Regular Article
Effect of Addition of CaO Component on the Oxidation Reaction of Wustite Particles in Sintering Bed
2015 Volume 55 Issue 5 Pages 940-946
Details
2015 Volume 55 Issue 5 Pages 940-946
This study has been performed to understand the reaction behavior of wustite particles and the effect of existing state of CaO component in the iron ore sintering bed for their effective utilization as an agglomeration agent. Changes in the structure and pressure drops of a sintering bed were measured by using a laboratory-scale sintering simulator.
When wustite particles were mixed with model pellet of raw materials, pressure drop of the sintering bed did not show significant change independent of average CaO content between 5 and 10 mass%CaO. This is because most of wustite particles and model pellets kept their initial shapes and therefore the structural change of the sintering bed did not occur. On the contrary, pressure drop of the bed changed drastically when wustite particles were mixed with CaO particles with 1.0–2.0 mm in particles size. In these cases, wustite and CaO particles were melted and agglomerated each other. Addition of CaO component at the vicinity of the wustite particles seems to decrease in liquidus temperature of local composition and promote the melt formation. In order to effectively utilize wustite containing materials as an agglomeration agent, it is essential to arrange sufficient amount of CaO component close to such particles for melt formation at lower temperature.
Efforts have been required for reduction of CO2 emissions to the iron and steel industry to mitigate the global warming. Iron ore sinter is the most commonly used blast furnace burden in the world. Amount of sinter produced in Japan is above eighty million ton a year and CO2 emissions from the sintering process account for about 3% of total domestic emissions. This emission is caused by the utilization of coke fine called agglomeration agent, which is main heat source of the process.
Mineral form of Fe of major iron ore utilized for the sintering process is hematite. Their prices have considerably increased as compared to several years ago,1) and development of new iron resources is progressing.2) In this situation, the increase in the production mass of magnetite ore, which contains Fe2+, is expected2) and its utilization would influence the sintering performance.
Other iron source containing Fe2+ is, for example, mill scale produced in the milling process. Mill scale contains lower oxidation state ferrous materials such as wustite and magnetite. The efficient utilization of their oxidation heat would have a potential to decrease the amount of coke utilized in sintering process. However, utilization of a large amount of mill scale often caused a decrease in the productivity of sinter.3) The reason of this seemed to be a decrease in permeability of the sintering bed, since the oxidation of mill scale did not give void like coke. Utilization partially reduced iron4) and steel can chip5) was also tried to apply the oxidation heat to the sintering bed. These results also suggest the utilization of iron containing materials in the sintering process leads to decrease in sinter productivity. However, a detailed studies on the reactivity of these materials and the effect on the structural change of sintering bed were not sufficiently made although it is necessary to efficiently utilize their oxidation heats in sintering process.
The purpose of this study is to clarify the high temperature reaction and sintering behavior in the bed when using wustite component as an agglomeration agent. Oxidation behavior of wustite particles was compared to that of coke and metallic iron particles. In addition, the effect of compositions and existing state of CaO components on the structural change of the sintering bed was examined.
Wustite (diameter: 1.0–2.4 mm), coke (Fixed carbon 87.5 mass%, diameter: 1.0–2.0 mm) and metallic iron particles (reagent, Fe 99.9 mass%, diameter: 1.0–2.0 mm) were used. Wustite was prepared by the reduction of hematite reagent (Purity 99.9 mass%, diameter: 2.0–5.0 mm) under the atmosphere of [CO: CO2 = 1:1] at 1000°C for 144 ks. Mineral phase was identified as wustite of FeO1.094 by XRD. The value was derived by the lattice parameter based on the method reported by R. L. Levin et al.6) Prepared wustite particles were screened between 1.0–2.4 mm.
In order to avoid the complexity in reactions led by the effect of particle shape and/or gangue composition, model pellets of the granules, called ACP (Alumina Cored Pellet) of raw materials were used. They are prepared as follows: powder mixture of Fe2O3 and CaCO3 was granulated around alumina balls (ϕ2 mm) by using a laboratory-scale pelletizer with addition of water. Prepared granule consisted of alumina ball as a nuclei and powder mixture as an adhering layer. Compositions of adhering layer were varied as shown in Table 1 on the basis after calcination. Prepared pellets were screened between 2.38–2.80 mm. Average CaO composition of the sample was adjusted by choosing the kind of model pellets. In order to discuss the effect of existing state of CaO component, CaO particles (LP, diameter: 1.0–2.0 mm) were prepared by crushing the CaO tablet made by calcination and sintering of CaCO3 tablet.
| Sample name | Fe2O3 (mass%) | CaO (mass%) |
|---|---|---|
| ACP-10 | 90.0 | 10.0 |
| ACP-15 | 85.4 | 14.6 |
| ACP-15 | 73.9 | 26.1 |
| ACP-15 | 61.2 | 38.8 |
Figure 1 shows schematic diagram of sintering simulator.7) Reaction tube was made of alumina, whose inner diameter was 35 mm. The height of sample bed was set at 20 mm. For pre-heating the gas and preventing the drop of melted sample, packed bed of alumina ball (ϕ2 mm) with 50 and 20 mm height were placed above and below the sample bed, respectively. After charging the sample and alumina ball beds, sample bed was heated to 900°C under N2 stream to prevent the oxidation of agglomeration agent. Gas flow rate was set to 0.45 Nm·s–1. When sample bed temperature became stable, gas was changed to N2-21 vol%O2 with the same gas flow rate. Then a heat profile was given to sample bed by the oxidation heat of agglomeration agent. Sample bed temperature at center and bottom, pressure drop of sample bed and alumina ball beds and gas compositions of CO, CO2 and O2 in the out let gas were measured by thermo couples set 10 and 20 mm below the top of sample bed, pressure drop meter and infrared gas analyzer, respectively.
Schematic diagram of sintering simulator.
The amount of model pellets packed in sample bed was 26 g when sample bed height was set at 20 mm. In order to keep average CaO composition in sample bed as 10 mass%CaO in the cases using coke, metallic iron and wustite, model pellets with different CaO compositions were chosen. Since preheating temperature was set at 900°C, calcination reaction of CaCO3 in the adhering layer seemed to be completed during the preheating process. Therefore, the calcination heat would not affect the oxidation reaction of the agglomeration agents.
Additive amount of agglomeration agents was based on the case of 5 mass% coke (1.31 g-coke in sample bed). In order to keep the total amount of generation heat, the amount of metallic iron and wustite in sample bed were calculated as 6.06 and 21.2 g, respectively. These values were obtained assuming complete reactions, i.e., all carbon and iron oxides in the agglomeration agent would be oxidized to CO2 and Fe2O3, respectively. When wustite particles are mixed with ACP, the height of sample bed did not fit within 20 mm because heat density of wustite was not large as other agglomeration agents. In this case, the bed height was set higher than other cases. In the case using CaO particles, they were mixed with 21.1 g of wustite particles. In the case using the CaO particles, they were mixed with 21.1 g of wustite particles. Base CaO concentration of the sample bed was set as 10 mass%CaO and it was changed to the higher (15 mass%CaO) and the lower (5 mass%CaO) values to examine the effects of existing state and CaO concentration on the structural change and pressure drop of the sintering bed.
Experimental conditions were summarized in Table 2.
| CaO source | Coke (g) | Metallic Iron (g) | Wustite (g) | Total CaO in sample bed (mass%) | Diameter of agglomeration agent (mm) | Sample bed height (mm) | |
|---|---|---|---|---|---|---|---|
| Coke | ACP-10 | 1.31 | 0.0 | 0.0 | 10.0 | 1.0–2.0 | 20 |
| Fe | ACP-25 | 0.0 | 6.06 | 0.0 | 10.0 | 1.0–2.0 | 20 |
| W | ACP-35 | 0.0 | 0.0 | 21.1 | 9.5 | 1.0–2.4 | 30 |
| W | ACP-15 | 0.0 | 0.0 | 21.1 | 4.3 | 1.0–2.4 | 30 |
| W | LP-15 | 0.0 | 0.0 | 21.1 | 15.0 | 1.0–2.4 | 20 |
| W | LP-10 | 0.0 | 0.0 | 21.1 | 10.0 | 1.0–2.4 | 20 |
| W | LP-5 | 0.0 | 0.0 | 21.1 | 5.0 | 1.0–2.4 | 20 |
Bed temperature profiles at center and bottom of sample bed in the cases of coke (Coke ACP-10), metallic iron (Fe ACP-25), wustite mixed with model pellets (W ACP-15 and W ACP-35) and CaO particles (W LP-10) are shown in Fig. 2. In the case using coke and metallic iron, the profiles shows similar behaviors, which were discussed in the previous reports.7,8) In the case using metallic iron (Fe ACP-25), discontinuous and jagged changes in the bed temperatures were observed by short-circuit of thermocouple due to the formation of conductive oxide melt. Sample bed temperatures at center and bottom are lower in the case of wustite mixed with model pellets (W ACP-15 and W ACP-35) than those of coke (Coke ACP-10) and metallic iron (Fe ACP-25) cases. Lower density of generation heat by larger volume of bed would cause such a behavior. In the case using wustite mixed with CaO particles (W LP-10), temperatures levels at the center and the bottom of sample bed are lower than the cases using coke (Coke ACP-10) and metallic Fe (Fe ACP-25). The peak value of the temperature at the bottom of sample bed is higher than that at the center. The differences in the bed temperatures between the center and the bottom were small in the case using wustite mixed with model pellets (W ACP-15 and W ACP-35).
Change in the sample bed temperature, a) center and b) bottom of sample bed, with time observed for the cases using Coke (Coke ACP-10), Metallic Fe (Fe ACP 25) and Wustite (W ACP-15) (W ACP-35) (W LP-10).
The temperature difference between the center and bottom of sample bed is discussed later with formed melt, taking account of the melt formation.
Changes in O2 concentration of outlet gas in the cases of coke (Coke ACP-10), metallic iron (Fe ACP-25) and wustite mixed with model pellets (W ACP-15), (W ACP-35) and CaO particles (W LP-10) and blank are shown in Fig. 3. Blank case indicates that sample bed was packed by alumina ball (ϕ2 mm) only and it shows the exchange characteristic when gas is changed from N2 to N2-21vol%O2. In order to elucidate the reaction ratio of each agglomeration agents, the amount of O2 consumptions by oxidation reaction of agglomeration agents is calculated from the difference between O2 concentration in outlet gas in each experimental conditions and blank case. Reaction ratios of agglomeration agent in each condition are in Fig. 4. Reaction ratio was defined as unity when all fixed carbons and all iron components were oxidized to CO2 and Fe2O3, respectively. The reaction ratio of coke reaches to about 0.8. On the other hand, that of metallic iron is 0.58, that of wustite mixed with model pellets (W ACP-15) and (W ACP-35) are 0.46 and 0.49, respectively, and that of wustite mixed with CaO particles (W LP-10) is 0.51.
Change in O2 gas concentration in outlet gas with time in the cases using Coke (Coke ACP-10), Metallic Fe (Fe ACP-25), Wustite (W ACP-15) (W ACP-35) (W 100 LP-10) and blank.
Change in reaction ratio of agglomeration agent with time in the cases using Coke (Coke ACP-10), Metallic Fe (Fe ACP-25) and Wustite (W ACP-15) (W ACP-35) (W LP-10).
As shown in Fig. 4, reaction ratios in the cases using iron containing agglomeration agents are lower than that in the case using coke. Difference of the reaction products would lead to such differences. When coke is oxidized, main product is CO2 gas and it diffuses to bulk gas. In this case, oxidation reaction will not suppressed with progressing the reaction. On the other hand, when iron containing particles are oxidized, solid and/or liquid oxide phase are formed at their surface. They will suppress the diffusion of oxygen and decrease the heat generation rate. As a result, reaction ratio does not reached to sufficient value.
Changes in the pressure drop of the sintering bed observed for the cases using different agglomeration agents are shown in Fig. 5. In the case using coke, pressure drop decreases significantly after the initial increase and continuously decreased. This phenomena can be explained as follows: The initial increase in pressure drop was led by rapid increase in sample bed temperature by starting of the coke combustion and succeeding decrease was led by increasing void fraction caused by disappearance of coke and rearrangement of the bed structure.7) In the case using metallic iron (Fe ACP-25), pressure drop decreases without initial increase and after that it started to increase showing peak value at 50 s. Initial decrease was led by agglomeration without increasing void fraction such as coke. The increase in pressure drop seems to be caused by the drop-down of formed melt to lower part of the bed and block of the voids for gas flow.8) In the case using wustite mixed with model pellets (W ACP-15) and (W ACP-35), pressure drops initially increases like the case of coke. However, they do not show significant change but gradual decrease only. CaO concentration of the model pellets did not seem to affect the pressure drop of the bed. On the contrary, in the case using wustite mixed with CaO particles case (W LP-10), pressure drop initially increases a little and then decreases significantly. The initial increase of pressure drop seems to be led by an increase in sample bed temperature like the case using coke. After that, it increases again but the value is kept smaller than the cases using metallic iron.
Change in pressure drop of sample bed and alumina bed with time in the cases using Coke (Coke ACP-10), Metallic Fe (Fe ACP-25) and Wustite (W ACP-15) (W ACP-35) (W LP-10).
In order to discuss effect of the existing state of CaO component on the sintering bed, vertical cross sections of (W ACP-35) and (W LP-10) are examined as shown in Fig. 6. In the case of (W ACP-35), white balls and black particles outlined with white dotted lines are mainly observed and these are alumina balls and wustite particles, respectively. Wustite particles are remained showing their initial shape and adhering layers of ACP does not seem to be deformed. In the case of (W LP-10), On the other hand, wustite particles are melted and agglomerated each other and residual of CaO particles are not observed. Further, large blocks are observed at bottom of sample bed, which are slags formed by melting of wustite and CaO.
Vertical cross sections of sample bed after reaction in the cases using Wustite (W ACP-35 and LP-10) as agglomeration agent.
Such structural change of sample bed may cause the temperature difference between the center and bottom of the bed in the case using CaO particles as shown in Fig. 2. In this case, formed melt flowed down with unreacted wustite particles to the bottom part of the bed and it tend to contact with the bottom thermocouple as shown in Fig. 6. Therefore, the oxidation heat of wustite will directly affected the measured temperature by the thermocouple. Further, it will moderate the temperature change.
The reason of a large difference in pressure drop profiles shown in Fig. 5 seems to relate a difference between sample bed structures shown in Fig. 6. In the case of (W ACP-35), bed structure did not change significantly and therefore pressure drop of the bed mostly depended on the change in the bed temperature. While, in the case of (W LP-10), melt formation and its flow led to remarkable structural change of the like the case of metallic iron.8) This seems to be the reason of the significant decrease in the pressure drop in the early stage of sintering. Succeeding increase may be caused by the blocking of the gas flow path by the large blocks formed at the bottom of sample bed. Excess amount of melt formation will bring about similar phenomena in the sintering bed.
The results shows that wustite particles hardly performed as an agglomeration agent like coke and metallic iron particles even though the expected amount of reaction heat were justified at the same level. However, addition of CaO particle promoted both the melting and oxidation reaction of wustite particle.
3.2. Effect of CaO Composition on the Reactivity and Structure ChangeFigure 7 shows the profiles of sample bed temperatures at center and bottom in the cases using wustite mixed with CaO particles varying the average CaO concentration in three levels. Some data smoothing was made to the case of (W LP-15), since significant fluctuation was observed between 25 and 60 s, which would be caused by short-circuit of thermocouple due to formed melt. In the case of 10 mass%CaO (W LP-10), sample bed temperatures at the canter and bottom reach to 1250 and 1320°C, respectively, and then, decrease gradually. When CaO composition is down to 5 mass% (W LP-5), maximum temperature at the center decreases to 1100°C and the maximum temperature at bottom reachs at approximately 1400°C. In the case of (W LP-15), sample bed temperature at the center increases to 1200°C and that at the bottom rapidly increases to 1300°C and continues to increase gradually showing longest holding time at high temperature. This behavior is different from other two results due to the difference of structural change of the sintering bed, which will be discussed later.
Change in the sample bed temperature, a) center and b) bottom of sample bed, with time observed for the cases using Wustite with 15 mass%CaO (W LP-15), 10 mass%CaO (W LP-10) and 5 mass%CaO (W LP-5).
Oxidation reaction ratio of wustite estimated using the data of change in O2 concentration in outlet gas is shown in Fig. 8. Oxidation reaction ratio of wustite increases with increasing amount of CaO addition. The case of (W LP-15) shows the highest reaction ratio.
Change in the reaction ratio of agglomeration agent with time observed for the cases using Wustite as agglomeration agent with 15 mass%CaO (W LP-15), 10 mass%CaO (W LP-10) and 5 mass%CaO (W LP-5).
In order to discuss the reason of the different temperature profiles between (W LP-10) and (W LP-15) in spite of similar reaction ratios, vertical cross sections of sample bed after reaction are shown in Fig. 9. In the case of (W LP-5), many un-melted residual particles outlined by white dotted lines are seen, while in the case of (W LP-10), particles were melted and agglomerated each other. Large blocks are also observed at the bottom part of sample bed. In the case of (W LP-15), melting and agglomeration of particles were further progressed and most of melt are moved down to the bottom of sample bed. This may be a reason of the lower peak temperature at the center part of the bed and the longer holding time at high temperature at the bottom. In addition, higher CaO concentration seems to lead to higher reaction ratio of wustite due to larger amount of melt formation as shown in Fig. 8.
Vertical cross sections of sample bed after reaction in the cases using Wustite with 15 mass%CaO (W LP-15), 10 mass%CaO (W LP-10) and 5 mass%CaO (W LP-5).
Changes in pressure drop of the sample bed with different CaO compositions are shown in Fig. 10. In all the cases, pressure drop first increases and then decreases followed by succeeding increases. Such changes are larger in higher CaO compositions. The second increase in pressure drop was also observed in the case using metallic iron particles (Fe ACP-25) and it could be explained as the blocking of the gas flow path caused by the flow down of formed melt to the bottom.8) Significant changes in the pressure drop observed for the larger amount of CaO particle addition are led by the larger amount of melt formation, since formation of melt promotes the structural change of sintering bed.
Change in pressure drops of sample bed and alumina bed with time observed for the cases using Wustite as agglomeration agent with 15 mass%CaO (W LP-15), 10 mass%CaO (W LP-10) and 5 mass%CaO (W LP-5).
Finally, the effect of existing state of CaO component on the melt formation behavior is discussed in order to seek the effective way to use wustite as an agglomeration agent for iron ore sintering process. When wustie particles were used as agglomeration agent, consideration of Fe2+ component in wustite particles, Fe3+ component produced by oxidation, Fe3+ and CaO components from model pellets and CaO components from CaO particles is necessary for discussion about melt producing behavior. Change in the local composition of wustite will be discussed by considering the CaO–FeO–Fe2O3 diagram estimated by an equilibrium calculation software FactSage (see Fig. 11). When wustite particles are used as an agglomeration agent, magnetite is first produced on the wustite surface. While the measured bed temperatures were lower than 1597°C, i.e., melting point of magnetite, the particles would not deformed/flow. However, in fact, melting of sample was observed in the case of CaO particle addition. This suggests that melt formation started in this case. One possibility of this formation is caused by a decrease in the liquidus temperature at the contact point between CaO particles and magnetite produced on the wustite particles. The melt formation and assimilation would further proceed along the line between Fe3O4 and Fe2O3-15 mass%CaO in the case of (W ACP-15) and between Fe3O4 and Fe2O3-35 mass%CaO in the case of (W ACP-35). On the contrary, in the case using CaO particles, the melt composition can move along the line between Fe3O4 and CaO. Expected paths of melt formation and assimilation drown on the CaO–FeO–Fe2O3 diagram are shown in Fig. 11.
Assimilation paths between Fe3O4 and CaO, Fe3O4 and Fe2O3-35 mass%CaO and Fe3O4 Fe2O3-15 mass%CaO and oxidation reaction paths of W-100 LP-15, W-100 LP-10 and W-100 LP-5 drawn on the CaO–FeO–Fe2O3 diagram.
Considering the melt region shown in Fig. 11, in the case using model pellets, the maximum bed temperature measured in the case of (W ACP-15) did not reach to the lowest liquidus 1350°C expected along the line between Fe3O4 and Fe2O3-15 mass%CaO. Similarly, in the case of (W ACP-35) the bottom temperature did not reach to 1200°C, which is the lowest liquidus temperature between Fe3O4 and Fe2O3-35 mass%CaO, and temperature at center was above the temperature for few time. Therefore, structural change of the bed hardly occurred in both cases. On the other hand, in the case adding CaO particles, the lowest liquidus temperature along the line between Fe3O4 and CaO is as low as 1170°C. As shown in Fig. 7, the measured bed temperatures were above such temperature in most cases. Therefore, there were chance to form melt at the contact point between wustite surface, i.e., Fe3O4, and CaO particles. In addition, higher CaO concentration promoted formation of melt and the structural change of the bed in the range between 5–15 mass%CaO as shown in Fig. 9. Further, melt region at lower temperature are seen about 20 mass%CaO in the wide range between FeO and Fe2O3. It seems to promote the melt formation and the structural change of the bed and also the oxidation reaction of iron in the melt. Therefore, it is necessary to control the local CaO composition at the vicinity of wustite particles to about 20 mass%. This will decrease the liquidus temperature of local composition and promote the oxidation reaction of wustite.
Effect of existing state of CaO component on the sintering behavior of wustite particles was experimentally examined to efficiently utilize materials containing lower oxides as agglomeration agent of iron ore sintering process. The results are summarized as follows.
(1) In the case that wustite particles were mixed with model granules prepared by alumina balls as core particles and powder mixture of Fe2O3 and CaCO3 as adhering layers, oxidation reaction of wustite did not proceed sufficiently. On the other hand, wustite particles were melted and oxidized effectively when they were mixed with CaO particles of 1.0 to 2.0 mm in particle size.
(2) Structural change of sintering bed was also promoted when wustite particle were mixed with CaO particles. Extent of the structural change became larger with increase in the amount of CaO addition, i.e., from 5, 10 and 15 mass%CaO. This behavior seems to be caused by decrease in the liquidus temperature with increase in CaO concentration.
(3) In order to effectively utilize wustite containing materials as an agglomeration agent, it is essential to arrange sufficient amount of CaO component close to such particles for melt formation at lower temperature.
