A comparison of a batch recycle reactor and an integral reactor with fines for scale-up of an industrial trickle bed reactor from laboratory data

Abstract

Observed reaction rates in trickle bed reactors are often strongly influenced by reactor hydrodynamics and transport limitations. The reaction engineer must account for these effects when designing a commercial scale reactor. This report describes the two different experimental approaches used to generate kinetic data for acetophenone hydrogenation to 1-phenylethanol over a copper catalyst. In the first approach, the active catalyst particles were diluted with inert fines. This diluted catalyst was used in a steady-state, single-pass, continuous reactor. The second experimental system used an alternative approach that measured apparent reaction kinetics under conditions of hydrodynamic equivalence to a full-scale reactor. In this approach, a batch recycle trickle bed reactor utilized complete recycle of the liquid reactor product, which was collected in a well-mixed vessel. The results of this work indicate that, for very active heterogeneous catalysts, the batch recycle trickle bed reactor is preferred to the integral reactor with fines for generation of apparent reaction kinetic data for full-sized catalyst pellets. In addition, this work outlines a methodology for developing and validating a kinetic model of an adiabatic trickle bed reactor. One byproduct of this effort is a closed form analytical expression for the effectiveness factor for a specific Langmuir–Hinshelwood rate expression in the limiting case where pore diffusion limitations are so great that one of the reactants is completely consumed before it diffuses to the end of the modeled pore.