Reduction of out-of-standard tramp element compositionReduction of main raw material costCu in hot metalEAFScrap

Data Science

Optimization of Scrap Blending by Estimation Model for Hot Metal Cu Concentration in Electric Arc Furnace

Controls Tramp Elements within Standards by Cost-Minimum Scrap Blending

Solution Point

Improvement of prediction accuracy of tramp element (Cu)
Improved prediction accuracy also results in optimization of main raw material blending calculations.

Features 01

Expected Problems

・Are you forced to downgrade or scrap products due to unexpected out-of-standard tramp elements?

・Isn't poor tramp element prediction accuracy causing excessive blending of high-grade scrap?

Features 02

Technology highlight

・In addition to traditional stacking management by scrap brand, a prediction model for tramp elements (in this case, Cu) using machine learning was constructed, and is also used in scrap blending.

Features 03

Proposed Solutions (Examples)

Possible to externally install on existing systems.

Links 01