Title:
Potential of a Machine Vision-Based Combustion Monitoring System in Optimizing Step-Grate Biomass Combustion
Author(s):
Garami, A., Tóth, P., Kókai, P.
Document(s):
Paper
Slide presentation
Abstract:
Biomass combustion is one of the oldest methods for energy generation. Recently, the demand for biomass fuels has been increasing due to the incentive to move towards sustainable energy production and its advantageous properties. On the other hand, regardless of being used as a pure fuel or co-combusted in retrofit technologies as part of a fuel mixture, biomass combustion poses several technical challenges mostly due to heterogeneous fuel quality and slagging and fouling issues. The most widely used technologies for biomass combustion are step-grate boilers. The sustainable and efficient operation of these boilers can be difficult, as the properties of the fuel can change considerably depending on the source and processing. The main objective of this work is to minimize emission levels and optimize thermal efficiency of a 3 MW, grate-fired biomass boiler without on-line fuel analysis systems by applying novel combustion process control based on routinely measured operating parameters and real-time flame image processing and machine learning. Two important tasks of image based combustion monitoring systems are providing alerts and predictions regarding the state of the system. The system issues alerts based on the location of the reaction zone and predicts boiler performance based on image and operating data.
Keywords:
biomass, combustion, step-grate boiler, grate monitoring, flame imaging, machine learning
Topic:
Industry Sessions
Subtopic:
Power & Heat processes and systems
Event:
25th European Biomass Conference and Exhibition
Session:
ICO.12.3
Pages:
1945 - 1952
ISBN:
978-88-89407-17-2
Paper DOI:
10.5071/25thEUBCE2017-ICO.12.3
Price:
FREE
