CFD analysis of bed textural characteristics on TBR behaviour: Kinetics, scaling‐up, multiscale analysis, and wall effects. (18th October 2018)

• Record Type: Journal Article
• Authors: Uribe, Sebastián; Cordero, Mario E.; Zárate, Luis G.; Valencia López, José Javier; Natividad, Reyna
• Is Part Of: Canadian journal of chemical engineering. Volume 97:Number 2(2019)
• Page Start: 485
• Page End: 499
• Publication Date: 2018-10-18
• Subjects: TBR CFD simulation -- hydrodesulphurization -- scale‐up/down process -- multiscale‐multiphysics approach; Chemical engineering -- Periodicals; Technology -- Periodicals; 660.05
• DOI: 10.1002/cjce.23298
• Languages: English
• ISSNs: 0008-4034

Abstract

Abstract: A simulation of a trickle bed reactor aided by computational fluid dynamics was implemented. With a Eulerian approach, geometrical characteristics were explicitly considered and two simultaneous heterogeneous reactions were included, hydrodesulphurization (HDS) and hydrodenitrogenation (HDN). This was performed in order to achieve the following: (1) attain further insight into a proper scaling‐up procedure to be able to obtain the same hydrodynamics and kinetics behaviour in two reactors of different length and diameter scales; (2) develop a multiscale analysis regarding the communication of information between scales through the construction of a porous microstructure model from which the geometrical information of the microscale is captured by the effective transport coefficients (which affect the overall reactor behaviour); (3) investigate the effect of operation condition variations on hydrodynamics and kinetics; and (4) assess the deviations and further differences observed from average to punctual conversion values and the assumptions from kinetic literature models through a preliminary multiscale analysis. The CFD results were validated against experimental pressure drop data as well as HDS and HDN conversion theoretical data. An excellent agreement was found. The model produces a significant improvement in hydrodynamic parameter prediction, achieving 5 times better accuracy in predicting pressure drops and 50 % improvement in holdup prediction. The fully coupled model predicts HDS conversion with 96 % accuracy and HDN conversion with 94 % accuracy. Results suggest that the best way to obtain similar kinetic and hydrodynamic behaviour in TBRs with different lengths and diameter length scales is by equaling the liquid holdup ( ϵ γ ) or the mass velocities (L‐G).