paper ADE.pdf
-
Upload
sebastian-castro -
Category
Documents
-
view
221 -
download
0
Transcript of paper ADE.pdf
-
Petroleum Science and Technology, 25:949965, 2007
Copyright Taylor & Francis Group, LLC
ISSN: 1091-6466 print/1532-2459 online
DOI: 10.1080/10916460500526981
Empirical Equations for Estimating ADE
of Crude Oils
M. A. Fahim
Chemical Engineering Department, Kuwait University, Safat, Kuwait
Abstract: There has been an increasing interest in the construction of asphaltene
deposition envelope (ADE) to determine a safe zone of operation during oil pro-
duction. Equations of state are usually used for calculations of ADE; however, the
method requires tuning some experimental Pressure Volume Temperature (PVT) data
of the reservoir fluid. The objective of this study is to develop a simple, accurate,
and reliable empirical equation for estimating the ADE by determining the upper
and lower onset pressures as well as the saturation pressures. Three simple empiri-
cal equations are developed to calculate these onset pressures of several crude oils.
Experimentally measured compositions, saturation, and onset pressures of 33 crude
oil samples, primarily from the Middle East, at different temperatures were used to
develop the equations. Another set of compositions and upper saturation and lower
onset pressures of different crude oil samples from the literature were used to test the
accuracy of the empirical equations. The results were also compared with equations of
state predictions. The results indicate that the method is accurate, valid, reliable, and
eliminates the splitting and characterizing of the heavy fraction, which is necessary
for the equations of state predictions. The method is useful for estimating the ADE
of crude oils where experimental data is not available.
Keywords: asphaltene, deposition, emperical, envelope
INTRODUCTION
The formation of asphaltene is a flow assurance problem experienced by a
large number of operators. It can be caused be thermodynamic considerations
such as pressure and temperature changes. It can also be caused by compo-
sitional changes during production. For example, it can be caused during
CO2 or gas injection operations. Gas breakthrough can cause it as well. The
stability of oils under reservoir conditions can be studied by construction of
Address correspondence to M. A. Fahim, Kuwait University, Department of
Chemical Engineering, P.O. Box 5969, Safat, Kuwait. E-mail: [email protected]
949
-
950 M. A. Fahim
what is called the asphaltene deposition envelope (ADE). A key evaluation in
asphaltene formation and stabilization in the determination of the pressure/
temperature phase envelope of asphaltene flocculation onset (ADE) will be
studied in this work.
Several previous attempts have been made to model asphaltene deposi-
tion with varying degrees of success. The colloidal method of Leontaritis was
used to calculate asphaltene flocculation conditions for CO2 C oil mixtures
but only at one temperature. Both Hirschberg et al. (1984), Chung (1992),
and Burke et al. (1990) have treated asphaltene as a pure liquid phase in
equilibrium with a solvent using the Flory-Huggins polymer solution the-
ory. The modeling of the temperature and pressure dependence of asphaltene
deposition, using an equation of state for vapor-liquid equilibrium, was rea-
sonably successful. Chung (1992) also used a simple regular solution theory
approach for solid asphaltene liquid equilibrium and was able to fit solubility
data for asphaltene in various solvents at ambient conditions, but this method
seems unlikely in giving good results under typical reservoir conditions. His
regressed asphaltene molar volume is an order of magnitude smaller than
Hirschbergs and seems suspect as the specific gravity of asphaltene is un-
acceptably high. Finally, Kawanaka et al. (1991) used a similar approach
to Hirschberg but with a continuous mixture for the asphaltene phase, and
obtained good agreement with solvent titration experiments and correct qual-
itative behavior at reservoir pressures.
Hirschbergs model appeared to be a good starting point but was not eas-
ily incorporated into multi-flash as it treats asphaltene-liquid equilibrium sep-
arately from vapor-liquid equilibrium, which is physically unrealistic. There-
fore, the approach chosen was to treat asphaltene as a solid phase in equi-
librium with the reservoir fluid, which was used with an equation of state.
Multi-flash can handle either a pure solid phase or a solid mixture. Although
in reality an asphaltene deposit is a complex mixture, given the lack of ex-
perimental data on asphaltene compositions, it was decided to treat it as a
pure solid pseudo component phase in the examples investigated to date.
DATA COLLECTION
The input data required to build the method are:
a compositional analysis of the reservoir fluid C7C molecular weight and gravity the amount of asphaltene and resin in the oil reservoir temperature.
A total of 33 samples of different crudes were collected from literature
and used. The following parameters should be available: compositional anal-
yses, C7C density and molecular weight, reservoir temperature, asphaltene,
-
Empirical Equation for ADE 951
and resin weight percentage. These parameters were used to calculate upper
and lower onset pressure and saturation pressure and constructing ADE. The
compositional data collected and onset experimental pressures and satutation
are shown in Table 1 (Fahim and Andersen, 2005).
EQUATION DEVELOPMENT
The molecular weight and specific gravity of the heptane plus-fraction are
used as an input variable distinguishing the plus-fraction. The density and
molecular weight of the total plus-fraction is important in the proposed
method because it reflects the paraflinic-naphthenic-aromatic (PNA) content
of the plus-fraction. A large density indicates a high content of aromatic com-
pounds and a low density indicates a high content of paraffinic and naphthenic
content. Extending the plus-fraction into C10 might increase the accuracy of
the proposed method. However, it would reduce its simplicity. No distinction
is made between iso-paraffins and normal paraffins for two reasons: first,
concentrations of these hydrocarbons are usually very small when compared
with methane and heptane plus-fraction, and second, the simulation results
have indicated minor change in the estimated pressure when iso and normal
are lumped as a single component. A similar treatment has been used by
Elsharkawy (2003) in predicting the saturation pressure by EOS.
The objective function in this case is:
P D f.H2S;N2;CO2;C1;C2; : : : ;C7C; C7C;MWC7C; aspht%;Resin%;T/
(1)
where P is the calculated pressures (upper, saturation, and lower) of the
crude oil sample. The objective is to minimize the average relative error
(ARE) between measured (P) and estimated pressure for all the experimen-
tally measured samples in this study. The average relative error is defined
as:
ARE D
X
.P Pcalc/=Pcalc
=n (2)
Multiple regression and error minimization resulted in the following em-
pirical equations.
For the upper onset pressure the following equation is obtained:
Ponsu D 4:3243H2SC 2:6047
N2 C 0:6890CO2 C 0:6503
C1
C 0:9273C2 C 0:0086C3 3:981
C4 0:5878C5 C 0:1817
C6
0:0811CC7 0:1305MWCC7 C 210:89
CC7 C 0:908Asph
0:2003Resin 0:8988TC 0:001T2 C 12:4148LN.T/ (3)
with r2 D 0:992.
-
Table 1. Crudes that were used in generation of ADE methods
Sample
no.
Mole%
H2S
Mole%
N2
Mole%
CO2
Mole%
C1
Mole%
C2
Mole%
C3
Mole%
C4
Mole%
C5
Mole%
C6
Mole%
C7
C7C
MW
C7C
Asph.
Wt%
Resin
Wt%
T,
K
Upper
press.,
MPa
Sat.
press.,
MPa
Lower
press.,
MPa
1 0.000 0.000 0.830 40.690 11.630 7.150 3.540 2.860 2.410 30.870 0.930 217.0 1.240 4.6 386 65.0 27.4 14.0
2 0.000 0.570 2.460 36.370 3.470 4.050 1.930 1.570 1.620 47.960 0.959 329.0 15.800 8.3 373 35.6 20.1 12.0
3 0.000 0.970 0.200 27.550 7.430 9.020 6.140 4.160 3.160 41.390 0.856 217.9 1.200 7.3 365 20.1 12.9 9.7
4 0.050 0.090 1.020 42.410 10.780 6.920 4.460 3.290 2.860 28.110 0.853 209.0 0.550 7.3 322 65.5 17.6 8.2
5 0.370 0.090 1.220 23.990 10.140 8.390 5.330 4.390 4.690 42.190 0.874 245.5 0.900 7.0 386 27.4 12.6 6.9
6 0.000 0.800 0.050 51.020 8.090 6.020 3.970 3.210 2.670 24.170 0.875 368.9 4.600 7.6 361 36.4 29.4 26.4
7 0.013 0.070 0.910 43.080 10.690 6.860 4.400 3.230 2.810 27.940 0.846 205.2 0.260 7.4 384 42.7 21.8 12.7
8 0.000 0.480 0.920 43.430 11.020 6.550 4.490 3.530 2.700 26.880 0.865 228.1 1.300 7.0 389 42.9 23.1 13.5
9 0.000 0.480 0.920 43.430 11.020 6.550 4.490 3.530 2.700 26.880 0.865 228.1 1.300 7.0 313 52.0 22.2 12.3
10 0.000 0.480 0.920 43.430 11.020 6.550 4.490 3.530 2.700 26.880 0.865 228.1 1.300 7.0 354 44.0 22.6 13.0
11 0.000 0.480 0.920 43.430 11.020 6.550 4.490 3.530 2.700 26.880 0.865 228.1 1.300 7.0 414 42.0 22.7 13.5
12 0.000 0.510 1.420 6.040 7.000 6.860 4.180 4.160 3.160 66.680 0.902 281.0 7.800 7.5 377 20.8
13 3.220 0.490 11.370 27.360 9.410 6.700 3.980 3.200 1.980 32.290 0.877 248.8 1.400 10.4 361 37.2 17.0 10.7
14 3.220 0.490 11.370 27.360 9.410 6.700 3.980 3.200 1.980 32.290 0.877 248.8 1.400 10.4 383 27.9 18.4 18.8
15 3.220 0.490 11.370 27.360 9.410 6.700 3.980 3.200 1.980 32.290 0.877 248.8 1.400 10.4 400 25.2 19.7 16.1
16 3.220 0.490 11.370 27.360 9.410 6.700 3.980 3.200 1.980 32.290 0.877 248.8 1.400 10.4 422 26.2 20.8 20.3
17 0.000 0.480 0.920 43.430 11.020 6.550 3.490 3.530 2.700 26.880 0.865 228.1 1.300 18.8 372 46.6 21.9
18 0.000 0.480 0.920 43.430 11.020 6.550 3.490 3.530 2.700 26.880 0.865 228.1 1.300 18.8 377 44.8 22.3
19 0.000 0.480 0.920 43.430 11.020 6.550 3.490 3.530 2.700 26.880 0.865 228.1 1.300 18.8 383 43.7 22.3
20 0.000 0.480 0.920 43.430 11.020 6.550 3.490 3.530 2.700 26.880 0.865 228.1 1.300 18.8 389 42.2 22.4 13.3
21 0.000 0.800 0.050 51.020 8.090 6.020 3.970 3.210 2.670 24.170 0.875 368.9 4.600 13.5 361 36.1 29.0 26.1
22 0.000 0.210 2.560 25.200 7.080 5.560 6.200 4.920 3.940 43.330 0.832 208.7 0.940 4.6 328 26.2 10.8
23 0.000 0.210 2.560 25.200 7.080 5.560 6.200 4.920 3.940 43.330 0.832 208.7 0.940 4.6 394 8.4
24 0.000 0.310 3.020 44.290 6.360 4.880 4.620 3.710 2.890 29.920 0.833 213.1 0.490 6.4 328 33.0 23.4
25 0.000 0.310 3.020 44.290 6.360 4.880 4.620 3.710 2.890 29.920 0.833 213.1 0.490 6.4 394 20.3
26 1.510 0.690 2.000 29.770 14.400 9.460 5.300 3.950 4.020 29.450 0.860 230.3 1.500 2.1 383 26.2 17.2 13.0
27 0.370 0.090 1.220 23.990 10.140 8.390 5.330 4.390 4.690 41.390 0.874 245.5 3.300 2.3 386 27.2 11.8
28 0.050 0.090 1.020 42.410 10.780 6.920 4.470 3.290 2.860 28.110 0.852 209.5 0.500 11.3 322 59.9 17.6 9.5
39 0.050 0.090 1.020 42.410 10.780 6.920 4.470 3.290 2.860 28.110 0.852 209.5 0.500 11.3 339 53.5 19.7 8.8
30 0.050 0.090 1.020 42.410 10.780 6.920 4.470 3.290 2.860 28.110 0.852 209.5 0.500 11.3 355 49.6 20.4 9.5
31 0.050 0.090 1.020 42.410 10.780 6.920 4.470 3.290 2.860 28.110 0.852 209.5 0.500 11.3 372 46.8 21.3 10.6
32 0.050 0.090 1.020 42.410 10.780 6.920 4.470 3.290 2.860 28.110 0.852 209.5 0.500 11.3 389 49.1 22.4 10.5
33 0.050 0.090 1.020 42.410 10.780 6.920 4.470 3.290 2.860 28.110 0.852 209.5 0.500 11.3 425 40.5 23.4 11.9
Experimental values not available.
952
-
Empirical Equation for ADE 953
For the saturation pressure the following equation is obtained:
Psat D 1:1107H2SC 1:5585N2 C 0:5764
CO2 C 0:469C1 C 0:0751
C2
C 0:7932C3 0:314C4 0:303
C5 C 0:763C6 0:0217
CC7
0:0284MWCC7 20:41CC7 C 0:011
Asph
0:025ResinC 0:0296T (4)
with r2 D 0:996:
For the lower onset pressure the following equation is obtained:
PonsL D 7:115H2S 17:603
N2 6:706CO2 4:842
C1 C 8:098C2
11:931C3 0:600C4 2:477
C5 37:166C6 C 0:201
CC7
C 0:569MWCC7 571:4:3CC7 2:422
Asph 0:052Resin
0:333TC 139:32LN.T/ (5)
with r2 D 0:992:
In all equations, P is the pressure in MPa, T is the reservoir temperature
in K, composition analysis of (N2;CO2;H2S;C1, :C6;CC
7 ) is in the mole
percent, MWC7C is the molecular weight of the heptane plus, andCC7 is
the specific gravity of the heptane plus fraction and Asph% and Resin% is the
weight percent of asphaltene and resin in the crude. It is important that the
summation of mole percentages of all components forming the hydrocarbon
must be equal to 100%. The degree of fitness of these equations are shown
in Figure 1. The range of variables used to develop Eqs. (3)(5) are given in
Table 2.
Estimation of Resin Content
Resin is a very important factor in asphaltene precipitation since it precipitates
simultaneously with asphaltene and its quantity should be known, but in
many cases resin content is not available for crude oil. A correlation has
been developed to estimate resin content of any crude by knowing asphaltene
Table 2. Summary of error analysis for the proposed empirical method
Upper pressure Saturation pressure Lower pressure
ARE 0.20 0.20 0.98
AAE 3.3 2.94 5.83
R2 0.992 0.996 0.991
-
954 M. A. Fahim
Figure 1. Comparison of experimental pressure and calculated pressure from empir-
ical equations: (a) upper pressure, (b) saturation pressure, (c) lower pressure.
content, sulphure content, and American Petroleum Institute (API) gravity of
crude oil. A total of 37 different crudes (with known resin contents) were
used (Speight, 1993) to generate the proposed correlation. The objective is
to minimize the ARE between measured resin wt% and estimated resin wt%
for all the experimentally measured samples in this study. Multiple regression
and error minimization resulted in the following empirical equations:
Resin wt% D 0:326375asph wt% 0:12388Oil wt%
C 6:089721S wt%C 0:493571API (6)
with r2 D 0:92:
It can be observed that properties used in this equation are measured at
ambient conditions, and we first assume that the resin and asphaltene % in
Eq. (6) will have the same values at high pressures. Resin and asphaltene that
precipitated on depressurizing redissolve at pressures lower than the saturation
pressures and that assumption might be justified.
-
Empirical Equation for ADE 955
COMPARISON WITH EQUATION OF STATE MODEL
The proposed model is compared with an equation of state model (EOS) and
with experimental data to show the accuracy of the proposed model.
The Soave-Redlich Kwong equation of state (SRK-EOS) is under consid-
eration in this work because it is commonly used by the petroleum industry
for predicting phase behavior and volumetric properties of hydrocarbon reser-
voir fluid mixtures. The equation is given by:
P DRT
V b
a
V.V C b/(7)
where P is the pressure, V is the molar volume, T is the absolute temperature,
and R is the universal gas constant.
This equation can predict the upper, saturation, and lower pressures and
vapor/liquid equilibrium calculations. For saturation pressure, it has reason-
able accuracy if the heptane plus-fraction of the hydrocarbon fluids is properly
characterized (Jhaveri and Youngren, 1988), but it has significant deviation
in predicting the upper and lower pressures.
The ADE is determined by performing flashes at varying pressures for
a number of temperatures. Flash calculations using a small pressure step
are used to locate the precipitation onset pressures. These calculated onset
pressures taken together define the asphaltene precipitation envelope.
TUNING OF EOS
It is important to know that when EOS is used to predict ADE (upper, sat-
uration, and lower pressures), it shows good results for saturation pressure
while the upper and lower pressure have a large deviation from experimental
values (maximum deviation for upper pressure was 165% and for lower pres-
sure 126%, while saturation pressure was 20%). The ranges of variables used
in this work are given in Table 3. To tune EOS, experimental onset pressures
at given temperatures were used. Then the critical properties (mainly critical
temperature and pressure) of the crude oils were changed. This will help in
adjusting the calculated values of upper and lower pressure by EOS. Table 4
and Figures 2 and 3 show the importance of tuning the EOS in the calculation
of upper and lower pressure.
Table 3. Ranges of variables used in the developing proposed empirical method
%,
H2S
%,
N2
%,
CO2
%,
C1
%,
C2
%,
C3
%,
C4
%,
C5
%,
C6
%,
C7
,
C7C
MW,
C7C
W%,
Asph
W%,
resin
Res,
T(K)
Min 0 0 0.05 6.04 3.47 4.05 1.93 1.57 1.62 24.17 0.832 205.2 0.26 2.08 322
Max 3.22 0.97 11.37 51.02 14.4 9.46 6.2 4.92 4.69 73.56 0.959 368.9 15.8 18.8 433
-
Table 4. Comparison of untuned and tuned EOS in predicting ADE with empirical method versus corresponding experimental values
Upper ADE Saturation Lower ADE
No. Exp
Empirical
method
EOS
not tuned
EOS
tuned data
EOS
tuned model Exp
Empirical
method
EOS
not tuned
EOS
tuned data
EOS
tuned model Exp
Empirical
method
EOS
not tuned
EOS
tuned data
EOS
tuned model
1 65.0 64.8 * 65.1 64.9 27.4 26.9 27.3 27.1 26.6 4.0 4.0 * 3.7 3.7
2 35.7 35.3 73.9 35.6 35.2 27.3 27.4 24.4 26.1 26.2 23.1 23.2 9.3 21.7 21.8
3 20.1 20.2 * 20.7 20.8 12.9 12.9 12.6 12.3 12.3 9.7 9.7 * 8.9 8.9
4 47.6 45.6 30.0 46.2 44.2 17.6 25.3 24.9 24.5 32.2 8.3 9.8 12.7 11.1 12.6
5 27.4 27.5 30.0 27.5 27.6 12.6 11.4 12.1 12.1 10.9 6.9 6.8 5.3 6.1 6
6 36.4 36.8 43.0 36.6 37 29.4 31.1 32.0 32.6 34.3 26.4 26.3 9.6 30.3 30.2
7 42.7 46.8 24.5 42.8 46.9 21.8 22.8 22.6 21.4 22.4 12.7 11.7 19.5 12.3 11.3
8 42.9 44.3 41.6 43.0 44.4 23.1 22.5 24.2 24.5 23.9 13.5 13.2 16.0 15.8 15.5
9 47.3 46.1 42.5 47.3 46.1 22.2 22.0 23.2 23.5 23.3 12.3 12.8 15.2 13.5 14
10 45.4 45.5 42.2 45.5 45.6 22.6 22.2 23.7 23.9 23.5 13.0 13.0 15.3 14.2 14.2
11 44.3 44.8 42.0 44.6 45.1 22.7 22.3 24.0 24.2 23.8 13.5 13.1 15.7 14.9 14.5
12 20.8 21.3 17.8 21.1 21.6 9.5 9.8 9.7 5.3 8.2 8
13 37.2 32.1 30.3 34.9 29.8 17.0 18.2 17.1 17.1 18.3 10.7 15.8 11.7 10.1 15.2
14 27.9 29.2 30.4 29.5 30.8 18.4 18.9 18.7 18.8 19.3 18.8 16.8 14.3 16.0 14
15 25.2 27.9 30.7 26.9 29.6 19.7 19.4 20.0 19.9 19.6 16.1 16.5 13.2 13.5 13.9
16 26.2 27.0 31.3 24.9 25.7 20.8 20.2 21.1 21.2 20.6 20.3 16.8 15.3 18.8 15.3
17 46.6 45.3 42.3 46.6 45.3 21.9 22.1 23.5 23.6 23.8 12.9 15.8 13.1
18 44.8 44.7 42.0 45.3 45.2 22.3 22.3 23.8 23.9 23.9 13.1 15.9 13.4
19 43.7 44.0 41.6 43.9 44.2 22.3 22.5 24.1 24.2 24.4 13.2 16.1 13.5
20 42.2 43.5 41.7 42.4 43.7 22.4 22.7 24.5 24.5 24.8 13.3 13.4 16.2 15.8 15.9
21 36.1 35.4 95.8 36.2 35.5 29.0 27.2 31.2 32.6 30.8 26.1 26.3 9.7 30.6 30.8
22 26.2 25.1 21.2 26.6 25.5 10.8 10.9 12.1 11.4 11.5 7.7 8.9 8
23 12.3 10.4 12.4 8.4 8.7 8.8 8.7 9 4.1 6.1 4.3
24 35.5 37.2 32.1 35.7 37.4 23.4 22.6 22.0 22.0 21.2 16.2 18.9 16
25 27.2 23.6 27 20.3 20.4 18.4 18.0 18.1 12.5 14.6 12.8
26 26.2 26.2 26.0 26.0 26 17.2 17.1 17.0 17.0 16.9 13.0 13.0 12.2 12.2 12.7
27 27.2 28.8 31.7 31.7 27.7 11.8 12.3 12.0 12.0 12.1 6.8 7.8 6.6
28 59.9 60.6 27.9 58.0 58.8 17.6 19.6 17.5 17.5 19.5 9.5 7.2 6.5 6.9 6.9
39 53.5 54.3 26.2 53.7 53.9 19.7 20.2 18.8 18.8 19.5 8.8 9.1 15.9 8.5 8.9
30 49.6 50.2 26.0 49.7 49.8 20.4 20.7 19.9 20.0 20.4 9.5 10.3 17.2 9.8 11.0
31 46.8 47.5 26.2 46.8 47.2 21.3 21.2 21.2 21.1 21.4 10.6 11.1 18.6 11.0 10.9
32 49.1 45.7 25.1 49.1 46.1 22.4 21.8 22.0 21.9 21.6 10.5 11.5 20.5 10.7 11.1
33 40.5 44.0 27.8 43.5 40.5 23.4 23.0 23.4 23.2 23.5 11.9 11.6 20.3 14.2 11.9
Experimental values not available.No convergence.
956
-
Empirical Equation for ADE 957
Figure 2. ADE comparison of proposed method with experimental, tuned, and un-
tuned EOS (Crude 28).
Figure 3. ADE comparison of proposed method with experimental, tuned, and un-
tuned EOS (Crude 9).
-
958 M. A. Fahim
The heaviest component can be divided into two componentsone is
nonprecipitating and the other is precipitating based on the weight of asphal-
tene in the crude. The mole fraction of asphaltene can be calculated according
to the following:
Zasph D Wt%asph Moil=MWasph: (8)
The asphaltene properties TC and PC are then tuned to satisfy the precipi-
tation condition. Values of P and f a can be found from extrapolation to 0
mole % precipitation.
RESULTS AND DISCUSSION
The accuracy of the empirical method is compared with the experimental
values reported from literature and with EOS (SRK). The statistical results
of various crude oils for the empirical method are shown in Table 2. This
table reports the average relative error (ARE) and the average absolute error
(AAE) for the 33 crude oil samples that were used to derive the equations. The
proposed method has the smallest errors for the measured saturation pressures
of the Middle East crudes that were used to develop the method. The current
method has the advantage of being simple and accurate, eliminating the need
for splitting the heptane plus-fraction.
It is important to know that as the number of points used to tune the
EOS increase, the accuracy of estimating ADE pressure increases. A sample
crude was taken and a different number of tuning points were used and shows
that with three points of tuning it gave ARE D 2.36 and the AAE D 6.52,
while using four points of tuning gave ARE D 2.75 and AAR D 4.69, and
using five points of tuning gave ARE D 0.031 and the AAE D 3.24. The
error analysis for EOS prediction is given in Tables 5 and 6.
A total of five samples of different crudes at different temperatures were
used for testing the proposed empirical equations (Table 7). Also three differ-
ent crude samples were used to test the effect of gas injection and were com-
pared with experimental values as shown in Table 8. The error analysis results
for crudes tested in Table 7 are shown in Table 9. Also an ADE for two differ-
ent crudes was generated by the proposed equations and compared with exper-
imental and tuned and untuned EOS as shown in Figures 2 and 3. As shown
Table 5. Summary of error analysis for ADE pressures calculated by
untuned EOS compared with experimental
Upper pressure
untuned EOS
Saturation pressure
untuned EOS
Lower pressure
untuned EOS
ARE 9.40 1.66 17.44
AAE 30.40 5.00 45.88
-
Empirical Equation for ADE 959
Table 6. Summary of error analysis for ADE pressures calculated by
tuned EOS compared with experimental
Upper pressure
tuned EOS
Saturation pressure
tuned EOS
Lower pressure
tuned EOS
ARE 0.60 1.44 0.49
AAE 1.70 4.66 11.90
in the figures, the proposed and tuned EOS are very close to experimental
data while the untuned EOS has a large deviation from experimental points.
Effect of Crude Properties and Composition on ADE
With many parameters such as gas injection, the changing of asphaltene
content has been studied and the effects of each one on ADE have been
studied. The crude oil stability increases as the asphaltene content increases.
Experience has shown that heavy precipitation of asphaltene occurs with
crude of low asphaltene content as in Hasi Masoud wells. The effect of initial
asphaltene content on ADE is predicted and compared with experimental data
as shown in Figure 4.
Figure 4. Effect of asphaltene mole% on ADE (Crude 28).
-
Table 7. Crudes used for testing the empirical method
No.
%,
H2S
%,
N2
%,
CO2
%,
C1
%,
C2
%,
C3
%,
C4
%,
C5
%,
C6
%,
C7
,
C7C
MW,
C7C
W%,
Asph
W%,
resin
Res,
T(K)
34 1.900 0.090 2.110 29.500 13.710 9.310 5.320 4.030 4.150 29.880 0.832 209.8 0.500 1.050 397.1
35 0.099 0.317 0.830 27.800 9.560 8.000 5.170 4.070 4.410 39.740 0.846 223.2 0.900 1.530 391.6
36 0.014 0.065 0.920 44.221 10.902 7.217 4.427 3.072 2.554 26.608 0.853 215.9 0.500 0.860 390.4
37 0.013 0.173 0.841 40.054 10.346 7.046 4.608 5.237 3.025 30.413 0.846 204.8 0.500 6.000 392.1
38 0.013 0.070 0.914 43.079 10.689 6.861 4.396 3.233 2.814 27.931 0.845 204.1 0.400 7.900 388.8
39 0.050 0.090 1.020 42.410 10.780 6.920 4.470 3.290 2.860 28.110 0.852 209.5 0.500 11.300 432.7
Table 8. Crudes used for testing the empirical method for gas injection
No.
%,
H2S
%,
N2
%,
CO2
%,
C1
%,
C2
%,
C3
%,
C4
%,
C5
%,
C6
%,
C7
,
C7C
MW,
C7C
W%,
Asph
Res,
T(K)
16 3.220 0.490 11.370 27.360 9.410 6.700 3.980 3.200 1.980 32.290 0.877 248.8 1.400 400
40 0.01 0.32 2.29 17.6 5.25 6.14 6.63 5.84 5.00 50.92 0.852 210 1.00 389
41 0.00 0.16 0.14 77.44 3.95 2.56 1.78 1.32 0.94 11.71 0.86 233.6 8.3 342
960
-
Empirical Equation for ADE 961
Table 9. Error analysis for crude used to test the empirical method
Pupper, Bar Psat, Bar Plower, Bar
No. Exp Calc Exp Calc Exp Calc
34 25.9 26.0 16.2 16.3 10.8 10.8
35 23.1 21.3 13.5 14.2 9.0 9.1
36 55.1 55.1
37 51.7 51.3 20.8 20.9 8.0 8.0
38 52.4 46.8 23.8 22.6 12.2 12.5
39 44.1 44.0 23.6 23.0 12.0 11.5
ARE 3.18 ARE 0.42 ARE 0.43
AAE 3.40 AAE 2.64 AAE 2.43
Table 10. Properties of injected gases
No.
%,
H2S
%,
N2
%,
CO2
%,
C1
%,
C2
%,
C3
%,
iC4
%,
nC4
%,
iC5
%,
nC5
%,
C6
%,
C7 MW
1 0.03 0.59 5.03 65.76 11.32 8.58 2.16 3.62 1.24 1.14 0.51 0.03 25.6
2 0.00 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 28.0
3 0.00 0.00 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 44.0
4 0.00 0.00 0.00 0.00 0.00 1.00 0.00 0.00 0.00 0.00 0.00 0.00 44.0
The upper onset pressure equation was used to investigate the effect of
injection of different gases such as scrubber gas (CO2) and H2S;N2;C3, and
natural gas. A list of gases used is given in Table 10. The results of the
injection of nitrogen, gas 1, and C3 on the upper part of ADE and saturation
pressure are shown in Figures 5, 6, and 7. In these figures the prediction by
the corresponding equation and EOS is also shown.
It is important to note that EOS can be tuned using the onset point
generated in this work, and consequently, it can be used to determine the
percentage of asphaltene precipitated. Figure 8 shows the effect of CO2 in-
jection on asphaltene precipitation obtained by EOS tuned by a proposed
model compared to experiential values.
CONCLUSIONS
The empirical methods that have been proposed to estimate upper, saturation,
and lower pressure of ADE for any crude oil have several advantages over the
equation of state. It is simple and gives an approximate estimation for ADE and
eliminates the need for splitting and characterizing the heptane plus-fraction.
-
962 M. A. Fahim
Figure 5. Effect of N2 injection on upper onset pressure and saturation pressures
calculated from a proposed model versus experimental (Crude 16).
Figure 6. Effect of gas (gas 1) injection on upper onset pressure and saturation
pressures calculated from a proposed model versus experimental (Crude 40).
-
Empirical Equation for ADE 963
Figure 7. Effect of C3 injection on saturation pressures calculated from a proposed
model versus experimental.
Figure 8. Effect of CO2 injection on asphaltene precipitation. Comparison of exper-
imental with EOS tuned with model (Crude 2).
-
964 M. A. Fahim
The proposed methods were tested with different crude oils and were also
tested against EOS and the result was satisfactory. The EOS has a large
deviation compared to the proposed method and experimental values and
tuning of the EOS is very important and will improve it. Injection of different
gasses also showed that the model gave a satisfactory result compared to the
EOS that requires splitting of C7C into different subtractions.
ACKNOWLEDGMENT
M. A. Fahim would like to thank Kuwait University for granting him a sab-
batical year (2002/2003), where he spent a portion of it at Denmark Technical
University.
REFERENCES
Burke, N. E., Hobbs, R. D., and Kashon, S. F. (1990). J. Petrol. Technol.
11:14401446.
Chung, T. (1992). SPE J. 24851.
Elsharkawy, A. M. (2003). J. Petrol. Sci. Engr. 1052:121.
Fahim, M. A., and Andersen, S. I. (2005). SPE J. 93517.
Hirschberg A., DeJong, L., and Meijer, J. (1984). Soc. Petrol. Eng. 24:283.
Huggins, M. (1941). Chem. Phys. 9:440.
Jamaluddin, A. K. M., Joshi, N., Iwere, F., and Gurpinar, O. (2002). SPE J.
74393.
Jamaluddin, A. K. M., Creek, J., McFadden, C. S., DCruz, D. Thomas, J.,
Joshi, N., and Pross, B. (2001). SPE J. 72154.
Jamaluddin, A. K. M., et al. (2003). SPE J. 80261 Sept:304.
Jamaluddin, A. K. M., et al. (2001). SPE J. 71546.
Jamaluddin, A. K. M., et al. (2000). SPE J. 87292.
Jhaveri, B. S., and Youngren, G. K. (1988). SPE (Reservoir Engineers),
3:10331040.
Kabir, C. S., and Jamaluddin, A. K. M. (1990). SPE J. 53155.
Kawanaka, S., Park, S. J., and Mansoori, G. A. (1991). SPE Res. Eng. 6:185.
Leontrites K. J., and Mansoori, G. A. (1987). International Symp. On Oil
Fuel and Chemistry. San Antonio, TX, February 46. SPE J. 16258.
Pederson, K. S., Fredenslund, A., and Thomassen, P. (1989). Properties of
Oils and Natural Gases. Houston TX: Gulf Publishing.
Speight, J. G. (1991). The Chemistry and Technology of Petroleum, 2nd Ed.
New York: Marcel Dekker, pp. 13431344.
Speight, J. G. (1993). Fuel 72:13431344.
Vafaie-Sefti, M., et al. (2003). Fluid Phase Equil. 260:111.
-
Empirical Equation for ADE 965
NOMENCLATURE
Dimensionless factor
Dimensionless factor
L Fugacity coefficient liquid phase
V Fugacity coefficient in vapor phase
Density
! Acentric factor
a First parameter in EOS
b Second parameter in EOS
f L Fugacity in liquid phase
f V Fugacity in vapor phase
K Equilibrium coefficient
MW Molecular weight
n Number of components
P Pressure
Pc Critical pressure
R Gas constant
r2 Correlation coefficient
T Temperature
Tc Critical temperature
V Molar volume
w Weight fraction
x Mole fraction in liquid phase
y Mole fraction in gas phase
z Mole fraction in feed