Hydrocarbon Processing - May 2022 - 64

Catalysts
TABLE 1. Qualitative comparisons between two experimental approaches, conventional pilot-scale vs. high-throughput system
Aspect
Conventional pilot-scale system
Feed and catalyst
shipment costs
Number of catalyst
loading schemes and
time requirement
Catalyst homogeneity
Feed storage
and control
Product analysis
Nitrogen slip control*
Recycle operation*
Testing fee
Higher costs from more feed and catalyst requirements, 150 l
(liters) of feed per one loading scheme (one test scenario)
Limited catalyst loading schemes, usually require
more time from limited reactors in parallel
Statistically less affected by non-homogeneity
due to a larger reactor volume
Has a dedicated feeding section for each reactor,
possible to process different feedstocks in parallel
Takes less time for the same sample amount:
an ideal choice when frequently varying process
conditions or analyzing detailed product properties
Directly measured from interstage sampling
Possible but may not be accessible by the majority of
refiners; primarily available in licensing/catalyst companies
Not necessarily more expensive
*Specifically for hydrocracking applications
Grading
Demetallization
Demetallization
Hydrotreating
Hydrotreating
Hydrocracking
Hydrocracking
Hydrocracking
activity) for a fixed reactor volume (see
an example of a typical loading scheme in
FIG. 2). As such, the refiner might miss out
on opportunities to process more difficult
feeds or increase the unit throughput from
unrealistic metal uptake requirements. Incorporating
past spent catalyst analyses
(for both metal and particle size distribution)
along with actual feed qualities will
help refiners develop realistic grading and
demetalization volume requirements.
Nevertheless, refiners should be careful
when balancing risks and opportunities.
Clearly communicate with catalyst
vendors about unit configurations and
constraints. The authors have seen a diesel
hydrotreating catalyst vendor that
once designed the catalyst loading scheme
without realizing there was no amine
scrubber in the recycle gas loop until the
catalysts were later put under operation.
FIG. 2. A simple loading scheme of a
six-bed hydrocracking reactor (not to scale),
where grading materials and demetalization
catalysts-highlighted in light blue and
green, respectively-protect the hydrocracking
reactor against high pressure drop and
rapid metal deactivation.
demetalization catalysts (lower activity),
which will, in turn, limit the volume of
the main hydrotreating catalysts (higher
64 MAY 2022 | HydrocarbonProcessing.com
Catalyst evaluation methods. While
some refiners still rely on vendor estimations/predictions,
independent catalyst
testing has become more popular as a tool
to reveal the actual catalyst performance.
This evaluation approach is particularly
crucial to a critical unit like a hydrocracker,
where a slight difference in product
yield can result in multi-million dollars of
profit/loss per year.
In contrast with a general notion,
comparing paper estimates/predictions
from different catalyst vendors is not an
apples-to-apples comparison; nonetheless,
this is prevalent among refiners due
to its simplicity. In fact, catalyst vendors
employ different design assumptions,
feed characterization techniques, kinetic
models and product property estimators,
such as basic to non-basic nitrogen ratio
or aromatics distribution. Consequently,
it is fundamentally incorrect to compare
estimates/predictions between catalyst
vendors. Sadly, many refiners are unaware
of this fact.
To make matters worse, some catalyst
vendors are more aggressive than others.
It is not uncommon to see catalyst vendors
distort results from their kinetic model to
make their proposal more attractive. The
authors worked with a diesel hydrotreating
catalyst supplier that proposed a closeto-nil
offgas yield without any logical explanation.
When this was challenged as
impossible, the supplier nonetheless confirmed
that it was a result of their kinetics
model. This is a perfect example of how
refiners should always be skeptical of data
presented in vendors' proposals and how
paper-based evaluation can be subjective.
Despite these shortfalls, it is still acceptable
to use paper-based evaluation for
less critical applications, such as a naphtha
hydrotreater, although the best practice
is to have your catalysts tested before
the actual reloading.
For refiners without an in-house catalyst
testing facility, several companies can
High-throughput system
Much lower, 20 l of feed per one loading scheme
(one test scenario)
Up to 16 parallel reactors: an attractive option when
catalyst loading schemes exceed four; usually require
less time as there are more available reactors in parallel
Practically no issue with good catalyst screening
Common feeding section, therefore the same
feedstock for all reactors in parallel (distributed
via the flow distribution device); distribution quality
affects mass balance errors
Takes more time for the same sample amount;
requires close monitoring and control to ensure
sample's uniformity
Indirectly measured via a mirror parallel pretreating
reactor1
Impractical, mainly from difficulties in flow control,
e.g., smaller recycle flowrate
Not necessarily cheaper
http://www.HydrocarbonProcessing.com
Hydrocarbon Processing - May 2022

Table of Contents for the Digital Edition of Hydrocarbon Processing - May 2022
