Catalyst Deactivation - Spent Hydroprocessing Catalysts - page 85

power form or Langmuir-Hinshelwood form. Reaction order is usually chosen to obtain the
best fit of the experimental results for a particular kinetic model. The model developed by
Long et al. [259,260] on a catalyst activity level showed very good fit of the experimental
results with those predicted by the model. However, such a good fit could only be obtained by
assuming that during very early stages, part of the V was deposited on the uncovered support,
thus having no adverse effect on active metals' phase. Therefore, this model does not assume a
uniform metal layer deposition contrary to some other models [267-269] . Moreover, metal
deposits (e.g., V 3 S 4 ) may exhibit some activity for hydroprocessing reactions. The
autocatalytic effect of deposits is usually overlooked during the development of models. The
autocatalysis may be, at least, partly responsible for deviation of the predicted results from
those observed experimentally. This may be evident particularly during the early stages of the
operation. The model tested by Melkote and Jensen [270] was among few in which the effect
of autocatalysis was considered.
A detailed account of the catalyst deactivation by metals was given by Tamm et al. [197] who
used five residues, metal content (V+Ni) of which varied from about 40 to almost 500 ppm.
Their model confirmed that the metal deposition patterns were feedstock dependent and
poisoning of active sites by metals and physical obstruction of pores by metals were
contributors to catalyst deactivation.
Surface area, pore volume and pore size distribution, size and shape of catalyst particles are of
the primary interest for designing catalysts for hydroprocessing of the metals and asphaltenes
containing feeds. For this purpose, parameters such as the effective diffusivity, efficiency
factor, distribution parameter, Thiele modulus, metal storage capacity, etc. are included in the
models. In addition, the development of models on the particle scale would be incomplete
without incorporating data on catalyst activity. This indicates the need of kinetic data and
catalyst deactivation pattern. Therefore, it may also be appropriate to refer to this level of
model development as the two-scale approach, i.e., an active phase and a single particle scales.
Thus, the applicability of the models on particles scale would be somehow limited without
including the effects of active phase on the catalyst performance.
The usefulness of the particle scale models for designing the catalysts for hydroprocessing of
heavy feeds was demonstrated in the study of Oyekunle et al. [271,272] . These authors
performed calculations for the three types of catalysts, i.e., microporous and macroporous with
the predominant portion of pores having APD < 100 D and APD= 100-250 A, respectively, as
well as the random pore distribution with the predominant APD between the microporous and
macroporous catalysts. They used the data on hydroprocessing of the heavy Maya crude
published by Ancheyta et al. [273,274] . Figure 4.24 showed a good fit of the process data with
those predicted by the models. The catalyst lifetime was then estimated by using the linear
regression analysis of the results in Fig. 4.24 . The total activity loss was predicted to occur after
462, 316 and 150 days for the macropore, random pore and micropore systems, respectively.