Hydrodynamic models for the interpretation of the liquid flow in trickle-bed reactors

Abstract

Two hydrodynamic models of the liquid in a trickle-bed reactor: PE and PDE model, have been considered, and the influence of the identification method on the parameter values was examined. The employed identification methods were: the moments method, supplemented by graphical determinations on the response curve, and the best fit method, where the minimum absolute and relative error between the calculated and experimental response curve was adopted as identification criteria. Except the total liquid hold-up, the values of the other parameters of both models were very sensitive to the identification method especially at low liquid flow rates. For this reason, much care must be taken in comparing literature results. The experimental runs were carried out with two tracer of rather different diffusivity; no influence of this on the parameter values was observed; this would mean, in particular, that the mass transfer between dynamic and stagnant zones is attributable to the movement of fluid streams between them.