Co-combustion characteristics of coal slime and moso bamboo based on artificial neural network modeling and TG-FTIR

The co-combustion behavior and gas product characteristics of coal slime (CS) and moso bamboo (MB) in air atmosphere were investigated by Thermogravimetric analyses (TGA) and Thermogravimetric-Fourier transform infrared spectrometer (TG-FTIR). Principal component analysis (PCA) was selected to determine the contribution rate and identify the main reactions in the co-combustion process. Three model-free methods (Kissinger-AkahiraSunose, Flynn-Wall-Ozawa, and Starink) and Coats-Redfern method were applied to calculate the combustion dynamics. The addition of MB improved the combustion performance of CS. CS interacted most significantly with MB when MB blending ratio was 30 %. When MB ratio was 50 %, the average apparent activation energy (E alpha) was the lowest, which was 19.7-19.8 % lower than that of CS combustion alone. The reaction mechanism of CS could be described by g(alpha) = 1-(1-alpha)1/4 model, and the reaction mechanisms of the blends had been changed during co-combustion. The emission rules of gas products were detected online by TG-FTIR, and the addition of MB was helpful to reduce the emission of CO2, CO, HCN, and SO2 in the high-temperature area. The artificial neural network (ANN) method was chosen to predict the effects of conversion rate and blending ratio on E alpha.