(557d) A Decoupling Methodology for Wood Gasifier Modeling: From Laboratory Scale Kinetic Measurements to Pilot Plant Simulation. Application to a Dual Fluidized Bed Gasifier

Different types of modelings have been developed in literature for gasifiers optimization. They could be simple equilibrium approach - in order to predict an outlet gas composition - or detailed kinetic approach reliable for gasifier scale-up. Until now, there are few available models taking into account close couplings between the chemical reactions, heat and mass transfer and hydrodynamics phenomena [1]. Indeed, such an approach requires the good knowledge of a great number of elementary processes. The determination of reliable kinetic constants is one of the main key points. Unfortunately, many values reported in the literature are only valid in the conditions where they have been measured [2]. In addition, analysis of all the phenomena is made complicated by the fact that they all occur simultaneously inside the gasifier. That is why it seems interesting to study the main involved elementary phenomena such as chemical reactions independently of each other.

Among the possible wood gasification technologies, the Dual Fluidized Bed (DFB) gasifier operating with steam as gasification agent may be used to produce a high quality syngas [3]. Typically, wet wood pellets enter the gasifier where they undergo heating and several physicochemical transformations. After pellet drying, primary pyrolysis produces char (carbon-rich residue) and volatile products (condensable vapors and gases) resulting in up to 80% wood weight loss [4]. Primary products are then subjected to secondary reactions occurring outside the pellets:

*vapors thermal cracking by homogeneous reactions occurring in the gas-phase and also through catalytic reactions with the bed material (e.g. olivine),

*gas reactions (water-gas shift and steam methane reforming),

*char-steam gasification.

In the present study, the main involved chemical reactions are decoupled and studied independently at the laboratory scale in thermal conditions close to those encountered in the DFB gasifier. For that purpose, original facilities are developed to follow the wood, vapors and char conversions.

Solid (i.e. wood pyrolysis) and gas/solid (i.e. char-steam gasification and vapors catalytic thermal cracking) reactions are carried out with an image furnace, a source of controlled concentrated radiation [4]. The solid sample (i.e. wood, char or olivine pellet) is submitted to heat flux densities representative of those prevailing in a fluidized bed (i.e. from 0.2 to 0.8 MW/m2). The sample is placed inside a transparent quartz reactor swept by an inert (N2) or reactive (N2/H2O, N2/H2O/volatile products) gas. The gaseous species issued from the sample surface are rapidly quenched by mixing with the sweep gas. The gas-phase temperature is significantly lower than that of sample surface in order to study only the reactions occurring at the sample level (quenching). All the products formed by the reactions are recovered and analyzed. Kinetic data for wood pyrolysis and char-steam gasification are extracted from a model taking into account couplings between chemical reactions and both internal and external heat transfers. Kinetic data for vapors catalytic cracking are derived from the experimental values of gas yields determined as a function of the olivine surface temperature. Experiments related to vapors thermal cracking in the gas-phase are carried out inside a Continuous Self Stirred Tank Reactor (CSSTR) where the gas phase temperature and composition are uniform [5,6]. The CSSTR cracking conditions (temperature: 973-1173 K, gas residence time: 0.3 s) are close to those encountered inside the gasifier. Kinetic data for vapors thermal cracking are simply derived from the experimental values of products yields (i.e. gases, vapors and water) measured as a function of the CSSTR temperature.

Chemical controlling steps in the DFB gasifier are then discussed on the basis of a characteristic time analysis. The gasifier is modeled at both pellet and gasifier levels (i.e. coupling between pellet conversion, secondary reactions, solid and gas-phase hydrodynamics and interfaces transfers). The reactor is represented by three different zones in series (i.e. dense, splash and freeboard zones). Finally, the decoupling methodology is validated from the comparison of model results with experimental measurements performed at the 8 MW (fuel power) DFB plant in Güssing/Austria. The good agreement between predicted gas composition (at the gasifier outlet) and temperature (in the fluidized bed) with experimental data proves the reliability of the decoupling methodology.

[1] A. Gomez-Barea, B. Leckner, Progress in Energy and Combustion Science, (2010), doi:10.1016/j.pecs.2009.12.002

[2] O. Authier, M. Ferrer, G. Mauviel, A.E. Khalfi, J. Lédé, 18th European Biomass Conference and Exhibition, Lyon (France), (2010).

[3] C. Pfeifer, R. Rauch, H. Hofbauer, Industrial & Engineering Chemistry Research, (2004), 43, 1634.

[4] O. Authier, M. Ferrer, G. Mauviel, A.E. Khalfi, J. Lédé, Industrial & Engineering Chemistry Research, (2009), 48, 4796.

[5] O. Authier, M. Ferrer, G. Mauviel, A.E. Khalfi, J. Lédé, 17th European Biomass Conference and Exhibition, Hamburg (Germany), (2009).

[6] S. Baumlin, F. Broust, M. Ferrer, N. Meunier, E. Marty, J. Lédé, Chemical Engineering Science, 60, (2005), 41.