Fluid Phase Equilibria 189 (2001) 99–110

Experimental determination and thermodynamic modeling of

methane and nitrogen hydrates in the presence of THF, propylene
oxide, 1,4-dioxane and acetone
Y.-T. Seo, S.-P. Kang, H. Lee∗
Department of Chemical Engineering, Korea Advanced Institute of Science and Technology, 373-1 Kusong-dong,
Yusong-gu, Taejon 305-701, South Korea
Received 16 October 2000; accepted 12 June 2001

Abstract
Three-phase hydrate equilibria of the methane + water + cyclic ether and nitrogen + water + cyclic ether systems
were measured by employing the temperature search method. Three cyclic ethers of THF, propylene oxide and
1,4-dioxane were chosen and their concentration was fixed at 3 mol% relative to water. The addition of cyclic ethers
caused the hydrate equilibrium pressure to be drastically lowered at a specified temperature and equivalently the
hydrate equilibrium temperature to be greatly raised at a specified pressure. This stabilization effect on hydrate for-
mation was found to be the highest for THF solution and in the order of THF > propylene oxide > 1,4-dioxane >
acetone. A new thermodynamic model was developed to describe the phase equilibria of mixed hydrates contain-
ing two guest molecules. The Soave–Redlich–Kwong equation of state (SRK-EOS) was used with the modified
Huron–Vidal (MHV2) mixing rule incorporated with the modified UNIFAC. Both the predicted and experimental
equilibrium data agreed well over whole temperature and pressure ranges investigated in this study. © 2001 Elsevier
Science B.V. All rights reserved.
Keywords: Gas hydrate; Model; Data; Methane; Nitrogen; Cyclic ether

1. Introduction

Gas hydrates are crystalline compounds that are formed by physically stable interaction between water
and relatively small guest molecules. At high-pressure and low-temperature conditions water molecules
are connected by hydrogen bonds and form various types of cavities. Low molecular-weight gas molecules,
like methane, nitrogen, and carbon dioxide are captured into these cavities. Those non-stoichiometric
crystalline compounds are divided into three distinct structures I, II, and H, which differ in cavity size
and shape [1].
∗
Corresponding author. Tel.: +82-42-869-3917; fax: +82-42-869-3910.
E-mail address: [email protected] (H. Lee).

0378-3812/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 3 8 1 2 ( 0 1 ) 0 0 5 8 0 - 5
100 Y.-T. Seo et al. / Fluid Phase Equilibria 189 (2001) 99–110

Information on the hydrate-forming conditions of natural gas components mainly composed of light
hydrocarbons is particularly important for natural gas transportation because the formed solid hydrate
particles block transmission lines and cause tubing and casing collapse. To prevent hydrate formation
simple physical methods may be adopted, which utilize the shift of hydrate formation condition caused
by changing either temperature or pressure. The other method is to add hydrate inhibitors such as alco-
hols, glycols, and electrolytes leading to higher pressure and lower temperature conditions than those
without the addition of inhibitors [2–6]. As a technological viewpoint, the importance of hydrates is also
recognized as a means for separating carbon dioxide from power plant flue gas and as a medium for
storing natural gas by fixation. Happel et al. [7] suggested a process and apparatus for rejecting nitrogen
from binary mixtures in excess of methane, which can be possibly applied to natural gas processing.
Kang and Lee [8] developed and thermodynamically verified a new process for effectively removing
and recovering carbon dioxide from large amount of multi-component flue gas. In their study, THF
was used as a hydrate stabilizer that can greatly reduce the required hydrate formation pressure and
moreover promote the corresponding hydrate formation rate. In addition, Ng and Robinson [9] found
that acetone, a hydrate former that is soluble in water, can also stabilize or suppress the formation of
hydrates composed of water and methane. These stabilizing chemicals might be effectively used for
storing natural gas in solid hydrate state because of their effect on the shift of hydrate-forming equilib-
rium temperature and pressure to milder ones and the large increase of storage capacity. Similarly, Saito
et al. [10] showed that a large group of water-soluble ethers can stabilize gas hydrate when added in
low concentration to water phase. Jager et al. [11] measured the equilibrium data on hydrate formation
for the methane + water + 1,4-dioxane system and confirmed that the addition of 1,4-dioxane up to
concentration of 5 mol% greatly reduced the equilibrium pressure of hydrate formation at a specified
temperature.
In the present study, the three-phase hydrate, water-rich liquid and vapor (H–Lw –V) equilibria of
methane + water + cyclic ethers and nitrogen + water + cyclic ethers systems were measured to examine
the stabilization effect of certain chemicals on hydrate formation. Three cyclic ethers of THF, propylene
oxide, and 1,4-dioxane were first selected as potential stabilizers to understand their structural effects on
specific molecular interactions between water and cyclic ethers. In particular, acetone, one of the simple
linear ethers, was also tested for comparison. A thermodynamic model was developed to predict hydrate
phase equilibria using statistical approach proposed by van der Waals and Platteeuw. The interaction
between the guest and the surrounding water molecules was described with the Kihara spherical core
potential. The fugacities of vapor and liquid phases coexisting with solid hydrate are evaluated from the
Soave–Redlich–Kwong equation of state (SRK-EOS) using the modified Huron–Vidal (MHV2) mixing
rule incorporated with the modified UNIFAC model.

2. Thermodynamic model

Several thermodynamic models dealing with hydrate phase equilibria have been derived and modified
to accommodate specific requirements appeared in a variety of hydrate mixtures [1,12–15]. All of these
models basically adopt statistical thermodynamic approach proposed by van der Waals and Platteeuw
[12]. According to the van der Waals and Platteeuw model, the fugacity of guest components in the
vapor phase is in equilibrium with that in the hydrate phase. Then, the final equilibrium conditions for
the hydrate-forming system are determined by comparing the equality between the fugacity of water in
 Y.-T. Seo et al. / Fluid Phase Equilibria 189 (2001) 99–110 101

the hydrate phase and that in any other phases. According to the van der Waals and Platteeuw model,
the chemical potential of water in the hydrate phase was related to that in the hypothetical empty hy-
drate lattice. The chemical potential difference between the empty hydrate and the filled hydrate phase
µMT -H (i.e. µMT − µH ) is obtained from
w w w
 



µMT w
-H = RT υ ln 1 +
m Cmj fˆjV  (1)
m j

where υm is the number of cavities of type m per water molecule in the hydrate phase, Cmj the Lang-
muir constant of component j on cavity type m, and fˆjV the fugacity of component j in the vapor phase
with which the hydrate phase is in equilibrium. The Langmuir constant considers the interaction be-
tween guest and water molecules in the hydrate cavities. In the present study, the Kihara potential
with spherical core assumption is used for the cavity potential function because it has been reported
to give better results than the usual Lennard–Jones potentials for calculating the hydrate dissociation
pressures [16].
Using the Lennard–Jones–Devonshire cell theory, van der Waals and Platteeuw presented the Langmuir
constant as a function of temperature:
  
4π R−a  (r) 2
Cmj = exp − r dr (2)
kT 0 kT
where T is the absolute temperature, k the Boltzmann’s constant,  (r) the spherically symmetrical cell
potential and r is the radial distance from the center of the cavity. McKoy and Sinanoglu [17] gave the
expression for  (r) as
12

σ a 11  σ 6
4 a 5
 (r) = 2zε δ + δ
10
− 5 δ + δ (3)
R 11 r R R r R
where

N1
r a −N
r a −N
δ = 1− − − 1+ − (4)
N R R R R
where N is equal to 4, 5, 10 or 11. The radius of the hydrate cage, R, the coordination number, z and
the number of cages per water molecule, υ, which are all the structure parameters of structure II hydrate
were cited from the literature. The Kihara hard-core parameter, a, is given in the literature, while the
energy and size parameters, ε and σ , are determined by regressing the experimental data of hydrate phase
equilibria with Eq. (1) and the following expression for chemical potential difference of water:
 T  P
µH µ0w hw υw
w
= − 2
dT + dP − ln γw xw (5)
RT RT0 T0 RT 0 RT
Holder et al. [14] reported the well-organized Kihara potential and reference parameter values for
simple and mixed hydrate systems. To calculate the solubility of guest molecules expressed as the last
term of Eq. (5), they used the Krichevsky and Kasarnovsky solubility derivation that was also incorporated
with Henry’s law constant.
102 Y.-T. Seo et al. / Fluid Phase Equilibria 189 (2001) 99–110

It should be carefully noted that this solubility approach may be not suitable for describing hydrate
formation behavior either at high-temperature region or for the mixed hydrates containing cyclic ethers
because the solubility of cyclic ethers in water is quite significant. In this connection, the last term of
Eq. (5) was replaced with water activity by using the proper EOS. The SRK-EOS incorporated with
a MHV2 mixing rule wad used to determine the corresponding fugacity. In addition, the use of the
equation of state has another advantage of providing a consistent representation of hydrate equilibria up
to high-pressures. In the present study, the SRK-EOS incorporated with the MHV2 mixing rule could be
effectively used for predicting the phase behavior of hydrate systems containing cyclic ethers.
Jager et al. [11] developed the thermodynamic model to describe the effect of 1,4-dioxane on methane
hydrate. Due to the low fugacity of 1,4-dioxane, they assumed that the vapor phase could be treated
as a pure methane phase and calculated the activity coefficients of fluid phase by using the van Laar
model. However, at this point, it must be noted that the vapor pressures of THF, acetone and propy-
lene oxide are much higher than 1,4-dioxane at temperature range interested in this study. Therefore,
their concentration in vapor phase should be considered through the rigorous flash calculation. The
hydrate stabilizers have two considerable effects on hydrate phase equilibrium that they act as guest
components forming the mixed hydrates and change the activity coefficient of water. To calculate the
activity of water component, the modified UNIFAC was chosen because no binary interaction param-
eters are available for the pairs consisting of a ether and a gaseous guest. All the UNIFAC parameters
used in this study were cited from [18]. The mixed hydrates containing a stabilizer form the structure
II hydrates. The assumption that the stabilizers can only be encaged in the large structure II cavity
changed the fractional occupancy of hydrate cavities compared with simple hydrates. For these mixed
hydrates, the Kihara parameter a was estimated with the corresponding state correlation proposed by
Tee et al. [19] and finally the corresponding parameter values of ε and σ were determined from the
experimental data of hydrate phase equilibria. All the property values of the hydrate cavity were obtained
from [1].

3. Experiments

The methane was supplied by the Scientific Gas Products Co. and had a purity of 99.99 mol%. World
Gas Co. supplied the nitrogen with a minimum purity of 99.9 mol%. The water was supplied from
Sigma-Aldrich Chemical Co. with a purity of 99.1 mol%.
The schematic diagram of the experimental apparatus used in this work is shown in Fig. 1. The equilib-
rium cell is made of 316-stainless steel. Its internal volume is about 50 cm3 and equipped by two thermally
reinforced sight glasses. The cell contents were agitated by a magnetic spin bar that was coupled with an
immersion magnet placed under the cell in the bath. The bath contained about 30 l of ethanol and water
mixture, which was controlled by an externally circulating refrigerator/heater. The actual operating tem-
perature in the cell was maintained with the PID temperature controller (JeioTech, MC-31) with ±0.1 K
accuracy and was measured by a K-type thermocouple probe with a digital thermometer (Cole-Parmer,
8535-26) of which the resolution is ±0.1 K (Every Ready Thermometer Co. Inc.). The Heise Bourdon
tube pressure gauge (CMM 104957, 0–200 bar range) having the maximum error of ±0.1 bar in the
full-scale range was used to measure the cell pressure in the system.
After the equilibrium cell was well cleaned, the overall inner parts were evacuated with vacuum
pump. An approximately 25 ml of water and ether mixture was introduced into the evacuated equilibrium
 Y.-T. Seo et al. / Fluid Phase Equilibria 189 (2001) 99–110 103

Fig. 1. Schematic diagram of experimental apparatus: (1) equilibrium cell; (2) magnet; (3) Heise pressure gauge; (4) thermometer;
(5) rupture disc; (6) check valve; (7) high-pressure hand pump; (8) gas cylinder; (9) line filter; (10) water bath; (11) external
refrigerator/heater; (12) magnetic spin bar; (13) regulator.

cell. Then, the equilibrium cell was charged with the hydrate-forming gas. After the cell was pressur-
ized to a desired pressure, the system was cooled to about 5 K below the anticipated hydrate-forming
temperature. Hydrate nucleation and growth was then induced and the system pressure was continu-
ously decreased. After the system pressure reached a steady state, the cell temperature was gradually
elevated to dissociate the formed hydrates. The external heater was used to increase the system tem-
perature at a rate of 1–2 K/h. As the system approach the equilibrium temperature, the dissociation of
hydrate phase caused the increase of system pressure. When the amount of minute crystals remained
and the system temperature was kept constant at least for 8–10 h after the system pressure was sta-
bilized, the resulting pressure was considered as an equilibrium dissociation pressure at the specified
temperature.

4. Results and discussion

Three-phase equilibria (H–Lw –V) of the methane + water + THF, methane + water + propylene oxide,
methane + water + 1,4-dioxane and methane + water + acetone systems at a fixed stabilizer concentration
of 3 mol% were measured and the overall results were given in Table 1 and Fig. 2. For the methane +
water + acetone system the present data showed a good agreement with those of Ng and Robinson [9].
As shown in Fig. 2, the presence of cyclic ethers induced hydrate-forming condition to be more favorable
by lowering pressure and raising temperature for forming the mixed hydrates when compared with the
corresponding condition of simple methane hydrate. At a fixed pressure, the equilibrium temperature
of the methane and THF mixed hydrate appeared to be the highest among these four systems. This
stabilization effect on hydrate formation was found to be in the following order: THF > propylene oxide >
104 Y.-T. Seo et al. / Fluid Phase Equilibria 189 (2001) 99–110

Table 1
Three-phase H–Lw –V equilibria of the ternary methane + water + cyclic ether (3 mol%) systems
System T (K) P (bar)

Methane + water +THF 292.16 20.2

297.06 40.2
300.66 68.4
302.46 89.1
Methane + water + propylene oxide 283.96 19.2
289.76 44.1
292.56 68.4
294.16 87.6
Methane + water + 1,4-dioxane 280.56 17.4
287.36 43.8
290.76 67.5
292.96 89.4
Methane + water + acetone 279.06 19.2
284.46 41.1
288.76 76.9

1,4-dioxane > acetone. Similar experimental results were obtained for the nitrogen + water + THF,
nitrogen + water + propylene oxide, nitrogen + water + 1,4-dioxane and nitrogen + water + acetone
systems and given in Table 2 and Fig. 3. Surprisingly, the equilibrium dissociation pressures of the mixed
hydrates containing the ethers was greatly reduced by 100–300 bar when compared with those of simple

Fig. 2. Experimental and predicted hydrate dissociation pressures for the binary and ternary systems of: (䊊) methane + water
[20]; (䊉) methane + water + acetone [9]; () methane + water + acetone; (䉲) methane + water + 1,4-dioxane; (䊐)
methane + water + propylene oxide; (䊏) methane + water + THF; (—) predicted.
 Y.-T. Seo et al. / Fluid Phase Equilibria 189 (2001) 99–110 105

Table 2
Three-phase H–Lw –V equilibria of the ternary nitrogen + water + cyclic ether (3 mol%) systems.
System T (K) P (bar)

Nitrogen + water + THF 287.35 43.30

289.05 56.40
290.15 66.00
291.05 78.26
292.62 94.75
293.55 108.72
Nitrogen + water + propylene oxide 279.05 40.40
281.65 60.90
283.25 79.02
284.55 97.22
285.85 117.40
Nitrogen + water + 1,4-dioxane 275.05 37.00
278.25 57.75
282.15 98.00
283.65 118.40
Nitrogen + water + acetone 274.05 44.20
275.95 58.05
277.25 69.85
278.85 87.10
280.85 113.85

nitrogen hydrates. The propylene oxide having an epoxide ring enhanced the hydrate stabilization than
1,4-dioxane. On the other hand, the acetone that is not a cyclic compound showed the lowest stabilization
effect. It might be, thus, concluded that the hydrogen bonding effect of open chain ether is weaker than
that of cyclic ethers. From these experimental results, it can be expected that the use of stabilizers greatly
reduces the operating pressure for storing or transporting natural gas in hydrate state and on the contrary
raises the operating temperature nearly to ambient temperature, which provides the considerable benefit
for cost reduction.
Among the used ethers THF was proven to be the strongest for stabilizing both simple methane and
nitrogen hydrates. Further, the concentration effect of THF on forming the mixed hydrates was examined
at four different concentrations of 1, 2, 3, and 5 mol%. And the measured data were presented in Table 3
and Fig. 4. As the concentration of THF increased, the equilibrium dissociation temperature also increased
continuously at the specified pressure. THF itself can play a role in forming hydrates as a guest molecule
to host water molecules and form structure II hydrate in which the THF molecules occupy only the
larger cages. The number ratio of large cages to water molecules in a structure II hydrates appears to
be stoichiometrically 1:17 (0.056) under full occupancy. Accordingly, the excess amount of THF over
5.6 mol% shows that its hydrate promotion effect is gradually decreased. The concentration effect of THF
on the equilibrium hydrate dissociation pressures was shown at three different isotherms in Fig. 5. The
pressure decreased rapidly up to 1 mol% of THF, but very slowly at THF concentrations of more than
1 mol%.
106 Y.-T. Seo et al. / Fluid Phase Equilibria 189 (2001) 99–110

Fig. 3. Experimental and predicted hydrate dissociation pressures for the binary and ternary systems of: (䊊) nitrogen + water
[21]; (䉱) nitrogen + water + acetone; (䉲) nitrogen + water + 1,4-dioxane; (䊐) nitrogen + water + propylene oxide; (䊏)
nitrogen + water + THF; (—) predicted.

Fig. 4. Experimental and predicted hydrate dissociation pressures for the ternary nitrogen + water + THF system: (䊊) pure water
[21]; (䊉) 1 mol% THF; (䊏) 2 mol% THF; (䉱) 3 mol% THF; (䉲) 5 mol% THF; (—) predicted.
 Y.-T. Seo et al. / Fluid Phase Equilibria 189 (2001) 99–110 107

Table 3
Three-phase H–Lw –V equilibria of the ternary nitrogen + water + THF system measured at four different concentrations of THF
Concentration (mol%) T (K) P (bar)

1 280.85 31.20
284.75 53.73
286.05 63.50
288.15 80.98
289.05 97.65
2 284.75 42.40
286.05 52.19
287.85 65.53
289.05 80.92
289.75 91.92
290.65 102.10
3 284.75 29.75
286.05 36.75
287.35 43.30
289.05 56.40
291.05 78.26
292.65 94.75
293.55 108.72
5 284.75 24.63
286.05 29.73
289.05 50.95
292.05 74.29
293.75 97.47

The equilibrium P–T data was used to determine the Kihara parameters of THF, propylene oxide,
1,4-dioxane and acetone and the resulting values were presented in Table 4. However, the value of
ε for THF was obtained by taking the average of parameter values optimized at four different THF
concentrations. Conversely, these calculated Kihara parameters were used to predict hydrate equilibrium
pressures of mixed hydrates containing a ether component. Both experimental and predicted results agreed
well for all the systems examined in this study.

Table 4
Optimized Kihara parameters of THF, propylene oxide, 1,4-dioxane and acetone regressed from the measured three-phase
H–Lw –V equilibrium data

a (Å) σ (Å) ε/κ (Å)

THF 0.9643 3.2038 243.27

Propylene oxide 0.9589 2.9817 274.08
1,4-Dioxane 1.0940 2.9557 285.97
Acetone 1.0310 3.3758 233.06
108 Y.-T. Seo et al. / Fluid Phase Equilibria 189 (2001) 99–110

Fig. 5. Hydrate phase equilibria of the ternary nitrogen + water + THF system measured at three different isotherms: (䊏)
284.15 K; () 286.05 K; (䊉) 289.05 K; (—) predicted.

5. Conclusions

The three-phase (H–Lw –V) equilibria were measured for the ternary mixtures composed of one host
and two guest components. As the primary guest molecules, the methane and nitrogen were chosen,
while the ethers of THF, propylene oxide, 1,4-dioxane and acetone were used as the secondary guest
molecules and at their concentration of 3 mol% relative to water. All the systems examined in this study
showed the substantial stabilization effect of ethers on hydrate formation resulting to the order of THF >
propylene oxide > 1,4-dioxane > acetone. The formed mixed hydrates were more stabilized when
the ether concentration increased, which indicates that the hydrate formation pressure can be much
lowered. This fact might be expected to provide considerable advantages to technological applications
utilizing hydrate formation/dissociation such as natural gas storage and transportation. The equilibrium
temperatures and pressures of mixed hydrates were predicted by the SRK-EOS with the MHV2 mixing
rule incorporated with the modified UNIFAC model. Both experimental and predicted values were found
to be in good agreement except a little deviation at low-pressure.

List of symbols
a radius of the spherical core
C Langmuir constant (bar−1 )
f fugacity (bar)
h enthalpy (J/mol)
 Y.-T. Seo et al. / Fluid Phase Equilibria 189 (2001) 99–110 109

k Boltzman’s constant (J/K)

P pressure (bar)
T temperature (K)
Greek letters
ε/κ depth of the intermolecular potential well
γ activity coefficient
ϕ fugacity coefficient
µ chemical potential
σ distance between the cores at zero potential energy
υ number of cavities, number of ions
 (r) spherically symmetrical cell potential
Superscripts
H clathrate hydrate phase
MT empty
V vapor phase
Subscripts
i, j components i and j
m cavity type
w water

Acknowledgements

This work was supported by Grant no. 98-0502-04-01-3 from the Basic Research program of the
KOSEF and also partially by the Brain Korea 21 Project.

References

[1] E.D. Sloan Jr., Clathrate Hydrates of Natural Gases, 2nd Edition, Marcel Dekker, New York, 1998.
[2] S.-P. Kang, M.-K. Chun, H. Lee, Fluid Phase Equil. 147 (1998) 229–238.
[3] K. Nasrifar, M. Moshfeghian, R.N. Maddox, Fluid Phase Equil. 146 (1998) 1–13.
[4] U. Hutz, P. Englezos, Fluid Phase Equil. 117 (1996) 178–185.
[5] P.D. Dholabhai, P. Englezos, N. Kalogerakis, P.R. Bishnoi, Can. J. Chem. Eng. 69 (1991) 800–805.
[6] J.P. Lederhos, J.P. Long, A. Sum, R.L. Christiansen, E.D. Sloan Jr., Chem. Eng. Sci. 51 (1996) 1221–1229.
[7] J. Happel, M.A. Hnatow, H. Meyer, Ann. New York Acad. Sci. 715 (1994) 412–424.
[8] S.-P. Kang, H. Lee, Environ. Sci. Technol. 34 (2000) 4397–4400.
[9] H.-J. Ng, D.B. Robinson, Ann. New York Acad. Sci. 715 (1994) 450–462.
[10] S. Saito, T. Kawasaki, T. Okui, T. Kondo, R. Hiraoka, in: Proceedings of the 2nd International Conference On National Gas
Hydrates, Toulouse, France, 1996.
[11] M.D. Jager, R.M. de Deugd, C.J. Peters, J. de Swaan Arons, E.D. Sloan Jr., Fluid Phase Equil. 165 (1999) 209–223.
[12] J.H. van der Waals, J.C. Platteeuw, Adv. Chem. Phys. 2 (1959) 1–57.
[13] W.R. Parrish, J.M. Prausnitz, Ind. Eng. Chem. Proc. Des. Dev. 11 (1972) 26–34.
[14] G.D. Holder, S.P. Zeets, N. Pradhan, Rev. Chem. Eng. 5 (1988) 1–70.
[15] P. Englezos, Ind. Eng. Chem. Res. 32 (1993) 1251–1274.
110 Y.-T. Seo et al. / Fluid Phase Equilibria 189 (2001) 99–110

[16] L. Lundgaard, J.M. Mollerup, Fluid Phase Equil. 70 (1991) 70.

[17] V. McKoy, O. Sinanoglu, J. Chem. Phys. 38 (1963) 2946.
[18] R.C. Reid, J.M. Prausnitz, B.E. Poling, The Properties of Gases, Liquids, 4th Edition, McGraw-Hill, New York, 1987.
[19] L.S. Tee, S. Gotoh, W.E. Stewart, Ind. Eng. Chem. Fundam. 5 (1966) 363–367.
[20] S. Adisasmito, R.J. Frank, E.D. Sloan Jr., J. Chem. Eng. Data 36 (1991) 68–71.
[21] J. Jhaveri, D.B. Robinson, Can. J. Chem. Eng. 43 (1965) 75.