Home Influence of Basicity and MgO on Fluidity and Desulfurization Ability of High Aluminum Slag
Article Open Access

Influence of Basicity and MgO on Fluidity and Desulfurization Ability of High Aluminum Slag

  • Ping Wang , Qing-min Meng EMAIL logo , Hong-ming Long and Jia-xin Li
Published/Copyright: September 17, 2015

Abstract

The viscosity of experimental slag, which was mixed based on the composition of a practical blast furnace slag, was measured in this paper. The influence of Al2O3 and MgO content, basicity R2 = w(CaO)/w(SiO2) on the fluidity of slag was studied. The stepwise regression analysis in SPSS was used to reveal the relationship between sulfur distribution coefficient LS and slag composition as well as furnace temperature. The results show that increasing of MgO up to 12% can decrease the slag viscosity. The w(MgO) should be controlled below 8% when there is 20% Al2O3 in the slag. Temperature of hot metal and content of CaO in slag are the two dominant factors on the desulfurization capacity of slag.

Introduction

Viscosity is an important parameter that indicates the liquidity of blast furnace (BF) slag. It influences the BF operation in many ways, such as desulfurization, coke consumption, and gas permeability in the furnace. In addition, melting temperature of the slag has a great influence on the BF slag fluidity [12]. When the slag temperature is below the melting point, the so-called short slag appears, which means the sedimentation of solid particles and a rapid increase of slag viscosity. Thus, to maintain BF slag in the liquid region is important to improve the fluidity of BF slag.

Many Chinese steel companies use iron ores from Australian and Indian, where the Al2O3 content is high: about 2.00% in Australian ore and 1.88% in Indian ore. By contrast, domestic ore contains generally less than 1% Al2O3. With the improvement of burden preparation technique and the increase of injection ratio of coal, the slag amount is increasingly reduced. These factors lead to a continuous increase of Al2O3 content in slag. In some companies, the Al2O3 content in slag reaches or exceeds 16%.

The fluidity and desulfurization capacity of high-aluminum slag have been investigated by numerous metallurgical scientists [36]. There are common opinions and controversies. The Kungliga Tekniska högskolan (KTH) model, which is developed by Swedish Royal Institute of Technology, predicts that the increase of Al2O3 improves the viscosity of BF slag. The study of Shen Fengman in China shows that viscosity of slag is in a low range and has a good stability [7], when the binary basicity(hereafter referred to as R2) of slag is within the normal range between 1.05 and 1.20 and Al2O3 content in the range of 7–15%.

Sulfur partition ratio is a common parameter to characterize the desulfurization capacity in BF iron-making process [8]. The study of Hu Xiayu in China shows that raising the content of MgO in high aluminum slag can increase the desulfurization capacity [9]. However, the thermodynamic model proposed by Shi Chengbin shows that the desulfurization capacity of free CaO is 97% at 1,500°C, while it is only 3% for free MgO [10].

In this paper, the viscosity–temperature curve was measured to investigate the influence of w(Al2O3), w(MgO), R2 and slag temperature on the fluidity of high aluminum slag. The relationship between sulfur distribution coefficient and slag composition as well as furnace temperature was analyzed through the stepwise regression method by means of SPSS program, in order to obtain key factors on the desulfurization capacity of high aluminum slag.

Investigation methods

Experimental and detection

Slag sample preparation

Practical slag from a 4,000 m3 BF in M steel was used as an experimental basis which has a composition shown in Table 1. With this composition, the influence of the contents of S, TiO2, FeO, and K2O on the slag fluidity can also be studied.

Table 1:

The composition of 4,000 m3 blast furnace slag in M steel.

CompositionSiO2Al2O3CaOMgOFeOTiO 2MnOK2ONa2OS
%32.4415.8338.218.940.3851.180.160.440.501.04

The BF slag was used as the main material to make up experimental slag, while the contents of CaO/SiO2, MgO, and Al2O3 were manipulated by adding chemical-grade reagents. The total weight percentage of SiO2, CaO, Al2O3, and MgO was controlled at 100%. Due to the high absorbent of CaO in air, Ca(OH)2 was used in the preparation of slag samples. The slag sample was heated to 1,530°C in a high-temperature furnace and hold for 30 min. Afterward it was put on an iron plate and rapidly cooled down and crushed.

Experimental equipment

Slag viscosity was measured with an RTW-08-type physical characteristics identifier, which is illustrated in Figure 1. The slag was heated to 1,500°C and hold for 30 min, then cooled down at a rate of 3°C/min.

Figure 1: Schematic diagram of experimental apparatus: 1, torque sensor; 2, rotating rods; 3, high temperature furnace; 4, graphite crucible; 5, probe; 6, thermocouple; 7, control cabinet; 8, computer system; 9, N2.
Figure 1:

Schematic diagram of experimental apparatus: 1, torque sensor; 2, rotating rods; 3, high temperature furnace; 4, graphite crucible; 5, probe; 6, thermocouple; 7, control cabinet; 8, computer system; 9, N2.

The temperature of molten slag cannot be measured simultaneously. The calibrating thermocouple was placed in the graphite crucible which was filled with high aluminum powders. Then the temperature was raised, stabilized, and decreased. The temperature difference was calculated from the two thermocouples that were located at the bottom and inside of the crucible. This value will be used for drawing the temperature–viscosity curve.

The data statistical analysis method

SPSS is currently the most widely used and internationally accepted statistical analysis software. The SPSS contains linear regression analysis and other statistical processes. Multiple stepwise regression analysis, which is one of the linear regression analyses, fits a set of variables. And variables are selected into the model basing on the significance of independent variables to the dependent variables. The variables in the model have significant effect on the dependent variable.

In the real slag, stepwise regression analysis was adopted to study the influence of each composition and the content of silicon in hot metal on the desulfurization capacity index LS = w(S)/w[S] to obtain the key influence factor of desulfurization capacity of high aluminum slag.

Experiments and field data investigation

Three principles of selecting components of the slag samples were followed: (1) The melting temperature of tapped slag should be below 1,400°C. Considering the influence of other components in the slag, this temperature limit can be raised to 1,420–1,430°C for a quaternary slag system. (2) When the slag temperature is below the normal operating range in BF, the viscosity of slag is preferably below 0.5 Pa s. In the worst case, it cannot exceed 2.0 Pa s. (3) The slag should have a good chemical stability (isothermal section of the diagram) and a good thermal stability (viscosity–temperature curve).

Single factor influence

The influence of w(Al2O3) on slag fluidity

Figure 2 illustrated the relationship between slag viscosity and temperature when R2 is 1.18, w(MgO) is 9.0% and w(Al2O3) is between 15.0% and 20.0%.

Figure 2: Effect of w(Al2O3) on viscosity of slag.
Figure 2:

Effect of w(Al2O3) on viscosity of slag.

Figure 3 shows the relationship between slag viscosity and w(Al2O3) at temperatures of 1,440°C, 1,470°C, and 1,500°C. The slag viscosity increases with the increase of w(Al2O3). Below 1,470°C, the viscosity is higher than 0.5 Pa s. Above 1,500°C, the viscosity is continuously decreased to 0.4 Pa s.

Figure 3: Effect of w(Al2O3) on viscosity of slag.
Figure 3:

Effect of w(Al2O3) on viscosity of slag.

The relationship between melting temperature and Al2O3 is shown in Figure 4. As seen in Figure 4, the melting temperature of slag increases with the increase of w(Al2O3). And when w(Al2O3) is greater than 18.0%, the melting temperature of slag increases sharply.

Figure 4: Effect of w(Al2O3) on Tm of slag.
Figure 4:

Effect of w(Al2O3) on Tm of slag.

The influence of w(MgO) on the fluidity of slag

Figure 5 shows the influence of MgO content and temperature on the slag viscosity with R2 of 1.18 and the content Al2O3 of 16%.

Figure 5: Effect of w(MgO) on viscosity of slag.
Figure 5:

Effect of w(MgO) on viscosity of slag.

Figure 6 shows the relationship between slag viscosity and w(MgO) at temperatures of 1,440°C, 1,470°C, and 1,500°C. The relationship between the melting temperature and w(MgO) is shown in Figure 7. From these two figures, it can be obtained that when the w(MgO) is in the range of 8–12%, the slag viscosity and the melting temperature increase with the content of w(MgO). However, when the w(MgO) is greater than 12%, the slag viscosity decreases. Therefore, w(MgO) should be controlled below 12%. As shown in Figure 8, the reason can be explained that increasing the MgO content of slag could improve the transform from silica composite anions (such as (Si6O18)12–, (Si4O12)8–, (Si3O9)6–) to melilite (such as (Si2O7)4–), which reduces the slag viscosity. However, when MgO content is higher than 12%, high melting point materials such as spinel (2,135°C) and periclase (2,800°C) are formed, increasing the slag viscosity.

Figure 6: Effect of w(MgO) on viscosity of slag.
Figure 6:

Effect of w(MgO) on viscosity of slag.

Figure 7: Effect of w(MgO) on Tm of slag.
Figure 7:

Effect of w(MgO) on Tm of slag.

Figure 8: CaO–SiO2–MgO–Al2O3 quadruple phase diagram (Al2O3% = 15%).
Figure 8:

CaO–SiO2–MgO–Al2O3 quadruple phase diagram (Al2O3% = 15%).

The influence of R2 on slag fluidity

When R2 = 1.12–1.25, w(MgO) = 8.5%, and w(Al2O3) = 16.5%, the relationship between slag viscosity and temperature is shown in Figure 9.

Figure 9: Effect of temperature on viscosity of slag.
Figure 9:

Effect of temperature on viscosity of slag.

The relationship between slag viscosity and R2 is shown in Figure 10. The slag has a minimum viscosity at an R2 of 1.15 in the range of 1.12–1.25. When R2 is greater than 1.20 and temperature above 1,470°C, the slag viscosity can also be controlled below 0.5 Pa s. The relationship between the melting temperature and R2 is shown in Figure 11. The melting temperature of slag increases with the increase of R2.

Figure 10: Effect of R2 on viscosity of slag.
Figure 10:

Effect of R2 on viscosity of slag.

Figure 11: Effect of R2 on Tm of slag.
Figure 11:

Effect of R2 on Tm of slag.

The fluidity of high aluminum slag

When w(Al2O3) is 20%, the relationship between slag viscosity and temperature is shown in Figure 12. The melting temperature and slag viscosity are shown in Table 2.

Figure 12: Effect of temperature on the viscosity of slag (w(Al2O3) = 20%).
Figure 12:

Effect of temperature on the viscosity of slag (w(Al2O3) = 20%).

Table 2:

Effect of MgO on the viscosity of slag.

Sample no.R2w(MgO)/%w(Al2O3)/%Viscosity (η/Pa s)Tm/°C
1,440/°C1,470/°C1,500/°C
161.304.020.00.740.500.401,385
171.256.020.00.620.480.391,381
181.258.020.00.700.510.421,400
191.208.020.00.630.500.451,375
201.2012.020.02.440.800.641,435
211.3012.020.01.841.210.981,450
221.008.020.01.471.140.95

The slags of sample nos 16, 17, 18, and 19 have a low viscosity. At 1,500°C, the viscosity is below 0.5 Pa s. At 1,440°C, the viscosity is below 0.8 Pa s. Melting temperature of no. 17 slag is 1,381°C. Melting temperature of no. 19 slag is 1,375°C.

The viscosity of nos 20 and 21 is greater than 0.60 Pa s, when the temperature is 1,500°C. The melting temperature of no. 20 is 1,435°C. And the melting temperature of no. 21 is 1,450°C; therefore, their chemical stability is poor. No. 22 slag (R2 equals 1.0) was acidic, whose viscosity–temperature curve is smooth and has no obvious turning point. But its viscosity is high, close to 1.0 Pa s at 1,500°C. In order to ensure a good flowability of this slag, a high-temperature operation should be adopted.

As a summary, w(MgO) in slag should be less than 12% for high aluminum slag, independent of R2. The w(MgO) should be controlled at less than 8% when Al2O3 is 20%.

Statistical analysis on the influence factors of slag actual desulfurization capacity

The BF slag and the composition of hot metal from no. 2 BF in M steel were taken as the research object. Multiple stepwise regression analysis method was used to analyze the relationship between the distribution coefficient of S(LS = w(S)/w[S]) and w(MgO), w(Al2O3), R2 (w(CaO)/w(SiO2)), R3 ([w(CaO) + w(MgO)]/w(SiO2)), R4 ([w(CaO) + w(Al2O3)]/[w(SiO2)+ w(MgO)]) of the slag, and w[Si] of the hot metal. Equation (1) was obtained as follows:

(1)LS=wS/wS=77.817+58.321×wSi+80.501×R2r=0.820

It can be obtained from eq. (1) that the temperature and the w[Si] of hot metal, and R2 of slag are the main influences on the slag desulfurization capacity. Other composition in the slag, such as MgO, has a relatively small influence, so they are weeded out in the stepwise multiple regression analysis. It is consistent with the results of thermodynamic calculation in a paper [10].

The high content of CaO will lead to increases of viscosity and melting temperature of the slag. For a high aluminum slag, if the melting point of slag is close to the discharging temperature of slag, the risk of the slag and iron clogging is raised, and even hearth freeze accident will occur.

Discussion of an optimal high aluminum slag system

Three kinds of slag systems are used worldwide. One is a high basicity (R2 = 1.20–1.25) with a low MgO (4–7.0%) slag system; the second is a medium basicity (R2 = 1.12–1.18) with a medium to high MgO content (8–10%); the last one is a low basicity (R2 = 1.00–1.10) with high MgO (11–13%) slag system.

The high basicity (R2 = 1.20–1.25) and low MgO (4–7%) slag system is short and has a poor stability. Chinese Baosteel and Japanese giant BF (both the BF effective volumes are larger than 4,000 m3) adopted this slag system. This slag system requires a low slag ratio and stable raw material and fuel. The temperature of hot metal can be raised to above 1,490°C with a smooth furnace operation. As we concluded from our experiment, the slag temperature is the most important factor on the slag fluidity. When the slag temperature is high than 1,470°C, the slag viscosity can be controlled below 0.5 Pa s, as demonstrated by no. 16 and 17 slag. When the slag ratio is below 300 kg/tHM, the high basicity and low MgO slag system should be adopted besides a high slag temperature. R2 should be controlled between 1.20 and 1.30 and w(MgO) below 8%.

In China, the compositions of raw materials and fuel fluctuate in a wide range. Meanwhile, the amount of slag is also high (the slag ratio is 320–400 kg/tHM), and the w(Al2O3) is in the range of 14–17%. Therefore, the medium basicity (1.10–1.17) with high to medium MgO (12%) slag system, such as nos 11 and 12 slag, is widely used. This slag system can ensure a hot metal with a low silicon and a low sulfur, when the hot metal temperature is higher than 1,460°C. This slag system should also meet the principle of R2 above 1.15, where the viscosity is minimized, and w(MgO) does not exceed 12%. At the same time, when the content of MgO is low, the binary basicity should be increased properly, vice versa.

In the former Soviet Union, the raw materials of BF are all from domestic, the amount of BF slag is generally large and w(Al2O3) is between 7% and 10%. In order to improve the slag fluidity, the slag system of higher w(MgO) (11–13%) and the lower basicity (R2 ≤ 1.0) was hence adopted. In North America, some BFs, which use the high proportion of pellets, also adopt this slag system. This slag system has a wider liquid-phase region, and hence it is stable and has low viscosity. But its desulfurization performance is poorer than the others.

Conclusions

  1. The increase of w(Al2O3) has an obvious effect on the viscosity of high aluminum slag and the melting temperature. The increase of w(MgO) plays a role in reducing the viscosity of the high aluminum slag and the melting temperature. When the R2 is 1.15, the slag viscosity is minimum. The rising of R2 increases the melting temperature of the slag. When the slag temperature is greater than 1,500°C, the viscosity of most slag systems can be controlled below 0.5 Pa s.

  2. When the w(Al2O3) is 20% in the BF slag, the w(MgO) should not be higher than 12%, independent of R2. The basicity should be adjusted to between 1.20 and 1.30, and the w(MgO) should be reduced to below 8%. The slag system with high basicity and high MgO is not suitable for a high aluminum slag. When w(MgO) is low, the basicity should be increased, vice versa.

  3. The temperature of hot metal, w[Si], and w(CaO) are the main factors determining the slag desulfurization capacity. MgO has a small influence. Meanwhile, maintaining a higher and stable temperature of hot metal is an important guarantee for a smooth operation for BF and also for a high slag desulfurization capacity.

Funding statement: Funding: We gratefully acknowledge the financial support for this work by the Foundation of National Natural Science Foundation of China (project no. 51474002).

References

[1] D. Ghosh, V.A. Krishnamurthy and S.R. Sankaranarayanan, J. Min. Metall. Sect. B-Metall., 46 (1B) (2010) 41–49.10.2298/JMMB1001041GSearch in Google Scholar

[2] K. Sunahara, K. Nakano and M. Hoshi, ISIJ Int., 48 (4) (2008) 420–429.10.2355/isijinternational.48.420Search in Google Scholar

[3] R.K. Joo, S.L. Young and J.M. Dong et al., ISIJ Int., 44 (8) (2004) 1291–1297.10.2355/isijinternational.44.1291Search in Google Scholar

[4] H.S. Seong, M.J. Sung and S.L. Young et al., ISIJ Int., 47 (8) (2007) 1090–1096.10.2355/isijinternational.47.1090Search in Google Scholar

[5] S.C. Du, J. Bygden and S.A. Seetharaman, Met. Mat. Trans. B., 25 (4) (1994) 519–525.10.1007/BF02650073Search in Google Scholar

[6] F.M. Shen, Angang Technol., 336 (6) (2005) 1–4.Search in Google Scholar

[7] A. Shankar, Ironmaking Steelmaking, 33 (5) (2006) 413–418.10.1179/174328106X113968Search in Google Scholar

[8] X.Y. Hu, L.Q. Lin and L.S. Li, China Metall., 16 (2) (2006) 36–41.Search in Google Scholar

[9] C.B. Shi, H.J. Guo and R.C. Ding et al., Chinese J. Process Eng., 10 (1) (2010) 159–164.Search in Google Scholar

[10] K.C. Mills, L. Chapman and A.B. Fox et al., Scand. J. Metall., 30 (6) (2001) 396–403.10.1034/j.1600-0692.2001.300608.xSearch in Google Scholar

Received: 2015-1-6
Accepted: 2015-7-6
Published Online: 2015-9-17
Published in Print: 2016-8-1

©2017 by De Gruyter

This article is distributed under the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Downloaded on 11.9.2025 from https://www.degruyterbrill.com/document/doi/10.1515/htmp-2015-0006/html
Scroll to top button