Iron Ore Reduction by Hydrogen Using a Laboratory Scale Fluidized Bed Reactor: Kinetic Investigation—Experimental Setup and Method for Determination
- 363 Downloads
Abstract
The reduction kinetics of hematite iron ore fines to metallic iron by hydrogen using a laboratory fluidized bed reactor were investigated in a temperature range between 873 K and 1073 K, by measuring the weight change of the sample portion during reduction. The fluidization conditions were checked regarding plausibility within the Grace diagram and the measured pressure drop across the material during experiments. The apparent activation energy of the reduction was determined against the degree of reduction and varied along an estimated two-peak curve between 11 and 55 kJ mol^{−1}. Conventional kinetic analysis for the reduction of FeO to metallic iron, using typical models to describe gas–solid reactions, does not show results with high accuracy. Multistep kinetic analysis, using the Johnson–Mehl–Avrami model, shows that the initial stage of reduction from Fe_{2}O_{3} to Fe_{3}O_{4}, and partly to FeO, is controlled by diffusion and chemical reaction, depending on the temperature. Further reduction can be described by a combination of nucleation and chemical reaction, whereby the influence of nucleation increases with an increasing reduction temperature. The results of the kinetical analysis were linked to the shape of the curve from apparent activation energy against the degree of reduction.
Introduction
The reduction of iron ore fines by means of fluidized bed technology has been of interest to the iron and steel industries for many years to produce direct reduced iron (DRI).[1] In the near future, the importance of DRI will increase drastically. Overall steel production has risen about 100 pc. over the past 20 years, from 0.85 to 1.7 billion tons of crude steel.[2] Consequently, the scrap return will also increase within the next few years, considering the fact that a typical life cycle time of steel products is 25 years. Thus, the amount of steel produced via the process route based on an electric arc furnace (EAF) will also grow in the near future. To obtain high-quality steel grades, the demand for DRI will also rise as a scrap substitute in the EAF. Aside from the increasing demand, direct reduction processes have advantages over conventional steel making via blast furnace and oxygen converter, especially in terms of environmental issues. Some direct reduction processes are able to operate with high hydrogen content in the reducing gas mixture. Therefore, accompanying CO_{2} emissions can be reduced. The use of fluidized bed technology brings one further advantage, as fine iron ores can be used directly without a prior agglomeration step. Examples of direct reduction processes using fluidized bed technology in industrial scale are the Finmet®[3,4] and the Circored®[5,6] processes. Both are based on natural gas as an energy source, whereby in the case of Finmet®, the required reducing gas mixture is produced by means of the reforming of natural gas. In contrast, the Circored® process only uses hydrogen as a reducing agent, also provided by means of the reforming of natural gas in combination with a shift reactor and a CO_{2} removal unit.
Investigations regarding the reduction kinetics of iron oxides using hydrogen as a reducing agent have been carried out by many authors.[11, 12, 13, 14, 15, 16, 17, 18, 19] The scope of this work was to investigate the reduction kinetics of iron ore fines during fluidized bed reduction using hydrogen as a reducing agent in a suitable temperature range for an industrial application. The determination of the apparent activation energy against the degree of reduction, model–fitting analysis of experimental results as well as multistep kinetic analysis were carried out. Conclusions were drawn concerning the shape of the apparent activation energy against the degree of reduction and the results from kinetic analysis.
Experimental Procedure
Experimental Setup
Before the experiment, the iron ore sample was placed into the reactor and purged with N_{2} to remove oxygen from the system. The heating period to the desired reduction temperature took place under constant N_{2} flow. After that, a temperature-equilibrium period of 10 minutes was chosen in order to reach a stable weight signal and a constant temperature. Then, N_{2} was replaced by the defined reducing gas mixture and the reduction procedure started. The change in the sample weight and the differential pressure through the grid and sample portion were measured during the experiment to evaluate the reduction and fluidization behavior. After the reduction period, the reducing gas mixture was substituted by N_{2} again, and the sample was cooled down to ambient temperature before discharging.
Experimental Conditions
Chemical Analysis of Iron Ore and Process Conditions
Analysis Iron Ore | Test Parameters | ||
---|---|---|---|
Fe _{tot} ^{a} | 63.6 wt pc. | input iron ore | 400 g |
FeO | 0.58 wt pc. | particle size | 250 to 500 µm |
SiO_{2} | 3.48 wt pc. | gas mixture | H_{2}/N_{2} 65/35 vol pc. |
Al_{2}O_{3} | 2.07 wt pc. | flow rate | 25.9 Nl min^{−1} |
LOI^{b} | 2.2 pc. | temperature | 873 to 1073 K |
BET^{c} | 11.83 m^{2} g^{−1} | pressure | 1.1 bar abs. |
Calculation Parameters of the Grace Diagram
Density Fluid 973 K | ρ _{F} | 0.076 | kg m^{−3} |
Density Solids | ρ _{S} | 3,500 | kg m^{−3} |
Kinematic Viscosity Fluid 973 K | η | 3.71E−5 | Pa s |
Minimum Fluidization Porosity | ε _{mf} | 0.39 | — |
Sphericity Solids | φ _{s} | 0.86 | — |
Results and Discussion
Experimental Results
Determination of Apparent Activation Energy Using the Model-Free Method
Investigations of Kinetics
The following sections detail the kinetical investigations carried out regarding the experimental data. To achieve accurate results, conventional kinetic analysis was done using only experimental data from 33 to 100 pc. degree of reduction, so only the reduction from FeO to Fe was taken into account. For the evaluation of the total reduction processes from Fe_{2}O_{3} to Fe, multistep kinetic analysis was performed.
Conventional Kinetic Analysis for the Reduction of FeO to Fe—Approach 1
Model | f(x) | g(x) | |
---|---|---|---|
Phase-boundary-controlled | |||
PBC1 | infinite slab | 1 | x |
PBC2 | contracting cylinder | \( 2\left( {1 - x} \right)^{1/2} \) | \( 1 - \left( {1 - x} \right)^{1/2} \) |
PBC3 | contracting sphere | \( 3\left( {1 - x} \right)^{1/3} \) | \( 1 - \left( {1 - x} \right)^{1/3} \) |
Diffusion models | |||
D1 | one-dimensional | \( 1/\left( {2x} \right) \) | \( x^{2} \) |
D2 | two-dimensional | \( \left( { - \ln \left( {1 - x} \right)} \right)^{ - 1} \) | \( x + \left( {1 - x} \right)\ln \left( {1 - x} \right) \) |
D3 | three-dimensional Jander | \( 3/2\left( {1 - x} \right)^{2/3} \left(1 - \left( 1 - x\right)^{1/3}\right)^{ - 1} \) | \( (1 - 1 - x)^{1/3} )^{2} \) |
D4 | three-dimensional ginstling | \( 3/2(\left( {1 - x} \right)^{ - 1/3} - 1)^{ - 1} \) | \( \left( {1 - 2/3x} \right) - (1 - x)^{2/3} \) |
Reaction-order models | |||
ROM1 | first order | \( 1 - x \) | \( - \ln \left( {1 - x} \right) \) |
ROM2 | 1.5 order | \( (1 - x)^{3/2} \) | \( 2\left({\left( {1 - x}\right)^{ - 1/2} - 1} \right) \) |
ROM3 | second order | \( (1 - x)^{2} \) | \( (1 - x)^{ - 1} - 1 \) |
Nucleation models | |||
NM1 | n = 1.5 | \( 2/3\left( {1 - x} \right)\left( { - { \ln }\left( {1 - x} \right)} \right)^{1/3} \) | \( \left( { - { \ln }\left( {1 - x} \right)} \right)^{2/3} \) |
NM2 | n = 2 | \( 2\left( {1 - x} \right)\left( { - { \ln }\left( {1 - x} \right)} \right)^{1/2} \) | \( \left( { - { \ln }\left( {1 - x} \right)} \right)^{1/2} \) |
NM3 | n = 3 | \( 3\left( {1 - x} \right)\left( { - { \ln }\left( {1 - x} \right)} \right)^{2/3} \) | \( \left( { - { \ln }\left( {1 - x} \right)} \right)^{1/3} \) |
NM4 | n = 4 | \( 4\left( {1 - x} \right)\left( { - { \ln }\left( {1 - x} \right)} \right)^{3/4} \) | \( \left( { - { \ln }\left( {1 - x} \right)} \right)^{1/4} \) |
Model-Fitting Analysis of Experimental Data for Reduction of FeO to Fe Showing Coefficient of Determination at Different Temperatures—Approach 1
Model | 873 K | 923 K | 973 K | 1023 K | 1073 K |
---|---|---|---|---|---|
PBC1 | 0.871 | 0.856 | 0.825 | 0.776 | 0.777 |
PBC2 | 0.979 | 0.968 | 0.846 | 0.919 | 0.905 |
PBC3 | 0.993 | 0.988 | 0.974 | 0.957 | 0.941 |
D1 | 0.958 | 0.943 | 0.921 | 0.881 | 0.868 |
D2 | 0.993 | 0.984 | 0.974 | 0.952 | 0.932 |
D3 | 0.937 | 0.963 | 0.974 | 0.980 | 0.967 |
D4 | 0.992 | 0.991 | 0.988 | 0.977 | 0.957 |
ROM1 | 0.942 | 0.972 | 0.980 | 0.979 | 0.964 |
ROM2 | 0.648 | 0.758 | 0.813 | 0.798 | 0.780 |
ROM3 | 0.351 | 0.447 | 0.544 | 0.420 | 0.461 |
NM1 | 0.981 | 0.989 | 0.980 | 0.970 | 0.955 |
NM2 | 0.980 | 0.978 | 0.960 | 0.945 | 0.932 |
NM3 | 0.957 | 0.949 | 0.920 | 0.899 | 0.89 |
NM4 | 0.935 | 0.925 | 0.890 | 0.863 | 0.861 |
Conventional Kinetics Analysis for the Reduction of FeO to Fe—Approach 2
Values of Reaction Rate k and Root Mean Square Deviation RMSD Resulting from Model-Fitting—Approach 2
Model | 873 K | 923 K | 973 K | 1,023 K | 1,073 K | |||||
---|---|---|---|---|---|---|---|---|---|---|
k (s^{−1}) | RMSD | k (s^{−1}) | RMSD | k (s^{−1}) | RMSD | k (s^{−1}) | RMSD | k (s^{−1}) | RMSD | |
PBC3 | 1.70E−4 | 0.016 | 2.10E−4 | 0.015 | 2.62E−4 | 0.025 | 3.29E−4 | 0.032 | 4.10E−4 | 0.029 |
D3 | 5.95E−5 | 0.098 | 7.37E−5 | 0.099 | 9.18E−5 | 0.099 | 1.16E−4 | 0.097 | 3.66E−1 | 0.108 |
ROM1 | 6.43E−4 | 0.022 | 7.89E−4 | 0.025 | 9.86E−4 | 0.026 | 1.23E−3 | 0.034 | 1.53E−3 | 0.039 |
NM1 | 6.13E−4 | 0.044 | 7.53E−4 | 0.040 | 9.49E−4 | 0.040 | 1.19E−3 | 0.040 | 1.47E−3 | 0.022 |
Comparison of Arrhenius Activation Energy Values
Determined Values of Apparent Activation Energy from Different Model-Fitting Analyses
Model No. | Approach 1 | Approach 2 |
---|---|---|
E_{a} (kJ mol^{−1}) | E_{a} (kJ mol^{−1}) | |
PBC3 | 29.13 | 34.33 |
D3 | 30.19 | 32.98 |
D4 | 30.14 | — |
ROM1 | 29.23 | 33.91 |
NM1 | 28.76 | 34.37 |
Values of Apparent Activation Energy for the Reduction of FeO to Fe, as Reported in Literature
Reference | Reduction Step | T-Range (K) | Ea (kJ mol^{−1}) | Experimental Method | Type of Experiment |
---|---|---|---|---|---|
Barde et al.[32] | FeO → Fe | 1073 to 1273 | 30.0 | isothermal (H_{2}) | fixed bed |
Kuila et al.[33] | FeO → Fe | 973 to 1273 | 55.0 | isothermal (H_{2}) | fixed bed |
Kuila et al.[31] | FeO → Fe | 973 to 1173 | 11.0 | isothermal (H_{2}) | fixed bed |
Muntenau et al.[27] | FeO → Fe | 298 to 1073 | 85.7 | non-isothermal (H_{2}) | fixed bed |
Jozwiak et al.[34] | FeO → Fe | 298 to 1173 | 104.0 | non-isothermal (H_{2}) | fixed bed |
Hou et al.[35] | FeO → Fe | 863 to 903 | 75.9 | isothermal (H_{2}-Ar) | fixed bed |
Multistep Kinetic Analysis of Total Reduction from Fe_{2}O_{3} to Fe Using a Parallel Reaction Model Based on the JMA Model
Weight Factors, Nucleation Rate Constants, Kinetic Exponents, and RMSD for Multistep Kinetic Analysis (Parallel) at Different Temperatures
873 K | 923 K | 973 K | 1023 K | 1073 K | |
---|---|---|---|---|---|
w _{1} | 0.0915 | 0.1032 | 0.1379 | 0.2691 | 0.2371 |
w _{2} | 0.8535 | 0.7607 | 0.6277 | 0.3865 | 0.0722 |
w _{3} | 0.0549 | 0.1362 | 0.2344 | 0.3443 | 0.6907 |
a _{1} | 0.0117 | 0.0140 | 0.0109 | 0.0286 | 0.0176 |
a _{2} | 5.313E−05 | 7.287E−05 | 1.394E−04 | 1.201E−04 | 5.382E−07 |
a _{3} | 2.064E−09 | 2.065E−09 | 2.065E−09 | 2.065E−09 | 8.614E−07 |
n _{1} | 1.17 | 1.07 | 1.05 | 0.69 | 0.84 |
n _{2} | 1.29 | 1.28 | 1.22 | 1.24 | 1.86 |
n _{3} | 2.61 | 2.65 | 2.76 | 2.86 | 2.04 |
RMSD | 0.0045 | 0.0048 | 0.0049 | 0.0061 | 0.0070 |
Conclusions
A laboratory fluidized bed reactor was used to investigate the reduction kinetics of a fine hematite ore to metallic iron by hydrogen at temperatures from 873 K to 1073 K. Different approaches for kinetic analysis were used to evaluate the rate-limiting step. Several conclusions can be drawn. First, the influence of kinetic limitation on the reaction rate of reduction reactions decreases drastically with increasing temperatures. At 1073 K, the reduction proceeds near the thermodynamic equilibrium, while at 873 K, the deviation between experimental results and thermodynamic equilibrium is much higher. Second, the reaction kinetics of hematite reduction by hydrogen cannot be described using only one simple gas–solid reaction model. The limiting mechanism varies with temperature and the degree of reduction. Therefore, the apparent activation energy is not constant during the reduction procedure. Third, a two-peak-shaped curve of apparent activation energy against degree of reduction was determined, whereby the apparent activation energy varies from 11 to 55, 55 to 23, and 23 to 42 kJ mol^{−1} for the reduction of Fe_{2}O_{3} to Fe_{3}O_{4}, Fe_{3}O_{4} to FeO, and FeO to Fe, respectively. Fourth, polished micro-sections show that at the beginning of metallic iron formation, the iron is formed uniformly and distributed in the whole particle, which indicates that diffusion of the reducing gas does not limit the reduction in this case. Lastly, multistep kinetic analysis using the JMA model shows that the initial stage of reduction might be controlled by first-order kinetics and diffusion, depending on the temperature. The reduction of FeO to Fe is limited by first-order kinetics and nucleation, whereby the importance of nucleation increases with rising temperatures. Moreover, diffusion is not important in the case of fluidized bed reduction, using hydrogen as a reducing agent during the reduction from FeO to Fe.
Notes
Acknowledgments
Open access funding provided by Montanuniversitaet Leoben. The authors would like to acknowledge the financial support from the project E^{3}-SteP (Enhanced Energy Efficient Steel Production), funded by the Austrian Research Promotion Agency (FFG).
References
- 1.T.F. Reed, J.C. Agarwal, and E.H. Shipley: J. Met., 1960, vol. 12, pp. 317–320.Google Scholar
- 2.World Steel Association: Steel Statistical Yearbook. https://www.worldsteel.org, 2017. Accessed 15 Oct 2018.
- 3.R. Lucena, R. Whipp, and W. Albarran: Stahl Eisen, 2007, vol. 127, pp. 567–90.Google Scholar
- 4.W. Hillisch and J. Zirngast: Steel Times Int., 2001, 3, 20–22.Google Scholar
- 5.D. Nuber, H. Eichberger, and B. Rollinger: Stahl Eisen, 2006, vol. 126, pp. 47–51.Google Scholar
- 6.J.P. Nepper, S. Sneyd, T. Stefan, and J. Weckes: Proceedings of the 6th European Coke and Ironmaking Congress, Düsseldorf, Germany, 2011, pp. 1–7.Google Scholar
- 7.L.v. Bogdandy and H.-J. Engell: Die Reduktion der Eisenerze: Wissenschaftliche Grundlagen und technische Durchführung, 1st ed., Springer, Berlin, 1967, pp. 18–38.CrossRefGoogle Scholar
- 8.M. Hanel, PhD-thesis, Montanuniversitaet Leoben, 2014.Google Scholar
- 9.Y.-W. Zhong, Z. Wang, X.-Z. Gong, and Z.-C. Guo: Ironmaking Steelmaking, 2013, vol. 39, pp. 38–44.CrossRefGoogle Scholar
- 10.B. Zhang, X. Gong, Z. Wang, and Z. Guo: ISIJ Int., 2011, vol. 51, pp. 1403–9.CrossRefGoogle Scholar
- 11.K. Piotrowski, K. Mondal, H. Lorethova, L. Stonawski, T. Szymanski, and T. Wiltowski: Int. J. Hydrogen Energy 2005, vol. 30, pp. 1543–54.CrossRefGoogle Scholar
- 12.K. Piotrowski, K. Mondal, T. Wiltowski, P. Dydo, and G. Rizeg: Chem. Eng. J., 2007, vol. 131, pp. 73–82.CrossRefGoogle Scholar
- 13.M. Bahgat and M.H. Khedr: Mater. Sci. Eng. B, 2007, 138, 251–258CrossRefGoogle Scholar
- 14.O. A. Teplov: Russ. Metall. 2012, 1, 8–21.CrossRefGoogle Scholar
- 15.P. Pourghahramani and E. Forssberg: Thermochim. Acta, 2007, vol. 454, pp. 69–77.CrossRefGoogle Scholar
- 16.R.J. Fruehan, Y. Li, L. Brabie, and E-J. Kim: Scand. J. Metall., 2005, vol. 34, pp. 205–12.CrossRefGoogle Scholar
- 17.Z. Wei, J. Zhang, B. Qin, Y. Dong, Y. Lu, Y. Li, W. Hao, and Y. Zhang: Powder Technol., 2018, vol. 332, pp. 18–26.CrossRefGoogle Scholar
- 18.A. Pineau, N. Kanari, and I. Gaballah: Thermochim. Acta, 2006, vol. 447, pp. 89–100.CrossRefGoogle Scholar
- 19.A. Pineau, N. Kanari, and I. Gaballah: Thermochim. Acta, 2007, vol. 456, pp. 75–88.CrossRefGoogle Scholar
- 20.VDI-Gesellschaft Verfahrenstechnik und Chemieingenieurwesen: VDI Heat Atlas, 2nd ed., Springer, Berlin, 2010, pp. 131–45.Google Scholar
- 21.D. Kunii and O. Levenspiel: Fluidization Engineering, 2nd ed., Butterworth-Heinemann, Stoneham, 1991, pp. 68–84.Google Scholar
- 22.D. Geldart: Gas Fluidization Technology, John Wiley & Sons, Chichester, New York, Brisbane, Toronto, Singapore, 1986, pp. 21–24.Google Scholar
- 23.A. Haider and O. Levenspiel: Powder Technol., 1989, vol. 58, pp. 63–70.CrossRefGoogle Scholar
- 24.P. Simon: J. Therm. Anal. Calorim., 2004, vol. 76, pp. 123–32.CrossRefGoogle Scholar
- 25.M. Pijolat, L. Favergeon, and M. Soustelle: Thermochim. Acta, 2011, vol. 525, pp. 93–102.CrossRefGoogle Scholar
- 26.H. Chen, Z. Zheng, Z. Chen, X. and T. Bi: Powder Technol,. 2017, vol. 316, pp. 410–20.CrossRefGoogle Scholar
- 27.G. Munteanu, L. Ilieva, and D. Andreeva: Thermochim. Acta, 1997, vol. 291, pp. 171–77.CrossRefGoogle Scholar
- 28.M. Shimokawabe, R. Furuichi, and T. Ishii: Thermochim. Acta, 1979, vol. 28, pp. 287–305.CrossRefGoogle Scholar
- 29.A. Ortega: Thermochim. Acta, 1996, vol. 284, pp. 379–87.CrossRefGoogle Scholar
- 30.A. Khawam and D.R. Flanagan: J. Phys. Chem. B, 2006, vol. 110, pp. 17315-28.CrossRefGoogle Scholar
- 31.S.K. Kuila, S. Chaudhuri, R. Chatterjee, and D. Ghosh: Proceedings of the International Conference on Chemical Engineering, Dhaka, Bangladesh, 2014, pp. 80–85.Google Scholar
- 32.A.A. Barde, J.F. Klausner, and R. Mei: Int. J. Hydrogen Energy, 2016, vol. 41, pp. 10103–19.CrossRefGoogle Scholar
- 33.S. K. Kuila, R. Chatterjee, and D. Ghosh: Int. J. Hydrogen Energy, 2016, vol. 41, pp. 9256–66.CrossRefGoogle Scholar
- 34.W. K. Jozwiak, E. Kaczmarek, T. P. Maniecki, W. Ignaczak, and W. Maniukiewicz: Appl. Catal. A, 2007, 326, 17–27.CrossRefGoogle Scholar
- 35.B. Hou, H. Zhang, H. Li, and Q. Zhu: Chin. J. Chem. Eng., 2012, vol. 20, pp. 10–17.CrossRefGoogle Scholar
- 36.H. Zuo, C. Wang, J. Dong, K. Jiao, and R. Xu: Int. J. Miner. Metall. Mater. 2015, 22, 688–96.CrossRefGoogle Scholar
- 37.W.A. Johnson and R.F. Mehl: Trans. AIME, 1939, vol. 135, pp. 416–58.Google Scholar
- 38.M. Avrami: J. Chem. Phys., 1939, vol. 7, pp. 1103–12.CrossRefGoogle Scholar
- 39.M. Avrami: J. Chem. Phys., 1940, vol. 8, pp. 212–24.CrossRefGoogle Scholar
- 40.M. Avrami: J. Chem. Phys., 1941, vol. 9, pp. 177–84.CrossRefGoogle Scholar
- 41.J. Malek: Thermochim. Acta, 1995, vol. 267, pp. 61–73.CrossRefGoogle Scholar
- 42.E. R. Monazam, R. W. Breault, and R. Siriwardane: Ind. Eng. Chem. Res., 2014, vol. 53, pp. 13320–28.CrossRefGoogle Scholar
- 43.E. R. Monazam, R. W. Breault, R. Siriwardane, G. Richards, and S. Carpenter: Chem. Eng. J., 2013, vol. 232, pp. 478–87.CrossRefGoogle Scholar
Copyright information
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.