US3124524A - Distillation - Google Patents

Description

March 10, 1964 B. s. FRIEDMAN 3,124,524

METHOD OF PRODUCING A HIGH OCTANE GASOLINE BY REFORMING A NAPHTHA IN TWO STAGES Filed June 1, 1959 26 STORAGE 32 TANK 24 J 1 HEATER 3o 29 g EXTRACTOR REACTOR REACTOR DISTILLATION 35 COLUMN 36 9 STORAGE TANK HEATER a? DISTILLATION 38 lo COLUMN 4| 42 u REACTOR 4lo 2 '4 is\ ,40 .3 REACTOR 47 sfs SEPARATOR 44 48 I8 T l5 SEPARATOR L I 2o DISTILLATION 43 46 f COLUMN SEPARATOR i zz DISTILLATION COLUMN 23 I 5o l FRACTIONATOR INVENTOR BERNARD S. FRIEDMAN ATTORNEY United States Patent 1 3,124,524 METHQI) 0F IRQDUCING A HIGH OCTANE GASDLINE BY REFGRMENG A NAIHTHA IN TWO STAGES Bernard S. Friedman, Chicago, lib, assignor, by mesne assignments, to Sinclair Research, Inc, New York,

N.Y., a corporation of Delaware Filed June 1, N59, Ser. No. 817,963-

4 Claims. (Cl. 298-65) 7 This application is a continuation-in-part of application Serial No. 628,666, filed December 17, 1956, now abandoned, and relates to a method for the conversion of hydrocarbon fractions boiling in the motor fuel range to obtain products of higher octane value. More specifically, my invention is directed to a multi-stage procedure involving the use of a plurality of different catalysts to transformthe constituents of straightrun naphthas to higheroctane components.

In recent years the automobile manufacturers have steadily increased the compression ratios of their spark ignition engines as a means of obtaining more power and greater efliciency. As the compression ratios of the engines increase, the hydrocarbon fuel employed must be of higher octane value to provide efiicient knock-free o eration notwithstanding that fuel octane can be increased through the addition of tetraethyl lead; and other undesirable aspects of engine operation, for instance preignition, can be overcome by the use of other additive components. Thus the problem remains for petroleum refiners to produce higher octane base hydrocarbon fuels under economically feasible conditions.

These refiners now have installed a substantial number of units for reforming straight run petroleum fractions in the presence of free hydrogen and over a platinum met-al alumina catalyst to obtain relatively high octane products. Primarily these products, frequently called reformates, are blended with other gasoline components such as thermal and catalytically cracked gasolines, alkylate, etc., and additives such as tetraethyl lead in obtaining present-day motor fuels. The reforming operation has a number of disadvantages. First, as the octane requirements of the blended engine fuels rise, the octane quality of the reformate must also increase if the blends be otherwise unaltered. increase results in a substantial reduction in yield particularly when obtaining reform-ates having octanes (RON neat) of the order of 90 to 95 or above. As the severity of the operation is increased, the platinum metal containing catalyst becomes fouled more often with carbonaceous deposits which requires more frequent regeneratio-ns and/or replacements. As the platinum metal-alumina catalysts are relatively expensive, either replacement or withdrawal from use during regeneration materially increases the cost of providing a given volume of reformate. These and other factors affecting the yield-octane number-cost relationship make it desirable for the refiner to consider various ways in which high octane hydrocarbon fuel components can be obtained by employing processing methods other than or in conjunction with the platinum metal-alumina catalyst reforming operation.

One method now under consideration by petroleum refiners for obtaining stocks of higher octane value involves the isomertization of parafiinic hydrocarbons boiling in the motor fuel range. In general, as the side chain branching of normal parafiins and of slightly branched isoparalfins increases, their octane ratings rise. A number of catalysts are known as being useful in this type of operation and such catalysts include hydrogen fluoride-boron trifluoride (see US. Patents Nos. 2,513,103 and 2,446,- 998) and platinumalumina.

In the present invention I have devised a multistage procedure involving the use of a plurality of dif erent catalysts to transform the constituents of straight run 3,124,524 Patented Mar. 10, 1964 naphthas to higher octane components. In the first reaction stage of my system the straight run naphtha is reformed over a platinum group metahalumina catalyst and a selected, low octane C -C fraction of the reformate is then treated with a hydrogen fluoride-boron trifluoride catalyst under particular conditions to give a further increase in octane rating.

Particularly, by operating my second stage reaction system under given processing conditions I have obtained material advantages in terms of an increase in the yieldoctane relationship of the gasoline boiling range hydrocarbons. Although my second stage system effects isomerization reactions, there is also considerable evidence at this time that other reactions such as disproportionation, polymerization and alkylation take place. By employing my improved process the operator can supply isoparafi'lns for various uses and particularly as the yield-octane relat-ionship of the gasoline boiling range product is exceptional my method affords an advantageous means for obtaining higher octane gasoline blending components which permit a petroleum refiner to elevate the octane rating of his gasoline pool.

The advantageous yield-octane relationships afforded by my method are due among other things to the charging to the hydrogen fluoride-boron trifiuoride reaction system of substantial amounts of isobutane. Thus I provide at least about 5 0% by Weight of isobutane based upon the charge of parafiinic feedstock boiling in the motor fuel range. The upper limit on the amount of isobutane charged is primarily an economic question involving factors such as the cost of distilling C s from the liquid product, the value of the liquid product, and the cost of increased consumption of butanes. However, in general there does not seem to be any advantage in employing more than about 600 Weight percent of the isobutane and preferably I charge about to 300%.

Also, it is important in my method that the hydrogen fluoride-boron trifluoride reaction system be operated so that there is not a net make of butanes. In fact, I find it advantageous if there'be a consumption of butanes, and preferably there is consumed in the reaction at least about 5 weight percent of butanes based on the C to C feed. The isobut'ane charged to my system can be relatively pure or mixed with other materials such as the various refinery hydrocarbons. Frequently, as found in the refinery, the isobutane is in admixture with at least normal butane, and I can employ such mixtures. Of course, the isobutane can be recycled from the reaction system effluent and usually at least a portion of any normal butane fed to the system is converted to isobutane which can be recycled and considered as part of the isobutane requirement charged to the reaction zone. The amount of isobutane in the light refinery streams could be increased as by contacts in an isomeriz-ation reaction stage with the catalyst passing to my HFBF conversion zone. The efiluent from this preliminary reaction is combined with my paraffin-rich feedstock and charged to the principal HFBF reaction system. The feed to the I-IF-BE, reaction system can also include materials such as propane, hydrogen or other light diluents. However, it is desirable to keep the concentrations of these materials low since they are not as beneficial as isobutane and heat may have to be supplied toany subsequent system to separate them from the reaction product. Also, these extraneous materials occupy space in the reaction system which could necessitate an increase in equipment size, and they may in effect retard the desired reactions.

It may be that there are a number of reactions which are effected in my second stage system. For instance, while the parafiinic feed boiling predominantly in the C to C range is undergoing isomerizati-on, there may be a disproportionation of the small amount of C to C and C hydrocarbons. Moreover, there is evidence that relat-ively high octane gasoline components are produced possibly by polymerization and by alkylation. These and other reactions might be taking place in my system but in any event I believe that the presence of the relatively large amount of isobutane and the paraifinic materials in the overall hydrocarbon feed is necessary in obtaining the desired result which is more than additive of that which would be expected from the conversion of the parafiins and isobutane in separate hydrogen fluoride-boron trifiuoride catalytic reaction systems.

The hydrocarbon feed to my essential HFBF catalytic process contains lost octane normal parafiins boiling predominantly in the C to C range. Such materials are found in various petroleum refinery streams and can be separated in more or less pure form or obtained in admixture with similar boiling isoparatfins and with materials such as olefins and aromatics. My hydrocarbon feedstocks contain at least about 15 percent of normal paratfin, preferably from 13 to 35 percent by weight. Thus the feed while boiling predominantly in the C to C range may include small amounts of n-pentane, n-hexane, nheptane or other C gasoline boiling range normal parafiins either alone, mixed with each other or with their isoparatfins or with similarly boiling olefins or aromatics.

As previously stated the feedstock for my hydrogen fluoride-boron trifluoride reaction system is a low octane par-aflinic concentrate effluent of reforming systems employing platinum-alumina catalysts. As an example most if not all of the aromatics of the reformate can be separated as by adsorption, extractive distillation, or any other procedure desired. The paraffinic materials resulting are useful in my process when in a feed containing at least about 15 weight percent of normal parafiins and these stocks may also contain small amounts of isoparaflins, olefins and aromatics. To obtain the desired selective paraflin-rich feed from the reformate, the aromatics can be adsorbed on silica gel; or separated by solvent extraction through the use of a solvent selective for aromatics, e.g. phenol, or by any other desirable procedure. A particularly useful method for accomplishing this separation is employed commercially and includes the use of a glycol-water extraction medium. As commercially licensed, one such system is known as Udexing by regulation of conditions such as the glycol to water ratio, the extraction and solvent stripping temperatures, and the character of the glycol, a Udex raffinate varying in paraflinicity is obtained. The manner of controlling these factors is known in the art and it sufiices to say that the preferred glycol materials are the diglycols such as diethylene and dipropylene glycols and their mixtures.

Thus in the present invention I have devised a highly attractive method for obtaining higher octane fuels for spark ignition engines which involves the use of separate reaction zones in which are employed catalysts of different properties. In this method a straight run gasoline or naphtha boiling range hydrocarbon is contacted with a platinum metal-alumina catalyst in the presence of free hydrogen under conditions which provide a substantial increase in the octane number of the petroleuin hydrocarbon material. A low octane parafiin-rich fraction boiling predominantly in the C to C range of the resulting reformate and isobutane are then contacted with the hydrogen fluoride and boron trifluoride catalyst according to a prescribed procedure to give a product boiling in the motor fuel range which is of substantially increased octane quality.

The hydrocarbon feedstocks charged to the reaction system containing the pl-atinum-metal-alumina catalyst are primarily the straight run petroleum fractions boiling in the gasoline and naphtha ranges, for instance in the range from about 175 to 450 F., but somewhat higher or lower boiling constituents can be included if desired. Preferably, the feedstock boils in the range of about 200 to 400 F. Although the hydrocarbons passing to the platinum metal-alumina catalyst reaction systems are composed of predominantly straight run naphtha material, minor amounts of additional components can be included such as olefins, thermal and catalytioally cracked stocks, recycled reformate and fractions of these cracked and reformed materials. The reaction conditions observed or maintained in the platinum metal-alumina catalyst system include those suggested for present commercial reforming operations such as temperatures from about 750 to 1000 'F., preferably about 825 to 975 F., and pressures from about 50 to 1000 p.s.i.g., preferably about 150 to 500 p.s.i.g. The free hydrogen supplied to this reaction system usually is in the form of hydrogen-rich recycle gases and generally provides about 2 to 20 moles of hydrogen per mole of hydrocarbon feed; preferably this ratio is about 4 to 10:1. The space velocity usually lies in the range of about 0.5 to 10 WHSV (weight of feed per weight of catalyst per hour) preferably about 2 to 5 WHSV.

The platinum metal-alumina catalysts employed in the method of this invention include a number of compositions. Generally, the platinum metal is a minor amount of the catalyst, e.g. about 0.1 to 1.5 weight percent of the final composition. Platinum is the most commonly employed metal present in these reforming catalysts although other useful platinum metals include rhodium, palladium, iridium which, along with platinum, are the face centered cubic crystallite types of the platinum family as distinct from the hexagonal types ruthenium and osmium which appear to be of lesser value.

These catalysts can be made by a number of procedures but a particularly eifective catalyst is one in which the alumina is obtained through calcination of an alumina hydrate containing at least about 65 weight percent of trihydrate and about 5 to 35 weight percent of alumina monohydrate and/or amorphous alumina forms, and advantageously having a surface area of about 350 to about 550 square meters per gram (BET method) when in the virgin state. The minor amount of platinum metal in the catalyst is usually present in finely divided form and is not detectable by X-ray diffraction techniques. Also, these catalysts are advantageously prepared to afford about 0.10 to 0.5, preferably about 0.15 to 0.3 cc./gram of their pore volume in pores of about to 1000 Angstrom units in size. Application Serial No. 489,726, filed February 21, 1955, describes the preparation of such catalysts. If desired, the catalyst can contain minor amounts of additional materials, for instance promoting components particularly those acidic in nature, such as silica and fluoride. Such promoting components are usually less than 10 weight percent of the final calcined catalyst.

The platinum metal-alumina catalyst can be employed in any type of reaction system desired, for instance moving or fluidized bed, regenerative or non-regenerative, etc., but advantageously the catalyst is disposed as a fixed bed. In the latter type of operation the size of commercial units is such that essentially adiabatic reaction systems must be employed and in View of this and the endothermic ature of the reforming operation the catalyst is placed in fixed beds in a plurality of reactors, each of which is preceded by means for heating its charge. In fixed bed operations, the catalyst is in macrosize form, that is, particles generally at least about A in length and diameter and preferably not exceeding about in diameter. Particularly 'when such particles are provided by extrusion, their length may be up to about 1" or more. If the platinum metal-alumina catalyst reforming system be of the regenerative type it can be arranged so that the catalyst of all of the reactors can be regenerated simultaneously or individually. Other variations in the platinum metal catalyst reaction system can be made according to the desires of the operator.

In the present invention, the essential feed to my hydrogen fluoride-boron trifl-uoride catalyst system is a paraffin-rich fraction boiling predominantly in the C to C range of the liquid reformate from the platinum metal catalyst operation, and as noted above, it includes at least 15 weight percent of normal parafiin constituents and usually less than about 10 weight percent of aromatics, preferably less than about 5 percent. This feedstock boils primarily in the C to C range, although lighter and heavier constituents can be included. The parafiin-rich feed boiling predominantly in the C to C range can be obtained as a Udex raffinate and frequently the feed contains a minor amount of olefins. The character of this rafiinate can be controlled by the boiling range of the reformate feed to the extractive distillation. As an example, if the reformate feed is of narrow boiling range, the rafiinate will also be of close boiling range.

A typical Udex raftinate A is predominantly of C to C hydrocarbons and analyzes as follows:

Gravity API, 60 F 67.4. AST M dist, -F.:

IBP 245. 50% 253.

Octane number (RON) 25.5 (neat), 60.1 (3 cc. TEL

added per gal).

Parafiins, vol. percent 95.5. 'Olefins, vol. percent 3.1. Naphthenes, vol. percent 0.0. Aromatics, vol. percent 1.4. Composition, vol. percent:

C l1 4 i-C 31 5 n-C 21.7 1-6 40.6 D-Cg In the hydrogen fluoride-boron trifluoride reaction system the conditions of treatment can vary widely depending upon feed composition, octane number desired for the final product, etc. I maintain the catalyst essentially in the liquid state even though when this is done it is quite likely that a portion of the boron trifluoride will also be present as vapor. Thus the pressure maintained must be sufiicient to provide essentially liquid phase reaction conditions as determined by the vapor pressure of the hydrogen fluoride, boron trifiuoride, the react-ants and other materials such as hydrogen and the reaction products present. Generally, the boron trifiuoride partial pressure will be at least about 200 psi. and preferably at least about 400 psi. in maintaining the liquid phase reaction. There seems to be little if any advantage in this partial pressure being above about 1500 psi. and at higher pressures propane formation may be excessive. Thus, the total pressure will be at least equal to the boron trifiuoride partial pressure and usually will not be above about 2000 p.s.i.g. The reaction temperature and reaction contact time are interdependent factors with a lesser time being required to provide a given result as the temperature increases. The reaction temperature will usually be in the range of about to 300 F, preferably about 75 to 200 F, with the time required ranging from about =1 minute to hours, preferably about 5 minutes to 3 hours. Time longer than 5 hours can be employed; however, no particular advantage would be derived thereby which overcomes the obvious economic disadvantages. Although the contact time and temperature employed can be selected as desired and can even be dependent upon factors such as the type of reaction system employed, I believe the following temperature-time relationships provide the best results but I do not intend to be limited by them.

TABLE I Temperature, F.: Contact time, minutes 100 30 to 180 100 to 200 to 60 200 to 250 5 to 30 250+ 1 to 10 In general, I prefer to select conditions which avoid production of substantial amounts of propane and long contact times at high temperatures.

In the reaction zone enough hydrogen fluoride is added so that when combined with the boron trifluoride a catalyst layer separate from the hydrocarbon can be obtained. Usually, this requires at least about one mole of hydrogen fluoride per mole of motor fuel range paraflmic hydrocarbon feed. Preferably, this ratio is not greater than about 10 to -1 as larger amounts of hydrogen fluoride require excessive handling facilities. The amount of boron trifluoride added to the reaction system has been discussed above with reference to its partial pressure which insures adequate combination with the hydrogen fluoride. The hydrogen fluoride and boron trifluoride catalytic components can be added separately but preferably they are introduced into the reaction Zone in admixture. The hydrogen fluoride-boron trifiuoride reaction system is usually conducted essentially in the absence of water to avoid having to increase the amount of boron trifluoride, but frequently there are minor amounts of water present such as those derived through the use of commercially available hydrogen fluoride and boron trifluoride.

The hydrocarbon product and the hydrogen fluorideboron trifluoride catalyst layers can be separated in any manner desired. When agitation of the reaction mixture is stopped it will separate into two phases in the reactor or in any other vessel into which it is transferred as in a continuous, semi-continuous or batch operation. These phases can be separated by simple decantation. The reaction mixture could be allowed to separate into a lower layer of catalyst containing aromatics which can be recycled to the reaction system in whole or in part. Usually, I keep the aromatic content of the predominantly C to C hydrocarbon feed at less than about 10 Weight percent, preferably less than about 5 percent. In designating amounts of aromatics, I refer to the values obtained by the Fluorescence Indicator method commonly known as the ETA. method involving chromatography on silica gel. The upper hydrocarbon layer formed in my system could be freed from catalyst by distillation and/ or washing with water or passed through a column of basic ion exchange resin or other solid adsorbent such as charcoal, potassium sulfate, sodium sulfate, etc. Aromatics appearing in the catalyst layer could be separated as by distillation of the catalyst, and the aromatics might then be combined with the hydrocarbons of the upper layer to provide a higher octane product. However, the recovered aromatics may be heavier than gasoline. Small traces of fluoride remaining in the hydrocarbon material can be removed as by passage over aluminum or alumina at 200 to 400 F. Various drying procedures could be employed to separate water from the hydrocarbon materials and such materials could be stabilized, for instance by the removal of C s and lighter constituents.

In the drawing I have illustrated a simplified flow sheet of one operation conducted in accordance with my method.

In this system straight run naphtha is charged by way of line 1 to heater 2 and then through line 3 to the top of an initial reactor 4 which contains a fixed bed of platinum-alumina catalyst. The efiluent from reactor 4 is passed by way of line 5 to heater 6 and then through line 7 to the top of a second reactor 8 containing a fixed bed of platinum-alumina catalyst. The platinum-alumina catalyst reaction section 'is of the adiabatic type and more than two reactors can be provided if desired and in fact usually at least three catalyst beds in separate reactors will be employed with each reactor having associated therewith a 'feed preheater. A third heater 9 and third reactor 10* containing a fixed bed of platinum-alumina catalyst are shown in the drawing.

The reformate from the bottom of reactor 10 is passed by way of line 11 to flash drum 12' which separates C s and lighter materials which are passed through line 13 to separator 14. The separator provides for removal of 7? C to C hydrocarbon constituents through line 15 and hydrogen and methane are recycled by way of line 16 to line 1. Excess hydrogen and methane can be removed from line 16 by way of line 17.

The liquid reformate from flash drum 12 is passed through line 18 to an intermediate portion of distillation column 19 and a light gasoline is taken overhead through line 20. The bottoms fraction from column 19 is carried by line 21 to an intermediate portion of distillation column 22 and heavy gasoline is removed as bottoms from this column through line 23. The overhead from column 22 is passed by way of line 24 to storage tank 25. Liquid hydrocarbon is withdrawn from the storage tank through line 26 and passed to an intermediate portion of extractor 27. Entering near the top of extractor 2'7 through line 28 is a glycol-water extractive medium. The raffinate produced in the extraction operation is taken overhead by line 29 and transported to storage tank 2%. The extract passes by way of line 30 to an intermediate portion of stripper 31. The stripped extractive medium then returns to extractor 27 through line 28. The overhead from stripper 31 is returned by line 32 to the lower portion of extractor 27.

A side stream from stripper 31 is charged to distillation column '34 and a toluene-containing overhead is removed by line 35. The bottoms from column 34 pass by way of line 36 to an intermediate portion of distillation column .37 from which xylenes are removed as overhead by line 38. The bottoms fraction from column 37 contains polymers and is removed by way of line 39.

The raffinate from storage tank 29a is charged through line 41 to reactor 40 after the addition of isobutane by way of line 41a. The hydrogen fluoride-boron trifiuoride catalyst mixture enters reactor 40 through line 42. The reaction efiluent is carried to separator 43 where a hydrocarbon phase and a catalyst phase are formed. The catalyst phase can be recycled to the reactor through lines 44 and 42 while the hydrocarbon phase is passed to the intermediate portion of fractionator 45. C minus overhead -from the fractionator goes to separator 46. In this separator, the C hydrocarbons are obtained and then recycled by way of lines 47 and 41 to reactor 40. C and lighter materials are removed by line 48 from separator 46. A gasoline fraction is taken as a side stream from fractionator 45 by way of line 49 while heavier hydrocarbons are withdrawn from the fractionator in bottoms line 50.

Although the drawing provides an illustration of a typical process I can employ, it is not to be considered limiting; for instance, the parafiinic fraction of catalytic reformate can be charged to the hydrogen fluoride-boron trifiuoride reaction system in admixture with small amounts of extraneous relatively close cut hydrocarbons such as n-pentane, n-hexane, n-heptane or mixtures of these normal paraflins with their isomers. As an example, the catalytic reformat-e might be flashed to remove C and lighter hydrocarbons and a C to C fraction separated by distillation. The resulting C reformate can be treated to obtain a paraiiind'ich fraction of pre-v dominantly C3 to C which is then charged to the hydrogen fluoride-boron trifiuoride reaction system. The isobutane is provided by recycle from the reaction zone and in addition extraneous normal and isobutanes can be added to the reaction zone as desired. The motor fuel boiling range products would then comprise essentially the gasoline obtained from the paraflin-rich portion of the reformate and isopentane produced in the hydrogen fluoride-boron trifluoride system due to the charging of n-pentane in the C to C fraction of the retormate. This reformate fraction would also contain isopentanes which could be isomerized to greater degrees of branching or merely carried through the hydrogen fiuoride-boron tri fluoride reaction system. These isopentanes would also appear in the motor fuel boiling range product.

8 Example I A straight run naphtha is obtained by distillation from crude oil, and the naphtha typically has an ASTM distillation boiling range of about 209 to 381 F., a RON (neat) of about 47.2, and a gravity API 60 F. of about 56.7. This naphtha is fed to a reforming unit containing three essentially adiabatic reactors each having a fixed bed of a platinum-alumina reforming catalyst. This system is equipped with means for heating the charge to each reactor and the heaters and reactors are arranged for serial flow. The catalyst employed is a platinum-alumina reforming catalyst containing about 0.6 weight percent platinum, and manufactured in accordance with application Serial No. 489,726, listed above. The inlet temperatures of the feed to each of the three catalyst beds are 940 F while the pressure is about 500 p.s.i.g. Free hydrogen is supplied to the feed passing to the heater before the first reactor and the hydrogen is obtained by recycle from the third reactor efiluent stream. The molar ratio of hydrogen-rich recycle gas (72.7% H to hydrocarbon feed is approximately 5.5 to 1, while the overall space velocity is about 2.34 WHSV. The efiluent from the last reactor is conveyed to a flash drum operating at 500 p.s.i.g. and is then treated or depropanized to remove C and lighter hydrocarbons by distiallation. Inspection on the resulting reformate is as follows:

Gravity API, 60 F 53. ASTM distillation, F

IBP 112.

EP 397. Octane number (RON) 84.9 (neat), 95.3 (3 cc. TEL

added/ gal.) Percent aromatics 51.2 (by F.I.A.). Percent olefins 0 (by F.I.A.).

To facilitate an understanding of this example the extractive distillation operation will be described with reference to the drawing. Thus, 6.71 parts by volume of the 112 to 397 F. boiling range reformate are passed at a temperautre of 222 F. to an intermediate portion of distillation column 19. The column top temperature is 215 F. and the column bottom temperature is 336 F. In column 19, 2.54 parts by volume of light gaso line are separated as overhead and this gasoline has a gravity API 60 F. of 71.1 and a boiling range of about 108 to 216 F. 3.17 parts by volume of the bottoms from column 19 are charged at 316 F. to an intermediate portion of distillation column 22. This column has a top temperature of 279 F. and a bottom temperature of 348 F. The overhead from column 22 is 2.24 parts by volume and the bottoms fraction is 0.93 part by volume of a heavier gasoline fraction. The overhead from column 22 has a gravity API 60 F. of about 46.2 and a boiling range of about 250 to 284 F.

The feed to the intermediate portion of extractor 27 is 1.622 parts by volume of the column 22 overhead. The tower top and bottom temperatures of the extractor are 280 F. and there results 0.838 part by volume of raffinate overhead from the extractor. The bottoms from the extractor is passed to the intermediate portion of stripper 31 which has a top temperature of 231 F. and a bottom temperature of 297 F. 8.16 parts by volume of extractive medium are separated as bottoms from stripper 31 and passed to the top of extractor 27. This extractive medium contains about 17% by volume of dipropylene glycol, 75.5% by volume of diethylene glycol and 7.5% by volume of water.

0.78 part by volume of a side stream from stripper 31 are charged at 292 F. to column 34. The top temperature of this column is 232 F. and the bottom temperature is 299 F. The overhead is 0.186 part by volume of a fraction consisting essentially of toluene. 0.598 part by volume are withdrawn as bottoms from column 34 and passed at 291 F. to column 37 which has a top temperature of 285 F. and a bottom temperature of 305 F. The overhead from column 37 is 0.597 part by volume of a fraction consisting essentially of Xylenes and the botl conducted through a Dry Ice cooling trap, safety trap, water scrubber, gas sampler and wet test meter. The hydrocarbon layer is separated from the ice water and the former is washed three times with separate 500 cc.

toms is 0.001 part by volume of polymer. portions of water. The washed hydrocarbon is dried by A 1750 ml. stainless steel Magne-dash bomb having contact with potassium carbonate. The products obtapered Walls to give maximum thickness in the bottom tained are 191 grams of liquid hydrocarbon, 237 grams half of the bomb is evacuated with a vacuum pump. of condensible gas and 3.74 liters (STP) of dry gas Cold liquid hydrogen fluoride (124.6 grams is drawn into (mostly air). The condensible gas and liquid are then the bottom of the bomb through a copper tube. 180- combined and distilled through a 12" glass helices vacuum butane (219 grams) 1S pressured into the bomb from a jacketed distillation column to separate 205.5 grams of pressure cylinder. 3.92 moles of boron trifluoride are C -C wet gas, 175 grams of initial to 435 F. overhead charged into the bomb by pressuring from a 2-liter cylingasoline, and 31 grams of still residue boiling above der. The amount of boron trifiuoride introduced is esti- 435 F. The yield of gasoline when corrected for the matfid y interpolation from the a of K p k and approximate 9% handling and mechanical losses is 100.1 Luborsky, 76 5865 The bomb P volume per-cent based on the rafiinate feed and inspecting sure is 530 p.s.i.g. and the temperature is 80 F. The as f ll contents or the bomb are stirred while pressuring in from Gravity APL 5 R a blow case 310 cc. (226 grams) of the overhead from Octane number (RON) 852 (neat). extractor 27. Thus, the isobutane to rafiinate Weight ratio Percent aromatics 0.0 (by FlA.) is 0.97 to 1. The charging of the overhead is over a Percent olefins 147 (b 5 a y period of one minute and the temperature rises to 90 Bromine number 135 F. while the pressure goes to 575 p.s.i.=g. Inspection of the RVR 12 6'lbs raffinate feed which contains at least about 15% n-paraf- Th t d M g T I t fins (water washed and dried extractor overhead) is: 3 Congo 6 We propane 1S v0 1 Percen 0 while the corrected yield of hydrocarbon boiling above GraVtY Am 9 435 F. is 11.2 volume percent. The total butanes re- ASTM dlstlnalon, 4 covered are less than those charged to the hydrogen IBP 2 fluoride-boron trifluoride reaction zone by 7 volume per- 50% cent based on the rafiinate feed and the recovered butanes 90% contain 90 mole percent of isobutane. Thus about 7 EP volume percent of butanes are consumed in the reaction Octane number (RON) 32.0 (neat), 64.7 (3 cc. basgd upon the raffinate feed TELaddw/gam- The yield-octane advantage obtained by charging the substantial amount of isobutane to my operation can be Percent aromatics (by readily seen by comparing several runs made in which Percent Olefins 23 (by the amount and nature of the lighter hydrocarbon portion Stirring of the contents of the bomb is continued for of the feed are varied. In these runs, the feed was the an additional 51 minutes and the temperature drops to U'dex rairinate of Example I; and, except for run 5, the 85 F. and the pressure goes to 450 p.s.i. g. While stir- HFBF reaction system was conducted essentially as ring, the contents of the bomb are discharged into a miX- described in Example I under the conditions noted in ture of ice and water through the bottom take-off valve. the following table. The procedure and conditions of The discharging requires about 9 minutes. Gases are run 5 are also given in this table, that is, run 5 was pro- TABLE Run 1 2 3 Ex. I 4 5 9 6 7 8 Temp, F. Initial/Final 01' Initial/Maximum/Fin 90/80 79/76 160/169/155 90/85 77/70 79/71 84/80 85/80 Pressure, p.s.i.g., Initial/Final or In Maximum/Final 585/405 590/440 760/770/725 575/450 1400/1420/1390 500/480 520/435 580/450 610/475 gondtact Time, min 62 60. 5 61 60 59 60 6 60 60 Udex Ralfinate:

g.. 436 343 311 226 197 o 54. 6 237 229 230 cc 600 469 426 310 270 75 325 315 315 Isobutane:

g 106 96 219 573 321. 5 b 278 199 230 cc 176 159 348 954 582 475 355 410 Catalyst HF, g 122. 4 127 124. 6 182 110. 9 125. 1 115 108 13m, mo1cs 3. 74 3. 76 3. 92 5. 62 3. 34 3. 92 3. 88 3. 88 Products:

Liquid Hydrocarbon, g 346 284 158 191 148 15 174 209 172 Condensiblc Gas, g 59 148 286.5 237 583.5 331 305 185 234 Dry Gas, liters STP 2 92 2. 77 4.08 3. 74 4. 72 5 5.13 2.1 1. 3 Recovery, Wt., Percent 76 86 97 91 94 91 94 68 Total 115.0 105.3 106.4 113.4 Butsnes Consumed 0 0 7 Percent Isobutane in C4 Fraction 93 91 36 90 Gasoline:

o.N. (R.M. neat) 75. 2 81. 2 82. 8 85. 2 FIAAromatics 0. 0 0. 0 0. 0 0. 0 Olefins--. .4 18.9 15. 8 14. 7 R.V.P 10. 75 12. 6

Gallons per gallons ralfinate in feed. n-Butane.

9 Isobutane HF and B F3 were all in bomb and Udex Raflinate was charged at the rate 0f1% (ac/min. during 1 hour.

d Based upon combined materials of Runs 7 and 8.

e The paraflins of the Udex rafiinate are about 49.9 weight percent CB and about 49.4 weight percent C9, sec Example I for the other characteristics of the ratfinate. Raflinate A described above could be used in the examples.

1 l cedurally similar to Example *1 except for the charging of the bomb.

The data of this table show that in run 1 conducted in the absence of isobutane the gasoline product had an octane number of 75.2 and the yield was only 56.8 percent whereas in Example I, this yield was 100.1 percent at 85.2 octane number. Runs 7 and 8 closely check EX- ample I in these respects, and in these systems the isobutane to ralfinate Weight ratios were in the range from about 0.87 to 1/ 1. In runs 2 and 3 where this ratio was only around 0.3/1 the gasoline yield was considerably reduced and there was a butanes make instead of a consumption as in Example I and runs 7 and 8. In run 6 where n-butane replaced the isobutane, the octane of the product was good but the yield of gasoline was low and there was a net production of butanes.

In another operation 70.6 grams of liquid hydrogen fluoride and 251 grams of isobutane are placed in the Magne-dash bomb of Example I. 2.3 moles of boron trifluoride are charged into the bomb by pressuring from a 2-liter cylinder, and this amount is estimated according to the method previously noted. The bomb pressure is 430 p.s.i.g. and the temperature is 63 F. While stirring its contents, the bomb is heated to 250 F. over a period of 1 hour and the pressure rises to 1615 lbs. This temperature is held for 1 hour and the products are discharged into ice water. The resulting separated hydrocarbon product analyzes as follows:

Component Weight percent C s 0.4 Propane 1.2 N-butane 8.7

Isobutane 88.6

Isopentane 0.8 N-pentane 0.1

It is seen that even under these severe reaction conditions little conversion of the isobutane is accomplished.

The advantageous yield-octane number relationship resulting from my method, for instance in Example I, is far better than would have been expected from the results received when treating the Udex raftinate (see run 1 of the foregoing table) and the isobutane with the hydrogen fiuoride-boron trifluoride in separate reaction systems. It is therefore apparent that my use of the paramnic hydrocarbons boiling predominantly in the C to C range in conjunction with the substantial amount of isobutane under conditions where there is not a net make of butanes is responsible for my highly improved results.

I claim:

1. In a method of converting a straight run hydrocarbon fraction boiling in the motor fuel range, the steps comprising contacting said hydrocarbon fraction with a platinum metal-alumina catalyst in the presence of free hydrogen at a temperature of about 750 to 1000 F. and a pressure of about 50 to 1000 p.s.i.g. to provide a product boiling in the motor fuel range of increased octane value, separating from this product paraffinic hydrocarbons consisting essentially of parafiinic hydrocarbons boiling predominantly in the C to C range, said paraflinic hydrocarbons containing at least about 15 weight percent of normal parafiin and having less than about 10 weight percent aromatics, contacting in the liquid phase isobutane and the separated paraffinic hydrocarbons with a catalyst consisting essentially of hydrogen fluoride and boron trifluoride at a temperature of about to 300 F. and at a boron trifiuoride partial pressure of at least about 200 p.s.i. with there being a net consumption of butanes of at least about 5 weight percent based upon platinum-alumina catalyst in the presence of free hydrogen at a temperature of about 825 to 975 F. and a pressure of about to 500 p.s.i.g. to provide a product boiling in the motor fuel range of increased octane value, separating from this product paraffinic hydrocarbons consisting essentially of paraffinic hydrocarbons boiling predominantly in the C to C9 range and having less than about 10 percent aromatics, said parafiinic hydrocarbons containing about 15 to 35 weight percent of normal paraffin, contacting in the liquid phase isobutane and the separated parallinic hydrocarbons with a catalyst consisting essentially of hydrogen fluoride and boron trifluoride at a temperature of about 75 to 200 F. and at a boron trifluoride partial pressure of at least about 4-00 p.s.i. with there being a net consumption of butanes of at least about 5 weight percent based upon the said parafiinic hydrocarbon, said contacted isobutane being about 75 to 300 weight percent of said paraflinic hydrocarbon, and separating a hydrocarbon boiling in the motor fuel range.

3. In a method of converting a straight run hydrocarbon fraction boiling in the motor fuel range, the steps comprising contacting said hydrocarbon fraction with a platinum metal-alumina catalyst in the presence of free hydrogen at a temperature of about 750 to 1000 F. and a pressure of about 50 to 1000 p.s.i.g. to provide a product boiling in the motor fuel range of increased octane value, separating by means of a glycol-water extraction medium from this product parafiinic hydrocarbons consisting essentially of parafiinic hydrocarbons boiling predominantly in the C to C range, said parafiinic hydrocarbons containing at least about 15 weight percent of normal paraflin and having less than about 10 weight percent aromatics, contacting in the liquid phase isobutane and the separated parafiinic hydrocarbons with a catalyst consisting essentially of hydrogen fluoride and boron trifiuoride at a temperature of about 0 to 300 F. and at a boron trifluoride partial pressure of at least about 200 p.s.i. with there being a net consumption of butanes of at least about 5 weight percent based upon the said parafiinic hydrocarbons, said contacted isobutane being about 50 to 600 Weight percent of said paraflinic hydrocarbon, and separating a hydrocarbon boiling in the motor fuel range.

4. In a method of converting a straight run hydrocarbon fraction boiling in the motor fuel range, the steps comprising contacting said hydrocarbon fraction with a platinum-alumina catalyst in the presence of free hydrogen at a temperature of about 825 to 975 F. and a pressure of about 150 to 500 p.s.i.g. to provide a product boiling in the motor fuel range of increased octane value, separating by means of a glycol-water extraction medium from this product paraffinic hydrocarbons consisting essentially of paralfinic hydrocarbons boiling predominantly in the C to C range and having less than about 10 percent aromatics, said paraffinic hydrocarbons containing about 15 to 35 weight percent of normal parafiin, contacting in the liquid phase isobutane and the separated paraffinic hydrocarbons with a catalyst consisting essentially of hydrogen fluoride and boron trifluoride at a temperature of about 75 to 200 F. and at a boron trifluoride partial pressure of at least about 400 p.s.i.g. with there being a net consumption of butanes of at least about 5 weight percent based upon the said paratlinic hydrocarbon, said contacted isobutane being about 75 to 300 weight percent of said parafiinic hydrocarbon, and separating a hydrocarbon boiling in the motor fuel range.

References Cited in the file of this patent UNITED STATES PATENTS 2,583,740 Kemp Jan. 29, 1952 2,740,751 Haensel et al Apr. 3, 1956 2,781,298 Haensel et al. Feb. 12, 1957 2,877,173 Thorne et al Mar. 10, 1959 2,880,164 Viland Mar. 31, 1959 2,917,449 Christensen et al Dec. 15, 1959 2,938,853 Amer et al. May 31, 1960

Claims (1)

1. IN A METHOD OF CONVERTING A STRAIGHT RUN HYDROCARBON FRACTION BOILING IN THE MOTOR FUEL RANGE, THE STEPS COMPRISING CONTACTING SAID HYDROCARBON FRACTION WITH A PLATINUM METAL-ALUMINA CATALYST IN THE PRESENCE OF FREE HYDROGEN AT A TEMPERATURE OF ABOUT 750 TO 100*F. AND A PRESSURE OF ABOUT 50 TO 1000 P.S.I.G. TO PROVIDE A PRODUCT BOILING IN THE MOTOR FUEL RANGE OF INCREASED OCTANE VALUE, SEPARATING FROM THIS PRODUCT PARAFFINIC HYDROCARBONS CONSISTING ESSENTIALLY OF PARAFFINIC HYDROCARBONS BOILING PREDOMINANTLY IN THE C8 TO C9 RANGE, SADI PARAFFINIC HYDROCARBONS CONTAINING AT LEAST ABOUT 15 WEIGHT PERCENT OF NORMAL PARAFFIN AND HAVING LESS THAN ABOUT 10 WEIGHT PERCENT AROMATICS, CONTACTING IN THE LIQUID PHASE ISOBUTANE AND THE SEPARATED PARAFFNIC HYDROCARBONS WITH A

Images