HP March 2022 eBook—Energy Transition - 69

Catalysts
100
80
90
60
50
70
40
30
20
10
11/5/17
Other SOx
2/13/18
additive
5/24/18
9/1/18
12/10/18
Proprietary SOx additiveb
3/20/19
Proprietary SOx
6/28/19
additivea
Uncontrolled SOx
100
80
90
60
50
70
40
30
20
10
11/5/17
Other SOx
2/13/18
additive
5/24/18
FIG. 3. SOX emissions level (normalized to highest uncontrolled SOX
80
90
100
60
70
50
40
20
30
10
11/5/17
Other SOx
2/13/18
additive
5/24/18
9/1/18
12/10/18
Proprietary SOx additiveb
3/20/19
Proprietary SOx
6/28/19
additivea
FIGS. 4 AND 5. PUF plotted against time and % SOX
SOx
additive performance at BP
additive on an ongoing
emissions limadditive
,
later
80
90
100
60
70
50
40
20
30
10
0%
Other SOx
20%
additive
reduction.
Rotterdam. The BP Rotterdam refinery
uses an SOx
basis to comply with SOx
its. In March 2018, the refinery elected
to change from a general SOx
to the proprietary SOx additiveb
changing to the newly developed additivea
how
these SOx
ed SOx
. The following case study describes
additive changes impactemissions
control.
The BP Rotterdam FCCU is a challenging
operation in terms of SOx
additive
performance. Specifically, the low
oxygen level in the regenerator is likely
to be a limiting factor for SOx
additive
performance due to the oxidation of SO2
to SO3
being a bottleneck in the SOx
reduction
mechanism. This is reflected in
the relatively low pick-up factor (PUF)
values observed. It is also worth noting
that the Rotterdam FCCU's highly variable
feed properties can make the evaluation
of SOx
additive performance more
challenging. For this reason, a multivariable
regression analysis was performed.
This statistical analysis seeks
to establish the relationship between
dependent and independent variables.
Developing a robust multivariable regression
allows the evaluation of additive
performance despite many changes
occurring simultaneously.
age rate from November 2017 through
the SOx
FIG. 2 shows the daily SOx
additivea
additive ustrial.
Using a sevenadditive
rate, as it
additiveb
day (7-d) rolling average is a useful way
to look at the SOx
helps to smooth out large differences in
the daily addition rate, enabling a more
meaningful comparison of PUFs. The
average usage rate of the SOx
was slightly higher than the previous
additive due to pre-blending with fresh
FCC catalyst, which coincided with a
period of higher catalyst addition rate.
The average addition rate for the latest
SOx
additivea
the SOx
general additive period.
The uncontrolled levels of SOx
level that would be obtained
without using an additive) are calculated
using a correlation based on the
wt% sulfur in the net slurry product.
FIG. 3 shows the uncontrolled SOx
vs. actual SOx
tained at low levels during the use of the
proprietary SOx
additivesa,b
additivea
levels
levels observed during the
different periods (the values are normalized
to the highest uncontrolled SOx
level). The actual SOx
levels were main.
The overall
SOx reduction increased from 67% to
70% with the use of one SOx additiveb
.
,
then increased further to 73% using the
newest SOx
As a performance indicator when as40%
SOx
60%
reduction
Proprietary
SOx additiveb
Proprietary SOx
additivea
80%
100%
9/1/18
12/10/18
Proprietary SOx additiveb
).
3/20/19
Proprietary SOx
6/28/19
additivea
Actual SOx
80
100
60
40
20
11/5/17
2/13/18
5/24/18
Competitor
Regression
9/1/18
12/10/18
3/20/19
Proprietary SOx additiveb
Proprietary SOx
additivea
FIG. 6. Comparison of actual SOX
levels vs.
calculated SOX from regression analysis.
sessing SOx additive performance, the
PUF is a calculation of mass of SOx
captured
per mass of SOx
is a measure of SOx
additive used, and
additive effectiveness.
A PUF value for a given application
can fluctuate depending on various
factors. For example, the PUF would be
expected to decrease when achieving a
higher percentage of SOx
the concentration of available SOx
reduction as
molecules
becomes lower.
Therefore, with the higher percentage
SOx reduction achieved with the proprietary
SOx additivesa,b
, a lower PUF might
have been expected. However, FIGS. 4
and 5 show that the expected drop in
PUF was not observed. In fact, for a given
percentage of SOx
reduction, the additives
even resulted in a slightly higher
PUF, which is testament to the SOx
reduction
performance being achieved.
As previously mentioned, multi-variable
regression analysis can be useful for
evaluating SOx
was 15% lower than the
(i.e.,
additive performance, as
it considers changes that are occurring in
the operation.
Such an analysis was performed to
establish a correlation for calculating the
expected SOx
emissions during the period
of using a general additive. The analysis
identified three key variables or predictors
for SOx
sulfur (wt%), SOx
and excess oxygen (vol%). The formula
established below results in a high R2
value
of 90%, indicating that this regression
analysis can accurately explain the variation
of SOx
predictors. The definition of R2
) is explained
emisby
the three key variables.
These three key variables are wellknown
in terms of impact on SOx
sions. Slurry sulfur levels typically correlate
well with coke sulfur levels, which is
Hydrocarbon Processing | SEPTEMBER 2021 | HydrocarbonProcessing.com
emissions using the specified
being
the percentage of the response variable
variation (calculated SOx
emissions (ppm): slurry
additive rate (kg/d)
6/28/19
PUF, normalized
PUF, normalized
Normalized SOx
http://www.HydrocarbonProcessing.com
HP March 2022 eBook—Energy Transition

Table of Contents for the Digital Edition of HP March 2022 eBook—Energy Transition
