Upgrade FCC performance

New ceramic feed distributor offers ultimate erosion protection

L. M. WOLSCHLAG and K. A. COUCH, UOP LLC, a Honeywell Company, Des Plaines, Illinois
F
luid catalytic cracking (FCC) technology has been a part of
the petroleum industry since the 1940s. Despite being a very
mature technology, continued development is vital, especially
as many refiners move their FCC operations from fuels produc-
tion to higher-value products. Advanced diagnostic and design
tools are accelerating process developments.
Through the development and commercialization of world-
scale FCC units, technical discoveries have emerged that provide
opportunities for improvements across all units, independent of
size. Using sophisticated engineering tools, such as computa-
tional fluid dynamic (CFD) modeling combined with radioac-
tive tracer and tomography, will streamline physical inspection
reports and commercial yield analysis. The article highlights
advancements in regenerator technology for higher capacity
through existing assets, emissions reduction and feed distribution
systems for large-diameter risers.
Dual-radius feed distributors. As refiners look to capital-
ize on economies of scale, design throughputs of FCC units have
reached record levels. At these scales, opportunities have emerged
from the background noise of the data to improve FCC technol-
ogy. Through pushing multiple constraints to design limits on one
particular unit, yields and conversion deviated from benchmark
performance, with gasoline selectivity lower, conversion lower and
dry gas higher than benchmark performance. To get more out of
the existing asset, an intensive program was undertaken to achieve
benchmark performance.
The riser for a particular FCC unit
has an inner diameter (ID) of 6.6 ft at the
point of feed injection, which expands to 9
ft immediately above. The feed is injected
into the riser through a set of circumferen-
tially positioned distributors. The combi-
nation of low conversion and high dry gas
yield seems counter-intuitive, given tradi-
tional FCC operations. A hypothesis was
raised that the large riser diameter might
be preventing the feed from adequately
distributing across the full cross-sectional
area of the riser. To test this hypothesis,
a CFD model of the riser was created to
analyze the fluid dynamics of the system.
Results of the model supported that raw
oil feed would only penetrate the riser a
finite distance, thus creating a vapor annulus, and that much of
the catalyst flowing up the riser would form a high-density core.
Based on CFD results, a tomographic analysis (gamma scan) of
the riser was completed. The scan results confirmed the CFD
model prediction as illustrated in Fig. 1.
Radioactive tracer work was also completed on the 9-ft ID
riser. Irradiated Krypton-79 gas was injected into the riser base.
Detectors were positioned along the riser length and reactor to
measure the tracer as it moved through the system. The results
indicated that the time of flight of the krypton gas from one
detector to another did not provide a sharp response peak. An
early peak followed by a secondary peak which was skewed a high
degree is shown in Fig. 2.
A mathematical evaluation was performed to determine what
type of continuous stir tank reactor (CSTR) response would be
needed to emulate the measured data. To accurately reproduce the
field data plot, a composite plot modeled 100, 40 and 15 CSTR
responses (Fig. 2).
Unit performance, CFD modeling, tracer and tomography
tests, and mathematical analysis all indicated the same pathol-
ogythe feed was not adequately accessing the full cross-sec-
tional area of the riser leading to the presence of a high-density
core of catalyst and a low-density annulus, which caused low
conversion and high dry gas and coke make. One solution to
this problem would be to install two, smaller diameter risers to
match more conventional FCC sizes. However, installing dual
CFD prediction and gamma scan of 6.6-ft riser. FIG. 1
REFINING DEVELOPMENTS SPECIALREPORT
Originally appeared in:
September 2010, pgs 57-65.
Used with permission.
HYDROCARBON PROCESSING SEPTEMBER 2010
REFINING DEVELOPMENTS SPECIALREPORT
risers, even with new construction, is substantially more expen-
sive. For an FCC unit of 200,000 barrels per stream day (bpsd),
the estimated cost difference between a single, large-radius riser
and a pair of smaller risers has a cost estimated at $60 million.
A substantially lower cost solution with an implementation of
dual-radius feed distributors was developed (Fig. 3). This design
ensures optimal feed distribution across the entire riser, while
avoiding adjacent spray impact that could cause undesirable
spray interference.
Another CFD model that incorporates the dual-radius feed
distributors was created. Fig. 4 shows catalyst density profiles of
an axial slice of the riser, both with and without dual-radius feed
distributors. The riser on the left side without the dual radius feed
distributors shows the high-density core of the catalyst; the CFD
model with the dual radius feed distributors indicates that the
catalysts dense core is effectively eliminated.
The dual-radius feed distributors were installed on a FCC unit
designed with an 8-ft-diameter riser at the point of feed injection.
The unit was commissioned in May 2009. Results indicate that dry
gas yield and conversion and gasoline selectivity were within expec-
tations. The risers gamma scans indicate that the catalysts high
density core was effectively eliminated. The catalyst density profile
of the riser at approximately 1 pipe diameter above the point of
dual radius feed injection, indicates that core annular flow has been
achieved with an evenly distributed catalyst density profile (Fig. 5).
Additional tomography scans were completed at varying feed ratios
to optimize distribution of oil and steam across the riser.
Erosion of the inner feed distributors was a client concern.
This was mitigated by using ceramic feed distributors. Ceramic
offers the ultimate in erosion protection, and feed distributors
with ceramic tips can withstand highly erosive environments with
zero discernable erosion.
CERAMIC FEED DISTRIBUTORS
Development. FCC feed distributor tips are subjected to a
high-temperature, high-velocity erosive environment. To func-
tion in this harsh environment, FCC feed distributors have his-
torically been fabricated from various erosion-resistant materials.
While these materials are proven effective at reducing rates of
erosion, most erosion-resistant materials are, by their nature,
generally hard and brittle and can be susceptible to brittle frac-
ture. Erosion and brittle fracture have been an industry-wide
issue, and can be induced mechanically or by thermal shock.
This must be considered in the design of FCC feed distributors
as erosion and brittle fracture can occur when relatively cold oil
and/or steam are rapidly introduced to the system in which the
tips are hot from circulating catalyst.
These issues were addressed in many ways with a distribution
system. Following proper operating procedures will avoid thermal
shock and brittle fracture. However, erosion is more a function of
operating environment as opposed to improper operation.
Designs. Advanced design feed distributors include three pri-
mary designs: standard, weld overlay and ceramic. The standard
designthe new distributor for most FCC applicationsbal-
ances the erosion issue and the possibility of cracking due to ther-
mal shock. The tip incorporates a more erosion-resistant metal
alloy, changing the geometry and reducing stress concentrations.
Incorporating orifice extensions extends the flashing hydrocarbon
feed further away from the metal tip. Additional protection can
be provided by applying a very hard diffusion coating over the
cobalt-based (Co-based) alloy.
0
1,000
800
600
400
200
0
1 2 3
Seconds
March 14, 2007, gas injection
R
e
s
p
o
n
s
e
4 5 6 7
Riser ex top avg
40 CSTR response
Composite
100 CSTR response
15 CSTR response
Early peak
feed plug
ow core?
Centroid 77.18s
t-res = 1.32s
velocity 9.3 m/s
Late peak
wall holdup?
Mathematical composite. FIG. 2
Schematic of dual-radius feed distributors. FIG. 3
CFD models of the riser catalyst
profiles with and without dual-
radius feed distributors.
FIG. 4
HYDROCARBON PROCESSING SEPTEMBER 2010
REFINING DEVELOPMENTS SPECIALREPORT
The weld overlay design is applied to resolve chronic prob-
lems with wet steam and installations that have a high risk of
thermal shock. The erosion-resistant weld overlay is applied to a
softer, more ductile base metal for superior thermal shock resis-
tance. To further combat erosion, this tip incorporates orifice
extensions to move the flashing hydrocarbon feed further away
from the distributor tip. While a very hard diffusion coating is
used to provide additional protection against erosion, the pri-
mary design goal is resistance to thermal shock, and, therefore,
it is recommended only for FCC operations that have proven to
be particularly susceptible to thermal shock.
Finally, the ceramic design represents a step-change improve-
ment for superior erosion resistance. Determining the erosion
potential of FCC feed distributors is based on the physical prop-
erties of the feedstock. The ceramic design is used in applications
where erosion is forecast to be higher than normal or in units that
have previously exhibited high erosion rates. Even though the
ceramic material is very hard, quench testing in the laboratory
and commercial application have indicated that new ceramic tips
are no more susceptible to thermal shock than traditional fabrica-
tions with co-based alloys. Fig. 6 shows three new tip designs, as
well as older versions.
Ceramic tipsdesign challenges. Ceramic materials are
widely accepted and proven to be more resistant to erosion than
metallic materials. The characteristics that impart erosion resis-
tance also tend to make these materials more brittle. Successful
application of ceramics in FCC feed injection required that two
technical challenges be overcome: 1) selecting a suitable ceramic
material that can be fabricated into the required geometry and
2) developing a means to connect the ceramic tip to the metallic
base assembly of the distributor.
The geometry used for the ceramic distributor tip was the
same as the traditional elliptical-feed distributor. The same prin-
ciples and considerations applied to reducing mechanical stresses
and improving thermal shock resistance in metallic tips were
applied to address the brittle nature of ceramics. The ceramic
tips were subjected to laboratory quench testing to simulate the
unique temperature profiles in the feed-injection system. Quench
testing was used to help select the proper ceramic material, and it
confirmed that the final material was no more susceptible to brit-
tle fracture than previous FCC feed distributor metallic tips.
The large differences in thermal expansion coefficients
between the materials provided the next challengea means
of attaching the ceramic tip to the metallic base assembly.
The attachment should provide a liquid-tight seal at design
pressure drop across the distributor, while accommodating
a wide range of feed and steam temperatures experienced
across startup, normal operation and FCC unit shutdown.
Creative engineering, stress modeling, full-scale prototyping
and therma-cycle testing were all used to develop a proprietary
mechanical connection. With an acceptable ceramic identi-
fied and a means of connecting the ceramic to the metal base
assembly, the next step was to demonstrate new distributors
in a commercial application.
Ceramic tipsa commercial experience. As the design
details for the new ceramic tip and connection were finalized,
an opportunity presented itself in which two ceramic tips could
Gamma scan of 8-ft ID riser with dual-radius feed
distributors.
FIG. 5
FCC feed distributor tip designs. FIG. 6 Metallic and ceramic feed distributor tips after 18 months
in operation.
FIG. 7
HYDROCARBON PROCESSING SEPTEMBER 2010
REFINING DEVELOPMENTS SPECIALREPORT
be installed in the same reactor riser at the same time as metal-
lic tips, providing an ideal side-by-side commercial test.
1
The
subject FCC had a history of aggressive feed distributor tip ero-
sion, and a trial installation of the ceramic tips was welcomed.
Final design details regarding tip connection were addressed. In
April 2007, ceramic tips were commissioned in a commercial
FCC reactor riser.
After 18 months of operation, the ceramic tips were inspected
and were free of erosion and cracking, while the adjacent metal-
lic tips exhibited signs of erosion. In Fig. 7, the metallic tip
shows significant erosion, while the ceramic tip shows zero
discernable erosion.
The viability and benefit of using ceramic tips for the feed
distributor were confirmed. The expected life of the distributors
in this application was revolutionized, from imminent failure
(with an average run life of 23 years), to potentially a life with
perpetual success.
Since January 2010, FCC ceramic feed distributors have been
delivered to three refiners in addition to the trial installation. The
second installation was placed into service on May 17, 2009, and
it continues to perform well with two additional project ship-
ments pending. Ceramic distributors are currently recommended
and supplied as the premiere offering to improve reliability in
installations with aggressive distributor tip erosion.
Elephant trunk arm combustor riser disengager.
The market drive to maximize returns through economies of
scale can present technical challenges with respect to scale-up. A
phenomenon occurred on a large combustor style FCC regenera-
tor in which flue-gas catalyst losses appeared to increase at the
higher end of superficial velocities that are typically stable for
smaller designs. In this case, the refiner was interested in achiev-
ing a higher capacity through an existing asset.
The inside of the upper regenerator has two major pieces of
equipment: cyclones and a combustor disengager. The combus-
tor disengager provides the first-stage inertial separation of cata-
lyst from the combustion products, and the cyclones provide the
final separation. Layout of this particular regenerator is unique
in that the cyclone pairs are configured on two different radii
(Fig. 8). While this has been a common plan view layout for
bubbling-bed regenerators, this was the first time it was applied
to a combustor-style unit.
To start the evaluation, a CFD model of the regenerator was
created to study the unit-specific gas flow paths in the upper
regenerator. The model demonstrated that the gas flow exiting
the standard tee disengaging arms was in the range of 49 m/s
(Fig. 11). This velocity range is between 50%100% higher at
a 15% lower superficial velocity compared to the next largest
combustor-style regenerator. The model also indicated that
the jet length projected from the disengaging arm was long
enough that the high-velocity gas stream moved horizontally
in the area of the dipleg termination. This resulted in fines re-
entrainment with preferential flow to the inner-radius cyclone
pair, at a rate that exceeded the catalyst discharge capacity of
the cyclones. This result was initially difficult to believe, as the
primary cyclone inlets on the two different radii were only 18
in. apart. However, the preferential flow was readily apparent
upon internal unit inspections at the turnaround six years
after commissioning. A slight change to the base design had a
profound impact on the equipment performance.
Solution. The solution developed was a variation on what
is called the elephant trunk disengager (Fig. 12). While basic
elephant trunk disengagers were used in FCC reactor riser
disengagers in the late 70s and early 80s, the regenerator appli-
cation required substantial engineering work to ensure that
the proper gas flow paths and catalyst separation efficiencies
Dual-radius cyclone system in regenerator. FIG. 8
Tee and elephant trunk arm disengage. FIG. 9 CFD model of the gas profiles in the upper regenerator
with tee and elephant trunk arm disengagers.
FIG. 10
HYDROCARBON PROCESSING SEPTEMBER 2010
REFINING DEVELOPMENTS SPECIALREPORT
were achieved. The disengager arm was curved to lower the
impact transition, reduce catalyst attrition and improve lin-
ing reliability. The shroud was extended to direct the catalyst
more into the catalyst bed, but was limited in length so as not
to provide excessive separation efficiency that would lead to
increased afterburn and high dilute phase temperatures. The
outlet area was optimized to ensure that the combustion gases
bleed off horizontally with minimal cross-wind at cyclone
dipleg terminations (Fig. 9).
The CFD model of the final design indicated that at a super-
ficial velocity of 1.05 m/s, slightly higher than the base case
model, the gas velocities exiting the arms of the elephant trunk
disengager were significantly lower than the gas velocities for the
tee disengager, with peak gas velocities reduced by 25% and the
horizontal gas velocities at the dipleg outlets reduced to nearly
zero (Fig. 10).
With the original design, 10 out of 11 inner cyclones holed
through after six years of operation. With the elephant trunk
disengage installation, the fines entrained to the inner-cyclone
set were reduced sufficiently to reasonably expect a 10-year ser-
vice life. This enables the refiner to either significantly reduce
maintenance costs and realize greater onstream reliability, or to
push the system harder for greater operating margin.
CFD model validation. CFD models have historically
met with substantial skepticism in mixed-phase fluidized bed
systems. To validate the CFD modeling efforts, multiple oper-
ating regenerators were modeled, and the results compared
with turnaround field inspection reports. The CFD modeling
has proven to be predictive with respect to erosion of both the
cyclones and the external support braces when compared with
field inspection reports.
To further evaluate the accuracy of the CFD modeling and
determine the proper boundary conditions for the models,
multiple radioactive tracer tests were completed on regenera-
tors with the tee disengager and elephant trunk disengager. The
downward gas flow predicted with the tee disengager was vali-
dated, and the residence time of the flue gas within the upper
regenerator was within 6% of the CFD model. Tracer studies
of the elephant trunk disengager confirmed a greater amount
of gas dispersion, eliminating regions of high gas velocity, and
effectively using regenerator volume.
The first commercial combustor riser elephant trunk disen-
gager was commissioned in 2009. Initial results have been very
promising. Catalyst containment is very good and continues
to be closely monitored. The flue gas residence time in the
upper regenerator increased by as much as 26%substan-
tially improving regenerator performance. The unit design
and operation resulted in extremely low delta coke operation
and a regenerator average dense-bed temperature as low as
1,198F. Even with this low regenerator temperature operating
at maximum throughput, the average afterburn is only 8F.
This is a step-change advancement in regenerator combustion
performance and it supports that the modeled increase in flue-
gas residence time was achieved.
2
The elephant trunk disengager was developed to improve
the performance of a very large combustor. CFD modeling,
tracer work, unit inspection and operational data collectively
contributed to its creation, proof of principle and commer-
cialization. However, by using these sophisticated tools, other
benefits were discovered that are applicable to all sized units.
Eliminating the high-velocity regions reduces erosion to inter-
nals and associated catalyst attrition. The increased residence
time improves the burning capacity of the regenerator, enables
lower excess oxygen operation and directionally reduces NO
x

emissions. Now, the elephant trunk arm disengager has become
the standard design for all new combustor-style regenerators,
with several revamp and new unit designs in progress.
SPENT CATALYST DISTRIBUTOR
Problem. Engineering tools and associated skills used to
solve the previously discussed problems for very large FCC
units can be used on FCC units of all sizes and types, to sup-
port operating and reliability needs of individual refiners. In
one example, an 80,000-bpsd FCC unit with a bubbling bed
regenerator exhibited a regenerator cyclone outlet temperature
differential of 100F from one side of the regenerator to the
other. This afterburn differential resulted in a localized hot
spot that limited the throughput of the unit against a main
air-blower constraint. The regenerator was an older design
that used a gull-wing spent-catalyst distributor design. Catalyst
maldistribution in the regenerator causes fuel-rich areas in
the dense phase, with localized hot spots directly above in the
dilute phase. Hot spots can be completely invisible within a
unit depending on where instrumentation is placed in relation
Cyclone 1
Cyclone 2
Cyclone 3
Cyclone 4
Cyclone 5
Cyclone 6
Cyclone 7
Cyclone 9
Cyclone 10 14%
12%
10%
8%
6%
4%
2%
0%
Catalyst tracer results for a bubbling bed regenerator with
a gull-wing design.
FIG. 11
Gull-wing and piped spent-catalyst distributor. FIG. 12
HYDROCARBON PROCESSING SEPTEMBER 2010
REFINING DEVELOPMENTS SPECIALREPORT
to the spent-catalyst inlet.
To validate the temperature data, catalyst tracer work was
completed on the regenerator to evaluate the flow distribution
in the unit. With ideal distribution, a radar plot of the detec-
tor signals would show perfect symmetry. The actual unit data
showed that the catalyst was heavily skewed to one side, which
was not a surprise (Fig. 11).
Solution. The typical spent catalyst distributor installed in
a bubbling-bed regenerator of this vintage was the gull wing
design with an external lift riser. Fig. 12 is a schematic of the
distributor. Air maldistribution in this type of regenerator
design results from two sources. First, the external riser lift
air discharges vertically out of the disengager, resulting in an
oxygen-rich environment in the dilute phase. Second, high
localized catalyst density and resultant hydraulic head caused
a preferential flow of combustion air to the opposite side of
the regenerator.
To achieve a more even catalyst density and uniform coke
distribution, the piped spent catalyst distributor was developed
(Fig. 12). The piped distributor was designed to radially dis-
tribute both the lift air and spent catalyst across the regenerator
bed through a set of side arms. The size and orientation of the
distributor arms were designed in an iterative process with CFD
modeling to ensure as much even catalyst and air distribution
as possible within the back-pressure limitations of the existing
lift air blower.
CFD models of the gull-wing distributor and the piped spent
catalyst distributor were created to predict the catalyst distribu-
tion, gas flow paths and bed-density profiles in the bubbling-bed
regenerator. With the gull wing distributor, the catalyst was
concentrated in the bed center. With the piped spent catalyst
distributor, the catalyst distribution was much more uniform
throughout the bed (Fig. 13).
Results. The piped spent-catalyst distributor was commis-
sioned in December 2006. Post-revamp tracer tests were con-
ducted on the regenerator. The actual catalyst distribution is
very close to the ideal distribution as illustrated in Fig. 14.
Operational data also indicate a significant improvement
in the regenerator performance. The dilute phase temperature
differential was reduced from 100F pre-revamp to about 15F
following the implementation of the piped spent-catalyst dis-
tributor. As a result, the refiner was able to lower the excess
oxygen level in the flue gas from a pre revamp minimum of 2
mol% to a post-revamp 1 mol%, enabling a higher capacity
through existing assets and saving on utility consumption. HP
A
new catalyst regenerator technology was developed to
improve coke operations. Since improved equipment
technology and catalyst offerings have resulted in progres-
sive decreases in coke, refiners have continued to debottleneck
unit constraints within the capacity of existing major equipment.
Usually with the constraint of not replacing main vessels or large
rotating equipment, i.e., main air blowers (MABs), wet gas com-
pressors (WGCs), regenerator shells, reactor shells, stripper shells
and catalyst circulation standpipes. Maintaining this philosophy
through several technology upgrades can result in several opera-
tional and reliability concerns:
Excessive catalyst loading to regenerator cyclones
Cyclone erosion, fines generation
High catalyst circulation and catalyst flux
Insufficient regenerator and stripper residence time
Hydraulic instability or limitation in the catalyst
standpipes
Excessive catalyst fines to the flue gas system
Increased particulate matter (PM) emissions.
Improvements in FCC technology to achieve lower coke
operations not only present the concerns previously listed, but
also cause refiners to face several dilemmas between wants
and often conflicting also wants, (Table 1). Optimum prod-
uct selectivities, conversion and throughput are traditionally
opposite to lower coke yield, optimum coke combustion, and
retaining existing equipment.
CFD model of the catalyst densities in a regenerator with
gull-wing and piped spent-catalyst distributors.
FIG. 13
14.00%
12.00%
8.00%
6.00%
4.00%
2.00%
0.00%
10.00%
Catalyst tracer results for bubbling-bed regenerator with
piped spent-catalyst distributor.
FIG. 14
HYDROCARBON PROCESSING SEPTEMBER 2010
REFINING DEVELOPMENTS SPECIALREPORT
Combined technology effects. Fig. 15 illustrates the
expected benefits to be obtained from a step-wise improvement
in FCC reactor technology. The individual steps include:
Replacing an elevated distributor with elevated feed distri-
bution system
Replacing a reactor stripper with an advanced fluidization
stripper technology
Replacing a tee riser termination device (RTD) with a vortex
separation (VSS) RTD.
With each progression, the refiner gains the yield and selec-
tivity benefits inherent of increased catalyst-to-oil and lower
coke. The combined technology improvements resulted in a
23% increase in catalyst circulation and a regenerator dense-bed
temperature reduction to 1,260F. The technology benefits are
summarized in Table 2.
As hydraulic limits are reached in the catalyst circulation
between the reactor and regenerator, relieving one aspect of a
hydraulic constraint through a slide valve or standpipe modifi-
cation can be tempting. However, not paying attention to the
overall design can simply result in the relocation of the constraint,
introducing greater risk to the system reliability. With the onset
of clean fuels initiatives, these issues can be further strained by
increasing the percentage of hydrotreated feed to the unit and
further decreasing the coke and regenerator temperature.
Refiners moving toward more hydrotreated feeds are often
placed in the position of opting for less coke-selective catalyst, fir-
ing the air heater or firing torch oil to add heat to the regenerator.
Using a less-coke-selective catalyst (more coke produced) negates
yield and selectivity benefits previously gained from equipment
upgrades. Firing the direct-fired air heater for long durations can
lead to erosion and failure of the air distributor as the distribu-
tor jets are pushed beyond design exit velocities. These high exit
velocities can also lead to excessive catalyst fines generation. Firing
torch oil to keep the regenerator hot essentially burns high-value
feedstock as fuel, while at the same time damaging the catalyst
activity. The burning of any fuel in the regenerator apart from
coke on the circulating catalyst inventory is an economic loss.
While these issues present problems that the refiner must over-
come, they also present some excellent operational, financial and
environmental opportunities.
New technology recycles spent catalyst. The problems
of low coke operations (low regenerator temperature) have cre-
ated an opportunity that was uniquely addressed. Traditionally,
the catalyst from the bottom of the reactor stripper has com-
monly been referred to as spent. However, modern catalyst
systems can accumulate appreciable quantities of coke and still
maintain a significant amount of activity. The term carbonized
was adopted to describe this catalyst. The activity characteristics
of carbonized catalyst are not only usable, but are, in certain cases,
preferred. Coke deposition preferentially attenuates the strongest
catalytic sites, providing for more selective cracking with carbon-
ized catalyst.
In a traditional FCC unit, increasing the catalyst-to-oil ratio
to increase conversion also increases the coke yield and catalyst
circulation to the regenerator. The new technology provides the
ability to increase both conversion and selectivity by recycling a
portion of the carbonized catalyst back to the base of the reactor
riser (Fig. 16). The carbonized catalyst circulated from the stripper
back to the base of the riser is effectively at the same temperature
as the reactor. Since the recycle catalyst adds no heat to the system,
the recycle is heat-balance neutral. For the first time, the catalyst
circulation up the riser can be varied independently from the
catalyst circulation rate to the regenerator and is decoupled from
the unit heat balance.
With new technology, a portion of the stripped catalyst
(~1,000F) is directed through the recycle catalyst standpipe to the
chamber where it is combined with the hot regenerated catalyst
(~1,300F). The lower contact temperature between the combined
catalyst and raw oil feed results in higher product selectivity with
less dry gas and coke production, and a substantial increase in
conversion due to the higher riser catalyst-to-oil ratio. Similar to
a conventional FCC unit, the balance of the carbonized catalyst
that is not recycled travels through the stripper to the regenerator
1,250
1,270
1,290
1,310
1,330
Elevated
premix
Elevated feed
distributor
+ AF
Stripper
+ VSS
RTD
R
e
g
e
n
e
r
a
t
o
r

d
e
n
s
e

b
e
d
,

F
6
8
10
12
14
16
C
a
t
/
o
i
l
Regenerator temperature
Cat/oil
(0)
(2)
(1)
(3)
Technology upgrade effects. FIG. 15
Recycling spent catalyst technology. FIG. 16
HYDROCARBON PROCESSING SEPTEMBER 2010
REFINING DEVELOPMENTS SPECIALREPORT
where the coke is burned off before it returns to the base of the
riser. The catalyst flowing to the regenerator carries a higher coke
content, which, in turn, raises the regenerator temperature and
enables the easing of constraints in the system.
The ability to control the catalyst flow up the riser indepen-
dently of the heat balance adds increased flexibility to the FCC
unit to more easily handle changes in feed quality and shifting
product slates. This aspect is particularly useful in units that peri-
odically switch from gasoline to olefin or distillate mode, through-
out the year. In a conventional FCC unit, a shift in operating mode
is accomplished by a change in reactor temperature and a change
in the rate or activity of the catalyst makeup. With new technol-
ogy, a change in catalyst activity in the riser can be accomplished
by merely changing the amount of carbonized-catalyst recycle.
Result: The change from gasoline mode to/from olefin mode can
be rapid. This application has an even greater impact for refiners
that use liquefied-petroleum-gas olefin additives, i.e., ZSM-5,
and refiners have traditionally had to shift their catalyst inventory
over several weeks to reap the full financial benefits of a change in
product slate.
As of February 2010, four new technology designs have been
commissioned and are operating successfully. Four more units are
now in design, and five more are currently under construction.
Integrating the new technology provides numerous benefits in
both revamp and new unit applications, providing the refiner with
the ability to accomplish the following:
Increased conversion
Increased gasoline yield
Decreased dry gas yield
Increased propylene yield with additive use
Reduced ZSM-5 additive consumption
Decreased coke yield at constant conversion
Increased regenerator dense-bed temperature
Increased regenerator residence time
Lower regenerated and spent-catalyst standpipe flux
Lower regenerator emissions.
The expected benefits for the presented case study with the
implementation of the new technology is shown in Fig. 17 and
Table 3.
The ability of this new technology to improve conver-
sion and selectivity provides the refiner with a tool to achieve
improved yield targets with less coke make. The reduced coke
make translates to reduced air blower demand, as well as a
reduction in CO
2
emissions.
Catalyst consumption rates are also very competitive with, if
not better than, typical FCC units. For one refiner operating a new
unit, catalyst consumption was 0.14 lb of fresh catalyst per barrel
of feed processed. A second unit operation demonstrated approxi-
mately 35% lower catalyst consumption than a comparative unit
operated by the same refiner without the new technology.
Summary. Although FCC technology is over 65 years old, there
is still much to learn and improve upon particularly as refiners
maximize throughput on existing assets and move from fuels pro-
duction to higher-value products. Advanced diagnostic and design
tools are accelerating the development and creation of state-of-the-
art technology. This article showcased three of these innovations
that have emerged by using sophisticated tools.
CFD modeling, combined with radioactive tomography and
commercial data analysis enabled the development of the dual-
radius feed distributors, and effectively shattered the paradigms
on large riser design, extending well beyond previously published
operating envelopes. Using these tools, along wih radioactive trac-
ing and physical inspection reports, led to the development of the
TABLE 1. Conflicting wants
Wants Also wants
Optimum product But Sufficient regenerator temperature for
selectivities Improved coke burn kinetics
Higher conversion But Lower coke yield for:
Improved selectivity
Lower CO
2
Less dry gas But Sufficient regenerated catalyst temperature for:
Improved coke burn kinetics
Higher throughput But Retain existing equipmentMin. CAPEX:
MAB, WGC, standpipes
TABLE 2. Technology upgrade effects
Base case premix feed Plus VSS
distributor (point 0) (point 3)
Cat-to-oil 8.27 10.14
Regenerator temperature, F 1,324 1,260
Conversion, lv% (90% at 380F) Base +0.7
Gasoline, lv% (90% at 380F) Base +1.8
Coke, wt% 5.6 5.6
coke 0.68 0.55
TABLE 3. Effects of new technology on unit
performance
Base case premix Cumulative to new fad
feed distributor distribution technology (point 4)
(point 0) Riser Regenerator standpipe
Cat-to-oil 8.27 15.6 7.8
Regenerator temperature, F 1,324 1,312
Conversion, lv%
(90% at 380F) Base +3.8
Gasoline, lv%
(90% at 380F) Base +4.9
Coke, wt% 5.6 5.4
coke 0.68 0.35 0.70
+ AF
stripper
1,250
1,270
1,290
1,310
1,330
Elevated
premix
Elevated feed
distributor
New feed
distributor
technology
+ VSS
RDT
R
e
g
e
n
e
r
a
t
o
r

d
e
n
s
e

b
e
d
,

F
6
8
10
12
14
16
C
a
t
/
o
i
l
Regenerator temperature
Cat/oil
(0)
Riser cat/oil (4)
Regen.
S.P. cat/oil
Base
Effect of new technology on unit performance. FIG. 17
HYDROCARBON PROCESSING SEPTEMBER 2010
REFINING DEVELOPMENTS SPECIALREPORT
elephant trunk combustor arm disengage, which is applicable to
combustor-style regenerators of all sizes. CFD modeling, tomog-
raphy and commercial data analysis also allowed for the successful
development and implementation of the piped spent-catalyst
distributor, which reduced the excess oxygen level and decreased
NO
x
emissions for bubbling bed regenerators. HP
ACKNOWLEDGMENTS
The authors thank the following individuals for their assistance in providing
data and/or support that made this article a realityPeter J. Van Opdorp, UOP,
who provided the yield estimate comparisons between the design case and outer
maximum case; Reza Mostofi-Ashtiani, Mechanical Engineering and Materials
Engineering Center, for providing his assistance and expertise with the CFD
models; and Dave Ferguson, Justin Tippit, Benjamin Chang, Pannatat Trikasem,
Brian Octavianus and Nurudin Sidik at Tracerco, for their dedication and effort
that contributed to a successful project.
LITERATURE CITED
1
Mitchell, T. P. and K. A. Couch, Optimix (ER) Commercialization
Ceramic Tips, July 2009.
2
Couch, K. A., K. D. Seibert and P. J. Van Opdorp, Controlling FCC Yields
and Emissions, NPRA Annual Meeting, March 2004.
Lisa Wolschlag is senior manager of the FCC, alkylation and treating develop-
ment department for Honeywells UOP business located in Des Plaines, Illinois. She
has 18 years of experience working in various areas of UOP including research and
development, field operating service, technical service and process development. Ms.
Wolschlag received a BS degree in chemical engineering from the University of Illinois
and an MBA from the University of Chicago.
Keith Couch is senior business leader of BTX/aromatic derivatives for Honeywells
UOP business located in Des Plaines, Illinois. He has worked for UOP for 18 years in
manufacturing, research and development, field operating service, technical service,
sales support and process development. Mr. Couch received a BS degree in chemical
engineering from Louisiana Tech University and is pursuing an MBA from the Univer-
sity of ChicagoBooth School of Business.
Article copyright 2010 by Gulf Publishing Company. All rights reserved. Printed in U.S.A.
Not to be distributed in electronic or printed form, or posted on a website, without express written permission of copyright holder.