2020 Volume 60 Issue 7 Pages 1395-1400
Recently, demands for reduction of CO2 gas emission in iron making process are increasing. For further reduction of CO2 gas emission, a method of capturing carbon dioxide from blast furnace exhaust gas has been studied. The oxygen blast furnace using pure oxygen for blast does not contain nitrogen in the exhaust gas and that is more advantageous than conventional blast furnace in the point of view of CO2 separation energy. Although the oxygen blast furnace has been studied with the experimental furnace, the experimental furnace was not sufficiently investigated on the properties of cohesive layer because of the small load of burden materials. Therefore, in this study, the properties of cohesive layer of the oxygen blast furnace were studied.
The properties of the cohesive layer were evaluated under blast furnace conditions and nitrogen-less conditions in a load-softening test. As a result, the properties of cohesive layer were remarkably improved in nitrogen-less conditions. As a result of discussion, improvement of the properties of cohesive layer was quantitatively explained due to suppression of contraction of the sintered ore and the decrease of slag liquid because the nitrogen-less atmosphere promoted reduction reaction.
In recent years, improvement of productivity and energy efficiency at steel works has become a social obligation due to the growing demand for reduction of CO2 gas emissions as a countermeasure against global warming. Because the iron making process consumes a large amount of coal not only to reduce and smelt iron ore, but also to supply energy to the steel works, demand for reduced CO2 in the iron making process is particularly high. In the blast furnace process, sintered ore and coke are charged from the top of the blast furnace and hot air is blown into the lower part of the furnace to achieve high energy efficiency and high productivity by a countercurrent moving bed type reaction. So far, in addition to developing technologies such as pressurizing operation, auxiliary reducing agent injection, burden distribution control, etc.,1,2,3,4,5) upsizing the inner volume of the blast furnace1) enhanced productivity and reduced heat loss for a given amount of production. As a result, blast furnace heat loss no longer exceeds 10% of input energy, and the margin available for reduction of heat loss is also slight.6)
Under these circumstances, technologies that use the hydrogen contained in coke oven gas and that separate and capture CO2 from blast furnace gas7) are being studied as further efforts to reduce CO2 emissions. Although capture and underground storage of CO2 or conversion into a chemical product have been proposed, a large amount of energy is required for CO2 separation due to the low CO2 concentration of blast furnace gas. On this point, in the case of an oxygen blast furnace using pure oxygen for blowing, it is not necessary to separate nitrogen, and improvement of the reduction rate and productivity is also expected due to the increase in the concentration of CO in the furnace. In the 1980s, some studies were conducted on the oxygen blast furnace, including experimental blast furnace operation,8,9) but the gas permeability of the cohesive layer was not evaluated in the experimental oxygen blast furnace because the burden load was small and the formation of the cohesive layer was insufficient. The permeability of the cohesive layer, which account for 1/3 of the pressure drop in the furnace, is important for stable descent of burden materials in commercial blast furnaces.
Therefore, in this study, the effect of the nitrogen-free reducing atmosphere assumed for an oxygen blast furnace on the properties of the cohesive layer was evaluated.
In blast furnace operation, a small pressure drop is preferable because trouble such as blow-through and unfavorable descent of burden materials tend to occur when the pressure drop in the furnace becomes large. When the temperature inside the furnace reaches a high temperature of about 1200°C, the ore deforms due to the load of the burden materials and softening of the ore itself, the porosity of the ore layer decreases, and as a result, pressure drop increases. In the blast furnace, ore and coke are charged in layers, so there is a cohesive zone consisting of a deformed ore layer (cohesive layer) and a coke slit in the high temperature area inside the blast furnace. Although the properties of the cohesive zone are determined by the properties of both the cohesive layer and the coke slit, the experiments and discussion in the present study focus on the cohesive layer.
In this test, a load-softening test was carried out in order to evaluate the properties of the cohesive layer in the blast furnace. A schematic diagram of the experimental apparatus and the method of charging the sample are shown in Fig. 1. A sintered ore and coke were charged in layers in a graphite crucible having a diameter of ϕ50 mm, and a reducing gas was flowed from the lower part of the crucible while applying a load to the upper part of the sample with a push rod. During the test, the furnace temperature was controlled with a heater. In the lower part of the crucible, 13 holes with diameters of ϕ5 mm are provided for gas passage and melt dripping. In this study, dripping was confirmed in all cases. Figure 2 and Table 1 show the time-dependent changes in the load, gas composition and temperature used in the experiment. These time changes were set by simulating the situation in which the burden materials descend from the upper part of the blast furnace to the cohesive zone in an actual furnace. The experiments were carried out under two conditions, one simulating conventional blast furnace and the other the nitrogen-less condition assuming an oxygen blast furnace. In order to investigate the influence of changing the atmosphere to a nitrogen-less one in an oxygen blast furnace, the temperature and load pattern under the nitrogen-less condition were set to be the same as those in the conventional blast furnace condition, and in the gas pattern, nitrogen was removed from the blast furnace condition. Since the gas flow rate greatly affects pressure drop, the flow rates of CO and CO2 under the nitrogen-less condition were adjusted so that the amount of flowing gas was equal to 6.0×10−3 Nm3/min under both the blast furnace condition and the nitrogen-less condition.
(a) Schematic drawing of testing system. (b) Schematic view of sample in crucible.
Experimental conditions of temperature, load, and gas composition. (Online version in color.)
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/60_1395_tbl_1.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
|
In the load-softening test, the pressure drop of the ore layer and the temperature change of the shrinkage rate were measured. The pressure drop was measured from the difference between the pressure of the inlet side gas and the atmospheric pressure. The shrinkage of the ore layer was measured from the displacement of the push rod during the experiment, assuming that the coke layer does not shrink as the temperature increases. In this study, 80 g of sintered ore was charged in a crucible, and 10 g of coke was charged above and below the sintered ore. The sample height of the upper and lower coke layers was 10 mm, and that of the sintered ore was 30 mm. Table 2 shows the sample and gas conditions used in the experiment, and Table 3 shows the chemical composition of the sintered ore used in the experiment. Since the reduction rate of the sintered ore during the experiment was difficult to measure by gas analysis, the interrupted experiment was conducted and the reduction rate was measured appropriately by chemical analysis.
| Sinter | Diameter | mm | 4.75–6.5 |
| Weight | g | 80 | |
| Coke | Diameter | mm | 9–11 |
| Weight | g | 20 | |
| Crucible | Diameter | mm | 50 |
| Height | mm | 80 | |
| Material | – | Carbon | |
| Gas | Composition | – | CO, CO2, N2 |
| Volume | Nm3/min | 6.0×10−3 |
| Compositions (mass%) | ||||||
|---|---|---|---|---|---|---|
| T.Fe | M.Fe | FeO | SiO2 | CaO | Al2O3 | MgO |
| 57.3 | 0.0 | 7.0 | 5.1 | 10.8 | 1.7 | 0.9 |
A load-softening test was carried out under the blast furnace condition and nitrogen-less condition, and the pressure drop and change in the shrinkage ratio of the ore layer were measured.
The change in pressure drop during the experiment is shown in Fig. 3. The horizontal axis represents the estimated temperature in the crucible and the vertical axis represents the pressure drop. The estimated temperature in the crucible was obtained by using the relationship between the control temperature measured in another experiment and the temperature in the crucible. The solid line shows the measurement result for the blast furnace condition, and the dotted line is the result for the nitrogen-less condition. Under the blast furnace condition, several interruptions occurred in the experiment because the pressure drop reached the design upper limit of 20 kPa in the middle of the experiment. In this test apparatus, the charged state of the sample may have some influence on the experimental results, especially when the pressure drop is large. However, regardless of whether or not the pressure drop reached the design upper limit, the data for tests which could be conducted to the end were adopted since the reproducibility of the temperature dependency of the pressure drop was high. Under both conditions, pressure drop trended at about 0.2 kPa from the start of the experiment, but then started to rise after 1200°C, peaked at about 1350°C and decreased thereafter. Since the pressure drop at the start of the experiment depends on the charging state of the sample, it varied by about 0.05 kPa in each experiment. The pressure drop increased to 19 kPa or more under the blast furnace condition, but in contrast, it increased only to 2 kPa under the nitrogen-less condition. Thus, the maximum pressure drop of the cohesive layer decreased greatly due to the nitrogen-less condition in the reducing atmosphere.
Measurement results of pressure drop with temperature in blast furnace condition and N2-less condition.
Here, the temperature at which the pressure drop increased by 0.1 kPa or more, which was sufficiently larger than the variation depending on the charging state, was defined as the pressure drop starting temperature Ti (°C). Ti was 1220°C under the blast furnace condition and 1266°C under the nitrogen-less condition. In the blast furnace, an increase in Ti means that the property of the cohesive layer improves because the thickness of the cohesive zone becomes thinner and total pressure drop is reduced.
Figure 4 shows that relationship between the temperature in the crucible and the contraction rate of the ore layer during the experiment. The contraction rate of the ore layer was indicated based on the ore bed height at 900°C. Although both conditions showed almost the same contraction rate up to around 1200°C, after that temperature the contraction rate of the nitrogen-less condition was smaller than that of the blast furnace condition. This confirmed the fact that the contraction rate was suppressed by the nitrogen-less atmosphere.
Measurement results of contraction rate with temperature in blast furnace condition and N2-less condition.
Numerous papers on the load-softening test have been reported so far. For example, Kokubu et al. reported that addition of hydrogen to reducing gas promoted ore reduction, and Ti increased and the maximum pressure drop decreased.10) Nishimura et al. suggested that when the reduction degree of ore is low, a slag liquid composed mainly of FeO is formed in the voids of the packed ore bed, and as a result pressure drop may increase.11) According to these past findings, it is estimated that the change in the properties of the cohesive layer due to the nitrogen-less atmosphere in this report was also due to promotion of the reduction of the sintered ore.
In order to confirm this promotion of ore reduction, the experiment was interrupted at 1350°C, which showed the maximum pressure drop in the load-softening test, and the reduction degree of the interrupted sample was measured by chemical analysis. As a result, the reduction degree of the sintered ore was 88% under the blast furnace condition and 93% under the nitrogen-less condition. This confirmed that reduction of ore was promoted under the nitrogen-less condition.
4.2. Experiment with Pre-reduced Sintered OreThe influence of the reduction degree of the sintered ore on the properties of the cohesive layer was investigated by a load-softening test using pre-reduced sintered ore. The pre-reduced samples were prepared by raising the temperature to 900°C in an Ar atmosphere, holding the sample at that temperature for 30 min, and then reducing the sample with CO/N2 = 50%/50% gas at a constant temperature until reaching the target weight. The reduction degrees of each sample were finally determined by chemical analysis. Samples adjusted to several different reduction degrees in this way were subjected to a load-softening test under an inert atmosphere, and the temperature change of the pressure drop and the contraction rate was measured. The temperature and load pattern were the same as those already described. The chemical compositions of the pre-reduced sintered ore samples are shown in Table 4. The sintered ore was the same as that used in Chapter 3.
| Composition (mass%) | |||
|---|---|---|---|
| R | Fe2O3 | FeO | M.Fe |
| 64 | 9.0 | 34.6 | 34.2 |
| 77 | 5.2 | 23.9 | 47.5 |
| 99 | 0.6 | 1.2 | 72.0 |
Figure 5 shows the temperature change of the pressure drop during the load-softening test for samples with pre-reduction degree R = 64%, 77%, and 99%. As the pre-reduction degree increased, Ti increased and the maximum pressure drop decreased. This is consistent with the trend of the load-softening test results under the blast furnace condition and nitrogen-less condition. This result supports the speculation that the change in the cohesive zone properties under the nitrogen-less condition is mainly due to promotion of the reduction of the sintered ore.
Measurement results of pressure drop with temperature using pre-reduction sinter.
Figure 6 shows the temperature change of the ore layer contraction rate in the load-softening test using the pre-reduced samples. The contraction rate of the sintered ore layer is suppressed by an increase in the reduction degree. Contraction started rapidly at 1100°C with the samples with pre-reduction degree R = 64% and 77%, and difference between the samples widened after 1200°C. The sample with the pre-reduction degree R = 99% behaved differently from the other samples, and contraction began sharply from 1300°C. Although it is considered that the reduction degree dependence of contraction is strongly related to the softening viscosity of the phase in the sintered ore, this was not formulated in the present experiment.
Measurement results of contraction rate with temperature using pre-reduction sinter.
Next, we attempted to quantitatively evaluate the relationship between the reduction degree and the properties of the cohesive layer. According to Ichikawa et al., the pressure drop of the cohesive layer is expressed by the following equation taking into account the effects of ore contraction and liquid slag formation.12)
| (1) |
Assuming that the initial porosity is ε0 and the volume of the ore layer does not change, the porosity of the ore layer after contraction is expressed by Eq. (2).
| (2) |
Further, ht can be obtained from the liquid phase density and the amount of liquid slag. The liquid phase density ρL (kg/m3) was defined as the slag density of the multicomponent system at 1400°C as a function of the weight fraction of FeO, as shown in Eq. (3).14)
| (3) |
The composition used for the calculation was prepared based on the results of a chemical analysis of Fe2O3, FeO, SiO2, Al2O3, MgO, and M. Fe in the sintered ore adjusted to several reduction degrees. The chemical compositions are shown in Table 5. As the reduction degree increases, the proportion of FeO in the sintered ore decreases. Figure 7 shows the relationship between the reduction degree of the sintered ore and the amount of liquid slag for each temperature. It can be understood that the amount of liquid slag decreases as the reduction degree increases at temperatures of 1300°C or higher.
| Reduction degree (%) | Compositions of sinter (mass%) | ||||||
|---|---|---|---|---|---|---|---|
| Fe2O3 | FeO | SiO2 | CaO | Al2O3 | MgO | M.Fe | |
| 40 | 12.9 | 55.4 | 5.8 | 12.2 | 1.9 | 1.0 | 10.8 |
| 50 | 11.1 | 46.8 | 6.0 | 12.5 | 2.0 | 1.1 | 20.6 |
| 60 | 9.1 | 38.5 | 6.1 | 12.9 | 2.0 | 1.1 | 30.3 |
| 70 | 7.0 | 29.7 | 6.3 | 13.3 | 2.1 | 1.1 | 40.5 |
| 80 | 4.8 | 20.4 | 6.5 | 13.7 | 2.1 | 1.2 | 51.3 |
| 90 | 2.5 | 10.5 | 6.7 | 14.1 | 2.2 | 1.2 | 62.8 |
| 100 | 0.0 | 0.0 | 6.9 | 14.5 | 2.3 | 1.3 | 75.0 |
Relationship between the amount of liquid slag and reduction degree of sinter. (Online version in color.)
Table 6 shows the actual measured values of the maximum pressure drop under both conditions and the calculated values according to Eq. (1). The contraction rate was obtained from experimental values. It can be understood that the experimental and calculated values of the maximum pressure drop under the nitrogen-less condition are in good agreement.
| BF condition | N2-less condition | ||
|---|---|---|---|
| Maximum pressure drop | Measured | > 19.0 kPa | 2.0 kPa |
| Calculated by Eq. (1) | (flooding) | 1.6 kPa |
On the other hand, it was assumed that flooding occurred under the blast furnace condition because ε-ht<0, and agreement between the experimental and calculated value was not found. However, this was consistent with the fact that the pressure drop exceeded the upper limit of the apparatus under the blast furnace condition and the experiment was interrupted several times.
Figure 8 shows Ti in the load-softening test using the pre-reduced samples. The measured values are represented by × in figure, the values calculated from only the contraction rate of the ore layer are indicated by ○, and the values calculated from both the contraction rate and the amount of liquid slag are indicated by ●.
Relationship between Ti and reduction degree of sinter (Measured and calculated). (Online version in color.)
The horizontal axis represents the pre-reduction degree of the sample. There was no correlation between the reduction degree and Ti when calculated from only contraction the rate. However, when both the contraction rate and the amount of the liquid slag were considered, the tendency that Ti rises as the reduction degree increases was obtained. Although the absolute value deviates slightly in the actual measurement and calculation results, this is conceivably due to an increase in the apparent contraction rate of the ore layer owing to softened sintered ore entering the coke layer, or the fact that a uniform composition was used in the calculation of the amount of liquid slag.
From above discussion, it was concluded that the decrease of the maximum pressure drop and the increase of Ti under the nitrogen-less condition were mainly due to suppression of contraction of the ore layer and reduction of the amount of the liquid slag by promotion of the ore reduction reaction.
The influence on the gas permeability of the cohesive zone due to a nitrogen-less atmosphere was evaluated by a load-softening test simulating the condition inside the blast furnace. The results of the test using pre-reduced samples were discussed, and the following conclusions were obtained.
(1) From the results of the load-softening test, it was confirmed that the pressure drop starting temperature Ti increased and the maximum pressure drop greatly decreased under the nitrogen-less condition compared to the conventional blast furnace condition.
(2) Improvement of the properties of the cohesive layer was quantitatively explained by suppression of contraction of the sintered ore and a decrease in liquid slag because the nitrogen-less atmosphere promoted the ore reduction reaction.
