Multiobjective Optimization of Top Gas Recycling Conditions in the Blast Furnace by Genetic Algorithms

Abstract

Limited natural resources and a growing concern about the potential effect of carbon dioxide emissions on the world's climate have triggered a search of ways to suppressing the emissions of CO₂ in primary steelmaking. A possible future solution is to strip CO₂ from the blast furnace top gas, feeding back the gas to the tuyere level. The work reported in this article explores states of an integrated steel plant that arise if both production costs and emissions are simultaneously minimized. This multiobjective problem is tackled by genetic algorithms using a predator–prey strategy for constructing the Pareto-frontier of nondominating solutions. Four alternative ways of treating the top gas recycling problem are explored, and the resulting solutions are analyzed with respect to the two objectives and to the internal states of the plant they correspond to. Conclusions are drawn concerning the solutions in terms of technical feasibility and complexity.

Keywords:

• CO₂ emissions
• Genetic algorithms
• Multiobjective optimization
• Predator-prey method
• Top gas recycling

ACKNOWLEDGMENTS

Financial support from the Academy of Finland to the GreenSteel project within the Sustainable Energy program is gratefully acknowledged. We also thank Rautaruukki, Finland, for access to process data.