Achieving precise chemical composition in steelmaking is critical for quality and cost control. Traditionally, operators used experience or spreadsheets to add Ferro alloys to adjust chemistry levels, leading to inconsistent final chemistry, excessive use of costlier alloys, and longer production times. This lack of standardization affected product quality and efficiency.
To address this, a Python-based Data Science model using Linear Optimization was developed. Integrated via a Flask web app, it recommends optimal Ferro alloy quantities in real time based on recovery rates, chemistry targets, and alloy prices. This ensures accurate composition with minimal adjustments and lower costs during live production. A Power BI dashboard, supported by Azure services, offers insights into alloy consumption, costs, inventory, and compliance. Data from multiple sources is standardized for visualization and monitoring.
The solution has delivered cumulative savings of ₹133.64 Crores (up to October 2024) across multiple plants, with an ROI of less than one month.