Note: Descriptions are shown in the official language in which they were submitted.
12~?31~
i-
~ 936-153 a
MIDDISTILLATE PRODUCTION
TECHNICAL FIELD
One of the most important characteristics -that a
modern petroleum refinery must have is flexibility. The
ability to use different feedstocks, ranging from shale
oils and heavy oils to light oils, to produce differen-t
product slates in response to changin~ consumer demands is
crucial to profitability. Feedstocks, for example, are now
shifting to higher boiling, lower quality mixtures having
I0 more metal, nitrogen, and sulfur contaminants than feeds
previously used. Forecasts of consumer demand are predicting
shifts from gasoline-range hydrocarbons to heavier, higher
boiling products such as diesel fuels, fuel oils, and turbine
fuels.
Hydrocracking, used either in a one-step process
or in multistep processes coupled wi:~hhydrodesulfurization
and hydrodenitrogenation steps, has been used extensively
to upgrade poor-quality feeds and to produce gasoline-range
materials. Over the years, much development work has gone
into finding improved hydrocracking conditions and catalysts.
Tests have used catalysts containing only amorphous materials
and catalysts containing zeolites composited with amorphous
materials.
Among the zeolites disclosèd in the literature
are Y (U.S 3,130,007); decationized Y (U.S. 3,130,006);
ultrastable Y (U.S. 3,293,192 and U.S. 3,449,070); and
ultrahydrophobic Y (U.K. 2,014,970, published September 5,
1979). Disclosures have appeared which relate -to modifying
zeolites. U.S. 3,367,884, Reid, February 6, 1968, discloses
~Z~?331 9~
--2--
reduclng the activity of superactive zeolites by calcining,
leaching procedure. The catalyst is disclosed as being
especially useful for cracking yas oils to gasoline. U.S.
3,506,400, Eberly et al., April 14, 1970 and U.S. 3,591,488,
Eberly et al., July 6, 1971, disclose improving the stab-
ility of the crystalline lattice of zeolites by a steam-
ing and acid extraction process. The product, a stabilized
zeolite, may be used in hydrocracking and is disclosed as
having improved selectivity as shown by higher gasollne
yield and lower coke make.
Other research has resulted in disclosures rela-
ting to producing midbarrel products. U.S. 3,853,72, Ward,
December 10, 1974, discloses hydrocracking high-boiling
feeds to produce midbarrel products using a catalyst con-
taining a steamed zeolite. U.S. 4,120,825, Ward, October
17, 1978, also discloses improving the production of mid-
barrel products by hydrocracking with a catalyst containlng
a steamed zeolite.
I have discovered that middistillate products
can be selectively produced by hydroprocessing with an
expanded-pore zeolitic catalyst.
TECHNICAL DISCLOSURE
My discoveries are embodied in the process for
selectively producing middistillate hydrocarbons, comprising:
(a) contacting under hydroprocessing conditions
a hydrocarbonaceous feed boiling above about 600 F (316 C)
with a catalyst comprising a hydrogenation component and an
expanded pore zeolite which consists of a faujasitic zeolite
which has been steamed then dealuminated; and
~)3~9~L
-2a-
(b) recovering a hydrocarbonaceous effluent of
which more than about 40 percent by volume boils above about
300F (149C) and below about 700F (371C).
I have discovered that faujasitic zeoli-tes which
have been dealuminated after high-temerpature steaming have
surprising stability and activity for producing middistill-
ates from higher-boiling feeds. By faujasitic zeolites is
meant crystalline aluminosilicates, synthetic or natural,
which have the crystalline structure of the large-pore zeo-
lite, faujasite. These zeolites include faujastie, zeolite
X, zeolite Y, and zeolites derived from them. For example,
there are numerous processes known to the art for treating
zeolite Y to produce "decationized Y", "ultrastable Y",
"Z14-US", and others. The preferred faujasitic zeolite is
zeolite Y, as well as derivative
-
~2~3~g~
01 _3_
zeolites having the crystal lattice characteristics of
zeolite Y. The most pre~erred faujasitic zeolite is an
05 ultrastable Y zeolite having a sodium content of less than
about 0 5 wt. % (as Na2O).
As prepared, the large-pore zeolites typically
contain significant amounts of alkali metal. Because the
alkali metals tend to poison the acid sites of the zeo-
lite, standard ion-exchange procedures are used to remove
them. The alkali metal content (calculated as oxide) is
preferably reduced to below 5 weight percent before any
heat treatments, and to less than about ~00 ppm by weight
in the final zeolite
The faujasitic zeolite being treated is calcined
at high temperatures in the presence of water. During
this high-temperature steaming, it is desired that at
least 2 weight percent of the atmosphere above the zeolite
be water, preferably more than 10 weight percent, more
preferably greater than 25 weight percent. The most
convenient way to calcine the zeolite is to place the
zeolite which has undergone aqueous ion exchange into an
autoclave and allow steaming to take place under auto
genous pressure. The temperature of the steaming step is
normally above 1000F t538C~, preferably above 1200F
(649C), and most preferably above 1~00F (760C)~ The
time of the steam calcining can range from one-half hour
to twenty-four hours or more.
It appears that the steam calcining causes
3~ aluminum to be removed from the crystal lattice and
silicon to be volatilized to repair the holes left in the
lattice by the aluminum. Thus, the integrity of the
lattice is largely maintained and total collapse is
avoided. Nevertheless, there is some loss o~ crystal-
linity. The steaming also creates gross cracks andfissures in the crystalline particles. I~he aluminum
removed from the lattice appears to form amorphous alumina
deposits in the lattice pores and channels. These
amorphous deposits are removed by the dealumination
procedure. Dealumination typically involves leaching the
~03191
01 _4_
steamed zeolite with organic chelating agents such as EDTA
or with organic or inorganic acids. Dilute inoryanic
05 acids, particularly hydrochloric acid and sulfuric acid,
are most preferred. Where acids are used, the pH of the
leaching solution is preferably below about 2. It can be
appreciated that if the pH is too high, dealumination will
take inconveniently long, while if the acid concentration
1~ is too high, the zeolite's crystal lattice can be
attacked. Typical acid solutions are from about 0.01 N to
about 10 N.
The final zeolitic product will have a smaller
crystal lattice and a higher silica:alumina mole ratio
than is normally obtained. Steamed, dealuminated zeolite
will typically have a cubic cell constant less than
about 24.40 Angstroms and a silica-alumina mole ratio
greater than about 10:1, and most preferably greater than
about 20:1. The final dealuminated product will also have
a higher surface area than the starting material. The
steaming/calcining treatment surprisingly also improves
the catalytic characteristics of ultrastable Y zeolites.
Even though the alumina content of the steamed/leached
faujasitic zeolites is very low compared to the starting
materials, they retain surprising activity and they gain
significant selectivity for the valuable middistillates.
The final catalyst composite includes both the
faujasitic zeolite and an inorgànic oxide matrix. Inor-
ganic oxides are standard supports for zeolites used in
hydroprocessing and ~an include alumina, silica, magnesia,
titania, and combinations thereof. The preferred support
is alumina. A wide variety o~ procedures can be used to
combine the zeolite with the refractory oxide. For
example, the zeolite can be mulled with a hydrogel of the
oxide followed by partial drying if required and extruding
or pelletizing to form particles of the desired shape.
Alternatively, the refractory oxide can be precipitated in
the presence of the zeolite. This is accomplished by
increasing the pH of the solution of a refractory oxide
precursor such as sodium aluminate or sodium silicate. As
-
~L2~3~9~
5--
01
described above, the combination can then be partially
dried as desired, tableted, pelleted, extruded, or formed
05 by other means and then calcined, e.y., at a temperature
above 600F (316C), usually above 800F (427C).
Processes which produce larger pore size supports are
preferred to those producinq smaller pore size su~orts
when cogelling. Additionally, if the steamed zeolite is
added to an acidic solution of inorganic oxide precursor,
the leaching step can be carried out in situ in the
cogellation mixture without a separate leaching step.
The catalyst should contain less than about 50,
preferably less than about 30 weight percent of the
lS zeolite based on the dry weight of zeolite and refractory
oxide. However, æeolite content should exceed 0.5 and ls
usually above 2 weight percent.
The final catalyst composite includes at least
one hydrogenation component. The hydrogenation component
is typically a transition or Group IV-A metal, and is
usually a Group VI-B or VIII metal or combination of
metals or their oxides or sulfides.
The hydrogenation components preferably are
molybdenum, tungsten, nickel and cobalt metals, oxides and
sulfides. Preferred compositions contain more than about
5 weight percent, preferably about 5 to about ~0 weight
percent molybdenum or tungsten or both, and at least about
0.5, and generally about 1 to about 15 weight percent of
nickel or cobalt or both, determined as the corresponding
oxides. The catalysts are often presulfided before use as
sulfide form of these metals tends to have higher
activity, selectivity and activity retention.
The hydrogenation components can be added by any
one of numerous procedures. They can be added either to
the zeolite or the support or a combination of both. In
the alternative, ~he Group VIII components can be added to
the zeolite by comulling, impregnation, or ion exchange
and the Group VI components, i.e., molybdenum and
tungsten, can be combined with the refractory oxide by
~ impregnation, comulling or co-precipitation.
~L2~)3~L9I
Ol -6-
The hydrogenation components can be incorporated
at any one of a number of stages in the catalyst prepara-
05 tion. For example, metal compounds such as the sulfides,oxides or water-soluble salts such as ammonium hepta-
molybdate, ammonium tungstate, nickel nitrate and cobalt
sulfate can be added by comulling, impregnation or preci-
pitation to either the zeolite or the refractory oxide or
both before the zeolite is finally calcined and combined
with the support or after its final calcination but be~ore
combination with the refractory oxide. These components
can be added to the finished catalyst particle by impreg-
nation with an a~ueous or hydrocarbon solution of soluble
compounds or precursors.
The hydrocarbonaceous feeds used in these
processes boil primarily above about 600F (316~C).
Preferably, at least about ~0 percent of the feed will
boil between about 700F (371~) and about 1200F
(649Cj. Feedstocks having these characteristics include
gas oils, vacuum gas oils, coker gas oils, deasphalted
residua and catalytic cracking cycle stocks. The feed to
the hydrocracking zone generally contains at least about
5 ppm and usually between about 10 ppm and Ool weight
percent nitrogen as organonitrogen compounds. It can also
contain substantial amounts of mono- or polynuclear
aromatic compounds corresponding to at least about 5, and
generally about 5 to about 40 volume percent aromatics.
Although the catalysts used in these methods
exhibit superior stability, activity and midbarrel selec-
tivity, reaction conditions must nevertheless be
correlated to provide the desired conversion rates while
minimizing conversion to less desired lower-boiling
products. The conditions required to meet these objec-
tives will depend on catalyst activity and selectivity andfeedstock characteristics such as boiling range, as well
as organonitrogen and aromatic content and structure.
They will also depend on the most judicious compromise of
overall activity, i.e., conversion per pass and selec-
tivity. For example, these systems can be operated at
12~3~ ~
01 ~7~
relatively high conversion rates on the order of 70, ~0 oreven 90 percent conversion per pass. However, higher
05 conversion rates generally result in lower selectivity.
Thus, a compromise must be drawn between conversion and
selectivity. The balancing of reaction conditions to
achieve the desired objectives is part of the ordinary
skill of the art.
1~ Reaction temperatures generally exceed about
500F (260C) and are usually above about 600F (316C),
preferably between 600F (316C) and 900F (482C).
Hydrogen addition rates should be at least about ~00, and
are usually between about 2,000 and about 15,000 standard
cubic feet per barrel. Reaction pressures exceed 200 psig
(13.7 bar) and are usually within the range of about 500
to about 3000 ~sig (32.4 to 207 bar). Liquid hourly space
velocities are less than about 15, preferably between
a~out 0.2 and about 10.
The overall convèrsion rate is primarily
controlled by reaction temperature and liquid hourly space
velocity. However, selectivity is generally inversely
proportional to reaction temperature. It is not as
severely affected by reduced space velocities at otherwise
constant conversion. Conversely, selectivity is usually
improved at higher pressures and hydrogen addition rates.
Thus, the most desirable conditions for the conversion of
a specific feed to a predetermined product can be best
obtained by converting the feed at several different
temperatures, pressure, space velocities and hydrogen
addition rates, correlating the effect of each of these
variables and selecting the best compromise of overall
conversion and selectivity.
The conditions should be chosen so that the
overall conversion rate will correspond to the production
of at least about 40 percent, and preferably at least
about 50 percent of products boiling below about 700F
(371C) per pass. Midbarrel selectivity should be such
that at least about 40, preferably at least about 50
percent of the product is in the middistillate range.
~2S)3~9~
Ol -8-
The process can maintain conversion levels in excess of
about 50 percent per pass at selectivities in excess of 60
05 percent to middistillate products boiling between 400F
(204C) and 700F (371C).
My process can be operated as a single-stage
hydroprocessing zone. It can also be the second stage of
a two-stage hydrocracking scheme in which the first stage
l~ removes nitrogen and sulfur from the feedstock before
contact with the middistillate-producing catalyst. My
process can also be the first stage of a multistep
hydrocracking scheme. In operation as the first stage,
the middistillate-producing zone also denitrifies and
desulfurizes the feedstock; in addition, it allows the
second stage to operate more efficiently so that more
middistillates are produced overall than in other process
configurations, This method of operating with the
middistillate-producing zone first~ followed by at least
one further hydroprocessing zone is especially preferred
for increasing middistillate production.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. l illustrates the difference in ~ouling
rate and activity between a standard amorphous catalyst
used to produce middistillates and the catalyst used in my
invention.
FIG. 2 illustrates the effect of the amount of
acid used in the leaching solution on the silica:alumina
mole ratio of the zeolite.
FIG. 3 illustrates the effect of the amount of
acid used in the leaching solution on the crystallinity of
the product zeolite as compared to the reactant steamed
zeolite.
FIGS~ 4, 5, and 6 illustrate the superiority of
my invention in producing middis~illate. Catalysts con-
taining steamedjleached, steamed, and untreated ultra-
stable Y zeolites were contrasted. FIG. 4 illustrates the
higher middistillate diesel yields of the steamed/leached
catalyst; FIG. 5 illustrates the lower heavy naphtha
4~ yields of the steamed/leached catalyst; and FIG. 6
~2()3~9~
01
_9_
illustrates the lower aromatics content of the middis-
tillate produced by the steamed/leached catalyst. The
05 yields are plotted against conversion to below 670F
(354C).
FIG. 7 contrasts the pore size distribution of a
steamed and steamed/leached Y zeolite.
Example 1
A catalyst containing steamed, leached Y zeolite
was compared to a nickel-tungsten, nonzeolitic catalyst
~silica/alumina/titania cogel base) used to prepare mid-
distillates to compare activity and fouling rates. The
zeolitic catalyst contained 15 weight percent zeolite
steamed at 1475F (802C) for 1 hour and washed with 1 N
hydrochloric acid. The zeolite was mulled with alumina
and had a final metals content of 3.9 weight percent
nickel and 20.5 weight percent tungsten. The feed was a
straight-run Arabian Heavy gas oil having the following
characteristics:
AP~ 21.8
Aniline Pt. 173F (78C)
S, Wt. % 2.62
N, ppm 846
Distillation (D-1160),
C:
St/5 332 / 388
10/30 401 / 436
~51
70/90 468 / 4g9
95/EP 514 / 540
Reaction conditions included an LHSV of 0.75; 1400 psig
(96.5 bar); and 5000 SCF H2/bbl feed. The results are
shown in FIG. 1. The zeolitic catalyst is significantly
more active with a significantly lower fouling rate than
the nonzeolitic catalystO
3~L9~
01 -10-
Example 2
A series of experiments was performed to examine
OS the effect o~ acid washing on silica:alumina mole ratio
and product zeolite crystallinity. The starting material
was NH4y having a sodium content of less than 1 percent
and a silica:alumina mole ratio of S.l:l. The zeolite was
steamed for 1 hour at 800C and then acid washed. The
washing solutions ~lere prepared on the basis of volumes of
concentrated hydrochloric acid per gram of the zeolite
(0.8 ml concentrated HCl i5 the stoichiometric amount of
acid required to raise the Sio2/Al2o3 mole ratio from 5:1
to 30:1), the measured acid was then diluted to about 1.1
lS M or 8.5 M and the zeolite was washed ~ith the resulting
solution. The silica:alumina mole ratio was determined by
neutron activation analysis; percent crystallinity was
measured as percent X-ray diffraction intensity relative
to Na-Y. As can be seen from FIGS. 2 and 3, Sio2/Al2o3
mole ratio and percent crystallinity appear to vary
linearly and unexpectedly with the amount of acid used to
prepare the wàshing solution rather than with the strength
of the washing solution~ It is preferred to retain at
least 50 percent crystallinity measured as percent X-ray
2~ diffraction intensity relative to Na-Y in the product as
compared to the starting material.
Example 3
A series of experiments was performed to compare
the products obtained with different zeolitic hydrocrack-
ing catalysts. ~atalysts A and B were steamed for 1 hourat 800C and Catalyst B was leached with 1 N hydrochloric
acid to remove the alumina debris. The zeolites were
composited with alumina by comulling and then extruded.
The extrudate was impregnated with the hydrogenation
3S metals using standard procedures. The catalysts had the
following characteristics:
1~13~91
0 1
Catalyst A B C
Zeolite Steamed Steamed/ Ultra-
05 Ultra- Dealuminatedstable Y
stable Y Ultra-
stable Y
Wt. ~ Zeolite 15 15 15
Wt. ~ Ni 6.9 4.0 3.4
Wt. % W 17.2 21.0 20.0
Cell Constant, A 24.44 24O36 24.58
SiO2:A12O3 5:1 2901 5:1
The catalysts were tested in a one-step hydrocracking
process using a straight-run Arabian Light vacuum gas oil
feed having the foilowing characteristics:
API 22.9
Aniline Pt., F (C) 177.5 (81C)
Sr weight percent 2.15
N, ppm 877
Distillation (D-1160),
C:
St/5 339 / 374
10/30 381 / 413
448
70/90 467 / 503
95/EP 511 / 536
Reaction conditions included an LHSV of 1.2; 1470 psig H2
(101 bar); and 5000 SCF/bbl feed recycle gas.
The highly desirable product characteristics
produced by my process are illustrated in FIGS~ 4, 5, and
6. The steamed/leached zeolite produces significantly
more middistillate ( FIGo 4) of lower aromatics content
(FIG~ 6) than the steamed and untreated Y zeolite.
Additionally, cracking of the feed to naphtha-range
materials occurs to a significantly lower extent (FIG~ 5)
with the steamed/dealuminated zeolite catalyst.
1~319~
-12-
01
Example 4
A steamed (1375F; 746C) Y zeolite having a
05 silica~alumina mole ratio of 5.1:1 was washed with an
approximately 1 M ~Cl solution (0.7~ cc concentrated HCl/g
zeolite). The leached product had a silica:alumina mole
ratio of 9.6:1. The pore size distributions for both
materials (relative pore volume dV/d (log d) as a function
of pore diameter) are contrasted in FIG. 7.
2~
