Application of Genetic Algorithm (GA) to estimate the rate parameters for solid state reduction of iron ore in presence of graphite

In the present article, Genetic Algorithm (GA) has been applied to estimate the rate parameters from experimental data during solid state reduction of iron ore in presence of graphite. Iron ore is charged and reduced as iron ore-graphite composite pellets in a laboratory scale packed bed reactor. The reduction of iron ore has been considered in three elementary steps, namely hematite to magnetite, magnetite to wustite and wustite to iron. Assuming all the above conversions follow the first order kinetics, mass balance equations have been setup to define the rate of conversion of various phases of iron oxide, as well as metallic iron. The unknown parameters in mass balance equations, namely three sets of frequency factors and values of activation energy, have been estimated by coupling mass balance equations with GA. The optimum values of the rate parameters have been obtained by minimizing the difference between predicted and experimental degree of reduction. After optimization, reasonable agreement between experimental and predicted degree of reduction is obtained. The model has been utilized to calculate the evolution of various phases of iron oxide as well as metallic iron during reduction. Based on model predictions, it is found that hematite depletes fastest in the pellet, followed by magnetite, which passes through a maximum. Wustite initially builds up and gradually decreases while iron increases steadily. Temporal evolution of various iron oxide phases and metallic iron is also obtained through XRD and microstructural studies, which are found to be consistent with that predicted through kinetic model. The application of GA to estimate the rate parameters for iron ore reduction is unique and applied successfully for the first time.