Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
~L~7~ 3fi348/36357
This invention relates to the treatment of
petroleum and petroleum residues; more specifically, this
invention pertains to the treatment of petroleum, petroleum
residues in general, but especially high boiling residues
according to ASTM D-1180 and D-1160 method, or residual
fractions having no boiling points according to these
methods or which are destructively distilled with formation
of carbon residues. Further, this invention concerns an
interdependent treatment of component parts of the above
petroleum and petroleum residues for improving the overall
yields and space-time velocities including the use of
somewhat higher pressures, such as pressures up to 150 psi
to aid further in the recovery of the desirable products.
PRIOR ART
In general, the prior art has attempted the
upgrading of heavy petroleum fractions with limited success.
Many of the petroleum components, as residues, have been
especially refractory to further treatment. By stripping
all of the oil components from the bottoms, such as by steam
stripping and the like, or by further treating of these such
as by hydrogen transfer reactions, partial coking, delayed
coking and the like, these bottom fractions have been
improved and/or further components obtained therefrom. For
example, by uslng a hydrogen donor solvent, some of! the
~2337~7 36348/36357
-- 2 --
hydrogen from the solvent is transferred to the higher
boiling petroleum residue fractlons and thereby the
viscosity of a part of a high boiling fraction is improved
such that lighter products are obtained. Typically these
reactions are carried out at very high temperatures and with
or without the presence of hydrogen under high pressure.
Other methods have been used where~y hydrogen
obtained by stripping hydrogen from part of the very
refractory residues have 'oeen used in a parallel stream
employing hydrogen as a reactant gas and the coked product
thereafter employed for other purposes or burned. The
excess hydrogen may also be used SUCh as in the donor
solvent system or elsewhere in the refining cycle used to
improve the residues~
In my previous work,published on ~ovember 18,
1981 in Great Britain as U.K. Patent Application 2,075,542,
I have disclosed a process for hydrotreating petroleum
residue materials in which the residue material is con-
tacted with steam and with alkali metal hydrosulfides or
the empirical monosulfides or polysulfides and/or the
hydrates thereof and mixtures of the foregoing to hydro~
crack, hydrogenate, hydrotreat, denitrogenate, and/or
demetallize and/or desulfurize the carbonaceous material.
According to the process disclosed in my above application,
the treatment of petroleum and
36348/36357
~L2~377~
-- 3 --
petroleum residues has been considerably improved based on
yield, conversion per pass, space-time velocity, API number,
etc., such that the process produces upgraded products of
greater value.
In my V.S. Patents 4,366,044 and 4,366,n4s to
dif~erent processes which relate to the treatment of coal, a
number of prior art references are also mentioned. The
references as cited in these two patents and as found of
record therein provide a background information to the
alkali sulfide chemisty and also include, in part,
references related to the treatment of crude oil or
petroleum refining residues. These references show various
attempts in the field to improve the residues or treat
various carbonaceous materials.
Other prior art, based on U.S. patent literature,
has been summarized in M.J. Satrian, Editor,
"Hydroprocessing Catalysts for Heavy Oil and Coal~, Noves
Data Corporation, Park Ridge, New Jersey, U.S.A. (1982),
e.g. pp. 62 to 78. In the U.S. patent literature the
following patents have been noted (even though these may
have been mentioned inter a _ in the above art): U.S.
Patents 3,520,825; 3,775,346; 3,850,840; 4,117,099;
4,119,5Z~ and 4,203,868. This prior art, however, does not
disclose the herein described invention, which represents a
further discovery and an invention over my process mentioned
.
'`
3634~/36357
~33~
in published U~Ko Application 2,075,542.
Other less relevant prior art is also found, e.g.
in Fuel, Vol. 62, FebO 1983, which issue is devoted
exclusively to fundamental studies of catalytic coal and
carbon gasifications (cf. Rikuchi et al., "Supported Alkali
Catalysts for 5team Gasiflcation of Carbonaceous Residues
from Petroleum, ibid., pp. 226 et ~., who report carbon
deposits on supported catalysts, generally attempt
substantial gasification, and had no cognizance of the role
sulfide series of alkali metals play in these reactions).
Still other art has been disclosed by Sikonia et
al., "Flexibility of Commercially Available UOP Technology
for Conversion of Resid to Distillatesn, Publication
AM-81-46 of National Petroleum Refiners Association (NPRA),
1981, Washington, D.C., ~.S.A.; Ritter et al., "Recent
Developments in Heavy Oil Cracking Catalysts", Publication
AM-81-44, NPRA, 1981; Bartholic et al., ~Utilizing
Laboratory Equipm~nt in New Residual Oil Development~,
Publication AM-81-45, NPRA, 1981.
BRIEF DESCRIPTION OF THE PRESENT INVENTION
As part of the process disclosed herein, it
encompasses a method of usinq an ebullating bed reactor and
supported catalysts, as well as a process for treatment of
the especially refractory petroleum residues, for example
36348/36357
~33~7~
those boiling over 850F and especially asphaltenes or the
ORA fractions, that is a mixture of oil, resin and
asphaltenes left over from the petroleum refining and
characterized as boiling at 1000F and higher according to
~STM D-1180 and D-1160 method. That aspect of the invention
represents a further more successful conversion of the
especially refractory bottom or residual products.
A further aspect of the invention is to provide a
further contribution to the process briefly described
immediately above. That aspect includes a more facile and
advantageous treatment of petroleum or crude oil, as such,
as a starting feed as well as the less completely refined
petroleum or crude oil residues, such that when treating
crude oil, substantially no residue rem~ins and the starting
material is converted substantially completely to lower
boiling point products. This result is accomplished
because, according to my process, the severe treatment ~hich
produces conventional residua is not applied to the crude
oil.
By a speci~ied boiling point, as discussed herein,
is meant a boiling point such that the products will be
substantially completely distilled at that boiling point
(according to the defined methods) including up to that
temperature, and no suhstantial amount of residue remains
unless so speci~led.
36348/3h357
~2~33777
BRIEF DESCRIPTION OF THE INVENTION
The present process distinguishes from the prior
art in the specific catalyst species which are being used
preferably in a supported form, to attack the extremely
refractory petroleum residue components such as boiling over
1,000F and usefully convert these residue components into
highly desirable lighter viscosity products in a highly
efficacious manner.
My present process further distinguishes from the
prior art in that the reaction is specifically attacking the
most refractory components of the residue such as
asphaltenes, while at the same time avoiding, due to the
discovery of the catalytically aided thermal decomposition
by the herein used catalysts, the un~avora~le effects of
coke formation which may occur if the process is not carried
out properly. Moreover, further process advantages are
realized by the employment of an ehullient hed reactor, by
carrying out the process continuously and at higher
temperatures such as up to 650C, yet at the same time
quenching the conversion product to ohtain a desired end
product in another reactor(s) in comhination with the first
reactor.
In accordance with the present process, the O~A
fraction obtained from the residues such as from petroleum
refining or refining of any other carbonaceous source
36348/3h357
~23377~
yielding these fractions are being treated advantageously by
the present methodO ~owever, with outstanding results the
present process is useful for treatment of asphaltenes, e.g.
of solvent extracted asphaltenes when these are being
treated in one or two stages to further cleave and/or
hydrogenate in one or two stages this especially refractory
product and thus to improve the overall yield obtainable
from a barrel of oil.
This improved process is also characterized by the
ability of the specific catalyst to convert the Ramsbottom
or Conradsen carbon into usefully hydrogenated products
without affecting, to any noticeable degree, the process as
practiced herein.
Still further, the present process also provides
for especially advantageous catalyst support combinations
which can be used such as in an ebullient bed reactor and
produce the conversion products in an especially
advantageous manner without the catalyst support combination
being affected by the unwanted metal constituents found in
petroleum and accumulated predominantly in the ORA fraction.
Still further, it has been found that after the
first stage conversion with the active catalysts, a second
quench stage may be provided where the cleaved product may
be appropriately tailored by a specifically selected
catalyst in the second reaction stage to produce the
36348/36357
predetermined and/or highly desired product cuts. ~owever,
the second stage reaction is interdependent and based on the
specific catalyst in the first stage, and based on the
properly carried out reaction in the first stage reactor.
Moreover, the present process allows also a
recycle of the product not adequately reacted in the first
reactor and its subsequent conversion into the desired end
product.
Also, it has surprisingly been found that my
process can be further improved by the parallel treatment of
the various products first recovered from the ebullating or
a fluidized bed reactor and split in at least two streams as
further described herein. It has been found surprisingly
that the split stream treatment procedure (which is a
straight through treatment for the first gaseous or va~or
phase, i.e. top products), for the bottoms thereof which are
obtained by collection in a further separating device such
as cyclone, contributes further to the overall conversion of
the starting material by this separate treatment of the
heavy fractions carried over such as from the ebullating
bed. This is due, in part, to the application of different
catalyst combinations to overcome some shortcomings of the
catalyst compositions when these are used to treat the
highly diverse components of the original feed or the
recycle thereofO Moreover, the bottoms being pre-treated to
.~ ~
,,
'
7~
g
a certain degree (e.g.) as characterized by lower API
values) benefit from this parallel treatment.
This separate treatment in at least one
parallel stream of a smaller, yet initially upgraded
S fraction coming over from, e.g. an ebullating bed
reactor, provides a number of benefits. The catalyst
movement is about 2 to 40 inches/second/10 to 30
minutes of residence time of particle feed with about
17 inches/second/20 minutes of residence time of
particle of feed being average. (The catalyst
movement is defined in accordance with ebullating bed
technology.) Additional ebullation is caused by
steam and feed being fed into the reactor. At the
same time this treatment makes it easier to obtain
improved throughput rates, yields, conversions, type
of products, etc., for the top products from the
first stage such that the overall flexibility of the
process for treating a wide variety of feed is
remarkably aided by this parallel treatment as
disclosed herein, especially allowing the split
proportions to be tailored to accommodate the various
feeds with different component fractions, as well as
the mixed product feeds of various source materials
which are especially difficult to treat.
That aspect of the process where the heavy
bottom fractions may be recycled but without parallel
treatment, the results are outstanding, but further
improvememt has now
363~8/36357
~:33~7~ ,
-- 10 --
been found contrary thereto. That aspect of the process is
called a ~recylce method~. The improved aspect is called a
"parallel methodn. Thus, a separate treatment without
recycling is more advantageous in the overall improvement of
the products and product yields to be obtained from the wide
variety of heavy crude oils, petroleum refining fractions,
and the especially refractory component parts. In fact, the
parallel method as disclosed herein causes almost an
entirely complete recovery of all crude oil components as
useful products.
The above process variations are characterized by
and their accomplishments made evident from the improved
product being obtained from the very poor starting material,
the higher hydrogen content of the starting material, the
lower viscosity and smaller molecular size of the cleaved
product and the amena~ility of the treated product to
further conventional treatment ~teps.
Inasmuch as the present process accomplishes
considerable hydrogenation which improves the yieldsr
improves the product, and provides smaller molecules and
thus a less viscous product and eliminates and/or minimizes
to a very significant degree the presence of free Conradsen
or Ramsbottom carbon, the attained end result shows an
outstanding àchievement in the continuous search for
utilizing all fractions of a refinery product which
`:
..: .
;
36348/36357
33~
heretofore could not be advantageously upgraded to the
degree such as now disclosed herein.
The process as disclosed herein, i.e. consistinq
of the "recycle method" and "parallel method", is, thus,
especially useful because the petroleum fractions which
previously were reacted according to prior art under very
severe conditions can be reacted by means of the disclosed
catalysts in a less severe, straight through fashion, and
either a recycle or in a parallel method. The very
refractory bottom products (by conventional practice~,
constituting a smaller part of the total mixture and which
are the hard to treat, are either pre-treated and recycled
or, instead of being initially more drastically attacked,
e.g. bottom reboiled and/or recycled, can now be treated in
a parallel treatment with the necessary degree of severity.
This treatment severity for the recycle or parallel method
is still markedly and significantly less than that necessary
in the prior art, but accomplishes far better results as
characterized by more valuable product components, more
product recovery and simpler recovery steps of components
and other beneficial features further eYplained herein.
Further, the process as disclosed herein also
applies to other products such as Gilsonite (also known as
uintaite, a black lustrous asphalt occurring in Utah), pitch
(natural and/or refinery product), tars such as from tar
36348/36357
~337~
- 12 -
pits found in California, asphalts (natural) such as found
in asphalt lakes in Trinidad and of the type found in Utah
and like products found in nature or obtained after refinery
treatments which concentrate oils, resins and asphaltenes in
the resldues; slurry oil found in refineries which oil is
often without a distillation point; bottom oil from coal
gasification plants, e.g. solvent extraction products of
coal, or coal extracts such as shown by Satrien, above, p.
206, or bottom fractions of retorted shale oils; heavy
bottoms from the Lurgi gasifier process which are produced
in operation of the SASOL process in South Africa; delayed
cocking products obtained during bottom distillation of
these products; resins both hard and soft and asphaltenes.
It is inadvisable to mix some of the feeds, but that can be
determined by experiment at the actual reaction conditions.
By heavy crude oils is meant crude oils of an API number of
about less than 22 including the minus number API crudes.
The combination process as disclosed herein
provides further that the ~ore;volatile fractions from the
petroleum and the like hydrocarbon can be treate~ in a
straight through fashion, but the heavier bottoms, that is
the heavier components, especially the more refractory
components, are most advantageously treated in a recycle or
parallel run, depending on the severity of treatment
required. For example, the feeds requiring the most severe
,,
.
~337~ 36348/36357
treatment are best treated by the parallel method~ ~ence,
the method most suitable for the feed is selected which will
not impede or detract from the overall, first reactor stream
yields, and yet improve the recovered bottom products by a
recylce or in a parallel treatment sequence for improved
overall yields.
While the "parallel method" aspect has been
disclosed with respect to a two reaction train or two stream
parallel treatment, equally applicable are further parallel
treatments where the refractory products obtained from the
second reaction train or stream may be further reacted
without recycling.
As a result of the above disclosed combination of
steps, the present invention provides for advantageous
space-time velocities, yields on an absolute basis, yields
per respective parallel pass or combined yields.
As an example for a parallel treatment, the
products obtained from a mixed heavy bottom fraction of
especially refractory products have yielded end products
having a X factor of about 11.71, i.e. highly paraffinnic
products, which is achieved by the ~reater ease the tops
from the first reactor are subsequently treated, e.g. to
reduce the bromine number and the like.
. . .
:
36348/36357
33~7~7
I~LUSTRATION OF THE PROCESS
BY A FLOW SHEE~ AND DESCRIPTION O~ TH~ DRAWINGS
In the drawings herein:
Figure 1 illustrates schematically the recycle
method in an ebullating bed reactor, and
Figure 2 shows schematically a continuous process
wherein by the parallel method the various source materials
are converted into useful products.
As illustrated in Figure 1, the first stage
reactor is an ebullating bed reactor 10. It consists of a
reaction vessel 11 with a catch funnel 15 at the top and a
pump 14 for recirculating the fluid 8 products undergoing
the reaction. The catalyst 9 which is in a dispersed-
supported form is ebullating with the fluid 8. A fluid 8
undergoing the reaction including the catalyst 9 therefor
may be made to circulate by the pumping of the fluid into
the reactor and appropriately distributing the same.
Typically, the circulating fluid might be introduced at the
bottom, but introduction may be elsewhere in the reactor at
one or more places. On a smaller scale, and as a close
approximation of an ebullating bed reactor, a stirred tank
is appropriate, provided the supported catalyst material is
placed in a cage(s) or baskets such as four stainless steel
mesh envelopes and these attached to a suitable frame driven
by an external motor. By varying the speed of the rotation,
36348~36357
~%3~1t7~
e.g. 20 to 180 rpm, more typically from 50 to 15~ rpm,
reactions very closely approximating those in an ebullient
bed reactor are achieved. Again, it is important that
adequate steam-catalyst-fluid contact take place to assure
the desired result.
Circulatory fluidized bed reactors where the
supported catalyst circulates with the fluid or fluidized
bed reactors may also be suitable. Simllarly, a fixed bed
reactor with the fluid downflowing or upflowing may be used
for that purpose.
A continuously introduced pre-heated feed charge
such as asphaltenes or an ORA cut are introduced via pipe
12. Steam 13 may be introduced with the feed or it may also
be distributed in reactor 10 throughout the fluid from the
bottom of the reactor 10.
The obtained lighter viscosity fluids are
conducted by a large diameter type or conduit 15a and
appropriately cooled if needed in a quench zone 1~ and then
introduced in a second stage reactor 17. Although one
reactor has been shown in the Figure, a number of reactors
in series or in parallel may also be used. These reactors
may be fixed catalyst bed, gaseous or vapor-phase reactors.
A column where the catalyst is shown as supported on the
trays 17a is typical. Other well known devices may be
employed for this purpose such as trickle bed reactors and
36348/36357
- 16 -
the likeD The bottoms from the second reaction stage are
collected in the collection zone 18. These are classified
as No. 1 bottoms. These bottoms may be recirculated
entirely or partially into the ebullating reactor with the
fluid collected by the ebullating bed funnel conduit 15. If
necessary, a pump 14A may be used for that purpose. Part of
the product may also be diverted and recovered for further
processing. The top fraction from the second stage reactor
17 may be refluxed via reflux boiler 19 and the products may
be diverted from this reflux boiler and the gaseous products
therefrom further worked up in a second reactor such as
depicted for 17 but now shown herein; these are all called
second stage reactors as distinguished from the first stage
reactor 10 where the cleavage of the source material is
undergoing. The second stage reactor products and other
gaseous products may be treated in a further reactor or such
as by bubbling through an appropriate bath to remove any
unwanted constituents such as hydrogen sulfide. This may be
accomplished in a vessel 20 in which potassium hydroxide has
been dissolved.
; The gaseous products are thereafter recovèred in a
conventional manner.
As illustrated in Figure 2, the first stage
reactor 110 is an ebullating bed reactor. It may also be a
stirred basket or cage type reactor or a fluidized bed
'
36348/36357
~233'77~7
reactor. The rotating~ i.e. stirred cages or baskets, have
not been shown herein. ~owever, inasmuch as the ebullating
bed and stirred cage or basket type reactors are very
closely related in their actual behavior and performance,
these will be discussed as being the same.
The reactor 11~ consists of reaction vessel 111
with a catch means such as a funnel 115 at the top and a
pump 114 for recirculating the fluid 108 products under~oing
the reaction. The feed or fluid 108 may be petroleum or
light petroleum residues boiling below ahout 8~0F, but may
also be the ORA (Oil, resin, asphaltene) boiling above that
temperature, e.g. above +1,050F, i.e. having no boiling
point. A catalyst 109 which is in a dispersed, supported
form is ebullating with the fluid 108. A screen lO9a
prevents the catalyst from upflowing with the ebullating
liquid 108. However, the fluid undergoing the reaction
including the catalyst therefor may or may not circulate by
the pumping into the reactor through pump 114 because the
catalyst may stay suspended in the fluid (for circulating
fluid hed reactors) or only a li~uid phase freed of the
catalyst may be caught in the funnel 11~. As shown in the
drawing, the ebullating fluid pumped through pump 114 may be
introduced anywhere in the reactor, although typically it is
introduced near or at the bottom thereof.
The feed 108, of the above-described type, is
~ -- .
:,
:
3634~/36357
~2337~
- 18 -
joined in a single introduction port with steam and/or water
113 or the introduction ports for either o~ these may be
circumferentially around the vessel 111. The refractory
components of petroleum, having a longer residence time,
undergo the catalyzed reaction and are carried over by the
exit conduit 115a. This conduit then leads directly into a
cyclone 11~ which separate the vaporous phase products from
the liquid phase and/or entrained products and products
carried over with steam from the ebullating bed. More than
one cyclone 116 may be used. The cyclone bottom products
112a are treated in a separate parallel stream.
The top products coming from the cyclones 11~ may
be quenched in a quenching zone 116a as shown in the
drawing, and thereafter immediately reacted in a reaction
vessel 117. The products may not be quenched prior to
introduction into reaction vessel 117 as quenching may take
place therein. The reason for auenching will be explained
laterO
After the intro~uction into vessel 117, an
appropriate catalyst, as further described herein, supported
or unsupported, is suspended, such as schematically shown,
on tray 117a or by other means. The catalyst will cause the
reaction ;n the gaseous or liquid stream to produce again a
lighter fraction and a heavier fraction or the gaseous steam
may be hydrogenated. These hottom productc may be recovered
363~8/36357
~33~
-- 19 --
such as in a vessel 118, and further treated in a
conventional fashion or, as disclosed further herein, may be
treated separately in still another parallel stream.
Although not shown in the drawings, additional
reaction vessels in the same stream of a type such as 117
may be provided. The final recovery of the gases from
reflux condenser 119 or other condensers (not shown) may be
accomplished in the conventional fashion such as by cooling
and/or low temperature chilling and/or pressure condensation
and need not be discussed.
It has been found advantageous that the gaseous
products of the initial straight through product, after
treatment in vessel 117, be further treated at a temperature
such as at about 130C to about 240C in a reaction vessel
of the type as shown in the figure at 117. If more
saturated products are desired, a reflux column 119,
appropriately selected and adjusted, may be employed to
collect these products.
The bottoms 112a, from the reaction product
separated such as in the cylones 11~, are generally the more
refractory components of the petroleum or petroleum
products. These products 112a, although believed to be
partlally reacted in reactor 110 (e~g. as characterized by
the improved API number), are best further treated in a
first parallel pass. As these bottom products have reacted
36348/36357
~L~33777
- 20 -
to a certain degree due to the catalyst 109 suspended in the
ebullating bed, the further reaction is best carried out
with another specific catalyst appropriately modified, if
necessary, to suit the particular feed.
This second feed stream is thus the bottoms 112a,
fed to a reaction vessel 130 conjointly with steam, and this
co-fed stream lends itsel~ to a catalytic treatment of the
still hot product. The feed 112a is more amenable to
hydrogenation and thus further improvements are better
achieved for obtaining lighter viscosity components of the
heavier fractions of petroleum. Feed stream products 112a
may be advantageously treated with a different catalyst
composition in the presence of ~team in the same ratio
ranges for steam as for the feed fed to the first reactor
111. Various catalyst compositions have been found to he
more advantageous in the subsequent reactions in contrast
from the catalyst which has been originally used in the
ebullating bed reactor and designated as 109 therein. The
use of a different catalyst is due to the partial reaction
of the more refractory components in vessel 111. Again, the
different catalyst composition for treatment of the variou~
bottom fractions will be further discussed herein.
As the reaction vessel 13n may still be followed
with another reaction vessel (not shown), the top products
recovered in reflux column 133, from reaction vessel 130,
~33~7~ 36348/36357
- 21 -
may be further treated. Conventionally, the overhead
products from reflex column 133 may be scru~bed in vessel
135 to remove unwanted constituents, e~g. H2S and the
gaseous products thereafter recovered and reused. ~he
reaction vessel 130 may be used in a similar manner as
reaction vessel 117, i.e. as an upflow or do~7nflow reactor,
with a supported or unsupported catalyst suspended on a bed
or on trays 130a, or as a fixed bed catalyst. A liquid
catalyst, such as hydrated melts of the various alkali
sulfides, liquid at the temperature at which these are desired
as previously disclosed by me, e.y. in my U.S. Patent
4,366,044 may also be employed in the subsequent reactors.
If no further reduction of the molecular size of
the overhead distillate product from the first reactor 110,
is desired, the overhead distillate product should be
hydrotreated in vapor phase reactors, immediately, as it
emerges from the cyclone 116 which separates the bottoms
(the heavy, partially reacted o~ unconverted recycle stock),
from the overhead distillate which is the product.
Quenching such as in quencher 116a prevents recondensation
of the cleaved components. Hydrotreating is aided by the
absence of the heavier fractions (removed by cyclone 116).
By definition, the hydrotreatlng is done in the absence of
either thermal or catalytic cracking and includes the
addition of hydrogen to a molecule. As the hydrogenation
~ . . .
: .,
~: :
36348/36357
~337~
- 22 -
reaction is exothermic, the previously described quencher
116a helps to control the temperature and prevents product
damage if excessive temperatures are encountere~.
Fixed bed reactors may be used for this
hydrotreating, or liquid catalyst reactors may be used.
~owever, at present, the liquid catalyst reactor appears
most effective at temperatures below ~80C, due to the
solidification of the catalyst above 280C, as a result of
the interconversion of the hydrates.
~ he fixed bed reactors are less efficient in terms
of the degree of hydrotreating achieved from the liquid
reactors.
Turning now to the description of the feed,
typically an ORA fraction, that is oil, resin, asphaltene
fractiont is described as one which boils at atmospheric
pressure at a temperature above 1000F. Although this is a
rough description because the amount of oil, resin and
asphaltene are not necessarily ascertained, it is a
convenient measure for this the most refractory component in
an oil. It is known that asphaltenes can be solvent
extracted from petroleum residues or from the ORA fraction
to allow the oil and resin residue to be further treated.
Asphaltenes, o~ course, are especially intractable to
further treatment such as by catalytic or other means, and
thus constitute a fraction which can only be usefully burned
....
36348/36357
~233~
or coked by the prior art methods to strip all available
hydrogen therefrom.
A convenient characterization of the asphaltene is
that it is that portion of the asphalt or bitumen which is
soluble in carbon disul~lde, but insoluble in paraffins,
e.g. heptane, paraffin oil, or in ether. Resins from the
ORA fraction may be extracted with propane. Bitumens are
also soluble in carbon disulfide. Carhenes, which are
constituents of bitumen, are insoluble in carbon
tetrachloride but soluble in carbon disulfide. Further, the
oily or soft constituent of bitumen is also named malthenes
or maltenes. These are soluhle in petroleum spirits.
Malthenes are pentane soluble compounds and asphaltenes are
pentane insoluble compounds.
Still further and in a broader sense, the natural
asphalts such as petrolene, mineral pitch, earth pitch,
Trinidad pi'ch, petroleum pitch, and native mixtures of
hydrocarbons such as amorphous solid or semi-solid fractions
produce~ by oxidatlon of residual oils are included within
the above definition.
Inasmuch as t~.ere is no a~reement on the exact
deflnition of these compounds such as malthenes or
asphaltenes, mixtures are often reported as one or the other
in the prior art. Moreover, the solvents used and the
extraction and precipitatlon techniques practiced affect to
3~348/36357
~23377~7
; - 24 -
a lesser or greater degree the end product properties. For
this reason, the solvent extracted asphaltenes such as
~ carbon disulfide extracted asphaltenes precipitated from
- heptane are still not considered pure compounds as these
have no specific melting points but only softening points.
Asphaltene softening points may be up to 400F and higher.
As a result, a convenient measure is to define the ORA
fraction as one boiling at 10U0F and higher, although this
temperature limit is arbitrary and lower temperatures such
as 900F may be selected because all of the material may not
be desirably stripped away. Hence, a lower temperature of
800F merely characterizes a less intractahle composition.
For example, a high softening point solvent
extracted asphaltene will have a softening te~perature of
about 270F, a specific gravity at 60F of ahout 1.1149 and
~; a viscosity at about 275F of about 4060 poises. The
specific gravity at 275F is 1.026, and thus the viscosity
is 3,957 stokes or 395,700 c~T (centistokes) (stokes are
obtained by dividing the poise~ by a specific gravity at the
indicated temperature). Viscosity at 3nooF of the same high
softening point asphaltene is 877 poises with a specific
gravity at 300F of 1.016, and 86, oon cST. Viscosity at
325F for this asphaltene is 2fil.5 poises, and specific
gravity at 335F is 1.006 giving 26,000 cS~.
The analysis for the above solvent extracted
. . .
`:
.
3634~/36357
~2337~7
- 25 -
asphaltene is found in the following table.
ASPHALTENE ANALYSIS
Metals
Carbon 84.59% Fe 360 ppm
Hydrogen 8.80% Ni 147 ppm
Nitrogen 0.82% V 490 ppm
Sulfur 5.52% Na 4~7 ppm
Ash 0.27% K 4 ppm
Moisture 0.0%
Oxygen
Total 100.00~
B.T.U. content of asphaltene: 17,627
'
Although the above asphaltenes may be considered as
representative, various other asphaltenes, depending on the
source, may have different characteristics.
Based on the above product analysis, it is seen
that these products contain considerable amounts of metals.
These amounts vary based on the source of the material and
may range up to 6,000 parts per million (ppm) of vanadium,
but typically up to about 600 ppm. Nickel and other
constituents may also be present up to about the last named
amount. Consequently, these metals also affect the ability
of the residue to be treated by conventional methods of
..
;: :
:
'`
363~8/36357
~Z33~7~7
petroleum residue treatment.
Based on the various analyses, typically the
hydrogen content of the ORA fraction may range from 13.5% to
about 7%, and lower by weight, but again this is not a
precise characterization. In a petroleum residue boiling
over l,000F, the hydrogen content will be about 12.5% and
lower. A considerable percentage of "free" carbon is also
found (as Conradsen carbon), e.g. up to 45% by weight. The
free carbon is defined as Conradsen carbon or Ramsbottom
carbon, but these analyses are not identical because
different methods are used to define the the "free" carbon
which, in fact, may not be "free". In the ORA fraction,
Conradsen carbon may range up to 40+%, by weight. In any
event, the carbon residue is amenable to conversion
according to the process as disclosed herein.
Any cracking, either thermal or catalytic,
produces smaller molecules than were present in the
feedstock being cracked. A smaller molecule will have a
lower temperature boiling poin~ than a larger molecule.
Hydrotreating appears not to produce cracking (and thereby
form smaller molecules). The boiling point temperature
range seems to be approximately the same before and after
hydrotreating.
~ ydrotreating is effective, for example, in the
hydrogenation of olefins to paraffins, and in the
~ ,
'`
36348/36357
~ ~3377~
- 27 -
elimination of diolefins and triolefins in the product by
hydrogenating these diolefins and triolefins to either mono-
olefins or paraffins. Aromatics are also converted to
naphthenes by hydrotreating. True paraffins appear not to
show a rea~tion during hydrotreating.
~ ydrotreating may also reduce the sulfur and
nitrogen content of the product. However, if the
asphaltenes, which contain most of the nitrogen, have not
previously been cracked, the nitrogen of said asphaltenes is
not available for removal by hydrotreating. The degree of
sulfur removal and nitrogen removal is e~uivalent to the
hydrotreating of an unsaturated bond (with no hydrogen
present at the position at which the hydrogen is to be
inserted into the molecule). Those sulfur and nitrogen
atoms, in an exposed position in the chemical structure of
the product, and which nitrogen or sulfur atoms can be
replaced by hydrogen, in a thermodynamically favorable
exchange, with the formation of ammonia or hydrogen sulfide
and the formation of more saturated molecule of the product,
are the sulfur and nitrogen compounds that may be removed by
hydrotreating.
The bromine nu~ber is a measure of the degree of
unsaturation of the product. The lighter unsaturated
product cuts will have the highest bromine numbers. The
hydrotreating of the product will bring the bromine number
, ,
`~: ' ' ,,
~ .
363~8/36357
~2337~7
- 28 -
down to acceptable levels, e.g. a bromine number of 40-50
for the initial boiling point to 400F product (obtained by
distillation within that range of the product by the
method(s) disclosed above).
As previously mentioned, the bottom products 31
may form a still separate parallel path and may be used as a
feed, but these can likewise be recycled because the further
reaction pass only introduced further complications.
Nevertheless, if the bottom products 31 are difficult to
treat, by a further spli~, these may be sub~ected to still
another treatment stream of the type as shown for feed 12a,
and specifically in Figure 2 herein.
Inasmuch as most of the overhead products from
reaction vessel 11 are treated in one path straight through,
the bottom fractions 12a such as collected in cyclone 1~ or
in the collection vessel 18, constitute products, but of a
sma~ler proportion depending on the component composition of
the petroleum. These bottom products may be thus treated in
'1er vessels based on the proportion o ~he components of
the petroleum products heing treated. Accordingly, various
mixtures of petroleum can now be usefully treated depending
on the bottom product 12a components. The size of the
reactor 30 may range in proportion to the bottom components
in the petroleum such as those boiling at a temperature of
1000F and above.
..
36348/3~357
~233~7~
- 29 -
Steam is being introduced in the reactors 110 or
130 at a rate such that the product sought to be obtained
dictates, to a certain extent, the amount of steam being
used. The steam may be from about 50 and up to 100 to 130%
of the product recovered, by weight. ~team may also be
expxessed on a basis defining a lower limit, namely such
that coking does not occur to any suhstantial degree due to
absence of steam, but which would occur in the reactors 110
or 130 if insufficient steam were present. Thus as a lower
limit, the amount of steam introduced must be such that any
significant coking is avoided in reactors 110, 117, 130,
etc. Additional steam may be supplemented such as for
reactors 117, 130, etc. Steam is low pressure, typically
waste steam.
The temperature in the reactor generally may be up
to 650C, although the most advantageous operating
temperature is below 450C, such as below 425C. The
temperature, however, in the subsequent reaction vessels,
e.g. 117, etc., may be lower, and lower temperatures seem to
produce a more hydrogenated product, such as characterized
by a lower bromine number. ~owever, the more refractory
component treatment in reactor 13~ may require the same
temperature as in reactor 110.
The quenching may be carried out at a temperature
of 390C and lower. The third stage reactor, not shcwn
~...~.
:
36348/36357
~Z33'77'7
~ 30 -
of 390C and lower. The third stage reactor, not shown
herein, may have a temperature as low as in the range of
170C and lower, i.e. down to 125C (but without
condensation of steam). Lower temperatures are desirahle
because hydrogenation is exothermic and most of the
hydrogenation is carried out in the downstream reactors,
whereas most of the cracking is carried out in reactor 110,
and to a lesser degree in reactor 130.
The pressure in the vessel 111 and downstream
thereof may be maintained at an atmospheric pressure.
Although subatmospheric pressures are possible, these are
not as convenient. If the pressure is between about 100
psig or below 150 psig (although up to 25 atm. may be used),
it seems that the reaction may be improved. When higher
pressures are used, between 60 psig and below 150 psig for
some unknown reason suggest the best range. Steam
condensation must be a~oided at the higher pressures.
Although the exact reason is unknown and the conversion is
not understood, it seems to indicate a better hydrogenation
in that range, but the increase in pressure carries with it
a certain trade-off due to more complicated equipment and
more capital investment.
Further, instead of cyclone separation such as
depicted in the drawing herein, i.e. one or more of the
cyclone 116 being employed, other separation means may be
`:
~; .
~33~77 36348/36357
- 31 -
employed to recover the liquid and/or entrained products and
thus to accomplish the parallel path treatment of the
petroleum components; such other means may be centrifuge
separators, knock-out drums or the like.
With respect to the materials which may be
subjected to the treatment, these are various crude oils
from whatever source and of whatever viscosity. These crude
oils, as long as these are adequately liquid at the
treatment temperature, may be treated and produce the
outstanding results as disclosed herein. Atmospheric
residues and bottoms at atmospheric conditions obtained rom
conventional petroleum refining may also be included. These
source materials thus are characterized as products which
have a boiling fraction of less than about 850F. However,
their component parts or reaction products ~from reactor
110) which have a boiling point above 850F are best treated
in a parallel path as disclosed herein.
The further treatment in the second parallel
reaction path will produce products of outstanding and
improved viscosity, improved hydrogen content and improved
chemical structure (e.g. products are of a hi~her
paraffinnic content with high K values, or aromatic
polycyclic compound which can still thereafter subsequently
be treated to produce the desirable, less olefinically
unsaturated cyclic compounds and linear or hranched
~233~7~ 36348/36357
- 32 -
paraffinnic products). Hence, the gaseous products obtained
by practicing the process herein and the conversion of the
unsaturated products to more saturated products is a further
contribution to my process previously described.
In hydrotreating the products of th i5 process, in
order to reduce the bromine number and to further reduce the
nitrogen and sulfur content of said products, it is
necessary to keep in mind that the lower temperature
hydrotreating aspect does not appear to "crack" the product
to any significant degree and thereby form smaller
molecules, which in turn would reduce the temperature of the
boiling point range. Hydrotreating of the product will
render further upgrading, i.e. by cracking (as determined by
a lower temperature boiling point range such as the
formation of smaller average molecular sizes), more
difficult due to the apparent ina~ility in this process to
crack paraff)ns.
The hydrotreating should, therefore, be the final
part of the process as disclosed herein and not be used
until the desired molecular size has been achieved by the
"cracking" aspect of this process~ Thus the vacuum resi~ua
feedstocks are initially cracked into smaller molecular
size, under less severe conditions, and then the refractory
residues, i.e. feed stream 112a, under the necessary but
appropriately severe conditions which do not affect the
. , :
. ~
`
~233~7'7
-33-
mainstream products. The resultant product has a
lowered temperature boiling point range. Accom-
panying this molecular size reduction is a reduction
in the sulfur and nitrogen content of the top product
compared to that of the feedstock.
The resulting smaller molecules which form
the "cracked" products do not contain the high
amounts of nitrogen of the feedstock. Nitrogen is
normally concentrated in the heavy "asphaltene" end
of the petroleum residua. Metals are bound through
the nitrogen bond to the methylene groupings of the
porphyrin structure of the asphaltenes, and are thus
usefully separated in the parallel reaction path.
Therefore, the metals and the nitrogen remain with
the unreacted central porphyrin structure of the
asphaltenes and are either not presen-t or present in
very limited quantity in the overhead distillate
product derived from the vacuum residua containing
the asphaltenes, which have been concentrated rela-
tive -to the asphaltene content of the crude which
formed said vacuum residua. The metals and the
nitrogen are more concentrated in the No. 1 bottoms,
e.g. 112a, which do not form overhead distillate
product. The sulfur content is not necessarily
concentrated in the asphaltene portion of the petro-
leum crude or residuum. The resin portion of the
residue may contain as much as S~, by weight, of
sulfur.
~23~ 363~8/3~357
- 34 -
; With respect to the catalysts which are sometlmes
also called reagents, these are employea in the form such
that these produce the deslred results, namely sufficient
cleavage and/or scision of the bonds in the first and
further parallel stream reactors or the hydrogenatlon
ultlmately deslred. The results are obtained by first
appropriately severely, but not overly severely, treatlnq
the first treated product with the further herein described
catalysts, and these will be described with reference to the
results achieved. ~econdly, the pre-treated (in reactor
110) refractory components are thereafter appropriately
severely treated in the parallel path. Mixtures of the
catalysts are also suggested as highly des~rable for the
crude oils and their residues, or for mixtures of crude oils
based on the various crude oil component fractlons.
A suitable severe catalyst ~or reactor 110 is
designated as Catalyst A, and it is prepare~ as follows. A
mole of potasslum hydroxide is dissolved in either ethanol,
methanol, particularly best in ethanol, or an ethanol-
methanol mixture, or less advantageously, because of ~esser
solubility therein, in l-propanol or l-butanol. ~olubility
of the catalyst product is lower in these last two and
larger amounts of the alkanol must be used and subsequently
separated. The alkanols may be absolute alkanols, although
these may be such as 95% ethanol. The potassium hydroxide
.
:
36348/36357
~L;;~3377~7
is dissolved in this solution and is then reacted with
hydrogen sulfide bubbled through the solution. After
thorough saturation, the catalyst-alcohol mixture is
recovered and the alcohol separated by vacuum aspiration.
The residue is the catalyst. For a mixture of solvents,
typically one mole of potassium hydroxide is dissolved in
200cc of ethanol and 130cc of methanol. Typically,
analytical reagent grade pellets of potasslum hydroxide
(about 86% KOH), absolute ethanol or 95% ethanol and
absolute methanol are used. As mentioned above, the
proportions of an ethanol-methanol mixture may be varied.
The solution is evaporated under vacuum untll no more
residual alkanol can be removed. These catalysts appear to
be best suited for cleavage, and especially for cleavage of
the refractory components and, therefore, are used in the
appropriate proportions (or solely) with others to achieve
the deslred cleavage.
Catalyst may be unsupported in the first stage
reactlon, but most advantageously for the first or
subsequent stages it is deposited on supports and calcined.
Turning now to the catalyst supports which have
been employed, these have been employed mostly for the
purpose to obtain increased surface area. The catalyst
supports are spinels and such as chromite spinel ~CrO) and,
most advantageously, porous metal, i.e. stainless steel of
36348/36357
3~7
- 36 -
the available AISI grades, and the like. The last are
obtained by sintering very fine sized uniform, powdered
metallurgy particles or are produced as thin plates obtained
by leaching out leachable constituents in the thin, te.g.
one eighth of an inch) metal plate, providing thereby
intercommunicating passages. Other metal supports are such
as are obtained by sintering very fine wires, about 0.2 to 5
mm thick, and cutting these to length, e.g. 2 to 5 mm.
Still other supports are such as alumina with sizes of the
o o
pores ranging from 50A to 350 and even up to 1,000A, but
these may need to be protecte~ as further explained herein.
Although the treatment in subsequent reactors may be less
demanding based on support characteristics, the treatment in
the original first stage reactor in accordance with this
invention is best carried out with a strong, inert support
such as the porous metal supports which have a size range of
the pores, e.g. up to 3,500A and larger, i.e. the metal may
be from 10~ metal and 90~ the pores, by volume, although
metal may be up to ahout 25%, py volume.
In all the reactions the catalyst is allowed to
react with the exclusion of atmospheric oxygen and thus in
absence of oxygen. Similarly, the deposltion of the
catalyst on the support is in absence of oxygen as is the
driving of~ of the volatiles from the support.
The catalyst and the support, after the volatiles
: ` ,
36348/3~357
~233~77~
- 37 -
the available AISI graaes, and the like. The last are
obtained by sintering very fine sized uniform, powdered
metallurgy particles or are produced as thin plates obtained
~y leaching out leachable constituents in the thin, ~e.g.
one eighth of an inch) metal plate, providing thereby
intercommunicating passages. Other metal supports are such
as are obtained by sintering very fine wires, ahout 0.2 to 5
mm thick, and cutting these to length, e.g. 2 to 5 mm.
Still other supports are such as alumina with sizes of the
o o
pores ranging from 50A to 350 and even up to 1,000A, but
these may need to be protected as further explained herein.
Although the treatment in subse~uent reactors may be less
demanding based on support characteristics, the treatment in
the original first stage reactor in accordance with this
invention is best carried out with a strong, inert support
such as the porous metal supports which have a size ranqe of
the pores, e.g. up to 3,500A and larger, i.e. the metal may
be from 10% metal and 90% the pores, by volume, although
metal may be up to about 25%, ~y volume.
In all the reactions the catalyst is allowed to
react with the exclusion of atmospheric oxygen and thus in
absence of oxygen. Similarly, the deposition of the
catalyst on the support is in absence of oxygen as is the
driving off of the volatiles from the support.
The catalyst and the support, after the volatiles
,,
:
:
.
3~348/36357
~2; 33777
- 38 -
have been driven off, are heated to an appropriate
temperature such as between 320C and up to 450C or even up
to 550C. The catalyst tightly adheres to the support and
may thus be used such as in a spinning-cage (also called
spinning basket) reactor, ebullating bed or fluidized bed
reactor.
If the support is unduly attacked by the catalyst,
such as alumina in the first stage reactor, then the
following method is used. The above catalyst is evaporated
to considerable dryness, dissolved in alycerol, and the
glycerol-catalyst mixture deposited such as on an alumina
support. Other less resistant supports to the attack hy the
catalyst such as a molecular sieve supports are treated
similarly. Typically these molecular sieves may be of the Y
and X, e.g. YL-82, type, with low sodium content (availa~le
from Union Carbide, Danbury, CT. or comparable supports from
Mobil Oil, New York, NY). The molecular sieves function,
however, as supports for the catalyst, i.e. to increase the
contact area for the catalyst.
The glycerol catalyst mixture after depositing on
the reagent is then progressively heatefl such as up to 5~0C
th the vo~atiles being driven off.
Glycerol may also be first fleposited, heated up to
about 200C, and then the catalyst deposlted thereon 5after
the support has been cooled), and then heated to the flesirefl
,
~: :
. ~ .
36348/36357
33~
- 39 -
temperature.
The reaction in the first reactor may be at a
higher temperature, and may range from about 320C to about
450C although temperatures up to 560 have been used, even
up to 650C. For asphaltenes, the preferred temperature
range is from about 360C to about 430C; it appears that
between 390C and 425C is a very good operating range.
Inasmuch as for these catalysts the reaction must
at all times be conducted in the presence of steam to
facilitate the hydrogenation, hydrocracking, etc., steam i~
used in a ratio such that it is, at minimum, about 27%, by
weight, based on the weight of feed such as the ORA
fraction, crude, the residues, etc., charged to the above
reactors. Conversely, the amount of water charged in the
form of steam at the operating temperature may be increased
or diminished based on the degree of hydrogenation desired
(which also may take place in the first reactor to a certain
degree). If more hydrogenation is sought to be achieved,
more steam is being introduced, but typically steam does not
exceed about 85% weight percent of feed, although 130~, by
weight, may be used based on the hydrogenated product being
obtained, i.e. withdrawn (if gaseous fraction is being
produced then it is converted to a liquid equivalent).
Stated on another basis, the amount of water used is
determined by subtracting the hydrogen content of the
~Z33~77 363~8/3~357
_ 40 -
feedstock from the hydrogen content of the desired product,
on weight basis and multiplying the amount by 9 (as water is
1/9 by weight of hydrogen). Typicallyr up to a 30% excess
is injected in the first reactor.
If water is not being introduced in the reactor,
such as in the form of steam, or is interrupted for one
reason or another, then coking is apt to occur; thus carbon
is being deposited or generated by a process somewhat
similar to catalytic thermal crackingr but in this event the
catalyst acts a thermal cracking catalyst, albeit with some
advantage (because this catalytic thermal cracking is at a
fairly low temperature, e.g. 320C), but vastly less
efficiently than when it functions in the presence of steam
as a hydrogenation and/or cleavage catalyst. ~ence, as
previously mentioned slight coking may also be taken as a
lower, although less desirable, limit.
Intermittent or insufficient steam introduction
will also cause production of especially heavy product in
the reactor 110. It is important that steam is introduced
at all times, in a proper manner in the reactor and
thoroughly dispersed (without any steam and/or reagent free
space).
Nevertheless, it must be mentioned that excessive
amounts of steam also prevent the reaction from beinq
carried out appropriately, apparently by unduly entraining
-
~233~ 36348/36357
- 41 -
the partially converted products.
If carbon is being laid down for one reason or
another, typically it is on the hot spots such as heated
reactor walls or catalyst support. Hence, the reactor 10 is
preferably operated adiabatically. Carbon deposits on the
catalysts can be driven off, that is, converted back into
useful product by exposure to steam for a period of time
without introducing additional feed, after which the
catalyst is useful again and can be used for the production
of the desired product cut. Intermittent introduction of
hydrogen sulfide or sulfur may be helpful in general and for
low sulfide content feedstocks, e.g. sweet crude oils.
Still further, it has been found that if the
temperature, such as with the above catalyst, is increased
to about 320C to 420C, depending on the feedstock
composition (for example its asphaltene composition) an
exothermic reaction may take place, e.g. at 440C an
exothermic reaction sets in. The exothermic reaction may
reach temperatures up to 600C, but it also depends on the
amount of steam being introduced. More steam would tend to
produce lighter carbon pro~ucts. Excessl~e temperatures are
not desired, and temperatures below 440C are preferred.
In the second stage reactor in Figure 1, 17, or
117 in Figure 2 where further reactions take place,
advantageously the products from the first reactor are
~ .
. . .
.
36348/36357
~23;~7~7
rapidly cooled to about 250C and in the presence of a
catalyst, generally between temperatures of 250C and 390C.
Quality of product is increased when the process is operated
at temperatures up to about 430C, preferably 425C, but the
conversion apparently will not increase after 390C.
Cooling in ~uencher 16a is at such a rate that steam does
not condense and interfere with the reaction. The light
ends, of course, that is hydrogenated products, will not be
condensed.
~ he catalyst in the second reactor 17 or 117 again
is preferably a supported catalyst but it can be an
unsupported catalyst, and the treatment of the vaporous or
gaseous products. A typical catalyst for the second stage
reaction, i.e. reactor 17 or 117 is a less severe Catalyst
B. This catalyst is produced by dissolving a technical or
analytical grade of a potassium hydroxide which is
approximately 86~ potassium hydroxide in absolute or 95%
ethanol or methanol (preferahly ethanol) and saturated with
hydrogen sulfide but without boiling off the alkanol,
collecting and trapping anv alkanol given off in a
downstream vessel. Other vessels in which the reaction
takes place may be further downstream to catch the hydrogen
sulfide. When the last vessel containing KOH shows a
reaction, the reaction is stopped in all upstream vessels.
If the reaction is carried out in a further
~2~3~7 36348/3~357
- 43 -
reactor(s) 17 or 117, i.e. secona stage reactors, the
advantages of the process reside in the combination obtained
by the immediate quenching of the reaction products from the
first stage reactor to about 300C but preferably 250C, in
the presence of catalyst, and then conducting these
reactions of the first stage products in the second stage.
For this purpose, it has been found especially advantageous
to support the catalyst on a suitable support. These
supports may be the same as in the first stage, but in any
event these supports must be inert under the reaction
conditions in the particular reactor, e.g. Figure 1, 17, or
Figure 2, 117 and 130. These second stage reactors 17 or
117 may be used as fluidized bed (circulatory fluidized bed,
partially circulating or confined fluidized bed), fixed bed
or liquid bed reactors. ~eactor 30 may also be an
ebullating bed reactor, or a fluidized bed or circulating
fluid bed reactor.
It has been found acceptable for the second stage
reactors, e.g. 17 or 117, to use the supports of a type
commonly available such as alumina-alumina silicates of a
fixed zeolite type, i~e. molecular sieve type, with sodium
or potassium in the zeolite exchanged with ammonia. Type X
and Y zeolites (10 and 13) are suitable. Type Y molecular
sieve zeolites are preferred; of these, the low sodium ratio
sieves are especially desirable (i.e. about less than 1.0~
~.233~7 36348/36357
- 44 -
Na2O~. The molar ratio of silica to alumina of these is
about greater than 3 to 1; about 5 to 1, etc.; Na2O is about
0.2 weight percent. These are available such as from
commercial sources, in forms such as powder spheres,
cylindrical and other extrudates, etc., of suitable size
such as 1/8 of inch extrudates or spheres. Although these
have been alleged to be poisoned or destroyed by alkali
metals, as worked up by the below-described procedure, these
supports are useful despite the deposition thereon of the
herein described alkali sulfide reagents. These supports
may also be used in the first stage reactor 10.
Other zeolites are EL7.-L zeolite of the potassium
type as described in U.S. Patent 3,216,789, and silicalite
material as described in U.S. Patent ~,Ofil~724. The last
has a pore dimension of about 6 Angstrom units. Other
supports are such as those described in British Patent
1,178,18~, i.e. the very low sodium type--less than 0.7
percent, by weight, e.g. ELZ- -6, or ELZ-E-~, E-8, or E-10.
Other supports are mordenites ~nd erionites with very low
sodium content obtained by ammonia exchange and of the
calcined type. Of the above molecular sieves, the type Y
very low sodium, e.g. 0.15, by weight, ammonia exchanged
supports available under Trademark LZ-Y82 from sources such
as Linde Division, Union Carbide Corporation, Danbury, CT,
Mobil Oil Corporation, New York, NY, and other sources are
36348/36357
~233~
- 45 -
preferredO In any event, the stability and durability of
these molecular sieves used as supports are tested under the
reaction conditions and are established by the performance
in the second stage reactor.
The preparation procedure for the second stage
supports is as follows. The low sodium ammonium exchanged
zeolite extrudates, such as powders, or of shapes such as
cylinders, saddles, stars, rings, spheres, etc., of powder,
or extrudates of about 1/~ to 5/32 or 3/16 inch size are
treated with glycerol or like polyhyaroxy alkane compounds,
such as partially reacted polyhydroxy compounds including up
to hexahydric alkanes, by first impregnating these shapes in
a reactor which is kept closed. Thereafter, e.g. when using
glycerol, by heating and removing decomposition products
from these powders, or shapes, from room temperature up to
265 to 280 and even up to 560C, an appropriate, but
unknown, reaction takes place. The thus reacted support is
then screened, drained, and cooled in a closed and tightly
sealed container if the temperature has been brought up to
s6noc.
When cold, the above-described support is then
impregnated with a reagent-catalyst of the general formula
R2Sl.s (empirical); this catalyst is acceptable, ~ut it is
not outstanding for cleavage. Although the catalyst is
designated by an empirical formula, a particular preparative
3~34%/36357
~233~77
- 46 -
method determines the catalyst behavior. Consequently,
while various catalysts may be represented by the same
formula (empirical~, vastly different properties are
observed and different feed stocks being treated will
produce different products. This catalyst, designated as
Catalyst C, is obtained by dissolving 6 moles of KOH in 9
1/2 to 7 1/2 moles of H20 without external heat being
applied, and thereafter with a small amount of alkanol, e.g.
2 to 2.5 cc of methanol or ethanol being added per mole of
KOH. Then 4 moles of elemental sulfur are added to the
foregolng solution which react exothermlcally. Thereafter,
an appropriate amount of sulfur is added for ad]usting the
catalyst to the desired sulfur level by addition of
supplemental sulfur to form the empirical sulfide, ire. from
K2S but K2Sl.l to K2S2.s, including up to K2Ss is useful,
depending on the desired product cut. For more gas in the
product, less sulfur saturated species are used. ~or more
liquids in the product, more sulfur saturated species are
used.
Another catalyst, Catalyst D, is prepared as
follows. One mole of KOH is disssolved in 1.0 moles of
water with vigorous stirring. Then 2 ml of methanol or
ethanol are added immediately after KOH has dissolved.
Immediately thereafter 2/3 moles of elemental sulfur are
added and are allowed to react by a vigorous reaction. The
36348/36357
~:33~77
catalyst is adjusted to the deslred empirical sulfur content
by adding appropriate amounts of sulfur by further stirring,
e.g. one quarter of 2/3 moles of sulfur adds 0.5 to the
empirical sulf~r content of K2S; i.e. 1/4 of 2/3 moles of
dissolved sulfur gives K2Sl.s; 1/2 of 2/3 moles gives
K2S2.0, etc., including other appropriate fractions. Thus
the catalyst may range from R2Sl.l to K2S2.~ or even up to
K2Ss. The lower sulfur content species, e.g. K2~1.s, are
good cleavage catalysts when admixed with Catalyst A, and is
a good catalyst in general for the less severe treatment of
crudes. This catalyst is also a hydrogenation catalvst.
Catalyst A is generally admixed from 3% to 25~ but typically
less than 10~, by weight, to Catalyst D.
When the catalyst has been thus prepared, it is
vacuum evaporated to a flowing slurry. It is then poured
over the glycerol treated, cooled extrudate as described
above (i.e. if the support had heen heated up to 300C. or
higher), and under very low vacuum, agitated and aspirated
until dry. Then the catalyst is further screened when dry
and introduced immediately in the second stage reactor which
has been purged of air oxygen when used as hydroyenation
catalyst.
As another method for protecting the support, if
the glyercol treated support is heated between 260C to a
decomposition point (indicated by appreciable slowing down
.
-,
3~348/36357
~2337~
- 48 -
of a liquid condensate being collected), then the above
described catalyst slurry is added and the vessel is covered
and heated up to at least ~40C, including up to 5hOC. In
any event, the catalyst is calcined above the temperature at
which the catalyst is used in the process.
Still another method is to mix the glycerol, e.g.
about 88 ml of glycerol, to about the one mole (K basis) of
the catalyst, in solution, or by admixing the above
catalysts or mixtures thereof to glycerol. Then the
catalyst-glycerol mixture is heated to drive off water
and~or alcohol leaving a glycerol solution of the catalyst.
Temperature is brought up to 190C for the foregoing. The
mixture is then poured over the support and with agitation
brought up to at least 450C and even up to 560C. This
supported catalyst gives a very unpleasant odor. It must be
prepared under well isolated conditions.
In use for a gallon-si2ed first stage reactor in
conjunction with a second stage reactor, up to about 2/3 of
mole of supported catalyst (K basis) is charged to the
second reactor. As an example, alumina supported
K2Sl 5 (empirical) catalyst may be charged to the second
stage reactor.
Another catalyst, Catalyst E, is a nonsupported or
supported catalyst capable of decreaslng the molecular size
of the product in a first stage reactor (or used in a
~3377~ 363~8/36357
- 49 -
further second stage reactlon). Catalyst E is obtained by
adding a dried KHS powder or slurry in appropriate mole or
weight percent increments (based on the desired size of the
product) to any of the above-descrlbed reagent mixtures A,
B, C or D. Either unsupported or supported forms of the
catalyst may be used. That is from 1/5 to 1/3 moles on
molar basis of K of KHS is added to e.g. K2S (empirical)
sulfide, or to K2Sl 5 (empirical), and the molecular slze of
the product is decreased by these additions of K~S.
When the process is run with the thus supported
catalyst in the second stage reactor, appropriate
adjustments may be made, e.g. K2Sl 1 or K2S1 5 give more
hydrogenation, and K2S2 gives larger molecules (also more
distillate, less gases). These reactions are run in a
temperature range from 113C to 440C. Similar catalyst
adjustments may be maae in other reactors, e.g. when more
than one second stage reactor 17 or 117 is used.
In any event, the first stage reaction, however,
is carried out with the specified reagent to accomplish the
desired degree of cleavage of the refractory, intractable
initial source material components, e.g. resldues of crude
oil, the ORA fraction, and especia~ly asphaltenes. In the
parallel treatment then these not yet reacted intractable
compounds are then treated catalytically to accomplish the
desired degree of cleavage. The total process combin~tion
363~8/36357
~2337~
50 -
in the further stages, that is secon~, third, fourth, etc.,
stages, depends on the specified first stage reaction and
removed portion of unreacted constituents and is thus
interdependent.
The amount of catalyst deposited on the support is
from about 4M of catalyst (R basis~ to about 0.5M or as low
as 0.lM (K basis) per 500 cc of support. On another basis,
not identical basis, the amount of catalyst is about 20
grams per 300cc of support, but it may range from 3 grams or
5 grams of reagent per 100cc of support to about 25 grams to
30 grams per 100cc of support.
The vanadium metal may be entirely removed from
the petroleum feedstock and the heaviest product may contain
essentially all of the vanadium. If run with the presently
supported catalysts, the vanadium is analyzed in very small
quantities or as non-detecta~le in the No. 2 bottoms, i.e.
product 18. The removal of the nickel is aided if some of
the reagent be present in the hydrosulfide form. There is
no reaction between the alkali metal sulfides and nickel
sulfides ~ut there is a solubility reaction when alkali
metal hydrosulfide and nickel sulfides are present. ~ickel
(and iron1 form complexes like ferrites with the alkali
metal sulfides-hydrosulfides. These complexes are
immediately hydrolyzed in liauid water to form the
precipitates of iron or nickel hydroxides. In liquid water,
363~8/3fi357
~2~7~
the vanadium complex with the catalyst is highly water
soluble and water stable. Iron is normally present in the
residue, after distillation range determinations, in amounts
between 3 and 5 ppmr but the amount depends also on the
amounts in the initial stage.
In general, the catalysts for the second and
further sta~e reactions herein are the hydrosulfides and
sulfides, that is, monosulfides and polysulfides of the
Group IA elements of the Periodic Table other than hydrogen
prepared from the alkanol solution as mentioned above.
Although for the stated purpose sodium, potassium, rubidium
and lithium may be used, far and away the most advantageous
are sodium and potassium. ~f these two, potasslum is
preferred. Although rubidium compounds appear to be
acceptable, rubidium, the same as lithium, is not
cost-advantageous. However, for the first stage reactor,
rub-idium may be very advantageous in a blend of rubidium,
potassium and sodium, in the following proportlons: 14%
rubidium, 26~ potassium, and 60% sodium sulfi~es, i.e. the
various species thereof, on basis of the elemental metal,
by weight. The ratio ranges for the preceding mixture are
1:1.5-2.5:3.5-4.5, respectively, but these compositions must
be prepared in the manner as defined according to the
procedure described for Catalyst A. The catalysts used are
typically used as the hydrates, but a small portion of the
`:
~ ,
36348/3~357
'~ 2~ 77
_ ~2 -
catalyst is apparently in the form of an alkanolate tthe
hydrate analogue)~ iOe. up to about 15% but typically less
than 10% or even less than 5~, by weight. Alkali metal
thionates are also present as transitory intermediates.
Hydrates (and alkanolates) of these compounds are very
complex and undergo a number of transitions during the
start-up and as these achieve the reactlon conditions. It
appears that the presence of mixed hydrates and alkanolates
are necessary for the outstanding results. No attempt has
been made to elucidate the nature of these transitions or
the actual structure at the reaction conditions for the
sulfides, hydrates, alkanolates, thionates or the mixtures
of each. It is sufficient to indicate, however, that the
charged catalyst can be a mixture of a number of hydrates
and/or alkanolates or a eutectic mixture of various hydrates
and/or alkanolates, but in any event the specific cleavage
catalyst must be use~ on the refractory compounds.
Similarly, during the reaction, as there is
interconversion of the sulfur-containing forms of the
sulfides, no attempt has been made to characterize this
interconversion. However, as mentioned before, if the first
stage feed contains sizeable amounts of the refractory
components, i.e. asphaltenes of varlous sizes, resins, and
oils such as slurry oils, i.e. more than 50%, by weigh~, and
up, but including up to 95~ of a fraction boiling above
363~8~36357
~233~
~1000F, the first stage reactlon requires at least a
portion of the specific catalyst, e.g. Catalyst A, as
defined above~ As the percentage of the more refractory
compounds in the feed decrease, less of Catalyst A needs to
be used. Similarly for feed 12a, Catalyst A is again
recommended. It may also be added to other catalyst
compositions depending on the desired bond scission (as
characterized by API number, viscosity, etc.), as that
composition appears best to achieve bond scission or
cleavage without excessive gas production. For the No. 1
bottoms, i.e. the feed stream 12a, the preferred reaction
vessel is an ebullating bed reaction vessel or the stirred
cage or basket reactor, further a fluid bed reactor and
still further a fixed bed reactor. Another catalyst for
Eeedstream 12a is Catalyst B with 5%, by weight (on
elemental R basis) of KH~ admixed thereto or replaced in the
same proportions including up to 95% of Catalyst A. In
general, the order of the catalysts for bond scissions or
cleavage of the refractory compounds is as follows:
Catalyst A (ethanol prepared species); Catalyst A (methanol
prepared species); Catalyst A (methanol-ethanol prepared
species~; Catalyst A + B (mixture 50~ or more Catalyst A on
R basis moles); Catalyst D; Catalyst E (about 1/3 moles of
KHS added); Catalyst B, and lastly Catalyst C. Of course,
increaslng the amount of KH~ additions to the catalysts will
....
.
~.
3h348/36357
- 54 -
increase the cra~king ability.
The order of the catalysts for hydrogenation is as
follows: 1/2 Catalyst A + RO~ on molar basis, then add the
other half of A (very exothermic hydrogenation reactions);
Catalyst C with K~S mixture ~based on needed aegree of
cleavage); Catalvst B + A (50% or more of Catalyst B on R
mole basis), and Catalyst B.
It must be remembered that in the above
discussion, the refractory nature of the carbonaceous source
materials and their components is that assigned to these
according to the prior art. By comparison, the present
process is carried out at lower temperatures and lower
pressures and is capable of converting all of the prior art
refractory materials with great advantage and facility.
However, as the carbonaceous source materials and their
components have a different "refractory" nature according to
the present process, e.g. the paraffins are the most
refractory while the asphaltenes and aromatlcs are not the
invention resides in the abili~y to treat, in a proper
sequence, according to this process, with the proper
catalysts, all the components of these carbonaceous
materials as characterized by heretofore unachievable yields
when treating these carbonaceous source materials.
In the following Examples, various reactior~s are
described. There is no intent to l;mit the invention by the
36348/36357
3~77
- 55 -
B amples but merely to illustrate its applicability.
EXAMPLE I
A high softening point asphaltene 270F as
described below and of a solvent extracted type was treated
with the following reagent to obtain product A. The
catalyst was Catalyst A previously described. When the
product from the first stage treated asphaltenes were
reacted in a second stage, the product was identified as ~.
The second stage catalyst was the same as in the first
stage. Both catalyst compositions were unsupported.
36348/3~357
~ ~337~
- 56 -
Feed Distillate Distillate Blend Residue
of A~B
A B A~B 600
Gravity, API Q60F -4.6 31.3 36.7 36.08.7
Kin. Visc. @210F, cSt -* - 0.94 58.0
Con. Carbcn Res., wt% 39.5 - - - 0.20 16.4
Aniline Foint, C -- 44.6
FIA, vol~
Aromatics - - 71.5
Olefins
Saturates - - 28.5
Bromine N~. - - - 53
Carbon, wt% 84.24 83.77 84.81
Hydrogen, wt% 8.50 - - 12.52 9.95
Nitrogen, wt% 0.75 - - - 0.10 0.58
Sulfur, wt% 6.1q 2.31 4.46
Ash, wt% 0.30 - 0.08 0.20
Moisture, wt% Nil 0.05 Nil
Oxygen, wtg 0.02 - 1.17 0.00
Nickel, ppm(w) 71 - - 33
Vanadium, ppm(w) 174 - - -- 160
Iron, ppm(w) LBl - - - 24
Heptane Insoluble 16.7
(IP Method)
* 270F - Softening Point
Arcmatics and olefins were not clearly separated in the
oolumn probably due to a heavy tail abcve 600F.
3~348/36357
- 57 -
The above data clearly indicate the consldera~le improvement
in the viscosity as well as the gravity of the products, the
dramatic increase of the hydrogen content and the
considerable removal of the metals present Erom the later
fractions.
The following examples show the results obtained.
The feedstock charge material was solvent-extracted
asphaltenes of the type identlfied above and in Example I
hereln .
All of the runs ~were made as batch-process runs,
in a st1r-tank reactor. The stir-tank reactor has an inslde
volume of 6.24" diameter and 10~ height. The stir-tank
reactor is fitted wlth an agitator and a steam sparger.
A steam generator, directly connected to the Clty
water supply, forms steam at 40 lb/sq.in. pressure. The
steam passes through 3/8~ inside diameter lines to the
sparger and is at atmospheric pressure. However, tne
reactor may operate from 1/2 atm. to about 5 atm. or even
higher as previously discussed.
The sparger is approximately 3~" in diameter and
has a series of sparger holes, all of the holes direct the
steam upwardly. The sparger is located on the bottom of the
reactor.
An agitator is provided or the reactor when
unsupported catalyst is used. The motor is mounted directly
.
,.
363~8/36357
~ ~3377~
- 58 -
above the reactor. A seal seals the area through which the
agitator rod passes into the reactor. The agitator is of
twin circles connected by angled, curved blades. The
agitator may be replaced by baskets holding supported
catalyst as further descrlbed herein. Four baskets
containing supported catalyst are mounted on the agitator
shaft. The bas~ets plus the agitator shaft has a total
diameter of almost 6.25". The baskets are 6" high and are
approximately ~" thick. The unmounted basket is a ~" deep
rectangle.
The top of the reactor contains the seal through
which the agitator shaft turns, the riser, which exits
overhead distillate from the reactor, a pressure relief
line, which consists of a valve which opens above 30
lbs/sq.in. pressure and vents the contents of the reactor to
a hoo~. This pressure relief fitting is also used to fill
the reactor with solid feedsto~k charge.
Usually, two thermocouples are fitted into the top
of the reactor. One thermocouple measures tne ternperature
at the bottom half of the reactor and the upper thermocouple
measures the temperature in the upper half of the reactor.
The riser conslsts of a llne approximately gn high
and having an inslde diameter of approximately 3/4n.
The second stage reactor is a tube reactor having
an inslde diameter of 1~" and is 12" long. The capac~ty of
.
3h348/36357
~337~
- 59 -
this reactor is 347.5cc. This reactor i5 fitted with three
wrap-around heaters. A thermocouple controls each of the
heaters througn the controllers, which are mounted on a
portable stand.
Gases and vapors passlng the up-flow second stage
reactor are then conaucted downward througn a 16" glass
bubble condenser. This condenser is not water cooled.
The first con~enser is mounted vertlcally and the
bottom of the condenser holds a 500cc collecting flask. The
flask is normally maintained at 2~0C by a mantel type
heater. The bottom of the flask has a stop cock for
collecting product.
A secon~ condenser rises from the a~ove flask and
is parallel with the first condenser. The second condenser
is not cooled by water. The second condenser is also a
glass condenser with bubble type coollng areas.
The downward slanting tube from the second
condenser connects to a water-cooled condenser. This water-
cooled condenser is mounted vertically and is approximately
18" long; it ~its into the top of an unheated 500cc flask
which is fltted with a stop-cock at the bottom. A parallel
vertical water-cooled condenser rises from the second
fittlng in this flask. Another water-cooled condenser is
fitted directly above this condenser.
The top water-cooled condenser is fitted with a
.~ .
'
36348/36357
~233t7~
- 60 -
12" long line which angles upwardly. This line has a
diameter between ~" and 3/4n. This line connects to an ice
cooler.
The ice cooler is a twin wall vessel, ordinarlly
used to trap vapors before they can enter a vacuum pump.
The center container contains a water-ice mixture while the
gases and vapors pass through the external container
sectlon. The gases and vapors enter at the bottom of the
vessel and exit at the top of the vessel. The bottom of the
vessel contains a 50cc collector. The collector is fitted
with a stopcock, for removing product.
The remaining gases and vapors are sent to another
cooler, similar to the ice cooler. This cooler is cooled by
a mlxture of solld carbon dioxide and 2-propanol. The
product is again collected in a 50cc vessel below the cooler
and this vessel is fitted ~7ith a stopcock.
The remainlnq gases and vapors are then washed in
a solution of potassium hydroxide. The solution contains 6
moles of KOH dissolved in 360 ml of water. The gases are
then measured by passage through a wet test meter. After
this measurement, samples are periodically collected and the
uncollected gases are vented to the hood.
For the following runs, 1300 grams of the solid
asphaltenes are weighed out and crushed to a size to charge
the react~r. Liquid or solid catalyst protected from oxygen
363~8/3~357
~2~3~
- 61 -
is then added to the reactor, previously purged, e.g. with
helium. About 40 grams of theoretical anhydrous Catalyst A
is charged to the reactor.
The secondary reactor is charged, again with the
same precaution, usually with about 300 cc of a supported
catalyst. The secondary reactor is initially heated, in
order to drive out the water content of both the zeolite
support and the catalyst.
After the water content of the secondary reactor
space has been reduced by bringing the temperature of the
second reactor to above 300C, the primary reactor is
heated.
Solvent-extracted asphaltenes having melting
points of either 200 or 400F were used. The melting point
determines the particular form of the asphaltenes.
After temperature adequate to melt the asphaltenes
were reached in the primary reactor, the agitator was turned
on. Normally the agitator is initially operated at
approximately 30 to 60 rpm.
Steam is normally introduced when the primary
reactor reaches a temperature of 220C. ~y this time, the
second stage reactor should have reached or leveled off to
424C.
Helium is normally sparged through the sparger
prlor to the introduction of steam to the system in order to
:
36348/36357
3~
- 62 -
keep the sparger holes clear and the system free of oxygen.
The helium is sparged at approximately 200 cc/minute.
EXAMPLE II
1300 grams of the solvent-extracted asphaltenes
were reduced in size so that they could pass through the 1"
opening in the top of the reactor. The asphaltenes were not
heated but were charged to the reactor in solid form. After
the charge, the reactor was sparged with helium.
The catalyst used was another verslon of Catalyst
A prepared as follows~ To the previously described initial
solutlon of KOH was added a solutlon of one mole KOH
dissolved in 30 cc of H2O and then the solution mixture
saturated with hydrogen sulfide. The solution separates in
two layers about 1/3 top layer and 2/3 bottom layer. The
layers are separated and dried and then the two proportions
reblended. The reblending may be in the same proportions as
obtained (as it was in this Example), or the proportlons of
the two catalysts mav be varied. The catalyst may also,
upon reblending, be dissolved or dispersed for deposition on
a support. On a theoretlcally anhydrous con~ition, the
weight of charged catalyst was approximately 40 grams.
The second stage reactor had been charged with
zeolite supported catalyst during apparatus assembly. The
second stage reactor contained approximately 300 grams of
36348/3~357
~L23377~7
- 63 -
support and catalyst. The zeolite support was L2-Y82 and
the catalyst was catalyst D, i.e. R~Sl.s (empirical), to
enhance the hydrogenation of the cleaved product. The
start~up procedure requires that the secondary reactor be
brought to at least 175C before the primary reactor is
heated.
The first stage reactor was then heated. Only the
bottom one-half of the reactor is heated, the top half of
the reactor is not heated. At 220C, a small amount of
steam was added to the reactor through the bottom sparger.
At approximately 320C, in the bottom of the first
stage reactor, there began a steady but slow production of
hydrocarbon product, which was condensed in the flask below
the twln water-cooled condensers. However, tnis product was
much heavier than the product obtained at
process-temperatures, in the 390C to 424C range.
When the bottom of the first stage reactor reached
360C, there was a cons1derable improvement in the rate of
product productlon. The react~on became exothermlc and rose
rapidly an~ leveled off at about 415C. This temperature
was maintained from that time forwar~ in the bottom of the
reactor. The top of the reactor had reached 3~0C.
The temperatures in the second stage reactor~ are
in the 220OC to 4~0C range.
When the contents of the first stage reactor were
3fi3d~8/36357
~233~77
- 6~ -
in contact with the agitator, the process ran uni~ormly at
about 415C in the first stage reactor and with variations
between 440 and 460C in the second stage reactor.
The amount of steam was estimated at approximately
20 cc of water converted to steam/minute. At the end of the
run, the top and bottom temperatures in the primary reactor
were allowed to rise to 440C.
Catalyst A variatlon above (as described above~
gave almost no gas through the wet test meter. The amount
of gas was less than 6 liters.
The bulk of product was the No. 2 bottoms,
collected below the water-cooled condensers. This product,
when combined with the product collected below the water-ice
trap totalled 458 grams. This product had an API number of
23 (sp.gr. @60F 0.9158).
The No. 1 bottoms totalled 38 grams and had a
gravity of 0.96587 (API number 15) @ 60F. The No. 1
bottoms were collected below tne air-cooled condensersO
The amount of bottoms collected below the dry
ice-2-propanol cold trap measured 44 cc in the calibrated
trap. However, when collected, only 28 cc were obtained due
to the evaporation of these light ends. The API gravity of
these light ends was 81 @ -10C (sp.gr. @-10C = 0.6553).
Due to rapid evaporatlon this gravity is very imprecise~
An extraordinarlly light coke was formed and
:` :
.. . .
36348/36357
~233~
- 65 -
formed 2~ thick layers in the reactor. This coke measured
1800 cc but had a weight of 513 grams.
The dead space below the agitator causes a delayed
coking operation below the true reaction zone.
EXAMPLR III
~ his example was carried out in a simllar manner
to that of Example II with the exception of a different form
of the catalyst and a more rapid start-up heating of the
first stage reactor.
Catalyst A for this example was a single layer
catalyst and did not require the layer separatlon durlng the
drying phase that was used for the catalyst used in Example
II above. The sustaine~ temperature of this run was 420C.
In the 42 minutes o~ this run, after achieving
process-temperature (at about 420C), the catalyst,
approximately 40 grams on a theoretically anhydrous basis,
converted the following ~ottoms from the initial 1300 gram
solvent-extractea asphaltene charge:
a) No. 1 bottoms, collected below the air-cooled
condensers, totalled 2~ grams of a hydrocar~on, having an
API grav1ty of 11.5 (sp.gr. of 0.98Y5 @ ~0F).
b) No. 2 bottoms, collected below the water-cooled
condensers and the water-ice condensers, totalllng 398 grams
of hydrocarbon, having an API gravity of 29 (sp.gr. @ 6UF =
, ~
36348/36357
~ ~337~7
0.8816).
c) No. 5 bottoms, collected below the dry
ice-2 propanol cold trap, totalled 63 cc with a gravity of
83 at ambient temperatures. A substantial part of the No. 5
bottoms were lost in determlnlng this API gravity.
dj A total of 135 liters of gas were prod~ced.
The gases were measured following the alkali hydroxide wash
of the gas-vapors (following the dry ice-2-propanol cold
trap). These gases were not collected for analysis, but the
average analysis of simllar runs produced approximately 5
(volume percent) non-hydrocarbon gases, such as hydrogen,
carbon monoxide and carbon dioxide. The remaining
hydrocarbon gases have an average molecular weight of 49; on
this basis a total weight of 280 grams of hydrocarbon can be
asslgned to the gas obtained.
e) The same light coke as formea in Example II was
observe~ in the reactor fcllowlng this run. The weight of
the coke was 48~ grams.
The accounted for weights are:
No. 1 Bottoms 29.0 grams
No. 2 Bottoms 3g8.0 grams
No. 5 Bottoms 41.5 grams (Cold trap)
Gases 280.0 grams
Coke 489.0 grams
Total 1,238.0 grams Accountabilty = 95.23%
Converslon = Nos. 1,2,5 bottoms + gas/feedstoc~ charge =
57.57~
` ~ .
36348/36357
~3377~
- 67 -
For both Examples II and III the supporte~
catalyst of the secona reactor was completely clean and free
of pitch, carbon, etc.
The total accounta~ility for products obtained by
Example II was 1,045.7/1300 = 80.42%. The conversion, i.e.
the Nos. 1,2,5 bottoms ~ gases = 532.7/1300 = 4u.97~.
The principal di~ference between the two Examples
was the much higher gas productlon in Example ~II.
In these two examples, the bottoms completely
separatea from the water, condense~ from the steam and no
emulslon was formed.
EXAMPLE IV
In this Example, a blend o~ catalyst was use~,
i.e. about 2/3 of the catalyst was that described of
Example II, but not separated, 1/3 catalyst of Example III
(K basis). The varlous proportlons may be changed,
including the proportlons of the catalyst layers in Example
II. The catalyst was unsupported an~ was about 4~ grams on
a theoretlcally anhydrous basls.
The reactor reached a temperature of 42no durlng
this run.
36348/36357
37~7
- 68 -
The products obtained durlng this run were:
No. 1 Bottoms 195.8 grams (sp.gr. 0.9793 @ 60F or API 13)
No. 2 Bottoms 309.5 grams (sp.gr. = 0.86 ~ 60F or API 33)
No. 5 Bottoms 28.0 grams
Gas 276.4 grams (133 llters X 0.95/22~4 X 49 =
27~.4 grams)
Coke 473.0 grams
Total 1,282.7 grams
Accountabilty = 1,282.7/1300 = 98.67~
onverslon (Nos. 1,2,5 bottoms + gas = 809.7 grams/1300
grams) = 62.28%.
It was apparent that the coke formation was from
the 480 cc of space below the agitator and some of this
space is also below the sparyer.
EXAMPL~ V
The prlnclpal di~rerence between this Example and
the previous Examples was the use of a supported catalyst
instea~ of unsupported catalyst being added prior to the
beginning of the run. The catalyst was supporte~ on
stainless steel slnterea mesh in four baskets. The
stainless steel was 1/8" thick and had been cut into 1/8"
strlps which were in turn cross-cut for 3/16~ sizes. The
support size was therefore 1/8" X 1/8" X 3/16n. The
catalyst was the same catalyst as used in Example II.
The supported catalyst was placed in 1/4" X 2. ~n X
6" baskets, 4 baskets were used an~ the bas~ets were
supportea and turne~ by the agitator shaft. The baskets
363~8/36357
~3377~f
- 69 -
became the agitator. A mesh held the supported ca~alyst in
place and the wire mesh baskets were supported by a frame.
Wlth the same catalyst as in Example II (and
conslderably less of the catalyst in the supported form) the
amount of gas produce~ decreased from 135 liters to 90
liters. Most of this gas was produced during the end of the
run when the temperatures rose to 440C.
The speed of the agltator which spun the baskets
was initlally 60 rpm and later in the run was increased to
120 rpm.
as 187.0 grams (90 llters of gas X o.95/22.4 X
4q = 187.0)
No. 1 bottoms 407.0 grams
No. 2 bottoms 368.0 grams
No. 5 bottoms 43.0 grams
Coke 19Y . 5 grams
Total 1,204.5 grams
The fee~stoc~ charge of the solvent-extracted
asphaltenes was 1290 grams. AccountaDility = 1204.5
grams/1290 grams = 93%. COnverSlon = Nos. 1, 2, an~ 5
bottoms + gas / 1290 = 77.0~.
The API number of the combined No. 2 and No. 5
bottoms was 32.0 @ 60F. Of the No. 1 bottoms, a division
was made between the par~ which was llquid at 200C and that
which was not llquid at 200C. The liquid portion had a
calculate~ API number of 5 and this portlon constlt~ted 228
grams of the 407 grams of total No. 1 bottoms, or 5~.03~.
36348/36357
~L233~77~
- 70 -
The remainder appear to be sllghtly upgraded forms of the
solvent extracte~ asphaltenes. Recalculation of the
conversion based on the liquid portlon @ 200C of the NoO 1
bottoms give a conversion of 64~.
In all these converSlonS and accountability
estlmates, the gas produce~ is calculated at 95~
hydrocarbon, of a hydrocar~on of an average molecular weight
of 49.
It is evident that, when the feedstock charge is
below the sparger, a competltive process is operational. It
involves a decreased temperature catalytic thermal cracking
with an improved threshold llmlt for thermal cracking, i.e.
when steam is not present along wlth tne catalyst. When the
catalyst is supported, there is no coke formatlon if steam,
the asphaltene feed, an~ the catalyst are in intlmate
contact wlth each other. If insufficient steam, or no steam
reaches the asphaltene feed, then the reactlon turns into a
catalytically aided thermal cracking. Although carried out
at lower temperature then normal thermal cracking, at about
atmospherlc pressure, and at about a rate 10 times faster
than normal thermal cracking, the catalytic hydrocracking-
hydrogenatlon is vastly more deslrable because or the high
yields, high space velocities, deslrable, adjustable product
composition and the reaction conditions.
3~3~8/36357
~2337~
EXAMPLE VI
In accordance Wlth the above described procedure
as it concerns Example II, a mixed res1due feed containing
various types of resldues deslgnated as FHC-353 was used.
Analysis for the feed is as follows:
weight % of below 1000F cut 8.~%
API number 6.6
specific gravity at 60F 1.0243
sulfur, wt. % 3.91
nitrogen, wt. % 0.478
hydrogen, wt. % 10.35
carbon, wt. ~ 84.72
oxygen, wt. ~ 0.471
Total elements: 99.9%
~ils, wt. % 29.6%
reslns, wt. % 57.8
aspaltenes, wt. ~ 12.
Recovery, wt. % lOu%
Ramsbottom carbon, wt. % 21%
Metals:
vanaaium 228 ppm
nickel 52 ppm
iron 14 ppm
sodium 4 ppm
Viscoslty 143u cts at lOO~C
2~n cts at 135C
pour point 130F
The residue feed was grad~ally heate~ up so as not
to cause coking untll the mixed resldue became pumpable or
reached the reactlon temperature of about 425C, at which
~.
3634~/36357
~2337~77
temperature the reactlon was conducted in the presence of
catalyst and steam. Total feed sample was 15,650 grams, and
recovery was 98.1~ of the feed charge based on the various
condensates obtained.
A boiling point redetermlnation of this feed
residue according to ASTM D-1160 and D-1180 was also carrled
out on a 200cc sample of the feed in accordance with the
described metho~ and the data obtained were as follows
(reported as temperatures observed durlng the
determinatlon) .
' :
:
r~ ~7 36348/36357
3~
- 73 -
~ample % GC temp. at C temp. Corrected temp~ Pressure in
cc Reçovery top of column of pot (D-1160; ~-1180? column-mm. of ~9
2~0 275 975 ~.2
253 286 810 4.2
263 295 835 4.6
273 304 858 4.2
283 310 882 4.2
290 316 4.2
100 50 297 324 907 4.2
110 55 305 330
120 ~0 312 338 938 4.2
130 65 320 344 ~53 4.2
140 70 326 351 4.2
160 80 319 347 990 2.0
170 85 330 2.0
180 90 339 364 1030 2.0
188 94 347 1050 2.0
347
190 99 351 380 1090 1.0
, .
~3~7~ 36348/36357
- 74 -
This product was converted in a stlrred basket reactor
having a supported catalyst. The support was a Y-82 zeolite
from Union Carbide, Danbury, CT in the form of a 1/8 inch
extrudates, e.g. flat shapes. The catalyst has an empirical
formula of K2S1 5, i.e. prepared as catalyst D above, with
5%, by weight, of Catalyst A ad~ed thereto. The amount of
catalyst was 20.16 grams per 220cc of the described support.
Initlal converSlon of product for first product
(top product from reactor 110) was 71% (for combined product
except the very volatile, dry ice and iso-propanol cold trap
materlals and gases). This convers1on also does not include
the distllled materlal in the feed recoverable below 1050F.
The top products plus recovered bottoms were accounted for
and 98.1% of initial feed charge, the remainder apparently
being the gas and volatile distillates.
A 1,392 gram test sample contalning 7 grams of H2O
of top products of reactor 110 and 117 except for the
con~ensates mentioned above had the following component
distrlbution: ,~
nitial boillng point to 360F 71.6 grams or 5~24 weight ~
360F to 650F 203.4 grams or 14.89 weight %
650F to 1000F lG84.1 grams or 79.36 weight %
Total distllled: 1359.1 grams.
The product of reactor 110 and 117 had an initial
boilng point to 360F product which had the following
characterlstlcs:
36348/3~357
~33~77
API number 54.5
specific gravity .7608 at 60F
ViSCoSlty 0.72 cts at 100F
0.568 cts at 150F
ash content .009%, by ~eight
nitrogen content 0.05%
sulfur content 1.0%.
The fractlon boillng from 360F to ~40F has the
following charact~ristlcs:
API number 34.2
speci~lc gravity 0.8540 at 60F
water 0.002%, by weight
pour point -20F
ViSCoslty 2.57 cts at 100F
1.61 cts at 160F
sulfur 2.18~
ash 0.01%, by weight.
The fractlon boillng to 650F to endpoint ha~ the
followlng characterlstlcs:
API number 17.8
specific gravity 0.9478 at 60F
pour point +115F
Viscoslty 37.34 cts at 150F
12.05 cts at ?.10F
carbon content 85.03%, by weight
hydrogen 11.55%, by welght
nitrogen content .2~, by weight
Conradson carbon 2.56~, by weight.
The recovered res1due, i.e. remaining after
+1050F stage had been reache~ had:
vana~lum 5.6 ppm
iron 48.6 ppm
sodium 8.2 ppm
potasslum 22.3 ppm
nickel 0.48 ppm.
.....
36348/36357
~3~77~
- 76 -
For the vacuum gas O11 fractlon, i.e. 650F and
above fractlon, the characteristlcs were as follows.
Initial boiling point 652F and recovery as follows:
10% at 732F
20% at 78~F
30% at 83uF
40% at 87~F
50% at 894F
60% at 92~F
70% at 948F
80% at 976F
90% at 1018F
end point at1050F
at 94.4% recovery.
The K factor for completely treated top products
from reactor 110 for the 650F plus fraction (hydrogenation
as a last step~ was 11.71.
The bottoms from the first stage reactor 110
recovered as 112a were rerun on a batch basls when
sufficient amount were recoverea. The catalyst for feed
112a was Catalyst A supported on a steel mesh particles of a
size of 1/8 X 1/4 X 3/lfi, of total volume of 200cc which
contained 15.7 grams of catalyst. ~otal sample run was 1286
grams. Total recovery was 750.7 grams (which included also
8 grams from the dry-ice trap but no gas). About 29~ of
FHC-353 feed were recovered as hottoms. Feed analysis of
,
36348/36357
337~7
- 77 -
112a bottoms was:
carbon 84.78~, by weight
hydrogen 8.74g, by weight
ash 0.98~, by weight
sulfur 4.64%, by welght
Conradson carbon 4~.57%, by weight.
About 58.4% of feed was converted. On a COntlnUOUS basls
higher conversion would be reacl)ed, but on batcn basis the
dead space in the reactor underneath ~he spinning cage or
basket did not allow complete reactlon space, as well as the
holdup in transfer llnes prevent completion of the reaction.
Combined recovered products had the following
characterlstlcs:
carbon 82.24%, by welght
hydrogen 11.~5%, by weight
oxyqen 0.64%, by weight
water 2.8%, by weight.
The 360 to 650F boillng point fraction has the
following characterlstlcs:
API number 27.2
pour point -15F
viscosity 8.533 cts at 100F
4.028 cts at 150F
nitrogen .17%, by welght
water 2.8%, hy weight
36348/36357
~2~37~'7
- 78 -
The fraction boiling 650C to endpoint ha~ the
following characteristlcs:
API number 17.4
specif lC gravity .944
pour point +95F
viscosity 21.54 cts at 150F
7.989 cts at 210F
carbon 84.7~
hydrogen 11.27%
oxygen - .66%
Conradson carbon 2.2%, by weight
nitrogen 0.27%, by weight
vanadium 3.5 ppm
iron 7.6 ppm
sodium 1.6 ppm
potasslum 10.3 ppm
nickel .45 ppm.
Distlllation resldue:
carbon 85.22%, by weight
hydrogen 11.06%, by weight
oxygen .49%, by ~eight
sodium 3.07 ppm
vanadium N. D.
nic~el .344 ppm
iron 4.6 ppm
potassium 27.4 ppm.
The 650F+ product (from feed 112a) has the
following distlllation characteristlcs for a 191 gram charge
of which 163 grams were overhead; it is a 175 ml sample at
60F. The recovered fractlons were as follows:
10~ at 650F
20% at 77~F
30% at 830F
40~ at 87uF
50% at 907OF~
80% at 1030F
85~ at 1050F
36348/36357
~23377~
- 79 -
This distillation and resldue product had the
folluwing characteristics: API number of 20.0 at ~0F and
16.4, respectlvely. Moreover, the low metal content and the
fairly high API number make tnis resldue a good feed, e.g.
as a catalytic crac~er feed.
While the resldue was about 15%, it would
constltute the bottoms in the reactor vessel, e.g. 130, and
these again may be su~iected to further treatment, e.g. in a
further parallel pass, i.e. the same as for feed 112a. The
analysls of the reactlon vessel resldue product was:
carbon 83.99%, by weight
hydrogen 6.8~%, by weight
oxygen 2.9%, by weight
water .05%, by weight.
Based on the combined 58.37~ (recovered product
divided by weight of feed stock with accountability of 83%)
conversion (rerun of bottom products from reactor 11~ feed
112a) and the top products of reactor 110 but not counting
in for the last the dry ice trap liquids and not the
recovered gases for the top products of reactor 110 and not
the supposed distlllahle products ~suppose~ly stlll in
FHC-353 feed distllling below +1050F), the total converslon
is 87.9%; however, if gases ana dry ice cold trap products
are countea on an average molecular weight basls of 4~ and
the volume converted to weight, including the actual
distillate content of FHC-353, then the actual recovery is
3~348/36357
~33~
80 -
even higher.
The yields in the above examples have been
calculated on a basls to ren~er the interpretation o~ these
yields consistent wlth the conventlonal practice. The
converslon on a space-time velocity may also be calculated,
i.e. the startlng material treated per unlt volume of the
catalyst for that starting material per a given unit of
time.
The yields are conventionally calculated on the
basls of the total amount of startlng material used and
converted with a fraction of a component boiling at a
boiling point of less than +1,050F included. The boillng
points are determlned by the previously described methods,
which are the conventlonal methods used in the petroleum
industry. In petroleum industry practice tne limit of
+lO~uF is used to describe tne very refractory crude or
distlllatlon product components according to the
conventlonal methods.
The yields may also be base~ on a proportlon of
the components converted to a llghter Viscoslty product such
as on a degree of API product basls. The residues such as
collected by the cyclones 116 and used as feed 112a are
calculated into the yields as additlons to the initial
amount of starting materlal treated and recovered. The
residue from vessel 130 boi11ng at or over 1,0~0F is then
`:~
36348/36357
~2;~37~7
- 81 -
subtracted from the overall collected materlal provided this
residue is not further treated.
In some practice the yields are based on the
treatment of resldues having certain boiling points, namely
residues boillng more than at 650C and less than 8S0C or
like limlts, and then to consider only the convertible
fractlon by conventlonal processes (whatever that process is
for the particular refiner). Inasmuch as that basis is not
useful for evaluating the present process because the
present process converts nearly all the previously
unconvertlble residues, it is believed best to characterize
the products which are recovered in the straight through run
and the first and second (if necessary) parallel run
proaucts distillable at less than 1050F belng additions.
That is after the first run products and second or third
parallel run products are added together including any
further reactlon vessel, e.g. 117, etc., including all
condensed products, etc., and these are combined and a
distlllation is carrled out as described and the resldues
boiling above ~1050F are subtracted for the conversion
calculations, the yields may then be determlned.
For best practice, the gaseous hydrocarbon
products in the feed distlllable at less than 1050F, the
low temperature condenslbles and gas are included. Gas can
be included on a calculated basls such as average molecular
:
.
3~348/363~7
~337~'7
- 82 -
weight calculated for the volume collected~ converted to a
weight percent to provide for accountability of all
pro~ucts.
Accordlng to the above description and pursuant to
the Example VI as disclosed herein, it is believed that the
specific treatment of the split or separated fractions from
petroleum crude, low boiling point residues, up to 850Ff or
high boil1ng point resldues up tv and over 1050F, in a
straight through fashion at low pressures and low
temperatures allows the cOnverSlon of almost the entlre
crude Oll or petroleum or other hydrocarbon values into
usable products if a second or further treatment of the
bottom fractions is practlce~ wlth the specifically selected
catalysts.
Heretofore conslderahle effort had to be devoted
to the conversion of the intractable portions o~ crude oll
or petroleum resldues into useful components, but the yields
have been low and numerous recycles had to be conducted to
achieve acceptable yields. If it is remembered that about
up to 50% of tnese intractable portions had to be subjec~ed
to the very severe treatment, the present parallel method,
which provi~es conslderahly higher yields an~ better end
products, high space-time velocity, etc., while utillzing
poorer starting product or separately treatlng the split
product at less drastic conditlons, then it is eviclent that
36348/3~3S7
~233~t7
- 83 _
the present parallel method stands out as a great
contrlbution in the contlnuous search for obtaining all
useful products from a given volume of crude oil, or its
resldue.
~ ,
