Optimization of Continuous Steel Annealing Operations Using Model Predictive Control at Tata Steel, India.

This article focuses on the development and implementation of a model predictive control (MPC) technology-based supervisory solution for continuous annealing furnaces (CAL) used in steel manufacturing, specifically at Tata Steel, India, to improve the quality and efficiency of automotive-grade cold-rolled steel strips. The MPC system uses a data-driven dynamic model combined with a Kalman filter to predict and optimize furnace temperature setpoints in real time, addressing challenges such as slow furnace dynamics, steel grade transitions, and multivariable interactions within furnace zones. Since becoming fully operational in January 2023, the MPC solution has increased the proportion of steel meeting tight temperature quality bands from 30% to 50%, reduced fuel consumption by 8% per ton of steel, and prevented the reprocessing of approximately 13,000 tons of material annually, translating to estimated annual savings of US$2.5 million and a reduction of 10,000 tons of CO₂ emissions. The article also discusses the integration of the MPC with existing legacy control systems, change management for plant operators, and plans to extend this data-driven predictive control approach to other steel manufacturing processes.