Infrared-based slag monitoring and detection system based on computer vision for basic oxygen furnace

Steel making is performed in large refractory-lined vessels known as basic oxygen furnaces (BOF). During processing, a layer of non-metallic, heterogeneous liquid, called slag is produced. Steel is tapped through a side opening. Slag comes at the end of tapping. Slag carried over during tapping has a negative influence on quality. Iron oxides contained in slag make de-oxidation more difficult and cause re-phosphorisation of the steel. To produce higher steel grades, secondary metallurgy is essential. With this, demand has come-up for reliable methods to avoid slag carry-over during tapping. Detecting the initial presence of slag in the tapping stream and stop tapping at correct time, helps in maximising the amount of steel removed from vessel while minimising the amount of slag carried over to the ladle. The work presented here describes the infrared-based method to detect the presence of slag in the tapping stream and generate an alarm to aid the operator.