Improving Accuracy in the Estimation of Kinetic Frequency Factors from Laboratory Data To Model Industrial Catalytic Cracking Risers

• Published in: Industrial & engineering chemistry research Vol. 50; no. 5; pp. 2736 - 2745
• Main Authors: Jiménez-García, Gladys, Quintana-Solórzano, Roberto, Maya-Yescas, Rafael
• Format: Journal Article
• Language: English
• Published: American Chemical Society 02.03.2011
• Subjects: Kinetics, Catalysis, and Reaction Engineering
• ISSN: 0888-5885; 1520-5045
• DOI: 10.1021/ie1012773

Fluidized-bed catalytic cracking of heavy petroleum distillates is one of the most important processes in petroleum refining; because of its large scale, estimation of the kinetic rate constants is performed at the laboratory level. Several experimental devices are commonly used, including fixed-bed reactors, fluidized-bed reactors, and riser emulator reactors. Despite the different hydrodynamics of the reactors mentioned, when compared to the industrial unit, the production of the main products (gasoline, gases, and coke) can be predicted and emulated accurately in the laboratory. Taking advantage of this fact, prediction of the kinetic rate constants and catalyst activity parameters has been performed at the laboratory level; however, frequency factors are larger than those necessary to simulate the industrial unit. This situation could arise because of the lack of fundamental modeling of transport phenomena and catalyst activity decay due to coke deposition; these phenomena have been modeled using side functions associated with “time-on-stream”, “coke-on-catalyst”, or “feedstock conversion”. The objective of this work is to improve accuracy during estimation of the frequency factors at the laboratory level, via the estimation of catalyst activity, based on the effectiveness factor; which has been included in the mathematical model of a fixed-bed laboratory bench-scale reactor. It is shown that following activity evolution along the experiment time helps to calculate frequency factors that are very similar to those needed to simulate industrial units; moreover, this model estimates different catalyst activity for each reactant and each reaction, and, therefore, it is selective. Frequency factors for a lumped kinetic scheme obtained at the laboratory level are used to simulate an industrial unit, showing good agreement during prediction of the product yields.