In catalytic cracking, catalysts selection is not a trivial task since a catalyst should exhibit a good hydrothermal stability as well as product distributions and qualities according to the necessities of the refinery. Cracking experiments at the bench scale in a fluidized confined-bed reactor at 789 K in the catalyst to oil range of 3.0-7.5 g/g using two industrial feeds have coupled to a five-lump kinetic model for a quantitative study of a set of cracking catalysts. To a gain insight into the catalysts performance, the assessment of raw rate coefficients has been complemented with a reaction pathways analysis. Primary cracking reactions are 1 order of magnitude higher than secondary ones, the gasoil to gasoline conversion being the fastest reaction. Up to 23% of the formed gasoline may undergo secondary cracking, more than 90% of it ending up in LPG. At the investigated conditions, coke formation is practically fully formed out of gasoline via secondary reactions and not from the gasoil. Cracking results are in concordance with catalyst properties, i.e., specific surface areas, and catalyst composition, e.g., the rate of secondary cracking of gasoline to LPG decreases with the zeolite and rare earths content.

No abstract

No abstract

Besides offering an adequate product distributions, a cracking catalyst must exhibit a reasonable resistance to the combined effect of high temperature, steam and metals. Two commercial catalysts, poisoned with Ni ? V (*2,500 ppm), are subjected to a large scale cyclic deactivation with steam, then characterized using mostly basic techniques and finally tested via bench scale and pilot scale cracking experiments. As examples of industrial deactivation, corresponding equilibrium catalysts (E-CATs) are investigated in parallel. Depending on their specific formulation, catalysts deactivate differently due to the exposition to high temperature steam and metals. The intrinsic deviations in the catalytic properties of cyclic deactivated catalysts (CD-CATs) related to E-CATs are further biases by the particular history of the latter and the particular reactor configuration and hydrodynamics, such deviations being magnified in the bench scale testing. Relative differences in activity and product yields (except for coke and hydrogen) of CD-CATs referred to E-CATs are below 10% at the pilot scale, and within 23% at the bench scale. Relative biases in hydrogen and coke yield of may be as high as 75 and 26% in the pilot plant and as much as 150 and 32% at the bench scale.