DESARROLLO DE UN MODELO PARA LA PREDICCION DEL ESPUMADO DE LA ESCORIA EN UN CONVERTIDOR DE ACERO

Abstract

Obtaining of steel from pig iron coming from blast furnaces is mainly done in special furnaces called BOF (basic oxygen furnace) converters. During operation, a foamy slag is created to improve the converter performance. Problems arise when the foam level exceeds the height of the vessel and overflows, causing metal loss, process disruption and environmental pollution. This article presents works carried out to analyze the process data in order to predict the emergence of this undesirable phenomenon before it is visible, so that the operator can take the appropriate measures. Data mining techniques has been used to define levels of risk from the initial conditions of each heat. Moreover, the characterization of decarburization curves obtained from spectrometers measurements has been used to generate a predictive model capable of generating alarms beforehand the emergence of the phenomenon. Keywords: Steel manufacturing, BOF converter, foaming slag, data mining