A comparison of fuzzy logic and cluster renewal approaches for heat transfer modeling in a 1296 t/h CFB boiler with low level of flue gas recirculation

What is it about?

The interrelation between fuzzy logic and cluster renewal approaches for heat transfer modeling in a circulating fluidized bed (CFB) has been established based on a local furnace data. The furnace data have been measured in a 1296t/h CFB boiler with low level of flue gas recirculation. In the present study, the bed temperature and suspension density were treated as experimental variables along the furnace height. The measured bed temperature and suspension density were varied in the range of 1131-1156K and 1.93-6.32kg/m3 respectively. Using the heat transfer coefficient for commercial CFB combustor, two empirical heat transfer correlation were developed in terms of important operating parameters including bed temperature and also suspension density. The fuzzy logic results were found to be in good agreement with the corresponding experimental heat transfer data obtained based on cluster renewal approach. The predicted bed-to-wall heat transfer coefficient covered a range of 109-241W/(m2K) and 111-240W/(m2K), for fuzzy logic and cluster renewal approach respectively. The divergence in calculated heat flux recovery along the furnace height between fuzzy logic and cluster renewal approach did not exceeded ±2%.

Featured Image

Why is it important?

Experimental research is needed to elucidate the heat transfer characteristic during the recirculation of flue gases back to the furnace through star-up burners. The fuzzy logic and cluster renewal approach were used to predict bed-to-wall heat transfer coefficient. This work provides a correlation between heat transfer coefficient and bed temperature whether or solid suspension density. The heat transfer data obtained in identical operative conditions for fuzzy logic and cluster renewal approach afterwards were compared with each other. These data are useful in the design and scale-up of heat exchangers operated in a fast fluidization flow regime within the combustion chamber. This work would stimulate discussion and development of these approaches for heat transfer modeling in large-scale CFB combustors.