20%CO2 (0%H2), and N2-12%CO-17.7%CO2-8%H2 -2.3%H2O (8%H2). “8%H2” condition was used because disintegration of sinter proceeds more significantly under 8%H2 condition.13) The gas concentration of CO and CO2 under 0%H2 condition were determined by considering conventional evaluation method of RDI. And, the gas concentration of CO and H2 under 8%H2 condition were determined by considering estimated gas condition of shaft region of hydrogen enriched blast furnace used COURSE50 project. The concentrations of CO2 and H2O gas were determined by the equilibrium conditions of the water–gas–shift reaction expressed in Eq. (1) at 500°C, after fixing the concentrations of CO and H2 gas. After the reduction experiment, the reducing gas was changed to N2 again, and then the reduced sample was cooled down.

  
H 2 +C O 2 ⇆ H 2 O+CO (1)

Table 2. Reducing gas composition of the reduction experiment (vol% at 300 K).

N2H2COH2OCO2
0%H260020020
8%H2608.0122.317.7

The reduction degree was calculated from the weight change of the sample before and after reduction using Eq. (2)

  
Reduction   degree= ( W 0 - W 1 ) W S ×100 (2)

where W0 and W1 are the weights of the sample before and after reduction, respectively. WS is the weight of oxygen coursed by iron oxide in sample.

Disintegration test was conducted on the basis of ISO 4696-2. Reduced samples were charged into a tumbling drum with an inner diameter of 130 mm and rotated for 1.8 ks with 30 rpm. After that, the sample was sieved with 2.8 mm and determined the weight of the sample. The RDI value was calculated using the following Eq. (3)

  
RDI= W 3 W 2 ×100 (3)

where W2 and W3 are the weights of the sample after the reduction experiment and that of particles smaller than 2.8 mm respectively.

After the disintegration test, the samples were sieved to the particle size of −0.25 mm, 0.25–2.8 mm, 2.8–8.0 mm, and +8.0 mm. The particle size distribution was obtained from its weight percentage. The particles sizes of −0.25 mm and 2.8–8.0 mm were defined as fine powder and intermediate grain, respectively. Disintegration tests were also conducted for 0.6 ks and 1.2 ks to discuss the formation mechanism of the fine powder and the intermediate grain.

X-Ray Diffraction (XRD) was carried out to identify the mineral composition of samples before and after reduction, using the Fe anode target with the wavelength of 0.194 nm (scan step: 0.01 deg, diffraction speed: 3.0 deg/s).

The pellet sample mainly consists of hematite, and magnetite formed after reduction. Hematite and magnetite phases can be distinguished by the optical microscopy because the color of these phases is obviously different. Therefore, the distribution of reduction degree in the pellet sample was determined. This reduction degree is defined as brightness reduction degree (BRD) and calculated according to Eq. (4). The center and edge of cross section of sample are defined as normalized radius ranging from 0% to 50% as shown in Fig. 2. “1/9” in Eq. (4) means that the reduction degree reaches 11.1% when all hematite is reduced to magnetite.

  
BRD= Area   of   Mag. Area   of   Hem.+Area   of   Mag. × 1 9 ×100(%) (4)

Fig. 2. Conceptual image of BRD measurement. (Online version in color.)

The formation of the micro- and macro-cracks by the reduction in the sample was analyzed. Microcracks is cracks which generate in the primary particles. Macrocracks is cracks which can lead volumetric destruction. Microcrack was defined as the crack width is smaller than 5 μm, and it was observed at the area within 0.5 mm from the sample surface by SEM. Microcrack density, Pmicro was calculated according to Eq. (5).

  
P micro [/μm]= l ' total [μm] A[μ m 2 ] (5)

Where l’total is the total length of microcracks in the image, and A is the cross-sectional area of samples in the image excluding the pores and cracks.

Macrocrack was defined as the crack length and width were longer than 0.5 mm and 10 μm, respectively. It was determined by inner structure observation using CT images which was obtained using the BL28B2 beamline at Spring-8, Japan. Comparing CT images of same sample before and after reduction, the propagated cracks during the reduction were observed. Then, macrocrack density, Pmacro was calculated according to Eq. (6).

  
P macro [/mm]= l total [mm] A BRD [m m 2 ] (6)

Where ltotal is total length of macrocracks formed by the reduction, excluding the crack length identified before reduction, and ABRD is cross-sectional area of the unreduced sample analyzed using CT image, excluding the pores and cracks.

3. Results and Discussion

3.1. Disintegration Behavior of Basic Pellet

The relationship between reduction degree and the RDI value of the pellet reduced at 500°C is shown in Fig. 3.17) Until reduction degree reaches about 4% in 0%H2 reduction and 2% in 8%H2 reduction, RDI shows the constant value at approximately 4%. The RDI value tends to increase parabolically in both gas conditions as the reduction exceeds. At the same reduction degree, the RDI value of 8%H2 reduction is higher than that of 0%H2. The RDI value of 8%H2 reduction is twice as large as that of 0%H2 reduction at 11% of reduction degree which corresponds to all hematite is reduced to magnetite. The trend of accelerating the disintegration of the pellet by H2 addition is agree with the previous research using the pellet.14,15) However, effect of H2 addition is different on reduction disintegration of the sinter.18) Furthermore, the increasing trend of the RDI value of 8%H2 reduction is divided into two stages. The effect of the reduction time on the particle size distribution after disintegration test of the pellet reduced by 0%H2 and 8%H2 is shown in Fig. 4.17) The amount of fine powder increases with proceeding the reduction. At the same reduction time, the amount of fine powder was larger in 8%H2 reduction than in 0%H2 reduction. In both conditions, the particle smaller than 2.8 mm, which is directly related to the RDI value, are almost consist of the fine powder. It means that the reduction disintegration mechanism of pellet is different from that of sinter12) which has wide particle size distribution. In addition, the percentage of the intermediate grains slightly increases with increasing reduction time, and it is more significant in 0%H2 reduction. In particular, an increase in the ratio of the intermediate grains in 0%H2 reduction for the reduction time from 1200 s to 1800 s. It means that some of reduced pellets caused the volumetric destruction. At this reduction stage, the reduction degree shows 4%–6%. And this behavior is corresponding to the change in the RDI value as shown in Fig. 3. This is considered that the formation of fine powder was amplified from not only the pellet surface but also the exposed surface of the intermediate grains formed during reduction.

Fig. 3. Change in the RDI value of the basic pellet with reduction degree at 500°C. (Online version in color.)

Fig. 4. Effect of reduction time on change in particle size distribution of basic pellet reduced at 500°C.

The change in the particle size distribution of the pellet reduced by 0%H2 and 8%H2 for 1800 s with the rotating time of disintegration test is shown in Fig. 5. The intermediate grains form in the early stage of the test, and the weight ratio of the intermediate grains shows almost constant value. On the other hand, the weight ratio of fine powder increases with passing the rotating time. This proposes that the intermediate grains form by the impact in the early stage, and the fine powders form during the disintegration test. Although the increase in the formation of intermediate grains is recognized in 8%H2 reduction for rotating time of 1200 s to 1800 s, this reason may not be derived from the volumetric destruction but the formation of the particle with diameter of smaller than 8.0 mm due to the abrasion from the sample surface. In this experiment, only one spheroidal particle with the size of 6.7–8.0 mm was obtained for the rotating time of 1800 s, and the weight ratio change of this particle was 1.57%. This is corresponding to the increase in the weight ratio of the intermediate grains at this time.

Fig. 5. Effect of rotating time for reduction disintegration test on change in particle size distribution.

3.2. Reduction Behavior of Basic Pellet

Figure 6 shows the change in reduction degree with the time of the sample reduced by 0%H2 and 8%H2 conditions. “0 s” means the time when the gas was changed to reducing gas. Reduction degree of 0%H2 condition starts to increase earlier than that of 8%H2. It is reported that H2 reduction take longer time for atomic iron exchange and nucleation of Fe than CO reduction in the temperature range of 350°C to 600°C.19) This may contribute to the delay of reduction under 8%H2. In both gas conditions, reduction degree reaches 11% after 3600 s and it means the reduction from hematite to magnetite completes. The BRD distribution of the sample reduced by 0%H2 and 8%H2 conditions for 1800 s are shown in Fig. 7. Under 0%H2 condition, the BRD at the outer layer is higher than that at the inner layer. It indicates that reduction reaction of 0%H2 condition topochemically proceeds. On the other hand, the BRD of 8%H2 condition is almost same value regardless of the distance from sample surface. It implies that the reduction reaction uniformly proceeds. This is because the molecular size of H2 is smaller than CO, and H2 molecular easily diffuses in the pellet sample which has a lot of open pores. The mutual diffusion coefficient of H2–N2 is about four times as large as that of CO–N2 at 500°C.20) In addition, H2 reduction take longer time for atomic iron exchange and nucleation of Fe,19) and it may delay reduction start time. These are the reason why 8%H2 reduction uniformly proceeds.

Fig. 6. Change in reduction degree of the basic pellet with time at 500°C under 0%H2 and 8%H2 conditions. (Online version in color.)

Fig. 7. Distribution of optical reduction degree of the basic pellet by 0%H2 and 8%H2 reduction at 500°C for 1800 s. (Online version in color.)

3.3. Microcrack Density and Fine Powder Formation

Figure 817) shows the SEM images near the surface of the reduced pellet by 0%H2 and 8%H2 condition for 3600 s. Both of phase are mainly consist of magnetite. Microcracks are indicated by the white arrows. More microcracks forms in the sample reduced by 8%H2 than that of 0%H2 condition. Figure 9 shows the relation between reduction time and microcrack density near the surface of the pellet. No microcracks were recognized before reduction. In contrast, microcracks were recognized in the sample reduced for 1200 s in both gas condition. It seems that the formation of microcracks is not significant at the early stage of the reduction when the RDI value does not change, as shown in Fig. 3. The RDI value starts to increase at the reduction degree of 3.8% in 0%H2 condition and 1.6% in 8%H2 condition, and the corresponding reduction time are approximately 800 s in 0%H2 condition and 1000 s in 8%H2 condition from Fig. 6. The starting point of the dotted line shown in Fig. 9 is corresponding to the time when the RDI value starts to increase. In both gas conditions, microcracks density increases as passing reduction time. Microcracks density does not change after reduction time of 1800 s in 0%H2 condition. At this time, the reduction degree reaches 9%, and the value indicates that reduction to magnetite is completed near the surface of the pellet sample. Therefore, it seems that microcracks density does not increase.

Fig. 8. SEM images of near the surface of the pellet after reduction by 8%H2 for 3600 s (microcracks are indicated by the white arrow).

Fig. 9. Change in Microcracks density with reduction time. (Online version in color.)

In contrast, microcracks density in 8%H2 condition slightly increase until the reduction time of 3600 s. It is confirmed that reduction reaction uniformly proceeds from BRD distribution shown in Fig. 7. It is considered that the reduction from hematite to magnetite continues near the surface of sample. After reduction degree reaches 11% for 3600 s reduction, microcracks density is considered no longer increase because after reduction from Magnetite to Hematite is completed. The dotted line shows the estimated microcracks density of longer reduction time.

The relationship between the weight ratio of fine powder and microcracks density is shown in Fig. 10. In both gas conditions, the weight ratio of fine powder increases as microcracks density increases. However, the weight ratio of fine powder reduced in 8%H2 condition is higher than that of 0%H2 condition even if they had almost same reduction degree. It indicates that other factors, for example, mechanical strength of reduced sample may affect the generation of fine powder from the sample.

Fig. 10. Change in the ratio of fine powder with Microcracks density. (Online version in color.)

3.4. Relationship between Macrocrack Density and Intermediate Grain Generation

Figure 11 shows CT images of samples before and after reduction for 1200 s in 0%H2 and 8%H2 conditions, respectively. Although many micropores are recognized in the samples before reduction, macrocracks which have several millimeters lengths does not exist. Some macrocracks indicated by the white arrows are recognized inside of sample reduced by 0%H2 condition for 1200 s. In contrast, few macrocracks are recognized in sample reduced by 8%H2 condition.

Fig. 11. CT images before and after reduction (500°C-1200 s).

Figure 12 shows the relationship between reduction time and the macrocracks density of sample reduced in 0%H2 and 8%H2 conditions. In 8%H2 condition, no macrocracks are recognized until reduction time of 1200 s and the macrocracks density rise after reduction for 1800 s. In 0%H2 condition, the macrocracks density increase as reduction proceeds, and the macrocracks density of 0%H2 condition is higher than that of 8%H2 condition even if they had almost same reduction degree. Figure 13 shows the relationship between the macrocracks density and the weight ratio of intermediate grains. In both gas conditions, the weight ratio of intermediate grains increases with increasing the macrocracks density. However, the slope of both conditions is different. It indicates that other factors, for example, the mechanical strength of reduced sample may affect the generation of intermediate grains.

Fig. 12. Change in macrocracks density with reduction time. (Online version in color.)

Fig. 13. Change in the ratio of intermediate grains with macrocracks density. (Online version in color.)

3.5. Mechanism of Low Temperature Reduction Disintegration of Basic Pellet

Based on the above results and discussion, it is considered that microcracks and macrocracks affect the formation of fines powder and intermediate grains, respectively. It is known that the main factor of the disintegration of iron ore burden is the volumetric expansion of the iron oxide phase due to the reduction from hematite to magnetite at low temperatures such as 500°C.21) Therefore, the reduction disintegration mechanism of the pellet is discussed based on the formation mechanism of microcracks and macrocracks.

First, the mechanism of microcracks is discussed. Figure 14 shows the schematic image of the pellet before and after reduction. The red arrows indicates that the stress applied by the reduction. The pellet consists of the network structure of the primary particles, and reduction reaction proceeds much more at the surface of the primary particles. When the difference of reduction degree become large between the primary particle surface and the center, then tensile stress applies at this point. This stress would lead to the formation of microcrack.

Fig. 14. Conceptual images of Microcrack Initiation Mechanism. (Online version in color.)

Second, the mechanism of macrocracks is discussed. The mechanism of macrocrack formation by 0%H2 reduction is shown in Fig. 15(a). The red arrows indicates that the stress applied by the reduction. The darker the color of the iron oxide particles means the higher reduction degree. Reduction reaction proceeds much more at the surface of the pellet because of topochemical reduction. The stress applied at the surface become larger than that of the center. When the difference of reduction degree become large between the sample surface and the center, the large tensile stress applies at this point. This stress would lead to the formation of macrocrack. At the same time, this macrocrack formation results in the stress relaxation at the primary particles. Accordingly, the microcrack formation is suppressed, compared with the case of 8%H2 reduction. In case of 8%H2 reduction, the macrocrack as shown in the case of 0%H2 reduction does not form. However, the macrocracks was observed in the case of 8%H2 reduction for 1800 s as shown in Fig. 11. The magnified CT images at the area where the macrocrack was observed in the pellet for 1800 s reduction in both gas cases are shown in Fig. 16. The macrocrack in the 8%H2 reduction is thinner than that in the 0%H2 reduction. Furthermore, the former is bend in tens of micrometers while the latter crack is straight in hundreds of micrometers. It means that a macrocrack forms by the connection of many microcracks in the case of 8%H2 reduction as shown in Fig. 15(b).

Fig. 15. Conceptual images of macrocrack Initiation Mechanism. (Online version in color.)

Fig. 16. Difference in macrocracks of the pellet by reducing gas. (Online version in color.)

4. Conclusions

In order to achieve a significant reduction of CO2 emissions from the ironmaking process, technology development is underway to actively utilize H2 as blast furnace reduction agent. In this study, the reduction disintegration behavior of the basic pellets is examined. Then, reduction test was conducted under the reducing gas with and without hydrogen at 500°C to clarify the reduction disintegration mechanism. The findings of this study are summarized below.

(1) In both gas conditions, the reduction disintegration of the pellets proceeds by the formation of fine particles smaller than 0.25 mm. The amounts of fine particles increase with the addition of H2.

(2) Compared with CO reduction, the hydrogen reduction results in decrease in the formation of the intermediate grains with the size from 2.8 to 8.0 mm.

(3) The reduction reaction with the hydrogen addition uniformly proceeds. As a result, the stress by the reduction from hematite to magnetite also accumulate uniformly in the pellet and many microcracks forms at the primary particles of iron oxide. It results in the fine particles formation.

(4) CO reduction without hydrogen proceeds topochemically in the pellet. It leads to the distribution of reduction degree in the pellet. The large tensile stress generates at the point where the difference of reduction degree becomes large. This stress leads to the formation of macrocrack. It results in the formation of the intermediate grains.

Statement for Conflict of Interest

The authors declare no conflicts of interest associated with this manuscript.

Acknowledgements

CT imaging was conducted at the BL28B2 of SPring-8 with the approval of the Japan Synchrotron Radiation Research Institute (JASRI) (Proposal No. 2022A1510).

This work is supported by ISIJ research Group “Lumpy zone control for next generation hydrogen enriched blast furnace”.

References
 
© 2025 The Iron and Steel Institute of Japan.

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs license.
https://creativecommons.org/licenses/by-nc-nd/4.0/
feedback
Top