Non-isothermal pyrolysis of xylan, cellulose and lignin: A hybrid simulated annealing algorithm and pattern search method to regulate distributed activation energies

Abstract

Biomass transformation into fuels, platform chemicals and materials contributes to developing sustainability and carbon neutralization. Pyrolysis is one of the most important approaches to valorize biomass. The distributed activation energy model (DAEM) is a comprehensive kinetic model for describing biomass pyrolysis kinetics. The estimation of the DAEM parameters is difficult because of the complex structure of the DAEM equation. This work formulated the numerical calculation approach of the DAEM and proposed a hybrid simulated annealing (SA) algorithm and pattern search (PS) method for optimizing the DAEM parameters. The DAEM kinetic parameters were evaluated by simultaneously fitting the experimental kinetic data of cellulose, xylan and lignin pyrolysis at various heating rates with the DAEM using the proposed hybrid optimization method. The results showed that the proposed hybrid optimization method could effectively and accurately determine the DAEM parameters and the DAEM with the optimal parameters could reproduce the experimental kinetic data of xylan, cellulose and lignin pyrolysis very well. The distributed activation energies of xylan, cellulose and lignin predicted by hybrid optimization were centered at 162.8, 222.8 and 236.7 kJ mol−1, respectively, with standard deviations of 4.6, 0.8 and 25.9 kJ mol−1, respectively. The reaction orders of xylan, cellulose and lignin pyrolysis are 1.7, 1.1, and 2.5, respectively.