Investigation of the Oxygen Bottom Blown Copper Smelting Process

Abstract

The oxygen bottom blown copper smelting process is a new technology which has been widely applied to the copper production in China. In this work, a computational thermodynamics model for this technology has been established, based on smelting mechanism and theory of Gibbs free energy minimization . The calculated results from the model agree well with the actual industrial data, indicating that the model can be used for the predictions under different operating conditions. The tendencies of the key parameters (such as Cu losses and Fe₃O₄ content in slag ) and the distribution ratios of the minor elements (such as Pb, Zn, As, Sb and Bi) can be predicted by adjusting the oxygen/ore ratio charged into the bottom blown copper smelting furnace . The model can be used to monitor and optimize the industrial operations of the oxygen bottom blown copper smelting process.

Introduction

The oxygen bottom blown copper smelting process is a newly developed intensifying smelting process, which has been widely applied to the copper productions in China in the past 10 years [6, 23]. This process was industrially tested initially in the Shuikoushan (SKS) smelter in 1990 and was named originally as “SKS process ”, which is also called BBS (bottom blown smelting ) or BBF (bottom blown furnace ) at present. As the SKS process is being widely applied in recent years [3, 4, 5, 26], fundamental research related to the SKS process is urgently required. Chen [2] studied the slag chemistry of the bottom blown copper smelting furnace at Dongying Fangyuan, and analysed the copper losses in the industrial smelting slags. Shui et al. [18, 19] studied the bath surface wave and mixing phenomena in the bottom blown copper smelting furnace , which provided a better understanding of blowing patterns. Liu et al. [10] studied the phase equilibria of the ZnO–“FeO”–SiO₂ and ZnO–“FeO”–SiO₂–Al₂O₃ systems at Po₂ 10⁻⁸ atm, which is close to the SKS smelting condition, and found that the presence of ZnO or Al₂O₃ in slag significantly increases the liquidus temperature . Zhang et al. [28, 29] analysed gas-liquid multi-phase flows in a SKS furnace to optimize the oxygen lance structure parameters using a CFD method. Yan et al. [28] studied the influence of the lances arrangement on the SKS process , and found an optimal lance inclination angle and spacing distance. Guo et al. [4, 5, 7] studied the mechanism, multiphase interface behavior, and optimization of the SKS process .

Up to date, no work on thermodynamics modeling of the SKS process has been reported. In the present study, a computational thermodynamics model of the SKS process has been developed and validated. The model can then be used to predict matte grade, copper losses in slag , Fe₃O₄ content in slag , the distribution of minor elements , and the results can be compared with the production data. This work is part of a comprehensive research program to gain more understanding of this new technology .

Method

Mechanism of SKS Process and Behavior of S₂

We have developed a mechanism model of the SKS copper smelting process, which is the basis of the computational thermodynamics model . In the mechanism model , the SKS furnace section from top to bottom is divided into seven functional layers, while the SKS furnace length is divided into three functional regions, as shown in Figs. 1 and 2 respectively. These layers and regions play different roles in the smelting process. After the copper sulfide concentrates and flux are fed into the top of the furnace , the concentrates decompose into Cu₂S, FeS, and S₂ (and other minor species) in the mineral decomposition transition layer. The Cu₂S and FeS sink to form the matte layer, and the decomposed S₂ goes directly into the gas phase layer. The oxygen-enriched air is blown into the matte layer with oxygen lances from the bottom of the furnace and almost all oxygen is consumed in the bath. S₂ can hardly get in contact and react with the oxygen. As a result, S₂ passes into the flue directly and then reacts with the oxygen from air infiltration in the flue rather than inside the furnace . Theoretically, a chemical multiphase equilibrium calculation should be based on the assumption that all substances are accounted for. However, the fact is that a portion of the sulfur from the concentrate (about 15% of total S in the smelting system according to [9] passes directly into the flue without reaction in the SKS furnace . This phenomena has been taken into consideration in the thermodynamic model to achieve a more precise and reliable computational results.

../images/468727_1_En_36_Chapter/468727_1_En_36_Fig1_HTML.gif — Fig. 1
SKS mechanism model-I [4, 5]

../images/468727_1_En_36_Chapter/468727_1_En_36_Fig2_HTML.gif — Fig. 2
SKS mechanism model-II [4, 5]

Thermodynamic Model

The SKS process is a typical multiphase and multicomponent system coupled with various chemical reactions. According to the second law of thermodynamics , spontaneous reaction always proceeds towards the direction that the total Gibbs free energy decreases. An isothermal and isobaric chemical system is at equilibrium when the total Gibbs free energy is minimized [8]. It is assumed that the SKS process is under isothermal and isobaric conditions and reaches thermodynamic equilibrium, so the total Gibbs free energy of the SKS system attains its minimum.

The total Gibbs free energy function [13] for a multiphase and multicomponent system can be expressed as Eq. 1 [15].

$G(n,T,P) = \sum\limits_{j = 1}^{{N_{p} }} {\sum\limits_{i = 1}^{{N_{c} }} {n_{ij} \mu_{ij} } } = \sum\limits_{j = 1}^{{N_{p} }} {\sum\limits_{i = 1}^{{N_{c} }} {n_{ij} [\Delta G_{ij}^{o} + RT\,\ln \left( {\frac{{f_{ij} }}{{f_{ij}^{o} }}} \right)]} }$

(1)

Where N_p and N_c are, respectively, the number of phases and the number of components in the system, n_ij is the mole fraction of component i in phase j, μ_ij is the partial molar Gibbs free energy , i.e., the chemical potential of component i in phase j in the system, which is comprised of two parts to modify its non-ideality. One part is $\Delta G_{ij}^{o}$ —the standard Gibbs free energy of formation of component i in phase j under the system temperature and standard pressure. $\Delta G_{ij}^{o}$ modifies the impact of temperature on the standard Gibbs free energy and can be calculated by the Eq. 2 as follows:

$\Delta G_{ij}^{o} = A_{ij} + B_{ij} T$

(2)

The other part is $R \cdot T \cdot ln(f_{ij} /f_{ij}^{o} )$ , which modifies the influence of pressure and concentration under non-ideal condition on the standard Gibbs free energy ; R is the ideal gas constant and T is the temperature in Kelvin, f_ij is the partial fugacity of component i in the phase j, and $f_{ij}^{o}$ is the fugacity of component i in the phase j at reference state. The value of the fugacity is dependent on the system temperature and pressure. In the SKS process , the gas phase is treated as an ideal gas. The high temperature melts are considered to be non-ideal liquid solutions. Therefore, activity coefficients are introduced to substitute fugacity and modify the chemical potential of the constituents in the real solution against the inapplicability of the ideal solution. The fugacity of components in the molten phase and gas phase can be expressed by Eqs. 3 and 4, respectively.

$RT\,\ln \left( {\frac{{f_{ij} }}{{f_{ij}^{o} }}} \right) = RT\,\ln (\lambda_{ij} x_{ij} )$

(3)

$RT\,\ln \left( {\frac{{f_{ij} }}{{f_{ij}^{o} }}} \right) = RT\,\ln (y_{ij} P)$

(4)

where λ_ij and x_ij are, respectively, the activity coefficient and mole fraction of component i in phase j, y_ij is the partial pressure of component i in the gas phase, and P is the total pressure of the gas phase.

At the same time, according to the principle of mass conservation law, the mass of input species or elements to the system should be equal to the mass of the output species and elements and the mole number of each component should be above zero. In this manner, the multiphase equilibrium thermodynamic model is established and component contents at equilibrium can be obtained by finding the minimum value of the total Gibbs free energy under the specific operating conditions with the input and output constraints for all species and elements.

In the SKS process model , the elements Cu, Fe, S, O, N, H, Si, As, Sb, Bi, Pb, Zn, Mg, Ca, and Al are taken into consideration for the multiphase equilibrium calculation to reflect the actual plant production. When the system is close to equilibrium, three independent stable phases, matte, slag and gas, are formed. The chemical components formed after thermodynamic equilibrium is reached are listed in Table 1.

Table 1

Chemical components in SKS copper smelting process [25]

Phases	Chemical components
Gas	SO₂, SO₃, S₂, O₂, N₂, H₂O, PbO, PbS, Zn, ZnS, As₂, AsO, AsS, SbO, SbS, BiS
Slag	FeO, Cu₂S, Cu₂O, Fe₃O₄, FeS, PbO, ZnO, As₂O₃, Sb₂O₃, Bi₂O₃, SiO₂, CaO , MgO, Al₂O₃
Matte	Cu₂S, Cu, FeS, FeO, Fe₃O₄, Pb, PbS, ZnS, As, Sb, Bi

The SKS system is assumed to be under isothermal and isobaric airtight conditions. The smelting temperature is around 1473 K (1200 °C).

Thermodynamic Data

The activity coefficient of each component are selected from previous publications and are listed in Table 2 [1, 21], which are crucial in the SKS process multiphase equilibrium calculation. The standard Gibbs free energy of each component formation is listed in Table 3 [16, 22, 20].

Table 2

Phase entrainment coefficient in the SKS process

Components	Phase	Activity coefficient
Cu₂S	Matte	1
FeS	Matte	$0.925/\left( {{\text{N}}_{{{\text{Cu}}_{2} {\text{S}}}} + 1} \right)$
Cu	Matte	14
FeO	Matte	$exp\left[ {5.1 + 6.2\left( {ln{\text{N}}_{{{\text{Cu}}_{ 2} {\text{S}}}} } \right) + 6.41\left( {ln{\text{N}}_{{{\text{Cu}}_{ 2} {\text{S}}}} } \right)^{2} + 2.8\left( {ln{\text{N}}_{{{\text{Cu}}_{ 2} {\text{S}}}} } \right)^{3} } \right]$
Fe₃O₄	Matte	$exp\left[ {4.96 + 9.9\left( {ln{\text{N}}_{{{\text{Cu}}_{ 2} {\text{S}}}} } \right) + 7.43\left( {ln{\text{N}}_{{{\text{Cu}}_{ 2} {\text{S}}}} } \right)^{2} + 2.55\left( {ln{\text{N}}_{{{\text{Cu}}_{ 2} {\text{S}}}} } \right)^{3} } \right]$
Pb	Matte	23
PbS	Matte	$exp[ - 2.716 + 2441/{\text{T}} + \left( {0.815 - 3610/{\text{T}}} \right)\left( {80 - \left[ {{\text{Pct}} \cdot {\text{Cu}}} \right]_{\text{mt}} } \right)/100]$
ZnS	Matte	$exp[ - 2.054 + 6917/{\text{T}} - \left( {1.522 - 1032/{\text{T}}} \right)\left( {80 - \left[ {{\text{Pct}} \cdot {\text{Cu}}} \right]_{\text{mt}} } \right)/100]$
As	Matte	$8.087 - 0.128\left[ {{\text{Pct}} \cdot {\text{Cu}}} \right]_{\text{mt}} + 0. 0 1 4\left[ {{\text{Pct}} \cdot {\text{Cu}}} \right]_{\text{mt}} \times \,\lg \left[ {{\text{Pct}} \cdot {\text{Cu}}} \right]_{\text{mt}}$
Sb	Matte	$- 0. 9 9 6 { } + 2.42\left[ {{\text{Pct}} \cdot {\text{Cu}}} \right]_{\text{mt}} - 1. 2 6\left[ {{\text{Pct}} \cdot {\text{Cu}}} \right]_{\text{mt}} \times \,\lg \left[ {{\text{Pct}} \cdot {\text{Cu}}} \right]_{\text{mt}}$
Bi	Matte	$10^{{(1900/{\text{T}} - 0.464)}}$
FeO	Slag	$1.42{\text{N}}_{\text{FeO}} - 0.044$
SiO₂	Slag	2.1
Fe₃O₄	Slag	$0. 6 9\,{ + }\, 5 6. 8 {\text{N}}_{{{\text{Fe}}_{ 3} {\text{O}}_{ 4} }} \,{ + }\, 5. 4 5 {\text{N}}_{{{\text{SiO}}_{ 2} }}$
Cu₂O	Slag	$57.14{\text{N}}_{{{\text{Cu}}_{ 2} {\text{O}}}}$
FeS	Slag	70
Cu₂S	Slag	$exp\left( {2.46 + 6.22{\text{N}}_{{{\text{Cu}}_{ 2} {\text{S}}}} } \right)$
PbO	Slag	$exp\left( { - 3330/{\text{T}}} \right)$
ZnO	Slag	$exp\left( {920/{\text{T}}} \right)$
As₂O₃	Slag	$3.838exp\left( { 1 5 2 3 / {\text{T}}} \right) \times {\text{P}}_{{{\text{O}}_{ 2} }}^{ 0. 1 5 8}$
Sb₂O₃	Slag	$exp\left( {1055.66/T} \right)$
Bi₂O₃	Slag	$exp\left( { - 1055.66/T} \right)$

Table 3

Standard Gibbs free energy of formation of each component (J·mol⁻¹)

Components	State	$A_{ij}$	$B_{ij}$
Cu₂S	Liquid	−145349	43.06
FeS	Liquid	−135556	43.06
PbS	Liquid	−151881	79.67
ZnS	Solid	−391434	203.08
Cu₂O	Liquid	−137139	54.25
FeO	Liquid	−259244	62.38
Fe₃O₄	Solid	−1097693.74	305.93
As₂O₃	Liquid	−1215325.18	457.37
Sb₂O₃	Liquid	−687438	237.86
Bi₂O₃	Liquid	−563470	257.66
PbO	Liquid	−196818	79.15
ZnO	Solid	−475260	208.63
SiO₂	Liquid	−912677	180.92
SO₃	Gas	−459543	165.15
SO₂	Gas	−361500	72.49
As₂	Gas	−415418	113.24
AsS	Gas	−184465	45.88
AsO	Gas	−257759	46.12
SbO	Gas	−126601	−60.35
SbS	Gas	103194	−59.91
BiS	Gas	−0.057	96.74
PbO	Gas	60860	−54.39
PbS	Gas	57812	−53.83
ZnS	Gas	13200	32.15

Physical Entrainment

Copper concentrates smelting and further copper matte converting produces slag containing copper in dissolved form as well as mechanically entrained form. Usually, the amount of dissolved copper is much less than that the amount of mechanically entrained matte. It is worth mentioning that the mechanically entrained matte is influenced by many factors, such as the operation parameters, the size of the matte droplets, the specific gravity and viscosity of the melts, the settling time, etc. Since the multiphase equilibrium model does not depict the mechanically entrained matte, we drew lessons from Nagamori et al. [11] and established a copper matte mechanical entrainment model that suits the SKS process and that is represented by Eq. 5 for slag entrainment in the matte phase and Eq. 6 for matte entrainment in the slag phase.

$M_{slag}^{ap} = (S_{mt}^{sl} \cdot S_{sl}^{mt} \cdot M_{matte} + S_{mt}^{sl} \cdot S_{sl}^{mt} \cdot M_{slag} - S_{sl}^{mt} \cdot M_{matte} )/(S_{sl}^{mt} + S_{mt}^{sl} - 1)$

(5)

$M_{matte}^{ap} = (S_{mt}^{sl} \cdot S_{sl}^{mt} \cdot M_{matte} + S_{mt}^{sl} \cdot S_{sl}^{mt} \cdot M_{slag} - S_{mt}^{sl} \cdot M_{slag} )/(S_{sl}^{mt} + S_{mt}^{sl} - 1)$

(6)

In Eqs. 5 and 6, M_slag and M_matte are the calculated mass of slag and matte phases at equilibrium, respectively, $M_{slag}^{ap}$ and $M_{matte}^{ap}$ are the calculated apparent slag mass to the matte phase and matte mass to the slag phase, respectively, and $S_{sl}^{mt}$ and $S_{mt}^{sl}$ are the entrainment coefficients for the slag to matte phase and the matte to slag phase, respectively, which are derived from the actual plant production data. The entrainment coefficients for the slag to matte phase and the matte to slag phase are listed in Table 4.

Table 4

Phase entrainment coefficients in the SKS process

Matte grade	Phase entrainment coefficient
Matte grade	$S_{sl}^{mt} (\% )$	$S_{mt}^{sl} (\% )$
50	2.2	2.8
65	2.9	3.2
75	3.2	3.8

Process Data

The initial conditions and operation parameters used in the model are those from Dongying Fangyuan Nonferrous Metals Co . Ltd., which are crucial for the model calculation. The composition of the mixed concentrates charged into the furnace are listed in Table 5 and the operation parameters are listed in the Table 6.

Table 5

Composition of the mixed concentrates charge

Elements/Species	Cu	Fe	S	Pb	Zn	As	Sb	Bi	SiO₂	MgO	CaO	Al₂O₃	Others
Content (wt%)	24.4	26.8	28.6	0.96	1.9	0.37	0.10	0.10	6.4	1.9	2.4	2.3	3.9

Table 6

Operation parameters at Dongying Fangyuan Nonferrous Metals Co . Ltd.

Operation parameters	SKS plant data
Charging speed of dry mixed concentrates (t/h)	66
Water percent in the mixed concentrates (%)	10.21
Charging speed of flux (t/h)	5.277
Smelting temperature (K)	1475
Negative pressure in furnace (Pa)	50–200
Volume of pure oxygen (Nm³/h)	10885
Volume of air (Nm³/h)	5651
Volume of O₂ in oxygen-enriched air (%)	73
Oxygen efficiency (%)	99

Calculation Procedure

A particle swarm optimization algorithm, using C# computer programming language and Microsoft Visual Studio, was used to develop the thermodynamic model of the SKS process (SKSSIM). The calculations are carried out by adjusting the ratio of oxygen to ore using the SKSSIM software and the results can be saved in a Microsoft Excel spreadsheet format.

Results and Discussion

Model Validation

The Dongying Fangyuan smelter has produced a high grade matte greater than 70% Cu and achieved autogenously smelting . The multiphase equilibrium calculation is based on Dongying Fangyuan initial conditions and operation parameters. Industrial data under stable operation conditions are compared with the calculated data as shown in Tables 7 and 8. Moreover, the mechanical entrainment coefficients in the model are obtained by analysed results in stable operation specifications, which is believed to be similar to the current operational condition.

Table 7

Comparison of predictions with actual plant data of matte and slag compositions in the SKS process

Composition (wt%)		Cu	Fe	S	Pb	Zn	As	Sb	Bi	SiO₂
Plant data	matte	70.77	5.52	20.22	1.73	1.07	0.07	0.04	0.06	0.51
Plant data	slag	3.16	42.58	0.86	0.43	2.19	0.08	0.13	0.02	25.24
SKSSIM	matte	70.31	4.80	20.38	1.69	1.02	0.07	0.04	0.06	0.82
SKSSIM	slag	2.93	42.07	0.73	0.37	2.08	0.07	0.12	0.02	25.18

Table 8

Comparison of predictions with actual plant data of the minor elements distribution in the SKS process

Phase	Plant data (%)					SKSSIM (%)
Phase	As	Sb	Bi	Pb	Zn	As	Sb	Bi	Pb	Zn
Matte	5.91	12.31	19.10	55.61	17.76	6.23	12.58	18.74	56.71	17.35
Slag	12.08	71.05	11.40	24.91	64.86	11.06	72.30	11.13	23.47	66.46
Gas	82.01	16.64	69.50	19.48	17.38	82.71	15.12	70.136	19.82	16.19

Tables 7 and 8 provide a comparison of the calculated and actual plant data in the SKS process . Table 7 shows that the predicted matte and slag compositions are in good agreement with the industrial values. The calculated and industrial matte grade are 70.3% and 70.8%, respectively. The calculated and industrial Fe in matte are 4.8% and 5.5%, respectively. Table 8 shows that the calculated and actual plant data for the minor elements distribution between the matte, slag , and gas phases are in good agreement. The calculated content of minor elements in the matte, slag , and gas phases are 6.23%, 11.06%, 82.71% for arsenic, 12.58%, 72.30%, 15.12% for antimony, 18.74%, 11.13%, 70.13% for bismuth, 56.71%, 23.47%, 19.82% for lead , and 17.35%, 66.46%, 16.19% for zinc. With these calculated data, the minor elements distribution trends in the matte, slag , and gas phases can be clearly seen.

The above comparison shows that the agreement between the multiphase equilibrium model predicting data and the industrial data is excellent. Consequently, the reliability of the present multiphase equilibrium model is validated and the model can be further used to predict the element distribution trends or to optimize operating parameters of the SKS process .

Relationship Between Copper Losses in Slag and Matte Grades

Copper losses in slag are always one of the most concerning issue in copper production, and reducing copper losses can greatly improve the copper recovery and reduce the slag cleaning load. The relationship between the copper content in slag and the matte grade is often used to evaluate the performance of a smelting operation. We used our computational thermodynamics model to predict the copper losses in the SKS smelting slag . The calculated results are shown in Fig. 3. Since the model prediction values at equilibrium are very close to those of the plant data under normal operations of the SKS process , we can conclude that the predicted results by this model are credible.

../images/468727_1_En_36_Chapter/468727_1_En_36_Fig3_HTML.gif — Fig. 3
Copper losses to slag in SKS smelting process

The copper losses in slag are commonly considered to originate from two sources: (1) Cu⁺ ions dissolved in the liquid slag that may be present in Cu₂O or Cu₂S form; and (2) entrained matte droplets in slag . Various factors affect the copper losses to slag , including both chemical and physical aspects. Dissolved copper loss to slag has been discussed by Yazawa et al. [27], Chen et al. [2], and Shimpo et al. [17] based on laboratory test data. With matte grades above 70%, the results obtained with our model are close to Yazawa’s results. However, with matte grades around 75%, the results are closer to Chen’s results, which are situated between Yazawa’s and Shimpo’s results.

Figure 3 shows that, as the matte grade increases, the total copper loss in iron silicate slag increases. When the matte grade is greater than 70%, the copper loss in slag increases more rapidly. The changes in physical and chemical properties of the slag could account for this phenomenon. As the matte grade increases, the oxygen potential of the SKS smelting system also increases, and more FeO in slag is oxidized to Fe₃O₄ , which can increase the viscosity of the slag and further reduce the settlement velocity of matte droplets in slag , causing large amounts of copper loss to slag in the form of physical entrainment. This hypothesis was proved by the results of Chen et al. [2] that entrained copper losses account for more than 70% of the total copper losses , with an Fe/SiO₂ ratio of 1.7 and a matte grade around 75%.

The quenched slag sample collected from Dongying Fangyuan Nonferrous Metals CO ., Ltd was analysed by Electron Probe X-Ray Microanalysis (EPMA) at the University of Queensland (UQ). Figure 4 shows the microstructure of the quenched slag sample and Table 9 provides the compositions.

../images/468727_1_En_36_Chapter/468727_1_En_36_Fig4_HTML.gif — Fig. 4
SEM backscattered image of quenched slag sample with a matte grade of 72.1% from plant

Table 9

Compositions of quenched slag sample (wt pct)

Fe₃O₄	FeO	Cu(total)	Cu(dissolved)	Cu(entrained)
30.6	31.2	3.2	0.4	2.8

The FeO and Fe₃O₄ contents in slag as a function of the matte grade are given in Fig. 5, compared with those reported by Wang and Zhang [24]. As the matte grade increases up to 57%, the FeO content in the slag slightly increases, resulting from the oxidation of FeS in the matte. However, as the matte grade increases above 57%, the FeO content in the slag decreases as the oxygen potential is sufficient enough to oxidize more FeO to Fe₃O₄ in the slag . Consequently, the Fe₃O₄ content in slag increases consistently, which increases both the solid fraction in slag and the slag viscosity . This suggests that an optimum matte grade should be chosen in a SKS process to enhance the copper direct recovery and reduce the load of slag to the cleaning process.

../images/468727_1_En_36_Chapter/468727_1_En_36_Fig5_HTML.gif — Fig. 5
Contents of FeO and Fe₃O₄ in SKS smelting slag

Minor Element Distributions

As copper concentrates become more and more complex, the control of minor elements is an important issue for all copper smelters. In the present study, the minor elements discussed include lead , zinc, arsenic , antimony and bismuth . The minor elements distribution in the SKS process are compared with the Isasmelt process [12] and the Flash smelting process [23].

Lead is both an hygiene and environmental concern while also affecting the copper cathodes quality. The distribution of lead is given in Fig. 6. In normal plant operational conditions, the distribution of lead between the off-gases, slag , and matte is 19.48%, 24.91% and 55.61%, respectively, as reported in Table 8. The predicted results are 19.82%, 23.47% and 56.71%, which are very close to the plant data. In the SKS process , most of the lead enters the slag phase in the form of PbO. As the matte grade increases, the proportion of PbO in slag becomes larger, whereas the proportions in both the off-gas and matte decrease. When the process deals with copper concentrates with high lead contents, lead can be removed by an iron silica slag and by increasing the matte grade or the amount of oxygen blown into the furnace .

../images/468727_1_En_36_Chapter/468727_1_En_36_Fig6_HTML.gif — Fig. 6
Lead (Pb) distribution in SKS smelting process

Zinc is an associated element in copper sulfide minerals . The distribution of zinc between gas, slag and matte as a function of matte grade is presented in Fig. 7. With a matte grade of 70%, two thirds of the zinc reports to the slag in the form of ZnO. As the matte grade increases, ZnO in slag becomes larger, similarly to lead . However, the presence of ZnO in the copper smelting slag significantly increases the liquidus temperature in the spinel primary phase field, which further increases the slag viscosity and therefore the copper losses to slag . It is difficult to recover the zinc in slag . Only a small proportion of zinc in the dust can be recovered through hydrometallurgical method. Based on the above reasons, the content of zinc in copper concentrates should be limited as much as possible.

../images/468727_1_En_36_Chapter/468727_1_En_36_Fig7_HTML.gif — Fig. 7
Zinc (Zn) distribution in SKS smelting process

As₂O₃ is a toxic substance, which can cause severe environmental pollution and serious harm to people’s health. With more restrictive environment protection regulations, the industry experiences a growing interest in the deportment of arsenic between various phases during the processing of copper concentrates. Figure 8 presents the variation of the calculated equilibrium distribution of arsenic between the gas, slag and matte as a function of matte grade. In the SKS smelting process, more than 80% of the arsenic reports to the gas. Its distribution between slag and matte is about 12% and 6%, respectively. Concentrating the arsenic to the gas phase is a feature of the SKS smelting process while the flash smelting process does just the opposite. With a increase of matte grades, the proportion of arsenic reporting to the matte phase also increases. Therefore, a high matte grade is not favorable to the removal of arsenic . Dealing with complex copper concentrates of high arsenic content, the matte grade should be reduced appropriately. In Yantai Hengbang smelter [14], the low-grade matte (around 50%) is adopted to deal with complex copper concentrates with 1.5–2% As, which results in about 90% of the arsenic reporting to the gas.

../images/468727_1_En_36_Chapter/468727_1_En_36_Fig8_HTML.gif — Fig. 8
Arsenic (As) distribution in SKS smelting process

The distribution of antimony is given in Fig. 9. The antimony distribution between gas, slag and matte does not show an obvious trend as a function of matte grades. About 70% of the antimony reports to slag . Therefore, it is difficult to remove antimony from the matte by only adjusting the matte grade in a SKS smelting process. However, antimony can affect the physical property of the anode copper and the performance of the electrolytic refining . For these reasons, the method to reduce the effect of antimony on copper production is to restrict the content of antimony in the mixed copper concentrates before feeding into SKS furnace by blending raw materials.

../images/468727_1_En_36_Chapter/468727_1_En_36_Fig9_HTML.gif — Fig. 9
Antimony (Sb) distribution in SKS smelting process

The calculated distribution of bismuth is given in Fig. 10 and shows a good match with the industrial data. Under normal operational conditions and with a matte grade of 70%, most of the bismuth reports to the gas phase, which is similar to arsenic . The proportion of bismuth reporting to the matte phase increases with an increase of matte grade, which is also similar to the variation trend of arsenic . Bismuth can increase the brittleness of anode copper and reduce its ductility, which leads to cracking of the copper anodes during the casting and electrolytic processes. When dealing with copper concentrates of high bismuth content, the matte grade could be reduced by adjusting the oxygen to ore ratio.

../images/468727_1_En_36_Chapter/468727_1_En_36_Fig10_HTML.gif — Fig. 10
Bismuth (Bi) distribution in SKS smelting process

Conclusions

A computational thermodynamics model for the SKS process was developed based on the mechanism characteristics of the SKS process and the theory of Gibbs free energy minimization . A good agreement was obtained between the calculation results and the actual plant data, which validated the reliability of the our model of the SKS process . The model was successfully used to predict the copper losses to slag and the distribution of several minor elements (such as Pb, Zn, As, Sb and Bi) between the off-gas, iron silicate slag , and matte phases as a function of copper matte grade by adjusting the ratio oxygen to ore charged in the SKS furnace . We believe our model helps to deepen the understanding of the SKS process and is of great significance to further optimize the operation parameters and to regulate production when the smelter deals with complex copper concentrates with high minor elements content.

Acknowledgements

The authors would like to acknowledge the National Nature Science Foundation of China (No. 51620105013) for financial support, and Dongying Fangyuan Nonferrous Metals Co ., Ltd. for providing the production data.

References

1.
Chaubal PC, Sohn HY, George DB (1989) Mathematical modeling of minor-element behavior in flash smelting of copper concentrates and flash converting of copper mattes. Metall Trans B 20(1):39–51Crossref
2.
Chen M, Cui ZX, Zhao BJ (2015) Slag chemistry of bottom blown copper smelting furnace at Dongying Fangyuan. Paper presented at the 6th international symposium on high-temperature metallurgical processing, Orlando, Florida, USA, 15–19 March 2015
3.
Cui Z (2016) Two-step copper smelting process at Dongying Fangyuan. Paper presented at the 9th international copper conference (Copper 2016), Kobe, Japan, 13–16 November 2016
4.
Guo XY, Wang QM, Tian QH (2014) Mechanism and multiphase interface behavior of copper sulfide concentrate smelting in oxygen-enriched bottom blowing furnace. Nonferrous Met Sci Eng 5:28–34
5.
Guo XY, Wang QM, Liao LL (2014) Mechanism and multiphase interface behavior of copper sulfide concentrates melting in oxygen-enriched bottom blowing furnace. Nonferr Met Sci Eng 5(5):28–34
6.
Guo XY (2016) Advanced copper smelting technologies used to quadruple China copper production between 2000 and 2015. Paper presented at the 9th international copper conference (Copper 2016), Kobe, Japan, 13–16 November 2016
7.
Guo XY, Wang QM, Tian QH, Zhang BJ (2016) Performance analysis and optimization of oxygen bottom blowing copper smelting process. Chin J Nonferr Met 26:689–698
8.
Lide DR (2004) CRC handbook of chemistry and physics. CRC Press Inc., Boca Raton
9.
Liao LL (2016) Development and application of simulation platform in the oxygen bottom blowing copper smelting process. Central South University, Changsha
10.
Liu HQ, Cui ZX, Chen M (2016) Phase equilibrium study of ZnO–‘‘FeO’’–SiO₂ System at Fixed Po₂ 10⁻⁸ atm. Metall Mater Trans B 47:164–177Crossref
11.
Nagamori M, Mackey PJ (1989) Thermodynamics of copper matte converting: Part I. Fundamentals of the noranda process. Metall Trans B 9(1):255–265
12.
Nagamori M, Errington WJ, Mackey PJ (1994) Thermodynamic simulation model of the Isasmelt process for copper matte. Metall Mater Trans B 25B:839–853Crossref
13.
Néron A, Lantagne G, Marcos B (2012) Computation of complex and constrained equilibria by minimization of the Gibbs free energy. Chem Eng Sci 82:260–271Crossref
14.
Qu S, Dong Z, Chen T (2016) Distribution of minor elements in complex copper concentrates in oxygen-enriched bottom blown smelting process. China Nonferr Metall 3:22–24
15.
Rossi C, Cardozo-Filho L, Guirardello R (2009) Gibbs free energy minimization for the calculation of chemical and phase equilibrium using linear programming. Fluid Phase Equilibr 278:117–128Crossref
16.
Shimpo R, Watanabe Y, Goto S, Ogawa O (1983) An application of equilibrium calculations to the copper smelting operation. Adv Sulfide Smelt 1:295–316
17.
Shimpo R, Goto S, Ogawa O (1986) A study on equilibrium between copper matte and slag. Can Metall Q 25:113–121Crossref
18.
Shui L, Cui ZX, Ma XD (2015) Mixing phenomena in a bottom blown copper smelter: a water model study. Metall Mater Trans B 46B:1218–1225Crossref
19.
Shui L, Cui ZX, Ma XD (2016) Understanding of bath surface wave in bottom blown copper smelting furnace. Metall Mater Trans B 47:135–144Crossref
20.
Seo KW, Sohn HY (1991) Mathematical modeling of sulfide flash smelting process. Part III: volatilization of minor elements. Metall Trans B 22(6):791–799Crossref
21.
Takeda Y, Ishiwata S, Yazawa A (1983) Distribution equilibria of minor elements between liquid copper and calcium ferrite slag. Trans Jpn Inst Met 24(7):518–528Crossref
22.
Tan PF, Zhang CF (1997) Computer model of copper smelting process and distribution behaviors of accessory elements. J Cent South U T 4(1):36–41Crossref
23.
Wang S (2016) Copper smelting: 2016 world copper smelting data. Paper presented at the 9th international copper conference (Copper 2016), Kobe, Japan, 13–16 November 2016
24.
Wang C, Zhang CF (2016) Study on choosing the best grade matte copper smelting process. World Nonferr Met 9:21–22
25.
Wu W (2007) Mathematical model research and system development of multiphase equilibrium in copper flash smelting. Jiangxi University of Science and Technology, Ganzhou
26.
Yan J (2016) Recent operation of the oxygen bottom-blowing copper smelting and continuous copper converting technologies. Paper presented at the 9th international copper conference (Copper 2016), Kobe, Japan, 13–16 November 2016
27.
Yazawa A, Nakazawa S, Takeda Y(1983) Distribution behaviour of various elements in copper smelting systems. In: Advances in Sulfide Smelting, Warrendale, PA, US, pp 99–117
28.
Zhang ZY, Chen Z, Yan HJ (2012) Numerical simulation of gas-liquid multi-phase flows in oxygen enriched bottom-blown furnace. Chin J Nonferr Met 22:1826–1834
29.
Zhang ZY, Gao Q, Liu L (2012) Influence of lance arrangement on bottom-blowing bath smelting process. Chin J Nonferr Met 22:2393–2400