Experimental investigations on visualization of three-dimensional temperature distributions in a large-scale pulverized-coal-fired boiler furnace

Abstract By using a novel flame image processing technique, the 3-D temperature distributions of combustion in the pulverized-coal-fired boiler furnace of a 200 MW power generation unit were visualized experimentally. With the assumption of gray radiation, multiple color flame image detectors were used to capture approximately monochromatic radiation intensity images under the visible wavelengths of red (R), green (G), and blue (B), and a blackbody furnace was used to calibrate the color images. A way to calculate the radiation intensity using the Monte Carlo method was used to establish the relationship between the radiation flame images and the 3-D temperature distribution inside the furnace, given the radiative parameters such as the absorption and scattering coefficients of the combustion medium. A modified Tikhonov regularization method was used to reconstruct the 3-D temperature distribution from the 2-D flame temperature images transformed from the color flame images. Eight flame image detectors were mounted close to the four corners in four layers in different heights in the furnace, the 3-D temperature distribution in 1200 discrete meshes in 12 layers along the height of the furnace was reconstructed from the eight flame images, and the visualization result was updated within 5 s. The errors between some local temperatures visualized by the present system with those measured by a suction thermocouple were within ±5%, showing applicability of the method. On-line monitoring results demonstrated that the average temperature inside the furnace changed directly proportionally with the load of the unit. The preliminary results show that this technology will be useful for the combustion diagnosis and control in coal-fired power generation plants.