Dynamic Optimization of Basic Oxygen Furnace (BOF) Operation

In recent years, there have been several publications that address dynamic modelling of the Basic Oxygen Furnace (BOF)1–5. Dogan et al.1 modelled the carbon content trajectory in a 200-ton BOF with reasonable accuracy. This process model focused on the decarburization rates in various zones and incorporated flux dissolution and scrap melting while required slag data is entered as an input. Sarkar et al.2 were able to better predict the decarburization rates for the same furnace by assuming a significantly higher droplet generation rate. Their model accounts for slag formation and refining reactions. Rout et al.5 built upon Dogan et al.1’s model by including other refining reactions and an additional zone, thereby accounting for slag formation. For these three models the bath and slag temperatures are assumed to increase in a linear fashion. Lytvynyuk et al.4 developed a complete, semi-empirical model for the BOF and validated it against two industrial data sets. Even though the model seems to accurately predict the slag and metal bath compositions at the end of the blow, insufficient information is provided about the calculation of the parameters.