ISSN:
1542-6580
Source:
Berkeley Electronic Press Academic Journals
Topics:
Process Engineering, Biotechnology, Nutrition Technology
Notes:
A two-stage, micro trickle-bed reactor (for studies of the effects of hydrogen sulfide on hydrodesulfurization (HDS) and hydrodenitrogenation (HDN) of Athabasca bitumen-derived heavy gas oil over commercial NiMo/Al2O3 catalyst) has been simulated. One dimensional homogeneous mass transfer and a two dimensional heat transfer models were developed. The essence of the simulation was to enhance the understanding of the effects of hydrogen sulfide in the hydrotreating catalyst bed in a two-stage mode and also to predict the catalyst requirements for deep HDS and HDN processes. The kinetic model used in the simulation was based on the Langmuir-Hingshelwood method of rate determination. Adsorption constants were estimated by non-linear least squares method. The kinetic models were tested on independent set of data and found to predict the experimental data satisfactorily. The mass transfer simulation considered the effects of variables such as temperature and catalyst loading or liquid hourly space velocity (LHSV) on the trends of hydrogen sulfide generation and, sulfur and nitrogen conversions along the catalyst bed. The model was numerically solved using a fourth-order Runge-Kutta technique. The 1:3 wt/wt catalyst loading with inter-stage hydrogen sulfide removal was found to give the best HDN and HDS activities. Simulated results showed that doubling the present catalyst mass and operating at 653 °C with inter-stage hydrogen sulfide removal would give 6 and 179 ppm product sulfur and nitrogen, respectively. On the other hand, without hydrogen sulfide removal, only 49 and 302 ppm product sulfur and nitrogen could be attained, respectively. The heat transfer simulation compared temperature profiles in the two-stage process to a single stage process for the 1:3 wt/wt catalyst loading at 653 K. The temperature regime in Stage II was found to be more uniform unlike Stage I and the single stage. Crank Nicholson algorithm was used to solve the 2-D partial differential equations.
Type of Medium:
Electronic Resource
URL:
http://www.bepress.com/ijcre/vol4/A20
