US2273661A - Process of refining mineral oil - Google Patents

Description

; Feb. 17, 1942. J. w. POOLE 2,273,661

PROCESS OF REFIN ING MINERAL OIL Original Filed June 9, 1935 2 Sheets-Sheet l Fig.1. 8%

Fig.4.:

LogT

8? Fig.5. a7

S0 80 so 100 ZAcefone lnvenTor. Uohn Poole ATTys.

Feb.'l7, 1942. J w POOLE 2,273,661

PROCESS OF REFINING MINERAL OIL Original Filed June 9, 1933 2 Sheets-Sheet 2 Fig 17 I F1910 Fig.

Fig.9

40 S0 60 mo ZMyIhyl Ceflosolve 'ln C oTonuldehyde l n v e n Tor.

\John W. Poole b mwkw Patented Feb. 17, 1942 Substituted for abandoned application Serial No.

675,040, June 9, 1933. This application January 27, 1938, Serial No. 187,260

7 Claims.

This invention relates to improvements in processes of refining mineral oils, and the principal object of the invention is to provide an economical solvent process possessing the advantages, which are inherent in an efficient solvent process, but which'is flexible and controllable to such anextent as to be universally adaptable with respect to the quality and yields of various products recoverable from. the numerous typesof oils which are encountered in practice.

The present application is a substitute application for my application Ser. No. 675,040, filed June 9, 1933, for Processes of refining mineral oil.

In order to accomplish thisresult it is a further object of the invention to provide a solvent which will be sparingly soluble in the oil in order that the natural ease of solubility of the various types of hydrocarbons and hydrocarbon derivatives present in the oil being treated may not be disturbed.

A further object of the invention isto provide a solvent which may be universally adaptable at ordinary atmospheric temperatures and pres-' sures and which will possess a proper volatility to be easily and non-destructively removable from the oil phase.

A further object of the invention is to provide' a solvent of the character above specified which when mixed with the mineral oil will efiect the formation of an oil phase and a solvent phase offsuch different specific gravities that they may be readily separated.

More specifically the invention relates to a process of refining mineral oil by solvent action which comprises first preparing from known' data a solvent blend of two solvents, each sparingly soluble in said mineral oil, one possessing solvent power so great as to dissolve excess quantities of hydrocarbons of highhydrogen-carbon ratio, the other possessing insufiicient solvent power to properly dissolve undesirable constituents of said oil, correlating the relative proportions of the solvents of said blend to the temperature at which the oil is to be treated andto the qualtity or the yield of the oil phase to be produced, or to a condition intermediate of the maximum yield and the best qualtity of the oil phase to be produced, or to the quality of the extract to be produced, second, thoroughly mixing the solvent with the oil to produce an oil phase and a solvent phase and thereafter separating said phases and removing the solvent from the oil phase to produce, for instance, a

lubricating oil of predetermined quality and/or" a yield or an extract. of predetermined quality.

The mathematical principles upon which the invention has: been developed are illustrated in the accompanying drawings, in which,

Fig. 1.is a graph employing the usual Cartesian coordinates illustrating diagrammatically the mathematica} similarity of the solubility of a group of comparatively pure hydrocarbons in all solvents in which their solubility is a function of temperatures;

Figs. 2, 3, 4, and 5, illustrate the effect of variouslpercentages of water blended with crotonaldehyde and the quality of the extract and of the rafiinate to such a degree as is expressed by specific gravities designed respectively as S0 for the original oil, Se for the extract, and. Sr for the raffinite, and also illustrates the corresponding effect upon the yield designated as Y. These figures vary only with respect to the viscosity of the oil and the necessaryaccompanying characteristics thereof Figs. 6, '7, 8, 9, and 10 (and also Fig. 3), illustrate the approximate similarity of all solvent blends consisting of two solvents each sparingly soluble in the oil and respectively of excessive and insufficient solvent power and demonstrate that any solvent power intermediate of two of such selected solvents may be secured by the proper blend with the accompanying effect on yield, and thatto a; great degree the quality of the raflinate' and of the extract for any one solventblend is consistent with the others for a given yield. The apparent inconsistencies of Figs; '7 and 8-shown by the dip of the yield curve are in fact correlated to differences indissolved solvent.

Fig. 11 is a comparative graph showing that when a solvent not sparingly soluble in the oil is blended with another solvent, the purpose of mixing the solvents is frustrated by the loss of selectivity due to increased solution in the oil; carbon tetrachloride being highly soluble in oil progressively decreases. both the yield and the quality of the raflinate as its content increases in the solvent blend.

Asa result of a vast number of experiments with numerous oils and solvents, and wide investigation of published data, I have become convinced that selectivity of a solvent is brought about by two very specific, necessary and sulficient requirements.

First, with respect to the mixture to be separated, the solvent must possess a limited, although ample, solvent power for' hydrocarbons and derivatives of hydrocarbons.

separation of the two resulting phases and evapo-l rating the solvent therefrom, the desired result would be accomplished. Such a result would be only possible because of the limited solvent power which caused the rejection by the solvent It, therefore, followsof the desirable material. that limited solvent power is necessary.

On the other hand, the result would be only possible provided the solvent possessed sufiicient solvent power to dissolve all the undesirable material present in the mixture; consequently, it is correct to say that ample solvent power is necessary. Thus the first requirement is-established.

Proof of the necessity of the second requirement and of the sufiiciency of the two requirements may not be as simply demonstrated. If, for instance, an unlimited amount of solvent could be dissolved in the hydrocarbon mixture it is self-evident that it would be impossible to saturate the oil phase. With many solvents, this is essentially what happens.

If the oil phase could not be saturated, it would be impossible to form a second or solvent phase, and if such a phase is not formed, obviously a process dependent upon the existence of two phases cannot be employed. Even if the solvent and oil were not completely miscible, it will be recognized that if a very large volume of solvent were dissolved in the hydrocarbon mixture, such a condition would occasion the use of unusual quantities of solvent, the solvent dissolved being manifestly useless for extraction work, thereby producing an uneconomicalcondition. Furthermore, limited solubility in the hydrocarbon phase is necessary for other reasons. There is a rule of organic chemistry, used as an approximation by physical chemists, to the effect that like dissolves like, or in other words, the more nearly alike are two organic liquids, the more miscible will they be. It therefore follows that as more of the solvent, which constitutes the major part of the solvent phase, becomes dissolved in the other phase, the more nearly similar will these two H phases become, and as a consequence the more easily would all portions of the oil phase dissolve in the solvent phase. This is graphically shown in Fig. 11 of the drawings.

. It therefore follows that the solvent must not be excessively soluble in the hydrocarbon mixture and that the lower such solubility the greater will be the selectivity. I havefound that given a definite hydrocarbon or even a cut of reasonably narrow boiling range, consisting essentially of members of the same hydrocarbon series, that 1 the solubility of such material in any solvent may be expressed (within experimental error) by the equation:

Log Ws equals A (log TKs) in which Ws isthe amount dissolved per one hundredgrams of solvent,,T isthe absolute temperature, A is a constant, for the hydrocarbon material and Ks is a constant for a given solvent, but possessing different values for diiferent solvents.

The manner in which exact values may be predicated by the use of this equation is shown in a paper published by me in collaboration with Theodore A. Mangelsdorf in volume 24, page 1215, of the Industrial and Engineering Chemistry, November 1932, Figure 8 of which shows the agreement between the calculated and observed values for a narrow cut of parafiin wax.

The meaning of this equation is that if logarithmic values of solubility are plotted on the conventional Cartesian coordinates, that solubility temperature curves for any one hydrocarbon-or closely related groups of hydrocarbonswill consist'ofstraight parallel lines, one for each solvent, as shown graphically in Fig. 1 of the drawings. In other words, every solvent below the region of critical solution, will act exactly the same as will any other solvent except for a temperaturecorrection. If this fact is true for all" hydrocarbonsand all pertinent data obtained by" me confirms me in that beliefevery solvent must'at some definite temperature 'dissolve from any hydrocarbon mixture exactly the same material dissolved by any one other solvent at some other temperature. Consequently, every solvent if used at the correct temperature would be theoretically equally selective as any other solvent. Such variationsfrom the above rule as may occur are, in my opinion, due to the solution of the solvent in the oil phase, thereby causing the oil phase as a whole to become more and more miscible with the solvent phase. If so, any means, such as adding a material which will lessen the solution in the oil phase willtend to restore the normal selective power to the solvent. If then a solvent is employed which dissolves sparingly the oil and this solvent is thoroughly mixed with a hydrocarbon mixture, the solvent may be expected to dissolve from each class of compounds according to the normal solubility in such solvent. If allhydrocarbons dissolve in every solvent according to the above equation, it is reasonable to assume that for a given oil similar percentages dissolved from thatoil by equal weights of solvents will be essentially alike in composition. This is graphically illustrated in Fig. 1 which shows how the temperature solubility curves of different steps of solvents will appear whenthe logarithm of solubility and the logarithm of temperature are plotted along-the respective coordinates. 1

, X0 represents a point of the desired solubility at temperature To. On the curves A, B, C, D, points Xa, Xb, X0, X1, also represent pointsof the same desirable solubility. On the other hand, none of these points-fall at the desired-temperature and will necessitate either refrigeration or heating if one of the solvents to which they pertain is to be used in the process.

My work has shown'that without doubt solvents ofa group comprising nitrobenzene, dichloro'ethylether, crotonaldehyde, ethylenedichloride, dioxan (diethyleneoxide), which may be designated as group 1, and which are typified by thecurves C and -D-, are-solvents sparingly soluble in'oil, and possess high solvent powers which will. give goodresults at the lower temperatures, and that solvents such as acetone, acetaldehyde, water,.methyl alcohol, ethyl alcohol, and monomethyl, etherv of ethylene glycol, etc.,,.which may be designatedas group 2, and. which are typified by thejcurves A and B, necessitate higher temperatures to possess solvent power, X, and' na'y or v may not be satisfactory at increased temperatures.

Therise in temperature probably will result in increased solubility in the oil, such increase causing the solvent to be reduced in selectivity. It will in consequence be doubtful if solvents A and B can be utilized at the temperature corresponding to Xa and Xb. Furthermore, it may not be economical to use solvents C and D at the temperatures corresponding to X and Xd since the refrigeration required to preserve the materials at such temperatures will often prove to be extremely expensive.

However, the problem may be considered as one in which we have, in order to secure the proper solubility, moved these curves A, B, and C, D, respectively down or up along the temperature axis in order to cause these curves to coincide with the curve S which may be considered the curve of the ideal solvent. Thi has been a correction by means of temperature adjustment. On the other hand, it is possible to cause these same curves A, B, and C, D, to coincide with the curve S of the ideal solvent by causing them to move along the other axis, that of solubility, and

thereby do away with the necessity of operating likely that there will be any increase in material dissolved in the oil. It may then be determined by experiment what proportion of the solvent A must be added to solvent C to cause the curve of solvent C to coincide with the curve S.

Another and practical manner in which this operation may be regarded i that of producing a blend of solvents A and C which will be intermediate of A and C in solvent power and if properly proportioned will coincide with the ideal solvent S.

The above reasoning is that by which I originally arrived at a conception of the present invention.

The accuracy of the above theory has been verified byme by numerous experiments with different solvents and with different hydrocarbon basic materials. In some instances apparent discrepancies have been observed, but in each case this has been found to be accompanied by either an increase or decrease from the anticipated degree of solubility of the solvent in the oil. This,

however, may be adjusted to approximately rethe same solvent power as a blend comprising 43% of dichloroethylether and 57% acetone. Reference to the curve of rafiinate gravity Sr shows that the first blend will result in a raffinate gravity of approximately .8825, whereas the second blend will result in a raffinate gravity of approximately .8865.

Figs. 2, 3, 4, and 5, of the drawings illustrate graphically how oils of the samegeneral-typabut varyingzin viscosity, :may be treated by this invention. From these figures there may be determined the percentages of water and crotonaldehyde to be used in a solvent blend which will be suitable to procure a raflinate of a particular quality, and the yield of such raffinate. These figures also show the manner in which the yield for eachvblend may be increased or decreased by varying the relative proportions of the components of the solvent. Figs. 2, 3, 4 and 5 aid in determining the proper blend of solvent to be employed which will produce an economical balance between the quality and the yield and give a commercially desirable lubricating oil at an economical price.

Figs. 6, '7, 8, 9, and 10 illustrate the application of'the invention to oils of substantially th same basic quality derived from a common source and having approximately the same viscosity as demonstrated by my experiments in which blends of different solvents were employed.

Fig. 6 illustrates the curves of the type above described resulting from blends in different proportions of dioxan and acetone.

Fig. *7 illustrates similar curves resulting from blends in different proportions of dichlorethylether and acetone.

Fig. 8 illustrates similar curves resulting from the use of dioxan and acetaldehyde.

Fig. 9 illustrates similar curves resulting from the use of monomethylether of ethylene glycol and crotonaldehyde.

Fig. 10 illustrates similar curves resulting from the use of acetone and crotonaldehyde.

The curves illustrated herein were plotted from data actually obtained from numerous experiments and the principles of the invention have been confirmed not only thereby, but also from data similarly obtained by the use of other solvents .mentioned herein, together with numerous other solvents having limited solubility in the oil treated.

Furthermore, it may be stated that with all the solvents used, I have found that no pair of solvents having the characteristics above described failed to act within the principles herein set forth.

Ihave for a long time recognized that oxygenated compounds of low hydrogen-carbon ratio are extremely likely to prove desirable solvents for oil refining. Through my knowledge of this fact I have developed to the stage of commercial utility processes using unsaturated aldehydes and their isomers which group of solvents is included in the above classification of oxygenated and low hydrogen-ratio compounds, and also can be successfully employed in the process herein disclosed. Diethylene oxide also falls Within the scope of I, the above classification of solvents,

inasmuch as it is both oxygenated and of low hydrogen-carbon ratio.

Concurrently with the general development of the invention herein described, I have discovered that diethylene oxide, a commercial grade of which has recently become available under the trade name of Dioxan, possesses proper solvent power, eifectively limited solubility in oil, proper specific gravity, and volatility as well as such other characteristics as are necessary for an effective selective solvent. This compound in itself apparently possesses a solvent power suitable for many oils and in consequence may, depending on the oil to be treated, be blended eitherto increase or to decrease its solvent power. For a certain range of mineral oils, particularly neutral oils, it has proved to be very Satisfactory without blending of any description. It is unusually effective as a decolorizing agent. To the best of my knowledge the use of. this solvent as a selective solvent has been heretofore unknown.

Due to the fact that a selective solvent, to be commercially useful, must conform to a number of requirements, this invention becomes particularly important. Among the requirements which must be met by a selective solvent of commercial value for refining oils are, first, the price of the solvent must be sufliciently low to make operation of the process profitable; second, the solvent must be usefully selective not for any mixture of hydrocarbons, but for the specific mixture to be processed. The solvent must dissolve the undesirables and to avoid dissolving the desirable components must be sparingly soluble in them. Third, the solvent must be of such volatility as to be easily removable from the resulting phases, and yet not require superimposed pressure to maintain it in a liquid state. Fourth, the compound must be reasonably stable in order to avoid loss of solvent by chemical changes. Fifth, it should possess such specific gravity as to result in oil and solvent phases easily separable each from the other.

The above discussion has stressed the fact that primarily a solvent to be used as a refining agent must possess selectivity. This quality is probably the most difficult of all requirements to meet.

The present invention affords an easy and inexpensive method of meeting this requirement by the use of a considerable number of blends of different solvents.

Instances in which the quality of efficient selectivity is found with single solvents are comparatively rare.

In consequence, it is difiicult to select a suitable solvent which will have all of the required characteristics to produce a proper separation of the oil into distinct phases. By this invention it is possible to select such solvents from a comparatively large field for blending as herein described which are inexpensive, which possess the proper volatility, which are resistant to chemical changes, and which when mixed with the oil will produce an oil and solvent phase of such different specific gravity that each will be easily separable from the other. Since for any solvent the solvent power is practically certain to be either too high or too low for the material to be refined, the utility of the present invention is obvious. Furthermore, since mineral oil, particularly petroleum, is an extremely variable material, the solvent power most effective for one petroleum fraction may be quite unsatisfactory for another. From a given raw petroleum, each distillate will vary in natural ease of solution from every other out and corresponding cuts from different petroleums will also differ.

In the performance of the process samples are taken of the mineral oil which it to be treated. Two solvents, one possessing solvent power sufficiently great to dissolve excess quantities of hydrocarbons of high hydrogen-carbon ratio, and the other possessing insuflicient solvent power to properly dissolve undesirable constituents of the oil, are then blended in such proportions as to produce a proper solvent value at a desired temperature and the quality of the oil and yield noted. Different solvents of high and low solvent power may be compared to determine the best to be employed in the process. When a proper solvent solution is obtained the relative values may be plotted on Cartesian coordinates as heretofore described and the relative proportions of the constituents of the blended solvents can thereupon immediately be determined by reference to the chart to produce a desired quality or a desired yield, or a commercial valuable condition of quality and yield intermediate of the best quality, and the maximum yield.

It has heretofore been stated that the object of the invention is to provide a solvent which at ordinary atmospheric temperatures and pressures will possess proper volatility to be easily and non-destructively removed from the oil phase, and that the solvents employed are blended in such proportion as to produce a, proper solvent value at a desired temperature and the quality of the oil and yield noted in the manner above specified. It is of course understood that it is not feasible to permit a solvent blend to fluctuate in temperature according to diurnal fluctuations of temperature or even to fluctuations over somewhat greater periods and that if a temperature of, for example degrees F. is once arrived at and imposed by atmospheric conditions, that approximately such temperatures will in all likelihood be necessarily maintained in the solvent blend. If temperatures were allowed to fluctuate according to diurnal fluctuations, it is evident that it would be constantly necessary to alter the composition of the solvent blend utilized.

For example, if the greatest atmospheric temperature reached under summer conditions were to be F., it is quite reasonable to believe that economics would requir operation at that certain temperature. Choosing such a temperature would avoid the necessity of resorting to refrigeration to maintain constant conditions within the system and heat exchange apparatus could be simplified to the extent that only a small amount of heating would be required.

By the use of my invention a proper blend of selected solvents may be employed to produce a predetermined raffinate, extract, or yield, or a desirable product intermediate of maximum quality and maximum yield. Furthermore, by my invention, use may be made of two solvents which in their pure state would be valueless for commercial operation, in that one might possess extremely great power to dissolve hydrocarbons and the other insufficient power to dissolve the undesirable compounds, while the blend of two of such solvents in the manner above described would be satisfactory.

Another advantage of the present invention in control of solvent power lies in the fact that it is frequently desirable to produce more than a single pair of products from a specific raw material; for example, it may be that on occasions it is desirable to improve a Mid-Continent oil to the point Where it meets the specifications for Pennsylvania oil, or an oil which is extremely light in color. In such cases it is usually necessary to remove more material than when merely desiring to produce an oil of proper color, carbon residue and emulsion tests. Such differences in dissolved quantities necessitate different solvent powers which may be obtained by adjusting the relative proportions of suitable mixed solvents.

A specific illustration which has been effected is as follows:

Certain Mid-Continent neutral oil of approximately 250 Saybolt Universal viscosity at F. was found to possess considerable quantities of material similar to Pennsylvania oil and also Approximatc.

a =la-rge volume of oil much-like so cajlled' Coastal on. It also contained theflusualdeleterious matter detrimental to all higher grade-oils Because of -move nearly 50% of the total mas er ,Inthis "case the raw 'stock was first treated with a solvent blend oi high solventpowercomprising {33% crotonaldehyde and- 17% acetone in accordance withtheprinciples of th'e presentinvention; The

solvent was evaporated from thesolv'ent phase av n n a to app atel 4 o thl extracted with a solvent blend similar tothe first exceptthat'waterwas added to decrease itsmiscibi'l ityand solvent'power. As a result nearly allot "the originally extracted material was .jrecovere'd as'a commercial on. of the Coastal type. 'The Of the products thus produced a large amount of the more valuablePennsy-lvania type of oil was-obtained, as well as--oil' -o'f the Coastal type which is also a valuablef commercial product. Even the heavy extract'is valuable asa'lubrict'nit for slow moving machinery, or asa rawmaterial for the production of high anti-knockgasoline. i

The process above defined was performed at a uniform temperature of 63 F., whereas if a pure solvent, such as crotonaldehyde, had been employed, the original extraction would necessarily have been made at approximately 32 F., and the secondary extraction at approximately 0 F., if it could have been made at all.

In view of the fact that at'60 F. all solutions are very fluid and separation easily effected, difficulties of separation caused by low temperatures may be avoided by the present invention.

My invention has been found to possess utility at temperatures somewhat greater than those normally imposed by atmospheric conditions in the case of waxy or viscous oils. For example, processing a heavy lubricating stock which after extraction produced a rafiinate of 120 F. pour point, it was found that settling and general mechanical operation was greatly improved if a temperature in the region of 130 F. were maintained in the final stages of countercurrent apparatus. Due to this somewhat elevated temperature, the solvent which at normal atmospheric conditions possessed a desirable solvent power was found to be too miscible with the oil phase and in consequence a greater amount of solvent of insufficient solvent power proved necessary. However, by properly proportioning the solvent blend, it was possible to produce such a blend to provide an efiicient and economical extracting agent.

While in the foregoing disclosure reference is made to the use of two solvents having the respective characteristics herein described, it will be understood that one or both of the solvents may be a solvent mixture or blend having the characteristics above described as exemplified in the treatment of the Mid-Continent stock above described.

original amount, This extract, obviously eing highly miscible in the original solvent blend was It will be understood that the present tion in its broad aspect comprises the use of any solvent or solvents having the characteristics herein described and used in the performance of the process as 'herein s et forth within the meaning and scope of the following claims' Iclaim: Y 7 L'The process of refining a mineral oil'containing desirablelubricating oil by selective solvent action of suitable solvents substantially liquid at usual atm'ospheric conditions under which conditions none of' thesolvents used will by itself approach maximum selective efi'iciency 'which comprises using as a solvent a blend of crotonaldehyde and another .solvent which is sparingly soluble in said mineral 0i1, ,is substantially liquid under said conditions, and possesses insufficient solvent power to properly dissolve hydrocarbons of low hydrogen carbon. ratio of said oil under said conditions, the relative proportions of the blended solvents being correlated .to the nature of the oil being treated. and to said c'onditionsto cause the resultant blend .to possess av solventpower of approximate maximum .selec-' .tive eflicieney intermediateof the solvent powers of the respective solvents, contacting the solvent blend with the oil to produce a fluid oil'phase and a fiuid solvent phase, thereafter separating the phases andremoving the solvent fromthe oil phase to produce a lubricating oilof (predetermined quality and yield. I I I 2. The process of refining a mineral oil containing desirable lubricating oil by selective solvent actionof suitable. solvents'substantially liquid at usual atmospheric conditions under Iwhich conditions none ofthe solvents used will by itself approach maximum selective efii'cienoy which comprises using as va solvent a blend of crotonaldehyde and water, the relative proportions of the blended solvents being correlated to the nature of the oil being treated, and to said conditions to cause the resultant blend to possess a solvent power of approximate maximum selective efiiciency intermediate of the solvent powers of the respective solvents, contacting the solvent blend with the oil to produce a fluid oil phase and a fiuid solvent phase, thereafter separating the phases and removing the solvent from the oil phase to produce a lubricating oil of predetermined quality and yield.

3. The process of refining a mineral oil containing desirable lubricating oil by selective solvent action of suitable solvents substantially liquid at usual atmospheric conditions under which conditions none of the solvents used will by itself approach maximum selective efficiency which comprises using as a solvent a blend of crotonaldehyde and 'acetaldehyde, the relative proportions of the blended solvents being correlated to thenature of the oil being treated, and to said conditions to cause the resultant blend to possess a solvent power of approximate maximum selective efiiciency intermediate of the solvent powers of the respective solvents, contacting the solvent blend with the oil to produce a fluid oil phase and a fluid solvent phase, thereafter separating the phases and removing the solvent from the oil phase to produce a lubricating oil of predetermined quality and yield.

4. The process of refining a mineral oil containing desirablelubricating oil by selective solvent action of suitable solvents substantially liquid at usual atmospheric conditions under which conditions none of the solvents used will by itself approach maximum selective efficiency which comprises using as a solvent a blend of crotonaldehyde and acetone, the relative proportions of the blended solvents being correlated to solvent blend with the oil to produce a fluid oil phase and a fluid solvent phase, thereafter separating the phases, and removing the solvent .from the oil phase to produce a lubricating oil of predetermined quality and yield. 5. The process of refining mineral oil containing desirable lubricating oil to produce a lubri-' eating oil -,of predetermined quality and yield, by selective solvent action of suitable solvents substantially liquid at usual atmospheric conditions .under which conditions none of the solvents used will by itself approach maximum selective efiiciency which comprises using as a solvent a mutually soluble blend of an unsaturated aldehyde having the essential solvent properties of crotonaldehyde and another solvent which is sparingly soluble in said mineral oil, is substantially liquid under said conditions, and possesses insuficient solvent powerto properly dissolve hydrocarbons of low hydrogen-carbon ratio of said oil under said conditions, the relative proportions of the blended solvents being correlated to the nature of the oil being treated, and to said conditions to cause the resultant blend to possess a solvent power of approximate maximum selective efiiciency intermediate of the solvent powers of the respective solvents, contacting the solvent blend with the oil to produce a fluid oil phase and a fluid solvent phase, thereafter separating the phases, and removing the solvent from the oil phase to produce a lubricating oil of predetermined quality and yield.

6. The process of refining hydrocarbon oil containing constituents of low and high hydrogencarbon ratio by selective solvent action of suitable solvents substantially liquid at usual atmospheric conditions under which conditions none of the solvents used will by itself approach maximum selective efficiency which comprises extracting the oil with an unsaturated aldehyde as a primary selective solvent having the essential solvent properties of crotonaldehyde under said conditions in the presence of a secondary solvent which is sparingly soluble in said oil and is adapted to reduce the solvent capacity of said primary solvent for said constituents of high hydrogen-carbon ratio, the relative proportions of the primary and secondary solvents being correlated to the nature of the oil being treated and to the conditions of operation to cause the resultant blend to possess a solvent capacity of a proximate maximum selective efilciency intermediate of the solvent capacities of the respective solvents, forming a fluid oil phase including oil constituents primarily of high hydrogen-carbon ratio and a fluid solvent phase including oil constituents primarily of low hydrogen-carbon ratio, and separating the said phases.

'7. The process of refining a mineral oil as defined by claim 4 in which the solvent is removed from the solvent phase to leave an oil extract, and re-extracting said extract with a blend of crotonaldehyde, acetone and Water to produce a lubricating oil therefrom.

JOHN WARD POOLE.

Images