Abstract: Humus is the main object of forest underground fire shaded combustion, and analyzing its pyrolysis characteristics and calculating kinetic parameters are of great theoretical significance for predicting and extinguishing underground fires. In this study, humus from larch woodland was taken as the object of study, and pyrolysis experiments were carried out under three kinds of heating rates using a thermogravimetric analyzer, and a multi-peak fitting model was proposed for the calculation of kinetic parameters of the experimental data. The model was based on the weight loss rate and acceleration rate curves to obtain the characteristic points, and the fitted peaks were calculated at the characteristic points, and four fitted peaks were obtained for each curve, which were utilized to calculate the kinetic parameters. The results showed that: the activation energy of the fitted peak Ⅰ was low, corresponding to the first peak of the DTG curve; the activation energies of the fitted peaks Ⅱ and Ⅲ were in the middle level, corresponding to the first shoulder and the second peak of the DTG curve; the activation energy of the fitted peak Ⅳ was the highest, corresponding to the last shoulder of the DTG curve. The results of the model's parameter calculations coincide with the kinetic parameter intervals calculated by the Gorbatchev integration method. The residual and goodness-of-fit tests were performed on the model, and the P<0.000 1, root mean square error RMSE<0.005, and goodness-of-fit R2≥0.99 of the fit test (F-test) were obtained. The test results indicated that the multi-peak fitting model can effectively decompose the sub-reactions of humus pyrolysis and calculate the kinetic parameter values, which has certain theoretical significance and practical guidance value.

Key words: Larix gmelinii, humus, kinetics, pyrolysis kinetics, multi-peak fitting model