See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/229309567

Hydrate dissociation conditions for gas

mixtures containing carbon dioxide, hydrogen,
hydrogen sulfide, nitrogen, and...

Article in The Journal of Chemical Thermodynamics · March 2007

DOI: 10.1016/j.jct.2006.07.028

CITATIONS READS

31 45

6 authors, including:

Xiao-Sen Li Shuanshi Fan

Chinese Academy of Sciences South China University of Technology
157 PUBLICATIONS 2,322 CITATIONS 208 PUBLICATIONS 9,058 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Shuanshi Fan on 19 March 2014.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue
are linked to publications on ResearchGate, letting you access and read them immediately.
 J. Chem. Thermodynamics 39 (2007) 417–425
www.elsevier.com/locate/jct

Hydrate dissociation conditions for gas mixtures containing

carbon dioxide, hydrogen, hydrogen sulfide, nitrogen,
and hydrocarbons using SAFT
a,*
Xiao-Sen Li , Hui-Jie Wu a, Yi-Gui Li b, Zi-Ping Feng a, Liang-Guang Tang a,
Shuan-Shi Fan a
a
Guangzhou Institute of Energy Conversion, The Chinese Academy of Sciences, Nengyuan Road, Wushan, Tianhe District, Guangzhou 510640, PR China
b
State Key Laboratory of Chemical Engineering, Department of Chemical Engineering, Tsinghua University, Beijing 100084, PR China

Received 25 May 2006; received in revised form 19 July 2006; accepted 26 July 2006
Available online 17 August 2006

Abstract

A new method, a molecular thermodynamic model based on statistical mechanics, is employed to predict the hydrate dissociation
conditions for binary gas mixtures with carbon dioxide, hydrogen, hydrogen sulfide, nitrogen, and hydrocarbons in the presence of aque-
ous solutions. The statistical associating fluid theory (SAFT) equation of state is employed to characterize the vapor and liquid phases
and the statistical model of van der Waals and Platteeuw for the hydrate phase. The predictions of the proposed model were found to be
in satisfactory to excellent agreement with the experimental data.
 2006 Elsevier Ltd. All rights reserved.

Keywords: Hydrate; SAFT; Equation of state; Phase equilibria; Gas mixture

1. Introduction
Gas hydrates are crystalline inclusion compounds that
consist of water and at least one other compound, usually
small molecules like methane, nitrogen, and carbon diox-
ide. Water molecules are connected by hydrogen bonds
and form various types of cavities. Low molecular-weight
gas molecules are captured into these cavities. Hydrates
are non-stoichiometric compounds that form generally in
either of three distinct structures, namely structures I, II,
and H, which differ in cavity size and shape [1].
It is of great importance to understand the hydrate dis-
sociation conditions for the rational and economic design
of processes in the chemical, oil, gas, and other industries
where hydrate formation is encountered. Hence, obviously,
predicting the conditions in which hydrates are dissociated
would be valuable. van der Waals and Platteeuw [2] were
*
Corresponding author. Tel./fax: +86 20 87057037.
E-mail address: lixs@ms.giec.ac.cn (X.-S. Li).
the first who developed the basic statistical theory for com-
putation of gas hydrate dissociation pressure. Later, Par-
rish and Prausnitz [3] developed a generalized method
based on the statistical model of van der Waals and Plat-
teeuw [2] to predict the hydrate dissociation conditions in
the presence of pure water. To expand the range of appli-
cability, Ng and Robinson [4], Holder et al. [5] and John
et al. [6] improved this model. To predict the effects of
inhibitors on the hydrate formation conditions, Hammers-
chmidt developed the first method used in the industry for
predicting the inhibiting effect of methanol [7]. Anderson
and Prausnitz [8] developed a method based on the identity
of fugacities to predict hydrate dissociation conditions with
methanol. Englezos et al. presented a methodology based
on the Trebble–Bishnoi equation of state to calculate the
inhibition effects of methanol [9,10].
So far, there are various thermodynamic models dealing
with the phase equilibria for hydrate systems developed or
modified to accommodate specific requirements appeared
in a variety of hydrate mixtures [2–13]. In short, in these

0021-9614/$ - see front matter  2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jct.2006.07.028
418 X.-S. Li et al. / J. Chem. Thermodynamics 39 (2007) 417–425

List of symbols

A Helmholtz free energy (J) e/k energy parameter of dispersion (K)

C Langmuir constant (MPa1) eAB/k energy parameter of association between sites A
Cp heat capacity (J Æ mol1 Æ K1) and B (K)
d hard-sphere diameter (1 · 1010 m) jAB bonding volume
f fugacity (MPa) DAB association strength between sites A and B
g radius distribution function l chemical potential
k Boltzmann constant (J Æ K1) q molar density (mol Æ m3)
mi effective number of segments in component i qn number density (m3)
(i = 1, 2, . . . , n) r soft-sphere diameter (1 · 1010 m)
M number of associate sites vm number of cavities of type m
n number of components in mixture
nc number of hydrate forming substances Subscripts
N number of molecules i, j, k components
NA Avogadro constant (6.02217 · 1023 mol1) m type of cavity
r radial distance from center of hydrate cavity (m) w water
R gas constant (8.3143 J Æ mol1 Æ K1)
Rm type m spherical cavity radius (m) Superscripts
P pressure (MPa) assoc association interaction
T absolute temperature (K) A, B association site
v molar volume (m3 Æ mol1) chain hard-sphere chain
W(r) cell potential function (J) disp dispersion interaction
xi mole fraction of component i in liquid phase hs hard-sphere
(i = 1, 2, . . . , n) res residual term
X Ai mole fraction of molecule i not bonded at site A H hydrate
yi mole fraction of component i in vapor phase L liquid
(i = 1, 2, . . . , n) L pure liquid water
Z compressibility factor MT empty lattice
 reference conditions of 273.15 K and zero abso-
Greek letters lute pressure
b 1/kT V vapor

traditional models as well as ones developed in recent
years, most of them applied either cubic equations of state
or activity coefficient models, which all are empirical or
semi-empirical models, for vapor and liquid phases. The
parameters in the models usually have little physical mean-
ing. The statistical associating fluid theory (SAFT) based
on Wertheim’s first-order thermodynamic perturbation
theory for associating fluid [14] has been developed very
rapidly in recent years [15]. The molecular-based equations
of states with salient physical meaningful parameters are
generally more reliable than empirical models for extrapo-
lation and prediction. Consequently, the SAFT has been
used to calculate successfully a wide variety of the thermo-
dynamic properties and phase equilibria for industrially
important fluids containing n-alkane mixtures, alcoholic
aqueous solutions and other mixture systems [15–18].
Recently, we first successfully used the SAFT equation in
conjunction with the van der Waals–Platteeuw model to
predict the thermodynamic inhibiting effect of methanol
and glycols on single gas hydrate formation [19]. It is noted
that the vapor and liquid phases were modeled using
SAFT, and the solid hydrate phase was modeled with the
van der Waals–Platteeuw model. In the present work, with
predictive success of our previous work, the above model
was extended to the prediction of the hydrate dissociation
conditions for binary gas mixtures.
2. Thermodynamic model
For the three phase vapor (V)/liquid (L)/solid hydrate
(H) system, the thermodynamic equilibrium condition is
described by:
fiL ¼ fiV ði ¼ 1; . . . ; N Þ; ð1Þ
fjH ¼ fjV ðj ¼ 1; . . . ; nc Þ; ð2Þ
where f is the fugacity of component i or j; N is all the com-
ponents; nc is the hydrate forming components including
water.
In the above equations, the fugacities in vapor, liquid
and solid phases may be calculated using a suitable thermo-
dynamic model. In this work, the SAFT equation of state is
employed for vapor and liquid phases. The van der Waals–
Platteeuw model is used for the solid hydrate phase.
 X.-S. Li et al. / J. Chem. Thermodynamics 39 (2007) 417–425 419

2.1. SAFT equation of state 2.1.3. Dispersion term

This term based on the Lennard-Jones potential is calcu-
The residual Helmholtz free energy for an n-component lated by use of the expression of Cotterman et al. [22].
mixture of associating chain molecules can be expressed as Xk
the sum of hard-sphere repulsion, hard chain formation, Adis 1
¼ xi mi ðAdis dis
1 þ A2 =T R Þ; ð11Þ
dispersion and association terms: NkT i¼1
T R

Ares ¼ A  Aid ¼ Ahs þ Achain þ Adisp þ Aassoc ; ð3Þ where

id
where A is the free energy of an ideal gas with the same Adis 2 3
1 ¼ qR ð8:5959  4:5424qR  2:1268qR þ 10:285qR Þ;
density and temperature as the system, Ahs is the free energy
of a hard-sphere fluid relative to the idea gas, Achain is the ð12Þ
free energy when chains are formed from hard spheres, Adisp Adis 2 3
2 ¼ qR ð1:9075 þ 9:9724qR  22:216qR þ 15:904qR Þ;
and Aassoc are the contributions to the free energy of disper-
ð13Þ
sion and association interactions, respectively.
T R ¼ kT =ex ; ð14Þ
2.1.1. Hard-sphere repulsion term 6
qR ¼ pffiffiffi f3 ; ð15Þ
The hard-sphere term Ahs is described with the Boublik– 2p
Mansoori–Carnahan–Starling–Leland equation [20,21]. X n X
n

P " # ex r3x ¼ y i y j eij r3ij ; ð16Þ

Ahs 6 ni¼1 xi mi 3f1 f2  f32 =f23 f32 =f23 f32 i¼1 j¼1
¼ þ 2
þ 2 lnð1  f3 Þ 
NkT pqs 1  f3 ð1  f3 Þ f3 n X
X n
r3x ¼ y i y j r3ij ; ð17Þ
Xn
i¼1 j¼1
xi mi lnð1  f3 Þ; ð4Þ
i¼1 xi mi
y i ¼ Pn : ð18Þ
where j¼1 xj mj

p X n In the above equations, the following combining rules

fl ¼ qn xi mi d lii ðl ¼ 1; 2; 3Þ; ð5Þ are used to calculate the cross parameters between different
6 i¼1
segments rij and eij
X
n
qs ¼ qn xi mi : ð6Þ rij ¼ ðrii þ rjj Þ=2; ð19Þ
i¼1 pffiffiffiffiffiffiffiffiffi
eij ¼ ð1  k ij Þ eii ejj ; ð20Þ
Here, qn is the total number density of molecules in the sys-
where kij is binary interaction parameter.
tem, and dii is the hard-sphere diameter of segment i. Its
relationship with the soft-sphere diameter (rii) proposed
2.1.4. Association term
by Cotterman et al. is based on the Barker–Henderson per-
The Helmhotz energy due to association is calculated
turbation theory and is expressed as follows [22]:
with the expression of Chapman et al. [24].
d ii 1 þ 0:2977kT =eii " #
¼ ; ð7Þ Aassoc X X 1
rii 1 þ 0:33163kT =eii þ 0:001047ðkT =eii Þ2 ¼ xi Ai Ai
ðln X  X =2Þ þ M i ; ð21Þ
NkT i Ai
2
where eii is the energy parameter of the L-J potential.
where Mi is number of associating sites on molecule i. The
2.1.2. Hard chain formation term term X Ai is defined as the mole fraction of molecules i not
The chain term Achain was given by Chapman et al. [23] bonded at site A, in mixtures with other components, and is
given by:
Achain X k
" #1
¼ xi ð1  mi Þ lnðghs
ii ðd ii ÞÞ; ð8Þ
NkT XX
i¼1 Ai Bj Ai Bj
X ¼ 1 þ NA xj qX D ; ð22Þ
where j Bj
P
gseg hs
ij ðd ij Þ  g ij ðd ij Þ ¼ where Bj expresses summation over all sites on molecule
 2 P
1 3d ii d jj f2 d ii d jj f22 j, Aj, Bj, Cj, . . . , j means summation over all components,
þ 2
þ 2 : q is the total molar density of molecules in the solution,
1  f3 d ii þ d jj ð1  f3 Þ d ii þ d jj ð1  f3 Þ3
and DAi Bj is associating strength and is given by
ð9Þ seg
DAi Bj ¼ d 3ij gij ðd ij Þ jAi Bj ½expðeAi Bj =kT Þ  1: ð23Þ
Equation (9) for like segments becomes
In equation (23), jAB is the bonding volume and eAB/k as
1 3d ii f2 d 2ii f22
ghs
ii ðd ii Þ ¼ þ þ : ð10Þ the associating energy. For cross-associating mixtures, we
1  f3 2ð1  f3 Þ2 2ð1  f3 Þ3
use the following mixing rules [18]
420 X.-S. Li et al. / J. Chem. Thermodynamics 39 (2007) 417–425

jAj Bi ¼ jAi Bj ¼ ðjAi Bi þ jAj Bj Þ=2; ð24Þ TABLE 1

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Segment parameters for pure fluids in the SAFT equation
eAj Bi ¼ eAi Bj ¼ eAi Bi eAj Bj : ð25Þ Fluid m r (1010 m) e/k (K) eAB/k (K) jAB
Water 0.982 2.985 433.91 1195.20 0.038
2.2. Model for hydrate phase Methane 1.186 2.990 160.84
Ethane 1.437 3.193 199.73
Propane 2.367 3.078 174.07
The statistical mechanics model of van der Waals and Butane 2.669 3.068 184.12
Platteeuw [2] is used for the fugacity of water in the hydrate Isobutane 2.166 3.794 203.97
phase in this work. It is described as follows: Ethene 1.843 3.248 150.63
  Propylene 2.287 3.082 171.13
H MT DlMT–H
W
fW ¼ fW exp ð26Þ CO2 1.833 2.654 165.80
RT N2 1.112 3.473 90.84
where H2 0.639 3.884 32.99
!! H2S 1.006 3.680 293.88 534.18 0.009
DlMT–H X 2 X nc
W
¼ mm ln 1 þ C mj fj ; ð27Þ
RT m¼1 j¼1
in the SAFT equation of state. These parameters are the
where DlMT–H
W ¼ lMT H
W  lW , and it represents the difference L-J potential well depth (e/k), the soft sphere diameter of
between the chemical potential of water in the empty lattice segments (r), the number of segments in a molecule (m),
(MT) and that in the hydrate lattice (H). Cmj are the Lang- the bonding volume (jAB) and the association energy
muir constants and represent the gas–water interactions. between sites A and B (eAB). In the case of mixtures, the
They may be obtained from the following equation given van der Waals one-fluid mixing rules with the binary inter-
by John and Holder [25] for a single spherical water shell action parameter, kij, for the dispersion interactions are
Z Rm  
4p W mj ðrÞ 2 used in the SAFT. In this work, the segment parameters
C mj ¼ exp r dr: ð28Þ required in the SAFT for methane, ethane, propane,
KT 0 KT
butane, isobutene, propylene, and CO2 are taken from Li
In the above equation, Rm is the type m spherical cavity ra-
and Englezos and are shown in table 1 [17,18]. The param-
dius; Wmj(r) is the function for the cell potential, and it has
eters for ethene, N2, H2, and H2S were obtained by simul-
been developed by Mckoy and Sinanoghu [26] using the
taneously fitting the experimental saturated vapor
Kihara potential function. In general, the Langmuir con-
pressures and liquid densities available in the literature
stants, Cmj, are calculated with equation (28). However,
[27,28] as given in table 1. It is noted that the binary inter-
for temperatures between 260 K and 300 K, Cmj may be
action parameters for the studied systems, kij, were taken as
obtained alternatively from a simple empirical correlation
zero except for the interaction parameter between CO2 and
of Parrish and Prausnitz [3]
water molecules in H2/CO2/water hydrate system,
C mj ¼ ðAmj =T Þ expðBmj =T Þ; ð29Þ kij =  0.0452, available in reference [18].
where Amj and Bmj are the fitted constants, which have been
given from Parrish and Prausnitz [3]. 3. Computational procedure
In this work, the temperature range of the studied sys-
tems all lay between 260 K and 300 K. Accordingly, we In equation (27), fj are the fugacities of the hydrate
do not use equation (28), in which the potential, Wmj are forming gases in the vapor phase. Hence, the (vapor + hy-
given from Mckoy and Sinanoghu [26] but we use equation drate) equilibrium conditions for nc hydrate forming sub-
(29) for all the calculations for Langmuir constants. stances, given by equation (2), are implicitly incorporated
The fugacity of water in the empty hydrate lattice, fWMT , is in equation (27). Because we search the hydrate dissocia-
obtained from the difference in the chemical potential of tion conditions, the equilibrium criteria given by equation
water in the empty lattice and that of pure liquid water, (1) may be satisfied by performing (isothermal + isobaric)

DlWMT–L ¼ lMT L (vapor + liquid) equilibrium calculations. The computa-
W  lW , using the expression given by Holder
et al. [5]: tional scheme is given as shown in figure 1. To calculate
 L
 the hydrate dissociation pressure (or temperature) at a
MT L DlMT–
W
fW ¼ fW exp ; ð30Þ given temperature (or pressure) for a system (1) enter the
RT temperature (or pressure), feed composition, and initial

where fWL is the fugacity of pure water, and it can be calcu- guess for pressure (or temperature); (2) carry out the tem-
lated from the SAFT equation of state. The calculation of perature–pressure flash calculations; (3) calculate the

DlMT–L
W is elsewhere [5,17]. fugacity of water in the hydrate using equation (26); (4)
check if f HW is equal to the fugacity of water in the liquid
2.3. Parameters for SAFT phase or vapor phase, the tolerance used is 1012. If this
fugacity is equal to the fugacity of water computed by
Three parameters for a pure non-associating fluid and the flash calculations, the assumed pressure (or tempera-
five parameters for a pure associating fluid are required ture) is the hydrate dissociation pressure (or temperature);
 X.-S. Li et al. / J. Chem. Thermodynamics 39 (2007) 417–425 421

H2). Table 2 also presents information about the ADD(P),

Start %. As seen from the table, the predictions are excellent.
The total AADs for methane/CO2, methane/N2, meth-
ane/H2, and methane/H2S are 0.35%, 1.41%, 1.29%, and
Input T (or P), feed composition and initial guess 0.64%, respectively. The minimum deviation for all above
for P (or T) hydrate systems is only 0.10% and the maximum one is also
only 2.35%.

4.2. CO2/water/ethane
Perform T-P flash calculation
We also employed the SAFT equation to predict the
hydrate dissociation pressures of the CO2/ethane gas
mixture with 40% to 96% of CO2. The results are shown
fwH Update P or (T)
in table 2 and figure 6. The total AAD is only 0.22. The
maximum deviation is also only 0.41% at 63% of CO2.
The prediction is in quite excellent agreement with the

( )
2 No experimental data. It is clear that the SAFT equation per-
ln f wH f wL < Tol formed a great job for such a gas mixture.

4.3. Gas/water/propane
Yes
Incipient hydrate formation P (or T) Table 2 and figures 7 and 8 present the predictions of the
hydrate dissociation conditions for CO2/propane gas mix-
ture with 40% to 91% CO2 and for N2/propane gas mixture
with 25% to 46% N2, respectively, using the SAFT equa-
Stop tion. As seen, the agreement with the experimental data
for the N2/propane is quite good. The total AAD is
0.57%. The predictions for the hydrate dissociation condi-
FIGURE 1. Computational flow scheme for the prediction of gas hydrate tions for CO2/propane gas mixture both at 41% and 90%
equilibria.
CO2 are excellent and the deviations are 0.54% and
0.31%, respectively. However, at 75% CO2 the model over-
otherwise, return to step 2 after updating the pressure (or predicts the data at one lower temperature (280.2 K).
temperature).
4.4. CO2/water/butane
4. Predictions of hydrate dissociations
The hydrate dissociation condition from CO2/butane
In the work, 11 hydrate systems, methane/CO2/water, gas mixture was also computed using SAFT. The predicted
methane/H2/water, methane/N2/water methane/H2S/ results are given in table 2. Figure 9 also shows the predic-
water, ethane/CO2/water, propane/CO2/water, propane/ tion with the experimental data. The total ADD is 0.23%.
N2/water, buthane/CO2/water, isouthane/CO2/water, The prediction compares quite well with the experimental
CO2/N2/water, and CO2/H2/water, were examined for the data.
prediction of the hydrate dissociation pressure at the given
temperature. The results are given in table 2. The following 4.5. CO2/water/isobutane
is the expression of the absolute average deviation of pre-
dicted pressure (AAD(P), %): The SAFT equation was also used to predict the hydrate
 X N P  
 dissociation pressures of the mixture of isobutane, the iso-
AADðP Þð%Þ ¼
1 P cal  P exp   100; ð31Þ
N P i¼1  P exp  i mer of butane, with concentrations of CO2 from 66% to
98%. Table 2 and figure 10 give the predicted values and
where NP is the number of data points. experimental data with the ADDs. The SAFT performs
an excellent prediction function as shown. The total
4.1. Gas/water/methane ADD is 0.65%.

The experimental data along with predictions are given
in figures 2 to 5 for the hydrate dissociation pressures of
different binary gas mixtures containing methane/CO2
(8% to 68% CO2), methane/N2 (13% to 50% N2), meth-
ane/H2S (3% to 22% H2S), and methane/H2 (22% to 36%
4.6. CO2/water/N2
The hydrate dissociation pressures for CO2/N2 mixture
with 91% and 97% CO2 are shown in figure 11. It can
be seen from the figure that the SAFT has a slight
422 X.-S. Li et al. / J. Chem. Thermodynamics 39 (2007) 417–425

TABLE 2
Predictions of the hydrate dissociation pressures
Gas component 1 Gas component 2 T-range/K P-range/MPa AAD(P)/% Data source
Methane 8% CO2 277.8 to 285.1 3.83 to 8.40 0.98 [29]
13% CO2 276.9 to 284 3.24 to 7.17 0.10
39% CO2 283.1 to 287.4 5.43 to 9.78 0.11
68% CO2 278.2 to 280.2 2.58 to 3.28 0.20

12.7% N2 282.8 to 295.2 7.40 to 31.31 2.35 [30]

26.9% N2 273.2 to 294.4 3.90 to 34.33 0.61
49.75% N2 273.2 to 291.8 4.96 to 33.19 1.27

3% H2S 278.7 to 287.6 2.83 to 6.65 0.26 [31]

6% H2S 279.3 to 287.1 2.21 to 4.79 0.64
11% H2S 281.5 to 292.1 2.07 to 6.00 1.10
22% H2S 279.8 to 287.6 1.03 to 2.10 0.56
22.13% H2 274.3 to 278.2 3.72 to 5.34 1.07 [32]
36.18% H2 274.3 to 278.2 4.46 to 6.63 1.50
Ethane 40% CO2 280.2 to 287.8 1.3514 to 3.8266 0.13 [33]
63% CO2 274.8 to 278.7 0.8894 to 1.4203 0.41
82% CO2 279.4 to 283.0 1.9581 to 3.1509 0.23
96% CO2 275.2 to 280.6 1.4824 to 2.8544 0.09
Propane 40% CO2 274.8 to 279.7 0.324 to 0.793 0.54 [34]
75% CO2 280.2 to 283.8 0.979 to 1.917 8.63
91% CO2 278.9 to 283.5 1.455 to 3.034 0.31

25.0% N2 274.5 to 278.7 0.256 to 0.676 0.48 [35]

45.8% N2 274.2 to 280.3 0.332 to 1.190 0.66
Butane 95% CO2 273.7 to 277.0 1.3172 to 1.9931 0.11 [33]
98% CO2 273.7 to 277.0 1.3379 to 2.0414 0.35
Isobutane 66% CO2 273.7 to 277.6 0.2068 to 0.4413 0.53 [33]
79% CO2 277.6 to 280.9 0.5378 to 0.8756 0.27
85% CO2 275.9 to 280.9 0.4964 to 1.0549 0.48
98% CO2 273.7 to 280.9 1.1376 to 3.1716 1.30
N2 91.0% CO2 273.4 to 279.1 1.37 to 3.09 5.33 [36]
96.5% CO2 273.1 to 279.4 1.22 to 2.89 6.65
CO2 16.7% H2 273.9 to 281.6 1.58-5.15 0.46 [37]
42.1% H2 274.6 to 281.4 2.77 to 8.31 0.29
60.8% H2 273.9 to 278.4 5.56 to 10.74 0.36

10.00 35.00
Symbols: Experimental Symbols: Experimental
9.00 8 % CO2 12.70 % N2
30.00
13 % CO2 26.90 % N2
8.00 49.75 % N2
39 % CO2 25.00 Predicted
7.00 68 % CO2
P / MPa
P / MPa

Predicted 20.00
6.00

5.00 15.00

4.00
10.00
3.00
5.00
2.00

1.00 0.00
276.0 278.0 280.0 282.0 284.0 286.0 288.0 272.5 275.0 277.5 280.0 282.5 285.0 287.5 290.0 292.5 295.0
T/K T/K
FIGURE 2. Gas hydrate dissociation condition prediction for methane/ FIGURE 3. Gas hydrate dissociation condition prediction for methane/
CO2/water system. N2/water system.
 X.-S. Li et al. / J. Chem. Thermodynamics 39 (2007) 417–425 423

7.00 3.20
Symbols: Experimental
3 % H2S 2.80 Symbols: Experimental
6.00
6 % H2S 91 % CO2
11 % H2S 2.40 75 % CO2
5.00 22 % H2S 40 % CO2
Predicted 2.00 Predicted
P / MPa

P / MPa
4.00
1.60

3.00 1.20

2.00 0.80

0.40
1.00
277.5 280.0 282.5 285.0 287.5 290.0 292.5 274.0 275.0 276.0 277.0 278.0 279.0 280.0 281.0 282.0 283.0 284.0

T/K T/K
FIGURE 4. Gas hydrate dissociation condition prediction for methane/ FIGURE 7. Gas hydrate dissociation condition prediction for propane/
H2S/water system. CO2/water system.

7.80
1.20 Symbols: Experimental
7.20 Symbols: Experimental
22.13 % H2 25.0 % N2
6.60 36.18 % H2 45.8 % N2
1.00
Predicted Predicted
6.00
P / MPa

0.80
P / MPa

5.40

4.80
0.60
4.20

3.60 0.40

3.00
0.20
274.0 275.0 276.0 277.0 278.0 279.0 280.0 281.0
274.0 275.0 276.0 277.0 278.0 279.0
T/K T/K
FIGURE 8. Gas hydrate dissociation condition prediction for propane/
FIGURE 5. Gas hydrate dissociation condition prediction for methane/
N2/water system.
H2/water system.

4.00 Symbols: Experimental 2.10

40 % CO2
2.00 Symbols: Experimental
3.50 63 % CO2
95% CO2
82 % CO2 1.90 98% CO2
3.00 96 % co2 Predicted
Predicted 1.80
P / MPa

P / MPa

2.50
1.70

2.00 1.60

1.50 1.50

1.40
1.00
1.30
0.50
274.5 276.0 277.5 279.0 280.5 282.0 283.5 285.0 286.5 288.0 273.5 274.0 274.5 275.0 275.5 276.0 276.5 277.0 277.5
T/K T/K
FIGURE 6. Gas hydrate dissociation condition prediction for ethane/ FIGURE 9. Gas hydrate dissociation condition prediction for butane/
CO2/water system. CO2/water system.
424 X.-S. Li et al. / J. Chem. Thermodynamics 39 (2007) 417–425

3.50 and 97% of CO2. However, the results are still reasonable
Symbols: Experimental
and satisfactory.
3.00 66 % CO2
79 % CO2
2.50
85 % CO2 4.7. CO2/water/H2
98 % CO2
Predicted
2.00 The predicted and experimental hydrate dissociation
P / MPa

pressures of the system, CO2/water/H2, at concentrations

1.50
of CO2 from 17% to 61% were obtained. The results are
1.00
shown in table 2 and figure 12. As seen, the prediction com-
pares quite well with the experimental data and the total
0.50 AAD is only 0.37%. This also illustrates that how well
the SAFT works for the prediction of the hydrate dissoci-
0.00 ation condition for the CO2/H2 gas mixture, the synthesis
273.00 274.00 275.00 276.00 277.00 278.00 279.00 280.00 281.00
T/K
gas in an integrated coal gasification combined cycle
(IGCC).
FIGURE 10. Gas hydrate dissociation condition prediction for isobu-
tane/CO2/water system.
5. Discussions

It is presented from the above calculations that the

4.00
Symbols: Experimental
SAFT performs satisfactory to excellent predictions of bin-
3.50 91.0 % CO2 ary mixture gas hydrate dissociation conditions. This indi-
96.5 % CO2 cates that not only can the SAFT be well applied to the gas
3.00 Predicted
hydrate system with inhibitors as presented in our previous
2.50 work [19], but also can work well in gas mixtures without
P / MPa

2.00 inhibitors. Compared to the methods of Sun and Chen

[38], CSMHYD software (1998) [38] and Yoon et al. [39],
1.50
It is found that the use of the SAFT has offered improve-
1.00 ment in predictions. In the ranges of the approximately
similar temperatures and the concentrations of the compo-
0.50
nents, the total ADDs for the studied systems in this work
0.00
273.0 274.0 275.0 276.0 277.0 278.0 279.0 280.0
are less than 1.5% except for the propane/CO2/H2O and
T/K
N2/CO2/H2O systems, for which the total ADDs are
3.2% and 6.0%, respectively. For the method of Sun and
FIGURE 11. Gas hydrate dissociation condition prediction for N2/CO2/ Chen [38], the CSMHYD software (1998) [38], and the
water system.
PSRK method of Yoon et al. [39], the total ADDs for
the methane/CO2/H2O system are 2.4%, 2.6% and 2.4%,
underprediction for the above two concentrations of CO2. respectively, those for ethane/CO2/H2O system are 2.6%,
Table 2 also gives the predicted results. As seen the pre- 6.9% and 5.6%, respectively, and those for propane/CO2/
dicted deviations are, respectively, 5% and 7% at 91% H2O system are 5.8%, 13.8% and 6.2%, respectively. For
the methane/H2S/H2O system, the total ADDs with the
12.00 method of Sun and Chen [38] and the CSMHYD software
Symbols: Experimental (1998) [38] are 8.0% and 6.5%, respectively. For the meth-
11.00
16.7 % H2
10.00 42.1 % H2 ane/N2/H2O and propane/N2/H2O systems, the total
9.00 60.8 % H2 ADDs with the PSRK method of Yoon et al. [39] are,
Predicted respectively, 12.3% and 5.4%. It can illustrates from the
8.00
7.00 above results that the SAFT gives better predictions for
P / MPa

6.00 the relevant studied systems than the method of Sun and
5.00 Chen [38], the CSMHYD software (1998) [38], and the
4.00 PSRK method of Yoon et al. [39]. Hence, it can be shown
3.00 that the SAFT equation of state in conjunction with the
2.00 van der Waals and Platteeuw model has rather reliable pre-
1.00 diction abilities. The SAFT is a molecular-based equation
0.00 of state. It is one of the perturbation theories based on sta-
273.0 274.0 275.0 276.0 277.0 278.0 279.0 280.0 281.0 282.0
tistic mechanics. Its advantage is that it has a robust
T/K
extrapolation and prediction capabilities. It has been used
FIGURE 12. Gas hydrate dissociation condition prediction for H2/CO2/ to model successfully a wide variety of the thermodynamic
water system. properties and phase equilibria for industrially important
 X.-S. Li et al. / J. Chem. Thermodynamics 39 (2007) 417–425
fluids [15]. In this work, although only some mixture gas
hydrate systems used in industry are predicted under a rel-
ative narrow range of the temperature conditions, we
believe that the proposed equation can also be applied to
other mixture gas hydrate systems under a wide range of
temperature conditions. Accordingly, the SAFT is very
promising to be applied in industry, where hydrate forma-
tion is involved.
6. Conclusion
In the work, we provide the SAFT equation of state in
conjunction with the van der Waals–Platteeuw model to
predict successfully the hydrate dissociation conditions of
11 binary gas mixtures. The good agreement of predicted
values with experimental data has been found. In the
SAFT, only pure-component parameters are used without
the binary interaction parameters for the prediction of gas
hydrate dissociation condition except for the interaction
parameter between CO2 and water molecules in H2/CO2/
water hydrate system.
Acknowledgment
The authors are grateful to the research program of the
One-Hundred Talent Project of Chinese Academy of Sci-
ence and Guangdong Natural Science Foundation
(06020461) for the financial support.
References
[1] E.D. Sloan Jr., Clathrate Hydrates of Natural Gases, second ed.,
Marcel Dekker, New York, 1998.
[2] J.H. van der Waals, J.C. Platteeuw, Adv. Chem. Phys. 2 (1959)
1–57.
[3] W.R. Parrish, J.M. Prausnitz, Ind. Eng. Chem. Process Des. Dev. 11
(1972) 26–35.
[4] H.I. Ng, D.B. Robinson, Ind. Eng. Chem. Fundam. 15 (1976) 293–
294.
[5] G.D. Holder, G. Gorbin, K.D. Papadopoulos, Ind. Eng. Chem.
Fundam. 19 (1980) 282–286.
[6] V.T. John, K.D. Papadopoulos, G.D. Holder, AIChE J. 31 (1985)
252–259.
[7] E.G. Hammerschmidt, Ind. Eng. Chem. 26 (1934) 851–855.
425
[8] F.E. Anderson, J.M. Prausnitz, AIChE J. 32 (1986) 1321–1333.
[9] P. Englezos, Z. Huang, P.R. Bishnoi, J. Can. Petrol. Technol. 30
(1991) 148–155.
[10] M.A. Trebble, P.R. Bishnoi, Fluid Phase Equilb. 40 (1988) 1–21.
[11] K. Nasrifar, M. Moshfeghian, R.N. Maddox, Fluid Phase Equilib.
146 (1998) 1.
[12] J.Y. Zuo, D. Zhang, H.-J. Ng, Energy Fuels 14 (2000) 19.
[13] W. Fürst, H. Renon, AIChE J. 39 (1993) 335.
[14] M.S. Wertheim, J. Chem. Phys. 87 (1987) 7323–7331.
[15] E.A. Muller, K.E. Gubbins, Ind. Eng. Chem. Res. 40 (2001) 2193–
2211.
[16] E.C. Voutsas, G.C. Boulougouris, I.G. Economou, D.P. Tassios,
Ind. Eng. Chem. Res. 39 (2000) 797–804.
[17] X.-S. Li, P. Englezos, Ind. Eng. Chem. Res. 42 (2003) 4953–4961.
[18] X.-S. Li, P. Englezos, Fluid Phase Equilib. 224 (2004) 111–118.
[19] X.-S. Li, H.-J. Wu, P. Englezos, Ind. Eng. Chem. Res. 45 (2006)
2131–2137.
[20] T. Boublik, J. Chem. Phys. 53 (1970) 471–472.
[21] G.A. Mansoori, N.F. Carnahan, K.E. Starling, T.W. Leland, J.
Chem. Phys. 54 (1971) 1523–1525.
[22] R.L. Cotterman, B.J. Schwarz, J.M. Prausnitz, AIChE J. 32 (1986)
1787–1798.
[23] W.G. Chapman, K.E. Gubbins, G. Jackson, M. Radosz, Fluid Phase
Equilb. 52 (1989) 31–38.
[24] W.G. Chapman, K.E. Gubbins, G. Jackson, M. Radosz, Ind. Eng.
Chem. Res. 29 (1990) 1709.
[25] V.T. John, G.D. Holder, J. Chem. Phys. 96 (1982) 455–459.
[26] V. Mckoy, O. Sinanoglu, J. Chem. Phys 38 (1963) 2946–2956.
[27] B.D. Smith, R. Srivastava, Thermodynamic Data for Pure Com-
pounds, Halogenated Hydrocarbons and Alcohols, Physical Science
Data 25, Elsevier, New York, 1986 (Part B).
[28] C.F. Beaton, G.F. Hewitts, Physical Property Data for the Design
Engineer, Hemisphere, New York, 1989.
[29] S. Adisasmito, R.J. Frank, E.D. Sloan, J. Chem. Data 36 (1991) 68.
[30] J. Jhaveri, D.B. Robinson, Can. J. Chem. Eng. 43 (1965) 75.
[31] L.J. Noaker, D.L. Katz, Trans. AIME 201 (1954) 237.
[32] S.-X. Zhang, G.-J. Chen, C.-F. Ma, L.-Y. Yang, T.-M. Guo, J.
Chem. Eng. Data 45 (2000) 908–911.
[33] S. Adisasmito, E.D. Sloan, J. Chem. Data 37 (1992) 343.
[34] D.B. Robinson, B.R. Mehta, J. Can. Petrol. Technol. 10 (1971) 33.
[35] H.-J. Ng, J.P. Petrunia, D.B. Robinson, Fluid Phase Equilib. 1 (1977)
283.
[36] S.-S. Fan, T.-M. Guo, J. Chem. Eng. Data 44 (1999) 829–832.
[37] R. Kumar, H.-J. Wu, P. Englezos, Fluid Phase Equilib. 244 (2006)
167.
[38] C.-Y. Sun, G.-J. Chen, Chem. Eng. Sci. 60 (2005) 4879.
[39] J.-H. Yoon, Y. Yamamoto, T. Komai, T. Kawamura, AIChE J. 50
(2004) 203.
JCT 06-139