Regeneration - Spent Hydroprocessing Catalysts - page 147

operating form of the spent NiMo/Al 2 O 3 catalyst containing about 16 wt.% of coke at 350 ◦ C
using 2% O 2 . In the fresh form, this catalyst had the shape of spheres of about 1mm radius.
Therefore, for such catalyst, the shape of curve at 350 ◦ C would approach that for the crushed
form of spent catalyst in Fig. 6.11 at 500 ◦ C. This indicates the importance of spent catalyst
pretreatment (e.g., de-oiling) and its particle size prior to the oxidative burn-off.

The diffusion controlled burn-off is associated with a continuous change in pore volume and
size distribution. These parameters were incorporated in the model proposed by Chang [399]
to describe the burn-off of spent hydroprocessing catalysts in the diffusion controlled region.
According to this model, a very low burn-off rate at the later stages was attributed to the loss of
reactive surface area due to the emergence of islands of catalyst surface not covered with coke.
The restoring pore network of the fresh catalyst would be an ideal case of oxidative
regeneration. This is however not achievable even in the case of a complete removal of carbon.
Almost certainly, other factors such as catalyst recrystallization occurring during the operation
and subsequent regeneration could have an effect on the pore structure of the regenerated
catalysts. Also, it is unlikely that all coke is removed during the regeneration.

6.2.3 Modeling of Oxidative Regeneration

The primary objective of the models is to predict temperature profiles during regeneration.
This information can be used for fine-tuning of the conditions applied during regeneration. For
this purpose, the rate of reactions, which generate heat, has to be estimated. This includes the
oxidation of carbon, hydrogen, and organic sulfur. In addition, the heat generated during the
oxidation of metal sulfides has to be considered. Therefore, kinetic measurements form a basis
for model development. A large volume of information is available on modeling of
regeneration of FCC catalysts. However, compared with spent hydroprocessing catalysts, the
prediction of temperature profiles is much simpler because carbon is far predominant
component of coke. Also, the amount of coke in spent FCC catalyst is much smaller compared
with spent hydroprocessing catalysts. Furthermore, metal sulfides in the former FCC catalysts
are present in very small quantities. Then, incorporating only carbon burn-off gives reasonable
information for regeneration of spent FCC catalysts.

The model for predicting ignition temperature of the spent catalyst particles during
regeneration was developed by Klusacek et al. [400] . In this case, ignition temperature was
defined as a sudden temperature rise compared with the temperature of oxidizing gas entering
regeneration system. Both combustion of carbon and sulfur were considered in the model. The
investigated catalyst was the spent CoMo/Al 2 O 3 used during hydroprocessing a conventional
distillate. This ensured the absence of V and Ni as well as an even radial distribution of coke in
catalyst particles. For this case, the following heat balance equation was derived:

− C ps (d T s /d t ) = r c q c + r s q s + 3 h ( T s − T g )/( d p a )