Biomass pyrolysis: Kinetic modelling and experimental validation under high temperature and ﬂash heating rate conditions Capucine Dupont a, *, Li Chen a , Julien Cances a , Jean-Michel Commandre b , Alberto Cuoci c , Sauro Pierucci c , Eliseo Ranzi c a CEA, 17 rue des Martyrs, 38054 Grenoble cedex 09, France b RAPSODEE, UMR-CNRS 2392, Ecole des Mines d’Albi-Carmaux, 81013 Albi CT cedex 9, France c CMIC Politecnico di Milano, Piazza Leonardo da Vinci, 32, 20133 Milano, Italy 1. Introduction Biomass is a renewable, CO 2 -neutral energy resource, widely available and increasingly used as an alternative to fossil fuel for energy supply. The thermal conversion of biomass to produce fuel gas (mainly CO and H 2 ) via gasiﬁcation, of which pyrolysis is the ﬁrst step, is considered as a very promising process. It is well known that biomass pyrolysis is a complex process that involves mass and heat transfer phenomena as well as chemical reactions. Furthermore, the chemical reactions during pyrolysis can be focused on three different aspects:  Biomass devolatilization: i.e. the decomposition of the solid into permanent gases, condensable vapours (tars) and solid residue (char);  Secondary gas phase reactions of the released gas and tar species;  Heterogeneous reactions between solid and gas. The way these reactions occur, and therefore the ﬁnal product yield, is strongly related to operating conditions and possible heat and mass transfer resistances. Several researchers have studied the intrinsic kinetic of biomass devolatilization with small particles (100 mm). Network biomass devolatilization models describe accurately the chemistry of the devolatilization process [1]. The FG (functional group)-Biomass model [1–3], the Bio-Flashchain model [4], and the bio-CPD (chemical percolation devolatilization) model [5,6] are examples of these network models. They assume that the biomass macromolecule is constituted of different lumped groups, and the macromolecular fuel structure changes during the devolatilization process to produce gas, tar, and char. These models, initially developed for coal devolatilization, have been recently extended to biomasses, but their availability and validation are still limited. Research efforts have also been devoted to the pyrolysis of large (said as ‘‘thermally thick’’) biomass particles [7–10] in order to study the inﬂuence of heat and mass transfer limitations and the evolution of solid physical properties. Usually the chemistry of biomass devolatilization is roughly sacriﬁced. As clearly stated by Di Blasi [11], as well as by Janse et al. in modelling the ﬂash pyrolysis of a single wood particle [12], the accurate knowledge of the reaction kinetics appears to be a crucial parameter for a reliable modelling of the pyrolysis process. Besides, the main conclusion of a recent literature review on the modelling of biomass pyrolysis [13,14] underlines that the available knowledge on kinetics and transport phenomena has J. Anal. Appl. Pyrolysis 85 (2009) 260–267 ARTICLE INFO Article history: Received 30 June 2008 Received in revised form 18 November 2008 Accepted 19 November 2008 Available online 3 December 2008 Keywords: Biomass Pyrolysis Modelling Devolatilization Gas phase kinetics Entrained ﬂow reactor ABSTRACT This work analyzes and discusses the general features of biomass pyrolysis, both on the basis of a new set of experiments and by using a detailed kinetic model of biomass devolatilization that includes also successive gas phase reactions of the released species and is therefore able to predict the main gases composition. Experiments are performed in a lab-scale Entrained Flow Reactor (EFR) to investigate biomass pyrolysis under high temperatures (1073–1273 K) and high heating ﬂuxes (10–100 kW m 2 ). The inﬂuence of particle dimensions and temperature has been tested versus solid residence time in the reactor. The particle size appeared as the most crucial parameter. The pyrolysis of 0.4 mm particles is nearly ﬁnished under this range of temperatures after a reactor length of 0.3 m, with more than 75 wt% of gas release, whereas the conversion is still under evolution until the end of the reactor for larger particles up to 1.1 mm, due to internal heat transfer limitations. The preliminary comparisons between the model and the experimental data are encouraging and show the ability of this model to contribute to a better design and understanding of biomass pyrolysis process under severe conditions of temperature and heating ﬂuxes typically found in industrial gasiﬁers. ß 2008 Elsevier B.V. All rights reserved. * Corresponding author. E-mail address: capucine.dupont@cea.fr (C. Dupont). Contents lists available at ScienceDirect Journal of Analytical and Applied Pyrolysis journal homepage: www.elsevier.com/locate/jaap 0165-2370/$ – see front matter ß 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jaap.2008.11.034