INDUSTRIAL AND ENGINEERING CHEMISTRY
1280
Vol. 17, No. 12
Thermal Expansion of California Petroleum Oils' By Edward H. Zeitfuchs STANDARD OIL Co. (CALIF.),
RICHMOND, CALIF.
I t is generally believed t h a t California oil has a higher rate of thermal expansion t h a n do Central and Eastern oils. One purpose of this investigation has been to determine experimentally the rate of expansion of typical California petroleum oils and to compare the results with the values given in Technologic Paper 77 of the U. S. Bureau of Standards, which are based on determinations made with samples collected from all parts of the United States, including the states of Pennsylvania, New York, Ohio, Louisiana, Texas, Oklahoma, Kansas, Indiana, and California. Practically all the data on which existing tables and equations for the thermal expansion of petroleum are based were determined below 100" C.
From consideration of the data available on the thermal expansion of hydrocarbons present in petroleum oils, i t was certain t h a t the rate of expansion of the petroleum oils increased rapidly a t higher temperatures.2 Reliable data on the expansion u p t o temperatures as high as 400' C. (752' F.) were urgently needed for use in refinery practice. A further purpose of this investigation has been, therefore, to determine by direct measurement the thermal expansion of typical California petroleum oils a t high temperatures. These tests extended to temperatures as high as 400' C. on some of the lieavier products.
HE materials used in this investigation were samples
for (a),except that the expanded oil was removed by unscrewing the cap, F . All temperatures were measured with thermometers calibrated by the Bureau of Standards. The temperatures up to 200" C. were measured with certainty of 0.05" C. and those above this temperature with a certainty of 0.3" C.
.. .. .. . . .. .. . .
T
of regular refinery products from typical California crude oils as received from the various departments, with no especial precaution taken to insure more than ordinary purity. The products tested ranged in specific gravity between 0.766 and 0.932 at 15.56"/15.56" C. (60"/60" F.) (53" and 20.5" A. P. I.), except in the case of 7 0 penetration asphalt, whose specific gravity was 1.0135 at 15.56"/15.56" C. Apparatus and Procedure The pycnometers are shown in Figure 1: (a)was made of Pyrex glass, (b) of steel. All determinations above approximately 200" C. (424" F.) were carried out with ( b ) . I n general, the procedure used in obtaining the data given in Table I was as follows: Refer to pycnometer (a),Figure 1. The empty pycnometer was weighed and the lower reservoir, A , was filled with oil through B to the top of the capillary B by means of a glass funnel with a small stem. The pycnometer was then immersed in a constant-temperature bath of kerosene at 25" C. (77" F.). The oil which expanded through B into the upper reservoir, C, was removed by pipet, C, cleaned with petroleum ether, and the pycnometer with oil reweighed. The end of tube D was then sealed by means of a blow torch and the pycnometer suspended in the vapor of a liquid which was brought to boiling by means of a small electric heating element immersed in the liquid. The general arrangement of the vapor bath with pycnometer suspended in it is shown in Figure 1. The pycnometer was heated until no more oil expanded from A into C. It was then removed from the bath, the tip of tube D broken off, the oil removed from C by pipet$, cleaned, and the.pycnometer and removed tip reweighed. This procedure was then repeated with a bath a t a higher temperature. The liquids used and the approximate temperatures of the vapor baths were as follows: acetone, b. p. 57" C.; benzene, b. p. 80" C.; water, b. p. 100" C.; aniline, b. p. 184" C.; and naphthalene, b. p. 218" C. Determinations a t temperatures above that of the boiling point of naphthalene were made with the steel pycnometer completely immersed in a well-stirred bath of molten solder, whose temperature was held constant to within 0.3" C. by hand regulation of the gas heater. The procedure when the steel pycnometer (b) was used was the same as that outlined 1 Presented before the joint meeting of the Division of Petroleum Chemistry and the Section of Gas and Fuel Chemistry at the 70th Meeting of the American Chemical Society, Los Angeles, Calif., August 3 t o 8, 192,:.
Calibration of Pycnometers The internal volumes of the pycnometers were determined with mercury with the exception of the volume of the steel pycnometer at 25"C., which was determined with freshly distilled water. Calibrations were made at temperatures approximately the same as those a t which the volume of one gram of oil was to be determined and a straight line was drawn through the points from which the volume of the pycnometer could be read a t any temperature with the desired accuracy. The following are volumes read from the calibration curve: Temgerature C.
Internal volume pycnometer, ml. (a) (b)
Temperature Range of Determinations The approximate temperatures a t which volumes of 1 gram of the individual sample were determined were: 25", 57", 80", loo", 184", 218", 280", 342", and 410" C. The observed temperatures and the number of points at which determinations were made are shown in Table I. Calculation of Results From the weighings and the observed temperatures at each of the several points, the volume of 1 gram of each oil given in Table I was calculated as follows:
W = apparent mass of pycnometer filled with oil a t temperature, t P = apparent mass of empty pycnometer d = density of air at laboratory temperature dl = approximate density of the oil determined by hydrometer 2
Wilson and Bahlke, THISJOURNAL, 16, 120 (19%).
INDUSTRIAL AND ENGINEERING CHEMISTRY
December, 1925
8 . 4 = density of the brass weights V = internal volume of pycnometer M = mass of oil contained in pycnometer V, = volume of one gram (in vacuo) of oil at the observed temperature, t Reduction of Observations
Having determined the volume of 1 gram of the individual samples a t the several temperatures, the volumes were plotted against the temperatures on a large scale and curves drawn through the plotted points as shown on Figure 2 for two of the oils. It was assumed that the expansion of any sample could be represented by an equation having the form:
in which V, = volume of the oil at any temperature, t V p = volume of the oil at standard temperature, T t = temperature, ' C. QI, bi, A I , and B1 are constants to be determined for each sample
-- COLJENSER
-GlASS
1281
where V, = volume at any temperature, t o C. VIS., = volume at 15.56" C . (60' F.) V ~ S=Ovolume at 260" C . (600" F.) A = a B=- b C = G D e -d v1s.E.e
V16.56
vzso
v Z 60
The method of averages3 was used for determining the most probable values of A , B, C, D. An example illustrating the various steps in the operation of determining these constants for a given sample is shown in Table 11. The results obtained for the individual samples are given in Table 111. Table I-Experimental Results VOLUME OF 1 GRAMOB INDIVIDUAL OILS AT EXPERIMENTAL TEMPBRATURES. WEIGHINGS i n vacuo Temp., O C. Ml./gram Temp., C. Ml./gram Asphalt Gas Oil (1) 25.0
0.9884 1.0336 1.0342 1.0908 1.0913 1.1158 1.1164 1.1636 1.1767 1.2211 1.2380 1.3449
100.0 100.0 184.2 184.2 218.0 218.0 279.3 294.7 342.0 357.5 410.6 25.0 100.0 218.2
18' Tar
1.0610 1.1204 1.2248
25.0 25.0 56.4 56.4 100.0 184.2 218.0 218.0 278.3' 341.3 362.7 409.0 410.1
1.1332 1.1335 1.1623 1.1664 1,2050 1.3011 1.3470 1.3463 1.4447 1.5789 1.6347 1.8722 1.8782 Gas Oil (2) 25.0 1.1353 25.0 1.1354 25.0 1.1355 100.1 1.2072 100.3 1.2084 184.0 1.3034 218.1 1.3491 218.3 1.3497
Mineral Seal Distillate
TUBE
25.0 100.4 218.3
1.1666 1.2397
25.0 100.2 218.1
1.1771 1.2557 1.3937
1.3818 Mineral Seal Distillate (Treated)
W. W. Distillate
25.0 100.4 184.3 25.0 57.1 80.2
1.1869 1.2697 1.3842
Kerosene
1.2174 1.2530 1.2805
Kerosene (Highly Treated) 25.0 100.2 183.9
1.2389 1.3277 1.4520
Crude Naphtha
25.0 56.7 100.2
1.2809 1.3238 1.3893
Engine Distillate
I
25.0 56.8 100.2 100.3
CO77ONPACK/NG
I
i
25.0 37.5 56.7 56.9 79.9 100.2
iC23-m 0 0 4
(ai GIASS P I C N O ~ ~ E ~ R
Figure 1-Apparatus
VAPOR BATH
(blST€fL PICNOMETER
for Measuring Coefficient of Expansion of P e t r o l e u m Oils
It was found impossible to represent the data for any sample over the temperature range 15.56'to 400' C. (60' to 752' F.) by a single equation : V, = V, a1(t-T) b1(t-T)* This was due to the rapid change in the rate of expansion a t high temperatures, as shown by the curves in Figure 2. To represent the data over the range from 15.56' to 400' C. (60' t o 752' F.), therefore, two equations were used; the first covering the range 15.56' to 260' C. (SO' tJo 500" F.), and the second, 260' to 400' C. (500' to 752' F.). Then from the equation given above,
+
+
+
+ + +
Vis., ~(t-15.56) b(t-15.56)' Vg or V, = VIS.@ [I A(t-15.56) B(t-15.56)21 and similarly V, = VZW+.c(t-260) 4-d(t-260)2 or V, = Vteo[l C(t-260) D(t-260)*1
+
+
1.2809 1.3202 1.3853 1.3861
Gasoline
1.3193 1.3375 1.3667 1.3672 1.4052 1.4712
Plot of Values of A, B , C, D against Specific Gravity
I n Figure 3 the values of A , B , C, D taken from Table I11 are plotted against the specific gravity. The dotted line represents the values of A and B determined by the Bureau of Standards. Source of Errors
The errors entering into the determination of the volume per unit mass of oil by the method used are of two kinds: (a) errors in weighing, and ( b ) errors in temperature measure:Lipka, "Graphical and Mechanical Computations," 1918.
& Sons, Inc. 4 Bur. Standa7ds. Tech. P o p e 77.
John Wiley.
INDUSTRIAL AND ENGINEERING CHEMISTRY
1282
ments. The following is an illustration of the accuracy of the experimental work: Two determinations of the volume at 204.4' C. (400' F.) of 1 gram of the same sample differed by 0.0007 ml. The total expansion 25' to 204.4' C. was 0.1825 ml.; the error in the volume determined at 204.4' C. without averaging the results was therefore about 0.3 per cent. It was found that the volumes of 1-gram samples of different oils of the same specific gravitx at 15-56' C. (60' F.) differed by as much as 0.002 ml. a t 204.4 C. In view of this variation shown by different oils of the same gravity at 15.56' C., the experimental errors are evidently insignificant.
Vol. 17, No. 12
Example: Observed gravity, 50' A. P. I. Observed temperature, 23.9' C. (75" F.) Corrected by Bulletin 154 to 15.56' C. (60' F.) = 48.6' A. P. I. Correction applied = 1.4' 1.4 X 1.07 = 1.5 Corrected to 15.56' C. (60' F.) by the use of the results in present paper, 48.6' A. P. I.
--
T a b l e If-Examples of Reduction of Observation VlS.66 -k ~(t--15.56) f- b(l-15.56)' in which Vt = volume nf 1 gram of oil a t ID C. I temperature,-' C. Vir.ra = the volume per gram at 15.56O C., extrapolated from 25' to 15.56' on the curve Vt-Vis.~ a(!--15.56) b(t--15.56)' Vf V1r.u o b(l--15.56)
vf
--
Y 5.56 t.
+
+
-
Thio last is a straight-line function of-6 and ( f 15.56). and o and b I- 15.56 may be computed from it. Examplo-Material Is* Tar A = Vt V:obsd. vt- vlS.56 -VI VI calcd. calcd. obsd. f I-15.56 from plot (L- 15.56) 0 16.86 0 1.0543 1.0543 0 0.00077948 1.1204 100.4 84.8 1.1204 0 1.2248 0.00084114 202.7 1.2248 218.3 0 0.00077948 a 84.8 b 0.00084114 = o f 202.76 b = 0.523 X 10-1 a 0.735 X 10-8 A = 0.697 X 10-a B 0.496 X 10-6 For tar V T V l ~ . s ~ f [ l 0.697 X 10-8 ( t - 15.56) f 0.496 X 10-1 (1-15.58)'
-
-
I
I
I
I
I
+
FIGURE 2
I
I
ri/i
D e t e r m i n e d Values of A, E, C, D Range of temoerature Sp. gr. foi which A B C D established 15.56'/ Deg. 15.56' C. A. P. I. X 108 X 106 X 10: X 10' O c. 0.7662 53.2 1.123 0.820 15.56 t o 120
Table 111-Experimentally
OIL Gasoline Crude naphtha Engine distillate
- - -- _--
0.7886
47.9
1.048
1.481
15.56 t o 150
0.7991
45.6
1.060 0.904
15.66 to 150
0.8142 0.8242
42.3 0.886 3 9 . 3 0.889
0.8493
TYPOtPd
kerosene Kerosene W. W. distillate Treated
Comparison of Results
Comparison of the average curve shown in Figure 3 with the results taken from Technologic Paper 77 indicates that California petroleum oils have a slightly lower rate of expansion for gravities less than about 25' A. P. I.; for gravities above this there is a distinct deviation in the other direction, the deviation increasing with the gravity. At 50' A. P. I. the value of A is about 7 per cent greater. One example will show the order of magnitude of the error introduced by using the National Standard Petroleum Tables6 for correcting observed gravities to a basis of 60' F. (15.56' C.). Bur. Standards, Bull. 164.
~~~~~-
tillate Mineral seal distillate
0.710
1.163
15.56 t o 200 15.56 t o 120
35.1
0.824
1.270
15.56 t o 200
0.8564
1
33.7
0.821
0.664
15.56 to 260
10.8636
32.3
0.765
0.020
16.56 t o 260
0.9321
20.3
0.677
0.564
15.56 t o 120
0.9418 0.9485 1.0173 1.0173
18.8 0.689 0.078 17.9 0.697 0.496 7 . 6 0.580 0.447 7.6
15.66 to.120 15.56 to 260 15.56 t o 260
Gas oil Light lubricating oil distillate Tar Lubricating oil Cylinder oil stock Tar 18' Asphalt
{
0.729 0.577
260to400