Professional Documents
Culture Documents
a r t i c l e i n f o a b s t r a c t
Article history: Thermodynamic modeling of reservoir fluids behavior is of great importance for the oil and gas industry.
Received 5 August 2017 For design and safe operation purposes, it is necessary to have an accurate, simple and robust model for
Received in revised form the prediction of the phase equilibrium of reservoir fluids. In this work, the UMR-PRU model is applied in
4 October 2017
the prediction of dew points and liquid dropouts of gas condensate mixtures, and is compared with the
Accepted 24 October 2017
widely used cubic equations of state Soave-Redlich-Kwong (SRK) and Peng-Robinson (PR) as well as the
non-cubic Perturbed Chain Statistical Associating Fluid Theory (PC-SAFT) equation of state. The evalu-
ation of the models was based both on synthetic and real gas condensate mixtures. For real gas con-
Keywords:
Gas condensates
densates, a characterization method for the C7þ fraction based on the one proposed by Pedersen is
Dew point developed for UMR-PRU. The results reveal that UMR-PRU is a sufficiently accurate model for dew point
Liquid dropout and liquid dropout predictions of gas condensates.
Cubic equation of state © 2017 Elsevier B.V. All rights reserved.
UMR-PRU
https://doi.org/10.1016/j.fluid.2017.10.024
0378-3812/© 2017 Elsevier B.V. All rights reserved.
N. Novak et al. / Fluid Phase Equilibria 457 (2018) 62e73 63
[18e21]. compared with PR, SRK and PC-SAFT. Furthermore, the perfor-
Except for the classical two-parameter cubic EoSs, three- mance of UMR-PRU model in the prediction of dew points and
parameter cubic EoSs [22e26] have also been used for petroleum liquid dropouts of real gas condensates combined with a properly
fluid modeling, due to their improved density predictions [27,28]. adapted Pedersen characterization method is examined and
Three-parameter EoSs, like Patel-Teja and Patel-Teja-Valderrama, compared with PR and SRK EoSs.
have shown to be more accurate than PR and SRK for density and
compressibility factor predictions [28,29]. Dew point predictions 2. Thermodynamic models
however are not necessarily improved with the addition of a third
parameter [28,29]. All models considered in this study have been thoroughly pre-
Another class of equations of state, of non-cubic form, based on sented in the literature. SRK [2] and PR [62] have been applied in
the statistical thermodynamic principles of SAFT (Statistical Asso- this work coupled with the vdW1f mixing and combining rules
ciating Fluid Theory), has gained popularity in the industry after the using temperature independent binary interaction parameters
integration of the derived EoS in commercial process simulators. (BIPs) retrieved either from ASPEN-HYSYS process simulator
For example, the Perturbed-Chain SAFT EoS (PC-SAFT) is widely databank or determined in this work. The UMR-PRU model [61,63]
used for predicting asphaltene precipitation [30e39]. Recently, Yan has been applied using the UNIFAC interaction parameters pre-
et al. [39] compared dew point predictions of PC-SAFT with those sented by Louli et al. [61]. Pure component critical temperatures
obtained by PR and SRK and found that while all models are similar and pressures as well as acentric factors, needed for SRK, PR and
for natural gas dew points, for gas condensates PC-SAFT yields UMR-PRU, were taken from the DIPPR data compilation [64].
better results. Surprisingly, for mixture liquid densities, PC-SAFT For PC-SAFT [65,66], the three pure component parameters for
was found to yield on average similar results with the untrans- non-associating compounds: the segment number (m), the
lated PR, although it is clearly superior for pure components [40]. segment diameter (s), and the segment energy parameter (ε=k),
Gas condensate mixtures consist of numerous different com- were taken from Gross and Sadowski [66] and Ting et al. [37],
ponents, which makes a complete component analysis practically except for n-tetracosane and 2,2,4,4,6,8,8- heptamethyl-nonane,
impossible. The most common case of compositional analysis of which are not available in the literature. The parameters for these
reservoir fluids is to perform detailed analysis up to n-hexane, components were determined in this study by fitting pure
while the heavy end of the fluid is described by a single plus frac- component vapor pressure and saturated liquid density data and
tion (C7þ). EoS predictions for real gas condensates are greatly are tabulated in Table 1. PC-SAFT has been extended to mixtures by
influenced by the characterization method of the C7þ fraction, so a introducing mixture segment diameter and segment energy pa-
good coupling of an EoS with a characterization method is very rameters derived from the van der Waals one fluid theory [66].
important for accurate results [28,41e43]. The expected un- Similarly to cubic EoSs, PC-SAFT has been applied using tempera-
certainties of EoSs in gas condensate PVT and saturation point ture independent binary interaction parameters in the dispersion
predictions vary from one author to another, depending on the term of the EoS taken either from the literature or determined in
characterization method and the physical properties used for the this work.
pseudocomponents. The general conclusion that most researchers
agree on is that traditional cubic EoSs do not yield accurate phase
3. Results and discussion
equilibrium predictions, unless the fluid is somehow matched to
experimental data [29]. These data, such as PVT experiments,
3.1. Dew point predictions of synthetic gas condensate (SGC)
saturation points etc., can be used to manipulate the fluid proper-
mixtures
ties, like the MW of the plus fraction [44], or the binary interaction
parameters [45] in order to match the model predictions with the
The database of synthetic gas condensate (SGC) mixtures that
experimental data. This has also been done indirectly by calculating
was developed for the evaluation of the models in dew point pre-
EoS dependent pseudo-critical properties for the plus fraction
dictions, consists of eight mixtures that contain hydrocarbon (HC)
regressed on PVT data [46].
molecules of diverse chemical structure such as straight chain al-
Most characterization methods comprise of three steps: split-
kanes, branched alkanes and aromatic hydrocarbons, at various
ting the C7þ fraction into single carbon number (SCN) fractions,
compositions. Moreover, the experimental dew point data cover a
assigning physical properties to each SNC according to suitable
wide range of temperatures and pressures. The compositions of the
correlations based on combinations of molecular weight (MW),
SGCs are given in Table 2.
boiling point (Tb) and density or specific gravity (SG), and, finally,
Initially the effect of the binary interaction parameters (BIPs) in
lumping them together again into multiple carbon number (MCN)
the dew point predictions of gas condensate mixtures was exam-
fractions for computational reasons without losing the information
ined. A typical example is presented with PR in Fig. 1, where the
of the SCN approach [47e49]. Detailed review of physical property
effect of using different sets of BIPs is showcased, dividing the HCs
correlations is available in the literature [50e52]. The selection of a
into two categories. The first category includes HCs lighter than n-
specific correlation for physical properties of the SCN fractions has a
decane, hereby denoted as HCset1, and the second, HCs at least as
significant effect on the EoS predictions but no correlation is
heavy as n-decane, denoted as HCset2. The following scenarios
consistently better than the others [53]. Several characterization
methods have been proposed in the literature for cubic EoSs
[51,54e58], the most popular being the ones proposed by Whitson Table 1
[58] and Pedersen [51]. The same approach has been also proposed PC-SAFT pure component parameters regressed in this work. Experimental vapor
pressure (Ps) and saturated liquid molar volume (Vl) data were taken from DIPPR
for PC-SAFT EoS, however, the determination of the pseudo-
[64].
components parameters is still under investigation [39,40,59,60].
In previous publications, it has been shown that the UMR-PRU Compound T-range (K) m s (Å) ε/k (K) %DPs %DVl
model is superior to SRK, PR and PC-SAFT in dew point pre- n-Tetracosane 325e804 9.7400 3.9582 254.36 2.6 3.3
dictions of both synthetic and real natural gases [18e20,61]. In this 2,2,4,4,6,8,8- 200e690 5.0666 4.2832 265.35 0.8 2.0
study, the UMR-PRU model is further applied in the prediction of heptamethyl-nonane
PNDP abs ðPis;exp Pis:calc Þ PNDP abs ðVl;iexp Vl;icalc Þ
dew points and liquid dropouts of synthetic gas condensates and is %DP s ¼ NDP
100
i¼1 Pis;exp
.%DVl ¼ NDP
100
i¼1 Vl;iexp
.
64 N. Novak et al. / Fluid Phase Equilibria 457 (2018) 62e73
Table 2
Molar compositions of SGC mixtures.
Table 4
Average absolute deviations in saturation point pressure for synthetic gas conden-
sate mixtures. BIPs for SRK, PR and PC-SAFT were found in literature with the
exception of those for CH4/heavy HCs which were calculated in this work.
were examined: (a) non-zero BIPs for all binaries, (b) non-zero BIPs
Fig. 1. Phase envelope prediction with PR EoS for SGC 7. Diamonds: experimental data;
only between methane and heavy hydrocarbons (HCset2), (c) non-
solid line: non-zero BIPs for all binaries; short-dashed lines: non-zero BIPs only be- zero BIPs only between methane and all HCs, and (d) BIPs equal to
tween methane and HCset2; long-dashed line: non-zero BIPs only between methane zero for all binaries. It was shown that the BIPs that mainly affect
and all HCs; dotted line: all BIPs were set equal to zero. The BIPs used for CH4/HCset2 dew point predictions are the ones between methane and heavy
binaries are the ones calculated in this work.
hydrocarbons. Actually, the use of interaction parameters only be-
tween methane with the heavier hydrocarbons yields practically
Table 3
VLE predictions (kij ¼ 0) and correlations for binary CH4/HCset2 systems using SRK, PR and PC-SAFT EoSs.
CH4 with Reference NDPa T range P range %AADPb (kij ¼ 0) Optimum kij values %AADPb (opt. kij) %AADPb
(K) (bar)
SRK PR PC-SAFT SRK PR PC-SAFT SRK PR PC-SAFT UMR-PRU
n-Decane [75,103,104] 115 303e423 1e362 8.2 9.4 13.4 0.0349 0.0388 0.0213 3.4 3.5 4.0 5.9
n-Tetracosane [77,78] 122 330e440 20e971 21.9 25.3 27.1 0.0465 0.0556 0.0383 10.0 9.9 2.5 16.0
n-Hexadecane [79e81] 92 290e360 21e703 23.4 25.4 24.1 0.0558 0.0592 0.0319 5.4 5.3 3.7 9.5
1-methyl-Napthalene [82] 7 464 21e251 18.5 21.8 27.7 0.0938 0.1009 0.0662 1.4 1.4 1.3 13.8
Phenanthrene [83e85] 60 380e473 178e960 48.3 51.6 44.23 0.2004 0.2025 0.0854 7.6 7.5 4.2 13.3
Naphthalene [85] 12 373e423 19e87 25.0 29.8 26.1 0.1110 0.1220 0.0537 2.9 2.9 2.7 7.1
n-Tridecanec e e e e e e e 0.0330 0.0364 0.0233 e e e e
2,2,4,4,6,8,8
heptamethyl-nonanec e e e e e e e 0.0558 0.0592 0.0319 e e e e
Average 22.3 24.8 25.1 6.4 6.4 3.4 11.0
a
NDP: Number of experimental data points.
b 100
PNDP abs ðPiexp Picalc Þ
%AADP ¼ NDP i¼1 Piexp
, where P is the bubble point pressure.
c
No experimental vapor-liquid equilibrium data are available.
N. Novak et al. / Fluid Phase Equilibria 457 (2018) 62e73 65
Fig. 2. Phase envelopes for synthetic gas condensates mixtures. Diamond markers: experimental data; Blue line: PR; Black line: PC-SAFT; Green line: SRK; Red line: UMR-PRU. (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
66 N. Novak et al. / Fluid Phase Equilibria 457 (2018) 62e73
the same dew point predictions with those obtained when BIPs for limitations may be imposed by different equilibrium types like
all pairs containing methane or the complete set of BIPs are used. liquid-liquid splits [87] and/or solid precipitation [78,81,83,88e90].
Although, Fig. 1 presents results only for PR, the same stands for Deviations in saturation point pressure are presented in Table 4,
SRK and PC-SAFT. while Fig. 2 presents graphically the results in PT diagrams. The
Due to the sensitivity of the EoSs to the BIPs of methane with the introduction in PR, SRK and PC-SAFT of non-zero BIPs shifts the dew
heavier HCs, the corresponding BIPs were recalculated in this work point branch to higher pressures (Fig. 1), which becomes more
while the rest were taken from the literature. Specifically, for pronounced at high molar fractions of heavy HC, and usually leads
methane with HCset1, the binary interaction parameters, although to better dew point predictions than using zero BIPs. Only for
not important for dew point predictions, were taken from the mixtures 2, 4 and 8 the introduction of non-zero BIPs leads to worse
HYSYS® process simulator database for PR and SRK. For PC-SAFT, results than using zero BIPs.
the corresponding binary interaction parameters were found in SGC 1 that contains the highest molecular weight HC (n-tetra-
the literature [71,72]. For methane with hydrocarbons belonging to cosane) in this work is better described by UMR-PRU followed by
HCset2, binary interaction coefficients were determined by fitting PC-SAFT, while cubic EoSs underestimate the experimental data.
binary vapor-liquid equilibrium experimental data. The values of For SGC2 that is actually the same gas as SGC1 but contains
the BIPs are presented in Table 3 along with the deviations in naphthalene instead of n-tetracosane, although all models can
bubble point pressures. For two mixtures, methane with n-tride- capture the change of aromaticity by predicting a narrower phase
cane and methane with 2,2,4,4,6,8,8-heptamethyl-nonane, no envelope than that of SGC1, they overestimate the experimental
experimental VLE data were found in the literature. For n-tride- data with PC-SAFT EoS yielding the best results. For SGC3 that
cane, the mean value of the BIPs of CH4/n-dodecane and CH4/n- contains both n-tetracosane and naphthalene, SRK, PC-SAFT and
tetradecane was used, which were fitted to VLE data [73e75]. For especially UMR-PRU predict more accurately the dew points than
2,2,4,4,6,8,8-heptamethyl-nonane the value of its isomer, hex- for SGC2. However, both cubic EoSs, particularly PR, underestimate
adecane, was used instead. All BIPs used for PR, SRK and PC-SAFT the experimental data of SGC3. For SGC4 that contains phenan-
are given in the supplementary tables S1 to S3. threne, UMR-PRU gives very good results, unlike PC-SAFT and cubic
Bubble point pressure predictions with EoSs, i.e. all kij's were set EoSs that yield poor predictions. In the case of SGC5 that contains
equal to zero, get progressively worse with increasing carbon the same components as SGC4, but the aromatic phenanthrene that
number. In most cases SRK yields the best predictions. As far as the is replaced by n-hexadecane, UMR-PRU still gives very good results.
correlation ability of the models is concerned, PR and SRK yield Satisfactory results are also obtained by SRK and especially PR, in
similar results, inferior to PC-SAFT, which in accordance with pre- contrast to PC-SAFT that highly overpredicts the dew points. SGCs 6
vious studies exhibits good correlation ability independently of and 7 are practically identical. For these mixtures, cubic EoSs are
chain length and molecule complexity [76]. UMR-PRU offers a very more accurate as both UMR-PRU and especially PC-SAFT over-
good compromise between accuracy and predictive ability, since its predict the experimental dew points. Finally, for SGC 8, UMR-PRU
group interaction parameters between CH4 and CH2 groups were yields very good predictions, while cubic EoSs and, especially, PC-
fitted to VLE data for alkanes up to hexadecane, and it is reasonable SAFT overshoot the phase envelope.
to expect that the UMR model cannot compete with EoS specific To summarize, PC-SAFT systematically overpredicts the dew
binary mixture correlations [77e85]. point branch of asymmetric gas condensates especially when n-
Moreover, although as already mentioned BIPs between hy- hexadecane is present. Cubic EoSs fail to describe gas condensate
drocarbons other than those containing methane are not important mixtures that contain polyaromatic components. UMR-PRU does
and they could set equal to zero, in this work BIPs for these binaries not present any systematic deviations for specific type of mole-
were also used. For PR and SRK they were taken from the HYSYS® cules, composition or complexity and exhibits the lowest overall
process simulator database and for PC-SAFT from the literature. deviation in saturation point predictions.
The extension from binary phase equilibrium to phase equilib-
rium in multicomponent gas condensate mixtures is not an easy 3.2. Liquid dropout predictions for SGC mixtures
task. The regression of binary interaction parameters at conditions
where gas condensates are present is not always possible due to the The database used for the evaluation of the models in liquid
lack of VLE experimental data. Usually the available experimental dropout predictions of synthetic gas condensates is presented in
VLE data for methane mixtures with heavy HCs are limited to Table 5. Unfortunately, such experimental data are scarce, so the
relatively low pressures unlike the high pressures encountered in database is less diverse than that developed for dew points.
gas condensate phase equilibria [86]. Furthermore, other Methane molar concentrations ranges from 0.63 up to 0.82 and
Table 5
Molar compositions of VPL mixtures.
SGC VPL 1 VPL 2 VPL 3 VPL 4 VPL 5 VPL 6 VPL 7 VPL 8 VPL 9 VPL 10
[91] [91] [91] [91] [91] [91] [91] [69] [69] [69]
T (K) 366.5 338.7 366.5 366.5 338.7 310.9 394.3 333.2 353.2 383.2
temperatures from 310.9 to 394.3 K. Since liquid molar volume is involved in the calculations, volume
The liquid dropout on a volume basis is defined as following: translation should be used with cubic EoSs and UMR-PRU. How-
ever, volume translation does not significantly affect the liquid
100,ð1 bÞvliq dropout predictions, as its effect is almost canceled out due to the
VpL% ¼ (1) definition of the VpL%. Typical examples with and without Pene-
ð1 bÞvliq þ b,vgas
loux volume translation [15] are shown in Fig. 3 for PR and UMR-
PRU. The results for both models are almost identical with and
vliq is the liquid phase molar volume, vgas is the vapor phase molar
without volume translation, and, therefore, no volume translation
volume and b ¼ V/(V þ L), where V corresponds to the moles of the
was used in the rest of this work.
vapor phase and L to the moles of the liquid phase at equilibrium.
Fig. 3. Effect of volume translation on liquid dropout predictions. Diamond markers: experimental data; Blue line: PR; Red line: UMR-PRU; Solid line: vpl% without volume
translation; Cross markers: vpl% with Peneloux volume translation. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of
this article.)
Fig. 4. Liquid dropout curves for synthetic gas condensates mixtures. Diamond markers: experimental data; Blue line: PR; Black line: PC-SAFT; Green line: SRK; Red line: UMR-PRU.
(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
68 N. Novak et al. / Fluid Phase Equilibria 457 (2018) 62e73
VpL results obtained from isothermal flash calculations are comparison because no well-established characterization method
presented graphically in Fig. 4. The predictions for PR, PC-SAFT and exists for this model.
UMR-PRU are very satisfactory for the low-pressure part of the VpL Pedersen's method uses as input variables the molecular weight
curve up to the maximum VpL, with no model having an advantage and density of the C7þ fraction and consists of two steps, splitting
over the others. PC-SAFT yields the lowest maximum VpL pre- and lumping. At first, the plus fraction is split into single carbon
dictions between the models, and SRK the highest. The worst re- number fractions (SCN), from C7 up to C80. Eq. (2) assumes a linear
sults are obtained in most cases by SRK, which was expected since relationship between the carbon number of the SCN fractions
SRK yields worse liquid molar volume predictions for hydrocarbon (CSCN) and the natural logarithm of the SCN mole fraction (zSCN)
mixtures than the other models [40]. For lower temperatures, that within the Cþ fraction, and each SCN is assigned to a molecular
are usually closer to the critical point of the mixtures, deviations weight (MW) according to Eq. (4). In this equation, the constant 14
from the experimental data are more pronounced both at accounts for the molecular weight increment from each carbon
maximum liquid dropout and at the high pressure part of the curve, atom that is accompanied from two hydrogen atoms and the
since the predicted upper dew-point pressure lead to substantial term 4 accounts for the presence of aromatic structures in the
error in the VpL (VPL1, VPL2, VPL6 and VPL8). Moreover, the effect reservoir fluid. Constants A and B are determined by simultaneous
of temperature on liquid dropouts (VPL 8 to 10) is qualitatively solving of Eqs. (2), (5) and (6). In this procedure components with
captured by all models, with higher maximum dropouts as tem- molar fractions smaller than 108 are ignored.
perature decreases.
PRU, the method has been properly adapted, taking into account ZCþ ¼ zSCN (5)
SCN¼7
that UMR-PRU treats compounds of known chemical structure,
since it utilizes UNIFAC. PC-SAFT has not been included in this
Table 6
Molar composition and C7þ properties of RGC systems.
[92] [27] [27] [41] [41] [93] [93] [93] [93] [93] [94] [95] [95] [95]
Nitrogen 3.912 0.600 0.440 0.436 0.749 0.550 0.530 0.520 0.520 0.510 2.170 0.610 0.570 0.590
Carbon Dioxide 0.750 5.560 5.470 0.839 0.524 3.110 2.980 2.960 2.930 2.890 0.340 0.460 0.370 0.420
Methane 70.203 85.190 85.480 95.792 96.003 75.570 72.890 72.460 71.740 70.930 70.640 68.640 59.390 64.430
Ethane 9.220 3.570 3.500 2.309 2.101 9.100 8.960 8.930 8.880 8.850 10.760 13.900 13.780 13.820
Propane 2.759 1.540 1.460 0.211 0.202 2.780 2.810 2.810 2.820 2.830 4.940 6.890 7.580 7.240
i-Butane 0.662 0.370 0.370 0.056 0.048 0.540 0.550 0.560 0.560 0.560 e 0.660 0.790 0.730
n-Butane 0.981 0.570 0.580 0.072 0.058 0.600 0.620 0.620 0.630 0.640 3.020 2.660 3.280 2.980
i-Pentane 0.402 0.270 0.280 0.024 0.026 0.160 0.170 0.170 0.170 0.180 e 0.620 0.840 0.740
n-Pentane 0.422 0.230 0.230 0.019 0.023 0.470 0.620 0.650 0.690 0.730 1.350 0.940 1.310 1.130
n-Hexane 0.816 e e 0.029 0.038 0.460 0.620 0.640 0.690 0.730 0.900 1.140 1.800 1.480
C7þ 9.873 2.100 2.140 0.210 0.228 6.670 9.250 9.670 10.360 11.150 5.880 3.480 10.290 6.440
MW(Cþ) (g/mol) 192.80 129.81 143.60 156.08 151.99 193.50 193.50 193.50 193.50 193.50 153.00 152.30 177.10 166.40
density at 15 C (kg/m3) 803.0 772.8 785.5 804.4 799.4 807.3 807.3 807.3 807.3 807.3 809.3 776.3 800.2 790.6
[96] [48] [48] [97] [97] [97] [97] [97] [97] [98] [55] [55] [55] [99]
Nitrogen 2.440 2.260 2.010 0.280 1.150 0.950 0.770 5.820 0.100 3.340 0.221 4.141 1.688 2.120
Carbon Dioxide 0.080 5.670 5.620 e 0.320 e e e e 0.600 7.806 0.238 0.815 0.230
Methane 82.100 45.740 46.790 86.860 80.610 74.060 89.800 56.850 94.510 74.160 71.052 91.246 84.629 73.150
Ethane 5.780 11.470 12.650 2.370 4.600 7.720 3.810 11.090 1.680 7.900 10.565 2.711 5.898 9.390
Propane 2.870 7.590 5.870 1.500 2.380 4.850 1.510 8.580 0.920 4.150 4.823 0.609 2.875 4.860
i-Butane 0.560 e e e e e e e e 0.710 0.657 0.137 0.574 1.020
n-Butane 1.230 6.380 6.040 1.330 1.900 3.100 0.800 4.810 0.610 1.440 1.613 0.231 0.820 1.780
i-Pentane 0.520 e e e e e e e e 0.530 0.383 0.058 0.271 0.600
n-Pentane 0.600 4.310 3.920 0.860 1.160 1.750 0.400 2.210 0.310 0.660 0.592 0.075 0.236 0.490
n-Hexane 0.720 5.920 4.780 1.000 0.880 1.510 0.450 1.520 0.320 0.810 0.549 0.072 0.247 0.930
C7þ 3.100 10.660 12.320 5.800 7.000 6.060 2.460 9.120 1.550 5.690 1.738 0.482 1.946 5.430
MW(Cþ) (g/mol) 132.00 148.00 148.00 178.00 194.00 130.000 158.00 157.00 127.00 151.02 130.25 139.18 153.55 162.00
density at 15 C (kg/m3) 773.4 803.8 803.8 867.4 823.3 779.9 793.0 830.5 762.1 808.5 761.2 795.7 795.0 795.7
N. Novak et al. / Fluid Phase Equilibria 457 (2018) 62e73 69
Fig. 5. Saturation pressure cross plots for UMR-PRU, PR, SRK and Elsharkawy correlation for real gas condensate mixtures.
70 N. Novak et al. / Fluid Phase Equilibria 457 (2018) 62e73
Fig. 6. Liquid dropout curves for real gas condensate mixtures. Diamond markers: experimental data; Blue line: PR; Green line: SRK; Red line: UMR-PRU. (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of this article.)
N. Novak et al. / Fluid Phase Equilibria 457 (2018) 62e73 71
Fig. 6. (continued).
[45] D.L. Katz, A. Firoozabadi, in: Predicting Phase Behavior of Condensate/Crude- temperatures from 311 to 423 K and pressures to 10.4 MPa, J. Chem. Eng.
oil Systems Using Methane Interaction Coefficients, 1978. Data 37 (4) (1992) 516e520.
[46] K.S. Pedersen, A. Fredenslund, P. Thomassen, Properties of Oils and Natural [76] P.D. Ting, et al., Phase equilibrium modeling of mixtures of long-chain and
Gases, vol. 5, Gulf Pub Co, 1989. short-chain alkanes using PengeRobinson and SAFT, Fluid Phase Equil. 206
[47] C.H. Whitson, Characterizing Hydrocarbon Plus Fractions, 1983. (1e2) (2003) 267e286.
[48] K.H. Coats, Simulation of gas condensate reservoir performance, J. Petrol. [77] J.F. Arnaud, et al., Excess volumes and saturation pressures for the system
Technol. 37 (10) (1985) 1,870e1,886. methane þ n-tetracosane at 374 K. Representation by improved EOS mixing
[49] K.S. Pedersen, P. Thomassen, A. Fredenslund, Thermodynamics of petroleum rules, Fluid Phase Equil. 124 (1e2) (1996) 177e207.
mixtures containing heavy hydrocarbons. 3. Efficient flash calculation pro- [78] E. Flo€ter, T.W. De Loos, J. de Swaan Arons, High pressure solid-fluid and
cedures using the SRK equation of state, Ind. Eng. Chem. Process Des. Dev. 24 vapour-liquid equilibria in the system (methaneþ tetracosane), Fluid phase
(4) (1985) 948e954. Equil. 127 (1) (1997) 129e146.
[50] C.H. Whitson, Effect of C7þ properties on equation of state predictions, Soc. [79] M.P.W.M. Rijkers, C.J. Peters, J. de Swaan Arons, Measurements on the phase
Pet. Eng. J. 24 (6) (1984) 685. behavior of binary mixtures for modeling the condensation behavior of
[51] K.S. Pedersen, P. Thomassen, A. Fredenslund, Thermodynamics of petroleum natural gas: Part III. The system methane þ hexadecane, Fluid Phase Equil. 85
mixtures containing heavy hydrocarbons. 1. Phase envelope calculations by (1993) 335e345.
use of the Soave-Redlich-Kwong equation of state, Ind. Eng. Chem. Process [80] H.-M. Lin, H.M. Sebastian, K.-C. Chao, Gas-liquid equilibrium in hydrogen þ
Des. Dev. 23 (1) (1984) 163e170. n-hexadecane and methane þ n-hexadecane at elevated temperatures and
[52] H.A. Aladwani, M.R. Riazi, Some guidelines for choosing a characterization pressures, J. Chem. Eng. Data 25 (3) (1980) 252e254.
method for petroleum fractions in process simulators, Chem. Eng. Res. Des. [81] M. Glaser, et al., Phase equilibria of (methane þ n-hexadecane) and (p, Vm, T)
83 (2) (2005) 160e166. of n-hexadecane, J. Chem. Thermodyn. 17 (9) (1985) 803e815.
[53] C.H. Whitson, Effect of C7þ properties on equation-of-state predictions, Soc. [82] H.M. Sebastian, et al., Gas-liquid equilibrium in binary mixtures of methane
Pet. Eng. J. 24 (6) (1984) 685. with tetralin, diphenylmethane, and 1-methylnaphthalene, J. Chem. Eng.
[54] C. Li, et al., Prediction of the dew point pressure for gas condensate using a Data 24 (2) (1979) 149e152.
modified PengeRobinson equation of state and a four-coefficient molar [83] E. Flo€ter, et al., High-pressure phase equilibria in the systems methaneþ
distribution function, J. Nat. Gas Sci. Eng 27 (2015) 967e978. phenanthrene and methaneþ 1-phenyldodecane up to 400 MPa, Fluid Phase
[55] K. Folsta, G. Camargo, R. Espo sito, Gas condensate characterization from Equil. 134 (1) (1997) 1e19.
chromatogram areas and retention times, Fluid Phase Equil. 292 (1) (2010) [84] P.V. Malone, R. Kobayashi, Light gas solubility in phenanthrene: the hydro-
87e95. gendphenanthrene and methanedphenanthrene systems, Fluid Phase
[56] A. Shariati, C.J. Peters, M. Moshfeghian, Further evaluation of the shar- Equil. 55 (1) (1990) 193e205.
iatiepetersemoshfeghian C7þ characterization method, Fluid Phase Equil. [85] N.A. Darwish, K.A. Gasem, R.L.J. Robinson, Solubility of methane in benzene,
179 (1e2) (2001) 23e41. naphthalene, phenanthrene, and pyrene at temperatures from 323 to 433 K
[57] M. Sportisse, A. Barreau, P. Ungerer, Modeling of gas condensates properties and pressures to 11.3 MPa, J. Chem. Eng. Data 39 (4) (1994) 781e784.
using continuous distribution functions for the characterisation of the heavy [86] S. Stamataki, D. Tassios, Performance of cubic EOS at high pressures, Oil Gas
fraction, Fluid Phase Equil. 139 (1e2) (1997) 255e276. Sci. Technol. 53 (3) (1998) 367e377.
[58] C.H. Whitson, Society of Petroleum Engineers of AIME Report 11200, vol. [87] J.M. Prausnitz, R.N. Lichtenthaler, E.G. de Azevedo, Molecular Thermody-
1982, 1982. namics of Fluid-phase Equilibria, Pearson Education, 1998.
[59] K. Schou Pedersen, C. Hasdbjerg, PC-SAFT Equation of State Applied to Pe- [88] S.-P. Hong, K.A. Green, K.D. Luks, Phase equilibria of the mixtures methane þ
troleum Reservoir Fluids, Society of Petroleum Engineers, 2007. n-hexane þ n-hexatriacontane, methane þ toluene þ naphthalene, and
[60] X. Liang, et al., On petroleum fluid characterization with the PC-SAFT methane þ n-hexane þ naphthalene, Fluid Phase Equil. 87 (2) (1993)
equation of state, Fluid Phase Equil. 375 (2014) 254e268. 255e272.
[61] E. Voutsas, et al., Thermodynamic property calculations with the universal [89] E. Flo€ter, T.W. de Loos, J. de Swaan Arons, Hyperbaric reservoir fluids: high-
mixing rule for EoS/GE models: results with the PengeRobinson EoS and a pressure phase behavior of asymmetric methane þn-alkane systems, Int. J.
UNIFAC model, Fluid Phase Equil. 241 (1e2) (2006) 216e228. Thermophys. 16 (1) (1995) 185e194.
[62] D.B. Robinson, D.Y. Peng, The Characterization of the Heptanes and Heavier [90] J.J.B. Machado, T.W. de Loos, High pressure solidefluid and vapoureliquid
Fractions for the GPA Peng-Robinson Programs, Gas Processors Association equilibria in model hyperbaric fluids: the system methane þ tetracosane þ
Research Report RR-28, 1978. triacontane, Fluid Phase Equil. 226 (2004) 83e90.
[63] E. Voutsas, K. Magoulas, D. Tassios, Universal mixing rule for cubic equations [91] L. Yarborough, Vapor-liquid equilibrium data for multicomponent mixtures
of state applicable to symmetric and asymmetric Systems: results with the containing hydrocarbon and nonhydrocarbon components, J. Chem. Eng.
PengRobinson equation of state, Ind. Eng. Chem. Res. 43 (19) (2004) Data 17 (2) (1972) 129e133.
6238e6246. [92] T. Yang, W.D. Chen, T.M. Guo, Phase behavior of a near-critical reservoir fluid
[64] T.E. Daubert, R.P. Danner, Physical and Thermodynamic Properties of Pure mixture, Fluid Phase Equil. 128 (1e2) (1997) 183e197.
Chemicals, Data Compilation, Hemisphere, New York, 1994. [93] W. Yan, et al., A systematic experimental study on the phase behavior of
[65] J. Gross, G. Sadowski, Application of perturbation theory to a hard-chain complex fluid mixtures up to near-critical region, Fluid Phase Equil. 190
reference fluid: an equation of state for square-well chains, Fluid Phase (1e2) (2001) 159e178.
Equil. 168 (2) (2000) 183e199. [94] K. Coats, G. Smart, Application of a regression-based EOS PVT program to
[66] J. Gross, G. Sadowski, Perturbed-chain SAFT: an equation of state based on a laboratory data, SPE Reserv. Eng. 1 (03) (1986) 277e299.
perturbation theory for chain molecules, Ind. Eng. Chem. Res. 40 (4) (2001) [95] C.W. Donohoe, R.D. Buchanan Jr., Economic evaluation of cycling gas-
1244e1260. condensate reservoirs with nitrogen, J. Petrol. Technol. 33 (02) (1981)
[67] M.R. Jensen, et al., Crystallisation of heavy hydrocarbons from three syn- 263e270.
thetic condensate gases at high pressure, Fluid Phase Equil. 208 (1e2) (2003) [96] A. Firoozabadi, Y. Hekim, D.L. Katz, Reservoir depletion calculations for gas
247e260. condensates using extended analyses in the peng-robinson equation of state,
[68] P. Ungerer, et al., High pressure-high temperature reservoir fluids: investi- Can. J. Chem. Eng. 56 (5) (1978) 610e615.
gation of synthetic condensate gases containing a solid hydrocarbon, Fluid [97] S.M.J. Majidi, et al., Evolving an accurate model based on machine learning
Phase Equil. 111 (2) (1995) 287e311. approach for prediction of dew-point pressure in gas condensate reservoirs,
[69] F. Gozalpour, et al., Vapoureliquid equilibrium volume and density mea- Chem. Eng. Res. Des. 92 (5) (2014) 891e902.
surements of a five-component gas condensate at 278.15e383.15 K, Fluid [98] K.S. Pedersen, P.L. Christensen, J.A. Shaikh, Phase Behavior of Petroleum
Phase Equil. 206 (1e2) (2003) 95e104. Reservoir Fluids, CRC Press, 2012.
[70] F. Gozalpour, et al., Viscosity and density data for a six-component model [99] J.L. Vogel, L. Yarborough, The Effect of Nitrogen on the Phase Behavior and
fluid at temperatures to 423 K and pressures to 140 MPa, High. Temp. High. Physical Properties of Reservoir Fluids, Society of Petroleum Engineers, 1980.
Press. 34 (4) (2002) 483e486. [100] K.S. Pedersen, P. Thomassen, A. Fredenslund, Thermodynamics of petroleum
[71] P. Arce, M. Aznar, Modeling of critical lines and regions for binary and mixtures containing heavy hydrocarbons. 2. Flash and PVT calculations with
ternary mixtures using non-cubic and cubic equations of state, J. Supercrit. the SRK equation of state, Ind. Eng. Chem. Process Des. Dev. 23 (3) (1984)
Fluids 42 (1) (2007) 1e26. 566e573.
[72] D.N. Justo-García, et al., Calculation of critical points for multicomponent [101] E. Retzekas, et al., Prediction of physical properties of hydrocarbons, petro-
mixtures containing hydrocarbon and nonhydrocarbon components with leum, and coal liquid fractions, Ind. Eng. Chem. Res. 41 (6) (2002)
the PC-SAFT equation of state, Fluid Phase Equil. 265 (1e2) (2008) 192e204. 1695e1702.
[73] H. Nourozieh, M. Kariznovi, J. Abedi, Vaporeliquid equilibrium measurement [102] B.E. Poling, et al., The Properties of Gases and Liquids vol. 5, McGraw-Hill
and thermodynamic modeling of binary systems (methane þ n-tetradecane), New York, 2001.
Fluid Phase Equil. 318 (2012) 96e101. [103] H.-M. Lin, et al., Gas-liquid equilibrium in binary mixtures of methane with
[74] M.P.W.M. Rijkers, et al., Measurements on the phase behavior of binary N-decane, benzene, and toluene, J. Chem. Eng. Data 24 (2) (1979) 146e149.
mixtures for modeling the condensation behavior of natural gas: Part II. The [104] M. Rijkers, et al., Measurements on the phase behavior of binary hydrocar-
system methane þ dodecane, Fluid Phase Equil. 72 (1992) 309e324. bon mixtures for modelling the condensation behavior of natural gas: Part I.
[75] S. Srivastan, et al., Solubility of methane in hexane, decane, and dodecane at The system methaneþ decane, Fluid phase Equil. 71 (1e2) (1992) 143e168.