PC 2 2013 Rahuma 209
PC 2 2013 Rahuma 209
PC 2 2013 Rahuma 209
ISSN 1337-702
ABSTRACT
In this work the idea was to apply literature models to predict viscosity of crude oils. Fifteen different
oil samples from Libyan reservoirs have been used. Experimental data collected from the Libyan
Petroleum Institute database. Supplementary data were collected from published literature. Chew-
Connally, Beggs-Robinson, Labedi, Kartoatmodjo and Schmidt, modified Kartoatmodjo, and Elsharkawy
& Alikhan models have been investigated for saturated oil , while Beal , Vasquez and Beggs, Khan
and Ali, Labedi ,Kartoatmodjo and Schmidt, and Elsharkawy & Alikhan have been used to undersaturated
oil. The accuracy of models for investigated viscosity were applied to the acquired data set and a
comprehensive error analysis was performed based on a comparison of the predicted value with
the experimental value. The results obtained indicate, that modified Kartoatmodjo model is the
best and accurate for saturated oil viscosity, while Beal’s model is the best for undersaturated oils.
The less AARE % was calculated 23.9% for saturated oil, and 21.9 % for undersaturated oil.
Key Words: Viscosity; model; sample; saturated oil; undersaturated oil.
1. OVERVIEW
Viscosity is a strong function of the temperature, pressure, oil gravity, gas gravity,
and gas solubility. Viscosity could be measured by two methods semi-theoretical which
are derived from a theoretical framework, but involve parameters experimentally determined,
or empirical, which include a wide variety of equations used throughout the industry.
Empirical models can be classified to corrective, which is an experimental mixture data
used for evaluation of model parameters, and predictive which is the properties of pure
components are utilized.
The viscosity is usually reported in standard PVT analyses. If laboratory data are not
available, engineers may refer to published correlations, which usually vary in complexity
and accuracy depending upon the available data. The viscosity of hydrocarbon mixtures
and petroleum reservoir fluids is commonly measured by either the rolling ball viscometer or
the capillary tube viscometer. In the rolling ball viscometer, the time required by the steel
ball to travel through the fluid is correlated to its density and viscosity. Other nonconventional
methods include the laser light scattering and the piezoelectric quartz crystal techniques.
The objective of this study is to apply an accurate correlations to predict oil viscosity
at various operating conditions. In the literature, several empirical correlations have been
proposed for oil viscosity prediction (saturated, and undersaturated). Here, based on Libyan
oil reservoirs data, accuracy of these correlations has been confirmed by comparing the
obtained results of these correlations with experimental data for Libyan oil samples. Statistical
analysis is the criteria adopted for the evaluation in this study.
2. MATERIAL AND METHODS
PVT experimental data of fifteen samples include bubble point (saturated) pressure
Psi , API gravity and solution gas-oil ratio Scf/STB at reservoir temperature Fº. A typical
viscosity plot as a function of pressure is shown in Figure 1 (sample No 6). It can be seen
from Figure 1 that, the oil viscosity decreases with pressure reduction in single phase
(undersaturated condition). This reduction continues to bubble point. As the pressure
decreased below the bubble point pressure causes gas release. Therefore the oil viscosity
K. Rahuma, E. Edreder/Petroleum & Coal 55(2) 113-136,2013 134
has been increased. It can be concluded that the minimum value of the viscosity is at
bubble point pressure.
Statistical experimental data are shown in table 1. At pressures above bubble point
pressure, oil is at single-phase state, while its solution gas–oil is constant and it seems
that pressure will be the most effective in oil viscosity.
Table 1. Statistical experimental data of sample oils
3. CORRELATIONS LITERATURE
Six correlations were used to predict the viscosity of saturated oil, including: Chew-
Connally [1], Beggs-Robinson[2], Labedi [3], Kartoatmodjo and Schmidt [4], modified
Kartoatmodjo [4], and Elsharkawy & Alikhan [5] correlations, the accuracy of each correlation
of predicted saturated oil viscosity was checked with experimental data table2 reveals
average relative error (ARE%), and absolute average relative error (AARE%) for saturated
oil viscosity.
Table 2 ARE% and AARE% for saturated oil viscosity
Model ARE(%) AARE(%)
The modified Kartoatmodjo 6.51867 23.9795
Elsharkawy & Alikhan 20.2973 24.0984
Chew & Connally 13.4312 25.2184
Beggs & Robinson 24.4467 27.0632
Kartoatmodjo & Schmidt 24.3451 32.7524
Labedi 19.8468 35.901
K. Rahuma, E. Edreder/Petroleum & Coal 55(2) 113-136,2013 135
The contradiction may be clarified by referring to figure 2, which compares the scatter
diagram relationship between the measured and the predicted viscosity.
4. Conclusion
Several empirical models for estimating the viscosity of crude oils (saturated and under-
saturated) have been evaluated using viscosity data of crude oils from the selected Libyan
oil reservoirs. Good agreements between the predicted and experimental values have
been observed. It can be concluded from results that, the modified Kartoatmodjo, and
Elsharkawy & Alikhan correlations are the best and accurate for saturated oil viscosities,
while Beal’s, and Khan & Ali correlations are preferred to get accurate results for under-
saturated oils. The less AARE% was calculated 23.9% for saturated oil, and 21.9 % for
undersaturated oil.
REFERENCES
[1] Chew, J., Connally, C.A., 1959. Viscosity correlation for gas saturated crude oil.
Trans. AIME 216, 23–25.
[2] Beggs, H.D., Robinson, J.F: J. Pet. Tech., Sept. 1975, p. 1140-1141.
[3] Labedi: J. Pet. Sci. Eng. 8 . 1992 ,p. 221 – 234.
[4] Kartoatmodjo, F., Schmidt, Z.: Oil & Gas J. 4, 1994, p. 51-55.
[5] Elsharkawy, A.M, Alikhan, A.A. : Fuel, 78, 1999, p.891-903.
[6] Beal, C.: Trans, AIME, 165, 1946, p.94-112.
[7] Khan, S. A., et al., “Viscosity Correlations for Saudi Arabian Crude Oils,” SPE Paper
15720, Presented at the Fifth SPE Middle East Conference held in Manama, Bahrain,
March 7-10, 1987.