Issue 
Metall. Res. Technol.
Volume 116, Number 1, 2019



Article Number  112  
Number of page(s)  11  
DOI  https://doi.org/10.1051/metal/2018050  
Published online  18 December 2018 
Regular Article
Performance simulation and optimization of COREX shaft furnace with central gas distribution
Laboratory of Multiphase Transport and Reaction Engineering, Department of Ferrous Metallurgy, School of Metallurgy, Northeastern University, P. O. Box 312, Northeastern University,
311 Wenhua Road, Heping District,
110819
Shenyang,
Liaoning Province, PR China
^{*} email: zouzs@mail.neu.edu.cn
Received:
13
March
2018
Accepted:
22
May
2018
To improve the distribution of the reducing gas inside the shaft furnace, recently, a central gas distribution (CGD) device is installed at the bottom of the shaft furnace in Bayi Steel. In this work, a twodimensional steadystate mathematical model, including momentum, heat and mass transfers and chemical reaction, is developed to study the performance of the furnace with CGD. The model is validated by industrial production data of COREX3000, and then it is used to study the detailed phenomena in the furnace with CGD. The results show that, compared with the traditional COREX shaft furnace without CGD, the CGD device has a positive effect on improving the solid metallization rate (MR) and gas utilization rate (UR). The solid metallization rate (MR) increases from 0.601 to 0.628, the gas utilization rate (UR) increases from 0.3437 to 0.3519 and the standard deviation (STDEV) of solid MR decreases from 0.0820 to 0.0037, which means that the MR of the DRI from the furnace with CGD is more uniform than that without CGD. In addition, the effects of operation conditions and the CGD design on the solid MR, gas UR and STDEV of MR are discussed and the optimized conditions are suggested.
Key words: COREX shaft furnace / central gas distribution (CGD) / numerical simulation / gas utilization / solid metallization
© EDP Sciences, 2018
1 Introduction
With the development of ironmaking technologies and the increasing resource and environment demands, many alternative ironmaking processes have been developed in recent decades, such as COREX, Midrex, HYL, Finex, etc. [1–4]. The COREX process, which has been industrialized successfully, has been developed three different models of COREX1000 in South Africa, COREX2000 in Korea and COREX3000 in China [5,6]. However, differing from the traditional BF ironmaking process, the COREX3000 faces many problems especially in the shaft furnace due to the lack of production experience, such as low metallization rate of DRI, poor distribution of reducing gas and sticking of burden, etc. [7,8]. To overcome these problems, two parallel beams for areal gas distribution (AGD) were installed in the shaft furnace in Baosteel [9], but it caused other problems. The most urgent issue was the dust accumulation in the bustle pipe which directly affects the smooth operation of the shaft furnace. When one set of COREX was moved from Baosteel to Bayi Steel, one AGD beam, instead of two beams, was used to decrease the dust accumulation. However, new security problems arose as the AGD beam might not be strong enough to bear the gravity of burden. To solve these new problems, as well as to further improve the gas distribution, recently, a new technique, socalled the central gas distribution (CGD) device is installed at the bottom of the shaft furnace in Bayi Steel. Due to the short development history, the transport phenomena inside the shaft furnace with CGD are not clear. Therefore, it is necessary to make a thorough study on the transport and reaction phenomena in the furnace with CGD.
Because of the enclosed structure and severe environment in the furnace, it is difficult to get detailed information directly by measurement. With the development of computational fluid dynamics, the numerical simulation technique has been widely used to solve such problems. Dong et al. [10] developed a twodimensional mathematical model to investigate the transport phenomenon by the potential flow theory that assumed the velocity of solid burden as fixed value in the axial direction. Compared with the method of the potential flow theory, the Eulerian method was widely used to investigate the transport phenomena in a shaft furnace. Wu et al. [11] developed a twodimensional mathematical model by the Eulerian method, which considered the transfers of momentum, heat and mass between burden and reducing gas, to study the distribution of velocity, pressure, temperature and species and investigated the effect of the DRI down pipe gas on the burden metallization rate. Xu et al. [12,13] successively developed a twodimensional mathematical model and a threedimensional fullscale mathematical model by the Eulerian method, which considered the transfers of momentum, heat and mass, to study the reduction behaviors in the shaft furnace and obtained some optimal operation conditions. Based on Xu’s work, Kou et al. [14] developed a threedimensional quarter model by Eulerian method to study the effect of operational parameters on the characteristics of gas and solid flow inside the COREX shaft furnace. Furthermore, based on the above works, Wu et al. [15] developed a threedimensional mathematical model with the Eulerian method to investigate the effect of AGD on the characteristics inside the furnace, which considered the distribution of velocity, temperature and species, and recommended an optimal AGD arrangement. Additionally, Hou et al. [16] developed a twodimensional slot model to study the momentum and heat transfers in the shaft furnace with CFDDEM, but this method needed massive computation. Recently, Zhou et al. [17] used the DEM approach to investigate the solid flow in the shaft furnace with CGD, but they did not consider the chemical reaction and the transfers of momentum, heat and mass between the gas and solid phases. Even though many researches have been done on the shaft furnace, few researches are made on the momentum, heat and mass transfer behavior in the shaft furnace with CGD.
In this work, a twodimensional steady state mathematical model with the Eulerian method is developed to investigate the characteristics inside the COREX shaft furnace, which considers the momentum, heat and mass transfers between the gas and solid phases. The model is validated by the industrial production data of the COREX3000, and then it is used to study the transport phenomena and reduction reaction inside the shaft furnace with CGD. In addition, the effects of operation conditions and the CGD design on the solid MR, gas UR and the uniformity of MR are discussed and the optimized conditions are suggested.
2 Model formulation
Based on the characteristics of Eulerian twophase model, in this work, the gas and solid phases are both considered as continuous fluid phases. The transport of mass, momentum and heat, as well as species in steady state can be described by the general conservation equation of equation (1). (1)
In equation (1), the subscript p denotes the considered phase, D and S represent the diffusivity and source term, respectively. All of variables ϕ are listed in Table 1.
The chemical reactions considered in the model are listed in Table 2, which adopted the threeinterface unreacted core model. Based on the works of Hara et al. and Austin et al. [22,23], the reaction rate of the threeinterface unreacted core model is expressed by equation (2). The reaction rate constants, chemical equilibrium constants and effective diffusivities all come from the work of Takahashi et al. [24,25]. A modification is made that the intraparticle diffusivity for the reduction step from FeO to Fe should multiply onetenth in the range of the reduction degree over 75% [15]. All of the physical and thermodynamic data of involved substances come from the work of Perry et al. [26]. (2)
Chemical reactions and rate expressions in the model.
3 Boundary conditions
The geometrical structures of the traditional shaft furnace and the shaft furnace with CGD are shown in Figure 1. The main boundary conditions are based on the operation conditions of the COREX shaft furnace, which are listed in Table 3. The main feature of gas and solid movement inside the furnace is of a counterflow. The burden is charged from the top of the furnace continuously, then reduced by hightemperature reducing gas progressively to DRI in the furnace, and finally discharged by screws at the bottom of the furnace. The reducing gas is introduced from the bustle pipe and the CGD, and is discharged from the top of the shaft furnace. In the present model, all of the wall boundaries are considered as the noslip condition. The thermal boundary of the wall is equivalently set as temperature and heat transfer coefficient of 300 K and 20 W/(m^{2}K), respectively [11]. The density, viscosity, thermal conductivity and particle diameter of burden are set as 2273 kg/m^{3}, 6 kg/(m s), 0.8 W/(m K) and 0.0124 m, respectively [11,15]. The following assumptions are made for the model:

the porosity of the packed bed is 0.4;

the shape of solid phase is spherical;

the species of gas and solid phases include CO, CO_{2}, H_{2}, H_{2}O and Fe_{2}O_{3}, Fe_{3}O_{4}, FeO, Fe;

the chemical reactions considered in the model are R_{1}R_{7} as listed in Table 2.
The main numerical method in this model is the finitevolume method. The conservation equations of the mass, momentum and heat are discretized by the first order upwind scheme and the solving method is the phase coupled SIMPLE method. The numerical calculation in this model is carried out based on the platform of the commercial CFD software, ANSYSFluent, and considered as converged when the residuals of all the variables are less than 1 × 10^{−5}. In order to ensure the numerical stability and efficiency, the gridindependence test is conducted and the number of cells for case furnace without and with CGD is 10601 and 12559, respectively.
Fig. 1 Geometrical structures and boundary conditions of traditional shaft furnace and the shaft furnace with CGD. 
4 Results and discussion
4.1 Model validation
The comparison between the simulated results and the practical data is listed in Table 4, including top gas composition and solid MR. The solid MR and relative error are expressed respectively by equations (3) and (4) [15]. (3) (4)
As shown in Table 4, the maxin is 6.9%, which is the average volume fraction of CO_{2} at the top of the shaft furnace. The relative error of the average solid MR at the bottom of the shaft furnace is 1.5%. For the error between the simulated value and the measured value, the main reasons are discussed as follows. Firstly, in the condition of industrial production, the burden diameter and porosity of packed bed would change with the burden moving down. Secondly, in this model, the burden profile at the top of the shaft furnace is assumed as flat that is different from the practical condition. The burden profile would affect the distribution of gas at the top of the shaft furnace slightly. Lastly, the composition of reducing gas and burden in practical condition would include a small quantity of other species, such as CH_{4}, N_{2}, CaO, MgO and SiO_{2}, etc. However, even though the difference between the simulated results and the practical data exists, the model in this work is reasonably reliable and can be used to predict the inner characteristics of the shaft furnace with CGD.
Comparison between measured data and simulated results.
4.2 Comparison of inner characteristics between shaft furnaces with and without CGD
Based on the practical pressure condition of CGD and bustle pipe, the corresponding gas flow rates are calculated, arriving at that the volume fraction of reducing gas input from CGD is 12∼15%. Fifteen percent of reducing gas input via CGD is used in the following simulation study.
4.2.1 Velocity distribution
The velocity vectors of gas and solid phases in the two types of shaft furnaces are shown in Figure 2. In the traditional shaft furnace, all the reducing gas is introduced from the bustle pipe and the gas velocity decreases sharply after entering the furnace. A stagnant zone is formed at the bottom of the furnace and thus leading to the nonuniform distribution of gas flow. On the contrast, in the shaft furnace with CGD, because the reducing gas is introduced from both the bustle pipe and the CGD, the distribution of gas flow is more uniform and the stagnant zone at the bottom is much smaller.
For the velocity vectors of solid phase in the traditional shaft furnace, as the burden is charged from the top of the shaft furnace, the velocity of solid phase decreases with the inner radius of furnace increasing. On the contrast, the burden velocity increases gradually with the inner diameter decreasing below the bustle pipe level and reaches to about 0.0008 m/s at the bottom of furnace. Because of the “umbrella effect” of the CGD device, small inactive zones of solid flow occur above and below the CGD device. Actually, comparing the velocity vectors of the solid phase in the two types of the shaft furnaces, the difference is not very significant.
Fig. 2 Velocity vectors of gas and solid phases in the two types of shaft furnaces. 
4.2.2 Temperature distribution
With cold burden being charged continuously from the top of the shaft furnace, strong convective heat exchange occurs between the gas and solid phases at the top section of the furnace. Therefore, both the gas and solid phase temperatures change rapidly and the gas temperature is higher than the solid phase. With the burden moving down, the temperature difference between gas and solid phases is very small at the bottom of the furnace.
The temperature distributions of gas and solid phases in the two types of shaft furnaces are shown in Figure 3. Compared with the traditional shaft furnace, because part of reducing gas is introduced from CGD, the distribution of gas flow in the furnace with CGD is more uniform and thus reduction reactions R_{3} and R_{6} would occur more at the center of the furnace with CGD. Reaction R_{3} is a weak exothermic reaction while reaction R_{6} is strong endothermic. Therefore, a low temperature zone appears in the center of the furnace with CGD.
Fig. 3 Temperature distributions of gas and solid phases in the two types of shaft furnaces. 
4.2.3 Distribution of gas composition
The mole fraction distributions of gas species (CO, CO_{2}, H_{2} and H_{2}O) in the two types of shaft furnaces are shown in Figure 4. In the traditional shaft furnace, because all the reducing gas is introduced from the bustle pipe, the temperature near the bustle pipe is higher and the reduction reaction is faster than other places. Therefore, the mole fraction of CO and H_{2} decrease rapidly near the bustle pipe and the variation tendencies of CO_{2} and H_{2}O are opposite.
For the shaft furnace with CGD, due to part of reducing gas is introduced from CGD, compared with the traditional shaft furnace, high concentration regions of CO and H_{2} appear near the bustle pipe and the CGD device. With the reactions going on, the mole fractions of CO and H_{2} decrease from the bustle pipe and the CGD device, and correspondingly, those of CO_{2} and H_{2}O increase gradually.
Fig. 4 Gas species distributions in the two types of shaft furnaces. 
4.2.4 Distribution of solid composition
According to the features of the threeinterface unreacted core model, the solid burden is reduced as Fe_{2}O_{3} → Fe_{3}O_{4} → FeO → Fe successively by the reducing gas. Based on the thermodynamic characteristics of R_{1}R_{6}, the reaction equilibrium constants K_{1} and K_{4} are much bigger than the others. Therefore, with the solid burden descending down gradually, the burden is reduced from Fe_{2}O_{3} to Fe_{3}O_{4} rapidly on the top of the shaft furnace. Because the reaction equilibrium constants K_{3} and K_{6} are very small, the proceeding of reactions R_{3} and R_{6} needs high temperature and high concentration of reducing gas.
The mass fraction distributions of solid species (Fe_{2}O_{3}, Fe_{3}O_{4}, FeO and Fe) in the two types of shaft furnaces are shown in Figure 5. For the traditional shaft furnace, because all the reducing gas is introduced from the bustle pipe, FeO is mainly distributed in the center of the furnace and Fe is mainly distributed along the furnace wall under the bustle pipe. The maximum value of Fe mass fraction is 0.621 and the average value is 0.542 at the furnace bottom.
For the shaft furnace with CGD, because the reducing gas is introduced from both the CGD device and the bustle pipe, the distribution of reduction reaction is more uniform and Fe distributes more uniformly also at the bottom of the furnace. The maximum value of Fe mass fraction is 0.576 and the average value is 0.568.
Fig. 5 Solid species distributions in the two types of shaft furnaces. 
4.2.5 Distributions of solid metallization rate (MR) and gas utilization rate (UR)
The solid MR and gas UR are important indexes for the shaft furnace production. High solid MR and high gas UR mean that the shaft furnace is in a stable and efficient working state.
According to equation (3), the distribution of solid MR is similar to the distribution of the Fe. With the solid burden moving down, the solid MR increases gradually as more and more FeO is reduced to Fe. The distribution of solid MR in the two types of shaft furnace is shown in Figure 6 (a_{1}, b_{1}). In the traditional shaft furnace, the maximum value of solid MR at the bottom of the furnace is 0.678 and the average value is 0.601. On the other hand, due to part of reducing gas is introduced from the CGD device, the distributions of reduction reaction, and thus those of species are more uniform in the furnace, leading to that the maximum value of solid MR is 0.636 and the average value is 0.628 at the bottom of the furnace. The STDEV of solid MR decreases from 0.0820 to 0.0037, which means that the MR of the DRI from the furnace with CGD is more uniform than that without CGD.
According to equation (5), the distribution of mixture gas UR is opposite to the tendency of the mole fraction of CO and H_{2}. The distributions of gas UR in the two types of shaft furnaces are shown in Figure 6 (a_{2}, b_{2}).With the reducing gas moving up, CO and H_{2} decrease continuously and CO_{2} and H_{2}O increase correspondingly. Therefore, the gas UR is the highest at the top of the shaft furnace. In the traditional shaft furnace, the gas UR at the top of the furnace is 0.3437. For the shaft furnace with CGD, the gas UR is 0.3519. (5)
Fig. 6 Distributions of solid MR and gas UR in the two types of shaft furnaces. 
4.3 Influences of operation condition and CGD design
The effects of operation conditions (fraction and temperature of CGD gas input) and CGD design (height and diameter) on solid MR, gas UR and STDEV of MR are discussed in this section and the optimization results are also provided.
4.3.1 Effect of CGD gas input on solid MR, gas UR and STDEV of MR
The effect of CGD gas input on solid MR, gas UR and STDEV of MR is shown in Figure 7. With the CGD gas input increasing from 0 to 100%, the solid MR increases from 0.601 to 0.657, gas UR increases from 0.3437 to 0.3584, but the STDEV of MR decreases first from 0.0820 to 0.0088 at 20% CGD gas input and then increases to 0.0271. Especially, as the CGD gas input increases from 0 to 40%, the solid MR increases rapidly from 0.601 to 0.642. Simultaneously, the gas UR increases from 0.3437 to 0.3565 and the STDEV of MR decreases from 0.0820 to 0.0083. Further increasing of CGD gas input has much less effect. The results reveal that a reasonable CGD gas input is not only favor for reduction reactions in the shaft furnace but also the uniformity of the DRI MR. Therefore, the recommended CGD gas input is 40%.
Fig. 7 Effect of CGD gas input on solid MR, gas UR and STDEV of MR. 
4.3.2 Effect of CGD gas temperature on solid MR, gas UR and STDEV of MR
The temperature of CGD gas input is different from that of bustle gas, because the former comes directly from the melter gasifier. Therefore, it is necessary to understand the effects of CGD gas temperature on solid MR, gas UR and STDEV of MR. The effects of CGD gas temperature on solid MR, gas UR and STDEV of MR are shown in Figure 8. With the increasing of CGD gas temperature from 1000 K to 1300 K, the solid MR increases from 0.624 to 0.639 and the gas UR increases from 0.3508 to 0.3537. Temperature increasing is favorable for the kinetics of all reduction reactions, as well as the thermodynamics of reaction R_{6}. It should be noticed that the STDEV of MR decreases from 0.0159 to 0.0037 as the temperature increases from 1000 K to 1100 K and then increases from 0.0037 to 0.0213 as the temperature increases further from 1100 K to 1300 K. It means that too low and too high temperatures of CGD gas input are not good to the uniformity of DRI. In addition, the temperature of reducing gas is limited by the softening and sticking of burden in the furnace, and thus it should be controlled below 1200 K.
Fig. 8 Effect of CGD gas temperature on solid MR, gas UR and STDEV of MR. 
4.3.3 Effect of CGD height on solid MR, gas UR and STDEV of MR
The effect of CGD height on solid MR, gas UR and STDEV of MR is shown in Figure 9. With the increasing of CGD height from 3653 mm to 5653 mm, the solid MR decreases from 0.628 to 0.60, the gas UR decreases from 0.3519 to 0.3474 and the STDEV of MR increases from 0.0037 to 0.0065. With the increasing of CGD height, the effective volume of the furnace and thus the residence time of reducing gas decreases, leading to decreasing of solid MR and gas UR, and the increasing of STDEV of MR. Therefore, the recommended CGD height is as low as possible, i.e. 3653 mm of the current design.
Fig. 9 Effect of CGD height on solid MR, gas UR and STDEV of MR. 
4.3.4 Effect of CGD diameter on solid MR, gas UR and STDEV of MR
The effect of CGD diameter on the solid MR, gas UR and STDEV of MR is shown in Figure 10. With the increasing of CGD diameter from 2500 mm to 4500 mm, the solid MR increases from 0.628 to 0.630, the gas UR increases from 0.3519 to 0.3524 and the STDEV of MR decreases from 0.0037 to 0.0011. Although increasing of the CGD diameter has a little positive effect on the distribution of gas flow and the uniformity of DRI, too large CGD diameter would interfere the burden flow and thereby the smooth running of the shaft furnace. Therefore, the recommended CGD diameter is 3500 mm.
Fig 10 Effect of CGD diameter on solid MR, gas UR and STDEV of MR. 
5 Conclusions
A twodimensional steadystate mathematical model is established to analyze the influence of the central gas distribution device on the inner characteristics of the COREX shaft furnace, which included the momentum, heat and mass transfers and chemical reactions between the solid burden and the reducing gas. Comparing the inner characteristics of the two types of shaft furnaces, the effect of CGD on the performance of the furnace is discussed. The main conclusions are as follows:

compared with the traditional shaft furnace, the gas flow distribution in the shaft furnace with CGD is improved significantly and the stagnant zone at the bottom of the furnace is reduced. Because of the “umbrella effect” of the CGD device, small inactive zones of solid flow appear above and below the CGD device;

as part of the reducing gas is introduced from CGD, the reduction of FeO to Fe with H_{2}, which is strong endothermic, takes place significantly in the center of the furnace and thus leading to a low temperature zone in the furnace center;

the effect of CGD on solid MR and gas UR is significant. Compared with the furnace without CGD, the solid MR and gas UR of the furnace with CGD increase from 0.601 to 0.628 and from 0.3437 to 0.3519, respectively. The STDEV of MR decreases from 0.0820 to 0.0037, which means that the DRI product from the furnace with CGD is more uniform than that without CGD;

the effect of CGD gas input on the solid MR, gas UR and STDEV of MR are significant. With the CGD gas input increasing from 0 to 100%, the solid MR increases from 0.601 to 0.657, the gas UR increases from 0.3437 to 0.3584, but the STDEV of MR decreases first from 0.0820 to 0.0088 at 20% CGD gas input and then increases to 0.0271. The recommended CGD gas input is 40%;

the effect of CGD gas temperature on the solid MR and gas UR is monotonically positive. With the CGD gas temperature increasing from 1000 K to 1300 K, the solid MR increases from 0.624 to 0.639 and the gas UR increases from 0.3508 to 0.3537. However, neither too high or too low temperature is good to the uniformity of DRI. In addition, to prevent the burden sticking, the temperature of CGD gas should be controlled below 1200 K;

the height and diameter of the CGD device play important roles in the performance of the furnace, but the effect of diameter is smaller than that of height. With the height of CGD increasing from 3653 mm to 5653 mm, the solid MR decreases from 0.628 to 0.60, the gas UR decreases from 0.3519 to 0.3474 and the STDEV of MR increases from 0.0037 to 0.0065. With the diameter of CGD increasing from 2500 mm to 4500 mm, the solid MR increases from 0.628 to 0.630, the gas UR increases from 0.3519 to 0.3524 and the STDEV of MR decreases from 0.0037 to 0.0011. The recommended CGD height and diameter are 3653 mm and 3500 mm, respectively.
Nomenclature
C_{P,i}: Specific heat capacity of species i (J/(kg · K)
D_{i}: Mass diffusion coefficient of species i (m^{2}/s)
d_{s}: Diameter of solid phase (m)
E_{gs}: Coefficient of convective heat transfer (W/ (m^{2} · K)
: Drag force between gas and solid phases (N/m^{2})
: Gravitational acceleration (m/s^{2})
H_{p}: Enthalpy of phase p (J/kg)
ΔH_{n}: Enthalpy of reaction n (J/kg)
K_{n}: Equilibrium constant of reaction n (‒)
k_{i}: Thermal conductivity of species i (W/(m · K)
k_{n}: Rate constant of reaction n (‒)
M_{i}: Molecular weight of species i (kg/kmol)
M_{g}: Average molecular weight of gas phase (kg/kmol)
R_{n}: Chemical reaction rate of reaction n (kmol/(m^{3} · s)
Re_{s}: Relative Reynolds number based on solid phase diameter (‒)
S_{φ}: Source term of variable φ
T_{p}: Temperature of phase p (K)
W: Variable explained in reference [23]
Y_{i}: Mole fraction of species i (‒)
: Mole fraction of species i under equilibrium of reaction m (‒)
Greek Symbols
α_{n,m}: Variable explained in reference [23]
w_{i,j}: Mass fraction of species i in phase j (‒)
ϵ_{p}: Volume fraction of phase p (‒)
ρ_{p}: Density of phase p (kg/m^{3})
ϕ: General dependent variable in equation (1)
: Stress tensor of phase p (Pa)
φ: Gas composition in volume fraction (‒)
Subscripts
Abbreviations
Acknowledgement
The authors would like to thank the National Key Technology R&D Program during the “12th Fiveyear Plan” of China (Grant No. 2011BAE04B02) and the National Natural Science Foundation of China (Grant No. 51574064) for their financial support.
References
 B. Anameric, S.K. Kavatra, Direct iron smelting reduction process, Miner. Process. Extr. Metall. Rev. 30, 1–51 (2008) [CrossRef] [Google Scholar]
 C. Thaler, T. Tappeiner, J.L. Schenk, W.L. Kepplinger, J.F. Plaul, S. Schuster, Integration of the blast furnace route and the finexprocess for low CO_{2} hot metal production, Steel Res. Int. 83, 181–188 (2012) [CrossRef] [Google Scholar]
 X.L. Liu, G. Pan, G. Wang, Z. Wen, Mathematical model of lump coal falling in the freeboard zone of the COREX melter gasifier, Energ. Fuel 25, 5729–5735 (2011) [CrossRef] [Google Scholar]
 E. Ottenschlaeger, W. Kepplinger, G. Papst, R. Hauk, COREX process: Hot metal production on the basis of coal, Metall. Plant Technol. Int. 9, 24–30 (1986) [Google Scholar]
 N. Wang, X.M. Xie, Z.S. Zou, L. Guo, W.R. Xu, Y.S. Zhou, Analysis of material and energy consumption of COREX C3000, Steel Res. Int. 79, 547–552 (2008) [CrossRef] [Google Scholar]
 M.Y. Kou, S.L. Wu, G. Wang, B.J. Zhao, Q.W. Cai, Numerical simulation of burden and gas distribution inside the COREX shaft furnace, Steel Res. Int. 86, 686–694 (2015) [CrossRef] [Google Scholar]
 W.G. Li, Operation status and technical problems of COREX3000, The 3rd Baosteel Biennial Acad. Conf. Shanghai, A75–83 (2008) [Google Scholar]
 Q. Zhang, L. Guo, Analysis and improvement on Baosteel COREX C3000 shaft furnace operation, The 3rd Baosteel Biennial Acad. Conf. Shanghai, A219–223 (2008) [Google Scholar]
 J.J. Lin, W.G. Song, W.Y. Xia, Progress and improving directions of COREX3000 in Baosteel, The 5th Baosteel Biennial Acad. Conf. Shanghai, A499–506 (2013) [Google Scholar]
 X.F. Dong, ‘Simulation study on the process in the prereduction shaft furnace of COREX’, Master’s thesis, Northeastern University, Shenyang, China, 1998 [Google Scholar]
 S.L. Wu, J. Xu, S.D. Yang, Q. Zhou, L.H. Zhang, Basic characteristics of the shaft furnace of COREX smelting reduction process based on iron oxides reduction simulation, ISIJ Int. 50, 1032–1039 (2010) [CrossRef] [Google Scholar]
 J. Xu, S.L. Wu, X.Y. Guo, M.Y. Kou, Numerical simulation of thermal and iron ore reduction conditions in prereduction shaft furnace based on reducing gas composition and temperature, J. Shanghai Jiaotong Univ. (Sci.) 16, 375–379 (2011) [CrossRef] [Google Scholar]
 J. Xu, S.L. Wu, M.Y. Kou, K.P. Du, Numerical analysis of the characteristics inside prereduction shaft furnace and its operation parameters optimization by using a threedimensional full scale mathematical model, ISIJ Int. 53, 576–582 (2013) [CrossRef] [Google Scholar]
 M.Y. Kou, S.L. Wu, K.P. Du, W. Shen, X.D. Ma, M. Chen, B.J. Zhao, The effect of operational parameters on the characteristics of gassolid flow inside the COREX shaft furnace, J. Miner. Met. Mater. Soc. 67, 459–466 (2015) [CrossRef] [Google Scholar]
 S.L. Wu, K.P. Du, J. Xu, W. Shen, M.Y. Kou, Z.K. Zhang, Numerical analysis on effect of areal gas distribution pipe on characteristics inside COREX shaft furnace, J. Miner. Met. Mater. Soc 66, 1265–1276 (2014) [CrossRef] [Google Scholar]
 Q.F. Hou, M. Samman, J. Li, A.B. Yu, Modeling the gassolid flow in the reduction shaft of COREX, ISIJ Int. 54, 1772–1780 (2014) [CrossRef] [Google Scholar]
 H. Zhou, S.L. Wu, M.Y. Kou, Z.G. Luo, W. He, Z.S. Zou, Y.S. Shen, Discrete particle simulation of solid flow in a largescale reduction shaft furnace with center gas supply device, ISIJ Int. 58, 422–430 (2018) [CrossRef] [Google Scholar]
 X.F. Dong, A.B. Yu, J. Yagi, P. Zulli, Modelling of multiphase flow in a blast furnace: recent developments and future work, ISIJ Int. 47, 1533–1570 (2007) [CrossRef] [Google Scholar]
 S. Ergun, Fluid flow through packed columns, Chem. Eng. Prog. 48, 89–94 (1952) [Google Scholar]
 J. Yagi, Mathematical modeling of the flow of four fluids in a packed bed, ISIJ Int. 33, 619–639 (1993) [CrossRef] [Google Scholar]
 W.E. Ranz, W. R. Marshall, Evaporation from drops, Chem. Eng. Prog. 48, 141–173 (1952) [Google Scholar]
 Y. Hara, M. Sakawa, S. Kondo, Mathematical model of the shaft furnace for reduction of ironore pellet, TestutoHagané 62, 324–333 (1976) [CrossRef] [Google Scholar]
 P.R. Austin, H. Nogami, J. Yagi, A mathematical model for blast furnace reaction analysis based on the four fluid model, ISIJ Int. 37, 748–755 (1997) [CrossRef] [Google Scholar]
 R. Takahashi, Y. Takahashi, J. Yagi, Y. Omori, Operation and simulation of pressurized shaft furnace for direct reduction, Proc. of the 43rd Ironmaking Conf., Iron & Steel Soc. of AIME, Warrendale, PA, 43, 485–500 (1984) [Google Scholar]
 R. Takahashi, Y. Takahashi, J. Yagi, Y. Omori, Operation and simulation of pressurized shaft furnace for direct reduction, Trans. Iron Steel Inst. Jpn. 26, 765–774 (1986) [CrossRef] [Google Scholar]
 R.H. Perry, D.W. Green, J.O. Maloney, Perry’s Chemical Engineer’s Handbook, 7th ed., McGrawHill, New York, 1997 [Google Scholar]
Cite this article as: Xingsheng Zhang, Zhiguo Luo, Zongshu Zou, Performance simulation and optimization of COREX shaft furnace with central gas distribution, Metall. Res. Technol. 116, 112 (2019)
All Tables
All Figures
Fig. 1 Geometrical structures and boundary conditions of traditional shaft furnace and the shaft furnace with CGD. 

In the text 
Fig. 2 Velocity vectors of gas and solid phases in the two types of shaft furnaces. 

In the text 
Fig. 3 Temperature distributions of gas and solid phases in the two types of shaft furnaces. 

In the text 
Fig. 4 Gas species distributions in the two types of shaft furnaces. 

In the text 
Fig. 5 Solid species distributions in the two types of shaft furnaces. 

In the text 
Fig. 6 Distributions of solid MR and gas UR in the two types of shaft furnaces. 

In the text 
Fig. 7 Effect of CGD gas input on solid MR, gas UR and STDEV of MR. 

In the text 
Fig. 8 Effect of CGD gas temperature on solid MR, gas UR and STDEV of MR. 

In the text 
Fig. 9 Effect of CGD height on solid MR, gas UR and STDEV of MR. 

In the text 
Fig 10 Effect of CGD diameter on solid MR, gas UR and STDEV of MR. 

In the text 
Current usage metrics show cumulative count of Article Views (fulltext article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 4896 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.