Process Furnace Optimization – Hybrid Modeling & AI Analytics | SABIC 2024
- Challenge
- Limited visibility into remaining furnace run length, causing unplanned decoking
- Reactive monitoring delayed anomaly detection and corrective action
- High operating cost and yield losses due to unstable furnace performance
- AIGC Solution
- Hybrid AI model combining first-principle simulations with data-driven analytics
- Real-time condition monitoring with anomaly detection and advisory recommendations
- KPI and run-length forecasting integrated with process historians
- Interactive dashboards for furnace performance and diagnostics
- Realized Value
- Extended furnace run length and reduced decoking frequency
- Lower operating costs through early anomaly detection
- Improved process yields and operational stability