Bubble Point Pressure Correlation

MAGNOLIA PETROLEUM CO.

J. A. LASATER
DALLAS, TEX.

ABSTRACT tern and Mid-Continental United /1"

flf=---H (4)
States, and South America. no
11" +
A correlation of the bubble point
pressure for black oil systems is de- The number of mols of tank oil
CORRELATION DEVELOPMENT per barrel is a function of the "mc-
veloped using the standard physical-
chemical equations of solutions. The The basic assumption used in this lecular weight" of the tank oil. Al-
correlation is based on 158 experi- development is the same as employed though the tank oil is a complex mix-
mentally measured bubble point pres- by Standing', ture, it was assumed for the purposes
sures of 137 independent systems po = feR, Yo, t, r). (1) of this correlation that a unique mo-
and is expressed in terms of the There is a wide variety of ways to lecular weight could be assigned to a
usually measured .field parameters- combine these parameters; however, given oil. This was designated as the
flash separation gas-oil ratio, tank oil in this instance the combination was "effective molecular weight", and was
gravity, total gas gravity, and reser- made on the basis of Henry's law'. related to the oil gravity,
voir temperature. Accordingly, the relationship pro- M, = fer). (5)
The data were obtained on sys- posed is This empirical relationship was
tems produced in Canada, Western Po = y"H'. (2) developed simultaneously with the
and Mid-Continental United States, Although Eq. 2 defines an individual correlation by assuming values of Mo
and South America. The average er- system, it is of limited value since H' and working to obtain a smooth
ror (algebraic) in the representation is a function of gas-phase composi- curve for both the correlation and
is 3.8 per cent, and the maximum er- tion and the system temperature. the effective molecular weight. The
ror encountered is 14.7 per cent. It was observed that for the sys- relationship between the oil gravity
tems where the bubble point was and the effective molecular weight
measured at several temperatures that used in this correlation is shown in
INTRODUCTION
the ratio of the bubble point pres- Fig. 1.
In the absence of experimentally sures and the ratio of the correspond- The effective molecular weight is
measured properties of reservoir ing absolute temperatures (OR) were somewhat higher than the molecular
fluids, it is often necessary for the practically identical. Thus, for cor- weight of the C,+ fraction. The dif-
field engineer to make estimates re- relation purposes the bubble point ference between these values is larg-
garding the fluid properties based on pressure may be taken as a direct est for the low-gravity systems. It is
the usually measured producing pa- function of the absolute temperature. noted that this effective molecular
rameters. To aid in these estimations, This relation is valid only for those
various correlations have appeared in systems that are not near the critical
the literature in recent years. Among point. Accordingly, this correlation
the pertinent properties of interest is will be inadequate for systems in the
the bubble point pressure. A correla- region of the critical point.
tion for this parameter has been re- The solubility of the various hy-
ported by Standing'. However, this drocarbons found in the gas phase
correlation was based essentially on increases with the molecular weight.
California produced crudes and this Thus, the saturation pressure should
limitation was pointed out with its be inversely related to the gas grav-
presentation. The correlation pre- ity.
sented in this paper utilized data on Applying these principles to Eq. 2
crude oil systems from Canada, Wes- and rearranging terms gives:
(Po)(y,,) - H (3)
Original manuscript received in Society T - Y"
of Petroleum Engineers office Sept. 15. 1957.
Revised manuscript received April 3. 1958. The variables on the left side of EFfEctIVE MOI.EC1.\.AR WE1IHT Of TANK OIL.
Paper presented at Fall Meeting of Southern
California Petroleum Section in Los Angeles. Eq. 3 were designated as the "bub- FIG. 1- EFFECTIVE MOLECULAR WEIGHT
Oct. 17-18. 1957.
lReferences given at end of paper. ble point pressure factor". RELATED TO TANK OIL GRAVITY.

MAY, 1958 SPE 957-G 65

weight relationship corresponds TABLE I-SMOOTHED BUBBLE POINT FACTOR permits rapid graphical evaluation of
FUNCTION
c105ely to that given for crude oil Gas Mol Bubble Point the predicted bubble point. This cal-
Fraction Pressure Factor
systems with a UOP characterization culation chart is shown in Fig. 3.
0.05 0.17
factor" of 11.8. 0.100 0.30 The error distribution using the cal-
0.150 0.43
The relationship between the bub- 0.200 0.58 culation chart is essentially the same
0.250 0.75
ble point pressure factors calculated 0.300 0.94 as found for the correlation_
0.350 1.19
from the experimental data and the 0.400 1.47
gas mol fraction is shown graphically 0.450 1.74
0.500 2.10 CONCLUSION
in Fig. 2. Representative values of 0.550 2.70
0.600 3.29
the curve are given in Table 1. Since 0.6'0 3.80 The correlation is generally appli-
0.700 4.30
the representation of Fig. 2 is not a 0.750 4.90 cable to a large number of producing
0.800 5.70
linear function of the variables, H 0.850 6.70 areas and provides a rapid method of
is not a constant. Thus, a simple estimating the bubble point pressure
analytical expression was not ob- of crude systems with a reasonable
tained, and it is necessary to rely on degree of accuracy_
the graphical representation of Fig. yg = R/379.3 (6)
2 to obtain Pt from YQ' R/379.3 + 350 y,' NOMEN CLATURE
The bubble point pressure for a M,
given gas-oil system may be obtained The value of the gas mol fraction f = function
from the correlation by the utiliza- is applied in Fig. 2 to obtain the yg = total gas gravity (air = 1.0)
tion of Figs. 1 and 2. The effective bubble point pressure factor. The
molecular weight is established from bubble point pressure is calculated y, = tank oil specific gravity
the crude oil gravity, Fig. 1, and the by use of the following equation H' = general Henry's law constant
gas mol fraction is obtained from the Pb = (Pr)(t + 459.6) (7) H = specific Henry's law constant
following equation. yg (independent of gas com-
The 158 experimentally deter- position and temperature)
mined bubble point pressures of the M, = effective molecular weight of
6.4
I
0
0
137 independent crude oil systems
were compared with the values pre-
tank oil
'.0. dicted by the correlation. This com- ng = mols of gas

0
I parison showed an algebraic devia-
tion of 3.8 per cent. Approximately
n, = mols of tank oil
5.6
yg = mol fraction of gas

I
21 per cent of the data points show
5.2 a deviation of 0.5 per cent, and 80 p, = bubble point pressure, psia
0
per cent have a deviation of less Pt = bubble point pressure, factor
1
ow than 6.5 per cent. The maximum er-
4 .•
01 R = total flash separation gas-oil
09 ror encountered was 14.7 per cent.
,,"'0 ratio, cu ft/bbl (measured
... The ranges of the field measured

J
'---'-' 4.0.
J
1"
parameters covered in this correla-
tion were as follows. r =
at 60°F)
tank oil gravity, °API (cor-
rected to 60°F)
Bubble point pressures 48-5780 psia
3.6 t Flash gas-oil ratios 3-2905 cu ft/bbl t = temperature, OF
Tank oil gravities 17.9-51.1° API
\C#:J Total Gas gravities
(air = 1.0)
T = absolute temperature, OR
0.574-1.223

1
'.2 Reservoir temperatures 82-272°F
Separator stages
(stock tank = 1) 1-3 ACKNOWLEDGMENT
Separator temperatures 34-106°F
2 .• Separator pressures 15-605 psia
The author is indebted to C. A.

I
This correlation was based on sys- Connally and L. G. Sharp for their
2.'
tems essentially free of non-hydro-
many helpful suggestions pertaining
carbon material. The presence of
2.0.
large amounts of nitrogen, carbon to the phase behavior of fluids, and
dioxide, hydrogen sulfide, etc., will to the Magnolia Petroleum Co. for
°1
permission to publish this work.
I.•
010 result in the predicted bubble point
,0° being low. The following is given as
1.2 a guide to the effect of the presence REFERENCES
J of non-hydrocarbon materials.
0.0

°
I Component
Per Cent
of Gas
Error in
Predicted Pb
(Per Cent low)
1. Standing, M. B.: Drill_ and Prod.
Prac., API (1947) 275.
2. Hougen, O. A., and Watson, K. M.:
/
0.' Carbon dioxide -9-.1- 5.0
Hydrogen sulfide 3.1 Chemical Process Principles, John
Carbon dioxide 3.1 1.1
Wiley & Sons, Inc., N. Y. (1943) I,
0. Nitrogen 2.5
0. 0.2 0.. 0.. c.• 1.0
Carbon dioxide 0.3 2.7 146.
GAS MOL FRACTION
3. Watson, K. M., and Murphy, G. F.:
FIG. 2-CORRELATION OF BUBBLE POINT A calculation chart has been pre- Ind. Engr_ Chem. (1935) 27, 1460.
pared based on this correlation that

66
PRESSURE FACTOR.
***
JOURNAL OF PETROT·EUM TECHNOLOGY
 FIG. 3-CHART FOR CALCULATION OF BUBBLE POINT PRESSURE.
ill:
;.. 6000
><
5000
.... «
~ (j)
Cit 4000
0..
c:c
3000 II lIiI/I I 7 1 7 ,v' 7 -1 7 , 0 / fO •• 3000
W
2500 n:::
TANK OIL GRAVITU ::>
°API (60 0 F) _ (j)
2000

21 7~LL'
2 (j)

1500 l//IJ/ I I I J 1500

W
Ct:
0..

I-
«-" Z
1000 0
CoO 0..
\... 800
W
....J
(()v CD
(()
600 CD
::>
\ /17 o· 74,0;'-;>1;>;/ I 500 CD

{ 400

v-0
300
0'
J..........
~'<

0.:::'
/
c,
~ 7

EXAMPLE

REQUIRED:
BU B B LE POI NT PRESSURE
LIQUID AT 200· F HAVING A
GAS I 01 L RATIO OF 500 CU FT I BBL,
A TANK OIL GRAVITY OF 30· API, AND
A GAS GRAVITY OF 0.8.

PROCEDURE

10 ~~-+-L-L~4-L+ __~~__~____+-__-+____+-__-+__~~~~ STARTING AT THE LEFT SIDE OF THE CHART

PROCEED HORIZONTALLY ALONG THE 500 CU FT /
V I T/y§/
Ll:/:tZIICOR~lu~~IO~ I I . if~O> I
BBL LINE TO AN OIL GRAVITY OF 30· API. NOW
DROP VERTICALLY TO THE CORRELATING CURVE, THEN
GO HORIZONTALLY TO THE 200· F LINE, NOW RISE

~ V/~ VERTICALLY TO THE 0.8 GAS GRAVITY LINE THE

REQUIRED BUBBLE POINT PRESSURE IS READ
HORIZONTALLY AT THE RIGHT - 2625 PSIA