Note: Descriptions are shown in the official language in which they were submitted.
001 -1-
002 EXTENDED CYCLE REGENERATIVE REFORM~NG
004 FIELD OF T~E INVENTION
005 The present invention is dlrected to the catalytic
006 reforming of hydrocarbon fractions. More speciflcally, the
007 present invention is concerned with reforming in a plurality of
008 reaction zones in series to improve the octane rating of the
009 feed.
010 BACXGROUND OF T~E INVENTI~N
011 Reforming of a naphtha fraction is generally
012 accomplished by passing the naphtha through a plurality of
013 reaction zones in series, each zone containing a catalyst com-
014 prising a hydrogenation-dehydrogenation component supported on
015 a porous solid carrier. Typlcal catalysts include platinum on
016 alumina with or without such promoters such as rhenium, tin,
017 iridium, etc. The naphtha fraction to be reformed is contacted
018 in the first reaction zone with a platinum-containing catalyst
019 at reaction conditions to convert principally naphthenes to
020 aromatics. In addition to naphthene dehydrogenatlon, side
021 reactions such as isomerization, hydroisomerization and hydro-
022 cracking may also occur. Typically, the effluent from the
023 first reaction zone is heated prior to being introduced to a
024 subsequent reaction zone.
025 After a period of use in reforming, the catalyst
026 becomes gradually deactivated due to the deposition of coke on
027 the surface of the catalyst and consequently a decrease of the
028 octane values of the reformate product is observed.
029 If the octane requirements imposed upon the par-
030 ticular reforming system are to be continuously met, the
031 reaction temperature of the catalyst must be increased ln order
032 to compensate for the loss in activity due to the coke deposl-
033 tion. The fastest catalyst deactivation occurs in the reactor
034 where paraff in dehydrocyclization and hydrocracking are the
035 principal reactions. Consequently, even with a constant inlet
036 temperature, the average reaction temperature increases with
037 each successive reactor because the reactions in each
001 -2-
002 successive reactor are not as endothermic as in the preceding
003 reactor.
004 Coke deposltion on the catalyst not only decreases
005 the activity of the catalyst -but also results in a decrease in
006 the yield of C5+ gasoline product produced. Thus,.the yield of
007 C5+ gasoline product generally decllnes throughout the
008 reforming process untll it reaches an unacceptable level, at
009 which point common practice is to regenerate all or part o the
010 catalyst. Typical coke levels on the catalyst at the tlme of
011 regeneration are 10 to 12 weight percent or more on the
012 catalyst in the last reactor and 5 or 6 weight percent on the
013 catalyst in the first reactor. Coke levels on catalysts in
014 intermediate reactors will generally fall between these two
015 figures.
016 Regeneration procedures, whether continuous or batch
017 operations, are generally well known to the art. Such
018 procedures generally involve several steps: a car5On burn-off
019 by contacting the catalyst with oxygen-containing gas at an
020 elevated temperature until substantially all of the carbon is
021 removed; subsequent contacting of the catalyst with an
022 oxygen-containlng gas to redistribute the platlnum group metal;
023 a halogen adiustment and optlonally a reduction of the
024 regenerated catalyst with a hydrogen-containing gas prlor to
025 returning the catalyst to the reactor. See, for instance, U.S.
026 Patent No. 3,496,096.
027 SUMMARY OF T~E INVENTION
028 It is an object of this inventlon to provide a
029 reforming process having an extended operating cycle between
030 regenerations when compared with ordinary reforming processes.
031 It is another object of this invention to provide a method for
032 extending the effective life of a reforming catalyst which is
033 nearing the end of the run.
034 In accordance with one embodiment of the present
035 invention there is provided for a reforming process where~n a
036 naphtha feedstock is contacted at reforming conditions in the
037 presence of hydrogen with a reforming catalyst in a plurality
001 _3_
002 of reaction zones in series, a product of improved octane
003 rating is recovered from the effluent of the last reaction
004 zone, and during the course of said reforming coke deposits
005 upon sai_ catalyst thereby deactivating and necessitating
006 eventual regeneration or replacement thereof, the method of
007 extending the length of service of the catalyst in all reaction
008 zones downstream of the first whlch comprises maintaining the
009 level of coke on the catalyst in said first reaction zone at
010 less than 1~ by weight of the catalyst by adjusting the
011 frequency at whch the catalyst in said first reaction zone is
012 regenerated or replaced with fresh or regenerated catalyst.
013 In accordance with another embodiment of the present
014 invention, there is provided a multiple stage process for
015 catalytically reforminq a naphtha charge stock which comprises:
016 (a) reacting said charge stoc~ in the presence of
017 catalyst and hydrogen at catalytic reforming conditions in a
018 moving bed reaction zone through which the catalyst is movable
019 via gravity flow and recovering the resulting hydrocarbonaceous
020 effluent from the moving bed reactlon zone;
021 (b) maintaining the amount of coke deposited on the
G22 catalyst in said moving bed reaction zone below 1% by weight Dy
023 at least periodically introducing fresh or regenerated catalyst
024 into the upper end of said moving bed reaction zone and at
025 least periodically withdrawing an equivalent amount of coke-con-
026 taminated catalyst from the lower end of said moving bed
027 reaction zone;
028 (c) further reacting the effluent of said moving bed
029 reaction zone at catalytic reforming conditions in a flxed bed
030 reactor system containing at least one fixed bed reaction zone;
031 (d) recovering a normally li~uid catalytically reformed
032 product from the effluent withdrawn from the last fixed bed
033 reaction zone.
034 In accordance with yet another embodiment of the
035 present invention, there is provided a catalytic reformlng
036 process in which a naphtha fraction is contacted at reformlng
~37 conditions in the presence of hydrogen with a reforming
3Zl
001 -~-
002 catalyst in a p`lurality of reaction zones in series, a product
003 of improved octane rating is recovered from the effluent of the
004 last reaction zone, and during the course of said reforming
005 coke deposits upon said cata~yst and deactivates thereby
006 decreasing decreasing the yield of said product, tne method of
007 increasing said yield whlch comprises maintaining the level of
008 coke on the catalyst in said first reaction zone at less than
009 1~ by weight by adjusting the frequency at which the catalyst
010 in said first reaction zone is regenerated or replaced with a
011 fresh or regenerated catalyst.
012 BRIEF DESCRIPTION OF THE FIGURES
013 FIG. 1 illustrates the effect on reaction tempera-
014 ture, CS+ yield and hydrogen production of a catalyst bed
015 containing 0, 10, 20 and 30% fresh catalyst on top of a layer
016 of coked catalyst. FIG. 2 illustrates yield losses due to
017 variable levels of coke on catalyst at the top of a bed of
018 coked catalyst. FIG. 3 illustrates the results of a comparlson
019 between reforming with coked catalyst and with 10~ fresh
020 catalyst on top of coked catalyst.
021 DETAILED DESCRIPTION
022 The present invention is applicable to those
023 reforming systems wherein a plurality of reation zones in
024 series are used. Preheaters are preferably present between
025 reaction zones so that the temperature of the feed to each
026 reaction zone may be controlled. Preferably, each reaction
027 zone will be located in a separate reactor. The present
028 invention is concerned with those systems wherein at least two
029 reactors and preferably from 3 to 5 reactors are in series.
030 Although some or all of the reactors may be moving bed
031 reactors, the most preferred system is for all reactors to be
032 fixed bed systems or all reactors but the first to be fixed bed
033 systems with the first being a moving bed system.
034 In multi-reaction zone reforming systems, the
035 catalyst may vary in composltion in the different reforming
036 zones, although generally the catalyst is the same in all
037 zones. ~owever, the volume of catalyst generally differs from
,
~l~i;3'Zl
001 -S-
002 one reaction zone to the next. A typical catalyst loading ln a
003 three-reactor system may employ one-quarter of the total charge
004 of catalyst in the first reactor, one-quarter in the second
005 reactor and one-half in the last reactor. The first reactor
006 generally contains less catalyst because the highly endothermic
007 reaction taking place therein results in the rap~d coollng of
008 the feed. If a large volume of catalyst were present in the
009 first reactor, the temperature of the feed in the lower portion
010 of the catalyst bed would be too low for significant dehydro-
011 genation reactions to occur and thus the lower portion of the
012 catalyst bed would not be used effectively.
013 To achieve the benefits of the present invention, the
014 catalyst in the first reaction zone should be replaced or
015 regenerated with sufficient frequency to maintain a level of
016 coke in the catalyst less than 1% by weight preferably less
017 than 0.7% and still more preferably less than 0.5~ by weight.
018 The first reaction zone may contain up to 30% by volume of the
019 total catalyst in the reactor system, although preferabiy it
020 will contain only up to 20~ and still more preferably only up
021 to 15% of the total catalyst mass in the reactor system.
022 The temperature in each of the reactlon zones can be
023 the same or different, but generally it will fall in the range
024 from 700~F to 1050F and preferably within the range of about
025 850F to 1000F. The terminal reaction zone generally has the
026 highest average catalyst bed temperature. The pressure in each
027 of the reaction zones will usually be the same, eitner
028 atmospheric or superatmospheric. Preferably, the pressure will
029 be in the range of 25 to 1000 psig and more preferably between
030 50 and 750 psig. The temperature and pressure can be
031 correlated with the liquid hourly spaced velocity (L~SY) to
032 favor any particularly desirable reforming reactions and will
033 generally be from 0.01 to 10 and preferably from 1 to 5. It is
034 apparent that with different catalyst loading and different
035 reaction zones, the space velocities in the individual reactlon
036 zones can vary considerably.
~;
ll.~i;~l
001 -6-
002 Although reforming generally results in the
003 production of hydrogen, it is common to recycle hydrogen
004 separated from the effluent of any of the reaction zones,
005 usually the terminal reaction~ zone, to the first or subsequent
006 reaction zones. The hydrogen can be admixed with the feed
007 prior to contacting catalyst or simultaneously with the intro-
00~ duction of the feed to the reaction zone. The presence of
009 hydrogen serves to reduce formation of coke which tends to
010 poison the catalyst. ~ydrogen is preferably introduced into
011 the reforming reaction zone at a rate which varies from 0.5 to
012 20 mols of hydrogen per mol of feed. Hydrogen can be an ad-
013 mixture with light gaseous hydrocarbons.
014 The catalyst used in the reaction zones comprises a
015 platinum group component in association with a porous solid
016 carrier. Preferably the platinum group component is platinum
017 and the preferred porous solid carrier is a porous refractory
018 in organic oxide, for example, alumina. The platinum group
019 component will be present in an amount of from 0.01 to 3 weight
020 percent and preferably 0.01 to 1 weight percent.
021 Other components in addition to the platinum group
022 component can be present on the porous solid carrier. It is
023 particularly preferred that rhenium be present, for example in
024 an amount of 0.01 to S weight percent and more preferably 0.01
025 to 2 weight percent. Rhenium significantly improves the yields
026 obtained using a platinum-containing catalyst, and a
027 platinum-rhenium catalyst is more fully described in U.S.
028 Patent 3,415,737. Generally, the catalyst will be promoted for
029 reforming by the addition of a halide, particularly fluorlde or
030 chloride. The halide provides a limited amount of acldity to
031 the catalyst which is beneficial to most reforming operations.
032 The catalyst promoted with hallde preferably contains 0.1 to 3
033 weight percent total halide content and the preferred hallde is
034 chloride.
035 The catalyst within the first reactlon zone may be
036 replaced or regenerated in situ or ex sltu, with sufficient
037 frequency to maintain less than 1 weight percent coke on the
001 -7-
002 catalyst, and preferably less tAan 0.7 welght percent and more
003 preferably, below 0.5 weight percent. The amount of coke on
004 the catalyst should be calculated as the average amount of coke
005 on the entire volume of catalyst in the first reaction zone.
006 The catalyst in the first reaction zone may e1ther be disposed
007 in a fixed or moving bed. If the reactlon zone is a flxed bed,
008 it may be radial flow, upflow or downflow, and it may be
009 desirable for a swing reactor to be present and ready to be put
010 in service when the first fixed bed reaction zone is removed
011 from service for regeneration. Of course, if the first
012 reaction zone has a moving catalyst bed, the reaction zone c n
013 be maintained onstream while fresh or regenerated catalyst is
014 added to the top of the reaction zone and used coke-deactivated
015 catalyst withdrawn from the bottom. The remaining reaction
016 zones in the system can be either fixed bed or moving bed
017 reaction zones, b~t for the purposes of this invention, it is
018 advantageous if they are fixed bed reaction zones. The fixed
019 bed reaction zones can be either upflow, downflow or radial
020 flow with radial flow being preferred.
021 To determine the level of coke or carbon on the
022 catalyst of the first reactor, the catalyst can be sampled as
023 it is removed from the reactor in a moving bed reaction zone or
024 catalyst samples may be withdrawn from the reaction zone itself
025 without disrupting a normal reforming process. A variety of
026 means are available for removing catalyst samples from reactors
027 without involving shutdown of the reactor, for example, see
028 U.S. Patents 3,129,5gO and 3,319,469. The level of coke or
029 carbon deposited on the catalyst sample may be determined by
030 any suitable means such as combustion of the carbon and
031 measurement of the quantity of CO2 produced or by determining
032 the rate at which a combustion zone progresses through a bed of
033 coke-contaminated catalyst, such as described in ~.S. Patent
034 3,414,382.
035 The hydrocarbon feedstock employed in the retorming
036 operation of the present invention may be any suitable hydro-
037 carbon capable of being catalytlcally reformed at the stated
~14~3'Zi
oo1 ~
002 conditions. Preferably the feedstock is a naphtna fraction,
003 which is a light hydrocarbonaceous oil generally boiling within
004 the range from 70 to 550F and preferably from 150 to 450F.
005 The feedstoc~ may be, for example, either a straight-run
006 naphtha, a thermally cracked or catalytically cracked naphtha
007 or blends thereof. Generally the naphtha feed will contain
008 from about 25~ to 75% and preferably about 35% to 60~
009 paraffins, about 15~ to 65% and preferably about 25% to 55%
010 naphthenes and about 5% to 20% ~romatics, calculated on a
011 volume percent basis.
012 Catalyst regeneration procedures are well known to
013 the art and will generally include burning the carbon from the
014 catalyst by contacting the catalyst to an oxygen-containing gas
015 at an elevated temperature from about 700 to 1100F.
016 Preferably the oxygen concentration and temperature are
017 increased in stages as the regeneration progresses. Following
018 the carbon burn-off, a gaseous halogen may be introduced into
019 contact with the catalyst. Subsequently, the catalyst may be
020 dried and reduced before being returned to the reforming zone.
021 Examples of suitable regeneration processes are illustrated in
022 U.S. Patents 3,134,732 and 3,496,096.
023 EXAMPLES
024 The present invention will be further clarifled by
025 consideration of the following examples which are intended to
026 be purely exemplary and not limiting of this invention.
027 Example 1 shows that the presence of a small amount of fresh
028 catalyst on top of a bed of catalyst containing 10.9% carbon
029 acts to substantially increase the C5+ yield and hydrogen
030 production as well as to decrease the average catalyst
031 temperature required to ~ake a product of a predetermined
032 octane. These results indicate that the presence of a small
033 amount of fresh catalyst upstream of a larger mass of
034 coke-contaminated catalyst serves to significantly extend the
035 length of time which the total mass of catalyst can be
03~ maintained in reforming service before being regenerated.
037 Example 2 shows the adverse effect on yield due to the presence
3Zl
001 _9_
002 of an increasing amount of coke on a small mass of catalyst
003 situated above a larger mass of coke deactivated catalyst.
004 These results indicate that the less carbon that is present on
OOS the catalyst in the first reaction zone, the better the overall
006 yield. Example 3 is a side-by-side comparison of reformlng
007 with a catalyst bed containing 10% fresh catalyst on top of a
008 bed of test catalyst containing various levels of co~e which
009 illustrates the activity and yield advantages of having a small
010 layer of fresh catalyst present in the top of the catalyst bed.
011 Example 1
012 A mid-continent naphtha having the characteristics
013 shown in Table I was passed through a series of foul reactors
014 containing a platinum-rhenium reforming catalyst at reforming
OlS conditions including a pressure of 200 psig, a liquid hourly
016 space velocity of 2, a hydrogen to hydroccarbon mol ratio of 3
017 and a temperature adjusted to obtain a reformate product having
018 a research octane number of 98 clear.
020 TABLE I
022 Mid-Continent Naphtha
a24 Gravity API 55.0
025 D-86 Distillation
026 IBP - F 174
027 10% - F 214
028 30% - F 239
029 50% - F 263
030 70% - F 2~4
031 90% - F 342
032 EP - ~F 390
034 % Paraffins 43.1
035 % Naphthenes 46.8
036 % Aromatics 10.0
037
001 -10-
002 After 55 days onstream, a portion of the catalyst was
003 removed from the last reactor in the series. The catalyst,
004 averaging 10.9 weight percent coke, was tested ln a micro-sized
005 pilot plant reformer in three separate tests in which the top
006 10%, 20% and 30% of the used catalyst replaced by an equlvalent
007 amount of fresh catalyst. The results, as represented ln FIG.
008 1 and Table II show that by placing a layer of fresh catalyst
009 on top of a larger mass of coked catalyst, (1) the activity of
010 the total mass of catalyst increased significantly, by 19F for
011 10% fresh catalyst, by 22F for 20~ fresh catalyst, and by 32F
012 for 30% fresh catalyst; (2) the C5+ yield increased by 1.9
013 liquid volume percent, from approximately 78.6 to 80.5 LV
014 percent; (3) the hydrogen production increased by about li~,
015 from 1217 standard cubic feet per barrel of feed to 1356-1347
016 standard cubic feet per barrel of feed; and (4) CH4 production
017 decreased 23-26~, from 108 standard cubic feet per barrel of
018 feed to 80-83 standard cubic feet per barrel of feed. Thus,
019 the presence of a small amount of fresh catalyst on top of a
020 larger amount of coked catalyst serves to substantially
021 increase the activity, C5+ liquid yield and rate of hydrogen
022 production, far more than would ~e predicted just from the
023 small amount of fresh catalyst added.
025 TABLE II
027 Layered-Bed Tests on End-of-Run Catalyst
029 ~ C5+,LV% H2, SCF/B ~ , SCF/B
031 End-of-Run (EOR)
032 Catalyst, 10.9% C, 955 78.6 1217 108
033 10% Fresh over
034 90~ EOR Catalyst 936 80.6 1356 82
035 20% Fresh over
036 80% EOR Catalyst 933 80.5 1347 80
037 30% Fresh over
038 70% EOR Catalyst 923 80.5 1349 83
039 Example 2
040 A study was made to determine the effect on yield of
041 a varying amount of coke on the top 10% of c~talyst in a
1141;~Zl
O O 1 - 1 1-
002 cataiyst bed. FIG. 2 shows the effect on C5+ yield and H2
003 yield associated with an increasing carbon content on the
004 catalyst in the top 10% of the catalyst bed. Using as the
005 standard a catalyst bed containing 10~ fresh catalyst on top of
006 90% catalyst containing 13.3 weight percent carbon, a catalyst
007 bed with the top 10% of catalyst containing 2 weight percent
008 carbon loses 1% by volume of C5+ yield; a catalyst bed with the
009 top 10% of catalyst containing about 6 weight percent carbon
010 loses 2% by volume of C5+ yield; and a catalyst bed with the
011 top 10% of catalyst containing about 12 weight percent carbon
012 loses about 3% by volume of C5+ yield. Hydrogen yield loss
013 also increases in the same manner with increasing carbon
014 content in top 10% of the catalyst in the catalyst bed. Thus,
015 to obtain the maximum yield benefit from the process of this
016 invention, the amount of carbon on the catalyst in the top of
017 the catalyst bed should be kept as low as poss1ble.
018 Example 3
019 A test was conducted to compare the performance of a
020 bed of co~e-deactivated platinum-rhenium catalyst with an
021 equivalent volume of catalyst comprislng 10 volume ~ fres~
022 catalyst on top of 90% of the deactivated catalyst. Samples of
023 catalyst from a commercial reformer were obtained at approxi-
024 mately 0, 28, 61, 89 and 122 days on stream. One portion of
025 each catalyst sample was tested on the feedstock shown in Table
026 I at reforming conditions including a pressure of 200 psig, a
027 liquid hourly space velocity of 2, a hydrogen to hydrocarbon
028 mol ratio of 3 and a temperature adjusted to obtain a reformate
029 product having a research octane number of 98, clear. A layer
030 of 10~ fresh catalyst was put on top of a 9o% layer of catalyst
031 from the 61, 89 and 122-day samples, respectively, and then
032 tested under the same reforming conditions.
033 The results, as shown in Figure III, demonstrate that
034 the catalyst beds containing 10% fresh catalyst are far more
035 active (20F after 120 hours) and more selectlve (4% C5+ yleld
036 after 120 hours) than the beds containing only coked catalyst.
037 The fouling rate for the beds containing 10% fresh catalyst is
001 -12-
002 less than that for the bèds of coked catalyst -- 0.15Ftday vs.
003 0.33F/day, indicating that the effect of the fresh catalyst is
004 far out of proportion to its volumetric presence.
OOS From the foregoing, it may be seen that the pr,esent
006 invention operates in a novel and effective manner to increase
007 the activity, selectivity (C5+ yield) or both of the total mass
008 of catalyst in a reforming reaction zone, and thus it ~er~.its
009 the service life of the bulk of catalyst to be extended before
010 regenera~tion or replacement is necessary.
011 Although only specific arrangements and modes of
012 operation of the present invention have been descr~bed,
013 numerous changes can be made in those arrangements without
014 depar.ing from the spirit of the invention and also changes '
015 that fall within the scope of the appended claims are intended
016 to be embraced thereby.
