Influence of Vanadium Extraction Converter Process Optimization on Vanadium Extraction Effect

Abstract

In view of the problem of low bottom-blowing stirring intensity and easy blockage in the use of capillary bricks, the bottom-blowing gas supply element of circular seam was used. The results of water model experiments shows that the suitable bottom-blowing element was arranged on the circumference 0.45D away from the center of the furnace bottom (where D is the diameter of the converter bottom). The best blowing process parameters for vanadium extraction from hot metal were are obtained in different periods. The industrial test results shows that the average values of vanadium content in semi-steel and metallic iron content in vanadium slag were 0.033 wt% and 22.39 wt%, respectively, after process optimization, which were 0.003 wt% and 5.25 wt% lower than those before process optimization. The average value of vanadium oxide content in vanadium slag was 18.54 wt%, an increase of 0.2 wt% compared with the one before process optimization. This shows that after the process was optimized, the kinetic conditions of the molten bath were improved, and the vanadium oxide reaction was more sufficient. An additional 5210 tons of vanadium slag could be produced each year, and better economic benefits could be obtained.

1. Introduction

Vanadium extraction by means of oxygen blowing in a converter is currently one of the effective methods for extracting vanadium from vanadium-containing hot metal. That is, the vanadium in the metal is oxidized by blowing oxygen from the top lance, and the vanadium in the hot metal is oxidized to vanadium oxide as much as possible [1,2,3]. At the same time, the oxidized hot metal also needs to be used as a raw material for subsequent steelmaking. Therefore, the production of vanadium slag with high vanadium content and semi-steel with sufficient carbon content and temperature in the vanadium extraction furnace is a prerequisite to ensure the smooth production of the vanadium extraction converter and the decarburization converter.

Many scholars have performed a lot of research on the blowing process of vanadium extraction from hot metal [4,5,6,7,8,9,10,11,12,13,14,15]. The characteristics of vanadium slag in the process of vanadium extraction from hot metal have been studied, including how to control the composition of FeO-SiO₂-based vanadium slag [4]. The saturated solubility of magnesium oxide in vanadium slag and its influencing factors have been explored [5]. The relationship between slag basicity and iron oxide in vanadium slag mineral phase has been analyzed [6]. The influence of the mass ratio of calcium oxide to vanadium oxide on the yield of vanadium has been systematically studied [7,8,9,10]. The selection of the oxidation temperature between carbon and vanadium has been mainly studied in the blowing process [11]. How to improve the kinetic conditions of the molten bath [12] and establish a kinetic model for the oxidation of metallic vanadium has been studied [13]. At the same time, a new process of blowing CO₂–O₂ mixed gas into the molten bath has been developed [14,15,16].

The vanadium extracting converter of a plant uses capillary bricks in the vanadium extracting process, which often results in clogging. At the same time, the strength of the bottom-blowing gas supply is generally about 0.02 Nm³/t∙min, which cannot meet the agitation strength of the hot metal containing vanadium on the molten bath. The main problem in the vanadium extraction process is that the residual vanadium of semi-steel is relatively high, with an average residual vanadium content of 0.037 wt% and a fluctuation range from 0.012 to 0.087 wt%. The average content of metallic iron in vanadium slag is 27.64 wt%, and the fluctuating range is from 22.21 to 35.06 wt%. The average content of vanadium oxide in vanadium slag is 18.34 wt%, and the fluctuation range is from 16.84 to 20.18 wt%.

The bottom-blowing gas supply element of circular seam [17,18] has been widely used in the hot-metal smelting process, but they have been rarely used in vanadium-containing hot-metal smelting. Compared with capillary bricks, the circular seam gas supply element is supplied with gas by the circular seam, as shown in Figure 1. Its characteristics are that it is not easily blocked and that the stainless-steel high-strength ball head seal at the end is connected to the metal hose, which can prevent gas leakage in long-term use and further ensure that the bottom-blowing gas flow rate in the whole furnace meets the process requirements.

In this paper, a circular seam gas supply element was used to replace the capillary bricks. The process parameters of a vanadium extraction converter were optimized using the results of water model experiments and were implemented in an industrial test. Through experimental research, the purposes of increasing vanadium oxide content in vanadium slag and reducing metal iron content in vanadium slag and vanadium content in semi-steel were achieved.

2. Experimental Equipment and Methods

2.1. Water Model Experiment

2.1.1. Experimental Equipment

The 80 t vanadium extraction converter was selected as the prototype. According to the similar ratio of 1:4.7, the converter cold state model was made with plexiglass.

Table 1. Comparison of main dimensions of converter prototype and model.

Name	Prototype	Model
molten bath diameter/mm	3540	753
Converter height/mm	7140	1519
Converter mouth diameter/mm	2220	472

shows the relevant parameters of the converter prototype and model, and the experimental device diagram is shown in Figure 2.

2.1.2. Experimental Parameters

When establishing the kinetic similarity, it is necessary to ensure that the Froude criterion modified for the converter prototype is equal to the Froude criterion modified for the model, and the following equation is obtained [18].

$$F{r}^{\prime }=\frac{{\rho }_{\mathrm{g}}{\nu }_{\mathrm{g}}^2}{{\rho }_L\mathrm{g}{d}_0} (\mathrm{Froude} \mathrm{criterion} \mathrm{modified})$$

(1)

$$\frac{{\nu }_{air}}{{\nu }_{O_2}}={\left(\frac{{\rho }_{O_2}}{{\rho }_{air}}\right)}^{\frac{1}{2}}{\left(\frac{{\rho }_W}{{\rho }_i}\right)}^{\frac{1}{2}}{\left(\frac{d_M}{d_P}\right)}^{\frac{1}{2}}$$

(2)

where ${\nu }_{air}$ and ${\nu }_{O_2}$ are the gas flow velocity at the nozzle outlet of the model and prototype oxygen lance, m∙s⁻¹. ${\rho }_{air}$ and ${\rho }_{O_2}$ are the gas densities in the model and prototype, kg∙m⁻³. ${\rho }_W$ and ${\rho }_i$ are the liquid densities in the model and prototype, kg∙m⁻³. $d_M$ and $d_P$ are oxygen lance nozzle outlet diameter in model and prototype, m.

The equation for calculating the gas flow in the model and prototype is as follows:

$$\frac{Q_M}{Q_P}=\frac{{\nu }_{air}}{{\nu }_{O_2}}\times {\left(\frac{d_M}{d_P}\right)}^2$$

(3)

According to the actual operation data of the 80 t converter, the related parameters of the model kinetics were calculated as shown in

Table 2. Kinetic parameters related to the converter prototype and model.

Name	Gas Density /kg∙m−3	Liquid Density /kg·m−3	Nozzle Throat Diameter/mm	Nozzle Outlet Diameter/mm	Molten Bath Depth/mm
Prototype	1.43	7000	33.0	41.8	1131
Model	1.29	1000	7.02	8.89	240

. The number of nozzle holes in the oxygen lance in the experiment was three, the center angle was 11°, and the Mach number was 1.95.

According to Equations (1)–(3) and Table 2, the top blowing flow and bottom blowing flow in the experiment, as well as the arrangement of oxygen lance positions and bottom blowing, were calculated, as shown in

Table 3. Blowing parameters used in the water mold experiment.

Scheme Code	A1	A2	A3	A4	A5
Top blowing gas flow in prototype/Nm3∙h−1	11,000	12,000	13,000	14,000	15,000
Top blowing gas flow in the model/Nm3∙h−1	93.99	102.53	111.07	119.61	128.16
Scheme code	B1	B2	B3	B4
Bottom gas element arrangement	0.40D	0.45D	0.53D	0.60D
Scheme code	C1	C2
Bottom blowing gas flow in the prototype/Nm3∙h−1	160	260
Bottom blow gas flow in the model/Nm3∙h−1	1.25	1.63
Scheme code	D1	D2	D3
Oxygen lance position in prototype/mm	1200	1400	1600
Oxygen lance position in the model/mm	255	233	340

.

2.1.3. Experimental Method

In order to characterize the ability of the oxygen lance and bottom blowing to stir the molten bath during the blowing process, the mixing time of the molten bath, the impact depth and the impact area of the oxygen lance were measured in the experiment. In the experiment, the mixing time, impact depth and area were measured in the water model according to the changes of the set top blowing flow, bottom blowing flow, oxygen lance position and bottom blowing element arrangement (as shown in Table 3).

The method of determining the mixing time of the molten bath was to add NaCl aqueous solution to the molten bath, and measure the conductivity change in the water through the conductivity meter on the other side of the molten bath. When the conductivity was stable, it was regarded as the molten bath mixing. The time from the aqueous solution to the molten bath mixing was called the molten bath mixing time, and each blowing stage was repeated six times, and the average value was taken. The method of determining the impact depth was to wait for the oxygen jet to stabilize, and the height from the upper surface of the molten bath to the oxygen jet at the bottom of the impact pit of the molten bath was called the impact depth. In the experiment, the impact depth was measured three times with a ruler, and the average value was taken. The maximum projected area of the impact crater formed by the oxygen jet in the molten bath was called the impact area. In the measurement, the impact crater edge depth ≥ 10 mm was considered as the effective impact area. All impact area data taken in the experiment were the effective impact area.

2.2. Industrial Test

2.2.1. Test Equipment and Process

The test was carried out in an 80 t vanadium extraction converter, the diameter of the molten bath was 3870 mm, and the depth of the molten bath was 1131 mm. The oxygen lance was a three-hole Lava tube nozzle, the Mach number was 1.92, the throat diameter was 33 mm, the outlet diameter was 41.8 mm, the center angle was 11°, the oxygen pressure was 0.8–0.9 Mpa, and the oxygen supply flow was 11,000–16,000 Nm³/h. Capillary bricks were adopted in the bottom blowing gas supply device. The quantity was 4, and the bottom blowing gas intensity was 0.02 Nm³/t∙min. In the process of vanadium extraction from hot metal, capillary bricks were often blocked, and the bottom blowing flow was low, which cannot meet the requirements of hot metal extraction for vanadium agitation in the molten bath. Therefore, the circular seam gas supply element was used in place of capillary bricks, see Figure 1.

2.2.2. Test Scheme

The process of extracting vanadium from hot metal was first to desulfurize the hot metal containing vanadium, and then entered the vanadium-extracting converter to extract the vanadium. The composition of hot metal was a carbon content of 3.27 to 4.63 wt%, the silicon content was from 0.08 to 0.77 wt%, phosphorus content of 0.033 to 0.22 wt%, the sulfur content was from 0.002 to 0.05 wt%, and vanadium content of 0.20 to 0.56 wt%.

In the process of vanadium extraction from hot metal, iron balls and pellets were used to cool down the temperature of the hot metal. The purpose was to prevent the temperature of the hot metal from rising too fast, which will cause a large amount of oxidation of the carbon content in the hot metal. At the end of blowing, the carbon content in the semi-steel was required to be from 3.0 to 3.6 wt%, the vanadium content was from 0.025 to 0.032 wt%, and the temperature was from 1320 to 1360 °C.

The semi-steel was poured into the hot metal ladle as the raw material for smelting clean molten steel. The content of vanadium oxide in the slag at the end of blowing was greater than 18 wt%, and the slag was poured into a slag tank as a raw material for extracting metallic vanadium, thereby realizing the efficient utilization of vanadium containing hot metal.

According to the results of the water model experiment, the industrial test scheme was determined, and the process before and after optimization was shown in

Table 4. Changes of blowing process before and after process optimization.

Name	Bottom Blowing Gas Supply Element	Bottom Blowing Gas Flow	Top Blowing Gas Flow and Oxygen Lance Position	Coolant Addition Time
Before optimization	The capillary bricks	The bottom blowing flow of the whole process was 160 Nm3/h	The top blowing flow was from 11,000 to 15,000 Nm3/h, and the oxygen lance position was from 1.2 to 1.6 m	Iron balls and pellets were added in two times, every 2 min, and the addition is completed within 4 min.
after optimization	The circular seam bottom blowing gas supply element	0–220 s, bottom blowing flow was 160 Nm3/h; 220–280 s, the bottom blowing flow was 260 Nm3/h	0–120 s, the top blowing flow was 14,000 Nm3/h, and the oxygen lance position was from 1.4 to 1.5 m; 120–220 s, the top blowing flow was from 12,000 to 13,000 Nm3/h, and the oxygen lance position was 1.55 m; 220–280 s, the top blowing flow was from 12,000 to 13,000 Nm3/h, the gun position was 1.3 m	Iron balls and pellets were added in three times, every 1 min, and the addition was completed within 3.5 min.

.

After the process was optimized, the whole process was divided into three stages. The top blowing gas flow, the oxygen lance position and the bottom blowing gas flow were changed in each stage. The time interval for coolant addition has varied.

2.2.3. Test Methods

The test method was mainly to select steel samples and slag samples for the end point of vanadium extraction. The chemical composition of the steel sample was analyzed, and the slag sample was analyzed for its mineral composition by X-ray diffraction; then the mineral structure was observed under a microscope to obtain a semi-quantitative mineral phase composition.

3. Results

3.1. Determination of the Arrangement of Bottom Blowing Elements on the Bottom of the Converter

The bottom blowing element of the original design of the vanadium extraction converter was arranged on 0.60D. In order to optimize the design, the layouts of 0.40D, 0.45D and 0.53D were selected, respectively, as shown in Figure 3.

Figure 4 shows the effect of different bottom blowing element arrangements on the mixing time measured in the experiment. The meaning of A2C1D1 in the figure was that the top blowing gas flow was 12,000 Nm³/h, the oxygen lance position was 1.2 m, and the bottom blowing gas flow was 160 Nm³/h. A4C2D2 means that the top blowing gas flow was 14,000 Nm³/h, the oxygen lance position was 1.2 m, and the bottom blowing gas flow was 160 Nm³/h.

Under different oxygen jet and bottom blowing stirring conditions, when the bottom blowing elements were arranged at 0.40D and 0.45D, the melt bath mixing time was the shortest. However, considering that the vanadium slag in the vanadium extracting converter needed to be smelted three times before being poured out, a large amount of vanadium slag would be accumulated at the bottom of the converter.

If the bottom blowing element was too close to the bottom of the converter, a large amount of vanadium slag might be accumulated, causing the blockage of the bottom blowing gas element. Therefore, considering comprehensively, choosing 0.45D was more suitable for the process of vanadium extraction from hot metal.

3.2. Influence of Top Blowing Gas Flow, Oxygen Lance Position and Bottom Blowing Gas Flow on the Stirring Effect of Molten Bath

Figure 5 shows the effects of top blowing gas flow, oxygen lance position and bottom blowing gas flow on the mixing time of the molten bath. The bottom blowing gas flow had a great influence on the mixing time. The mixing time in the molten bath without bottom blowing gas stirring was obviously longer than that with bottom blowing stirring.

When the bottom blowing gas flow increased from 160 Nm³/h to 260 Nm³/h, the mixing time of the molten bath was further shortened. This shows that the bottom blowing gas agitation had a better uniform mixing effect on the molten bath, and with the increased bottom blowing gas flow, the uniform mixing time of the molten bath was shortened. When the top blowing gas flow was greater than 14,000 Nm³/h, the mixing time of the molten bath decreased significantly with the increase in the bottom blowing gas flow.

Figure 5 (right figure) shows the effect of oxygen lance position on the mixing time. When the oxygen lance position was 1.2 m, the mixing time of the molten bath could be significantly reduced. When the top blowing gas flow was changed from 11,000 to 15,000 Nm³/h, with the increase in top blowing gas flow, the effect on the mixing time was less significant than that of the bottom blowing gas flow.

The impact depth is an important index to measure the effect of the oxygen jet on the molten bath. With the increase in the top blowing gas flow, the impact depth becomes larger at different oxygen lance positions. When the oxygen lance position is lower, the impact depth is greater.

The impact area determines the degree of reaction of the oxygen jet to the slag-steel interface. When the oxygen lance position was higher, the impact area was also larger. When the top-blowing gas flow increased from 11,000 Nm³/h to 14,000 Nm³/h, the impact area had not changed much, but when the top blowing gas flow increased to 15,000 Nm³/h, the impact area increased significantly, as shown in Figure 6.

3.3. Effect of Vanadium Extraction from Hot Metal before and after Process Optimization

The bottom blowing process of vanadium extracting converter was optimized by using the circular seam bottom blowing gas supply element. After process optimization, with the increase in bottom blowing gas flow, the molten bath of vanadium extraction converter maintained a good stirring effect. The data before and after process optimization were compared, in which the data of the finished vanadium slag was 40 heats, and the data of semi-steel was 900 heats. The comparison of relevant process technical indicators is shown in

Table 5. Changes of main components and temperature of vanadium slag and semi-steel before and after process optimization.

Item	Value	Before Process Optimization	After Process Optimization
(V2O5) in vanadium slag	average value/wt%	18.34	18.54
	variation range/wt%	15.4–20.18	15.7–20.2
metallic iron in vanadium slag	average value/wt%	27.63	22.39
	variation range/wt%	22.2–35.06	17.04–35.28
[C] in semi-steel	average value/wt%	3.28	3.36
	variation range/wt%	2.45–3.99	2.38–4.16
[V] in semi-steel	average value/wt%	0.036	0.033
	variation range/wt%	0.012–0.087	0.014–0.066
Semi-steel temperature	average value/℃	1358	1351
	variation range/℃	1277–1458	1261–1449

.

After process optimization, the average vanadium oxide content of vanadium slag increased by 0.2 wt%. Under the condition that the vanadium content of hot metal was 0.33%, after the process optimization, the metal iron content of vanadium slag was significantly reduced, from an average of 27.64 wt% to 22.39 wt%. When the carbon content in the hot metal was 4.04 wt% and the oxygen consumption was 850 Nm³/heat, the carbon content in the semi-steel increased by 0.08 wt% on average after the process optimization. This shows that after the process optimization, the kinetic conditions of the molten bath were improved, and the vanadium oxygen reaction was more sufficient.

For the semi-steel composition, the vanadium content of the semi-steel was reduced by 0.003 wt% after the process optimization. Therefore, the bottom blowing stirring of the vanadium extraction converter was strengthened, and the vanadium content of the semi-steel was reduced, which was helpful to increase the vanadium content in the vanadium slag.

Before the process optimization, the bottom blowing element of the capillary bricks was easily blocked in use and erode rapidly. After the optimization of the process, the circular seam gas supply element was adopted. In terms of the use effect, there was no clogging phenomenon, and the flow and pressure of each branch pipe were relatively stable during use.

The output of vanadium slag was calculated as follows:

W₀ = W_s × (1 − W_i) × (V)/(V₀)

(4)

where W₀ was vanadium slag output, W_i was vanadium slag weight, V was vanadium slag grade, V₀ was standard vanadium slag grade.

The content of metallic iron in vanadium slag before and after optimization was 27.63 wt% and 22.39 wt%, the grades of vanadium slag were 18.34 wt% and 18.54 wt%, respectively, and the standard vanadium slag grade was 15%. Calculated based on the annual production of 70,000 tons of vanadium slag, after the process was optimized, an additional 5210 tons of vanadium slag can be produced each year, and based on the price of 1700 USD per ton of vanadium slag, economic benefits of 8,855,300 USD can be obtained.

4. Discussion

4.1. Variation of Molten Bath Stirring Energy after Process Optimization

The stirring effect of the oxygen jet on the molten bath is a combination of the top blowing gas flow, the bottom blowing gas flow and the oxygen lance position. Using the equation of molten bath stirring energy density, these factors are linked together to obtain the relationship between molten bath stirring energy and process parameters in the process of vanadium extraction from hot metal. The equation for stirring energy density [19,20] is:

$${\epsilon }_T=1.83\times {10}^{-5}\frac{Q_T^3}{d^{3}h{V}_l}$$

(5)

$${\epsilon }_B=14.25\frac{Q_B}{V_l}{T}_l\lg \left(1+\frac{H}{1.47}\right)$$

(6)

$$\epsilon ={\epsilon }_B+0.1{\epsilon }_T$$

(7)

where Q_T is the top blowing gas flow, Nm³/min. d is the throat diameter of the oxygen lance, m. h is the oxygen lance position, m. V_l is the liquid volume of molten bath, m³. T_l is the temperature of the molten bath, K. H is the depth of molten bath, m. ε_T is the top blowing stirring energy density. W∙m⁻³; ε_B is the bottom blowing stirring energy density, W∙m⁻³.

From Equation (7), it can be seen that the effect of the top blowing gas on the stirring energy of the molten bath was 1/10 of that of the bottom-blowing gas, which was consistent with the research results of Kohtani T [21] on the stirring energy. He believed that most of the energy generated by the top blowing gas was consumed by the impingement liquid surface, and only 6 to 10% was converted into effective stirring power.

According to the changes of process parameters before and after optimization in Table 4, it was calculated that the average stirring energy of the molten bath before process optimization was 31.0 W/m³, and the variation range was from 24 to 45.23 W/m³. After process optimization, the molten bath stirring energy was 43.21 W/m³ on average, and the range was from 30.2 to 52.51 W/m³. This shows that the molten bath stirring effect after process optimization was greatly enhanced. The average mixing time of the molten bath before and after the process optimization measured in the water model experiment was 50 and 45 s, which was basically consistent with the calculation of the stirring energy.

In the initial stage of the vanadium extraction from the hot metal, it was beneficial to increase the fluidity of the molten bath by appropriately increasing the top blowing gas flow and reducing the position of the oxygen lance to increase the temperature of the molten bath. At this stage, the top blowing gas flow was chosen to be 14,000 Nm³/h, the oxygen lance position was from 1.4 to 1.5 m, and the bottom blowing gas flow was 160 Nm³/h. In the middle of the blowing process, appropriately reducing the top blowing flow and increasing the oxygen lance position would help prevent the temperature of the molten bath from rising too quickly. At this stage, the selected top blowing gas flow was from 12,000 to 13,000 m³/h, the oxygen lance position was 1.55 m, and the bottom blowing gas flow was 260 Nm³/h. In the later stage of blowing, the use of lower top blowing gas intensity and proper oxygen lance position helped to prevent the carbon content in the hot metal from oxidizing too quickly, and at the same time reduced the iron oxide content and metallic iron content in the vanadium slag. At this stage, the selected top blowing gas flow was from 12,000 to 13,000 m³/h, the oxygen lance position was 1.3 m, and the bottom blowing flow was 260 Nm³/h.

According to Equations (5)–(7), the stirring energy of the molten bath at each stage of smelting and the measured mixing time are shown in

Table 6. Stirring energy calculated and mixing time of molten bath measured at each stage of smelting.

Stage	Top Blowing Gas Flow/Nm3/h	Oxygen Lance Position/m	Bottom Blowing Gas Flow/Nm3/h	Mixing Time/s	Stirring Energy/W/m3
Early stage of blowing (0–120 s)	14,000	1.4–1.5	160	42	39.45–42.15
Middle stage of blowing (120–220 s)	12,000–13,000	1.55	260	50	25.78–32.02
Later stage of blowing (220–280 s)	13,000–13,500	1.3	260	38	37.65–41.83

. The parameters in the process of vanadium extraction from hot metal were optimized by measuring the mixing time and calculating the stirring energy, which conformed to the chemical reaction and temperature changes in the process of vanadium extraction from hot metal, and improved the kinetic conditions of the molten bath as much as possible.

4.2. Analysis of Factors Affecting the Distribution Ratio of Vanadium

When the bottom blowing process of the vanadium extraction furnace was optimized, the transfer of vanadium in the hot metal to the slag was promoted. Therefore, the ratio of the vanadium content in the vanadium slag to the vanadium content in the hot metal can be used to evaluate the degree of vanadium in the hot metal entering the vanadium slag. The higher the distribution ratio of vanadium, the more vanadium in the hot metal entered the slag, which was more conducive to the generation of vanadium slag with high vanadium content, thereby improving the efficiency of vanadium extraction [16].

The distribution ratio of vanadium (L_v) was calculated as shown in Equation (8).

$$L_v=\frac{0.56\times W_s\times \left({\mathrm{V}}_2{\mathrm{O}}_5\right)}{W_h\times \left[V\right]}\times 100\%$$

(8)

where W_s was the weight of the vanadium slag, t. W_h was the weight of the semi-steel, t. (V₂O₅) was vanadium oxide content in the vanadium slag, %. [V] was the vanadium content in the semi-steel, %.

Figure 7 shows the distribution ratio of vanadium obtained in the experiment. The vanadium distribution ratio after process optimization was increased by one compared with that before the process optimization, and the vanadium in the hot metal after the process optimization was more likely to enter the slag.

Figure 8 shows the factors that affect the distribution ratio of vanadium, and the test data were selected from the data after optimization. With the increase of oxygen consumption, the distribution ratio of vanadium increased obviously. The effect of the carbon content at the end-point on the distribution ratio of vanadium was not significant, and it tended to increase slightly. The distribution ratio of vanadium decreased with the increase in the endpoint temperature and the added coolant.

Based on the above data, linear regression was performed on the influencing factors of the vanadium distribution ratio, and the following relational expressions affecting the vanadium distribution ratio were obtained.

L_v = 0.67W_O + 0.20[C] − 0.005T − 0.005W_C + 8.01

(9)

where L_v was the distribution ratio of vanadium. W_O was the oxygen consumption, Nm³/t. [C] was the carbon content at the end-point, wt%. T was the end point temperature, °C. W_C was the weight of iron balls and pellets, kg/t.

According to the target of the current process blowing end point, the end point temperature was set from 1370 to 1390 °C, the carbon content was from 3.5 to 3.7 wt%, the weight of the iron balls and pellets was from 10 to 30 kg/t, and the oxygen consumption was from 11 to 12 Nm³/t. According to the calculation of Equation (9), the obtained distribution ratio of vanadium was above 9.18.

4.3. Changes in Slag Mineral Phase Composition before and after Process Optimization

In order to analyze the vanadium extraction process of the vanadium extraction furnace, it was very helpful to understand the chemical composition of the vanadium slag and its mineral phase composition in the process of vanadium extraction.

The composition changes of vanadium slag before and after process optimization were shown in

Table 7. Changes of vanadium slag composition before and after process optimization/wt%.

Name	CaO	SiO2	MgO	FeO	MnO	V2O5	M.Fe
Before optimization	4.78	12.38	1.24	31.26	8.0	18.56	17.28
After optimization	4.57	11.78	1.44	24.31	7.8	21.66	15.48

. From the composition comparison, the content of iron oxide and metallic iron in the vanadium slag after the process optimization has been greatly reduced.

The XRD measurement result of vanadium slag is shown in Figure 9. Before process optimization, the mineral phase composition of vanadium slag was FeO∙V₂O₃ accounting for 76.1 wt%, 2FeO∙SiO₂ accounting for 19.2 wt%, and Fe₃O₄ accounting for 4.7 wt%. After the process optimization, the mineral phase composition of vanadium slag was FeO∙V₂O₃, accounting for 85.3 wt%, CaO∙FeO∙2SiO₂, accounting for 11.3 wt%, and FeO accounting for 3.4 wt%.

Figure 10 shows the mineral phase structure photos of vanadium slag before and after process optimization. Before the process optimization, the mineral phase composition of vanadium slag was mainly RO phase and FeO∙V₂O₅ phase. After the process optimization, the mineral phase composition of vanadium slag has not changed, and the metallic iron content contained in it has been greatly reduced.

According to the above analysis, the combination of vanadium oxide and iron oxide in vanadium slag can produce vanadium iron oxide. The reaction was as follows:

FeO_(l) + V₂O_3(s) = FeO∙V₂O_3(s)    ΔG⁰ = −45190 + 16.32T

(10)

According to Equation (10), the following equation can be obtained.

lgX_V2O3 = −2360/T + 0.85 − lga_FeO − lgγ_V2O3

(11)

According to the composition of vanadium slag, SiO₂ was from 17 wt% to 18 wt%, CaO was from 3.9 wt% to 4.1 wt%, Cr₂O₃ was from 0.7 wt% to 0.8 wt%, V₂O₃ was from 16 wt% to 25 wt%, FeO was from 30 wt% to 40 wt%, M.Fe was from 15 wt% to 25 wt%. The activity coefficient of iron oxide and V₂O₃ of vanadium slag was selected as 1. Thus, the relationship between the iron oxide content and the vanadium oxide content of the vanadium slag can be obtained. In the temperature range of 1380–1400 °C, as the iron oxide content of the vanadium slag decreased, the vanadium oxide content decreased. When the iron oxide content of the vanadium slag was reduced from 40 wt% to 30 wt%, the V₂O₃ content in the vanadium slag increases by 3.5 wt% to 23.0 wt%.

Theoretically, the mass ratio of FeO to V₂O₃ was one when FeO∙V₂O₃ was generated. However, the current mass ratio of FeO to V₂O₃ was 1.5. It could be seen that by improving the stirring conditions of the bottom blowing and increasing the top blowing stirring intensity, the content of iron oxide and metallic iron in the vanadium slag was obviously reduced, which helped to increase the content of V₂O₃ in the vanadium slag.

5. Conclusions

In this paper, the effect of bottom blowing stirring of vanadium extraction furnace on vanadium extraction effects from vanadium-containing hot metal was determined by the water model experiment and industrial test, and the following conclusions were obtained:

• The arrangement of the bottom blowing element was determined by the water model experiment. When the bottom blowing gas element was located at 0.45D (D was the diameter of the furnace bottom), the best stirring effect could be obtained.
• The best blowing process parameters in the blowing process were determined: in the early stage of blowing, the top blowing gas flow was 14,000 Nm³/h, the oxygen lance position was from 1.4 to 1.5 m, and the bottom blowing gas flow was 160 Nm³/h. In the middle stage of blowing, the top blowing gas flow was from 12,000 to 13,000 Nm³/h, the oxygen lance position was 1.55 m, and the bottom blowing gas flow was 260 Nm³/h. In the later stage of blowing, the top blowing gas flow was from 13,000 to 13,500 Nm³/h, the oxygen lance position was 1.3 m and the bottom blowing gas flow was 260 Nm³/h.
• After process optimization, the average value of vanadium content of semi-steel and metal iron content of vanadium slag were 0.033 wt% and 22.39 wt%, respectively, which were 0.003 wt% and 5.25 wt% lower than those before process optimization. The average vanadium oxide content of vanadium slag was 18.54 wt%, which was 0.2 wt% higher than that before process optimization. This shows that after the process optimization, the kinetic conditions of the molten bath were good, and the vanadium oxygen reaction was more sufficient.