U.S. patent application number 11/289565 was filed with the patent office on 2006-06-08 for catalytic sulfur removal from a hydrocarbon stream.
This patent application is currently assigned to The University of Western Ontario. Invention is credited to Hugo I. de Lasa.
Application Number | 20060118465 11/289565 |
Document ID | / |
Family ID | 36573005 |
Filed Date | 2006-06-08 |
United States Patent
Application |
20060118465 |
Kind Code |
A1 |
de Lasa; Hugo I. |
June 8, 2006 |
Catalytic sulfur removal from a hydrocarbon stream
Abstract
A process for the catalytic removal of sulfur from a hydrocarbon
stream such as gasoline comprising an organo-sulfur compound such
as a mercaptan or thiol. The catalyst is a silica based zeolite
such as ZSM-5. The process is preferably performed in a downer
reactor with a residence time of between 7 and 30 seconds and a
volumetric particle concentration of between 15 and 40%.
Preferably, substantially all of the sulfur that is removed from
the organo-sulfur compound is in the form of hydrogen sulfide.
Inventors: |
de Lasa; Hugo I.; (London,
CA) |
Correspondence
Address: |
ANISSIMOFF & ASSOCIATES;RICHMOND NORTH OFFICE CENTRE
SUITE 201
235 NORTH CENTRE RD.
LONDON
ON
N5X 4E7
CA
|
Assignee: |
The University of Western
Ontario
|
Family ID: |
36573005 |
Appl. No.: |
11/289565 |
Filed: |
November 30, 2005 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
60632560 |
Dec 2, 2004 |
|
|
|
Current U.S.
Class: |
208/208R ;
208/245; 208/248 |
Current CPC
Class: |
B01D 2257/30 20130101;
B01D 2257/306 20130101; B01D 53/8603 20130101; C10G 29/00
20130101 |
Class at
Publication: |
208/208.00R ;
208/245; 208/248 |
International
Class: |
C10G 29/00 20060101
C10G029/00 |
Claims
1. A process for removing sulfur from a hydrocarbon stream
comprising an organo-sulfur compound, the process comprising: a)
providing a downer reactor; b) providing a hydrocarbon stream
comprising an organo-sulfur compound in the downer reactor at a
temperature in the range of 300 to 500.degree. C.; c) providing a
catalyst comprising a zeolite that promotes removal of sulfur from
the organo-sulfur compound in the downer reactor; d) contacting the
hydrocarbon stream with the catalyst; and, e) converting at least
70% of the organo-sulfur compound to hydrogen sulfide in the downer
reactor.
2. A process according to claim 1, wherein the catalyst has a
residence time in the downer reactor and wherein the residence time
is from 7 second to 30 seconds.
3. A process according to claim 1, wherein at least 75% of the
organo-sulfur compound is converted to hydrogen sulfide in the
downer reactor.
4. A process according to claim 1, wherein the temperature is from
about 375.degree. C. to about 500.degree. C.
5. A process according to claim 1, wherein the catalyst is present
in the downer reactor in a volumetric particle concentration and
wherein the volumetric particle concentration is from about 5 to
about 40%.
6. A process according to claim 1, wherein the hydrocarbon stream
has a fluid velocity in the downer reactor and wherein the fluid
velocity is between 0.5 and 4.0 m/s.
7. A process according to claim 1, wherein the downer reactor has a
length of between 5 and 40 meters.
8. A process according to claim 1, wherein the downer is operated
such that the product of catalyst reactor density and residence
time is between 0.30 and 1.44 g.sub.crystscm.sup.-3.
9. A process according to claim 1, wherein the catalyst is a
microporous hydrophobic zeolite having acidic sites and a tortuous
pore path.
10. A process according to claim 1, wherein the catalyst comprises
ZSM-5.
11. A process according to claim 1, wherein the organo-sulfur
compound is present in the hydrocarbon stream at a concentration of
between 10 and 10000 ppm.
12. A process according to claim 1, wherein the organo-sulfur
compound comprises one or more sulfur atoms and from 1 to 12 carbon
atoms.
13. A process according to claim 1, wherein the organo-sulfur
compound comprises a mercaptan, a thio-ether, a thio-ketone, a
thio-aromatic, a thio-paraffin, or a combination thereof.
14. A process according to claim 1, wherein the organo-sulfur
compound comprises methyl mercaptan, ethyl mercaptan, or diethyl
sulfide.
15. A process according to claim 1, wherein the hydrocarbon stream
comprises a hydrocarbon having from 1 to 20 carbon atoms.
16. A process according to claim 1, wherein the hydrocarbon stream
comprises n-octane.
17. A process according to claim 1, wherein the hydrocarbon stream
is gasoline.
18. A process according to claim 1, wherein the process further
comprises circulating the catalyst between the downer reactor and a
regenerator.
19. A process for removing sulfur from a hydrocarbon stream
comprising an organo-sulfur compound, the process comprising: a)
providing a solids transport reactor; b) providing a hydrocarbon
stream comprising an organo-sulfur compound in the solids transport
reactor at a temperature in the range of 300 to 500.degree. C.; c)
providing a catalyst comprising a zeolite that promotes removal of
sulfur from the organo-sulfur compound in the solids transport
reactor; the catalyst having a catalyst reactor density from 0.02
to 0.12 g.sub.crystcm.sup.-3; d) contacting the hydrocarbon stream
with the catalyst; and, e) converting at least 70% of the
organo-sulfur compound to hydrogen sulfide in the solids transport
reactor in a residence time from 7 to 20 seconds, wherein the
product of catalyst reactor density and residence time is from 0.30
to 1.44 g.sub.crystscm.sup.-3.
20. A process for removing sulfur from a hydrocarbon stream
comprising an organo-sulfur compound, the process comprising: a)
providing a downer reactor having a length of from 10 to 36 meters;
b) providing a hydrocarbon stream comprising gasoline and
comprising 10 to 10000 ppm of an organo-sulfur compound comprising
a mercaptan in the downer reactor at a temperature in the range of
300 to 500.degree. C.; c) providing a catalyst comprising zeolite
ZSM-5 in the downer reactor at a volumetric particle concentration
of from 5% to 40%; d) contacting the hydrocarbon stream with the
catalyst; and, e) converting at least 90% of the organo-sulfur
compound to hydrogen sulfide in the downer reactor in a residence
time of between 7 and 20 seconds.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application claims the benefit of U.S. provisional
patent application Ser. No. 60/632,560 filed Dec. 2, 2004, the
entire contents of which are hereby incorporated by reference.
FIELD OF THE INVENTION
[0002] The invention relates to the catalytic removal of sulfur
from hydrocarbon streams comprising a sulfur containing
hydrocarbon. In particular, the invention relates to the removal of
sulfur compounds such as mercaptans and thiols from hydrocarbon
streams such as gasoline by catalytic conversion to hydrogen
sulfide using a zeolite catalyst such as ZSM-5. The invention may
be practiced using a solids transport reactor and is advantageously
practiced in a downer reactor.
BACKGROUND OF THE INVENTION
[0003] Sulfur in fuels for internal combustion engines is generally
undesirable. In the auto industry, high sulfur levels in gasoline
lead to poisoning of catalytic converters and to corrosion caused
by sulfuric acid compounds. Sulfur in fuel is also detrimental to
the environment. Sulfur dioxide (SO.sub.2) emissions can be traced
directly to the combustion of sulfur species in transportation
fuels. SO.sub.2 emissions are of particular concern given that they
are precursors of acid rain and sulfate aerosol formation, which
contribute considerably to total ambient fine particulate matter.
Recent environmental regulations establish strict sulfur levels in
transportation fuels. As a result, the refining industries are
under constant pressure to achieve more rigorous standards of
product specification for sulfur.
[0004] Fluid Catalytic Cracking (FCC) is used to produce gasoline.
FCC gasoline contributes from 25% to 40% of the total volume of the
"gasoline pool". Gasoline coming from the FCC unit may have sulfur
levels as high as 2000 ppm, and FCC gasoline accounts for more than
90% of the total sulfur in the "gasoline pool". Refineries must
therefore focus on reducing sulfur in FCC gasoline to achieve the
target sulfur levels.
[0005] The hydrotreating of FCC gasoline, or hydrodesulfurization
(HDS), is a post-FCC treatment process performed to achieve low
sulfur levels in gasoline. Catalysts used are
CoMo/.gamma.Al.sub.2O.sub.3 and NiMo/.gamma.Al.sub.2O.sub.3.
Hydrodesulfurization reactions are exothermic and occur
simultaneously with hydrogenation reactions. Hydrogenation
reactions decrease the quality of gasoline since the olefins, which
contribute to the gasoline octane number, are converted into
alkanes (paraffins). It is well known that there are no conditions
in the HDS process where hydrogenation can totally be excluded.
During HDS, sulfur is removed in the form of hydrogen sulfide
(H.sub.2S); however, H.sub.2S reacts with olefins in the fuel,
producing mercaptans. The use of hydrogen adds a significant cost
to gasoline production. Thus, a process that does not require
hydrogen to remove sulfur from FCC gasoline is desirable.
[0006] Some attempts have been made to address the need for an
improved process for removing sulfur from fuels. Methyl mercaptan
(methane thiol) conversion was studied in a fixed bed reactor
containing H-ZSM5 at 482.degree. C. and 1 h.sup.-1 LHSV obtaining
conversions of 99% (Chang C. D., Silvestri A. J., J. of Catalysis,
Vol. 47, 1977, pp 249-259). The products obtained were hydrocarbons
ranging from C.sub.1 to C.sub.11 and H.sub.2S. About 27% of the
carbon feed was undesirably converted to dimethyl sulfide, which
has no commercial value in fuels. This process is not commercially
viable due to the long residence time and the detrimentally high
rate of conversion of gasoline to dimethyl sulfide.
[0007] Collins et al. in U.S. Pat. No. 5,401,391 and U.S. Pat. No.
5,482,617 disclose a dense phase fluidized bed reactor with H-ZSM5
catalyst used to desulfurize hydrocarbons. Organic sulfur compounds
were converted to H.sub.2S. The catalyst used contained 25% of
H-ZSM5 zeolite with a Si/Al molar ratio of 25. The process could be
used to desulfurize either light gases or gasoline streams with
light olefins being upgraded to more valuable gasoline range
materials. Reaction conditions were: 370-450.degree. C., 50-250
psig, and 0.1-2 h.sup.-1 WHSV. Up to 61% of sulfur in the feed was
converted to H.sub.2S. This level of conversion is not high enough
to be commercially useful as an alternative to hydrotreating. The
low overall conversion may be due in part to the significant
non-uniformity of fluid distribution of the gas phase in a dense
phase fluidized bed. Dense phase fluidized beds normally have
catalyst residence times on the order of minutes and gas phase
residence times on the order of a few seconds. The gas phase
residence time in a dense phase fluidized bed is generally too
short to achieve significant levels of conversion with minimal
amounts of intermediate organo-sulfur compounds, especially when
considering a reaction network involving a series-parallel reaction
mechanism such as in the present invention. In addition, the long
residence time of the catalyst in the dense phase fluidized bed
leads to significant irreversible adsorption of reactants and
reaction products on the catalyst and causes the undesirable
formation of aromatic hydrocarbons on the catalyst. Aromatic
hydrocarbons negatively impact gasoline quality and are pre-cursors
to coke formation. Some aromatics, such as benzene, are undesirable
carcinogenic species. Aromatics have to be capped in gasoline at
levels of less than 10%. As a result aromatics are preferably
avoided in commercial gasoline production.
[0008] Desulfurization of thiophene on H-ZSM5 was studied in a
mini-fixed bed reactor using alkanes as co-reactants at
400-500.degree. C. (Yu, S. Y.; Waky, T.,; Iglesia, E., Appl. Catal.
A, Vol. 242, 2003, pp. 111-121). Propane, n-hexane, and n-decane
were used as co-reactants. It was found that thiophene
desulfurization increased with increasing alkane chain size. It was
observed that thiophene does not alter the nature of alkane
reaction pathways on H-ZSM5, but increases the selectivity to
aromatics as a result of the selective reaction of alkane-derived
species with unsaturated fragments formed during thiophene
decomposition and thiophene desulfurization. The aforementioned
problems of fluid distribution, mass transfer limitations, low
overall conversion, and adsorption of reaction products leading to
reduced gasoline quality and coke formation are all disadvantages
of fixed bed systems when used for catalytic de-sulfurization.
[0009] The need therefore still exists for an improved process for
catalytic removal of sulfur from hydrocarbon streams.
SUMMARY OF THE INVENTION
[0010] According to one aspect of the invention, there is provided
a process for removing sulfur from a hydrocarbon stream comprising
an organo-sulfur compound, the process comprising: providing a
downer reactor; providing a hydrocarbon stream comprising an
organo-sulfur compound in the downer reactor at a temperature in
the range of 300 to 500.degree. C.; providing a catalyst comprising
a zeolite that promotes removal of sulfur from the organo-sulfur
compound in the downer reactor; contacting the hydrocarbon stream
with the catalyst; and, converting at least 70% of the
organo-sulfur compound to hydrogen sulfide in the downer
reactor.
[0011] According to another aspect of the invention, there is
provided a process for removing sulfur from a hydrocarbon stream
comprising an organo-sulfur compound, the process comprising:
providing a solids transport reactor; providing a hydrocarbon
stream comprising an organo-sulfur compound in the solids transport
reactor at a temperature in the range of 300 to 500.degree. C.;
providing a catalyst comprising a zeolite that promotes removal of
sulfur from the organo-sulfur compound in the solids transport
reactor; the catalyst having a catalyst reactor density from 0.045
to 0.12 g.sub.crystscm.sup.-3; contacting the hydrocarbon stream
with the catalyst; and, converting at least 70% of the
organo-sulfur compound to hydrogen sulfide in the solids transport
reactor in a residence time from 7 to 20 seconds, wherein the
product of catalyst reactor density and residence time is from 0.36
to 1.44 g.sub.crystscm.sup.-3.
[0012] According to yet another aspect of the invention, there is
provided a process for removing sulfur from a hydrocarbon stream
comprising an organo-sulfur compound, the process comprising:
providing a downer reactor having a length of from 10 to 36 meters;
providing a hydrocarbon stream comprising gasoline and comprising
10 to 10000 ppm of an organo-sulfur compound comprising a mercaptan
in the downer reactor at a temperature in the range of 300 to
500.degree. C.; providing a catalyst comprising zeolite ZSM-5 in
the downer reactor at a volumetric particle concentration of from
5% to 40%; contacting the hydrocarbon stream with the catalyst;
and, converting at least 90% of the organo-sulfur compound to
hydrogen sulfide in the downer reactor in a residence time of
between 7 and 20 seconds.
[0013] Surprisingly, the process of the present invention provides
high conversion of sulfur compounds to hydrogen sulfide at
relatively short residence times with no undesirable thio-paraffin
intermediates such as diethyl sulfide. Short residence times
decrease the size of the reactor, improving the economics of the
process, and reduce the likelihood for secondary olefin
condensation reactions on the catalyst. This in turn reduces the
production of aromatic hydrocarbons, which desirably limits any
detrimental effect of the process on gasoline quality and limits
the formation of coke on the catalyst. The process of the present
invention is particularly advantageously carried out in a downer
reactor, which permits optimization of process conditions over a
range that achieves desirably high conversion while maintaining
high selectivity.
[0014] The process of the present invention may be operated in a
regime that provides the following key advantages: [0015] 1. High
conversion of orgno-sulfur compounds using only a catalyst,
obviating the need for hydrogen; [0016] 2. High selectivity of
converted organo-sulfur compounds to H.sub.2S; [0017] 3. High
levels of olefins due to negligible gasoline cracking and the
surprising formation of low molecular weight olefins due to intra-
and inter-molecular dehydrosulfidation; [0018] 4. Negligible olefin
condensation and coke formation; and, [0019] 5. All of the above
achieved within practical reactor sizes and process conditions.
[0020] The hydrocarbon stream may comprise from 1 to 20 carbon
atoms and may comprise alkanes, alkenes, alkynes, aromatics, or a
combination thereof. Preferably, the hydrocarbon stream comprises a
mixture of paraffins, olefins, cyclo-paraffins and aromatics having
from 6 to 12 carbon atoms. The hydrocarbon stream may comprise, for
example, n-octane. The hydrocarbon stream may comprise gasoline or
light cycle oils such as diesel fuel. Most preferably, the
hydrocarbon stream is gasoline. The gasoline may be the product of
an FCC process.
[0021] The organo-sulfur compound may comprise one or more sulfur
atoms and from 1 to 12 carbon atoms. Preferably, the organo-sulfur
compound comprises one or two carbon atoms and from 2 to 10 carbons
atoms, more preferably from 2 to 8 carbon atoms, more preferably
from 2 to 4 carbon atoms. The sulfur may be bonded to one or two
carbon atoms. The organo-sulfur compound may comprise a mercaptan,
a thio-ether, a thio-ketone, a thio-aromatic, a thio-paraffin, or a
combination thereof. Preferably, the organo-sulfur compound
comprises methyl mercaptan, ethyl mercaptan, or diethyl sulfide.
The organo-sulfur compound may be present in the hydrocarbon stream
at a concentration of from 10 to 10000 ppm, preferably from 50 to
5000 ppm, more preferably from 75 to 2000 ppm, still more
preferably from 100 to 1000 ppm, yet more preferably from 120 to
500 ppm, even more preferably from 150 to 350 ppm.
[0022] The temperature in the reactor is any suitable temperature
for promoting catalytic dehydrosulfidation of the organo-sulfur
compound. Dehydrosulfidation is the process by which sulfur is
removed from an organo-sulfur compound as hydrogen sulfide, for
example, in a manner similar to the removal of oxygen from methanol
as water in a de-hydration process. Generally, higher temperatures
lead to faster reaction rates, but favour undesirable catalytic
cracking of the hydrocarbon stream. There is therefore an optimal
temperature range for the dehydrosulfidation process. The
temperature may be from 300 to 500.degree. C., preferably from 350
to 475.degree. C., more preferably from 375 to 470.degree. C.,
still more preferably from 400 to 460.degree. C., even more
preferably from 425 to 455.degree. C., most preferably about
450.degree. C.
[0023] The catalyst comprises a zeolite that promotes removal of
sulfur from the organo-sulfur compound, preferably through
dehydrosulfidation. The catalyst includes the zeolite crystallites
and may include a supporting matrix. Any suitable zeolite
crystallite or combination of crystallites may be used as catalyst.
The catalyst may include noble metals, transition metals or
transition metal oxides. The catalyst may include phosphorous. The
catalyst is preferably a microporous hydrophobic zeolite having
acidic sites and a tortuous pore path with a pore diameter of from
5-7 .ANG., preferably about 5.4 .ANG.. The catalyst may comprise a
silico-aluminate comprising silica in an amount of greater than 85%
by weight, preferably greater than 90% by weight, more preferably
greater than 95% by weight. Most preferably, the catalyst comprises
the zeolite ZSM5, which is understood to include H-ZSM5. The
catalyst may comprise a combination of different zeolites.
[0024] The hydrocarbon stream is contacted with the catalyst in the
reactor. Preferably the reactor is a solids transport reactor, for
example a circulating fluidized bed, a riser, or a downer. Most
preferably, the reactor is a downer. The length of the downer may
be from 5 to 40 m, preferably from 10 to 36 m, more preferably from
12 to 30 m, still more preferably from 14 to 27 m. The fluid
velocity of the hydrocarbon stream in the reactor may be from 0.5
to 4.0 m/s, preferably from 1.0 to 3.0 m/s, more preferably from
1.5 to 3.0 m/s, still more preferably from 1.8 to 2.8 m/s.
[0025] The catalyst circulates through the reactor with the fluid
and has a residence time within the reactor. The catalyst residence
time and the fluid residence time in the reactor are similar,
preferably within about 20% of one another. The catalyst residence
time should be less than about 30 seconds to prevent undesirable
coke formation on the catalyst and should be greater than about 7
seconds to provide the desired level of conversion while allowing
reasonable quantities of catalyst to be used. The catalyst
residence time may be from 7 to 30 s, preferably from 8 to 20 s,
more preferably from 9 to 18 s, still more preferably from 12 to 15
s.
[0026] Generally, conversion is related to residence time and
catalyst reactor density, which is the mass of active catalyst
crystallites in a given volume of reactor. The catalyst reactor
density is preferably from 0.02 to 0.12 g.sub.crystcm.sup.-3, more
preferably from about 0.04 to 0.10 g.sub.crystcm.sup.-3, still more
preferably from about 0.05 to 0.09 g.sub.crystcm.sup.-3 The product
of residence time and catalyst reactor density is a constant for a
given level of conversion. Preferably, the product of residence
time and catalyst reactor density is from 0.30 to 1.44
g.sub.crystscm.sup.-3, more preferably from 0.60 to 1.00
g.sub.crystscm.sup.-3, still more preferably from 0.65 to 0.85
g.sub.crystscm.sup.-3, yet more preferably from 0.70 to 0.80
g.sub.crystscm.sup.-3.
[0027] The level of conversion of the organo-sulfur compound in the
reactor is at least 70%, preferably at least 75%, more preferably
at least 80%, still more preferably at least 85%, yet more
preferably at least 90%, even more preferably at least 95%, still
even more preferably at least 97%, most preferably at least 99%.
Catalyst reactor density and residence time may be adjusted within
the ranges provided above to achieve the desired level of
conversion in the reactor. Preferably, substantially all of the
sulfur that is removed from the organo-sulfur compound is in the
form of hydrogen sulfide. However, not all of the sulfur in the
organo-sulfur compound is necessarily converted to hydrogen
sulfide; some of the sulfur may remain in the organo-sulfur
compound (for example, thio-aromatics) or alternative reaction
products (for example, sulfur containing coke) may be formed in the
dehydrosulfidation reaction.
[0028] The volumetric concentration of particles, which is the
percentage of the reactor volume occupied by catalyst particles
(matrix and crystallites), may be selected to achieve the desired
conversion at a given residence time. The volumetric particle
concentration is related to the catalyst reactor density by the
apparent density of the catalyst particles. The volumetric
concentration of catalyst should not be so high as to create
non-uniformity of fluid flow, particularly choking in a riser
reactor. The volumetric particle concentration may be from 5 to
40%, preferably from 8 to 30%, more preferably from 10 to 25%,
still more preferably from 14 to 23%.
[0029] When the hydrocarbon stream contains a mixture of paraffins,
olefins, and aromatics, and particularly when the organo-sulfur
compound is a thio-aromatic, there may be a tendency for coke
formation on the catalyst. The catalyst may be regenerated through
combustion of coke. The regenerator may be a dense phase fluidized
bed reactor. The catalyst may circulate continuously between the
solids transport reactor and the regenerator.
[0030] Further features of the invention will be described or will
become apparent in the course of the following detailed
description.
BRIEF DESCRIPTION OF THE DRAWINGS
[0031] In order that the invention may be more clearly understood,
embodiments thereof will now be described in detail by way of
example, with reference to the accompanying drawings, in which:
[0032] FIG. 1 shows conversions of ethyl mercaptan versus time with
a feed composition of 10 wt. % (10000 ppm) EM in nC.sub.8;
[0033] FIG. 2 shows conversions of ethyl mercaptan versus time with
a feed composition of 5 wt. % (5000 ppm) EM in nC.sub.8;
[0034] FIG. 3 shows conversions of n-octane versus time with a feed
composition of 100 wt. % nC.sub.8;
[0035] FIG. 4 shows conversions of n-octane versus time with a feed
composition of 10 wt. % EM in nC.sub.8;
[0036] FIG. 5 shows conversions of n-octane versus time with a feed
composition of 5 wt. % EM in nC.sub.8;
[0037] FIG. 6 shows the relationship between catalyst density and
reaction time for a conversion of 95% at 450.degree. C.; and,
[0038] FIG. 7 shows a schematic representation of an embodiment of
a process according to the present invention.
DESCRIPTION OF PREFERRED EMBODIMENTS
[0039] Ethyl mercaptan reacts over H-ZSM5 as follows: [0040] 1.
Intra-molecular dehydrosulfidation. Ethyl mercaptan reacts via
intra-molecular dehydrosulfidation to give ethylene and H.sub.2S,
CH.sub.3-CH.sub.2SH.revreaction.CH.sub.2=CH.sub.2+H.sub.2S (1)
[0041] 2. Inter-molecular dehydrosulfidation between two mercaptan
reacting molecules. This reaction leads to the removal of H.sub.2S
molecule from two mercaptan molecules yielding diethyl sulfide
(DiE-S) and H.sub.2S:
CH.sub.3-CH.sub.2SH+CH.sub.2SH-CH.sub.3.revreaction.CH.sub.3-CH.sub.2-S-C-
H.sub.2-CH.sub.3+H.sub.2S (2) [0042] Moreover, following, this
first step, a second step involves further intra-molecular
dehydrosulfidation of diethyl sulfide (DiE-S) yielding an olefin
(butene) and H.sub.2S:
CH.sub.3-CH.sub.2-S-CH.sub.2-CH.sub.3.revreaction.CH.sub.3-CH.sub.2=CH.su-
b.2-CH.sub.3+H.sub.2S (3)
[0043] Equilibrium constants and equilibrium compositions of the
proposed set of three simultaneous reactions for the
dehydrosulfidation of ethyl mercaptan (EM) can be considered at set
operation conditions: temperature, pressure, reactant
concentration. The theoretically calculated chemical equilibrium
constants for the three reactions are reported in Table 1.
Reactions 1 and 3 are endothermic reactions while reaction 2 is
exothermic. All equilibrium constants are higher than one, thus all
three reactions are favored under the selected operating
conditions.
[0044] Regarding the three reactions involved it should be
mentioned that reactions 1 and 2 are of the competitive type while
reactions 2 and 3 are in series. Thus, the overall reaction scheme
of dehydrosulfidation is a combined parallel-in series reaction
called, in the context of this invention, "parallel-series"
reaction network with the three main reaction steps contributing to
the final product distribution. TABLE-US-00001 TABLE 1 Equilibrium
constants in the 350-500.degree. C. temperature range for the EM
dehydrosulfidation reactions (equations 1, 2 and 3). T [.degree.
C.] K.sub.1 (eq. 1) K.sub.2 (eq. 2) K.sub.3 (eq. 3) 350 4.4 3.0
412.7 375 8.1 2.9 695.0 400 14.7 2.8 1136.2 425 25.4 2.8 1808.8 450
42.8 2.7 2811.9 475 70.0 2.7 4278.6 500 111.7 2.6 6385.6
[0045] Observing the calculated equilibrium constants it can be
noticed that for the temperature range of interest 350-500.degree.
C., the equilibrium constants for reaction 1 are consistently
higher than the equilibrium constants for reaction 2, and the
equilibrium constants for reaction 3 much higher than those for
reactions 1 and 2. This is an indication that diethyl sulfide
species are formed and consumed readily in all cases.
[0046] Calculations of expected conversion indicate that high
conversion of mercaptans coupled with elimination of the high
levels of di-sulfide observed in prior art processes should be
achievable. However, the exact conditions needed to achieve the
desired results is not evident. In order to design a process that
achieves the desired conversion, an overall kinetic rate constant
should be determined experimentally under conditions similar to
those expected in actual production. The distribution of reaction
products should also be examined to determine whether the gasoline
is being cracked in the process of removing sulfur from the
hydrocarbon stream. The undesirable formation and accumulation of
aromatics should also be monitored. The goal is to determine a set
of process conditions that achieves high conversion of
organo-sulfur compounds to H.sub.2S without cracking gasoline that
can be implemented in a practically sized reactor.
Experimental Protocol
[0047] Experiments were conducted using the CREC Catalytic
Simulator described in U.S. Pat. No. 5,102,628 (de Lasa), which is
hereby incorporated by reference. The 52 cm.sup.3 CREC Catalytic
Simulator consists of two outer shells, lower section and upper
section that permits to load or to unload the catalyst easily. This
reactor was designed in such way that an annular space is created
between the outer portion of the basket and the inner part of the
reactor shell. A metallic gasket seals the two chambers, an
impeller located in the upper section. A packing gland assembly and
a cooling jacket surrounds the shaft supports the impeller. Upon
rotation of the shaft, gas is forced outward from the center of the
impeller towards the walls. This creates a lower pressure in the
center region of the impeller thus, inducing flow of gas upward
through the catalyst chamber from the bottom of the reactor annular
region where the pressure is slightly higher. The impeller provides
a fluidized bed of catalyst particles as well as intense gas mixing
inside the reactor.
[0048] The CREC Catalytic Simulator operates in conjunction with a
series of sampling valves that allow, following a predetermined
sequence, to inject hydrocarbons and withdraw products in short
periods of time. An Agilent 6890N gas chromatograph (GC) with flame
ionization detector (FID) and a mass selective detector (MSD,
Agilent 5973N) allows the quantification of reaction products using
a capillary column HP-5 Phenyl Methyl Siloxane with a length of 30
m, a nominal diameter of 0.32 mm, and a nominal film thickness of
0.25 .mu.m. More details of the mechanical design of the Catalytic
Simulator are given by Kraemer (Kraemer, D., Modelling Catalytic
Cracking in a Novel Riser Simulator, Ph.D. Thesis, University of
Western Ontario, London, Ontario, 1991) and Pruski (Pruski, J.,
Adsorption Phenomena During FCC in a Novel Riser Simulator, M.E.Sc.
Thesis, University of Western Ontario, London, Ontario, 1996).
[0049] A catalyst comprising ZSM-5 crystallites prepared following
the method of Gabelica, et al. (Gabelica, Z.; Blom, N.; Derouane,
E. G. "Synthesis and Characterization of ZSM-5 Type Zeolites",
Appl. Catal, Vol. 5, 1983, pp. 227-248) was provided in the CREC
Catalytic Simulator. No binder was utilized with the catalyst. The
total mass of catalyst crystallites was about 0.2 g.
[0050] Ethyl mercaptan C.sub.2H.sub.6S (Alfa Aesar, CAS number
75-08-1), was selected as a key chemical species to evaluate the
mercaptan conversion and to assess the reaction network. A
hydrocarbon stream comprising gasoline was simulated with n-octane
(EM Science, CAS number 111-65-9), a straight-chain hydrocarbon
with a boiling point falling in the middle of the gasoline boiling
range.
[0051] Mixtures of these compounds were reacted at different
concentrations (5 and 10 wt. %), at three temperatures (350, 400
and 450.degree. C.) and at four different contact times (10, 20, 40
and 60 s), using a catalyst to oil ratio of C/O=2.5. The total mass
of the injected stream was 0.08 g. All experiments were repeated at
least 3 times to secure reproducibility of results. These
conditions were selected as considered representative of the ones
that could be encountered in a potential industrial post-treatment
process of gasoline dehydrosulfidation.
[0052] Three type of experiments were conducted: [0053] 1. Thermal
runs with pure n-octane, [0054] 2. Thermal cracking runs with 90%
n-octane and 10% ethylmercaptan, [0055] 3. Catalytic runs with ZSM5
and 90-95% n-octane and 5-10% ehtylmercaptan Thermal Runs Using
Pure n-Octane
[0056] To assess the possible thermal effects on n-octane, the
paraffin hydrocarbon species used to model gasoline, thermal runs
were developed in the CREC Catalytic Simulator using a reaction
time of 60 s and three temperatures: 350.degree. C., 400.degree. C.
and 450.degree. C. A sample of 0.08 g of pure n-octane was injected
into the reactor in each run. These conditions were studied to
evaluate the influence of thermal reactions under severe process
conditions (e.g. highest possible conversions).
[0057] Results of the thermal experiments are reported in Table 2.
For each condition, both conversion and mass balance closure are
average values for three repeat runs. Experiments developed with
pure nC.sub.8 at both 350.degree. C. and 400.degree. C. and 60 s,
the largest reaction time studied, showed no chemical species out
of nC.sub.8. This result demonstrates that at 400.degree. C. and
lower thermal levels, there is no significant nC.sub.8 thermal
conversion. Furthermore, at 450.degree. C. and 60 s, there is an
indication of a very small amount of nC.sub.8 being converted to
ethane and propene, with this fraction being limited to 0.34%. In
this product fraction ethene is present in a larger fraction than
propene (0.21 and 0.13 wt. %, respectively, which is about a factor
of two) with this product distribution being characteristic of
thermal cracking where dominant .beta.-scission cracking promotes
ethene as the more abundant product. Table 2 also reports mass
balance closures with these balances that are well in the range of
typical closures achieved in the CREC Catalytic Simulator.
[0058] Thus, on the basis of the data reported, thermal cracking of
gasoline model species is neglected at 450.degree. C. and 60 s.
TABLE-US-00002 TABLE 2 Conversion of n-octane at different
temperatures for thermal runs. Reaction time: 60 s. Feed: 0.08 g of
n-octane. Each conversion and mass balance closure is an average
value of three repeat runs. Temperature Conversion of nC.sub.8 Mass
Balance (.degree. C.) (%) Closure (%) 350 0 5.47 400 0 3.12 450
0.34 7.80
Thermal Runs Using 10 wt. % Ethyl Mercaptan in n-Octane
[0059] In order to investigate EM thermal cracking, runs were
developed in the CREC Catalytic Simulator with a feed containing
both EM and nC.sub.8 (10 wt. % EM/nC.sub.8). Results are summarized
in
[0060] Table 3
[0061] From these runs, it was found that there was no evidence of
conversions of either nC.sub.8 or EM at 350.degree. C. and
400.degree. C. and 60 s. When the temperature was increased to
450.degree. C., there was an indication of some conversion of
nC.sub.8 but with nC.sub.8 conversion remaining below 1%; this was
consistent with the very low conversion observed for thermal
cracking when pure nC.sub.8 was fed to the reactor unit. For EM,
the thermal conversion at these conditions was, however, 7.2%. This
EM conversion was judged to be modest, especially considering that
these experiments were developed under severe conditions
(450.degree. C., 60 s, 10 wt. % EM concentration).
[0062] The foregoing indicates that any chemical species change, as
observed during the catalytic runs, is essentially the result of
the ZSM-5 catalytic activity and that there is minimal influence of
thermal effects. TABLE-US-00003 TABLE 3 Conversion of nC.sub.8 and
EM at different temperatures for thermal runs. Reaction time: 60 s.
Feed: 0.08 g of 10 wt. % EM/nC.sub.8. Each conversion and mass
balance closure is an average value of three repeat runs. Standard
Standard Conver- Deviation Conver- Deviation Mass Temperature sion
of for nC.sub.8 sion of for EM Balance (.degree. C.) nC.sub.8 (%)
Conv. EM (%) Conv. Closure (%) 350 0 0 0 0 9.12 400 0 0 0 0 10.51
450 0.60 0.11 7.23 0.90 8.20
[0063] In spite of this very low thermal reaction contribution it
was interesting to review the products formed, which are presented
in Table 4. For a run at 450.degree. C., 60 s, and 10 wt. %
EM/nC.sub.8 mixture, ethane, propene, and trans-butene are the
detected product species with ethene being the more abundant
species followed by propene and trans-butene.
[0064] It was also observed that ethene was now at higher levels
than in the case of the thermal conversion with nC.sub.8 alone:
0.61 wt. % versus 0.21 wt. %, or three times higher. This suggests
that a good fraction of ethene formed in this case was originated
via EM conversion, a result of intra-molecular H.sub.2S removal.
Propene, on the other hand, increased from 0.13 wt. % to 0.25 wt.
%, or twice, and this points towards a mild sharing of the EM
conversion via a reaction involving possibly intra-molecular
H.sub.2S removal and an alkylation step. Finally, these experiments
also show the formation of trans-butene species, not observed for
the nC.sub.8 thermal conversion, and this strongly also suggests an
EM conversion via inter-molecular removal of H.sub.2S. Thus, the
observed reactor products strongly support both intra-molecular and
inter-molecular reactions contributing to EM conversion under
thermal conditions as described with equations (1)-(3). The
surprising level of olefin formation advantageously results in
increased octane number in the gasoline, rather than the decrease
in octane number that is experienced in conventional HDS
processes.
[0065] Another surprising finding from these experiments is the
absence of diethyl sulfide species, which were expected as a result
of the inter-molecular dehydrosulfidation reaction step. However,
it appears that this species is very reactive, being formed and
consumed very quickly and remaining under non-detectable levels.
This result is in marked contrast to the prior art which produces
high levels of Die-S. TABLE-US-00004 TABLE 4 Product composition in
mass fraction (wi) for thermal run. Reaction conditions: T =
450.degree. C., t = 60 s. Feed: 0.08 g of 10 wt. % EM/nC8. Species
Mass fraction (w.sub.i) ethene 0.0061 propene 0.0025 trans-butene
0.0017 ethyl mercaptan 0.0928 n-octane 0.8947 hydrogen sulfide
0.0040
[0066] While H.sub.2S could not be detected with the FID, it was
positively identified using the GC-MSD and quantified via element
balances.
Catalytic Runs: Conversions of Ethyl Mercaptan
[0067] FIG. 1 reports the EM conversions for the concentration of
10 wt. % of EM in nC.sub.8. It can be observed that the EM
conversion increases progressively with reaction time in the 10-60
s range. In addition there is an increase of catalytic
dehydrosulfidation with temperature.
[0068] Furthermore, and in order to explore the reaction order of
the EM reaction, the EM in nC.sub.8 was decreased to 5 wt. %
keeping the C/0 at 2.5. Results of these experiments are reported
in FIG. 2. The lack of dependence of the EM conversions, at various
temperatures and reaction times, with the EM concentration
(comparison between FIGS. 1 and 2) suggests an overall first order
reaction for the catalytic EM conversion.
Catalytic Runs: Conversions of n-Octane
[0069] Conversions of n-octane are reported in FIGS. 3, 4 and 5.
These figures contain results of experiments developed using 100
wt. % of nC.sub.8, 95 wt. % of nC.sub.8 (5 wt. % of EM) and 90 wt.
% of nC.sub.8 (10 wt. % of EM). It can be noticed that the runs
showed an expected and progressive increase of the nC.sub.8
conversion with reaction time and temperature. Conversions levels,
however, for the nC.sub.8 remained at much lower levels than the EM
conversion. For instance, at 450.degree. C., 60 s and C/O=2.5 the
observed nC.sub.8 conversion was 22.87% (0.5 S.D.) versus 50.85%
(0.91 S.D.) for EM at the same conditions.
[0070] These results show that there is a strong competition for
the acid sites of the ZSM-5 catalyst promoting both
dehydrosulfidation and catalytic cracking. It appears that, given
the significant differences in gas phase concentrations between
nC.sub.8 and EM (between 10 and 20 times), there is either a
greater affinity of EM adsorption versus nC.sub.8 adsorption or,
alternatively, a much faster intrinsic rate of EM
dehydrosulfidation versus the one for nC.sub.8 cracking.
Catalytic Runs: Product Distribution
[0071] In the catalytic runs using pure nC.sub.8, the obtained
amount of propene was always higher than the amount of ethene, with
this being a typical characteristic of catalytic cracking.
Trans-butene was only formed at 400.degree. C. and 450.degree. C.
and 40 and 60 s and the amounts produced were very low.
TABLE-US-00005 TABLE 5 Product distribution of key species for
reactions using 10 wt. % EM/nC8. Compositions expressed in wt. %
Time (min): 10 20 40 60 350.degree. C. ethene 0.07 0.21 0.46 0.80
propene 0.41 0.71 1.18 1.47 trans-butene 0.00 0.04 0.10 0.19
H.sub.2S 0.34 0.58 1.10 1.42 400.degree. C. ethene 0.20 0.37 0.66
0.80 propene 0.78 1.46 2.80 3.83 trans-butene 0.04 0.08 0.16 0.22
H.sub.2S 0.63 1.12 1.99 2.58 450.degree. C. ethene 0.64 1.06 1.55
1.99 propene 1.35 2.71 4.88 7.14 trans-butene 0.14 0.25 0.40 0.51
H.sub.2S 1.06 1.92 2.86 3.78
[0072] Product composition, as reported in Table 5, for the
catalytic runs using mixture of EM and nC.sub.8 shows trans-butene
levels produced being higher than those obtained from runs using
pure nC.sub.8. This indicates that there is formation of
trans-butene via the inter-molecular de-hydrosulfidation reaction.
Hydrogen sulfide amounts increase with residence time and
temperature, with this being in agreement with the rise of EM
conversion with these two parameters. Furthermore, the amounts of
ethene observed were also higher for the catalytic runs using
mixtures of EM and nC.sub.8 than for the corresponding cases using
pure nC.sub.8. This gives a good indication of formation of ethene
via the intra-molecular dehydrosulfidation reaction.
[0073] The ratios of tC.sub.4.sup.=/C.sub.2.sup.=
(trans-butene/ethene), which provide valuable insights on the
relative importance of inter-molecular and intra-molecular
catalytic dehydrosulfidation, are reported in Table 6.
TABLE-US-00006 TABLE 6 Trans-butene to ethene
(tC.sub.4.sup.=/C.sub.2.sup.=) ratios for catalytic experiments. T
= 350.degree. C. nC.sub.8 EM (5 wt. %) EM (10 wt. %) Time (min)
tC.sub.4.sup.=/C.sub.2.sup.= tC.sub.4.sup.=/C.sub.2.sup.=
tC.sub.4.sup.=/C.sub.2.sup.= 10 -- 0 0 20 -- 0.1976 0.1839 40 --
0.3087 0.2187 60 -- 0.322 0.2373 T = 400.degree. C. nC8 EM (5 wt.
%) EM (10 wt. %) Time (min) tC.sub.4.sup.=/C.sub.2.sup.=
tC.sub.4.sup.=/C.sub.2.sup.= tC.sub.4.sup.=/C.sub.2.sup.= 10 --
0.3322 0.2031 20 0 0.395 0.2074 40 0.117 0.3934 0.244 60 0.2357
0.4197 0.274 T = 450.degree. C. nC8 EM (5 wt. %) EM (10 wt. %) Time
(min) tC.sub.4.sup.=/C.sub.2.sup.= tC.sub.4.sup.=/C.sub.2.sup.=
tC.sub.4.sup.=/C.sub.2.sup.= 10 0 0.324 0.216 20 0.1059 0.372
0.2354 40 0.1967 0.394 0.2544 60 0.2335 0.4128 0.2559
[0074] It can be noticed that catalytic nC.sub.8 conversion leads
to tC.sub.4.sup.=/C.sub.2.sup.= ratios much smaller than the ones
observed for the catalytic conversion of EM-nC.sub.8 mixtures. For
instance for the catalytic conversion of nC.sub.8 at 350.degree.
C., this ratio is zero with no tC.sub.4.sup.= being detected; this
ratio also remains at small levels at 400.degree. C. and
450.degree. C. Thus, it can be argued on this basis that at
350.degree. C. olefin dimerization under these conditions is
negligible, becoming somewhat more prevalent at 400.degree. C. and
450.degree. C.
[0075] On the other hand, the tC.sub.4.sup.=/C.sub.2.sup.= ratios
increase considerably when the EM/nC.sub.8 mixtures are contacted
with the catalyst and this difference suggests an increased
influence of the competitive conversion of EM via the
inter-molecular path, with more tC.sub.4.sup.= species being
formed. There is evidence that increasing the temperature leads to
a more significant inter-molecular EM conversion. The surprising
levels of olefin formation observed in the presence of the catalyst
is another advantage of this process. TABLE-US-00007 TABLE 7
Product composition for catalytic dehydrodesulfidation in the CREC
Catalytic Simualtor. T = 450.degree. C., t = 60 s, C/O = 2.5,
Reactants: 5% of EM in nC.sub.8 Chemical Species Compositon (wt %)
Ethene 0.015 Propene 0.0817 Iso-butane 0.028 n-butane 0.053
t-butene 0.00626 2-methybutane 0.0105 1-pentene 0.0014 n-pentane
0.0137 Ethyl-mercaptan 0.0266 Cis-1,2-dimethycyclopropane 0.052
2-methy-pentane 0.0026 1-hexene 0.0014 n-hexane 0.0023 Benzene
0.0015 Toluene 0.0093 n-octane 0.794 Xylenes 0.0088 Hydrogen
sulfide 0.0232
[0076] Referring to Table 7, the only reaction contributing
significantly to the formation of C.sub.2 and C.sub.4 species is
the catalytic dehydrosulfidation of gasoline with minimum
contribution of gasoline cracking (less than 5%). The C.sub.2 and
C.sub.4 species produced are predominantly olefins (eg: ethene and
propene) that have significant petrochemical and commercial value.
With nC.sub.8 as a feed stock, there were essentially no C.sub.6
aromatics (eg: benzene, toluene, xylene) produced, and no coke
formation observed. It is also worthwhile to note that essentially
no diethyl sulfide was produced, contrary to the prior art and
providing evidence of the desired high selectivity towards hydrogen
sulfide.
[0077] In a commercial process where the hydrocarbon stream is a
mixture of parafins, olefins and aromatics, there is some coke
formation expected. In a preferred embodiment, the process is
developed concurrently with catalyst regeneration where the coke,
deactivating the catalyst, is combusted in a dense phase fluidized
bed and the overall process is envisioned with the catalyst
continuously circulating between two twin reactors: a solids
transport reactor for catalytic dehydrosulfidation of gasoline,
preferably a downer or riser reactor; and, a dense phase fluidized
bed for coke combustion and catalyst regeneration. A schematic
representation of an embodiment of this process is provided in FIG.
7 with a downer reactor. FIG. 7 will be described in greater detail
hereinafter.
[0078] Furthermore, the process of catalytic dehydrosulfidation
should not be considered limited to mercaptan dehydrosulfidation.
This is quite critical given that there is a significant fraction
of sulfur contained in gasoline in the form of thio-aromatic
compounds. For instance, one could envision intermolecular
dehydrosulfidation of one mercaptan molecule reacting with one
thio-aromatic molecule such as benzo-thiophene. In this case the
products of dehydrosulfidation will yield molecules with a critical
molecular diameter larger than the narrow 5.4 .ANG. channels of the
ZSM5 zeolite, with these coke precursors remaining trapped in the
ZSM-5 zeolite porous network and increasing the need for catalyst
regeneration.
[0079] Given the foregoing experimental results and analysis, the
following general observations and conclusions can be reached:
[0080] 1. Gas phase and catalytic phase residence times should be
very close to achieve the desired high conversion of organo-sulfur
compounds and selectivity to H.sub.2S; [0081] 2. The catalytic
reactor operates under close to isothermal conditions; [0082] 3.
The EM reaction exhibits first-order kinetics; and, [0083] 4. The
series-parallel reaction network dominates the progress of the
catalytic dehydrosulfidation reaction. This implies long gas phase
residence times (>7 s) to achieve the desired conversion levels.
Reactor Design Calculations
[0084] The CREC Catalytic Simulator design equation can,
considering the various special design features of this unit (batch
operation, well fluidized catalyst contained between two grids,
isothermal operation, high re-circulation of chemical species) be
expressed as follows: d C EM d t = r EM .times. W c V r ( 4 )
##EQU1## where W.sub.c is the weight of the zeolite crystallites
and V.sub.r is the reactor volume (cm.sup.3).
[0085] If a first order overall kinetics for EM conversion, as
suggested by the experimental data in the CREC Catalytic Simulator,
is assumed it results: d C EM d t = kC EM .times. W c V r ( 5 )
##EQU2##
[0086] Equation 5 can be written in terms of ethyl-mercaptan
fractional conversion and following integration it gives: - ln
.function. ( 1 - X EM ) = k .times. .times. W c V r .times. t ( 6 )
##EQU3## where X.sub.EM is the fractional conversion of EM.
[0087] A linear regression can be applied to equation 6 to find the
slope (kW.sub.c/V.sub.r). Knowing that the loading of zeolite
crystallites was 0.2 grams and that the CREC Catalytic Simulator
volume was 52 cm.sup.3, the value of k may be calculated.
[0088] Table 8 reports the obtained values for k with 95%
confidence bounds. TABLE-US-00008 TABLE 8 Values of the kinetic
constant at 350.degree. C., 400.degree. C. and 450.degree. C.
Temperature k (cm.sup.3/g.sub.cryst s) .+-.C.L. (95%) R.sup.2
350.degree. C. 1.11 0.0424 0.9624 400.degree. C. 2.05 0.1169 0.8574
450.degree. C. 3.88 0.1956 0.9184
[0089] An Arrhenius relationship k=k.sub.0exp(-E/RT) can be adopted
to express the temperature effect. This equation can be modified
using re-parametrization to reduce cross-correlation between the
activation energy and the pre-exponential factor. This is
accomplished employing a central temperature (T.sub.0), the average
value of the three temperatures used in this research. Thus,
equation 6 becomes, X EM , model = 1 - exp .function. [ - W c V r
.times. t .function. ( k o ' .times. .times. exp .function. ( - E R
.times. ( 1 T - 1 T o ) ) ) ] ( 7 ) ##EQU4##
[0090] A non-linear regression with minimization of residuals
.SIGMA.(X.sub.EM,exp-X.sub.ETM,model).sup.2 was developed using a
MatLab program and initial guesses for k as reported in Table
1.
[0091] The resulting kinetic parameters with their confidence
limits at the 95% were activation energy (E=46.66.+-.34.84 KJ/mole)
and pre-exponential factor (k.sub.0.sup.1=2.136
cm.sup.3/g.sub.crysts).
[0092] The following design equations apply for first order
kinetics in the CREC Catalytic Simulator
EMConversion=1-exp(-k.delta.'.sub.catt) (8) and
k=k.sub.0.sup.1exp[-E/R(1/T-1/T.sub.0)] (9)
[0093] with k.sub.0.sup.1 representing the kinetic constant at
400.degree. C., E the energy of activation (KJ/mole), t the
reaction time (s) and .delta.'.sub.cat the catalyst reactor density
(g.sub.cryst/cm.sup.3 of reactor).
[0094] Using equation (9), with T=450.degree. C. and
T.sub.0=400.degree. C. results in an observed kinetic constant k
for the CREC Catalytic Simulator of 3.809 cm.sup.3/g.sub.crysts.
With 60 seconds total reaction time, C/O=2.5 and T=450.degree. C. a
conversion of 56% is calculated for the CREC Catalytic Simulator
using the above parameters. This compares favourably with the
actual conversion observed in FIG. 1.
[0095] The design of a full-scale reactor for use in the process of
the present invention can be based on the following assumptions:
[0096] 1. An essentially plug flow pattern for both gas and solid
phases; [0097] 2. A negligible velocity difference between gas and
particles, thus gas and particles can be assumed flowing through
the unit with the same velocities; [0098] 3. A negligible variation
of particle and solid velocities across the reactor cross section;
[0099] 4. An isothermal operation, given the relatively low total
enthalpy changes involved in the dehydrosulfidation process and the
high radial mixing of solid particles; and, [0100] 5. Similar
residence times for catalyst and reactants in the range of less
than 30 seconds.
[0101] Given the above constraints, there are two basic types of
solids transport reactors that can be chosen: risers and downers.
Both types of reactors circulate the catalyst continuously, but the
flow direction in risers is upward, whereas the flow direction in
downers is downward.
[0102] On this basis the EM balance for the reactor can be
expressed under steady state operation as follows: u g .times. d C
EM d z = r EM .times. .rho. c ( 10 ) ##EQU5## where .rho..sub.c is
the catalyst reactor density (expressed as mass of zeolite
crystallites per unit reactor volume) and z is the axial reactor
length in m and u.sub.g represents the gas interstitial velocity in
the reactor in m/s.
[0103] If a first order overall kinetics for EM conversion, as
suggested by the experimental data in the CREC Catalytic Simulator,
is assumed it results: d C EM d t R = - k .times. .times. C EM
.times. .rho. c ( 11 ) ##EQU6## where t.sub.R represents the gas
phase residence time.
[0104] Equation 11 can be alternatively written in terms of
ethyl-mercaptan fractional conversion and following integration it
gives: -ln(1-X.sub.EM)=k.rho..sub.ct.sub.R (12)
[0105] Thus, one can postulate that the EM conversion in the
reactor for catalytic dehydrosulfidation of gasoline can be
expressed as X EM , model = 1 - exp .function. [ - .rho. c .times.
t R .function. ( k o ' .times. exp .function. ( - E R .times. ( 1 T
- 1 T o ) ) ) ] ( 13 ) ##EQU7##
[0106] By using in equation 12 an EM conversion of 95% with k=3.809
cm.sup.3/g.sub.crysts (for the parameters T=450.degree. C.,
E=46.66.+-.34.84 KJ/mole, k.sub.0.sup.1=2.13.+-.0.05
cm.sup.3/g.sub.crysts), it can be determined that
.rho..sub.ct.sub.R is a constant equal to 0.786
g.sub.crysts/cm.sup.3. Thus, one could secure the desired 95% EM
conversion by selecting operating conditions that satisfy the
relationship .rho..sub.ct.sub.R=0.786 g.sub.crysts/cm.sup.3.
[0107] For example, if one chooses a residence time in the reactor
of 12 seconds (gas superficial velocity=1.8 m/s, total reactor
length=21.6 m) the required catalyst reactor density is 0.063
g.sub.cryst/cm.sup.3 reactor. Given than one should expect that, in
a catalyst particle, the ratio between the total weight
(matrix+crystallites) of 60 micron pellets and the weight of
zeolite crystallites to be about 3 g.sub.particle/g.sub.cryst, the
overall catalyst density is about 0.189 g.sub.particle/cm.sup.3
reactor. Even more, assuming that the density of the pellet is
about 1 g.sub.particle/cm.sup.3 catalyst, this yields an apparent
volumetric concentration of 0.189 cm.sup.3 catalyst/cm.sup.3
reactor. Finally, and considering that an estimated 10% of the gas
volumetric follow in the downer is inert gas used to carry the
particles from the regenerator via pneumatic transport, a 0.9
correction is needed. This results in a volumetric particle
concentration of 0.17 cm.sup.3 particle/cm.sup.3 reactor or 17% for
achieving 95% EM conversion.
[0108] This high volumetric particle concentration is not
achievable in a riser reactor, given the constraints imposed on the
volumetric particle concentration by the choking of the suspension.
Referring to Kerry and Knowlton ("Wall Solid Upflow and Down Flow
Regimes in Risers for Group A Solids", Proceedings of the
Circulating Fluidized Bed Technology VII Conference, Niagara Falls,
2002, eds J. Grace., J. Zhu and H. de Lasa, pp. 310-316), FIG. 6
shows a choking velocity correlation for 67 micron FCC particles of
group A; these particles are very similar to the ones considered
for use in the present invention. At a superficial gas velocity of
2.8 m/s, the highest volumetric particle concentration before onset
of choking is 3.5%. At a superficial gas velocity of 1.8 m/s, the
highest volumetric particle concentration before onset of choking
is 1.1%. This is nearly an order of magnitude lower than the
volumetric particle concentration of 17% needed to achieve 95% EM
conversion. A riser reactor is therefore unsuitable for achieving
the desired combination of process conditions.
[0109] However, this high volumetric particle concentration is
perfectly achievable in a downer reactor as reported by Jin, et al.
(Y. Jin, Yu Zheng, F. Wei "State of the Art Review of Downer
Reactors", Proceedings of the Circulating Fluidized Bed Technology
VII Conference, Niagara Falls, 2002, eds J. Grace, J. Zhu and H. de
Lasa, pp. 40-60). Particles move in a downward direction in
downers, the same direction as gravity, not against gravity as in a
riser unit. In a downer unit there is no possible choking of the
suspension, so very high solids/gas loading ratios and volumetric
particle concentrations may be achieved. A practical upper limit of
volumetric particle concentration in a downer is 40%, which permits
the 17% value calculated above for 95% EM conversion.
[0110] The relationship between catalyst reactor density and
residence time is illustrated in FIG. 6 for
.rho..sub.ct.sub.R=0.786 g.sub.crysts/cm.sup.3, which corresponds
to 95% conversion at T=450.degree. C. with k=3.809
cm.sup.3/g.sub.crysts. Limit 1 represents the practical limitation
on residence time that results in a practical reactor length. A
residence time limit of 20 seconds corresponds to a reactor length
of 36 m. A preferred reactor length for this process is between
about 14 and 22 metres. Limit 2 represents the maximum catalyst
reactor density that can be practically utilized in a downer
reactor while still satisfying the fluid dynamics assumptions
listed above. Limit 2 corresponds to about 0.12
g.sub.cryst/cm.sup.3 reactor, or about 40% volumetric particle
concentration in the reactor. For comparison, a volumetric particle
concentration limit of 3.5% in riser reactors before the onset of
choking corresponds to 0.001 g.sub.cryst/cm.sup.3, which places
Limit 2 far to the right of Limit 1. Therefore, 95% conversion is
simply not achievable in a riser reactor. Thus, the downer reactor
is the only practical reactor system that provides the unique range
of conditions (high solid concentration, moderate gas superficial
velocity compatible with a proper total reactor length) desired to
achieve the targets for conversion and selectivity.
[0111] Calculations were conducted using the above equations and
are summarized in Table 9, which discloses design parameters for
several embodiments of the process according to the present
invention. As can be seen from the Table, high conversions can be
achieved with reasonable residence times and reactor lengths. The
volumetric concentration of catalyst in all cases is suitable for a
downer reactor. By using the parameters disclosed in Table 9, high
selectivity towards hydrogen sulfide should also be achieved in
each embodiment. TABLE-US-00009 TABLE 9 Catalytic
Dehydrosulfidation Conditions in a Downer Reactor at various
conversions and gas residence times. T = 450.degree. C., k = 3.809
cm.sup.3/g.sub.cryst s. Volumetric Particle Reactor Conversion
t.sub.R Concentration .rho..sub.c .rho..sub.c t.sub.R Length (%)
(s) (%) g.sub.cryst/cm.sup.3 g.sub.cryst s/cm.sup.3 (m) 70 9 9.45
0.035 0.315 16.2 80 9 12.6 0.0468 0.412 16.2 90 9 18.1 0.067 0.603
16.2 95 9 23.4 0.0868 0.781 16.2 70 12 7.1 0.0263 0.3156 21.6 80 12
9.5 0.0352 0.422 21.6 90 12 13.6 0.0505 0.606 21.6 95 12 17.6
0.0655 0.786 21.6 70 15 5.69 0.0211 0.316 27 80 15 7.58 0.0281
0.421 27 90 15 10.91 0.0404 0.606 27 95 15 14.17 0.0525 0.787
27
[0112] Referring to FIG. 7, a process of the present invention is
schematically represented. Streams of untreated gasoline 5 and
ZSM-5 catalyst 10 are introduced into a downer reactor 15 near the
top where the streams are co-mingled. As the co-mingled streams
pass through the downer reactor, sulfur of organo-sulfur compounds
in the untreated gasoline is catalytically converted to hydrogen
sulfide.
[0113] The co-mingled stream exits the downer reactor 15 at the
bottom and is fed to a first cyclone separator 25 where treated
gasoline is separated from catalyst. Treated gasoline is recovered
in treated gasoline stream 30 whereas separated catalyst stream 35
is combined with a first carrier fluid stream 40 (e.g. an inert gas
or steam) and fed to a dense phase fluidized catalyst regenerator
45.
[0114] An air stream 50 is introduced into the regenerator 45 and
any coke is combusted in the presence of the air to regenerate the
catalyst. Combustion gases are removed from the top of the
regenerator in combustion gas stream 60. A regenerated catalyst
stream 65 exits the regenerator at the bottom and is combined with
a second carrier fluid stream 70 (e.g. an inert gas or steam) and
fed to a second cyclone separator 75 where most (about 90% or more)
of the carrier fluid is separated from the catalyst. Separated
carrier fluid is recovered in separated carrier fluid stream 80 and
catalyst is reintroduced into the downer reactor in catalyst stream
10.
[0115] Other advantages which are inherent to the structure are
obvious to one skilled in the art. The embodiments are described
herein illustratively and are not meant to limit the scope of the
invention as claimed. Variations of the foregoing embodiments will
be evident to a person of ordinary skill and are intended by the
inventor to be encompassed by the following claims.
* * * * *