Contribution to the modeling of packed bed reactors under plugging conditions in single and two phase trickle flow
|Advisor:||Larachi, Faïcal; Grandjean, Bernard|
|Abstract:||Trickle bed reactors (TBR) behave as deep bed filtration (DBF) units when the liquid feedstock is contaminated with fine particles. Solid retention causes an ever increasing pressure drop in the bed that leads to eventual halting of the installation. Industry response has been so far to change the plugged, but still active, catalytic bed with a fresh catalyst packing causing important profit losses of the process. In this work two levels of analysis are proposed for the DBF in single and two phase trickle flow conditions. At bed scale, an Eulerian-Eulerian CFD approach is used that provides the framework for the insertion of closure equations for the mass transfer in DBF. At pore scale, Trajectory Analysis (TA) is used as is an accepted procedure for the analysis of Deep Bed Filtration (DBF) in single-phase aqueous systems. In single phase flow through packed beds, the known TA based expression of Rajagolapan and Tien (1976) is used. By inserting this expression in the CFD approach it becomes possible to obtain valuable information about the transient structure and development of plugging. Benchmarking was obtained with the work of Narayan et al. (1997). In two phase trickle flow, no TA approach is known so far and an all new extension of this methodology is proposed in this work. Using a film model to represent the trickle bed reactor, TA analysis is performed in single phase, one-layer and multilayer deposition in TBR conditions. TA tendencies were akin to the analysis of Rajagopalan and Tien (1976) demonstrating that deposition mechanisms are of the same kind as in the single phase flow only modified by the presence of the gas phase. Results were compared with the data of Gray et al. (2002). Static liquid hold-up (SLH) is a parameter that, according to experimental observations, affects significantly solid deposition in multiphase conditions. An almost exhaustive collection of SLH values was constructed from the available experimental literature. With the SLH data and with the use of a minimum energy algorithm, average contact angles for a wide range of liquids and packing were obtained. Reinserting the calculated contact angles in neural network software, a correlation was obtained which outperforms all the available correlations. It is hoped that in future work, this last parameter, the SLH, will be included in the plugging model or at least in the trajectory analysis at the collector scale.|
|Document Type:||Thèse de doctorat|
|Open Access Date:||11 April 2018|
|Collection:||Thèses et mémoires|
All documents in CorpusUL are protected by Copyright Act of Canada.