US2173915A - Method of solvent refining - Google Patents

Description

Sept. 26, 1939.

D. F. RENSHAW METHOD OF SOLV'ENT REFINING Filed Dec. 18, 1936 the portion insoluble in the Patented Sept. 26, 1939 UNITED STATES METHOD 0F SOLVENT David F. Renshaw, Olean, N.

SoconyVacuum Oil New York, N. Y.,

Y., assigner to Company, Incorporated,

a corporation of New York Application December 18, 1936, Serial No. 116,508

2' Claims.

This invention is' directed to y single solvent methods for the refining of petroleum oils. These methods .of rening are those wherein a petroleum oil fraction, comprising constituent groups having differing chemical properties, is subjected to admixture, in liquid form, with a solvent liquid having preferential solvent powers for` one of the groups of constituents of the oil, followed by phase separation and removal of the solvent containing the extracted constituent group. The solvents as applied may or mayv not be mixtures of reagents, and the desirable fraction of the oil, to be obtained from the operation, may be either soluble in the solvent. In commercial adaptations of this type of refining, the solvent used is a reagent of the type of beta-beta-dichlorodiethyl ether, known as Chlorex, phenol, cresols, nitrobenzene, furfural, aniline, and numerous other organic compounds, many of whichare alcohols, aldehydes, and ethers, which, used alone, or in company with some other member of the group, has a preferential solvent power for the less desirable naphthenic constituents of the oil, is heavier than the desirable oil and settles therefrom in phase separation, and which removes the undesired oil constituent as an extract. This invention has specically to do with operations o f this type, and by way of example discusses an operation wherein the solvent used is Chlorex.

To more clearly understand this invention, ref'- erence is made to the drawing attached to this specification. In this drawing, the single figure shows a somewhat diagrammatic setup of an apparatus for the treatment of oil with Chlorex or similar solvent relining agent.

The treating system setup shown in the drawing consists of four settlers or separating drums, I, 2, 3, 4, each preceded by and connected to a mixer, designated by 5, 6, 1,-8. Oil to be treated enters through pipe 9, passes through cooler I0, pipe II, through mixer 5, and into settler I. Before its entry to the mixer, it has been joined by a ow of extract from pipe I2. Oil and extract settle apart in I and oil iiows through pipe I3, pump I4, mixer 6, into settler 2, and similarly through the subsequent pipes, pumps, settlers I5, I6, 1, 3, I'I, I8, 8,1-, to emerge in pipe I9 from the top of settler 4 as rafnate or refined oil to be removed from the system. Solvent enters through pipe 20, passes throufghpump I8, mixer 8, and settler 4, picking up in its travel therein in companywith the oil, some of the compounds to be removed, and becoming in the tersolvent or the portion mixers and Y (ci. 19e-13) minology of the art, "extract. Extract settled in I ows therefrom through pipe 2I'to move through pump I6, mixer settler 3, pick up more tar in that stage, emerge from 3 and iiow through 22, I4, 6, 2, and then, in accordance with the method in use prior to this invention, flows through 23, pump 2l, pipe I2, mixer 5, and into I. Settled tar and solvent, termed extract inclusively, separated in settler I, is removed through line 25 to storage. Flow through lines 2I, 22, 23, l and 25 are controlled respectively by iioat controls 26, 21, 28, 29 working on the interface between extract" and rafiinate in the respective settlers. It will be seen that in this operation, the supply of solvent for use in any one stage is dependent, not upon the need for solvent in that stage, but upon the ability of preceding stage to` reject extract. This is particularly acute in the rst stage of oil treatment, where the solvent available for flow with the incoming raw oil a@ through mixer 5 and settler i is dependent not upon the proportion of solvent which should be applied to the oil, but upon conditions existing in settler 2. If the proper amount of solvent is not applied in 5, raffinate going into 6 will fur- 55 nish more orless tar than optimum to settle in 2, and operation in 5 and I will be further deranged. It would seem that these variations should iron out in practice, but they do not and in fact such a system is subject to a continuous series of progressive disturbances and conditions of unbalanced operation working their 'way through the system. The net result of these is a decrease in throughput of finished oil, accompanied by a serious loss of good oil to -ex 35 tract, which must be recovered and retreated, thus further cutting down the capacity of the operation. This invention is based in part upon my discovery that these diculties can be avoided to a great degree by the provision of a method 40 whereby the supply of reagent for use with raw oil can be adjusted with regard to the needs of the oil across any range desired without deranging subsequent portions of the operation. To do this, I install by-pass line 30, and therein a control valve operated by liquid level control 28 in settler 2, but arranged in reverse, that is, to open when the level in 2 is decreasing instead of closing, as is normal. By this means, when insuilicient extract is collecting in 2, instead of 28 closing pipe 23 entirely, thus depriving raw oil. of any treatment, changing the nature of extract in I quickly, and losing good oil in extract from I before correction of the system, the valvein pipe 30 opens, to admit fextract from I or from the 55 storage tank floating on pipe 25 and the proper conditions are maintained in mixer 5 and settler I. Thus no change in character of raffinate occurs in unitsl B and 2, and the system is given a chance to iron itself out or adjust itself to new conditions.

Rates are not the only thing in which the operation on raw oil is dependent upon the rest oi the system.V The temperature oi the incoming oil can be controlled in the prior setup, but the temperature of the extract from 2 is dependent upon conditions of operation in 2, 3, 4, and this temperature is not always the best for conditions which should obtain in i. The other portlonof my invention is based upon the discovery that operation on raw oil could be greatly improved by controlling the temperature of the oil-extract mixture entering mixer 5- rather than by controlling only the temperature of the oil. To do` this, I install a valve in pipe I2, and valved pipe 3|, whereby the valve in l2 being closed, the extract flowing to mixer 5 is passed through cooler l in company with raw oil from pipe 9 and an accurate control upon temperature in mixer is thus obtained.

While the description above is directed to a setup having four settling stages, it is to be understood that the details of operation herein disclosed center around the two mixing and settling stages which are first in the order of oil ow and that the invention is applicable to any system of treatment making use of two or more stages of treatment, each stage consisting of a mixing step followed by a settling step. Mixers of any convenient form may be used, and settlers may, of course, be vertical, inclined, or horizontal.

To show the effect of these changes, the following data is presented. Before changing, when treating Pennsylvania cylinder stocks with 2.2 parts Chlorex to 1 oil, nearly 10% of treating time was lost in stoppage to overcome diillculties, about 28% of the charge escaped into extract", requiring subsequent recovery, and only about 70% yield was obtained on charge due to lost time and required rerunning. Upon the addition of the step of recirculating extract from A100 viscosity index, to raise #l settler to charge, the down time decreased to about 1%, although the charge lost to extract was still high due to lack of proper temperature control, and the recovery of treated oil on oil charged in a given period was raised to about 80%. Upon addition of temperature control to mixed feed to mixer 5, stoppages due to unbalanced operation were entirely eliminated, loss of charge to extract became-negligible and the yield of treated oil on oil charged during a given period raised to about 90%-, the same as the net yield, because of lack of necessity for rerunning.

I have found that when treating an oil of about it to about '7-110 viscosity index, that optimum operating conditions are obtained with a solvent oil ratio of about 2:1, a temperature on incoming solvent of about 65 F., and a temperature on feed to first mixer of about 65 F.

I claim:

1. A process of treating oil with a solvent reagent to selectively dissolve and'remove therefrom naphthenic constituents comprising the following steps: flowing oil and solvent in counter A current through a system comprising a plurality of stages each consisting of a contacting step and a settling step, removing from each settling stage an extract consisting of naphthenic constituents dissolved in a substantial amount of solvent, which extract is introduced into 'the stage next preceding in the direction of oil flow, removing an extract consisting of naphthenic constituents dissolved in a substantial amount of solvent from the stage in which oil is first introduced, admixing at least a portion of said extract last named with the "extract withdrawn from the next sucseeding stage in the direction of oil flow, and thereby initially contacting the oil to be treated with solvent at a solvent/oil ratio controllably increased over the ratio atwhich solvent and oil are fed to the system.

2. The method of claim 1 in which the mixture of extracts is admixed with oil to be treated and the resulting fluid is subjected to controlled cooling prior to entry to the treating system.

DAVID F, RENSHAW.

Images