(586g) Semi Analytical Solution of a Heat Transfer and Kinetic Models Applied in a Biomass Pyrolysis Reactor

Currently, fuels renewal and innovation has been intensively investigated by the scientific community. Nevertheless, thermal cracking (pyrolysis) is considered a promising technique in biofuels production. This paper proposes a mathematical model for heat transfer phenomena and chemical conversion applied in a biomass pyrolysis reactor. Equations set were solved semi analytically; using a combination of Finite Analytical and Laplace Transform methods. Thus, it was possible to incorporate non-linear terms in the model solution and a numerical procedure, in which time interval can be made arbitrarily small without stability and convergence problems. The thermochemical process is a decomposition reaction (reduction) occurring at higher temperatures, in an environment quasi or deoxygenated – a rupture of the original molecular structure of a given compound or mixture by heating. This is self-sustaining from an energy standpoint, since the break in the oxygen absence in a waste processing reactor produces surplus energy. The biomass pyrolysis reactor was simulated, based on works of heat and mass transfer models and solved by Walas [1959], modified by Leung et al. [1965] and also in pyrolysis reactors by Bertoli [2000], Meier et al. [2009] and Wiggers et al. [2009]. Reactor wall is the control system, perpendicular areas (inlet and outlet) are the control surface; steady-state, uniform average temperature, adiabatic reactor, inert particles, first order reactions were the simplifying hypothesis of the mathematical model. Semi analytical solution designed considering the nonlinearity of the equations system by isolating the reactor in finite intervals, keeping the linear terms and evaluating the temperature range at beginning. Finite Analytical Method is based on the numerical solution incorporating a local solution of partial or ordinary differential equations. For data comparison (semi analytical versus numerical solutions) Runge-Kutta-Fehlberg routine was used. The proposed method results agree with values obtained from literature.