Measurement and Evaluation of Experimental Data for Modelling Thermal Decomposition of Solid Materials

This paper studies a method of estimating the kinetic parameters for wood thermal decomposition as an input for pyrolysis modelling. Samples of birch and pine wood were firstly examined using Thermogravimetric analysis (TG) both in oxidative (air) and non-oxidative (nitrogen) atmosphere, studying the influence of the heating rate (5, 15, 25 degrees C.min(-1)) on the wood mass loss rate. For a single step thermal degradation reaction in non-oxidative atmosphere activation energy and preexponential factor was determined from experimental data using isoconversional method. The results were used as initial estimates for optimization technique - genetic algorithm (GA). Using GA, three thermal degradation schemes of wood were studied. GA was proved to be effective tool for estimating kinetic parameters of pyrolysis models. Contrary to isoconversional method GA does not require user's participation during the parameter estimation process and is able to handle several parameters at once.