Unrestricted.2016-11-142011-02-182016-11-142003-05http://hdl.handle.net/2346/12866Hydrocracking is used in the petroleum industry to convert low-quality feedstocks into highly-valued fransportation fuels such as gasoline, diesel, and jet fiiel. Hydrocracking is usually carried out in two stages. The first stage decomposes sulfurand nifrogen-containing compounds and hydrogenates the aromatics. The liquid fraction from the first stage is hydroisomerized and hydrocracked in the second stage. The primary objective of the present research is to develop a very detailed, fundamental, and molecular-level model for the second stage hydrocracking process. The modeling methodologies reported in the literature for the hydrocracking process thus far, describe the feed and product compositions based on the boiling range and the actual reaction network is reduced to a smaller number of reactions between the lumped species. The present approach applies the concept of single event kinetics to the hydrocracking process. In this approach, the various reactions involved in hydrocracking are considered in terms of fiindamental elementary steps involving carbocations. A computer algorithm, in which feed and product molecules, carbocations, and olefinic intermediates are represented by means of Boolean relation matrices and characterization vectors has been developed to generate the elementary reaction networks for paraffinic, naphthenic, and aromatic feed components. The standardized labeling algorithms for acyclic and cyclic hydrocarbon stmctures are developed. The network generation leads to a very large network of elementary reactions (>10^). However, due to the molecular nature of the approach, the number of rate parameters is kept within the tractable limits (<30) and the rate parameters are independent of the feedstock composition. Since the number of chemical species generated in the reaction network is very large, a certain degree of lumping is required to reduce the number of continuity equations for the components to be integrated along the reactor. The lumps should be chosen in terms of the present day analytical capabilities. The single event kinetic model, in the present work, considers the pure components and lumps according to the carbon number. Each lump is defined by its carbon number and the type of chemical structure that represents that lump. The type of chemical structures considered here are n-paraffins, iso-paraffins, mono-, di-, tri-, and tetra-naphthenes, mono-, di-, tri-, and tetra-aromatics, and naphtheno-mono-, naphtheno-di-, and naphtheno-tri-aromatics. Some lumps are individual molecules while most are collection of molecules. For the lump involving a collection of molecules, the properties of the lump are determined by averaging of the properties of each individual molecule comprising the lump. The model parameters are estimated from the synthetic product distribution data obtained from an industrial organization. A partially hydrogenated vacuum gas oil (VGO) is considered as the feedstock. The single event kinetic model is inserted into a homogeneous reactor model and the resulting continuity equations are integrated numerically along the length of the catalyst bed. The reactor simulation results are the temperature profile, composition profiles, and hydrogen consumption profile through the catalyst bed. The hydrogen consumption is calculated in a very rigorous way in the single event model, which is not possible with the lumped models. The reactor simulation results are consistent with industrial practice and published information. A profit optimization study is carried out to evaluate the aspects of the single event approach for process optimization. The molecular nature of the single event approach provides a framework to calculate important properties such as Reid vapor pressure (RVP) and octane number that are difficult to estimate using the lumped models.application/pdfengPetroleum -- RefiningCracking processPetroleum chemicalsDissertation