Optimization of Top Gas Recycle Blast Furnace Emissions with Implications of Downstream Energy

摘要

A number of studies have recently been reported on the potential of the top gas recycle blast furnace to reduce carbon emission. Different modeling approaches have been suggested for predictive modeling of the furnace behavior. The present paper is an extension of the stoichiometric modeling approach adopted by the authors wherein process optimization has been attempted in the context of integration of the top gas recycle blast furnace within an integrated steel plant. The considerations of downstream energy available from this process become important for retrofitting the furnace in an existing steel plant. The optimization approach includes use of non-linear Artificial Neural Network as well as linear regression applied to the outputs of the furnace model for a set of input variables. Both the approaches successfully predicted the outputs of the furnace model. Optimization of the output from non-linear and linear regression for different values of downstream energy is performed, leading to optimal values of CO2 emission, carbon rates, and productivity for different downstream energy values. The results may be utilized for choice of appropriate input parameters to attain a specified downstream energy value.