47th ASEAN Glass Conference (AFGM)

Abstract: Enhancing Decarbonization Efforts through Hybrid Melter Design, AI-Based Batch Monitoring, and Model-Based Predictive Furnace Control

The melting of batch and cullet marks the critical first stage of the glass production process. The stability of this phase affected by variables such as pull rate, cullet ratio, glass color, raw material composition, rheological properties, and batch charger behavior is essential for maintaining consistent glass quality. This is particularly true for hybrid melters, where a significant share of the melting energy comes from electric boosting, making the process more sensitive to fluctuations.

As decarbonization efforts drive higher levels of electric boosting, the resulting changes in thermal convection patterns can displace the batch further downstream, increasing the risk of instability. In cold-top electric furnaces, batch thickness plays a vital role in insulating the melt and reducing heat loss. However, fluctuations in pull and cullet levels can cause significant variations in batch thickness, until now, a parameter that has been difficult to measure and integrate into automated control systems.

This paper presents recent advancements in hybrid melter design and the integration of AI-based batch monitoring with model-based predictive furnace control. These innovations enable continuous 24/7 monitoring and dynamic adjustment of furnace parameters, improving process stability and product quality. The paper also explores the operational impact of these technologies on furnace corrosion and longevity, offering insights into how they support broader decarbonization goals across gas-fired, hybrid, and cold-top electric melters.
 

Image