Intelligent sensor for predicting the quality of reduced iron in direct reduction furnaces

Abstract

Direct Reduction Iron (DRI) furnaces are used to produce iron from iron ore oxides using natural gas. The furnace takes the iron ore in the form of spherical pellets and a mixture of hydrogen and carbon monoxide and produces reduced iron. Accurate estimation of the quality of the reduced iron is essential for proper control and efficient operation of the DRI furnaces. In order to understand the various factors influencing the quality of the produced iron a mathematical model from the literature was utilized for the calculation of the solid and gas flow characteristics inside the DRI furnace. The model presents the differential equations governing the variations of the substance and energy exchange inside the shaft furnace. The objective of this work is to determine the influences of the various operating parameters on the performance of the DRI furnace. In addition to the mathematical model, investigation is carried out to develop a Neural Network model for on-line estimation of the quality of the reduced iron product based on the available process measurements.