Factors in Doctor Sweetening Influence on Inhibitor Susceptibility of Cracked Gasoline 6
T
HE s w e e t e n i n g of C. D. LOWRY, JR., C. 0.DRYER, CHARLES WIRTH, 111, out. It consisted in shakAND R. E. SUTHERLAND gasoline with doctor ing together gasoline and Universal Oil Products Company, Chicago, Ill. doctor solution in a bottle or solution and sulfur is one of the oldest and most separatory funnel for a short widely used refining procperiod and then adding esses. The fuiidanieiital reactions are simple, and yet the side flowers of sulfur, a little a t a time, with vigorous shaking bereactions commonly occurring make the process exceedingly tween additions; the amount of sulfur was limited to the minimum necessary to break out the lead sulfide. The main point complex. It is doubtful whether any of the papers published on doctor sweetening covers all the changes occurring. of importance was avoidance of excess sulfur dissolved in the There have been serious objections to doctor sweetening as mixture. By making comparisons between plant-sweetened gasoline carried out in many refineries. Principal among these are ( a ) poor inhibitor susceptibility of the sweetened gasoline, ( b ) and gasoline sweetened in the laboratory by the method dehigh cost, and (e) lowered octane rating of the gasoline. scribed, the factors in plant doctor sweetening which affect inhibitor susceptibility were studied. The experimental work These disadvantages have sometimes been considered inevitable and caused the replacement of doctor treating by upon which this paper is based was done in a number of refineries; a large part of it, however, was carried out in a single other sweetening processes. It seemed worth while to derefinery, using a continuous doctor sweetening plant with a termine to what extent the objections to this process could be reduced by careful control and modification of treating procethroughput of 500 barrels (79,485 liters) per hour. In this plant the gasoline was mixed with doctor solutlon and sulfur dure. The points tabulated have all been studied in the by passage through a series of orifice mixers, and in part of present investigation. This paper deals only with inhibitor the tests a Turbo-Mixer provided additional mixing. The susceptibility, which was briefly discussed by Altschuler and Graves (1) and Schultz and Buell (IO)after the present study mixing period was varied from 30 seconds to 3 minutes. was begun. The refining treatment given a large proportion of the cracked gasoline marketed a t the present lime consists of doctor sweetening and Improper doctor sweetening, particularly the use addition of an oxidation inhibitor. No acid, of excess sulfur, has been found to influence adclay, or similar agent is used, and the gasoversely the effectiveness of oxidation inhibitors in line is generally dyed. Extensive storage tests (4) have proved that this simple refining prosweetened gasoline. By sweetening at a temperaduces gasoline which will not form gum or deture not below 85"C., properly mixing gasoline, docpreciate in antiknock value during long periods tor solution, and sulfur, and allowing adequate of storage. I n fact, such inhibited gasolines settling time, doctor sweetening can be carried are often superior in storage stability to those out with a minimum amount of sulfur and protreated with clay or acid. An obstacle to the extension of inhibitor use has duces a gasoline in which the inhibitors will be conbeen the variation in results sometimes obtained sistently effective. The harmful effect of excess in the treatment of gasolines cracked from the sulfur is apparently due to the formation of polysame stock. I n practical operations the inducsulfides, particularly those higher than trisulfides, tion period has a t times been high in the gasoline which remain in the gasoline. Equations are preproduced one day and but half its previous value on the next, although no change was made in the sented to account for the formation of these polyamount of inhibitor used. Sometimes for a sulfides. period of months the treating operation would run smoothly, and then for no apparent reason the induction period would drop and copper dish gum increase shardv. Such a change has beenibserved several times on the incideice of cold weather. Settling was in two horizontal and one vertical drums, the These irregularities have been traced in many cases to total settling time being about 2 hours. No water wash was faulty sweetening practice. Since it is an old refinery process, given. From the treater the gasoline passed to run-down sweetening with sodium plumbite is conducted in different tanks and was inhibited when being transferred after 5 to 12 refineries in a variety of ways, often without careful chemical hours of further settling. supervision. When the variables affecting inhibitor susceptiSulfur Addition bility are known and proper control measures instituted, it has been foun possible to minimize the irregularities in inThe general opinion among treaters has been that if a gasohibitor effecti eness. line is sweet and noncorrosive to the copper strip, its properAs a check on plant sweetening of cracked gasoline a proties leave nothing t o be desired. In the case of distillatesmore cedure for reproducible laboratory sweetening was worked than slightly sour, this is not the case. To such gasolines far
$!
1275
*
VOL. 30, NO. 11
INDUSTRIAL AND ENGINEERING CHEMISTRY
1276
more sulfur may be added than is necessary for sweetening without producing a corrosive product. The harmful effect of excess sulfur is shown in Table I, which gives the results of sweetening a distillate containing abodt 0.06 per cent mercaptan sulfur. OF SULFUR IN SWEETENING WESTTEXAS TABLE I. EFFECT CRACKED GASOLINE
Excesa sulfur, % Induction period, min.: No inhibitor 0.013% U. 0.P. inhibitor 4 Copper dish gum, mg./100 00.: No inhibitor 0.013% U. 0.P.inhibitor 4
0
0.005
0.0075
0.010
0.015
200 380
210 400
115 285
125 210
100 175
299 5
273 6
327 74
389 82
357 138
A larger amount of inhibitor was used in the samples shown in Table I than is required for most gasolines. The analyses were made by methods given in U. 0. P. Laboratory Test Methods for Petroleum and Its Products (1.2). The time of evaporation in determining copper dish gum was 2.5 to 3 hours. None of these sweetened samples, despite the high sulfur content, was corrosive by the A. S. T. M. test. The copper -strip test, therefore, is not sufficiently delicate to be used for ‘control of sulfur addition in doctor sweetening. The excess sulfur, as shown in Table I, greatly reduced the induction period either with or without inhibitor, and the copper dish gum was not, reduced nearly so well by the inhibitor in the presence of excess sulfur as when the gasoline was properly sweetened. I n the samples in Table I, sulfur was added to the gasoline before contact with doctor solution. Sulfur added to sweet gasoline does not give the same results. There is usually no marked influence on inhibitor susceptibility, and even a small amount of sulfur produces a corrosive product, as shown in Table 11. TABLE 11. EFFECTOF SULFURADDITIONTO WEST TEXAS REFORMED GASOLINE AFTER SWEETENING Added sulfur, % Induction period, min.: No inhibitor 0.013% U. 0.P.inhibitor 4 Copper dish gum, mg./100 cc.: No inhibitor 0.013% U. 0.P. inhibitor 4 Corrosion, A. S. T. M. .I
0
0.005
0.010
0.015
135 435
110 435
105 435
110 435
1.5
44 19 Very pos.
59 9 Very pos.
67 40 Very pos.
8 Neg.
.
Determination of active sulfur was made in a number of samples of plant-sweetened gasolines by the method of Wirth and Strong (IS) (U. 0. P. method H-19, IS). A known amount of butyl mercaptan is added to a gasoline, doctor solution is added to cause reaction of the sulfur, and, after removal of the doctor solution and acidification, the remaining mercaptan is back-titrated with silver nitrate. It is an exceedingly useful test in checking doctor plant operation. Many of the plant-sweetened gasolines tested contained as much as or more sulfur ip excess than the highest amounts added to the gasolines in Tables I and 11. Study was therefore made of the possibility of reducing the sulfur used in plant operation. Excessive sulfur is added by many treaters to improve the break and accelerate settling of the lead sulfide in the sweetening plant. Satisfactory results were obtained without use of excess sulfur by attention to (a) mixing, ( b ) settling, and (c) temperature of gasoline.
gasoline and doctor solutions are mixed. The mixing may be done by orifice plates, baffled pipes, pipes with right-angle bends, or mechanical mixing devices, and mixing time may vary from a few seconds to several minutes. When sulfur is added to a mixture of gasoline and doctor solution, the oil becomes orange-red in color; after a longer or shorter time, depending on the gasoline and the amount of sulfur added, the “break” takes place, when the color disappears and a red-brown to black precipitate, usually called lead sulfide, forms and settles. I n different plants, the break may be produced as the gasoline leaves the mixers or it may not occur until after the gasoline has entered the settling drum. If a plant has adequate mixing (the proper time of mixing will depend on the type of gasoline and somewhat on the degree of sourness), the gasoline may be allowed to break as it leaves the mixers, and the amount of sulfur required to do this will not be so great as to harm inhibitor susceptibility. However, if a plant has little mixing equipment (a common installation is three mixing nozzles in about 6 feet of line, which for most gasolines is much less mixing than is desirable), a break cannot be obtained in the mixers without using a considerable excess of sulfur, with consequent detriment to the inhibitor effectiveness. If, however, the break in such plants occurs in the first settler from 3 to 15 minutes after the gasoline leaves the mixers, the results from an inhibitor standpoint will usually be satisfactory. But when the break occurs in the settler, the gasoline being no longer in contact with doctor solution, it tends to produce a finer precipitate, settling more slowly than when the gasoline is broken out while still in contact with doctor solution. As plants deficient in mixing equipment are also often lacking in settling capacity, the treater in such cases faces a serious dilemma. If he uses excess sulfur, his gasoline settles well but inhibitor susceptibility iapoor; if sulfur is kept down, the gasoline breaks and settles slowly and lead sulfide leaves the plant suspended in the gasoline. There are some indications that the place of addition of the sulfur affects the speed of break. I n one series of tests the doctor plant was hooked up so that sulfur could be added either with the doctor solution or subsequent to it. The operator desired the break to occur as the gasoline left the mixers. When adding the sulfur after the doctor solution, this condition was readily obtained without using an excessive amount of sulfur. When the sulfur was added with the dootor solution, if a break a t the same point was desired, excessive sulfur was necessary with consequent loss of inhibitor effectiveness; but if a slower break was taken, results were just as good as if the sulfur was added later. Comparative results are given in Table 111. A possible explanation of this behavior is given later under “Reactions during Doctor Sweetening.” OF ADDITION OF SULFUR TABLE 111. EFFECTOF PLACE
-Induction inh%?tor Laboratory-sweetened Plant-sweetened sulfur added with doctor soln.: Minimum for break in mixers Slow break in settler Sulfur. added after doctor soh.: Minimum for break in mixers 0 U. 0. P. method H-19.
Period0.013%
u., 0.P.
inhibitor 4
Active Sulf up*
Min.
Min.
%
175
390
0.001
145 180
255 390
0.023
165
375
0.008
0.012
Mixing and Settling Doctor plants differ in the way in which doctor solution and sulfur are mixed with the gasoline and the length of time they remain in contact. It is commonxpractice to add the sulfur in gasoline solution before, a t the same time as, or after the
It has been found convenient to control the addition of sulfur in practical doctor plant operation by the widely used butyl mercaptan test. T o 30 cc. of gasoline, filtered free from lead sulfide and in a 4-ounce (118-cc.) oil sample bottle,
INDUSTRIAL AND ENGINEERING CHEMISTRY
NOVEMBER, 1938
are added 20 cc. of a dilute butyl mercaptan solution (1 to 1400 by volume in cleaners' naphtha, containing no free sulfur) followed by 10 cc. doctor solution. The bottle is shaken for 15 seconds. The sample should then be light yellowgreen in color and transparent. An orange or brown color denotes free sulfur in such form as to make the gasoline corrosive. On standing the sample will become opaque and may retain its green-yellow color or turn orange or brown, a solid precipitate finally forming. If the color of the sample and precipita,te remains yellow for 30 minutes, the gasoline contains a low enough concentration of active sulfur to give optimum results with inhibitor. If, however, the sample becomes orange within 30 minutes, too much sulfur has been used and the inhibitor susceptibility will be below its maximum value. This is shown in Table IV, which presents data from the sweetening of a West Texas-mid-continent cracked gasoline blend containing 0.05 per cent mercaptan sulfur. Used in this way, the butyl mercaptan test is not primarily a test for elementary sulfur. If sulfur is present even in minute amount in sweet gasoline, it will give an immediate dark color in this test. The slow reaction producing the orange color is probably due to the breakdown of unstable sulfur compounds, thought to be polysulfides. It is believed that this test does not detect all polysulfides, for gasolines which are satisfactory according to this test, from an inhibitor standpoint are sometimes lower in octane number than the corresponding sour samples. The quantitative test for elementary sulfur of Wirth and Strong (IS)is apparently a test for polysulfides as well as free sulfur, and can be correlated with the qualitative test already discussed. A sample of sweetened gasoline which in the qualitative butyl mercaptan test will not turn orange usually contains not over 0.003 per cent active sulfur by the quantitative teat. An alternative means of detecting excess sulfur is shaking with mercury; after several minutes of agitation i t should not give more than a slight precipitate in the gasoline. '
TABLE Iv. RELATION O F INHIBITOR SUSCEPTIBILITY MERCAPTAN TEST Sweetening High sulfur Low sulfur Minimum sulfur 0
-Induction
Period-
No
$08!p
inhibitor inhibitor.4 Min. Min. 130 185 170 240 190 305
U. 0. P. method H-19.
Active Sulfur=
TO
BUTYL
Appearance o f
Butyl Mercaptan
Test
% 0.019 0.009
0.003
Orange in 4 min. Orange in 14 min. Did not t u r n orange in 30 min.
If a gasoline being sweetened is of considerable sournessfor example, containing over 0.03 per cent mercaptan sulfurit is advantageous to add the sulfur in solution in sweet rather than sour gasoline. When minimum sulfur is used, results as far as inhibitor use is concerned are not greatly improved, but the break TB sharper and the plant is more easily operable when using the sweet gasoline. This is related to the statement made earlier that mixing doctor solution and gasoline before adding sulfur gives improved results. If a gasoline requires a large amount of sulfur to sweeten it, a considerable proportion of gasoline will be introduced with sulfur midway of the plant; if this is sour gasoline, i t will tend to nullify the effects of the prior mixing of doctor solution and gasoline. The advantage of sweet gasoline for solution of sulfur was demonstrated as follows: A plant treating unit was regulated so as to use no more sulfur than was necessary for easy operation with sulfur dissolved in sour gasoline. A sample of gasoline was then taken and found to have the properties shown in Table V (sample A-3).
1277
The active sulfur content was high, induction period low, and copper dish gum not well brought down by the amount of inhibitor used. The sulfur was then reduced to the minimum on which the plant could be operated for a short period with extremely close chemical control and sample A-7 drawn, which, i t will be noted, had good properties. Change was then made to use of sulfur dissolved in sweet gasoline and the plant again regulated for easy operation. It will be noted that sample C-3, taken a t this time, had properties definitely better than A-3. Further reduction in sulfur IN SWEET AND SOUR GASOLINE TABLE V. ADDITIONOF SULFUR
-Induction
Sample
No.
A-3
A-7
C-3 C-7
PeriodCopper Dish Gum 0 0077 0.007% No in- U: 0.g N o in- U..O: P. Description hibitor inhibitor 4 hibitor inhibitor 4 Min. Min. Mg.1100 cc. Sulfur in Sour Gasoline Usual operation 130 180 188 100 Minimumsulfur 200 345 164 18 Sulfur in Sweet Gasoline 220 385 194 23 Usual operation Minimum sulfur 235 380 136 19
Active Sulfur
% 0.015
0.002
0.003 0.002
was attempted, and sample C-7 was drawn, but the content of active sulfur was but slightly less and the induction period and copper dish gum were no better than C-3. Hence, for the same ease of operation and the same rapidity of "break" approximately the minimum amount of sulfur was used when dissolved in sweet gasoline whereas much more sulfur was required when dissolved in sour gasoline. This test is subject to human error and may not be considered conclusive. It does show, however, that with very sour stocks, the use of s u l h r dissolved in sweet gasoline gives a plant operable without difficulty on minimum sulfur. The settling of the lead sulfide a t thiB time was very good, only a trace remaining in the gasoline leaving the settlers.
Temperature The general opinion among practical doctor treaters is that sweetening is facilitated by having the gasoline warm. Temperatures between 80" and 90" C. are considered desirable, and sweetening below 65" C. is regarded as poor practice. Laboratory tests showed that less sulfur is usually required a t 80' than a t 60" C. A plant test showed that although more than a minimum amount of sulfur was as harmful a t 85" C. as a t lower temperatures, with minimum sulfur the break was sharper and settling definitely better a t the high temperature.
Reactions during Doctor Sweetening The reactions of Equations 1 and 2 have often been taken to cover the chemistry of doctor sweetening: 2RSH
+ NazPbOz (RS)IPb + 2NaOH + S +PbS + RS-SR
(1) (2) Ott and Reid (6) showed that basic mercaptides ar$e also formed according to Equation 3, RSH h'azPbOz+RS-Pb-ONa NaOH (3) and that the reaction products of the treatment of gasoline with doctor solution are mixtures of mercaptides of the two types, their proportion depending upon the concentration of lead and caustic in the doctor solution used. These workers ('7) also proved that the reaction of sulfur with lead mercaptides is not limited to the simple change of Equation 2 and that the precipitate thrown down is far from being pure lead sulfide. According to them, a normal mercaptide may add sulfur to give compounds of the types: (RS)zPb
+
+
and
Pb
