| About this Abstract |
| Meeting |
2022 TMS Annual Meeting & Exhibition
|
| Symposium
|
Cast Shop Technology
|
| Presentation Title |
Aluminium Melting Process Pptimization with SmartMelt, a Digital Tool for Real-time Operational Guidance |
| Author(s) |
Amin Rostamian, Mario Salgado, Marc Bertherat, Jean-Luc Desbiolles, Michel Rappaz |
| On-Site Speaker (Planned) |
Amin Rostamian |
| Abstract Scope |
Every year, about 70 million tonnes of Aluminium (Al) scrapes are melted worldwide, in gas fired reverberatory furnaces. The efficiency of the melting process in such furnaces is strongly influenced by human-driven decisions and some furnace issues. SmartMelt is a novel IoT based platform that combines real time collection of sensors data, loads data, operators input in a dedicated interface, and a very efficient physical model of the melting process to provide on-line guidance to execute operations at optimal timings while detecting furnace issues. Integrated into the daily operations of two furnaces of 30 and 70 tonnes capacity for periods of 16 and 6 months, respectively, the system has yielded a maximum increase of productivity of about 12% and a reduction of gas consumption of about 11%. SmartMelt successfully detects in real-time deviations of burners performance while also monitoring and recording delays influenced by external factors. |
| Proceedings Inclusion? |
Planned: Light Metals |
| Keywords |
Aluminum, Machine Learning, Other |