Professional Documents
Culture Documents
Advanced FCC Troubleshooting Via Ecat Data Interpretation
Advanced FCC Troubleshooting Via Ecat Data Interpretation
RAY FLETCHER
The FCC Analyst, LLC
T
he fluidised catalytic crack-
ing unit is a highly complex 29500
multivariable operation. The 27500
conversion of feed to petrochemi- 25500
cal feedstocks and/or motor fuels is
Total metals, ppm
23500
primarily a function of feed quality, 21500
independent operating parameters, 19500
and catalyst composition. Perhaps 17500
being deposited on the surface of the cessed in the unit (dark red data
Approximately 75% catalyst. points). Figure 5 indicates the nor-
of all yield selectivity The Ecat Ni-to-V ratio is a sim- mal response of conversion to the
ple method to monitor subtle Ni-to-V ratio. The lighter feedstock
shifts are the result changes in feed quality and corre- clearly does not follow the same
lates well with many yield selectiv- response curve as observed with the
of feed quality ities such as conversion, gasoline, refiner’s more typical feeds.
and slurry. Please note that each unit
variation for most Figure 4 demonstrates that the responds differently to changes in
typical Ni-to-V ratio for Refiner #2 the Ni-to-V ratio. It is essential that
units not involved in is from 1.0-1.4. However, begin- the process engineer determines the
ning at approximately day 95, a response, if any, to the typical varia-
a catalyst change-out lower density feed slate was pro- tion in their unit.
1.90
#3 to normal variations in the equi-
librium catalyst Z-to-M ratio. It is
1.80 important to note the degree of gas-
1.70
oline yield variation in this period
of normal operations. Analyses such
1.60 as these enable the FCC operator to
0 10 20 30 40 50 60 70 80
Days
understand the root cause in ‘typi-
cal’ yield variation or scatter.
64
lyst additions but is correlated to
62
an anticipated long-term increase
60
in regenerator temperature or equi-
58 librium catalyst metals is encour-
56 aged to discuss potential catalyst
54 reformulation with their catalyst
53 supplier.
1.75 1.80 1.85 1.90 1.95 2.00 2.05 As the regenerator increases
Equilibrium catalyst Z-to-M
in temperature, the hydrother-
mal deactivation of the catalyst
Figure 7 Refinery #3 gasoline yield vs Z-to-M increases. The zeolite portion of the
catalyst is much more sensitive to
Fresh catalyst composition light olefin yield, it is recommended regenerator temperature than the
Fresh catalyst composition is critical to plot propylene yield or propyl- active alumina, resulting in dealu-
to optimal conversion with desired ene olefinicity vs Ecat sodium. For mination of the zeolite crystal struc-
yield selectivities. However, the additional yields, the engineer is ture. This loss in activity leads to
fresh catalyst certificate of analysis recommended to plot the desired increased slurry yield. The process
(COA) often contains limited data, yield vs the Ecat zeolite-to-matrix engineer is encouraged to evaluate
making detailed catalyst quality (Z-to-M) ratio. As with all analy- the effect of regenerator temper-
analysis difficult. It is recommended ses, it is important to verify each ature effects on Z-to-M and yield
that the process engineer correlates unit’s response to these variable selectivity on their unit.
key Ecat properties and ratios vs shifts. Not all units respond alike.
important yield selectivities. It is also suggested that the refiner Catalyst attrition resistance
For refiners targeting maximum negotiate tighter upper and lower The refiner is recommended to
regularly monitor and record cat-
alyst fines for surface area and
10.0
metals content. An increase in Ecat
9.0
fines surface area combined with a
8.0
decrease in total metals is an indi-
7.0
cation of softer catalyst. An increase
in catalyst fines production not
Capture, wt%
6.0
5.0 demonstrating increased surface
4.0 area or decreased metals is an indi-
3.0 cator of a new attrition source in the
2.0 unit. Please note that reports of soft
1.0 catalyst by any catalyst manufac-
0.0
turer is very rare.
0 10 20 30 40 50 60 70 80 90
Particle size, µ Cyclone mechanical integrity
It is strongly recommended that
Figure 8 Normal catalyst fines particle size distribution regular catalyst fines samples be
Capture, wt%
The suggested PSD is a one-micron
scale for 0-10 µ particles (1, 2, 3 … 6.0
10), every two microns for 12-40 µ
and every five microns for particles 4.0
10.0
ing between approximately 10-25
8.0
µ is an indication of a crack or hole
in a primary cyclone or plenum. A 6.0
bimodal distribution with an abnor- 4.0
mal secondary peak appearing at a
PSD greater than the ‘normal’ peak 2.0