Determination of Minimum Miscibility Pressure
Using PVTi Software, Eclipse 300, and Empirical
Correlations
Vahid Karamnia1 and Siavash Ashoori2
1
2
M.S. Student, Department of Petroleum Engineering, Omidiyeh Branch, Isla
Associate Professor, Department of Petroleum Engineering, Ahwaz Faculty of Petroleum
Received: October 15, 2020; Revised: January 28, 2021; Accepted: February
03, 2021
Abstract
One of the most important factors in the miscible gas injection process is to determine
the minimum miscibility pressure (MMP). The MMP is the minimum pressure at which, at
a constant temperature, the oil and gas injected can dissolve into each other to form a single
phase. Laboratory methods like the slim tube test and the ascending bubble apparatus
test are widely utilized but are time-consuming and expensive. This study determines the
MMP of reservoirs using PVTi software, Eclipse 300 software, and empirical correlations,
comparing results with laboratory data. Eclipse 300 proved to be the fastest and most
accurate approach.
Keywords: Empirical Correlations, First-contact Miscibility (FCM), Gas Injection,
Multi-Contact Miscibility (MCM), Simulation
1
Introduction
Due to the decline in reservoir pressure from primary production, secondary and tertiary production methods, such as miscible gas injection, are critical for enhanced oil recovery. The
displacement efficiency in gas injection depends heavily on pressure, specifically the minimum
miscibility pressure (MMP), where oil recovery increases significantly. Laboratory tests like
the slim tube test are accurate but costly, prompting the use of computational methods. This
study evaluates PVTi, Eclipse 300, and empirical correlations to determine MMP, comparing
results with experimental data to identify the most accurate method.
2
Theoretical Section
2.1 Experimental Data
Laboratory data from ? ] were used, focusing on four reservoir fluids (F2, F3, F4, F5). The
slim tube test was simulated using commercial software.
1
2.2 Test Fluids
Table 1 summarizes the molar compositions of the reservoir fluids.
Table 1: Experimental components of the oil reservoirs.
Components
F2 (%)
F3 (%)
F4 (%)
F5 (%)
H2 S
N2
CO2
C1
C2
C3
iC4
nC4
iC5
nC5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
C20+
MWC20+ (g/mol)
SGC7+
0
0.2
1.34
23.64
8.56
6.68
1.25
4.05
1.78
2.67
4.03
4.57
4.28
3.88
2.93
3.15
3.19
3.05
1.16
1.98
1.72
1.6
1.16
1.1
12.03
530
0.9493
0
0.45
1.64
45.85
7.15
6.74
0.84
3.11
1.03
1.65
2.52
3.77
4.28
2.7
1.69
1.81
1.47
1.45
1.28
1.15
0.91
0.82
0.8
0.71
6.18
474
0.9253
0
0.35
3.14
54.26
8.57
5.72
0.76
2.45
0.75
1.2
1.53
2.6
3.02
1.2
1.74
1.36
1.1
1.11
0.95
0.86
0.68
0.6
0.56
0.51
4.08
418
0.905
0.383
0.45
2.07
26.576
7.894
6.73
1.485
3.899
1.937
2.505
3.351
4.311
4.133
3.051
2.033
2.635
2.285
2.364
2.038
1.752
1.589
1.492
1.263
0.812
12.962
450
0.956
2.3 Specifications of Slim Tube Apparatus
The slim tube had a length of 10 m, an inner diameter of 0.635 cm, a porosity of 38%, and a
permeability of 2000 mD. Gas was injected at 10 cc/h.
2.4 Experimental Procedures
The slim tube was saturated with reservoir oil, and gas was injected at various pressures. Oil
recovery and gas-to-oil ratio were measured after injecting 1.2 pore volumes. The MMP was
identified as the pressure at the break point in the recovery vs. pressure curve.
2
2.5 Modeling with Commercial Software
PVTi software was used to model fluid properties using the Peng-Robinson equation of state.
Regression adjusted parameters (ΩA , ΩB , Pc , Tc ) to match laboratory data.
3
Results and Discussion
3.1 Regression Results
Figures 1–7 (not reproduced here) showed adjustments between calculated and experimental
data for relative volume, oil density, gas-to-oil ratio, and compressibility.
3.2 Determining MMP with Software
PVTi calculated first-contact miscibility (FCM) and multi-contact miscibility (MCM) pressures
(Table 2).
Table 2: MMP results by PVTi software.
Reservoir FCM (bar)
F2
F3
F4
F5
551.1
665.2
594.5
460.0
MCM (bar)
340.4
430.9
380.1
280.8
3.3 Simulation with Eclipse 300
A 1D slim tube model was simulated at various pressures with 100, 200, and 500 blocks.
Recovery factors were extrapolated to infinite blocks to correct dispersion, and MMP was determined from the recovery vs. pressure plot (Tables 3–6).
Table 3: Recovery factors for F2 (gas G2).
P (bar) 100 grids 200 grids 500 grids
80
110
140
170
200
230
260
290
320
58.87
65.45
72.25
79.95
86.14
91.35
94.73
96.73
98.47
59.25
66.25
73.68
80.89
87.74
92.75
96.65
98.41
99.21
3
59.78
66.98
74.29
81.67
88.42
93.87
98.23
99.21
99.63
Infinite grids
60.15
67.16
75.61
82.18
89.71
95.82
99.92
100.68
100.25
Table 4: Recovery factors for F3 (gas G3).
P (bar) 100 grids 200 grids 500 grids
240
270
300
330
360
390
420
450
480
69.26
75.56
80.23
87.85
90.05
93.23
96.81
98.27
98.62
69.87
76.06
80.97
88.87
91.09
94.47
97.95
98.89
99.05
70.11
76.49
81.68
89.90
92.23
95.61
99.14
99.58
99.49
Infinite grids
70.73
77.20
82.99
91.43
93.75
97.56
99.95
100.16
100.12
Table 5: Recovery factors for F4 (gas G4).
P (bar) 100 grids 200 grids 500 grids
240
270
300
330
360
390
420
60.37
68.58
75.65
83.16
88.52
94.36
97.73
60.75
69.05
76.15
83.96
89.15
95.63
98.69
61.15
69.45
76.57
84.53
89.75
96.54
99.41
Infinite grids
61.70
69.85
77.35
85.35
90.85
98.25
100.02
3.4 Empirical Correlations
Empirical correlations for light gases and nitrogen (e.g., ? ], ? ]) and for pure/impure CO2 were
evaluated. Results are shown in Table 7.
3.5 Error Calculation
The mean absolute error was calculated using:
P
1 X MMPM
i − MMPi
× 100
n i=1
MMPM
i
n
Mean absolute error =
Results are shown in Table 8.
4
Conclusions
Eclipse 300 was the most accurate and cost-effective method for MMP determination, with
errors below 3%. Among empirical correlations, ? ] was accurate for high methane content,
and ? ] was reliable for CO2 injection.
4
Table 6: Recovery factors for F5 (gas G5).
P (bar) 100 grids 200 grids 500 grids
250
280
320
350
380
410
440
82.49
85.41
88.45
91.16
93.95
96.16
97.15
83.67
86.65
89.75
92.85
95.45
97.15
97.69
Infinite grids
84.47
88.15
91.34
94.63
97.41
98.48
98.81
86.19
90.31
93.85
97.41
99.97
100.45
100.24
Table 7: MMP results using different methods.
Reservoir Experimental (bar)
F2
F3
F4
F5
235
376
379
366
Firoozabadi (bar)
Hudgins (bar)
Eclipse 300 (bar)
PVTi (bar)
370.88
379.31
354.1
367.98
309.57
312.23
303.94
308.61
240
385
386
369
340.35
430.94
360.19
280.78
Table 8: Mean absolute error of different methods.
Reservoir Firoozabadi (%) Hudgins (%) Eclipse 300 (%) PVTi (%)
F2
F3
F4
F5
57.82
0.88
0.07
0.54
31.73
16.96
19.8
15.68
5
2.13
2.39
2.64
0.82
44.82
14.61
4.96
23.28