Bio-syngas Conversion by FT Synthesis with High Loaded Fe-based Catalysts: Kinetic Parameters Regression

• Published in: Computer Aided Chemical Engineering Vol. 38; pp. 589 - 594
• Main Authors: Comazzi, Alberto, Galli, Federico, Galimberti, Marco, Manenti, Flavio, Bianchi, Claudia L.M., Pirola, Carlo
• Format: Book Chapter
• Language: English
• Published: 2016
• Subjects: Catalysis; Fixed bed reactor; MATLAB toolbox; Process simulation; Robust regression; Catalysis; Robust regression; MATLAB toolbox; Process simulation; Fixed bed reactor
• ISBN: 0444634282; 9780444634283
• ISSN: 1570-7946
• DOI: 10.1016/B978-0-444-63428-3.50103-X

Biomass-to-liquid Fischer-Tropsch (FT) is an industrial process that converts bio-syngas in hydrocarbons ranging from C1 to C100. Bio-syngas, a syngas mixture produced from biomass, is characterized by a H2/CO molar ratio in the range 1.0–1.5. An iron-based catalyst supported on silica for CO hydrogenation with 30wt% of metal was prepared, characterized by BET, SEM, TEM, TPR, XRD and tested at different temperatures and H2/CO ratios in a FT bench scale plant using a Packed Bed Reactor (PBR). The experimental results demonstrated that this catalyst is suitable for low H2/CO ratios, since by increasing the inlet H2/CO ratio the CO conversion increases, the product selectivity remains largely unchanged. Moreover the catalyst shows a satisfactory stability as a function of time of stream (TOS). A nonlinear regression was performed based on a selected kinetic model in order to obtain the main kinetic parameters. The elaborated kinetic model is based on the hypothesis that both FT and Water Gas Shift (WGS) reactions are active on the catalyst. Initially the regression was performed using MATLAB®, but the algorithm could not lead to an acceptable solution because of the high number of local minima presented by the objective function. In order to overcome this problem, a procedure in C++ has been developed. Respect to the model found in the literature, this new one has been rearranged for minimizing the multicollinearities between the optimized parameters and a satisfactory fitting respect the experimental values has been achieved. Based on the experimental collected data and the regressed parameters, a rigorous multi-scale simulation of reactor behaviour was developed in order predict the reactor yield and conversion.