Catalyst Deactivation - Spent Hydroprocessing Catalysts - page 78

non-issue for distillate feeds because in these systems the oil phase is predominant. The
situation may be less clear for the DAO where the conditions of deasphalting and origin of the
feed may have a pronounced effect on the colloidal structure and stability of the feed.
It is evident that during hydroprocessing, the asphaltene entities in the same feed may exhibit a
wide range of reactivity. Thus, the most soluble part (the least polar) of asphaltenes may be the
most reactive, whereas the insoluble part is the least reactive. The latter part sometimes
referred to as carboids can be separated from the asphaltenes by solvent precipitation [1] . After
most of the reactive portion of asphaltenes was converted to lighter fractions, the remaining
carboids may physically deposit on catalyst surface and as such contribute to the coke
formation. Because the content of carboids in asphaltenes from different heavy feeds is
different, their coke-forming propensity will be different as well. Nevertheless, carboids, as the
least reactive component of the colloidal system, may be partly responsible for the increased
aromaticity of asphaltenes isolated from the products compared with that in the corresponding
feed, sometimes reported in the scientific literature.
The results on solubility of asphaltenes published by Matsushita et al. [247] complement the
mechanism proposed by Seki and Kumata [233,234] . The former authors introduced the
solubility index defined as the ratio of the H/C ratio of asphaltenes to that of the DAO obtained
from the same feed using the different solvent/feed ratios. The decreasing solubility index
would indicate the loss of alkyl chains in asphaltenes (increase in aromaticity), in agreement
with the observations made by Callejas et al. [181] . Thus, the paraffinic hydrogen in alkyl
chains is an important contributor to the total hydrogen. This would decrease the solubility of
asphaltenes in oil and enhance their deposition on the catalyst surface. Also, the precipitation
of asphaltenes from the products would be enhanced. The onset of the asphaltenes
precipitation can be established using the critical solubility parameters, which can be
determined by the flocculation onset titration method [248] . These results suggested that
fouling is much less dependent on the origin of catalyst than on the colloidal stability of the
feed. It was observed that coke deposition was slowing down with the increasing solubility
index, i.e., with the decreasing H/C ratio of resins (heptane solubles) [249] .
The studies of Mochida et al. [250,251] showed that the compatibility problem in the graded
hydroprocessing systems might be alleviated by optimizing operating parameters. Thus, an
extensive deposit formation in the one-stage system could be prevented using a two-stage
system. In the latter case, the first reactor was operated at 663 K under conditions favoring the
HYD and asphaltenes depolymerization. The high rate of asphaltenes conversion at relatively
short contact time was achieved in the second reactor, which was operated at 693 K. A large
pore NiMo/Al 2 O 3 catalyst was needed to achieve these results. These studies focused on the
sludge formation in products however, similar factors are involved during the deposit
formation on the catalyst surface.