CATALYTIC CRACKING

INTRODUCTION
The first commercial catalytic cracking unit was put into operation in 1936.
Before the introduction of catalytic cracking, thermal cracking was the major
petroleum conversion process. Basically, both processes are tools for the same
job; that is, they destroy heavy fuel oil to produce lighter, more valuable
products. Thermal cracking does this by heat alone while catalytic cracking
combines the action of heat and a catalyst. This combination gives higher yields
of more valuable products and, as a result, catalytic cracking replaced thermal
cracking.
The cat products become feed to other units, such as alkylation and
polymerization plants. High boiling liquid products are used to make lubes, and
the gas goes into the refinery fuel systems. Cat cracking feed stocks come from
atmospheric and vacuum stills, phenol extraction plants, hydrotreaters,
deasphalters and cokers.
THE PROCESS
A catalytic cracking unit brings together a heavy feed and an active catalytic
agent. In general, the objective of the process is to produce gasoline and heating
oil from this heavy feed with a minimum of coke and gas formation. Reactions
are carried out during which a carbonaceous material is deposited on the catalyst,
which is then regenerated to bum off the deposit and restore the activity. While
there are many reactions, such as cracking, isomerization, alkylation,
dehydrogenation, etc., taking place in a cat reactor, they are complicated,
interrelated and not fully understood.
13
14 Pressure Safety Design Practices
The usual feed is a virgin gas oil; that is, the part of crude oil boiling
between about 60 "F. and 1050F. Sometimes material below 600F will be
included into the cat feed; but more often, it is put into diesel fuel or home
heating oil. The heavy material above 1050F is not normally used as cat feed
because it often contains metallic compounds that contaminate the catalyst. Even
if metals are not present, there are sometimes tarry materials that end up on the
catalyst. This deposit increases the load on the regenerator, and, hence, the
1050 O F + material is less desirable than lower boiling feeds.
Broadly speaking, gas oils can be considered as mixtures of aromatic,
naphthene and paraffin molecules. Aromatic rings are very hard to crack and
when they are cracked they tend to deposit a lot of coke on the catalyst.
Isoparaffms, on the other hand, shatter easily, and the small fragments wind up
in the products as gas or C,'s. Normal paraffms are harder to crack than
isoparaffms or naphthenes, but they are easier to crack than aromatic molecules.
Higher boiling components have larger molecules which are more easily cracked.
Thus, the products and the conversion depend in part upon the type of feeds that
are used. The operating conditions of the reactor are also important. This may
be visualized if the feed is thought of as a chain, and the reactor as a hammer and
anvil. A light tap here and there will crack only an occasional link, and the chain
will be broken into a few big pieces. However, if each link of the chain is hit
hard several times and the broken pieces are again recracked, most of the original
chain will be destroyed, and a number of small pieces will be all that remains.
The strength of the chain corresponds to the feed crackability, and the hammer
represents the reactor intensity. The intensity is the combined effect of reactor
temperature, pressure, amount of catalyst, and catalyst activity. The amount of
conversion depends upon both the strength of the chain and how hard it is hit
(reactor intensity).
Measures of Conversion
The factors that affect conversion were mentioned above, but the measure of
conversion was not described. In a simple chemical reaction, conversion is easily
determined by measuring the products formed or the disappearance of the starting
material. A petroleum process, however, deals with a multitude of different
compounds, many of which still cannot be identified, let alone measured. This
makes the selection of a good measure of conversion difficult.
One common practice in petroleum processing is to define conversion as
(100 ! % of material boilings above gasoline). While this is somewhat vague, it
can be made more specific if gasoline is considered as hydrocarbons boilings
Catalytic Cracking 15
between C, and 430F on a true boiling or 15/5 still. Thus, % 430 conv. = 100 -
% products on feed above 430F. One important drawback with this measure is
readily apparent when the feed has material below 430F. For example, with 5 %
boiling below 430F in the feed, this definition would show 5% conversion
before the feed reaches the reactor. This drawback can be overcome by using the
so-called "corrected 430 conversion" which considers only the conversion of feed
boilings above 430F.
Corr. 430 Conv. = 100 - (% product based on feed above 430 "F x
100) + (100- % in feed boiling below 430 OF)
This is one of the most common measures of conversion.
While the corrected 430 conversion has proved very useful, it too does not
measure all the conversion taking place. If a feed boils between 800F and
1000"F, for example, it is possible to crack it so that a large quantity of the
product would boil between 430F. and 800F. This is really converted material,
but the 430 conversion does not define it as such.
Product Distribution
Mere conversion of heavy fuel oil to lighter products is not sufficient to guarantee
profitable operation of a catalytic cracking unit. The dstribution of products is
also important, with local refinery demands and price structure playing an
important role in dictating which product slate is desirable. Cat crackmg yield
patterns can be vastly different depending upon the amount of conversion
required and the type of products desired. Gasoline octane is most important.
Seasonal chances in gasoline sales and heating oil sales compel some
modifications to be made in conversion level. Therefore, the conversion pattern
of a given catalytic cracking unit can vary from season to season. In summer
operations, for instance, higher yields of motor gasoline are desired, both from
direct production of C5/430"FVT catalytic naphtha and also from conversion of
butylenes and isobutane to alkylate.
Catalyst Types
Many substances exhibit catalytic properties to a greater or lesser degree, but
only a very few compounds are satisfactory for commercial cracking. To be a
good catalyst, a compound must have a high activity so that small quantities will
do the job. High activity alone, however, is not enough. The catalyst must have
16 Pressure Safety Design Practices
the ability to produce desirable products. For example, a catalyst must not make
too much coke. High coke yields are bad; first, because whatever weight of feed
goes to coke is lost so far as useful products are concerned; and second, the coke
deposited on the surface of the catalyst lowers its activity. This coke deposit must
be burned off to regain activity and regeneration is an expensive process.
The catalyst must also be selective to valuable products. Gasoline is
desirable, so a lot must be produced, but it must be high octane gasoline. C,s
and C,s are sometimes required for polymerization, alkylation and chemical
production. Certain catalysts give high yields of these compounds, especially the
unsaturated components. Gases, such as methane and hydrogen, are undesirable
so the yield of these products must be suppressed.
A good catalyst is also stable. It must not deactivate at the high temperature
levels (1300 to 1400F) experienced in regenerators. It must also be resistant to
contamination. M l e all catalysts are subject to contamination by certain metals,
such as nickel, vanadium, and iron in extremely minute amounts, some are
affected much more than others. While metal contaminants deactivate the catalyst
slightly, this is not serious. The really important effect of the metals is that they
destroy a catalysts selectivity. The hydrogen and coke yields go up very rapidly,
and the gasoline yield goes down. While Zeolite catalysts are not as sensitive to
metals as 3A catalysts, they are more sensitive to the carbon level on the catalyst
than 3A. Since all commercial catalysts are contaminated to some extent, it has
been necessary to set up a measure that will reflect just how badly they are
contaminated.
Over the years, thousands of compounds have been tried as craclung
catalysts. These compounds fall into two general categories: natural and
syntheric. Natural catalyst, as the name denotes, is a naturally occurring clay that
is given relatively mild treating and screening before use. The synthetic catalysts
are of more importance because of their widespread use. Of the synthetic
catalysts, two main types are: amorphous and zeolitic.
For many years the most common catalyst was an amorphous or
noncrystalline type called 3A. Initially, all 3A catalyst contained 13% alumina
and 87% silica. To improve activity maintenance, the alumina content was
increased to 25%. Both 13 and 25% alumina grades continue to be used; the
choice at a given refinery is based on the specific situation. Another amorphous
type catalyst, containing silica-magnesia and called 3E, is also used.
A major step in catalyst development was the introduction of crystalline
zeolitic, or molecular sieve catalysts. Their activity is very high, some of the
active sites being estimated at 10,OOO times the effectiveness of amorphous silica-
Catalytic Cracking 17
alumina sites. Because the zeolite crystals are too small to be used directly, and
because of their extremely high activity, small amounts such as 3-25 % of zeolite
are impregnated on amorphous clay or silica-alumina base to make commercial
catalysts.
With the many types of cracking catalysts available today, catalyst selection
has become a more important part of refinery planning. For a given refinery
situation, both the level and type of conversion are important. For example, a
75% conversion level can be achieved at moderate reactor temperatures with
highly active catalyst, producing maximum yields of good octane gasoline; or,
a 75% conversion level can be achieved at high reactor temperatures with
moderately active catalyst, producing maximum yields of olefinic gases and a
gasoline product with very high octane. By changing catalyst types, a refiner now
has wide flexibility in choosing the conversion level and product slate that best
fit his particular requirements. In general, catalyst types are selected to suit
feedstock quality, desired conversion level, and regional and seasonal product
distribution requirements.
Catalyst Testing
Since the catalyst is so important to the cracking operation, its activity,
selectivity, and other important properties should be measured. A variety of fixed
or fluidized bed tests have been used, in which standard feedstocks are cracked
over plant catalysts and the results compared with those for standard samples.
Activity is expressed as conversion, yield of gasoline, or as relative activity.
Selectivity is expressed in terms of carbon producing factor (CPF) and gas
producing factor (GPF). These may be related to catalyst addition rates, surface
area, and metals contamination from feedstocks.
Correlations
Catalytic crackings operations have been simulated by mathematical models, with
the aid of computers. The computer programs are the end result of a very
extensive research effort in pilot and bench scale units. Many sets of calculations
are carried out to optimize design of new units, operation of existing plants,
choice of feedstocks, and other variables subject to control. A background
knowledge of the correlations used in the "black box" helps to make such studies
more effective.
An important part of any cat crackings correlations package is an accurate
feed characterization technique. Feed characterization is vital since it quantifies
18 Pressure Safety Design Practices
the relative ease with which a given feed will crack and the tendency of the feed
to form desirable products. For many years catalytic cracking feedstocks have
been characterized by routine physical inspections which are related to total
paraffins, naphthenes and aromatics. This technique was sufficiently accurate for
virgin gas oils, but it has proven less accurate for the blends of cracked gas oils
and deasphalted and hydrotreated feeds.
Correlations have been used as a tool for catalyst selection studies.
Predictions of the product yields and qualities possible with various catalysts can
provide the necessary information for a refiner to study the economics of
switching catalysts, for instance. With a good idea of the profitability of changing
catalyst types, the refinery can justify such a change in his cat crackings
operation.
Another important use of correlations is the optimization of existing unit
operations. Cat cracking correlations can provide the refiner with valuable
information for optimizing reactor temperature level, gasoline/distillate cut point,
and feed and recycle rates. The practical application of this information can mean
increased profitability for the cat cracking operation.
EQUIPMENT
The first cat cracking unit was a fixed bed reactor using natural clay pills or
pebbles. Oil vapor was passed through the bed, and the cracking reaction took
place on the surface of the catalyst. As the crackings reaction proceeded, coke
was deposited on the catalyst which lowered its activity. After about 10 minutes,
it was necessary to regenerate the catalyst by burning off the coke. The
regeneration took 20 minutes, so one cycle was completed every 30 minutes. An
approach toward a continuous process was made by building several reactors so
that at least one would be on stream as a reactor while the others were being
regenerated.
During World War 11, the fluid catalytic cracker was developed to meet the
high demands for aviation. Various design configurations developed in
subsequent decades, including a side-by-side reactorhegenerator configuration
with U-shaped catalyst transfer lines between the vessels. In 1971
FLEXICRACKING technology was introduced which features a similar side-by-
side configuration with a number of improvements including increased riser
cracking time and reactor operating flexibility, improved feed injection, more
efficient catalyst stripping, better regenerator air distribution and various
mechanical improvements.
Catalytic Cracking 19
Flexicracking Operations
The flow plan of a Flexicracking unit is shown in Figure 1. Catalyst is circulated
rapidly in U-bend transfer line between reactor and regenerator. The feed, which
is usually a heavy gas oil, is preheated and injected into the stream of
regenerated catalyst moving from the regenerator to the reactor. The hot catalyst
vaporizes the oil (there is usually 5 to 8 times as much catalyst as oil), and the
vapors help to fluidize the catalyst in the reactor. This mixture flows into the
reactor where the cracking takes place at temperatures usually ranging from 900
to 1000F. The reaction is endothermic and consumes the heat brought from the
regenerator by the catalyst.
The cracked products pass out through two stages of cyclones which collect
entrained catalyst and return it to the dense bed. Velocities at the outlet of the
dense bed are normally 2.0-3.0 ft . /sec. Upon leaving the cyclones, the vapors go
to the primary fractionator which separates the heavy products from the gasoline
and lighter components. The light products go on to the light ends recovery unit.
The heavy material is separated and either recycled to the reactor or withdrawn
from the system.
As in the fixed bed process, carbon is deposited on the surface of the catalyst
and covers the active sites. This deposit lowers the activity very rapidly.
Since the activity falls so rapidly, it is desirable to regenerate as often as
possible. In normal operation, the catalyst is held in the reactor only about 0.5
to 1.0 minute. This is accomplished by the rapid circulation of catalyst between
reactor and regenerator, which in the larger units may be as high as 40-50 tons
minute.
The spent catalyst is withdrawn from the bottom of the reactor and stripped
with steam to vaporize the hydrocarbons remaining on the surface. Stripping also
removes most of the hydrocarbon vapors which are entrained between the
particles of catalyst. Without stripping, hydrocarbon products would be carried
to the regenerator and needlessly burned consuming much of the regeneration air,
and decreasing yield of useful products.
The stripped catalyst is picked up by a stream of air and carried into the
regenerator where the carbon is burned at temperatures about 1100-1300F.
Entrained catalyst is again removed by cyclones and the flue gas goes out the
stack. The hot, regenerated catalyst leaves the regenerator and takes with it much
of the heat of combustion. This is carried over to the reactor to vaporize the feed
and to balance the endothermic heat of cracking. Thus, the process is heat
balanced.
2
0

P
r
e
s
s
u
r
e

S
a
f
e
t
y

D
e
s
i
g
n

P
r
a
c
t
i
c
e
s

I

3

3

3

z

3
0

H

0

6
:

t
4

e

0

c

M

G

r
(

0
)

&

t

Catalytic Cracking 21
Combination Unit
A combination unit is a special type of unit that was developed to reduce the
investment for a small refinery. In effect, one main distillation unit serves as a
crude fractionator as well as the cat unit primary fractionator. This same tower
also serves the naphtha reformer and visbreaker. A schematic diagram of a
combination unit is shown in Figure 2. Crude oil is topped (material boiling
below 650F is removed) in the atmospheric tower, and the topped crude is sent
to the combination tower along with cat products and naphtha reformer products.
These latter streams provide heat to distill the topped crude and also, being more
volatile than topped crude, provide a lifting effect which assists in vaporizing
more of the crude.
A sidestream of material boiling between about 650"F/925 OF is withdrawn
as cat feed. This stream is made up of virgin material from the crude and recycle
cat products. When it is available additional extraneous virgin feed also may be
blended into the cat feed stream. Fractionator bottoms are withdrawn and may
be sent to fuel oil.
While this unit is considerably cheaper, it also has certain disadvantages. For
example, changes or upsets in any one unit may be felt throughout the refinery
because of the changes in fractionator operation. However, the considerable cost
saving possible with the combination type unit has permitted many small
refineries to finance a catalytic cracking unit when they could not afford a
conventional model.
Various Design Configurations
Various design configurations for fluid catalyhc crackers are illustrated in Figure
3. Their distinguishing features can be summarized as follows:
Model 11: Regenerator at higher elevation and lower pressure than reactor.
Slide valves control catalyst circulation.
Model 111: Regenerator and reactor at approximately equal elevation and
pressure. Slide valves control catalyst circulation.
Model IV: Regenerator and reactor at approximately equal elevation and
pressure. Catalyst circulates through U-bends, controlled by pressure balance and
variable dense-phase riser.
UOP Regenerator at lower elevation and higher pressure than reactor. Slide
valves control catalyst circulation.
2
2

P
r
e
s
s
u
r
e

S
a
f
e
t
y

D
e
s
i
g
n

P
r
a
c
t
i
c
e
s

3

g

4

.
9

r
)

0

m

r
u

0

E

z

2

;
a

E

4

1
E
:

g

a

a

CatalyticCracking 23
Figure 3. Various designs for fluid catalyoc crackers.
24 Pressure Safety Design Practices
Flexicracker: Side-by-side configuration with U-bends similar to Model IV
design but featuring increased riser cracking, improved feed injection, and
various mechanical improvements are also used. Many of the reactors for the
units described above were designed originally as dense bed reactor units. Some
of them have been modified to include FLEXICRACKING features such as
minimized bed cracking and better feed injection.
Various companies worked on the development of Fluid catalytic cracking
units. During World War 11, the government requested some of the leaders in this
field to pool their knowledge so as to speed the production of aviation gasoline.
The fact that so many Fluid units were constructed and put into operation in such
a short time shows that this joint effort was successful. However, because of this
effort, many of the basic Fluid patents were held for many years in combination
with other companies, some of which also developed their own Fluid designs.
Other companies have brought out their own proprietary unit designs aimed
at capitalizing on the cracking characteristics of the zeolite catalysts. Some are
transfer line units, in which the riser becomes the reactor and the old reactor
becomes a separator vessel. Gulf, Texaco and Kellogg offer these "riser-cracker"
type designs. Separate risers may be employed for fresh feed and recycle. A
small dense bed at the top of the riser may be used to obtain the last bit of
conversion. The specific reactor design selected is related closely to the desired
product distribution and activity of the sieve catalyst used. The higher the
activity, the less time is needed for reaction. Most designs differ in the
"hardware" used, rather than in basic process characteristics.
Over the years, many improvements have been made to the catalytic
cracking process. Basically, the total development of process and mechanical
improvements in catalytic cracking since its inception can be divided into two
time periods: pre-zeolitic catalyst introduction and post-zeolitic catalyst
introduction. In the period from the early 40's to the early a's, there were some
improvements made in catalyst quality, but in general these catalyst changes did
not dictate mechanical and operational changes in the basic units. Rather,
significant improvements were made in unit service factor and flexibility. One
of the most important advances came from the ability to crack heavy feeds.
Original moving bed units cracked only light feeds that were completely
vaporized. The new fluid units could handle both vapor and liquid feeds of high
boiling points. Final cut points are now generally held to about 1050F to avoid
metals contamination of the catalyst. Many engineering improvements were
made, reducing investments, improving run lengths and service factors, and
lowering operating costs. The use of CO boilers to recover heat from the CO
content and sensible heat in regenerator flue gases was another important step in
Catalytic Cracking 25
improving the profitability of the catalytic cracking process.
When zeolitic cracking catalysts were first used, a new era in catalytic
cracking had begun. The results of these research efforts were more catalyst
types of higher and higher activity and selectivity. These improvements were of
such significance that they dictated new ways of contacting catalyst and oil to
achieve optimum catalyst performance. The increased use of riser or transfer line
type cracking was required, and revamps of existing units were directed toward
this goal. In addition, the desirability of low CRC (carbon on regenerated
catalyst) levels for these active zeolitic catalysts dictated improved regeneration
techniques. At the higher levels of regenerator pressure, flue gas expanders have
been found to be attractive, recovering the energy added by the air blower.
Emission Control
Fluid catalytic cracking units present formidable emission control problems.
Contaminants are present in both reactor product gas and regenerator flue gas.
The reactor product contains hydrogen sulfide, ammonia, and cyanides, plus
combined sulfur and nitrogen in the liquid products. Hydrogen sulfide, ammonia
and cyanides are handled as part of the overall refinery waste water cleanup. The
combined sulfur and nitrogen may be removed by hydrotreating.
Both particulate and gaseous contaminants are present in the regenerator flue
gas. Regenerator emissions regulations are set by the Environmental Protection
Agency (EPA). The most prominent particulate contaminant in the flue gas is fine
catalyst dust which escapes the regenerator cyclones. Although these cyclones are
99.9% efficient, losses still amount to about 3.5 tons/day for a 45 MB/D unit.
The escaping catalyst dust can be removed with electrostatic precipitators which
electrically charge the particles and collect them on large metal plates. The
gaseous contaminants in the flue gas streams include compounds of feed sulfur
and nitrogen in addition to the carbon monoxide formed by incomplete
combustion of coke in the regenerator. Harmful CO gas can be combusted to CO,
in CO boilers which utilize the heat of CO combustion to generate steam for use
withinthe refinery. SO, and NO, removal from flue gases can be accoplished by
hydrotreating of the catalytic cracker feed stocks.