US3151056A

US3151056A - Upgrading of high severity reformate

Info

Publication number: US3151056A
Application number: US141747A
Authority: US
Inventors: William F Wolff; Gordon D Mcleod
Original assignee: Standard Oil Co
Current assignee: Standard Oil Co
Priority date: 1961-09-29
Filing date: 1961-09-29
Publication date: 1964-09-29
Anticipated expiration: 1981-09-29

Description

United States Patent O 3,15L056 UPGRADING F HTGH SEVERITY REFORMATE William F. Wold, Park Forest, and Gordon D. McLeod,

Lansing, Ill., assignors to Standard Oil Company, Chicago, Ill., a corporation of Indiana Filed Sept. 29, 1961, Ser. No. 141,747 Claims. (Cl. 203-95) This invention relates to catalytic reforming of naphtha fractions to produce gasoline constituents of high antiknock quality. More particularly, it relates to a combination of high severity catalytic reforming with a novel process for obtaining high octane gasoline blending stock with greatly reduced tendency to form harmful intake manifold deposits when used in internal combustion engines.

Automotive industry progress has resulted in the development of gasoline engines of higher and higher compression ratios requiring gasoline fuels of correspondingly higher octane rating and resistance to formation of harmful deposits in the engines intake manifold. To meet this demand the petroleum industry has developed various processes for producing gasoline of increased octane rating. One of the more eilicient of such processes is known as platinum catalytic reforming wherein low octane naphtha is contacted under pressure in the vapor phase and at a temperature above 800 F. in the presence of substantial amounts of hydrogen. These platinum catalysts are supported on suitable carriers such as specially prepared alumina or specially treated silica-aluminas and can be activated with chloride or fluoride. ln the recent past these platinum catalytic reforming processes have been operated to produce high octane gasoline blending stocks in the order of 90 to 92 clear research octane number. However, more recently the need for catalytic reformates having octane numbers in excess of 95 has developed as the automobile industry has increased the compression ratio and performance of automotive engines.

To meet the requirements of the present and future high compression automotive engines it is now found necessary to operate these platinum catalytic reforming processes at increased severity of operating conditions to the extent that a deleterious amount of high boiling polynuclear aromatic compounds are formed and are recovered along with the high octane gasoline blending stocks. Under these severe conditions a considerable portion boiling above about 400 F. is usually produced; this portion is almost entirely aromatic and therefore of a very high octane number. However, this portion contains polynuclear aromatic compounds such as naphthalenes, pyrenes, anthracenes and oxygenated derivatives thereof. The presence in gasoline of highly condensed polynuclear aromatic compounds such as anthracene and pyrene, even in very low percentages, has been found `to be deleterious to the automotive engines in forming gums and harmful deposits when these gasolines are used as fuels.

Rerunning of the catalytic reformate produced from high severity operations has been practiced in an effort to remove the small amount of deleterious polynuclear aromatic compounds. However, since the heavy ends are high in octane number and are generally not harmful to the automotive engine, rerunning to cut out a heavy bottoms will also lead to unnecessary loss of high quality catalytic reformate. In addition, rerunning requires additional equipment and substantial capi-tal expenditure and continued expense of operation of such equipment.

We have found that small amounts of deleterious polynuclear aromatic compounds formed during high severity l platinum catalytic reforming can be selectively removed from a high octane reformate by contacting the reformed Patented Sept. 29, 1964 naphtha in the liquid or vapor state with an alkali metal or an alkali metal deposited on a support and recovering therefrom reformed naphtha high octane gasoline blending stock with essentially only the polynuclear aromatic compounds removed by the alkali metal treatment.

1t has been shown that the tendency of hydrocarbons to form deposits in the intake manifold of engines is related to both their boiling point and type, and in both these respects the highly condensed polynuclear aromatic compounds and their oxygenated derivatives appear to be the most harmful. An ASTM distillation of the high octane gasoline blending stock does not give an accurate picture of the highest boiling fraction of gasoline in which the polynuclear aromatics are present because the polynuclear aromatics are concentrated in the final 5% of the gasoline. The amount of polynuclear aromatic compounds in the gasoline tail can be determined only by very careful and accurate analytical means, such as by the use of ultraviolet absorption spectrometry very small amounts of naphthalenes, anthracenes and pyrenes can be determined. The amounts of these compounds provide a good cross-section of the bicyclic, tricyclic and tetracyclic aro- :iatics which are present in the gasoline and which are conducive to the formation of harmful deposits in automobile engines. Usual methods for determining the gumforming tendency of gasoline, as the ASTM method, can not be correlated with the amount of polynuclear aromatic compounds present in the gasoline and thus the ASTM gum method does not give a true indication of the intake manifold deposit-forming tendency of the gasoline. A laboratory bench method has been developed to be used in the study of induction-system-deposit formation. This method has been labeled Rotogum and will be described in detail hereinbelow. This test is particularly very sensitive to changes in the polynuclear aromatic compound content of the heavy ends of high severity catalytically reformed gasoline blending stocks.

The polynuclear aromatic compounds found in Rotogum deposits are more dependent upon hydrocarbon type than upon their original concentrations in the gasoline being tested. Polynuclear aromatics such as anthracenes and pyrenes are predominant in the Rotogum deposits even though their concentrations in the gasoline are ex- Ilcmely low. The distribution of polynuclear aromatics in the hydrocarbon portions of the Rotogurn deposits formed from catalytic reformates derived from high severity operations has been found to comprise about 30% pyrenes and about 20% anthracenes. It has been determined that the relative amounts of these two polynuclear aromatics along with the amount of naphthalene present in the high severity catalytic reformate are a direct measure of the induction-system-deposit formation tendencies of a gasoline blending stock.

For convenience in expressing the polynuclear aromatic compound content of gasolines, an arbitrary rating has been developed. This rating, called the NAP number, expresses both the relative concentrations of naphthalenes, anthracenes and pyrenes in the gasoline as well as the tendency of the gasoline to form intake manifold deposits and thus is correlated with the Rotogum deposit forma- 1tion. The arbitrary NAP number is determined as folows:

NAP number l0 =-3 [wt. percent naphthalenes -l- 100 (Wt. percent an thracenes-I- pyrenes) In the practice of our invention, a desulfurized naphtha is charged to a platinum catalytic reformer operated under high severity conditions so as to produce a reformate having a clear research octane number in excess of about used.

vUnder these severe conditions the essentially sulfur-free debutanized catalytic reformate recovered will contain at least about 1.5 Weight percent naphthalenes, 0.012 weight percent pyrenes, and 0.015 weight percent anthracenes. This catalytic reformate produced by high severity reforming will have an NAP number greater than 15. In order to reduce the polynuclear aromatic compound content and thus the NAP number, the catalytic reformate is brought in contact with an alkali metal, such as sodium or potassium or alloys of the two, under conditions such that the bulk of the polynuclear aromatic compounds is removed. The alkali metal may be deposited on a support or used alone. Under normal operations the treatment will be carried out in the range of 70 F. to 500 F. The alkali metal treatment can be carried out at subatmospheric or superatmospheric pressures. The catalytic reformate recovered will have a very low NAP number and thus a very low tendency to form harmful deposits in the intake manifold when used in internal combustion engines.

The advantages of our invention will be better understood by the following description of an illustrative example with reference to the drawings in which:

FIGURE 1 is a flow diagram of a catalytic reforming process, including the process of our invention, and

FIGURE 2 is a graph plot showing the etiect of reforming severity on the tendency of the reformate to form harmful intake manifold deposits.

In the illustrative example shown in FIGURE 1, desulfurized naphtha is `fed to `a catalytic reformer through line to preheater 11 along with recycle hydrogen introduced by line 12. This mixture of naphtha and hydrogen is preheated to above 800 F., usually about 900 to 950 F. or higher, and introduced by line 13 into reactor 14, thence by lines and 17 to reheat furnace 1.6, to reactor 1S and thence by

lines

19 and 21 toreheat furnace and then to reactor 22. Reactors 14, 1S and 22 contain one of several commercially available platinum catalytsts used in catalytic reforming. Examples of such catalytsts are fully described in US. Patents 2,479,109 and 2,659,701. It is to be understood that any suitable platinum catalyst capable of producing debutanized catalytic reformates having clear research octane numbers in excess of 95 can be used.

Reactors

14, 18 and 22 are operated at about 950 F. and a pressure between 200 and 500 p.s.i.g., depending upon the type of reforming process In any event, the conditions of temperature, pressure and space velocity are such that severe reforming of the naphtha takes place to produce a catalytic reformate having a clear research octane number of 102 and containing 2.1 weight percent naphthalenes, 0.03 weight percent anthracenes and 0.02 weight percent pyrenes. While no regeneration facilities are shown in FIGURE 1, the catalytic reforming process can be either a non-regenerative or regenerative type. If it is nonregenerative, the catalyst will be changed periodically as its activity is reduced by use. If the regenerative type reforming process is used, a swing reactor is normally utilizezd to replace a reactor taken off stream for regeneration. Regeneration techniques known in the art for removing carbon deposits, as well as rejuvenation techniques, are described in U.S. Patents 2,560,329, 2,773,014 and 2,916,433. The mixture of severely reformed catalytic naphtha and hydrogen is withdrawn via

lines

23 and 24, passing through cooler 25 to separator 26. In separator 26 make-hydrogen is removed and recycled to the naphtha charge in line 10 via

lines

27, 28 and 12. Excess hydrogen may be bled off through valved line 29. Reformate is withdrawn from the bottom of separator 26, passed through heater-cooler 31 via

lines

30 and 32, then via line 33 to treating bed 34 through drier 35. Drier 35 is optional since normally the water content of the reformate is sufficiently low that it will be removed by the alkali metal treatment. Since the cost of potassium is fairly expensive, it is normally found economically advantageous to use the drier when the alkali metal used in treating bed 34 is potassium. Drier 35 may be any one of a number of commercially known processes for removing trace amounts of water in hydrocarbons. For example, it may be a bed of sodium chloride or calcium sulfate. Conventional regeneration facilities, not shown, may be used to regenerate drier 35. In this case the use of additional drier beds will be advantageous in order that one may be taken off stream for regeneration while the others remain in use. In treating bed 34 the alkali metal may be deposited on a support such as ground glass, pumice, porcelain chips or other types of solid supports. Or the alkali metal may be used alone in the molten state or as a dispersion. Any of the several methods of contacting molten alkali metal with liquid hydrocarbon may be used. In this example potassium on ground glass is used as the treating agent in treating bed 34. The temperature of the bed is 300 F., although any temperature in the range from 250 F. to 500 F. may be used. The pressure is p.s.i.g. The severely reformed reformate may be contacted in the liquid or vapor phase with the alkali metal. At 300 F. and 100 p.s.i.g. it is in the liquid state. Space velocity of treating bed 34 is not critical so long as there is sufficient contact between the reformate and the alkali metal. Potassium has been found to give the best removal of very small amounts of polynuclear aromatic compounds in severely reformed reformates. In heater-cooler 31 the temperature is adjusted to the desired temperature for the operation of the treating step.

Since the feed charged to the process is a desulfurized naphtha, the sulfur content of the reformate being passed through line 33 into the treating bed 34 will contain essentially no sulfur. Thus treating bed 34 may be operated for fairly long periods of time before requiring replacement. The reformate treated in treating bed 34 with essentially only the polynuclear aromatic compounds removed is charged by line 36 to stabilizer 37 wherein stabilizer gas is taken overhead .through line 38 and the stabilized catalytic reformate, having an extremely low polynuclear aromatic compound content of 1.8 weight percent naphthalenes, 0.003 weight percent anthracenes and 0.002 weight percent pyrenes, is recovered through line 39. While in this example the high severity catalytic reformate has been passed to the treating step following the separation of the hydrogen-rich gas in separator 26, it is to be understood that it may be found advantageous to pass the high severity reformate from separator 26 directly to stabilizer 37 wherein the stabilizer gas is withdrawn overhead and a stabilized catalytic reformate is withdrawn from the bottom. By this' operation the stabilized reformate may then be passed to a treating bed (not shown) wherein the polynuclear aromatic cornpounds are removed in the same manner as described above for treating bed 34. The operating conditions will be different if the treating bed follows stabilizer 37 since the reformate will be at a lower pressure in the order of atmospheric to about 50 p.s.i.g. At these pressures, ternperatures in the order of 70 F. to 350 F. may be used. Thus it can be seen that by the use of our irnproved process the deleterious polynuclear `aromatic cornpounds which are present in the reformate obtained from high severity reforming can be removed and the bulk of the high boiling aromatic hydrocarbons having high octane number and boiling in the approximate same ASTM boiling range as the polynuclear aromatic compounds are not lost during the processing.

As the severity of platinum catalytic reforming is increased to meet future gasoline needs, more of the deleterious polynuclear aromatic compounds Will be formed. The presence of small amounts of these deleterious polynuclear aromatic compoundsI in reformates did not become a problem until reforming was increased to the high severity conditions. An examination of some thirty-four commercial and experimental catalytic reformates has shown that as the severity of the processing is increased,

there is an increase in the NAP number and thus an increase in the deleterious polynuclear aromatics. This is shown in FIGURE 2 Where the research octane numbers of catalytic reformates are plotted against NAP number. At current catalytic reforming severity levels which produce catalytic reformates having a clear research octane number up to 95, the average NAP number is about 12. However, at the 100-102 octane number level, the average NAP number of the reformate is from to 32. These high octane levels produce reformates which have a vastly increased tendency to form intake manifold deposits.

The advantages of our process will be more greatly appreciated by reference to the following examples:

Example 1 A 200 ml. ask was tted with a water-cooled reflux condenser, a nitrogen inlet and a Teflon-covered magnetic stirring bar. To the flask were added 0.528 g. of potassium and 120 ml. of a catalytic reformate obtained under severe reforming conditions land which contained 2.6 weight percent naphthalenes, 0.026 weight percent anthracenes, and 0.018 weight percent pyrenes. The mixture was reuxed for three hours with stirring under a nitrogen atmosphere. At the end of this period of time the mixture was cooled to room temperature and allowed to settle overnight under a nitrogen atmosphere. A portion lof the liquid was filtered and found to contain only y0.0004% anthracenes, 0.0009% pyrenes, and 1.2% naphthalenes, as measured by ultraviolet absorption spectroscopy.

Example 2 The same procedure described in Example 1 was repeated using another portion of the same reformate feed except -that only 0.07 g. potassium wasI used per 100 ml. of reformate and the mixture was refluxed for 5.5 hours. The product contained only 0.0006% anthracenes, 0.007% pyrenes, `and 2.3% napthalenes, as measured by ultraviolet absorption spectroscopy.

Example 3 A third run demonstrating the process of our invention was carried out in a ml. flask using an uncovered magnetic stirring bar. A twenty ml. portion of the same reformate feed as used in Example 1 was charged to the flask along with 2 drops of a sodium-potassium alloy containing about 25 weight percent sodium. This alloy was liquid at room temperature. The flask was stoppered and the contents stirred with a magnetic stirrer for 1 hour at room temperature. The product was then filtered to obtain a clear, slightly yellow filtrate containing only 0.003% anthracenes and 0.006% pyrenes.

Example 4 In a fourth run demonstrating the process of our invention, a 200 ml. :flask was fitted with a Water-cooled reux condenser, a nitrogen inlet and a Teflon-covered magnetic stirring bar. A one liter vapor expansion cham- Example 5 In the fifth run demonstrating our process, a bed of potassium deposited on charcoal was used for treating the reformate. Dried, activated coconut charcoal, 8-14 mesh, was mixed with potassium in the ratio of 13.90 grams of charcoal per 4 grams of potassium. This mixture was stirred for 40 minutes at 290 C. under a nitrogen atmosphere. A glass column having 28 mm. outside diameter and containing a nitrogen atmosphere Was charged with 20 cc. of 3 to 4 mm. glass beads and about 14 grams of the above potassium deposited on charcoal. The glass tube was placed in a 12-inch electric furnace and maintained at 275 F. by use of a temperature-indicator-controller having a thermocouple in a thermowell in the center of the potassium-on-charcoal bed. Another portion of the same severely reformed reformate feed used in Examples 1 through 4 was introduced into the top of the `glass tube at a space velocity of about 2.5 v./v./ hr. The product which was water-white, was withdrawn from the bottom of the tube and condensed in a water-cooled condenser. Some of the fractions withdrawn were analyzed by ultraviolet absorption spectroscopy to determine the polynuclear aromatic content. The results were as follows Cumulative Volume oi Volume of Percent Percent Percent Cut Cut, m1. Treated Anthra- Pyrenes Naph- Reforcones thalenes mate, ml.

A composite was made of the first twenty-three cuts. This composite was analyzed by ultraviolet absorption spectroscopy and Rotogum contents were determined. The Rotogum test was conducted by slowly injecting ml. of the fraction to be analyzed into the walls of a rotating, inclined, glass tube heated to between 350 F. to 400 F. The introduction of the reformate is continued during a 45-minute test period during which the light ends are carried out of the tube in a stream of air and gasoline bottoms remain in the tube and form varying quantities of adhering deposits. These deposits are removed and weighed at the end of each test. Correlation studies have shown that the deposit-forming performance of gasolines in the Rotogum test can be correlated directly with their behavior in the intake manifolds of automotive test en gines. The results of ultraviolet absorption analyses and of the Rotogum test are compared below with the values for the untreated reformate:

What we claim is:

1. A method of selectively removing deleterious polynuclear aromatic compounds from a high octane essentially sulfur-free gasoline blending stock obtained from severely reforming a petroleum naphtha in the presence of platium catalyst and hydrogen at temperatures above 900 F., said method comprising contacting the sulfur-free reformed naphtha after removal of hydrogen gas with an alkali metal selected from the group consisting of potassium and sodium-potassium alloys, said contacting being in the liquid phase and at a temperature above about 70 F., and recovering therefrom a reformed naphtha blending stock having essentially only the polynuclear aromatic compounds removed by said alkali metal treatment.

2. The method of claim 1 wherein the alkali metal is potassium.

3. The method of claim 1 wherein the alkali metal is an alloy of sodium and potassium.

4. The process of claim 1 wherein the alkali metal is deposited upon a solid support.

5. A method of selectively removing deleterious polynuclear aromatic compounds from a high octane essentially sulfur-free gasoline blending stock obtained from severely reforming a petroleum naphtha in the presence of a platinum-on-alumina catalyst and hydrogen at a temperature ,above 900 F. and .at a pressure above about 200 p.s.i.g., `said method comprising contacting the sulfur-free reformed naphtha after removal of hydrogen gas with a bed of potassium deposited on a solid support selected from the group consisting of ground glass, pumice, porcelain chips and sand, said bed being at a temperature above 70 F., and said contacting being in the liquid phase and recovering therefrom the reformed naphtha blending stock having essentially only the polynuclear aromatic compounds removed by said potassium on -solid support treatment.

6. A method of selectively removing deleterious polynuclear aromatic compounds from a high octane essentially sulfur-free gasoline blending stock obtained from severely reforming a petroleum naphtha in the presence of a platinum-on-alumina catalyst and hydrogen at a temperature above L900 F. and at a pressure above about 200 p.s.i.g., said method comprising contacting the sulfur-free reformed naphtha after removal of hydrogen gas with an alkali metal selected from the group consisting of potassium and sodium-potassium alloys in the molten state, and said contacting being in the liquid phase, and recovering therefrom the reformed naphtha blending stock hav- Us ing essentially only the polynnclear aromatic compounds removed by said alkali metal treatment.

7. The method of claim 6 wherein the alkali metal is potassium.

' 8. A method of selectively removing deleterious polynuclear aromatic compounds from a high octane essentially sulfur-free gasoline blending stock containing greater than 0.015 weight percent anthracenes, and 0.012 Weight percent pyrenes obtained from severely reforming a petroleum naphtha in the presence of a platinum-on-alurnina catalyst and hydrogen at a temperature above about 900 F., said method comprising contacting the sulfur-free reformed naphthaafter removal of hydrogen gas with an alkali metal selected from the group consisting of potassium and sodium-potassium alloys, said contacting being in the liquid phase, and recovering therefrom treated reformed naphtna blending stock having essentially only the polynuclear aromatic compounds removed by the alkali metal treatment.

9. The process of claim 8 wherein the alkali metal is potassium.

l0. The process of claim 8 wherein the alkali metal is deposited on a solid support.

'References Cited in the file of this patent UNITED STATES PATENTS 1,859,028 Cross May 17, 1932 2,697,064 Brown Dec. 14, 1954 2,749,225 Barnum et al. June 5, 1956 2,864,761 DOuville et al Dec. 16, 1958 2,910,426 Gluesenkamp et al. Oct. 27, 1959

Claims

1. A METHOD OF SELECTIVELY REMOVING DELETERIOUS POLYNUCLEAR AROMATIC COMPOUNDS FROM A HIGH OCTANE ESSENTIALLY SULFUR-FREE GASOLINE BLENDING STOCK OBTAINED FROM SEVERLY REFORMING A PETROLEUM NAPHTHA IN THE PRESENCE OF PLATIUM CATALYST AND HYDROGEN AT TEMPERATURES ABOVE 900* F., SAID METHOD COMPRISING CONTACTING THE SULFUR-FREE REFORMED NAPHTHA AFTER REMOVAL OF HYDROGEN GAS WITH AN ALKALI METAL SELECTED FROM THE GROUP CONSISTING OF POTASSIUM AND SODIUM-POTSSIUM ALLOYS, SAID CONTACTING BEING IN THE LIQUID PHASE AND AT A TEMPERATURE ABOVE ABOUT 70*F., AND RECOVERING THEREFROM A REFORMED NAPHTHA BLENDING STOCK HAVING ESSENTIALLY ONLY THE POLYNUCLEAR AROMATIC COMPOUNDS REMOVED BY SAID ALKALI METAL TREATMENT.