ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Share
Publications Copernicus
Download
Citation
Share
Articles | Volume X-G-2025
https://doi.org/10.5194/isprs-annals-X-G-2025-117-2025
https://doi.org/10.5194/isprs-annals-X-G-2025-117-2025
10 Jul 2025
 | 10 Jul 2025

Flame Length and Geometry as Key Indicators for Real-Time Thermal Stress Monitoring in a Multi-Burner Combustion Chamber

Matthias Arnold, Mareike Hoyer, and Sina Keller

Keywords: Industrial Burner Flames, Combustion Control, 2D Measurements, Flame Length, Deep Learning

Abstract. This study investigates the relationship between flame length measurements and the temperature of the refractory lining in combustion chambers with multiple burners interacting in terms of fluid dynamics. As this has not been done before, this study investigates the general feasibility and can serve as a baseline for future studies. The data stems from the post-combustion chamber of a hazardous waste incineration plant, where waste is burned at very high temperatures to minimize pollution. It was recorded by a near-infrared camera, which samples pixel-wise temperatures with a frequency of 50 Hz. Four burners mounted at the same height are considered within the chamber, producing four flames mutually influencing each other. The flames are segmented with a deep-learning model. The acquired flame masks serve as a basis to determine individual flame properties. We introduce burner axis flame length and medial axis flame length and show that the shape or bending of a flame can be deduced depending on the relationship between those two lengths. Combined with their magnitude, these lengths indicate whether certain parts of the refractory wall are at risk of exposure to thermal stress, making the implementation of an automatic, generalizable, and transparent combustion control system feasible. Finally, we also highlight that all burners must be considered for effective automatic control in a complex multi-burner geometry. This system optimizes burner settings based on the observed and measured flame geometry, maximizing throughput while minimizing refractory wear, thus improving sustainability.

Share