Molecular modeling of the catalytic hydrocracking of complex mixtures: Reactions of alkyl aromatic and alkyl polynuclear aromatic hydrocarbons

Increasing legislative and environmental concerns have focused attention on determining molecular composition of heavy oil feedstocks and their refined products. However, tracking the 10{dollar}sp3{dollar}-10{dollar}sp5{dollar} molecular constituents during heavy oil upgrading has posed a unique challenge to traditional process modeling, in which the molecules are grouped according to global properties, such as boiling point fractions and solubilities.; In an effort to incorporate chemically significant information into a process model, a molecular approach was used to simulate the hydrocracking chemistry of example heavy gas oil mixtures. This approach consisted of three elements: (1) the development of reaction pathways and kinetics for model systems, (2) the formulation of Linear Free Energy Relationships (LFERs) to account for reactivity dependence on reactant structure, and (3) the incorporation of the pathway analysis into a simulation of complex heavy oil mixtures.; The hydrocracking reaction pathways and kinetics of alkylbenzenes and alkyl polynuclear aromatics (PNAs), likely present in typical gas oils, were investigated over a NiW/USY zeolite. Langmuir-Hinshelwood-Hougen-Watson (LHHW) rate parameters were determined for an extensive model compound database that included hydrocracking reactions of alkyl aromatics with side-chains ranging from 0-19 and aromatic rings from 1-4. The principal reactions observed for alkylbenzenes and alkylPNAs were ring-dealkylation, side-chain cracking and ring-closure. Additionally, alkylPNAs underwent hydrogenation.; LFERs existed for ring-dealkylation, hydrogenation and ring-closure. The ring-dealkylation and ring-closure rate constant for C2- to {dollar}sim{dollar}C9 alkylbenzenes followed the intrinsic chemistry behavior. However, the intrinsic correlations failed beyond this length, which signified the onset of extrinsic effects. LFERs for the hydrogenation, ring-dealkylation and ring-closure of alkylPNAs were successfully identified. Although the decrease in reaction rate constants with ring number for dealkylation and closure followed the intrinsic correlations, these results were also consistent with the presence of extrinsic effects as well.; The rate parameters obtained from the reaction pathway analysis of alkylaromatics, alkylPNAs and LFERs were used as input into a molecular model. Monte Carlo techniques were employed to stochastically build and react example gas oil mixtures. The model predictions in all cases were in reasonable agreement to the experimental hydrocracking yields.