Catalytic Upgrading of Waste Oils to Advanced Biofuels – Deactivation and Kinetic Modelling Study
The demand for liquid hydrocarbons as transportation fuels is enormous and ever growing.
Advanced biofuels is one of the promising solutions to keep pace with the global transition
to cleaner energy by reducing greenhouse gas emissions from the transport sector. It is
possible to selectively remove oxygen from waste oils like tall oil, used cooking oil etc. via
a catalytic hydrodeoxygenation (HDO) process to produce advanced biofuels. These
biofuels have similar molecules as in the traditional fossil-based fuels and exhibit improved
performance. This thesis focuses on aspects of catalyst deactivation and kinetic modelling
of HDO reactions.
In the first study, the influence of iron (Fe) as a poison during HDO of a model compound
for renewable feeds (Oleic acid) over molybdenum based sulfided catalysts was
investigated. Fe is a potential contaminant in renewable feeds due to corrosion during
transportation and storage in iron vessels. A series of experiments with varying Fe-oleate
concentration in the feed over MoS2/Al2O3 and NiMoS/Al2O3 catalysts. There was a salient
drop in the activity of the catalysts. At higher Fe concentration, for the NiMoS catalyst, the
selectivity for the direct hydrodeoxygenation product (C18 hydrocarbons) increased.
However, it was opposite for the MoS2 catalyst. There was a decrease in the yield of direct
hydrodeoxygenation products and an increase in yield of decarbonated products. It was
proposed that Fe interacted with these two catalyst systems differently. Fe influenced the
critical step of creation of sulfur vacancies in a negative way which resulted in lower
activity. Microscopic analysis indicated that Fe was preferentially deposited close or
around the nickel promoted phase, which explained why the role of Ni as a promoter for
the decarbonation route was subdued for the NiMoS catalyst.
In the second study, the kinetics during HDO of stearic acid (SA) over a sulfided
NiMo/Al2O3 catalyst were explored to investigate the reaction scheme. Model compounds
like octadecanal (C18=O) and octadecanol (C18-OH) were employed to understand the
reaction steps and quantify the selectivity. A Langmuir–Hinshelwood-type kinetic model
was used to investigate the kinetics. The results from the proposed kinetic model were
found to be in good agreement with experimental results. In addition, the model could
effectively reproduce the observed experimental profiles of different intermediates like
C18=O and C18-OH and illustrate phenomena like inhibiting effects of the fatty acid.