Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
1 8nd International Conference on Physical and Numerical Simulation of Materials Processing, ICPNS’16 Seattle Marriott Waterfront, Seattle, Washington, USA, October 14-17, 2016 Numerical Simulation on the Combustion Characteristic of Iron Ore Sintering with Flue Gas Recirculation Gan Wang1, Zhi Wen1, 2, Guofeng Lou1, *, Ruifeng Dou1, Xunliang Liu1, Fuyong Su1, Sizong Zhang1 of Mechanical Engineering, University of Science and Technology Beijing, Beijing 100083, China. 2Beijing Key Laboratory of Energy Saving and Emission Reduction for Metallurgical Industry, University of Science and Technology Beijing, Beijing 100083, China. 1School ABSTRACT Flue gas recirculation sintering (FGRS) technology can reduce pollutant emissions and reuse waste heat effectively in iron ore sintering. The incoming gas conditions such as temperature, velocity, composition and contents, may differ across processes because the sources of recirculated gas vary. Numerical simulations have been developed to predict sintering behavior quantitatively. To model FGRS, reactions which O 2, CO2, CO, and H2O participate in as reactants and products should be taken into account. A relatively comprehensive mathematical model for a sinter bed, related to the application of FGRS and the comparison with the conventional sintering (CS), is built. Multiphase theory is employed. This model involves most of the major gaseous and gas-solid reactions that affected significantly by incoming gas conditions. Heat transfers are described in improved manners compared to the previously models. To date, six sinter pot tests based on FGRS technology have been used for model validation, and good agreements between simulated and measured results are obtained. Parametric studies focused on the quantitatively evaluation of the effects of various incoming gas conditions on the combustion characteristics in the sintering process, compared to CS. Results show FGRS can significantly increase the maximum temperature in the sinter bed, improve the uneven distribution of heat, but slightly reduce flame front speed. Thus, the quality of sintered ores, especially for the upper bed, are enhanced, while the productivity is restricted. Velocity exerts the most significant effect, followed by O2 content, and then, temperature. Keywords: Flue gas recirculation; Iron ore sintering; Numerical simulation; Multiphase; Gaseous reaction; Gassolid reaction; Sinter pot test; Combustion characteristics. 1. INTRODUCTION Iron ore sintering produces 20% pollutant of iron and steel industry. To reduce the emission of flue gas and reuse waste heat effectively, several types of flue gas recirculation sintering (FGRS) processes have been developed in last 20 years such as EOS, LEEP, Eposint. FGRS has been applied in industrial production in China since 2013, and five sets of systems have been built. The main incoming gas conditions for FGRS technology such as velocity, temperature, composition and contents, may differ across methods because the sources of recirculated gas vary. A series of effects in combustion characteristics are then generated during the sintering process, which is worth further study. Mathematical models have been developed to predict quantitatively the sintering performance. Most of these models are 1D transient process, and transfer phenomena along the directions of grate length and width are negligible. Shibata1) and Patisson et al.2) concentrated on predicting the moisture transfer process. Venkataramana et al.3) focused on the effects of operating parameters including suction applied, ignition time and ignition gas temperature, while Nath et al.4,5) and Pahlevaninezhad et al.8) analyzed on kinetic parameters such as coke contents, coke particles size, limestone particles size and velocity of the incoming air. Zhou et al.6) treated solid particles with characteristic size distributions. Zhao et al. 7) applied an granulation model to provide a novel description of coke positioning within granules. Ramos et al.9) described geometric changes by combining various reactions and gas-solid heat transfer with granule movement. Mitterlehner et al.10) regarded the increase in bed porosity as coke particle shrinkage with combustion. Yang et al.’s model11-13) considered * Corresponding author: Guofeng Lou. Tel.: +86 10 62332730; fax: +86 10 62329145; E-mail: [email protected]. 2 coke particle shrinkage, generation of internal pores, and porosity changes. Komarov et al.14) built a 2D model in which molten iron ores were regarded as the non-fluid medium. Ahn et al.15,16) conducted a commercial flowsheet process simulator to build a 2D model, which was the first reported research on the simulation study of FGRS process. Castro et al.17,18) developed a 3D model and used soft-melting experimental data to calculate geometric parameters. Yamaoka et al.19) built a 3D model in which five state parameters were selected to express the geometric structure of the bed. The current simulations lack of investigation on the effects of incoming gas conditions on sintering process. The purpose of this present study is to build a relatively more comprehensive mathematical model for a sinter bed related to the application of FGRS, with a special focus on the quantitatively evaluation of combustion characteristics in the sinter bed. Multiphase theory11-13,17,18), which treats sinter solid materials as multiple solid phases, is employed to describe heat transfer within/between different solid and gas phases in a better manner. This present model involves most of the major physicochemical changes that affected significantly by incoming gas conditions, many of these phenomena have not been considered in current models. To date, six sinter pot tests based on FGRS technology are used for validation studies. Then, four quantitative parameters, namely, maximum temperature (MaxT), flame front speed (FFS), combustion zone thickness (CZT), and melting zone thickness (MZT), are selected to study the differences of combustion characteristics between the FGRS and CS processes, as well as the effects of temperature, O2 content, and velocity of the incoming gas on it. This work lays the foundation for further optimization of the FGRS process. 2. Description of the Mathematical Model The sintering process in this study is considered a 1D transient process. Figure 1 shows that the 1D and 2D representations are undoubtedly6). Multiphase theory is applied, meaning each component of the solid phases, as shown in Table 1, has characteristic chemical compositions and particle sizes, different temperatures and physical properties. No temperature or species concentration gradient exists within a single particle because of the moderate thermal conductivity. The liquid phase, i.e., the molten iron ores, moves together with the remaining solid phase due to viscosity. Figure 1 Extension of 1D transient model to 2D steady model. 2.1 Governing Equations Table 2 shows the detailed governing equations for each phase, including continuity equations, energy equations, and the component equation simplified by Ergun’s pressure drop equation. The terms on the right side of solid energy equation comprise a diffusion term that includes conduction and radiation between solid phases, convection between solid and gas phases, the heat of various reactions, heat loss from the release of gas produced by the reactions, convection between solid phases, and radiation between solid particles (in the same solid phase), respectively. The terms on the right side of gas energy equation are similarly defined as the first four terms on the right side of solid energy equation. Table 1 Raw proportion and chemical composition of the solid phases (wt. %). Composition Solid Phases Proportion Fe total CaO MgO Al2O3 SiO2 MnO S Moisture Iron ore fines 62.48 59.62 1.42 0.76 1.60 4.73 0.24 0.02 7.95 Return fines 26.04 58.83 8.76 1.56 1.82 4.98 0.28 0.02 0.32 Burnt-lime 1.47 0.08 52.66 0.47 0.34 1.09 0.04 0.04 0 Limestone 4.15 0.11 52.91 1.07 0.57 1.89 0.02 0.05 5.2 Dolomite 2.56 0 31.47 20.65 0.38 0.74 0 0.01 3.67 Coke 3.3 0.42 0.49 0.05 3.96 5.45 0 0.75 13.62 3 Table 2 Governing equations for each solid and gas phase. Transfers Solid: Mass6,11) Gas: Equation 1 g Y j t g i t 1 g u s Yi g u g i x Solid: M j ,k R j ,k k j M i ,k Ri ,k k i 1 j s , j C ps , j Ts , j t x u s s , j C ps , j Ts , j x Ts , j s , j ,eff x x hconv, j Assa , j Tg Ts , j j M j ,k R j ,k H j ,k M j ,k R j ,k C ps , j Ts , j h jj j Assa , j Ts , jj Ts , j Q rad 11,12) k j Energy g C pg Tg t Gas: g C pg u g Tg jj j k j Tg hconv, j Assa , j Ts , j Tg g x x x k j 1 M i ,k Ri ,k H i ,k M j ,k R j ,k C ps , j Ts , j k Component20,21) P 323 H jd p 1 2 2 3 u g 3.78 2.2 Sub-models H2O content in sintering flue gas is as high as 10~14%, though mixed with ambient air before recirculation, the H2O content in recirculated gas should be higher than air, as similar to CO2 and CO. To model quantitatively and accurately the combustion characteristics of FGRS process and make some comparison with CS process, it is necessary to take reactions which O2, CO2, CO, H2O, H2 participate in as reactants and products into consideration in reaction sub-models. In this study, seven kinds of major gas-solid reactions and five major gaseous reactions that affected significantly by incoming gas conditions are considered to calculate the source terms of the conservation equations mentioned in Table 2. 2.1.1 Gas-solid Reactions Table 3 shows the seven major gas-solid reactions considered in this established model. Among these, drying of water in sintering raw mix is seen to occur in two distinct periods2, 6). The major solid-gas reactions in sinter bed are coke combustion and limestone calcination, which are regarded as unreacted-core shrinking models. A stoichiometric coefficient κ is proposed to characterize the degree of incomplete combustion of coke22). The kinetic parameters for coke gasification and combustion are listed in Table 4. Other reactions are simply described as functions of temperature and TGA conversion in previous studies24-27). Iron oxides are considered to be reduced by CO and H2 in a stepwise manner during sintering, and a three-step mechanism is postulated, given that the solid temperature lasts long above 570 °C. The kinetic jj j k j g 1 2 ug jd p 3 parameters for iron oxide reductions are listed in Table 5. Melting and solidification are controlled by bed temperature, whereas the initiation temperature of melting, Tm1, and its completion temperature, Tm2, are derived from the CaO-Fe2O3 phase diagram. Gas species participate in reactions as reactants or products. 2.1.2 Gaseous Reactions For a CS process, as ambient air used for the incoming gas, only convection heat transfer between high-temperature bed and low-temperature air occurs in the upper bed, i.e., the sintered zone. However, gaseous reactions in the gas phase in the sintered zone may also take place because of the presence of additive CO2, CO, H2O, NOx, and SO2 in the recirculated gas via FGRS process. Meanwhile, CO2, CO, H2O, H2, and CH4 generated via the coke reactions mentioned earlier flows downward through the sinter bed with the high-temperature gas. Then, the gaseous reactions in raw mix zone occur. The homogenous gas reactions are depended on the contents of the reactants and products in the gas atmosphere. The present model has considered wider range of these reactions, compared to the most of previous models, as listed in Table 65,8,10,28). 2.3 Heat Transfers In this work, one single solid particle has a representative temperature and gas concentration. Thus, the complicated modes of heat transfer in a sinter bed are summarized as convection/radiation between gas and solid phases, conduction/radiation among different solid phases or within the same solid phase, and conduction in the gas phases. 4 Table 3 Gas-solid reactions considered in the established model. Reactions Expression (1) Drying and Condensation H2O(l)↔H2O(g) Rdry H 2O Assa Rg Tg PH 2O PH 2,6) 2O , where, ω=W H2O/Wcr. When ω>=1, χ=1, when ω<1, χ=1-(1-ω)∙(1-1.796ω+1.0593ω2). Rconden H 2O Assa Rg Tg PH PH 2O 1 jd p k Sh D B e DO2 c ,i CO2 ,eff j Rcoke,i nC 4rc2 C i (2) Coke Gasification and Combustion5,8,11,22) C(s)+H2O(g)↔CO(g)+H2(g) C(s)+2H2(g)↔CH4(g) C(s)+CO2(g)↔2CO(g) κC(s)+O2(g)→2(κ-1)CO(g)+(2-κ)CO2(g) 2O , where, i represent O , 2 H2O, CO2 and H2, respectively, and rc r0 mc m0 1 3 , kc A Ts , j exp E Rg Ts , j , Sh 2 0.7 Re 0.7 nL 4rc2 CCO CCO 2 Rlim ist jd p 23) (3) Limestone Calcination CaCO3(s)→CaO(s)+CO2(g) Sc 2 r r 2 0 1.41 0 rc J DCO rc 2 Sh DCO2 ,eff 13 , r0 rc . 2 , where, CCO 2 4.1868K eq k R T c g s, j K eq 1000Rg Ts , j K eq 101325exp 7.0099 8202.5 Ts , j , kc 1.52103 exp 20143.4 Ts , j (4) Dolomite Calcination24) CaMg(CO3)2(s)→CaO(s)+MgO(s)+2CO2(g) 190.67 103 0.4 0.6 m m Rdolom 1.628107 exp c 0 Rg Ts , j (5) Combined Water Release25) Ca(OH)2(s)→CaO(s)+H2O(g) 280.4 103 P exp 138.5 103 Rg Ts , j Rdiss 1.18 1018 exp S 0 mc , where p* = Rg Ts , j Rg Ts , j 1.834 × 108 atm, S0 = (8.3 ± 1) × 103 m2∙kg−1. (6) Iron Oxides Reduction26,27) 3Fe2O3(s)→2Fe3O4(s)→6FeO(s)→6Fe(s) , Rreduc A exp E R g Ts , j m10 3 m1c 3 2 m10 3 M f minTs Tm1 Tm2 Tm1 ,1 while Ts Tm1 Rmelt 0.001 Ts Tm1 s (7) Melting and Solidification5, 6) Solid Mixture(s)→Molten(s,l)→Precipitated Minerals and Phase(s) Tm1 1380 21.22Al2O3 3.35SiO2 1.8 flux , where Al2O3, SiO2, and flux indicate the corresponding mass contents in raw mix. Table 4 Kinetic parameters for coke gasification and combustion. Vulue O2 H2O CO2 H2 A (s-1) 1.715 3.42 589 3.42×10-3 E (J∙mol-1) 74.83×103 129.7×103 222.82×103 129.7×103 Table 5 Kinetic parameters for different steps of iron oxide reductions. Fe2O3 → Fe3O4 Fe3O4 → FeO FeO → Fe Reducing substance A (s-1) E (J∙mol-1) A (s-1) E (J∙mol-1) A (s-1) E (J∙mol-1) CO 1.32×106 114~125.4×103 3.06×105 68.6×103 1.91×105 66×103 H2 3.49×106 75.9~137.9×103 3.63×107 96.2×103 2.19×106 75.5×103 Table 6 Rates of gaseous reactions considered in the model. Reaction Rgas (mol∙m-3∙s-1) A (s-1) E (J∙mol-1) (1) CO+0.5O2→CO2 A exp( E RgTg ) CCOCO0.52 CH0.52O 1.3×108 125.54×103 (2) CO2→CO+0.5O2 A exp( E RgTg ) CCO2 7.5×1011 386.6×103 (3) CO+H2O→CO2+H2 A exp( E RgTg ) CCOCH2O 2.78 12.55×103 (4) H2+CO2→CO+H2O A exp( E RgTg ) CH 2 CCO2 93.96 46.59×103 (5) H2+0.5O2→H2O A exp( E RgTg ) CH 2 CO2 1011 42×103 5 Table 7 Modes of heat transfer used in the established model. Heat transfers Expression hconv, j (1) Convection between solid and gas phase (2) Conduction and radiation between solid phases g Nu jd p g 1 1.5 1 j 2 Re0.6 Pr1 3 jd p s , j ,eff 1 j s , j 4 j d p , jTs3, j 29) h jj j 2 (3) Convection among solid phases30) Q rad Assa I I 31) (4) Radiation within the solid particles of the same phase The major modes of heat transfer considered are listed in Table 7. (1+1.5(1−εj)) is the correction factor based on the porous media assumption to calculate convective heat transfer between solid and gas phases. An effective thermal conductivity term λs,j,eff is used to calculate conductive and radiative heat transfer between solid phases29). A form of the convective heat transfer coefficient hjj-j is used to express heat exchange among solid phases30). The two-flux model is used to calculate radiative heat transfer within solid particles31). 3. Sinter Pot Test 1 j s , j s , j C ps , j j g g C pg 1 j Ts j Tg W m 3 1 j 1 jj until the flue gas temperature slightly increases to a peak value; time for cooling: after moving the recirculation hood and until flue gas temperature drops to below 100 °C; temperature for ignition: 1150 oC; suction negative pressure during ignition, sintering, and cooling: 8.83, 14.71, and 7.87 kPa, respectively. The velocity, temperature, and species concentrations of the simulated recirculated gas, the bed temperature profiles, as well as the temperature and species concentrations of the flue gas in the wind box are measured. In this study, the results for only two cases, i.e., CS and T200-19, are shown. In both the present study and the most of the previous work10-14), the mathematical models are validated by comparing the model predictions with the sinter pot test data. A detailed description of the conducted sinter pot tests is provided in the previous work32).The raw materials for tests are obtained on site from a single batch to ensure consistency in chemical composition and mix proportion, as also listed in Table 1. The schematic illustrations of the installation of the experimental facility and that of the sintering process are shown in Figure 2. The tests are categorized based on the temperature and O2 content of the recirculated gas by considering FGRS technology. The values are set as close as possible to the values used in industrial production conditions. And case CS is also investigated for comparison, as shown in Table 9. The following are the main laboratory scale operational conditions for the tests: time of ignition: 2 min; time for flue gas recirculation: after ignition and Figure 2 Schematic diagram of the facility installation for the sinter pot tests. Table 9 Experimental cases of the sintering pot tests. Case T150-19 T200-19 T250-19 T200-18 T200-20 CS Temperature (oC) 150 200 250 200 200 30 O2 content (Vol. %) 19 19 19 18 20 21 6 CO2 content (Vol. %) 3.88 3.64 4. Results and Discussion 4.1 Model Validation The experimental conditions of the sinter pot tests are used to validate the mathematical model in this study by defining the initial and boundary conditions. The contrasting results between the simulated and measured bed temperatures and the flue gas species concentrations for cases CS and T200-19, along with other fixed operating parameters, are shown in Figure 3. The simulated results of the bed temperatures are close to the experimental data with regard to MaxT at given locations and duration times at high temperatures. The times of temperature increase for the simulation show a slight difference with the measured ones at each corresponding location. This observation may be explained through the simplified analysis of bed shrinkage in this model, in which shrinkage presumably occurs only during melting and solidification. Bed shrinkage caused by combustion and decomposition reactions will be considered in a future work. The predicted flue gas species concentrations of O2, CO2, and CO also exhibit a reasonable agreement with the experimental data, even though a slight difference exists near burn through because of combustion rate overestimation. MaxT and FFS are also selected for the quantitative comparison of the modeling and measured results. MaxT is defined as the maximum temperature of a solid material at a given time, which indicates the degree of heat concentration in the bed, whereas FFS can be defined as11) Dis tan ce between po int A and B at 973K FFS Tim e consum ed for propagation m in cm where points A and B represent the locations of x = 525 mm and x = 225 mm, respectively, where thermocouples are installed in the sinter pot tests. The 973 K represents the initiation temperature of coke combustion. The comparison of the six cases between the modeling and measured results is presented in Figure 4. The modeling bed temperatures at the location of x = 525 mm, i.e., the upper bed, agree well with the measured data, whereas the relative error at the location of x = 225 mm, i.e., the middle and lower beds, increases significantly and is valued at ~5%. The reason for this finding is that the temperature 3.43 3.53 3.47 0 difference between the two points is not obvious in the sinter pot tests. In fact, the temperature difference between the two points is close to the modeling results. Two reasons may cause the experimental error: the sensitivity of the thermocouple sensor on the one hand and the influence of the gas flow in the combustion zone on the thermocouple sensor on the other hand. The temperature of gas flow in the combustion zone is lower than that of the solid material because of the violent exothermic reaction during coke combustion. Good agreements are achieved for FFSs, and the relative error is less than 5%. (a) 7 4.2. Combustion Characteristics in the Sintering Process (b) Figure 3 Comparison between simulated and measured data for cases CS and T200-19: (a) bed temperature profiles and (b) flue gas species concentration profiles. (a) (b) Figure 4 Comparison between the modeling and measured results: (a) MaxT and (b) FFS. The 1D transient model can be converted into a 2D steady process in the Cartesian coordinate system when the abscissa uses sintering time and the ordinate uses bed height. The 2D modeling results of bed temperature and melt fraction distributions within the sinter bed, for case CS, are presented in Figure 5. After ignition, sintering is sustained by coke combustion. The temperature of air/gas supplied to the combustion zone increases because of convection heat transfer from the heated particles in the upper bed of the high-temperature zone; the combustion zone tends to widen, and peak temperature increases as the combustion front proceeds downward12), as shown in Figure 5(a). This finding fully testifies to heat accumulation in the lower bed. The coke combustion rate also increases as bed temperature increases. Heat accumulation also causes additional melting of iron ores, which decreases bed permeability, as shown in Figure 5(b). This phenomenon is observed in practice, which implies that the sintered ore from the lower bed is stronger than that from the upper bed. Temperature is considered the most important parameter in studying combustion characteristics in the sinter bed. The most important quantitative parameters, namely, MaxT, CZT, and MZT, should be discussed in detail because they significantly affect bed permeability. CZT and MZT are defined as the thickness of the combustion zone and the melting zone whose bed temperature is over the initial temperatures of coke combustion and iron ore melting, respectively. As shown in Figure 6, MaxT gradually increases, which implies heat accumulation in the lower bed. CZT and MZT, which are determined to be highly dependent on bed temperature, also tend to increase as the process progresses. The increase in MaxT, CZT, and MZT in the lower bed indicate that the quality of the sintered ores enhance because increased strength should be guaranteed with increased temperature. However, they also mean that the bed permeability in the lower bed decreases since the increased degree of melt fraction. Thus, the total pressure drop of the combustion zone and the melting zone in the lower bed increase significantly. 8 (a) (b) Figure 5 Combustion Characteristics within the sinter bed: (a) bed temperature distributions and (b) melt fraction distributions. Figure 7 Schematic diagram of industrial FGRS system. 4.3.2. Simulation cases Therefore, the calculation cases in this study are performed for various sets of the three main operating parameters of the recirculated gas, namely, average temperature, O2 content, and gas velocity. The cases listed in Table 10 cover all ranges of the operating conditions that are available in industrial FGRS, in which CS is used as the reference case. Ambient air is used as the incoming gas in the exposure region, of which the velocity is equal to that of the recirculated gas in the recirculation region. Figure 6 Simulated results of MaxT, CZT, and MZT curves for cases CS and T200-19. 4.3. Parametric Study on FGRS Process 4.3.1. Boundary and initial conditions of an industrial FGRS process The industrial production conditions for the no. 2 sinter strand in Baoshan Stainless Steel Co., Ltd. in China are used to define the initial and boundary conditions. The schematic diagram of the industrial FGRS system of this manufacturer is shown in Figure 7. During ignition, the inlet gas temperature is set to 1150 °C for 4.0 m and then to 850 °C for 3.8 m (heat preservation). Subsequently, a recirculated gas temperature is set for 21.5 m, after which ambient temperature is set. Inlet gas velocity is set to the values determined from the surface of the 700 mmhigh sinter bed, at the bottom of which is the 30 mmhigh hearth ore. The traveling speed of the sintering strand is 1.24 m∙min-1. The volume fractions of all the gas species are specified, and inlet pressure is defined as 1.0 atm. The gradient of all the gas species concentrations at the outlet is set to 0. The chemical compositions of the raw mixtures, which are also used as initial conditions in this study, are presented in Table 1. 4.3.3. Combustion characteristics of the reference case To quantify the combustion characteristics, MaxT, FFS, sintering time (ST), and terminal temperature (TT) are introduced. To define ST, x = 75 mm is used as a base level, and the time when flame front reaches the point is calculated. TT is defined as the solid temperature when time reaches ST. The quantitative parameters for CS are below: ST = 1822 s, FFS = 2.320 cm∙min-1, MaxT for x = 525 mm, 375 mm, 225 mm, and 75 mm are 1553.45 K, 1598.38 K, 1632.87 K, and 1662.59 K, respectively, and TT are 464.73 K, 863.54 K, 1381.60 K, and 1662.59 K, respectively. Table 10 Calculation cases categorized under various the incoming gas conditions. Case names Temperature (oC) O2 content (vol. %) a Gas velocity (Nm∙s-1) CS 30 21 0.3 GasT-** 150~250, Δ=50 19 0.3 GasO2-** 200 17~21, Δ=1 0.3 19 0.24~0.38, Δ=0.2 GasV-** a 200 The total content of O2 and CO2 is 21 vol. %. 4.3.4. Effect of incoming gas temperature Figure 8 shows the modeling results of the bed temperature profiles for various incoming gas 9 temperature and for a fixed gas supply. The combustion zone appears broader at higher incoming gas temperature. Interestingly, temperature increase and MaxT at a specific location show hardly difference because of the convective heat transfer between the heated particles and the heating gas. However, high incoming gas temperature results in relieved cooling progress of the sintered ore in the upper bed, which slightly improves the unevenly distributed heat concentration. Figure 8 Modeling results of the bed temperature profiles at various incoming gas temperatures. Compared with the CS process, the TTs of the cases with various incoming gas temperatures at specific locations, as shown in Table 11, all demonstrate a significant increase, and more notable increase is observed at the location of x = 525 mm, i.e., the upper bed. Simultaneously, the coke combustion rate is limited by O2 diffusion at high temperatures, which slightly reduces FFS for higher incoming gas temperatures. Actually, a higher incoming gas temperature leads to a larger suction applied to obtain a consistent gas supply, which implies that the power consumption and loading of the main fan will increase. Therefore, considering the safety of equipment and costs, incoming gas temperature becomes unsuitable at over high levels and is frequently controlled at the range of 200~250 °C in industrial production. 4.3.5. Effect of O2 content in incoming gas Figure 9(a) shows that at low O2 content, bed temperature increase at a specific location is delayed, whereas that at the lower bed becomes increasingly significant, which implies that the flame front propagates slowly through the sinter bed. From Figure 9(b), an increase in CO2 content can be observed in the case of low O2 content. Meanwhile, an increase in CO content indicates that fuel consumption rate decreases and the incomplete combustion degree of fuel increases during sintering. Therefore, heat loss and carbon residues that exit the sintered ore will increase, whereas fuel heat efficiency will significantly decrease33-35). Coke combustion within the sinter bed is a complex reaction that is assumed as an unreacted-core shrinking model with characteristic size distributions. Theoretically, combustion rate proportionally decreases along with O2 potential. Therefore, FFS decreases along with O2 content in the incoming gas, whereas ST exhibits the opposite trend, as listed in Table 12. In particular, low O2 content decelerates FFS and extends ST during the entire process, which inhibit productivity. Coke combustion rate decreases along with O2 content, which reduces MaxT at a specific location of the sinter bed, as illustrated in Figure 10. A low MaxT in the sinter bed frequently deteriorates the tumble index of the sintered ore because additional silicate slags are produced in the reaction as a result of the reduced oxygen potential in the sinter bed. At x = 525 mm, MaxT of case GasO2-17 is extremely close to that of CS, which indicates that if O2 concentration continues to decrease, the melting conditions of the upper bed may be worse than that in CS, which contradicts the original intention of FGRS technology. Table 11 Quantitative parameters at various incoming gas temperatures. Quantitative parameter GasT150 GasT200 GasT250 x=525mm 1564.69 1567.17 1569.53 x=225mm 1662.37 1666.10 1669.85 x=525mm 553.71 589.41 626.76 x=225mm 1425.09 1436.53 1448.15 1838 1860 1880 2.293 2.256 2.218 MaxT (K) TT (K) ST (s) -1 FFS (cm∙min ) (a) 10 (b) Figure 10 Modeling results of MaxT under various O2 contents in incoming gas. Figure 9 Modeling results of (a) bed temperature profiles and (b) flue gas species emissions under various O2 contents in incoming gas. Table 12 STs and FFSs under various O2 contents in the incoming gas. Characteristic Temperature GasO2-17 GasO2-18 GasO2-19 GasO2-20 GasO2-21 ST (s) 1889 1875 1860 1844 1830 FFS (cm∙min-1) 2.189 2.224 2.256 2.289 2.314 4.3.6. Effect of incoming gas velocity Figure 11 shows the results of bed temperature and flue gas emissions under various gas velocities. For a high gas velocity, temperature increase at a specific location is significantly advanced, which implies that the flame front propagates rapidly through the sinter bed. According to Yang et al.33), coke particles in the upper bed requires a long time to burn out under a low gas supply, and thus, the lower bed is slightly affected. From Figure 11(b), CO2 content is decreased under low gas velocity, whereas coke consumption rate is decreased under slow combustion propagation. That is, higher gas velocity results in larger coke combustion rate. (b) Figure 11 Modeling results of (a) bed temperature profiles and (b) flue gas species contents under various incoming gas velocities. (a) MaxT at a specific location slightly increases, as shown in Figure 12(a), which indicates that the quality of the sintered ore will not be compromised by high gas velocity. Figure 12(b) shows that FFS proportionally increases along with incoming gas velocity, whereas ST exhibits the opposite trend. High gas velocity accelerates FFS and saves ST during the entire process to enhance productivity. Given that high gas velocity increases the fan power and capacity of the flue gas treatment system, system costs are obviously increased. By contrast, gas supply must not be too small for a designed sintering strand speed (1.24 m∙min-1 in this study). For a low incoming gas velocity and using 0.24 11 Nm∙s-1 as an example, the flame front does not completely penetrate into the bed, and when unloading sintered ore into the cooler, the quality of sintered ore in the lower bed may not satisfy production requirements. Therefore, a low incoming gas velocity can directly reduce productivity. MaxT and combustion zone thickness of the sinter bed. Therefore, gas flow resistance is slightly increased, whereas FFS is slightly decreased. These results imply that the operating parameters, with regard to FGRS technology, must be adjusted reasonably to maximize the utilization of the advantages of such technology, that is, to improve the quality of the sintered ore while maintaining stable productivity. 5. Conclusions (a) (b) Figure 12 Modeling results of (a) MaxT and (b) FFS and ST under various incoming gas velocities. As confirmed by the simulation results, the effect of incoming gas velocity is greater than those of O2 content and temperature. For a fixed O2 content and temperature of the recirculated gas, a low incoming gas velocity reflects a low O2 supply per unit mass of coke, which results in a low combustion rate. Figure 12(b) shows that FFS in case GasV-0.32 is higher than that in CS. Therefore, FFS becomes consistent in CS as velocity slightly increases. However, by producing additional fan power, an increase in incoming gas velocity directly increases costs. O2 content in the incoming gas for FGRS is inevitably reduced. When the velocity and temperature of the recirculated gas are fixed, the melting conditions of the upper bed may be worse than that in CS if O2 content is reduced to less than 17 vol.%, as shown in Figure 10. Therefore, O2 content must be maintained at higher than 17 vol.% as much as possible. In addition, for a fixed velocity and O2 content of the recirculated gas, the increase in temperature of the incoming gas can increase the This study introduces a comprehensive 1D mathematical model of the iron ore sintering process by considering FGRS technology. This multiphase theory-based model considers most of the primary phenomena that occur in the sintering process. The major gas-solid and gaseous reactions which O2, CO2, CO, and H2O participate in as reactants and products are taken into account. Many of these phenomena have not been considered in currently established models. Convective, conductive, and radiative heat transfer modes within/between different solid and gas phases are considered. Geometric changes are not involved. Six carefully controlled sinter pot tests based on FGRS technology are used to validate the model. Reasonable agreements are obtained. The simulated results show that the combustion zone tends to widen, and MaxT increases as the combustion front proceeds downward, testifying the heat accumulation in the lower bed, and MZT also tends to increase as the process progresses. The increase in MaxT, CZT, and MZT indicate that the sintered ore from the lower bed is stronger than that from the upper bed. However, the bed permeability in the lower bed decreases since the increased degree of melt fraction. Thus, the total pressure drop of the combustion zone and the melting zone in the lower bed increase significantly. Parametric studies are performed at various velocities, temperatures, and O2 contents of the incoming gas. The modeling results also show the improvement of the unevenly distributed heat concentration in the bed via the FGRS process, due to the additional heat supply by hot recirculated gas. The decrease in O2 content may worsen the melting conditions of the upper bed. The increase in incoming gas velocity reflects more O2 supply per unit mass of coke, which results in a higher combustion rate and MaxT. However, FFS requires further attention. And with the increase in temperature or velocity, more power consumption and larger loading of the main fan are needed. Velocity exerts the greatest effect, followed by O2 12 content. The incoming gas conditions of FGRS must be carefully determined, the temperature is suggested to be 200~250 oC, the O2 content must be controlled to be higher than 17 vol.%, and the velocity should be slightly increased compared to CS process. Work continues to further improve the generality and accuracy of the proposed model, as well as to further investigate the process optimization of the FGRS technology. t: Time, s T: Temperature, K u: Velocity, m∙s-1 Wcr: Critical solid moisture content, % x: Spatial coordinate along the direction of bed height, m Y: Mass fraction of solid and gas phases, ΔH: Heat of reaction, J∙kg -1 Acknowledgments This work was financially supported by the National Natural Science Foundation of China (No. 50876011). Meanwhile, the authors wish to express their thanks to EssayStar.com for providing language help. ΔP: Pressure drop across the sinter bed, Pa Nu: Particle Nusselt number, Pr: Particle Prandtl number, Re: Particle Reynolds number, Sc: Particle Smit number, - Nomenclature Sh: Particle Sherwood number, - A: Specific surface area, factor, s-1 m 2∙m-3; Pre-exponential B: Parameters related to the surface structure of coke, C: Molar concentration of gas phases, mol∙m-3 Cp: Specific heat, J∙kg-1∙K-1 Greeks χ: Polynomial correlation of the characteristic drying curve for iron ore particles, δ: Ash layer thickness, m dp: Solid phase mean diameter, m ε: Porosity of sinter bed or solid phases, -; Emissivity, - D: Mass diffusion coefficient of gas phases, m 2∙s-1 φ: Fraction of heat absorbed by solid, - E: Activation energy, J∙mol -1 h, hconv: Convection coefficient, γ: Volume fraction of solid and gas phases, W∙m-2∙K-1 κ: Stoichiometric coefficient, - H: Height of the sinter bed, m λ: Conductivity, W∙m-1∙K-1 I: Radiation intensity, W∙m-2sr μ: Gas dynamic viscosity, kg∙m-1∙s-1 kc: Reaction rate constant, m∙s-1 ρ: Density, kg∙m-3 Keq: Reaction equilibrium constant, - σ: Stefan-Boltzmann constant, W∙m-2∙K-4 m0: Mass density of the initial particle, kg∙m-3 ςj: Solid phase shape factor, - mc: Mass density of the un-reacted part, kg∙m-3 M: Molecular weight, kg∙mol-1 n: Particle number density, Subscripts and Superscripts 1∙m-3 g: Gas P: Pressure, Pa Q: Volumetric heat generation rate, s: Solid W∙m-3 r0: Radius of the initial particle, m rc: Radius of the un-reacted part, m R: Reaction rate, mol∙m-3∙s-1 Rg: Universal gas constant, J∙mol-1∙K-1 k: Reaction index i: Gas species index (i = N2, O2, CO2, CO, H2, and H2O) j, jj: Solid species index (j = sinter feed, return fine, coke, limestone, dolomite, hydrated lime) C: Coke 13 L: Limestone H2O: Vapor or solid moisture eff: Effective diffusion rad: Radiation ssa: Specific surface area *: Saturation vapor; Gas equilibrium concentration ω: Phase change factor dependent on factors REFERENCES J. Shibata. (1988). Analysis of Sintering Process by the Mathematical Model, Mathematical and Computer Modeling, 11, 956-961. F. Patisson, J.P. Bellot, D. Ablitzer. (1990). Study of Moisture Transfer during the Strand Sintering Process, Metallurgical Transaction B, 21, 37-47. R. Venkataramana, S.S Gupta, P.C. Kapur, N. (1998). Ramachandran. Mathematical Modelling and Simulation of the Iron Ore Sintering Process, Tata Search, 50-55. Heat front Propagation in the Iron Ore Sintering Process, ISIJ International, 44, 11-20. W. Yang, C. Ryu, S. Choi, E. Choi, D. Lee, W. Huh. (2004). Modeling of Combustion and Heat Transfer in an Iron Ore Sinter bed with Considerations of Multiple Solid Phases, ISIJ International, 44, 492499. W. Yang, C. Ryu, S. Choi, E. Choi, D.W. Ri, W. Huh. (2004). Mathematical Model of Thermal Process in an Iron Ore Sinter bed, Metals and Materials International, 10, 493-500. H. Kang, S. Choi, W. Yong, B. Cho. (2011). Influence of Oxygen Supply in an Iron Ore Sintering Process, ISIJ International, 51, 10651071. S.V. Komarov, H. Shibata, N. Hayashi, E. Kasai. (2010). Numerical and Experimental Investigation on Heat Propagation through Composite Sinter Bed with Non-Uniform Voidage: Part 1 Mathematical Model and Its Experimental Verification, Journal of Iron and Steel Research, International, 17, 01-07. N.K. Nath, A.J. Da Silva, N. Chakraborti. (1997). Dynamic Process Modeling of Iron Ore Sintering, Steel Research, 68, 285-292. H. Ahn, S. Choi, B. Cho. (2013). Process Simulation of Iron Ore Sinter bed with Flue Gas Recirculation, Part 1- Modelling Approach, Ironmaking and Steelmaking, 40, 120-127. N.K. Nath, K. Mitra. (2005). Mathematical Modeling and Optimization of Two-Layer Sintering Process for Sinter Quality and Fuel Efficiency Using Genetic Algorithm, Materials and Manufacturing Processes, 20, 335-349. H. Ahn, S. Choi, B. Cho. (2013). Process Simulation of Iron Ore Sinter bed with Flue Gas Recirculation, Part 2- Parametric Variation of Gas Conditions, Ironmaking and Steelmaking, 40, 128-137. H. Zhou, J.P. Zhao, C.E. Loo, B.G. Ellis, K. Cen. (2012). Numerical Modeling of the Iron Ore Sintering Process, ISIJ International, 52, 15501558. J.A. de Castro, Y. Sazaki, J. Yagi. (2012). Three Dimensional Mathematical Model of the Iron Ore Sintering Process Based on Multiphase Theory, Materials Research-IBERO-American Journal of Materials, 15, 848-858. J.P. Zhao, C.E. Loo, R.D. Dukino. (2014). Modelling fuel combustion in iron ore sintering, Combustion and Flame, 162, 1019-1034. M. Pahlevaninezhad, M.D. Emami, M. Panjepour. (2014). The Effects of Kinetic Parameters on Combustion Characteristic in a Sinter bed, Energy, 73, 160-176. M.V. Ramos, E. Kasai, J. Kano, T. Nakamura. (2000). Numerical Simulation Model of the Iron Ore Sintering Process Directly Describing the Agglomeration Phenomenon of Granules in the Packed Bed, ISIJ International, 40, 448-454. J. Mitterlehner, G. Loeffler, F. Winter, H. Hofbauer, H. Schmid, E. Zwittag, T.H. Buergler, O. Pammer, H. Stiasny. (2004). Modeling and Simulation of J.A. de Castro, N. Nath, A.B. Franca, V.S. Guilherme, Y. Sasaki. (2012). Analysis by Multiphase Multicomponent Model of Iron Ore Sintering based on Alternative Steelworks Gaseous Fuels, Ironmaking and Steelmaking, 39, 605-613. H. Yamaoka, T. Kawaguchi. (2005). Development of a 3-D Sinter Process Mathematical Simulation Model, ISIJ International, 45, 522-531. C.W. Crawford, O.A. Plumb. (1986). The Influence of Surface Roughness on Pressure Drop and Fluid Velocities through Beds of Particulates, Journal of Fluids Engineering-Transactions of the ASME, 108, 343-347. J. Hinkley, A.G. Waters, J.D. Litster. (1994). An Investigation of Pre-ignition Air Flow in Ferrous 14 Sintering, International Journal of Mineral Processing, 42, 37-52. M.L. Hobbs, P.T. Radulovic, L.D. Smoot. (1993). Combustion and Gasification of Coals in Fixedbeds, Progress in Energy and Combustion Science, 19, 505-586. R.W. Young. (1977). Dynamic Mathematical Model of (Iron-Ore) Sintering Process, Ironmaking and Steelmaking, 4, 321-328. M. Hartman, O. Trnka, V. Vesely, K. Svoboda. (1996). Predicting the Rate of Thermal Decomposition of Dolomite, Chemical Engineering Science, 51, 5229-5232. A. Irabien, J.R. Viguri, I. Ortiz. (1990). Thermal Dehydration of Calcium Hydroxide. 1. Kinetic Model and Parameters, Industrial & Engineering Chemistry, 29, 1599-1606. W.K. Jozwiak, E. Kaczmarek, T.P. Maniecki, W. Ignaczak, W. Maniukiewicz. (2007). Reduction Behavior of Iron Oxides in Hydrogen and Carbon Monoxide Atmospheres, Applied Catalysis A: General, 326, 17-27. A. Pineau, N. Kanari, I. Gaballah. (2006). Kinetics of Reduction of Iron Oxides by H2. Part I: Low Temperature Reduction of Hematite, Thermochimica Acta, 447, 89-100. C.D. Blasi. (2000). Dynamic behavior of stratified downdraft gasifiers, Chemical Engineering Science, 55, 2931-2944. H. Thunman, B. Leckner. (2001). Ignition and Propagation of a Reaction Front in Cross-current Bed Combustion of Wet Biofuels, Fuel, 80, 473481. P.R. Austin, H. Nogami, J. Yagi. (1997). A Mathematical Model for Blast Furnace Reaction Analysis Based on the Four Fluid Model, ISIJ International, 37, 748-755. D. Shin, S. Choi. (2000). The Combustion of Simulated Waste Particles in a Fixed Bed, Combustion and Flame, 121, 167-180. G. Wang, Z. Wen, G.F. Lou, R.F. Dou, X.W. Li. (2015). Experimental Study on the Effect of Waste Gas Recovery on the Iron Ore Sintering, Heat Transfer - Asian Research, (Published online), http://onlinelibrary.wiley.com/doi/10.1002/htj. 21202. W. Yang, S. Choi, E.S. Choi, D.W. Ri, S. Kim. (2006). Combustion characteristics in an iron ore sinter bed—evaluation of fuel substitution, Combustion and Flame, 145, 447-463. X.H. Fan, Z.Y. Yu, M. Gan, W.Q. Li, Z.Y. Ji. (2013). Influence of O2 Content in Circulating Flue Gas on Iron Ore Sintering, Journal of Iron and Steel Research, International, 20, 01-06. X.H. Fan, Z.Y. Yu, M. Gan, X.L. Chen, T. Jiang, H.L. Wen. (2014). Appropriate Technology Parameters of Iron Ore Sintering Process with Flue gas recirculation, ISIJ International, 54, 2541-2550.