About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Energy Technologies and CO2 Management: Resource Efficient Processes
|
| Presentation Title |
Prediction of Linz-Donawitz Converter Gasholder in Iron and Steel Enterprise |
| Author(s) |
Xiancong Zhao, Yong Ma, Peng Zhao, Song Gao, Wei Li, Fang Guo, Pengfan Ren, Zhanxin Wu, Yingqin Wang, Minghui Chi, Zhongheng Chen, Hao Bai |
| On-Site Speaker (Planned) |
Yingqin Wang |
| Abstract Scope |
Linz-Donawitz converter gas (LDG) generated during steel-making process is an important secondary energy source in steel plant. However, its generation subjected to significant fluctuations due to multiple factors such as process requirements and production schedules. To solve this problem, LDG gasholder served as buffer units for temporary gas storage. Given their limited capacity, gasholder frequently experience levels that are either too high or too low. Therefore, it is necessary to predict and control the LDG holder level in a future period, so as to reduce gas flaring or shortage. Previous studies based on the premise that the measurement of LDG consumption flow was accurate. However, in actual production, measurement deviations of LDG widely existed. If actual measurement data is used in the prediction model, may lead to the continuous accumulation of measurement deviation. In this paper, a real-time prediction model considered the measurement deviation of LDG consumption is proposed. On-field test results indicated that the accuracy for the proposed model at a 30-minute horizon and 60-min horizon is 92.4% and 87.8%, respectively. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Energy Conversion and Storage, Machine Learning, Recycling and Secondary Recovery |