Experimental study and numerical simulation on porosity dependent direct reducibility of high-grade iron oxide pellets in hydrogen

In this study, a novel application of machine learning (ML) is introduced to pellet modeling in the intricate non‐catalytic gas–solid reaction of direct reduction of iron oxide in the steel industry. Twenty ML models are developed using four algorithms: multilayer perceptron neural networks (MLPNN), radial basis function neural network (RBFNN), support vector regression, and random forest (RF). Hyperparameter optimization is conducted using Bayesian algorithms, random search, and grid search. The optimum model achieves a mean squared error test of 0.0052 with random RF for the larger dataset (872 samples), while smaller datasets (132, 225, and 242 samples) produce optimum models with MLPNN and RBFNN. Hyperparameters vary between the larger datasets and the smaller datasets. The models offer insight into the complex interactions among variables, including time, temperature, gas composition, hematite composition, pellet radius, and initial pellet porosity, influencing the metallization degree. In this study, the significant role of time and temperature is emphasized, as revealed by explainable artificial intelligence using Shapley additive explanation analysis that utilizes the game theory, and the effects of pellet modeling parameters are elucidated through 3D plots, particularly highlighting the impact of changing H2/CO proportion on metallization degree and carbon deposit.

Modeling of Gaseous Reduction of Iron Oxide Pellets Using Machine Learning Algorithms, Explainable Artificial Intelligence, and Hyperparameter Optimization Techniques

Hosseinzadeh,

Kasiri,

Rezaei

2024

Optimization of Hydrogen Utilization and Process Efficiency in the Direct Reduction of Iron Oxide Pellets: A Comprehensive Analysis of Processing Parameters and Pellet Composition

Perrone,

Cavaliere,

Sadeghi

et al. 2024

The article deals with the H2 consumption for different processing conditions and the composition of the processed pellets during the direct reduction process. The experiments are carried out at 600–1300 °C, with gas pressures of 1–5 bar, gas flow rates of 1–5 L min−1, and basicity indices of 0 to 2.15. Pellets with different compositions of TiO2, Al2O3, CaO, and SiO2 are analyzed. The gas flow rate is crucial, with 0–10 L min−1 leading to an H2 consumption of 0–5.1 kg H2/kg pellet. The gas pressure (0–10 bar) increases the H2 consumption from 0 to 5.1 kg H2/kg pellet. Higher temperatures (600–1300 °C) reduce H2 consumption from 5.1 to 0 kg H2/kg pellet, most efficiently at 950–1050 °C, where it decreases from 0.22 to 0.10 kg H2/kg pellet. An increase in TiO2 content from 0% to 0.92% lowers H2 consumption from 0.22 to 0.10 kg H2/kg pellet, while a higher Fe content (61–67.5%) also reduces it. An increase in SiO2 content from 0% to 3% increases H2 consumption from 0 to 5.1 kg H2/kg pellet. Porosity structure influences H2 consumption, with the average pore size decreasing from 2.83 to 0.436 mm with increasing TiO2 content, suggesting that micropores increase H2 consumption and macropores decrease it.

Effect of Temperature on CH₄/CO Carburization Process of Hydrogen‐Based Direct Reduced Iron Pellets

Liu,

Hu,

Zhang

et al. 2024

Cementite (Fe3C) has a high combined‐carbon content, which not only increases the carbon content but also provides a heat source and shortens the smelting time. Besides, the CO generated by carbon oxidation can stir the molten pool. It is beneficial to the formation of foaming slag and the removal of inclusions. To study the formation behavior of cementite and surface‐deposited carbon during the carburizing process of hydrogen‐reduction pellets under different conditions, the effects of temperature and carburizing gas composition on the carburizing rate of hydrogen‐reduction pellets are studied. The results show that the carburization rate of CH4 gas is proportional to the carburization temperature. The carburization rate of CO gas is inversely proportional to the temperature, and almost no carburization occurs at 900 °C. The contents of cementite and deposited carbon in different weight gain samples are compared by X‐ray diffraction (XRD) refinement and high‐pressure air gun injection. The analysis revealed that CH4 gas is easier to form deposited carbon in the later stage of carburizing at 900 °C. It provides a theoretical basis for the control of the carburization behavior of hydrogen‐based pellets.