The approximation method proposed by van Krevelen and others for estim
ating the reaction order of a temperature-programmed process was recon
sidered, and we point out a limitation of their method. By numerical i
ntegration using a computer, the major characteristics of the reaction
rate of the first-order reaction are clarified solving differential e
quations and expanding the applicable range of the method. In each dif
ferential curve during the pyrolysis of wood, a clear shoulder was obs
erved in the pre-maximum stage. We make clear that the curve with a sh
oulder is a composite curve of two first-order reactions and obtain ki
netic parameters for each stage. Considerable differences existed amon
g the kinetic parameters for untreated and treated woods, and woods of
beech and larch, which may be based on the difference of inorganic an
d organic compositions among them. The differential curve for bark has
a complex shape with some jags. We distinguish between the main and a
ccompanied reactions and obtain kinetic parameters for each reaction.
The differential curve for cellulose shows a very simple and nearly sy
mmetric shape, and activation energy E is different between the first
and last parts divided at the maximum rate temperature. We propose a m
ethod to obtain each kinetic value.