Study on the Reblow Model for Medium-High Carbon Steel Melting by Converter

Introduction

BOF (Basic Oxygen Furnace) steelmaking process aims to acquire the liquid steel with qualified temperature and composition. Therefore, the restrictive link of tapping is whether the endpoint temperature and composition meet the demand of direct tapping [1, 2, 3, 4]. Generally speaking, the endpoint control of medium-high carbon steel melting is relatively difficult. Therefore, in order to control the endpoint temperature and composition of medium-high carbon steel effectively, large converters usually employ the dynamic monitoring equipment (such as sublance [5, 6, 7], gas analysis system [8, 9, 10] and so on) to test the temperature and composition of liquid steel at the end of main blow and adjust the amount of oxygen consumption and coolant based on the test values. However, due to the limitation of furnace mouth, small- and middle-capacity converters could not equip with sublance, and most of small- and middle-capacity converters in China do not employ the gas analysis system [11, 12]. For these kinds of converters which lack of dynamic monitoring devices, some employ bombing detection technology [13, 14], and others still rely on the artificial experience to achieve high hit rate of endpoint temperature and composition for medium-high carbon steel.

Only if endpoint temperature and component of molten steel do not approach the target value after the blowing process, the reblow operation is normally applied to adjust the temperature and composition of liquid steel [15]. As a result of the limitation of fund and technology, most of small- and middle-capacity converters in China could not equip with dynamic monitoring devices or apply bombing detection technology, they still rely on the experience to predict the temperature and composition of molten steel, judge whether the reblow process is needed and determine the reblow time and the amount of coolant added into the furnace based on the turndown temperature and component of liquid steel. As is known to all, this method is inefficient and greatly affects the productive efficiency of medium-high carbon steel melting. For the sake of reducing the reblow times and achieving high hit rate of endpoint temperature and composition for medium-high carbon steel melting, this paper proposed an effective method for precisely controlling the reblow operation. Through carrying out the study on the variation of component and temperature of liquid steel during reblow process, the reblow model was established and applied to the medium-high carbon steel melting of a special steel plant.

Decarburization and temperature increase during reblow process

As is widely accepted by the metallurgist, the decarburization reaction during BOF steelmaking process could be divided into three stages [16, 17]. The variation of decarburization rate during the oxygen blowing process is shown in Figure 1. During the reblow process, the carbon content in liquid steel is reduced to a relatively low level, and the decarburization rate decreases with the decline of carbon content. At this stage, the decarburization rate is controlled by mass transfer of carbon, and the supplied oxygen would mostly react with iron to form iron oxide. The decarburization rate during the reblow process could be expressed as follows [18]:

Figure 1:

Classical three-stage decarburization theory for BOF steelmaking process.

where −dC/dt is the decarburization rate, %⋅s−1; k is the constant coefficient, s−1; t is the oxygen blowing time, s; C is the carbon content in liquid steel, %, and Ck is the lowest carbon content in steel bath, %. Equation (1) was separated variables and integrated on both side, then eq. (2) was obtained. After the simplification of eq. (2), eq. (3) was obtained.

where Ce is endpoint carbon content, %; Cd is turndown carbon content, % and t is the reblow time, s. The change of carbon content with the reblow time during reblow process could be revealed as eq. (3), and the carbon content decreases exponentially with the reblow time. For the medium-high carbon steel melting, eq. (3) could be used to calculate the carbon content during reblow process based on turndown carbon content; thus it could provide a good reference for operators to judge BOF endpoint accurately. In order to solve the constant coefficient k in eq. (3), the authors collected production data of an 80t BOF reblow process from one special steel plant, and the constant coefficient k was 0.0105 after the calculation with 659 heat data.

During BOF reblow process, the supplied oxygen mainly reacts with carbon and iron in liquid steel, and the heat generated by the oxidation reaction promotes the temperature increase of molten steel. Furthermore, the heat balance of the converter reblow process was analysed and eq. (4) was obtained.

where Qreaction is the heat generated by the oxidation reaction, kJ; Qsteel, Qslag and Qcoolant are the heat absorbed by molten steel, slag and coolant, respectively, kJ and Qloss is the heat dissipated by the gas, furnace body and so on, kJ. Ping He et al. [15] found that there was an obvious linear relation between oxygen consumption and temperature increase during reblow process, if the charged hot metal, scrap, pig iron keep stable and iron ore were not added into the furnace. Equation (5) showed this relation, where ΔT is the increased temperature of liquid steel, °C; ΔO2 is the amount of supplied oxygen, Nm³; Wst is the weight of molten steel, t; b1 is the temperature coefficient, °C · t/Nm³ and b is the temperature drop caused by heat loss of the converter, °C.

Considering that the oxygen consumption is a function of oxygen blowing time during the reblow process. The relation between temperature increase caused by oxidation reaction and reblow time was assumed to be linear in this paper. Then, eq. (5) could be transformed into eq. (6) by taking into account the effect of coolants simultaneously. Just as eq. (6) showed, λ is the heating rate of the molten steel, °C/s; t is the reblow time, s; mi is the weight of coolant i that is added into the furnace, kg and θi is the cooling effect of coolant i, °C/kg.

As for the neighbouring heats, Qloss in eq. (4) could be considered to be approximately the same value, and the temperature drop b could be regarded as constant. Moreover, the value of λ and θi could be fixed on the basis of previous production data and artificial experience. Similarly, the authors collected 659 heats data of an 80t BOF from one special steel plant and found that only the sintered ore was added into the furnace to lower the temperature during the reblow process. According to the calculation of the production data, the value of λ and θ was 0.98 and 0.078, respectively. Thus, the temperature of liquid steel after the reblow operation could be estimated with eq. (6).

Variation of phosphorus content during BOF reblow process

Removal of phosphorus is a reaction of utmost importance during BOF steelmaking process, and low temperature of liquid steel, high slag basicity and high FeO content would promote the dephosphorization process [19]. With aims to study the variation of phosphorus content during BOF reblow process intensively, the authors applied the data mining method to analyse the collected 659 heat data, and found an apparently nonlinear relation between endpoint phosphorus content, turndown phosphorus content and reblow time, which was presented in Figure 2. SPSS (Statistical Product and Service Solutions) is one of the most popular statistical packages available that can be used to perform data entry and analysis. Moreover, SPSS is so powerful that it is capable of handling large amounts of data, and can perform most of the analyses covered in the test. Based on above consideration, SPSS was applied to complete the fitting between endpoint phosphorus content, turndown phosphorus content and reblow time in this paper, and eq. (7) was obtained. In the evaluation of properties of the fitting, the correlation coefficient R was used in this paper. Generally speaking, when the value of correlation coefficient R is greater than 0.5, the relation of the model is considered to be acceptable, and the closer to 1 the correlation coefficient R is, the more obvious the relation is. Besides, the correlation coefficient R of this equation was 0.706, and the deduced relation of this equation could be regarded as obvious.

Figure 2:

Relation between endpoint phosphorus content, turndown phosphorus content and reblow time.

where Pe is endpoint phosphorus content, %; Pd is turndown phosphorus content, % and t is the reblow time, s. Based on the turndown phosphorus content and reblow time, the phosphorus content of liquid steel after the reblow operation could be calculated with eq. (7), and it could provide reference for the operators to accurately judge whether the phosphorus content is qualified.

Estimation of endpoint oxygen content in liquid steel

Endpoint oxygen content in molten steel greatly affects the yield of alloy and quality of liquid steel. With the growing demand of high performance steel, fine control of BOF endpoint oxygen content becomes more and more important. Therefore, this paper presented the estimation of BOF endpoint oxygen content. When the effect of bottom blowing maintains at a high level, obvious carbon–oxygen equilibrium in liquid steel was showed on the condition that the carbon content was decreased to a low level [20]. Generally speaking, the decarburization reaction of BOF steelmaking could be expressed by eq. (8).

Equation (9) showed the expression of thermodynamic equilibrium constant K.

where PCO is the partial pressure of CO, Pa; Pθ is the standard atmospheric pressure, Pa; T is temperature of liquid steel, K; fC and fO are activity coefficients of carbon and oxygen in liquid steel; [%C] and [%O] are mass percentages of carbon and oxygen in molten steel. It was assumed that the decarburization reaction in molten bath approached thermodynamic equilibrium after the reblow operation. Then, activity coefficients of carbon and oxygen could be simplified as fC≈fO≈1. Meanwhile, the partial pressure of CO was supposed to be equal to the standard atmospheric pressure. Based on above conditions, eq. (9) could be simplified as follows:

Once the endpoint carbon content and temperature of liquid steel was acquired, the endpoint oxygen content could be calculated by eq. (10). According to the actual production of the special steel plant, the endpoint carbon content and temperature of medium-high carbon steel melting were mostly controlled within 0.10 % ~ 0.20 % and 1,640°C ~ 1,660°C, respectively. Figure 3 showed the calculated oxygen content of the molten steel when the endpoint temperature was 1,640°C. It can be seen from Figure 3 that there was an inverse relationship between the endpoint carbon content and oxygen content. Especially when the endpoint carbon content of liquid steel increased from 0.10 % to 0.20 %, the oxygen content of molten steel decreased from approximate 250×10^–6 to 125×10^–6. Meanwhile, according to eq. (10), it could be concluded that the oxygen content gradually added with the rise of endpoint temperature when the carbon content was fixed.

Figure 3:

Relation between endpoint carbon content and oxygen content.

Establishment and application of the reblow model

With the purpose of providing a systematic guidance for BOF reblow process, this paper proposed the reblow model on the basis of eq. (3), eq. (6), eq. (7) and eq. (10). Equation (11) showed the expression of the reblow model.

In order to help operators accurately control the reblow process and assure the endpoint composition and temperature of liquid steel all achieve the target values, the authors intended to provide an accurate solution for reblow process of medium-high carbon steel melting according to eq. (11). Specifically, the solution for reblow process included the reblow time and the amount of coolant added into the furnace. Just as Figure 4 showed, the principle of determination of reblow time was pursuing the shortest possible time on condition that the endpoint composition and temperature of molten steel achieved the target values. Meanwhile, the principle of calculation of coolant was seeking for the minimum coolant addition only if the endpoint temperature of liquid steel met the requirement for direct tapping.

Figure 4:

Calculation diagram of the solution for reblow process.

With the assistance of reblow model proposed in this paper, it can easily judge whether the turndown composition and temperature of molten steel meet the requirements for direct tapping and determine the necessity of reblow operation. If the judgment indicates the heat does not need the reblow operation, then the reblow model will suggest the operators to directly tap. If the judgment implies the heat needs the reblow operation, then the reblow model will provide the corresponding solution for reblow process (reblow time and amount of added coolant). In a word, the reblow model can not only provide solution for reblow operation but also calculate the composition and temperature of liquid steel after the suggested reblow operation, which would save the time spent on sampling and analysis of endpoint molten steel, provide reference for alloying process and improve production efficiency.

For the sake of generating the real-time solution for BOF reblow operation of medium-high carbon steel melting, the authors developed the software for reblow process with C# program on the basis of reblow model and actual BOF production technology. Figure 5 is the interface of the software for BOF reblow operation. As is showed in Figure 5, only if the turndown composition and temperature are acquired, the software will provide real-time solution for reblow operation and calculate the composition and temperature of liquid steel after the suggested reblow operation.

Figure 5:

The interface of the software for BOF reblow operation.

When using this software for actual production, the steel grade and its corresponding requirements for directly tapping should be fixed at first. Second, the turndown composition and temperature of molten steel should be inputted into the software. Finally, click the “Calculate” button and the software would automatically judge whether the obtained composition and temperature meet the requirements for direct tapping and determine the necessity of reblow operation. In order to validate the effect of the reblow model, the authors applied the software to the reblow process of an 80t BOF from a special steel plant, and the results showed that the solution for reblow operation provided by the software was accurate and effective. Table 1 presented the partial results for medium-high carbon steel melting, and some technological parameters about the 80t BOF are listed in Table 2. As can be seen from Table 1, not only the actual endpoint composition and temperature after the suggested reblow operation met the requirements for direct tapping, but also calculated endpoint composition and temperature were almost consistent with the actual values. Specifically, the calculated errors were mainly within ±0.02 %, ±0.002 % and ±5°C, respectively, when the reblow model was used to calculate endpoint carbon content, phosphorus content and temperature.

Along with the stable operation of the software for one year, some technical and economic efficiency were achieved for the 80t BOF steelmaking process. Due to the auxiliary guidance for the reblow process, there was an obvious increase for the reblow efficiency for medium-high carbon steel melting, which was showed in Figure 6.

Figure 6:

Change of reblow rate with application of the reblow model.

As it is demonstrated in Figure 6, with the application of the reblow model, the reblow rate (reblow more than once) dropped from 3.99 % to 0, while the reblow rate (reblow once) rose from 69.45 % to 71.35 %. Moreover, the average tap to tap time for medium-high carbon steel melting by converter was reduced from 37.43 min to 36.35 min. Although the reblow model was established based on the reblow operation of the converters which do not equip with dynamic monitoring devices, it could also be applied to the BOF with sublance system. According to the component and temperature tested by sublance system, the reblow model would propose solution for the following operation. Thus, the reblow model could provide a good reference for BOF steelmaking process.

Conclusions

This paper carried out the study on decarburization reaction, temperature increase, variation of phosphorus and oxygen content of molten steel during BOF reblow process, and the main conclusions were summarized as follows.

1. Based on the classical three-stage decarburization theory and thermal equilibrium theory, the decarburization reaction and temperature increase were discussed, and the models for the variation of carbon content and temperature during BOF reblow process were built.

2. The data mining method was used to explore the variation of phosphorus and oxygen content during reblow process, and the models for change of phosphorus and oxygen content were established based on production data and thermodynamic theory.

3. The software for BOF reblow process was developed, continuous operation of the software indicated that the reblow rate (reblow more than once) dropped from 3.99 % to 0 and the average tap to tap time was reduced from 37.43 min to 36.35 min. It could provide a real-time and effective solution to BOF reblow process for medium-high carbon steel melting.