The SAFECI project has made important strides in understanding how electricity behaves inside smelting furnaces used to produce ferroalloys—materials essential for steelmaking and other industries. This new knowledge will help make Norwegian ferroalloy production more energy-efficient and environmentally friendly in the years ahead.
By the end of the project, two industrial partners have implemented new types of electrical measurements. These innovations will improve how furnaces are controlled, leading to better performance and reduced energy use.
SAFECI also contributed to building expertise in the field: 10 researchers and professionals received training, and the project led to 21 scientific publications and one successful follow-up research proposal.
The project was led by NORCE, with NTNU as a research partner, and included international collaboration with the University of Padua. The industrial partners were Elkem ASA, Eramet Norway AS, Finnfjord AS, and Wacker Chemicals Norway AS.
The project has combined physics-based modeling with data-driven approaches and novel measurements:
Physics-based finite-element models from the ElMet project (project number 247791) have been continuously updated and used to simulate issues such as large electrode movements, unbalanced electrical conditions, induction and voltage measurements, and magnetic measurements. Computationally lightweight models (metamodels) have been derived from the finite-element models and are available online at https://safeci.web.norce.cloud/. Metamodels are useful for operator training and for studying controller performance.
The project also explored new ways to predict electrical resistance in electrodes and started developing a framework for combining live process data with metamodel predictions. This work continues in the project SAPPHIRE (project number 358034).
SAFECI reviewed the electrical measurements currently used in industry. Additional electrical measurements have been proposed, tested, and are scheduled for industrial implementation. The project also used circuit models to explore how energy can be saved by making changes to the furnace power supply.
The project also carried out detailed laboratory measurements of the bulk resistivity of raw materials used in ferroalloy production. These measurements were performed across a range of temperatures and included both pure materials and partially transformed mixtures. The resulting dataset provides a valuable reference for understanding how electrical properties change during the smelting process. This knowledge is essential for improving the furnace models and will serve as a foundation for future research (SAPPHIRE, project number 358034) and industrial applications.
By the end of the project, two industrial partners have implemented new types of electrical measurements. These innovations will improve how furnaces are controlled, leading to better performance and reduced energy use.
The potential of two new types of electrical and magnetic measurements has been demonstrated through industrial measurement campaigns. These methods provide hereto unavailable information about the inner furnace conditions.
The project will address the following priority set out by ENERGIX:
• Improved automation and control systems for achieving major gains in energy efficiency
The industrial partners (Elkem AS, Eramet Norway AS, Finnfjord AS and Wacker Chemicals Norway AS, major players in the production of silicon and ferroalloys) have identified that improved electrical conditions in the Submerged Arc Furnaces (SAFs) lead directly to more stabilized operation, more optimal energy distribution and enhanced furnace efficiency (lower kWh/kg product), with corresponding energy savings.
Electrical Conditions in smelting furnaces have been studied in the KPN project "Electrical Conditions and their Process Interactions in High Temperature Metallurgical Reactors (ElMet)". This project has provided very good insight based on a wide range of first-principle models. It was also tested how the results from several large-scale FEM (Finite Element Method) simulations can be "concentrated" into metamodels. These are surrogate models that retains the same generalization capabilities as the original FEM models, while being computationally lightweight.
The project intends to explore such metamodels and combine them with data-driven modelling into "Digital Siblings" (a major step towards appropriate future Digital Twins), mirroring the typical electric behaviour of SAFs. This tool will then apply existing operational data, combined with some required new measurements, to identify the inner, hidden, electrical states within the furnaces.
To succeed, the project needs to be highly interdisciplinary, putting together:
• Metallurgical knowledge at university level
• Metallurgical know-how from metallurgists, operators, etc. from the partner companies
• Physics based mathematical modelling of SAFs, especially electrical conditions
• Data based modelling of processes
• Big Data Cybernetics, including artificial intelligence (AI) and machine leaning (ML)
• Measurement technology
