Note: Descriptions are shown in the official language in which they were submitted.
~0~3`~
This invention concerns the separation of particulate
matter which is entrained or suspended in a gas stream moving in
a conduit. The invention arose and finds its most immediate use
in connection with the separation of fine solid catalyst particles
; from the gases produced in hydroca~bon conversion processes such
as fluid ~ed catalytic cracking processes, and it is therefore
primarily described hereinafter in relation to that field of use.
In fluid bed catalytic cracking of petroleum, con-
version of heavy or residual oils to lighter hydrocarbon fractions
is effected by contacting the oil with a hot, particulate catalyst
- as a fluidized bed or flowing suspension. In one widely practiced
cracking process known as "riser cracking", this contacting is
carried out in a reactor in the form of an elongated upwardly
extending tube which is referred to in the industry as a "riser
tube".
In this type of process, oil at a temperature of about
500-800 ~. is mixed at the bottom of the riser tube, with hotter
catalyst at a temperature of about 1150-1350 F. Contact of the
hot catalyst with the oil results in the very rapid generation
of very large volumes of gas, which cause transport velocities
in the riser tube o~ xoughly 35 to 50 ~eet per second. The
cracking reaction continues as the gas-particle mixtu~e moves
upwardly in the tube and until the catalyst and gases are
disengaged.
, - ~ '`
.' . ~. : ,~` `
~ 2~
10~ 7U'~
In order to stop the cracking reaction at a desired
stage and to prevent degradation of desired products, it is
necessary very rapidly to disengage the catalyst from the re-
; action products after the desired period of contact. This is
commonly done in what is known as a disengaging chamber. For
effecting this separation it has been conventional practice to
use one or more cyclone separatoxs, gases being separated and
discharged through the gas outlet of the cyclone and solids
being discharged through a dipleg to the lower part of the dis-
engaging chamber. If the degree of separation achieved in a
single stage cyclone is not adequate the effluent. containing
a small portion of solid particles still entrained in it, may be
further separated in a second stage cyclone.
As utilized in hydrocarbon conversion processes of `
;~ the specific type just referred to, this invention is especially
concerned with the separation of the catalyst from the gas-
catalyst mixture as it comes from a riser tube to the disengaging
~; chamber. In such processes the efficiency of catalyst separation
has important consequences~ Catalyst solids which are not
``20 separated and which remain entrained in the cyclone effluent -~
gas are lost to the cracking operation and must be replaced, or
recovered and returned to the process, to maintain a given -~
catalyst/charge ratio and to minimize catalyst costs. Moreover,
: catalyst particles which travel downstream with cyclone effluent
~ cause erosion of processing apparatus. The need to limit catalyst
; ~losse an ltself become an operating lirit on oil charge rate,
~ 3_
~.
, .. . . .
! 7 ~3 ~9
and th oo capacity. Further, in operating at high throughputs,
temperatures in the disengaging chamber can become so high as
to constitute operating limits as metal stress limits are
approached.
Although the cyclones used for catalyst separation are
already efficient separating devices, being capable of separating
up to 99.995% of the catalyst solids, they must handle very heavy
loads: in refining operations, catalyst feed rates to the riser
may exceed 1,800,000 pounds per hour. It will be apparent that
separating inefflciencies of only .005~ can still mean substantial
losses in tenms of actual pounds of catalyst.
For these reasons, it has been the o~jective of this
invention to provide means for effecting the required gas/particle
separation which will permit operating limits to be raised and
lower particle losses to be incurred.
Prior Art
An early approach taken by the prior art to the problem
- of disengaging solids from gases in riser cracking i5 shown in
Slyngstead Patent No. 2,994,659. There the riser tube has a
. 20 plurality of discharge slots in its sidewall, below a closed upper
end. The entirety of the effluent from the riser is dischar~ed - ~ -
directly into a disenqaging chamber in which there is a reduction
in the superficial velocity of the gas, which permits some catalys :
to settle out. A two-stage series cyclone separator has its inlet
open t the diuengaging chamber.
' .
.:
.:
~ 9
Experience showed that that arrangement was inefficient,
and that the degree of disengagement occurring in the chamber
decreased rapidly with increases in the superficial gas velocity
in the chamber. Above a certain limit ~usually between 3.5 and
5 feet/second, depending upon catalyst particle density and size
range, geometry, gas density and other factors), gas flowing
at high velocity from the riser outlet through the disengager
to the cyclone inlet, simply maintained a large portion of the
solids in suspension and carried that load into the cyclones.
The system was effective at low rates, but highly ineffective
at high rates required for economical operation.
An alternative approach subsequently taken by the art
is that shown in Wickham Patent No. 3,152,066. There the riser
tube had a single outlet opening in its sidewall, directly
opposite the cyclone inlet. There was a small horizontal gap
between the riser outlet and the cyclone inlet, to permit
stripping steam in the disengager chamber to be exhausted through
the cyclone. The entirety of the riser effluent discharged
directly into the cyclone. The outlet gas from the ~irst stage
cyclone passed directly into a second stage cyclone. In practice, ~ ~`
that system was also found to be poor in terms of separation
results. The cyclone system was very sensitive to pressure
fluctuations in the riser J such that riser changes tended to
upset the cyclone operation. This was due, at least in part,
to catal t surges in the riser and thus in the cyclone. The
`~-
-5-
10~37S19
device shown in the patent was unsuccessful in the industry.
~ It was modified by discharging the first stage cyclone gas
; effluent to the disengaging chamber, and feeding the second
stage cyclone from the chamber rather than from the first stage,
but those changes made no substantial improvement.
Subsequently to the Wickham type of disengager, the
art has more recently gone to use of a "T" shaped header on
the end of the riser tube. ~he "T" has horizontally extending
arms with outlets opening downwardly to the disengaging chamber,
-` 10 away from the cyclone inlet. The degree of separation occurring
in the chamber upstream of the cyclone is improved, in comparison
to the Slyngstead system, and the superficial velocity limit
is higher, but it nonetheless remains a rather sharp limit once
reached. Also, the degree of separation tends to vary drastically
depending on the height o~ "T" outlet above the bed of catalyst.
The closer the bed to the "~", the poorer the separation and the
higher the cyclone loading. Moreover, the downblast of catalyst
at rather high velocity causes severe wear problems on the riser,
the dipleg and the flapper valve at the end of the dipleg.
Chipley Patent No. 2,648,398 shows an air cleaner
~- comprising an elongated chamber having an inlet in one sidewall
with unrestricted communication to the atmosphere, and a dust
outlet ening ~mal1er than ~he inlet and aligned opp~site it
.' .
.
.
in an opposite sidewall of the chamber. The outlet also opens
to atmosphere. Suction is applied an air outlet of the chamber,
to draw a~r laterally from the space between the inlet and
outlet. Dust particles move laterally to the longitudinal
axis of the chamber, across the chamber from inlet through and ; :
out the opposed outlet, while clean air is drawn off longitudinall f
Smith NoO 2,540,695 shows a fuel ~conomizer and air
cleaner for motor vehicles, in which a funnel-mouthed member,
mounted behind the radiator of an automobile, leads inwardly to
a tubular baffle which is enclosed within a concentric ilter.
The filter has a nozzle outlet opposite from the funnel-shaped
inlet, through which nozzle grit is discharged to atmosphere.
A carburetor inlet leads radial]y from an annular chamber sur-
rounding the filter.
Patent No. 3,597,903 shows a vacuum cleaner in which
the intake manifold has an endwise opening to a filter bag and
an upstream sidewall opening into a secondary filter bag.
The space surrounding the two filter bags is under negative
pressure. Dirt is entrapped preferentially, first in the in-line
f~ilteF bag, then, when that is full, in the sidearm bag.
, Summary of the Invention
; In accordance with this invention as practiced in
hydrocarbon riser cracking processes, the riser tube leads to
. a disengaging chamber and opens directly into the disengaging
chamber through an outlet opening. The outlet is preerably
.~:
:::
_7
'
.
0 i3~
essentially perpendicular to the longitudinal axis of the tube.
The riser has a second~outlet or port which is upstream of the
outlet opening, in the riser sidewall, generally parallel to
the tube axis. That opening communicates directly with the
inlet of a cyclone separator. The first stage cyclone gas
discharge may be fed to the inlet of a second stage cyclone.
The disengaging chamber is at a pressure greater than that
in the cyclone, and there is essentially no gas flow through
it. ,
The invention makes use of the high velocity of the
catalyst particles and gas moving in the riser. The gas, being
of low density in comparison to the catalyst, can make the
angular turn through the upstream sidewise port into the cyclone,
while the dense catalyst is transported by its momentum into
the disengager. Thus, the gas is d:irected into the cyclone but
the bulk of the particles are projected into the disengager, out
of the deflected gas stream. (This may be contrasted with past
systems wherein the entire gas/particle stream is directed into
~ the cyclone, and with other systems wherein the entire stream is
directed into the disengager.)
The disengaging chamber is essentially closed to the
~low of so large a volume Qf gas through it, and there is no
substantial flow of gas from the riser through the disengaging
chamber. A static back pressure is maintained in the disengaqing
chamber ich diverts ~he g~ angularly from the riser so that
: :
.
.. . . . . .
104 3;~09
it does not pass through the disengaging chamber, but instead
passes into the cyclone inlet. The solid particles, having
higher momentum by reason of their higher density, continue
traveling in the upward direction in which they were moving
in the riser tube, and are not deflected by the backpressure.
They thus exit through the outlet opening of the riser into
the disengaging chamber and accumulate as a bed at the bottom
of the disengager from which they are drawn for stripping and
recycling. The major portion of the solids are thereby by-
passed around the cyclone and do not enter the cyclone at all;
a minor portion, which may be of the order of 10-20% of the
catalyst, does enter the cyclone, and is separated there.
Several surprising consequences are obtained by
practice of the invention. An especially unique and advantageous
result is that significantly higher cracking temperatures can be
used. Moreover, the superficial velocity of gases in the dis-
,~ engaging chamber is eliminated as an operating limit; and, third,
there is a dramatic improvement in separating efficiency and
stability over a wider range of operating conditions.
Because the riser discharge gas does not flow through
the disengager, there is essentially no flow of the gas in the
disengager. That is to say, the superficial or space velocity
(defined as gas flow divided by cross sectional flow 2rea) is
, essentially zero. This factor, which had been a critical upper
limit i certain earlier configurations, is no longer a limit.
',:
', " .~
.
;'S)9
With no superficial velocity in the disengager, there is
essentially no re-entrainment of particles discharged from the
riser, regardless of operating rate. Moreover, since the
pressure in the disengaging chamber is greater than the pressure
in the cyclone, there is no tendency for the cyclone to discharge
gas downwardly through its dipleg, which in itself would tend
; to cause re-entrainment.
The temperature at which the disengager runs can -
surprisingly be raised by use of the invention. Many disengaging
chambers now in service have been fabricated of metals which will
withstand internal gas temperatures up to about 950 F. If
modified to incorporate the structure of this invention, it is
found that the same disengaging vessel shell can now be run
at temperatures of about 1050 F., i.e., about a 100 F. increase,
without exceeding metal stress limits. This is an important
advantage, since it has been found in recent years that these
higher temperatures are desirable in terms of their affects on
the cracking reactions. Thus, modification of the riser-cyclone
structure in an existing disengager shell structure enables the
process to be run at optimum but higher-than-original-design gas
temperatures~
The reason ~or this appears to be that a static gas
~oundary layer now overlies the vessel wall and ignificantly
reduces heat transfer from the gas to the shell. Thus the sensed ~ -
shell temperature is in fact less, at the same riser discharge
gas temperature. -
; ~ ' '", ~ ' ' :
The invention can best be further described by
reference to the accompanying drawings in which,
Figure 1 is a diagrammatic elevation of one common
type of riser cracker,
Figure 2 is a fragmentary ~ertical section of the
disengaging chamber of a riser cracker having disengaging
structure in accordance with a preferred embodiment of the
invention,
~` Figure 3 is a horizontal cross section taken on line
3-3 of Figure 2, and
Figure 4 is a fragmentary vertical section of a
modified form of the invention.
As previously suggested, the present invention finds
its most immediate application in the disengaging o~ catalyst
particles from gases in connection with the riser cracking of
hydrocarbon conversion processes. For that reason the invention
is shown in the drawings with specific reference to that field
of use, although this is not intended as limiting. ~ ;~
., In the common form of riser cracker structure, as
shown in Figure 1, the oil charge is pumped to the bottom of the
riser tube, where it mixes with incoming hot catalyst from the
regenerator. Contact of the hot catalyst with the oil rapidly
generates a very large volume of qas and cracking occurs as the
mixture rises in the riser. The elongated tubular riser conduit
¦ l-ads rtically or angularly upwardly to an elevated disengager
1~ 1
;'
:: . ' ` ` ` . '
1(3~`')q~9
vessel for separation of the catalyst from the gases. The
separated gaseous products are taken off to fractionation for
separation into gas, gasoline, light cycle oil, gas oil, and
other products. The catalyst accumulates in a bed as indicated
by the dotted line, in the lower or stripper portion of the
disengaging vessel. Steam is added to the vessel to strip
uncracked oil from the ca~alyst particles. The stripped but
coke-encrusted catalyst is returned from the stripper to the
regenerator wherein coke is burnt off by the addition of hot
combustor air, producing hot flue gas as a product. The hot
~; ca~alyst i5 then recycled. A hopper is commonly provided for
catalyst storage. For further description of riser cracking,
reference may be had to Hydrocarbon Processing, Vol. 51, No. 5,
; May 1972, pages 89 to 92; ibid. Vol. 53, No. 9, September 1974,
pages 118-121; or to "Fluidization and Fluid-Particle Systems",
Zenz and Othmer, Reinholt Publishing Corp., 1960, pages 7-15.
In the disengaging structure of this invention, as
shown in Figure 2, the riser tube 10 enters the disengaging
vessel 11 from below and extends, in the embodiment shown,
upwardly generally along the vertical axis of the vessel. The
space 12 within the disengaging vessel around and above the
riser tube is referre~ to as the disengaging chamber. At its
upper end, riser 10 is vented directly into cham~er 12 through
an ou et opening or port 13, which preferably is an endwise
-12-
~- . . . - .
~1 ~0~37Q5~ ~
: opening, perpendicular to the axis of the tube and to the axis
of the chamber 12. Above the open upper end 13 of the riser 10
a downwardly facing deflector cone 14 is mounted to the top of
the disengager vessel. The purpose of this deflector cone 14
is to deflect catalyst particles which are discharged through
riser outlet 13, thereby preventing them from abrading the upper
end of the vessel, and also to minimize any "fall back" of
particles back into the riser tube through the open end thereof,
which might cause re-entrainment.
Spaced a short distance below but adjacent to riser
outlet 13 is at least one port 17 in the sidewall of the riser.
The preferred embodiment shown is a balanced or symmetrical
arrangement in which the riser is provided with two sidewise ports
17, 17 which are diametrically opposite one another, each of
which feeds a separate two-stage series cyclone separation system
(best shown in Fig. 3~. Specifically, each sidewise port 17, 17
is connected via a lateral or transiverse conduit, designated
respectively at 18, 18 to the inlet of a first stage cyclone
19, 19. The cyclones may be generally in accordance with known
configurations, and the cyclones themselves do not comprise the
invention. It is important, however, to note that the first
stage cyclones are fed solely through the sidewise ports17, and
not through the chamber 12. The conduits 18, 18 feed particles
: tangentially into the respective cyclones, wherein a fjurther
gas/parti e separation }s made. Particles separated in the first¦
Il
, .
.'' , .
o ~ ~ 1 0~3
stage cyclones 19, 19 are discharged through downwardly extending
dipleys, one of which is shown at 20 in Figure 2, with the gaseous
effluent discharged at the top through gas outlet conduits 21, 21
which are connected to the respective cyclone bodies through
expansion joints as shown at 22.
The upper end of the riser is desirably provided with
external stiffening means designated generally at 25, to support
the cantilever load of the cylcone separator which hangs from it.
It is also useful to provide a shoe, as at 26, on the side of
the riser, to prevent the cyclone from coming to bear on the
riser wall.
The gas outlets 21, 21 of first stage cyclonesl9, 19
are respectively connected through conduits 27, 27 to the inlets
of second stage cyclones 28, 28 respectively. Where two
stage cyclones are used, each second stage cyclone can be con-
nected directly to the gas outlet 21 of a first stage cyclone.
The conduits 27, 27 constitute the sole inlets to the second
stage cyclones; that is, those cyclones are not fed through or
from chamber 12. Expansion joints are provided to accommodate
the differential expansion between the two cyclones. The second ~- :
s~age cyclone diplegs, one of which is shown at 29 in Fig. 2,
discharge particles separated in the second stage to the bottom
of the disengaging chamber. The dipleg should desirably
~,~ terminate above the bed so as not to be covered by it. The gas
outlets 30, 30 of the second staga cyclones extend through the
disengager vessel and are connected to a manifold leading to
fractionators, not shown.
iq
, .
-14-
. ,, . , , . .......................... . . - .
.
lO'~ 'J
The disengaging chamber is pressurized, being con-
nected directly to the riser, but the catalyst discharge port
is covered by the bed of catalyst which thus restricts escape
of gases from the disengager. Steam is admitted to assist in
the stripping operation (see Fig. 1). The steam flow is very
moderate, for example of the order of 1500 lbs./hr. at 150 psi.
Except for the minor flow of stripping steam which percolates
upwardly through the stripper, there is essentially no flow of
gases through the disengaging chamber.
In operation, the internal pressuri~ation of chamber
12 blocks significant flow of gases into it through riser endwise
vent 13. The catalyst particles, having relatively high density
and low volume, are carried by their momentum into the riser
chamber, but the gases are diverted angularly to the cyclone
through ports 17,17. By far the greater portion of the catalyst
is separated where the gases are directed angularly sideways
while the particles are projected out of the riser, and these
particles largely bypass the cyclone system. A minor portion
of the particles are not separated, or are re-entrained in the
gas and enter the cyclone system. They are largely separated
in the first or second cyclones, which carry a much smaller load
than in the prior art. There is a pressure drop of about 2 psi
i through the cyclones.
Abxuptness of change of direction of gas flow is
important to achieving separation, because the particles do not
change direction as rapidly as the gas does. In this connection,
,, ~
ll
-15- '
~i.
y~
it is further beneficial to increase the velocity of the
gas/particles stream just upstream of the sidewise ports 17, 17.
For that purpose the invention preferably employs nozzle means
in the form of a conical neck or restric~or in the riser, as
indicated by the step-down section at 32 in Figure 2. This
neck reduces the cross-sectional area of the conduit, so that
the stream is accelerated as it moves past.
Alternatively, or in addition, where an assymetrical
or an unbalanced cyclone construction is used, it is advantageous
to employ a baffle or deflector means, preferably in the form
of a deflector plate 33 (see Fiq. 4) which projects angularly
inwardly from the riser sidewall just upstream of the sidewall
outlet and in line with it, such that particles are deflected
away from the sidewall outlet. The plate is preferably angulated
¦ at an angle A, with respect to the vessel sidewall 10, of about
¦ 30, and projects about 15% of the way across the tube diameter. ~ ~;
This further improves efficiency of operation, as will be shown
hereinafter.
The following examples and comparisons with other
particle disengaging techniques will further illustrate the
practice and advantages of the invention.
The data for Runs 1-10 in Table I was obtained with
a prior art disengager in which the entire effluent from the -
riser tubé was discharged through a sidewise port in the riser,
~thro a lateral conduit directly into the inle~ of a fir6t
-16-
,'
~ .
~:
,: . ' ~ . , , ' , ,
;'0~ ~
stage cyclone. There was no endwise vent, and all of the
catalyst went into the cyclone system. Also, the gas outlet
of the first stage cyclone was vented to the disengager chamber,
and a second stage cyclone had its inlet open to the interior
of the chamber. Stripping steam from an external source was
su~pli~d ~ the dl-onya9 ~h~b~r
-17-
,
: -
v~ ~
u~ ~
O ~ ~ o~ ~ ~ ~ cn
. ~ ~I d~ oo o ~ ~ o
~ _l ~ ~ r ~ r _ ~
t~l
U~
O ~ ~ ~O `D O 0) ~ 1~
_l D ~ o~ l o o
~
~ tn
r-~
3 o\o ~ ~ t~
V~ ~1 ... - - .....
C4 ~ ,~ 4 o O~ ~ O O O : . '. '
. :' .
~;
~r~ ~ ~ O ~ 1~ ~ r~) ~ ~ ''.
Ul ~ In ~ O t~~1 0
~,~ ~n ~ ~ ~ ~ ~, r_ ~ ~ ~ ~ ,:
~ C ~ r5~ co oo oo oo co a~ ~ ~ 0~ O
:4
. ~ `D u~O ~ ~ I`
~d h ¢ e~ O ~1~ 0 o ~ co
. h
¢
~_
3 a p,~ o cn ~ ~ ,c. u~ o
¢ ~ ~ ~ a~ ~ n ~1 ~ ?~
'I ~ ~ ~E~ ~ ~ ~ o~
i~ ~ P r r r,
O _
_ ~ ~ ~
p.... ~1 r-l ~1 . .,
~_
.r~ C~ ~ ~ CO ~ ~ ~ ~ ~ ~ ~ ~ ~
d u~ t7 t` 3
O D . . . . . . -
a
,b4
h ¢ ~t d ~ ~O `D
.~ .rl
. O
r~ : '
oo o ~ ~ ~ ,~ ~o oo cn ~ oo 00
_ O ~ C~
I ~D ~ ~ ~1 ~O ~ OL~l ~ t` ~ 1- ~1
¦ ~ _ _ ~ ~ ~ r ~ ~ ~ ~
1 o ~ o .~ oo o o ~1 ~
.
.,
O O O O O O O O C> O ~ O ~:
o o o o o n o o o o c~
~: ~ O O O O O t~ ~ ~ O O O Ll~
~- ~ o~ ~ ~ ~ O ~ CS~
t_~ V~ U) .
O ~: ¢ ¢ O ~ ¢ ¢ O h
~. ~
~ ¢ E e
:, .
--1 8--
.
: : ' . : ' ~'; ;
The data in Table II was obtained after the structure
~rom which the Table I data was taken was changed to incorporate
the vention.
~1 ~
~.,
~,
~, . ;:~
,' . ~
1~ -19-
, ' ~
.. . . . . . . .. ..
.,1 ~ l~ P~
vl ~
O ~ 00 ~ ~ ~ ~ O 0
I ~ In O ~ n ~~
V~ ~ CO o ~ I~ o~ o ~
U ~ _ _ ~
O ~ ~ O e~
~ o o o _I o ~ o o o
~ '
dP 1~ V~
:~ ,
V~ ~,
:~ ooO t~OOOO
.
.~ ~ ~d
~rl ~ 00 Cr) ~0 0 ~ 00 S~
~n ~ l o c~
~ ~ oo oO oO a~
U~
.
h ~¢ O ~ c~ o Ul ~o ~ ~1
U
C~ ~ U ' .
,_
a
O ~:4 p,
~ ~ ~ ~ Ln t~ ~ ~ Lr~ Lr~ ~ o
ILl E-- O ~ E~ 0 E~
~ ~ ~ "
¢ ;~ _
~ _
._~ cd t
o~
Ll~ ~ ~d ~ ~`I ~ d 0~ 00 ~ ~1 ~0
C ~_~ Ln Ln In ~ Lr~ ~ ~ .
o.> a
bO
,~ ~ ~ ~ O L~ ~--O ~ ~ O
t~ ~d ~ ... .....
¢ ~ Vl ~ C~ ~ ~ t--
_~
~1
o
æ ~
C~ ,.
.
C~ ~ 00 ~ `D~O L~
_ ~ ~ ~ ~ o~
., e~ ~ LS~ 00 0 a~
~ ~ _ __ a _ _ ~ _ a
h ~ ~ Ln t~
~a
rC ~
~ :''' ,
:.
a~ ~ o o o o o o o o o o
i ~ ~ O O O O o o o o o Ln
O O O O O O O O O t~ ~ I
V~ O 00 ~et O ~ ~ ~ o t~ ..
D t`~ ~iJ L~ ~ O 0 ~.0 ~
t_) .~
.
O ~ o Ln O
:~ ~ ~ ~ ~
~:: ,~ 4 ~ ~ Ln ~ 0 h
Y ~ ~ ~¢ oo
--20--
1~ '0~ ~
Catalyst A was a silica-alumina equilibrium catalyst
in microspherical particles having an apparent bulk density of
.72 grams per cc. Catalysts B and C were of the same general type
as catalyst A but had apparent bulk densities of about .82 grams
per cc.
The data in both Tables I and II is taken from yield
summaries. Where a Run No. is followed by the character A, the
data given is the average for a week, rather than being an actual
day's data. Stream densities are from Petroleum Tables compiled
10 by E. W. Saybolt & Co., and are based on API gravity of stream
according to yield summary. In these tables "Cat. Rate" repre~
sents the rate of catalyst circulation through ~he riser;
"Fxact. Btms. Flow" represents total 10w out of the fractionator
bottoms stream; "Fract. Btms. BSW" is the volume percent of
catalyst in the fractionator bottom stream~ and "Cat. Loss"
s represents the amount of catalyst not recovered by the cyclones,
assuming that all catalyst entering the fractionator leaves
in the bottom stream. The catalyst loss in pounds per day
was computed by converting the fractionator bottoms flow to
gallons per day and multiplying by the volume percent of catalyst
in the stream density. The catalyst loss in pounds per barrel
of raw oil charge was computed by dividing the loss per day by the
raw oil throughput, converted from pounds per hour to barrels per
day.
~ y comparison of TablesI and II, it can be seen that
the invention markedly reduced average catalyst losses, while
,~ : ~ '
,~ , ~
~ '
~ 21- ~
., .
.. r . -. ~ .. . .
~ . . . .
at the same time enabling oil charge rate to be increased.
Comparing invention Runs 11, 12 and 13 with the prior art Runs
6~10, all of which were the same type of catalyst, it can be
seen that the average loss per barrel charge was reduced by
52%, at the same time oil charge rate was being increased by
13%. Moreover, chamber temperature could be increased to
1050 F., from the previous limit of 950 F. so that a better
quality product was obtained.
The data set forth in Table III following, was obtained
from a bench scale separator in which cracking catalyst was
suspended in air, rather than cracking gases, and the data
does not represent actual cracking runs. In Runs A and B of
Table III, the simulated riser tube discharged to the dis-
engaging chamber through a "T" shaped header at its upper end,
above the catalyst bed at the bottom of the disengager. The "T"
had side holes and a bottom hole through which gases were vented
directly to the chamber. A ~irst stage cyclone inlet opened
to the chamber, and a second stage cyclone inlet was fed directly
from the first st~ge cyclone gas outlet. In Runs C and D,
the riser discharged through a "T" having 45 baffles at the
outer open ends of the arms to deflect the discharged material
downwardly.
In Run E, the riser was vented through an open upper
end to the chamber and was connected through a sidewise port,
just below the end vent, directly to the cyclone inlet, in
accordance with the invention.
,,,.,, .
,,
;i . .
ll -22- 1 ~
,.' ~':
, ~ , ' ; ' ' , ~ ~ . ,
~ 3'7~
,~ ~ . ~ ~3
.. ,. o~o
~n ~ ~ 0 o v~
o~ o ~ o
. ~ o~oo~oo\O . . , ,, ~,,
U~
h ~ O
~ ~ ~ .
V~ t~ `P ~ P~
. 5,~ .
~ V~ ,:
::~
U~
~ r~ v~
:~ ~ a) ~ f~
o ~ o
o a~
U U~ -- U~ ` I~~ U
~ . ~ ~ ~O Q~
a~
~1 .,.~
~ ~ aJ :
.; ~
.; ~ ~ ~ :
*
O Q) ~ ~C7 U~
OJ
~ , ~ ~ ~,~ 0 ~
c~ ~
ca ~
~1
~ .
~` o ~ ~ u ~ ~ u~
o a.) ~ #
H S ~ ~ 1 ~ IJ~ ~t O ~r1 t
~-1. ~ ~ ~ ~, o
i ~ V~ ~ ~ ~ V~
~ ,~ g
E-~ h ~ ,; .,~
:: .,~ ~ ~o o o o ~ ~ ~
U~¢ ~ ~ ~ ~oo ~ ~ o ~ C
o
H
, ~ ~ ~ h
O~ ~ ~ ~ ~ ~ u a~ ~
t)td a ~ ~ . . . . . .,~
~:~ . ~ .C ¢ ~ ~:
.. C~ V~ ~ O P.
~11 ~ ~'
O r~ .
'i O ~ ~ X
a~ R ~ ~ ~
a~ _1 tq ~ ~ u ~
o .,~ o
' q> ~1 o
s, ~ ~ o
o~. ,_ .
~ ~ ~1 ~O ~0 ~ ~
,_1 0 cr: ", ~, ,
.,
.
--23--
3',CI~ ~
In the runs referred to in this table, as well as
the runs in Table IV following, the catalyst used was an FCC
equilibrium catalyst with the following typical particle size
distribution:
0-20 microns - 0 wto ~
0-40 microns - 8 wt. %
0-80 microns - 70 wt. %
Bulk density of the catalyst was .8 grams per cc. Separation
efficiency is 1 minus the quotient obtained by dividing the
catalyst flowing into the first stage, by the catalyst feed
rate to the riser.
In Table III, the amount of catalyst collected in the
first stage cyclone dipleg shows the completeness of disengage-
ment. The invention (Run E) achieved much more compiete recovery;
it had a 7.05~ first stage recovery, in comparison to recoveries -
in other systems of from 13.9 up to 42%. Most significantly,
the amount of catalyst remaining in the system for recovery in
the second stage cyclone was very low~-only .002 lbs.
:~,,
:`
;,
~ -24- ~
,: . "~
~ .:, ,.
.
.
., .
\
~43'iJ09
~: ~ '~ ~ 3 ~ ~ ~ '~
C O ~J~4 3 V-l
1~ c~ C 0 C C
~ ~ ~~ ~ O ~ 0 0 0 0 0 ~. O _- Vl O
C~~ ~ -- 0 ~ ~ ~ ~ C ~ O ~
~U O~r~ O ~ 0 C~ rl 0 ~J .C
C'~J U q ~ D 'O q ~ C~ l ~ n ~
~ ~ . .~
~, C4 0 0 D D
'` ~ c u C c " ~ , o O
~ O O ~ O ~ O X ~O ~ ~1 _I O ~
U ~ U`- % .,~ . . . . . Z C Cl
Co~ ~ .
.~ C ~
' ~ U C) ~ o ~ O .
;~: ~ ~ O '~ ~ 01~ ~ O O O O O
J
.'. .
.
'' V) 00 ~ 0
. ~ C ~ c o~ ~-r o ~ co ~ ~ O o
~ _~ O q` ~
0 ~_
O ~ X
' ~ UL~ ,~ ~ o ~ ; ;
' : : '
O U ~ . .
,:0C U -1 ~ ~ o o ~ ~ ~ ~
V . .
., ~~ . , ' '
.`',,_ . ' .
.'` ~ C ~, 0 .
0 n V~ OC 00
o o
.~3 ~ ` : :
o e~ : 1: .
~ O ,
.11'1 C ~ . ~ .
' "' ~ ~ ~
~ r~ - ~
:`
,,, . -25-
, . .. . .. .,
. . '~
,
,
i
3 ~ ~
Table IV illustrates the results obtained in the same
simulated system, comparing the invention ~Runs F and G) with
still other systems (Runs H thru P). The structure utilized in
Run F was the same as that of Run E. The structure used in
Run G was similar, except that a deflector baffle was incorporated
in the riser pipe set below the sidewall exhaust by about 1/4 x
the riser diameter, in the form of a plate extending at an angle
of about 4~ to the riser axis, and pro~ecting across about 1/4
of the riser diameter. The purpose of this plate was to deflect
particles away from the sidewise gas outlet. The separator used
in Runs H and I had the riser discharging only to the first stage
cyclone inlet. The riser was not vented to the disengaging
chamber, and the first stage cyclone gas outlet opened to the
disengaging chamber and the second stage cyclone had lts inlet
open directly to the disengaging chamber. The first stage dipleg -
length was one inch measured from the intersection of the cone
and dipleg. The disengager in Run J was similar to that in Runs
H and I, except that the dipleg length was 24 inches. The dis-
' engager in Runs K and L was the same except that the dipleg
length was 18 inches. In Runs M and N the riser discharged to
the chamber through a "T" fitting having downwardly facing ports.
The two stages of cyclones were connected in series, with the
intake opening to the chamber above the "T". Runs 0 and P were
simila o Runs ~ and ~ except that ~ "cross" was used on the
26-
,~ '
~'
,
~, :
10~70~:i
end of the riser xather than the "T". The cross had four short
horizontal arms, at right angles to one another, with downwardly
facing discharge openings, the riser being connected to the
center of the cross.
Comparison of Run F with Run G in Table IV shows that
the use of the deflector significantly increases the separation
efficiency to the first stage cyclone (80.6 to 93~). Both of
~; khose runs achieved high separations prior to the first stage
of the cyclone in comparison to Runs H, I, J, K and L, wherein
10 nothing was separated prior to the first stage cyclone (the
entire riser effluent being conducted directly into the first
stage cyclone with no separation taking place in advance). The
separation achieved prior to the flrst stage cyclone in Runs M,
N, O and P was good, however, the systems tested there
fluctuated greatly as to separ~ting effectiveness if the level
of the catalyst bed at the bottom of the disengaging chamber
was less than ~our inches below a cross or "T" on the riser end.
As a result of this instability, such sysbems would not display
.~ uniformly good separations if used in commercial practice where
20 the distance between the bed and the cross can almost inevita~ly
be expected to vary substantially in ordinary operations. In
comparison, the invention provides ~ood separations without
regard bed level, so long as the dipleg is uncovered.
-27-
" '~
., . .
.~ .
,. . .
lO~ t39
In the foregoing examples the riser tube entered the
disengaging chamber through an opening in the bottom, and the
cyclones were physically disposed in the chamber. Those skilled
in the art will appreciate, from what has been said herein, that
it is not necessary that the riser enter the cyclone through the
~ottom, and in fact the riser may enter through the side or even
he top, and that the cyclones may be physically disposed outside
~f the disengaging chamber, as may be convenient especially in
the systems other than hydrocarbon conversion systems. It is
~ot the physical disposition of the cyclones in relation to the
l isengaging cha~ber which is important, but rather the fact that
i the riser discharges through an endwise opening into the dis-
~ngaging chamber and that it feeds through a sidewise opening
just upstream of the endwise opening, to the inlet of a cyclone,
regardless of whether the cyclone is inside of, or outside of,
~ the disengaging chamber.
; The invention has been primarily described herein in
relation to hydrocarbon conversion processes. However, those
skilled in the art will recognize that the invention is useful
in other catalytic gas phase chamical reactions wherein catalyst
?articles are contacted with chemicals suspended in a fluid
chemical stream flowing in a reactor tube, as well as in ~ ;
other instances wherein particles (whethér solid or liquid) are
~o be di ngaged from gases.
: ~:
-28-
. ,
: ,
l~
' ~
Examples of other fields wherein it is believed that
this method and apparatus will be especially useful and which
~ show the wide scope of utility of the invention, include the
: gasification of coal, the desulfurization of solid fuels, and heat
; exchangers wherein hot catalyst particles are mixed with incoming
gases to heat the latter while cooling the catalyst.
Having described the invention, what is claimed is:
~, , ' ~ .
,~ .
-29-
-
",
.,
.~ ,. ,r , , _
