Reduced-Order Models of Furnace Tapping Systems—A Case Study from a Submerged Arc Furnace Producing Silicomanganese

With the advent of the fourth industrial revolution, advanced methods for control and automation of pyrometallurgical furnace plants are beginning to receive considerable attention. One aspect of such work is the development of predictive tools based on fundamental process modelling—such models are able to extrapolate furnace behaviour beyond what is available in historical data. In the present paper, a reduced-order model of tapping from a submerged arc furnace is developed for the case of silicomanganese production. The utility of the resulting model is assessed by comparison to tapping measurements obtained from an industrial furnace plant. Strengths and weaknesses of the model are subsequently identified and discussed in the context of the potential for such models to act as operator guidance and digital twinning tools.