Bitumen-derived light gas oil (LGO) was hydrotreated over commercial NiMo/Al 2 O 3 catalysts in a trickle bed reactor. Statistical design of experiments was used to develop response surface models for predicting percentage conversions of aromatics, sulfur, and nitrogen in the LGO feed from Athabasca oil sands. The statistical approach was also used to study the effect of process variables and their interaction on aromatic hydrogenation (AHYD), hydrodesulfurization (HDS), and hydrodenitrogenation (HDN) activities. The two-level interaction between temperature and pressure was determined to affect AHYD significantly, whereas the interaction between temperature and the liquid hourly space velocity (LHSV) was the most important parameter affecting both HDS and HDN activities. Optimal conditions for the conversion of aromatics were observed at a temperature of 379°C, a pressure of 11.0 MPa, and an LHSV of 0.6 h -1 . Under these conditions, a maximum conversion of 63% can be attained. The cetane index of the diesel fraction was affected by changes in the aromatic compounds, as well as by the temperature and pressure of hydrotreating. Product distribution and gasoline yield of the liquid products were also greatly influenced by the reaction temperature, with a slight impact from pressure and LHSV. The kinetics of AHYD was modeled using a singe-site mechanism form of the LangmuirHinshelwood rate of reaction, whereas HDS and HDN were best described by an irreversible pseudo-first-order power-law reaction. Results of the kinetic studies showed significant inhibition of hydrogenation by hydrogen sulfide (H 2 S) gas produced during the HDS process.

In this work, hydrogenation of aromatic compounds in light gas oil derived from Athabasca bitumen was carried out using a single‐ and two‐stage hydrotreating processes. Experiments were performed in a trickle‐bed reactor using two catalysts namely NiMo/Al2O3 and NiW/Al2O3. NiMo/Al2O3 was used in the first stage for nitrogen and sulphur containing heteroatoms removal whereas NiW/Al2O3 was used in the second stage for saturation of the aromatic rings in the hydrocarbon species. Temperature and liquid hourly space velocity (LHSV) were varied from 350‐390°C and 1.0‐1.5 h−1, respectively, while pressure was maintained constant at 11.0 MPa for all experiments. Results from single‐stage were compared with those from two‐stage process on the basis of reaction time. Kinetic analysis of the single‐stage hydrotreating process showed that HDA and HDS activities were retarded by the presence of hydrogen sulphide that is produced as a by‐product of the HDS process. However, with inter‐stage removal of hydrogen sulphide in the two‐stage process, significant improvement of the HDA and HDS activities were observed.

In this work, a systematic experimental and kinetic study of hydroprocessing of light gas oils (LGOs) such as vacuum LGO (VLGO), atmospheric LGO (ALGO), and hydrotreated LGO (HLGO) using NiW/Al 2 O 3 and commercial NiMo/Al 2 O 3 catalysts has been conducted. Experiments were performed by varying temperature from 340 to 390 °C, at a constant pressure and liquid hourly space velocity of 11.0 MPa and 0.6 h -1 , respectively. H 2 /feed ratio was maintained at 550 mL/ mL throughout the experiments. Appreciable hydrogenation of aromatics (AHYD) was achieved by the NiW/Al 2 O 3 catalyst at low temperatures and at high severities of hydrotreating. However, the hydrogenation activity of NiMo/Al 2 O 3 was superior to that of the NiW/Al 2 O 3 catalyst. For hydrodesulfurization (HDS) and hydrodenitrogenation (HDN) activities, higher conversions of 95-98.8 and 96-99 wt %, respectively, were attained for the commercial NiMo/Al 2 O 3 catalyst throughout the temperature range studied. Simulated distillation of the feed showed that VLGO contained the most complex and heaviest compounds followed by HLGO and ALGO. Diesel selectivity in both ALGO and HLGO increased with hydrotreating temperature, but in the case of VLGO, it decreased with temperature. Kinetics studies showed that dearomatization of the HLGO feed was the most difficult, followed by ALGO and then the VLGO. Kinetics of ALGO and VLGO were best described by a pseudo-first-order reaction mechanism while the 1.3 power law kinetics worked well with HLGO.