ControlEAF

In 2023, France’s total electricity consumption reached approximately 445 TWh, with the steel and metallurgy industries accounting for a significant share of industrial energy use. Electric arc furnaces (EAFs), essential for recycling steel and producing ferro-alloys, are also highly energy-intensive. ControlEAF aims to enhance the operational efficiency and stability of EAFs by optimizing production within operational constraints. The project will develop a flexible hierarchical control architecture to manage electrical parameters, account for key operations like casting and electrode slipping, and respond to disturbances such as gas blows and burden collapses. Pilot tests will take place on a 34 MW SiMn furnace at Comilog Dunkerque, with a production capacity of 60,000 tons and an annual energy consumption of approximately 300 GWh. This interdisciplinary project integrates advanced tools from Electrical and Control Engineering, Optimization, and AI-based techniques. It will design a robust Model Predictive Control (MPC) framework, enhanced with Mixed-Integer Programming and learning-based techniques, to ensure stable, efficient operations closer to the furnace’s limits. By optimizing power distribution and reducing losses from electrodes intensities, ControlEAF will improve furnace efficiency and boost production. ControlEAF promotes energy efficiency and sustainability in France’s industrial sectors. It aims to reduce inefficiencies in high intensity submerged EAFs, which currently account for 15-20% of energy consumption. By optimizing operational parameters, ControlEAF is expected to increase energy efficiency up to 4 %, leading to significant reductions in both energy costs and production expenses. For an EAF consuming 300 GWh annually, a 4 % efficiency improvement saves about 12 GWh per year - enough to power about 1000 homes. Additionally, this would reduce CO2 emissions by around 1000 tons annually, contributing to France’s 2050 Carbon Neutrality Plan.