Optimisation of the lance injection desulphurisation of molten iron using dynamic modelling

Abstract

Pneumatic injection of calcium carbide and, in some cases, fluxes via a submerged lance has become the accepted way to desulphurise crude molten iron, as tapped from a blast furnace or COREX furnace. A number of complex phenomena take place during the injection process which require an investigation of the kinetics, thermodynamics, transport processes and overall process dynamics of both the molten iron and the top-slag layer to predict the dynamic removal of sulphur from hot metal. A one dimensional model in terms of lance injection depth is presented, which is dynamically updated for the duration of the blow. The model takes into account the kinetics, thermodynamics and transport processes to predict the sulphur levels in the hot metal throughout a blow. A sensitivity analysis is performed to aid in the optimal selection of operating parameters. The model was implemented in MATLAB and calibrated with data from the Mittal Steel plant in Saldanha, South Africa. The modelling results are significant in that it allows the selection of operating conditions to minimize the processing time for desulphurisation, and the minimization of the heat loss (that has to be regained in the subsequent steelmaking operations).