Kinetic Modeling of the Hydrocracking of Fused-Ring Di-Aromatic Species

Abstract

The derivation of kinetic models for the development and simulation of hydrocarbon processes require detailed information on the reaction network. In the present work, experiments on the hydrocracking of fused-ring species, specifically naphthalene and 1-methylnaphthalene were performed. The feed was diluted with n-heptane, and the experiments were conducted in a fixed bed catalytic reactor loaded with Pt/Pd HY zeolite catalyst in the temperature range 275 °C to 350 °C, at pressures between 28 and 35 bar, space times between 21 to 144 gcat h/mol of reactant and with hydrogen/hydrocarbon molar ratios ranging from 100 to 340. With naphthalene as the feed as many as 16 components were detected and identified in the reactor effluent, with 1-methylnaphthalene up to 42. The effect of catalyst deactivation on the rates of different reactions and elementary steps was studied, and the effects of temperature, pressure and space time on the product distribution in the absence of catalyst deactivation were observed. Following the rules of carbenium ion chemistry detailed networks in terms of elementary steps were generated for each feed species. The network for naphthalene contains a total of 16 reactions on the metal sites and 46 elementary steps on the acid sites, that for 1-methylnaphthalene 123 reactions on the metal sites and 258 elementary steps on the acid sites. Typical pathways to a given product include dehydrogenation/hydrogenation of ring structures on the metal sites and protonation/deprotonation, ring contraction/expansion, PCP branching of side chains, and endocyclic β-scission on the acid sites. The presence of Pt in the catalyst also led to side chain and ring-opening hydrogenolysis.

The single-event modeling approach aimed to reduce the number of independent parameters to be determined from the experimental data for the large number of reactions and elementary steps in the networks. The frequency factors on the acid sites were modeled using single-event kinetics and the activation energies using Evans-Polanyi relationship. On the metal sites the nature of the reacting species and the type of reaction determined the number of parameters. The kinetic model contains hundreds of parameters. Of those, there are only 25 independent parameters for steps on the acid sites, and 16 independent parameters for reactions on the metal sites. An optimization strategy was developed by the sequential use of three different optimization methods to fit the experimental data.