Sunteți pe pagina 1din 9

Applied Energy 187 (2017) 180188

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Gas recovery from depressurized methane hydrate deposits with


different water saturations
Mingjun Yang, Zhe Fu, Lanlan Jiang, Yongchen Song
Key Laboratory of Ocean Energy Utilization and Energy Conservation of Ministry of Education, Dalian University of Technology, Dalian 116024, China

h i g h l i g h t s
 NGHs deposits with excess gas or water were both recoveried by depressurization.
 Water saturation and depressurization range are key factors for NGHs dissociation.
 Larger depressurization range accelerates NGHs dissociation for excess gas deposit.
 The obvious water mobility in excess water deposits hinders methane gas output.

a r t i c l e

i n f o

Article history:
Received 3 July 2016
Received in revised form 31 August 2016
Accepted 8 October 2016

Keywords:
Methane hydrate
Depressurization
Excess water
Excess gas

a b s t r a c t
Natural gas hydrates (NGHs) are new and clean energy resources with significant potential. Many studies
have investigated NGHs in an attempt to recover natural gas from NGHs deposits. Additional investigations are still needed to clarify the dissociation characteristics of NGHs to develop safe and efficient
recovery methods. In this study, two types of NGH deposits were simulated by forming methane hydrates
(MHs) in porous media: the first type was formed with excess gas, and the other type was formed with
excess water. The formed MHs were dissociated by depressurization methods. Magnetic resonance imaging (MRI) was used to monitor the liquid water distribution and quantify the MH amounts during formation and dissociation. The results showed that a larger depressurization range enhanced the average rate
of MH dissociation and gas production for excess gas conditions. For excess water conditions, the mobility of liquid water was dominant during MH dissociation and hindered methane gas output. Furthermore,
a larger depressurization range accelerated MH dissociation. When MH dissociations were compared for
various gas-water saturated porous media, liquid water saturation and depressurization range were identified as two key factors affecting MH dissociation.
2016 Elsevier Ltd. All rights reserved.

1. Introduction
Natural gas hydrates (NGHs) are typically found in permafrost
and marine sediment, which contains abundant hydrocarbon gas
and water at high pressures and low temperatures [1,2]. NGHs
have recently been considered as potential non-conventional fossil
fuel resources due to their huge reserves [35]. To utilize NGHs as
an energy source, NGHs are dissociated via phase change processes
(i.e., from solid to gas and liquid) and require gas recovery methods. However, to date, no economically viable gas recovery method
has been developed for NGH deposits. Due to a number of mechanisms and new technologies still being developed, the exploitation
of NGHs has remained at laboratory and short-term production
scales. Methane hydrates (MHs) are widespread worldwide and
Corresponding author.
E-mail address: songyc@dlut.edu.cn (Y. Song).
http://dx.doi.org/10.1016/j.apenergy.2016.10.029
0306-2619/ 2016 Elsevier Ltd. All rights reserved.

are typically measured in NGH dissociation investigations [610].


These studies have primarily focused on developing guidance and
recommendations for the safe and efficiency recovery of on-site
MH deposits.
The three traditional MH exploitation methods are depressurization, thermal stimulation, and inhibitor injection [11]. Compared with the other two methods, depressurization is
considered a more competitive method due to its low energy consumption and low environmental impact [12,13]. For the depressurization process, the pressure of a MH deposit is decreased to
levels below the hydrate phase equilibrium pressure to initiate dissociation [14,15]. Many laboratory-scale studies have explored CH4
recovery methods in MH deposit depressurization [6,1619]. These
investigations have focused on gas production rates, efficiency, and
output [7,20]. Other studies have focused on heat and mass transfer characteristics [21]. The gas production characteristics have
been reported to be influenced by the compositions and sizes of

M. Yang et al. / Applied Energy 187 (2017) 180188

reservoirs [16,17]. These experiments have typically used glass


beads, silica gels, and artificial sandstone cores to simulate the
deposits. Numerical simulations have also been used to model
MH recovery by depressurization [22,23,9,24]. Recently, combinations of depressurization with other methods have been proposed
for CH4 recovery [25,26].
Overall, gas recovery from MH deposits is complicated and
involves thermodynamic considerations, two phase flows, heat
and mass transfer, physics and chemistry [8,2729]. Hydrate saturation levels and permeability in MH deposits also affect CH4
recovery [30,31]. Permeability is especially a sensitive parameter
[32,33]. Different classes of NGHs deposits are known, and three
could be potentially exploited for gas recovery [3436]. These
three types of deposits cover under burden and are covered by over
burden. The first class contains a two-phase zone under a hydratebearing layer [37]. An underlying water zone is present in the second class [38]. The third class comprises only an isolated hydrate
layer [34]. Moridis et al. investigated gas recovery from Class 1
and Class 2 deposits using numerical modeling. They found that
gas can be recovered at high rates for long periods of time from
Class 1 deposits by depressurization method. Furthermore, the
group found that the permeability of the over burden significantly
affected gas recovery from Class 2 deposits [35,36,39]. Konno et al.
also investigated depressurization-induced gas recovery methods
using simulations on oceanic NGHs sediments [40]. It is necessary
to verify the simulation results using experimental data. While the
experimental investigation are absent for the comparison of gas
recovery characters from the three classes NGHs deposits. Meanwhile, the characters of NGHs dissociation from different class
deposits are essential for the assessment of NGHs exploitation
technology adaptability. That is to say, it is desiderated to investigate the gas recovery characters from different classes NGHs
deposits.
As briefly discussed, the depressurization method for MH recovery has been widely studied, and it is surely the potential way.
While most of these investigations are single for excess gas deposits or excess water deposits, and it is hard to make comparison
between these investigations. And a few reports have compared
the use of excess gas or excess water in recovery methods in MH
deposits. It is necessary to measure MH dissociation characters
from different class deposits under different backpressures, and it
is important for the development of depressurization NGHs
exploitation technology. The present study simulated NGHs deposits with different phase saturation, and the MH dissociation characteristics for deposits with different initial water and gas
saturation levels are investigated. These reported experimental
results may provide guidance for the exploitation of different
NGH classes.

2. Experiments
2.1. Apparatus and materials
A high-pressure, nonmagnetic, polyimide-based cylindrical vessel (38  314 mm, 12 MPa) was used to form MHs. A vessel
jacket filled with a circulating Fluorinert solution (FC-40, 3 M Company, USA) was used to regulate the vessel temperature. A thermostat bath (FL300, JULABO technology, Germany) was used to
cool or heat the Fluorinert solution. A separate container in the
bath was used to maintain the temperature of water injected into
the vessel. An MRI system (operating at 400 MHz, Varian, Inc., USA)
was used to visualize and quantify the 1H in liquid water and
thereby determine water distribution in the porous media in the
vessel. As liquid water reacted to form solid hydrates, changes in
the water distribution and therefore, MH formation and dissocia-

181

tion, were detected by the MRI system. A spin echo pulse sequence
was applied at TR = 500 ms, TE = 1.63 ms, and slice thickness = 4 mm. The field of view (FOV) was 30  30 mm, and the
MRI image matrix was 128  128 pixels.
Two high-pressure pumps (260D, Teledyne Isco, Inc., USA) were
used to inject water and methane gas into the vessel. A backpressure control valve (BP-2080-M, JASCO, Japan) was connected to the
vessel outlet to control the backpressure for MH dissociation. The
data acquisition system included temperature and pressure transducers, an A/D module (Advantech Co., Ltd., China) and a computer.
The vessel pressure was measured to an accuracy of 0.1 MPa using
a pressure transducer (Nagano Co., Ltd., Japan). A thermocouple
was used to monitor the jacket temperature. BZ-02 glass beads
(0.1770.250 mm, As-One Co., Ltd., Japan) of approximately
35.4% porosity were packed into the vessel to simulate the porous
media. Methane gas (99.99%; Dalian Special Gas Co., Ltd., China)
was used to form the MHs. Deionized water was used in all experiments. Additional details regarding the experimental apparatus
are provided in a previous paper [41].
2.2. Procedure
The glass beads (BZ-02) were tightly packed into the vessel.
After a leakage test, the vessel was placed into the magnetic scanning body of the MRI system. Free gas was evacuated from the vessel, and deionized water was slowly injected to saturate the glass
beads until a vessel pressure of 6.0 MPa. Once the pressure reached
a steady state, the glass beads were considered fully saturated, and
an MRI image was obtained to show the MI value for this state.
Then, high-pressure CH4 (approximately 2.5 MPa) was injected to
displace some of the pore water to make space for CH4 to form
the hydrates. The amount of displaced water was recorded for each
cycle. Then, the thermostated bath was set and kept at 274.15 K,
i.e., the experimental temperature for MH formation and dissociation. After the vessel temperature reached a steady state, CH4 gas
was slowly injected until 6.0 MPa. Images were continuously
obtained using MRI.
After the MH formation process, two experimental procedures
were tested. The first procedure simulated a natural MH deposit
using excess gas. The vessel was depressurized to initiate MH dissociation. The second procedure simulated natural MH deposits
using excess liquid water to discharge some of the residual gas
before MH dissociation. During the injection process, the water
was maintained at 273.65 K to avoid hydrate dissociation. After
approximately 5 min, water injection (at 5 mLmin1) was stopped
when excess water flowed out of the vessel and when the vessel
pressure was steady. The amount of injected water was approximately twice the pore volume but did not completely displace
the residual gas [42]. Based on MH phase equilibrium data (i.e.,
approximately 2.98 MPa at 274.45 K) [43] and on our previous
work [44], the vessel backpressures were set to 2.8, 2.6, 2.4, and
2.2 MPa to dissociate MH deposits. Details of the experimental
parameters are shown in Table 1.
3. Results and discussions
Two experimental processes were investigated in this study and
are defined by the different water saturation levels in the MH
deposits. Case 1 utilized excess gas conditions with approximately
1517% water saturation. Case 2 utilized excess water conditions
with approximately 4250% water saturation. Four different backpressures were explored for MH dissociation. The MRI system
monitored an arbitrary sagittal plane in the vessel to determine
the liquid water distribution during MH formation and dissociation
in both processes. MH saturation, which is defined as the pore vol-

182

M. Yang et al. / Applied Energy 187 (2017) 180188

Table 1
Experimental parameters and results.a
Exp. no

Shi

Swi

Pd/MPa

Pi/MPa

t/min

r/%min1

R/mlmin1

Swe/%

Case
Case
Case
Case
Case
Case
Case
Case
Case
Case

30.7
29.5
32.7
21.2
34.8
36.7
27.6
30.0
26.6
33.6

17.1
15.1
14.9
17.6
50.0
47.1
48.8
43.9
44.3
41.9

2.8
2.6
2.4
2.2
2.8
2.6
2.6
2.4
2.4
2.2

3.35
3.52
3.40
3.61
5.01
5.24
5.06
4.92
4.70
5.53

102
38
36
16
72
88
72
66
68
26

0.301
0.776
0.908
1.325
0.483
0.417
0.383
0.455
0.391
1.290

620.3
1599.2
1871.2
2730.6
995.4
859.4
789.3
937.7
805.8
2658.4

40.6
38.5
38.2
30.9
46.7
42.5
41.4
39.1
37.4
35.4

11
12
13
14
21
22a
22b
23a
23b
24

a
The symbols in this table are defined as follows: Shi is the initial MH saturation before hydrate dissociation (volume fraction); Swi is the initial water saturation after water
injection (volume fraction); pd is the designed back pressure for hydrate dissociation; pi is the initial pressure for MH dissociation (MPa); t is the elapsed time for MH
dissociation; r is the average MH dissociation rate; R is the gas production rate (the gas volume is present at standard temperature and pressure); and Swe is the final residual
water saturation after MH dissociation (volume fraction). All dissociations were carried out at 274.15 K. There is no blockage for all cases.

ume fraction of the sediment occupied by the hydrate, was quantified based on the mean intensity (MI) data collected by MRI using
Eq. (1). A prior study reported that 1 m3 of MH releases 0.8 m3 of
liquid water, i.e., one volume of fresh water can form 1.25 volumes
of hydrate at standard temperature and pressure (STP) [45]. Thus,
the MH saturation at i minutes can be calculated as:

Shi 1:25 

I0  Ii  SW0
 100%
I0

where Sw0 is the initial water saturation, and I0 and Ii are the MRI
MIs of liquid water at t = 0 and i minutes, respectively. The water
saturation level at time i can also be calculated using MI as shown
in Eq. (2):

Swi

Ii  SW0
 100%
I0

2
Fig. 1. MI and vessel pressure changes during MH dissociation at different back
pressure for Case 1.

3.1. MH dissociation at different pressure for Case 1


The MH formation process was reported in our previous work
[44]. Briefly, the initial pressures and temperatures were set to
6.0 MPa and 274.15 K to form the MH deposits in all experiments.
The MH dissociation characteristics for Case 1 are summarized in
this section. The effects of backpressure (2.8, 2.6, 2.4, and
2.2 MPa) are discussed in detail, and the experimental parameters
and results are shown in Table 1. The curves for changes in MI (D
MI) and pressure for Case 1 (4 cycles) are shown in Fig. 1. The D
MI data shows the differences in residual water content between
the MH dissociation process and after MH deposit formation. An
increase in D MI indicated that solid hydrate was converted into
liquid water; the rate of MH dissociation for this sample was calculated from the instantaneous slope of the curve at a given backpressure. The average MH dissociation rate (r) and gas
production rate (R) were calculated using the following equations:

Shi
t

R r  V  /  160

3
4

where Shi is the initial MH saturation before hydrate dissociation


(volume fraction), t is the elapsed time for MH dissociation), V is
the volume of the high pressure vessel, U is the porosity of the porous media and it is equal to 36.8%, and it is also assuming that 160
volume gas was contained in 1 volume hydrate. It can be found that
both of r and R values were lower when the backpressure was set to
2.8 MPa. The actual MH dissociation rate is shown in Table 1. The
results indicated that the actual dissociation rate was lower than
that of the other cycles. The studied 2.8 MPa backpressure is close
to the 2.98 MPa equilibrium pressure of MH at 274.45 K, i.e., at

2.8 MPa, the driving force was relatively low for MH dissociation.
The highest MH dissociation and gas production rate was observed
at 2.2 MPa for Case 1. The MH water saturation level was low for
Case 14, and the final D MI was lower than those of the other backpressure tests. Large fluctuations were observed in the D MI curve
for Case 13. A peak was observed at approximately 30 min. This
may have been due to rapid MH dissociation that resulted in water
aggregation during the first stage. Water movement out of the FOV
caused the D MI decrease. After additional slight fluctuations, the D
MI finally stabilized. As shown in Table 1, the water saturation
levels after MH dissociation (Swe) slightly decreased, which indicated the movement of water out of the FOV. A constant MI after
approximately 100 min indicated that the MH dissociation was
completed. As shown in Table 1, low backpressures resulted in rapid
MH dissociation. Slightly pressure fluctuations during MH dissociation indicated unstable water and methane gas flow in the porous
media, which may pose a problem for MH recovery.
Fig. 2 shows the liquid water distribution in the porous media
during MH dissociation. The red1 and dark blue areas indicate high
and low water content, respectively. Six images were selected from
each case were analyzed for water distribution at different times
during the MH dissociation process. As indicated by the color shift
from dark blue to red, the water content increased over time. At
10 min, the water content in Case 12, Case 13 and especially Case
14 increased, which implied that some of the MH deposits dissoci-

1
For interpretation of color in Fig. 2, the reader is referred to the web version of
this article.

183

M. Yang et al. / Applied Energy 187 (2017) 180188

Case 1-1

0 min

10 min

20 min

30 min

40 min

100 min

0 min

10 min

20 min

30 min

40 min

100 min

0 min

6 min

10 min

20min

40 min

100 min

0 min

8 min

10 min

20 min

40 min

100 min

Case 1-2

Case 1-3

Case 1-4

Fig. 2. Water distribution during MH dissociation process for Case 1.

ated. The water distribution changed between 10 and 20 min; most


of the MH deposits had dissociated in Case 12 and Case 13 and had
nearly completely dissociated in Case 14. As shown in Fig. 2, rapid
MH dissociation in Case 12 and Case 13 occurred under 30 min.
Furthermore, water movement was observed at 40 min, as indicated
by a few dark blue areas. This was similarly observed in Case 14
from 20 to 30 min. These observations confirmed the results shown
in Fig. 1. A rapid MI increase occurred from 30 to 40 min in Case 11,
which was later than those of the other cases; the MH dissociation
was completed at approximately 100 min. The MHs seemed to dissociate firstly near the outer core region for all cases, indicating that
backpressure had no effects on the MH dissociation mode in our
experiments. The MH deposits dissociated in a radial manner due
to heat transfer dynamics, as was reported in previous studies
[46,47]. We examined MH decomposition for different backpressures (2.8 MPa, 2.6 MPa, 2.4 MPa and 2.2 MPa) under excess gas conditions. The collected MRI data were used to analyze the influence of
different backpressures on MH decomposition characteristics. Under
excess gas conditions, the average rate of MH decomposition and gas
production was significantly affected by the different backpressures.
Larger pressure drops accelerated decomposition.
3.2. Typical MH dissociation characteristics for Case 2
The dissociation characteristics of MH deposits under excess
water conditions (41.950%) were studied by depressurizing to
2.6 MPa and 2.4 MPa. Both backpressure conditions were examined for two different MH saturation levels. Fig. 3 shows the
change in MI (D MI) and pressure curves for these cases. D MI is
the difference between the MI value when pressure decreased to
the set value and the MI value in real-time. For Case 22a, the
observed increasing D MI indicated MH dissociation. A fluctuation
at approximately 80 min indicated water movement in the vessel.

The pressure curve also showed similar fluctuations. This could be


due to the presence of a significant amount of liquid water in the
porous media prior to MH dissociation. The additional liquid water
and methane gas produced during MH dissociation resulted in
water movement. The dissociation was considered completed
when the D MI reached a steady value. The D MI and pressure
curves also showed fluctuations for Case 22b. These fluctuations
disappeared by approximately 70 min.
To analyze the MH dissociation process, 8 representative
images of liquid water distribution are shown for Case 22a and
Case 22b in Fig. 4. The lighter areas in the figure indicate the presence of liquid water. Water was added to the vessel after MH
deposit formation; correspondingly, the image at 0 min is lighter.
Liquid flow channels through the deposits are also present at
0 min. For both cases, these flow channels were primarily distributed nearer the vessel wall. After depressurization at 2 min,
no noticeable changes were detected. Ten minutes later, the
amount of liquid water significantly increased near the vessel wall
in Case 22a. Liquid water channels radially extended toward the
center of the vessel at 30 min. This indicated that the MH dissociation occurred in an inwardly radial direction. This observed pattern was due to two factors. First, the dissociation process was
primarily affected by heat transfer, i.e., the endothermic MH dissociation resulted in decreased temperatures at the core and thereby,
hindered local MH dissociation. Because the vessel wall was maintained at a constant temperature, the MH deposits nearer the wall
first dissociated. Second, because the specific heat capacity of liquid water is relatively large, the endothermic decomposition process by itself would not be capable of causing a temperature
decrease in the vessel. Thus the MH deposits dissociate easily at
the area with higher water saturation levels, as was the case at
48 min. At 52 min, the water signal slightly decreased in the lower
right corner of the image, which indicated liquid water outflow. For

184

M. Yang et al. / Applied Energy 187 (2017) 180188

Fig. 3. D MI and vessel pressure variation during MH dissociation for Cases 22a (a), 22b (b), 23a (c) and 23b (d).

Case 2-2a
0 min

2 min

10 min

30 min

48 min

52 min

60 min

100 min

0 min

2 min

14 min

20 min

40min

44min

60 min

100 min

0 min

2 min

8 min

10 min

24 min

40min

60 min

100 min

0 min

2 min

4 min

10 min

20 min

24 min

40min

100 min

Case 2-2b

Case 2-3a

Case 2-3b

Fig. 4. Water distribution in FOV during MH dissociation for Cases 22a and 22b, 23a (c) and 23b (d).

Case 22b, when comparing the images at 2 min and 14 min, MH


dissociation initiated near the middle upper position of the vessel
and from the vessel wall to the middle, i.e., a similar dissociation
pattern as that of Case 22a.
The DMI and pressure curves for MH dissociation at 2.4 MPa
backpressure are shown in Fig. 3(c) and (d). The DMI curve of Case
23a rapidly increased after dissociation, which indicated a signif-

icant amount of liquid water produced. Fluctuations were then


observed, and the value of DMI briefly decreased before continuing
to increase. A similar trend was also observed for Case 23b. The
pressure curve shown in Fig. 3(c) revealed fluctuations at the
beginning of MH dissociation. The pressure was slightly higher
than the experimentally set backpressure, i.e., the MH deposit
rapidly dissociated. In Fig. 3(d), the D MI and pressure curves

M. Yang et al. / Applied Energy 187 (2017) 180188

showed initial fluctuations, which indicated that the mobility of


gas and water were large. At this point, the MH deposit likely dissociated at multiple points. Near the end of the decomposition, the
pressure gradually reduced to the set backpressure. The liquid
water distribution profiles for Case 23a and Case 23b are shown
in Fig. 4. After MH dissociation for Case 23a, the water distribution changed at 2 min, 8 min, 10 min, 40 min and 60 min according
to the data in Fig. 3. Additionally, due to a significant amount of liquid water gathering at the vessel wall, the MH deposit dissociated
radially inward from the wall to the center of the vessel, as clearly
shown in the images at 2, 8, 40 and 60 min. A similar phenomenon
was observed at 2, 4, 10, 20 and 24 min for Case 23b. The MH dissociation characteristics for Case 22a and Case 23a were similar.
Under excess water conditions, some liquid water flowed out of the
vessel during MH dissociation. Furthermore, the liquid water had
high mobility, which may be related to the diffusion of methane
gas during MH dissociation. The higher water saturation levels also
hindered CH4 gas output.
Fig. 5 shows the D MI curves during MH dissociation under different backpressures. Because the liquid water saturation level was
high before MH dissociation, the movement of liquid water during
MH dissociation was expected. Most of the water flowed out of the
FOV. Thus, the instantaneity hydrate dissociation rate and amount
cannot be measured in this cases. D MI can be reasonably used to
schematic represent MH dissociation process. Assuming no
hydrate formation in the water injection process, we calculated
the average rate of hydrate decomposition and gas production
(Table 1). Because the initial MH saturation levels in Case 21, Case
22a, Case 23a and Case 24 were nearly constant, we used these
experiments to compare the MH dissociation characteristics under
different backpressures. In Fig. 5, for all cases except Case 21, the
D MI gradually increased as the MH deposits dissociated. In Case
21, the MH deposit dissociated during an initial period; however,
this dissociation rate rapidly decreased (10 min later) with two
large fluctuations. The DMI remained constant for the next
40 min. The more rapid dissociation for Case 21 at initial time
was attributed to the higher liquid water saturation levels (50%)
relative to the other cases for excess water condition, where the
high specific heat capacity of water occupied a dominant. Additionally, as shown in Fig. 5, at the end of MH dissociation, D MI
in Case 22a was higher than those of others. Furthermore, as
shown in Table 1, the initial MH saturation in Case 22a was higher
than those of the other cases. For Case 23a and Case 24, many
fluctuations were observed in the initial period, which indicated
a more rapid MH dissociation. In Case 24, MH dissociation finished at 26 min, and the pressure curve (Fig. 3(b)) slightly changed
after 20 min. As shown in Table 1, the average MH dissociation and
gas production rate in Case 24 was the highest due to the greatest
pressure drop between the sample pressure and set backpressure.
When compared with pressure drops of the other cases, the larger
depressurization range likely enhanced the average MH dissociation and gas production rate. Finally, we found that the water saturation level also affected MH dissociation.
Fig. 6 shows the liquid water distribution under different backpressures. Eight images at different times for each MH dissociation
case were selected. The image at 0 min is the liquid water distribution after water injection. Flow channels near the vessel wall were
clearly observed; the large amount of water will affect the MH dissociation pattern. In Case 22a and Case 23a, the MH deposits
began to dissociate near positions with high levels of liquid water.
As the MH deposits dissociated, liquid water levels increased
toward the center of the vessel. Because liquid water has a large
specific heat capacity, the MH deposits rapidly dissociated. In Case
21 and Case 24, large amounts of liquid water gathered on the
right side of vessel wall. The MH deposits correspondingly dissociated near the right side of the wall at 10 min. On the left side of the

185

Fig. 5. D MI (a) and pressure variations (b) during MH dissociation process for
Case 2.

vessel wall, the change of images from 0 min to 10 min also shown
the presence of hydrate dissociation. This indicated that in addition
to the influence of liquid water, the temperature of the cooling
jacket also affected MH dissociation. Generally, the cooling jacket
temperature kept nearly constant, and it could counteract the
endothermic hydrate dissociation. As shown in Fig. 6, the water
increase in FOV during hydrate dissociation process is small, and
it is obvious lower than that for excess gas conditions. This means
that the hydrate dissociation water move out of the FOV, and the
larger liquid water mobility during MH dissociation hindered
methane gas output.
3.3. MH dissociation characteristics in deposits with different water
saturation levels
We compared the results of this work with those of our previous study [44] and conducted a comprehensive comparative analysis of the MH dissociation characteristics under different liquid
water saturation levels. In this study, MH dissociation under excess
gas conditions (with liquid water saturation levels between 12.9%
and 17.5%) indicated that the depressurization process was carried
out after MH formation. For low liquid water saturation level
deposits, liquid water was injected into the MH deposits before
the depressurization process at constant pressure. Before the MH
dissociation, the liquid water saturation levels were set at 22.6
27.4% [44]. MH dissociation under excess water conditions (with
liquid water saturation levels between 41.9% and 50.0%) was based
on low liquid water saturation levels. The liquid saturation levels
within the porous media further increased after high-pressure
water displacement prior to depressurization. The vessel pressure
was approximately 5.0 MPa before MH dissociation. The average
rate of MH dissociation remained consistent under different backpressures. This may be due to the following factors: initial hydrate

186

M. Yang et al. / Applied Energy 187 (2017) 180188

Case 2-1
No water inj.

0 min

10 min

20 min

30 min

40 min

60 min

120 min

No water inj.

0 min

10 min

20 min

30 min

40 min

60 min

120 min

No water inj.

0 min

10 min

20 min

30 min

40 min

60 min

120 min

No water inj.

0 min

10 min

20 min

30 min

40 min

60 min

120 min

Case 2-2a

Case 2-3a

Case 2-4

Fig. 6. Water distribution changes during MH dissociation under different back pressures for Cases 21, 22a, 23a, and 24 (the first image indicates the water distribution
before water injection for each cases).

saturation levels, pressures before and after MH dissociation, and


liquid water levels in the porous media before MH dissociation.
As shown in Fig. 7, the MH dissociation characteristics were
compared between the excess gas condition (left figure) and low
liquid water saturation levels (right figure). The experimental
detail for low liquid water saturation was shown in our previous
work [44]. The vertical axis R, T and Swe denote the average rate
of MH dissociation and gas production, the MH dissociation time
and the residual water saturation level in the porous media after
dissociation, respectively. The different backpressure values are
on the horizontal axis. Above each cluster in the histogram, the liquid water saturation level prior to MH dissociation is marked. As
can be seen, in the excess gas condition, the average rate of MH dissociation and gas production gradually increased as the backpressure decreased, which indicates that a large pressure drop would
accelerate MH dissociation. For MH dissociation at low liquid
water saturation levels under 2.8 MPa, the dissociation rate was
slightly lower than those of the other three cases. This indicates
that a large pressure drop would contribute to a higher dissociation
rate. However, under low liquid water saturation conditions, the
average rates of dissociation of the other three cases (at 2.6, 2.4
and 2.2 MPa backpressures) were similar to each other. This
implies that the pressure drop has a significant impact on the rate
of MH dissociation and gas production after a certain threshold
value. As is also seen from the histogram, the average rate of MH
dissociation and gas production under excess gas conditions was
higher than that under low liquid water saturation conditions; furthermore, the dissociation time was shorter. This shows that the
increase in liquid water saturation levels decreased the rate of
MH dissociation. During the MH dissociation under low liquid
water saturation levels, the gas output will be hindered by the liquid water and will delay the MH dissociation. Thus, liquid water
saturation is an important parameter that affects MH dissociation.
As seen in Fig. 7, Swe under low liquid saturation levels is higher
than that under excess gas conditions.
Combined with the experimental data in Fig. 8, we analyzed MH
dissociation under excess gas and excess water conditions. The
average rate of MH dissociation and gas production for the latter

Fig. 7. Comparison of experimental results for MH depressurization dissociation


between excess gas and low water saturation deposits.

Fig. 8. Comparison of experimental results for MH depressurization dissociation


between excess gas and excess water deposits.

condition under 2.2 MPa was much higher than those under the
other three backpressures, i.e., a large pressure drop accelerated
hydrate dissociation. However, before the MH dissociation, the
actual pressure inside the vessel under excess water conditions
was higher than that under excess gas conditions, resulting in a

M. Yang et al. / Applied Energy 187 (2017) 180188

large drop in the MH dissociation process. This indicated that the


average MH dissociation rate under excess gas conditions
remained higher than that under excess water. Furthermore, the
dissociation rate was slightly higher, which further indicated that
the increased liquid water saturation caused the slow MH dissociation rate. This also shows that a high liquid water saturation level
is an important factor in inhibiting MH dissociation. Under excess
water conditions, the saturation level of the liquid water retained
within the porous media was slightly higher after MH dissociation
because the liquid water saturation level before the MH dissociation was already high.
Fig. 9 compares the MH dissociation characteristics under low
liquid saturation levels and excess water conditions. Rl, Swel and
Tl are the average rate of MH dissociation and gas production at
low liquid saturation levels, the liquid water saturation level after
dissociation and the dissociation time, respectively. Rh, Sweh and Th
are the corresponding parameters under excess water conditions.
The average dissociation rate increased when the backpressure
decreased under these two conditions, i.e., a greater pressure drop
promoted rapid MH dissociation. Rh of the excess water condition
was slightly higher than Rl of the low liquid water saturation levels,
especially at a backpressure of 2.2 MPa. For the simulated deposits
with low liquid water levels, to maintain a constant vessel pressure, the actual pressure during the water injection process and
before water injection was kept very small. For the simulated
deposits with excess water, we used high pressure water injection,
and the water saturation level increase also increased the vessel
pressure. For the depressurization process under excess water conditions, the pressure drop was much greater than the pressure drop
under low liquid saturation levels, leading to a rapid MH dissociation and gas production rate. The pressure drop is an important
parameter in MH dissociation by depressurization. For low liquid
saturation levels and excess water conditions, the pressure drop
significantly influenced MH dissociation. Regarding the dissociation time, the duration under the low liquid water saturation levels
was higher than that under the excess water condition. For Swel and
Sweh, after MH dissociation, both parameters decreased with
decreasing backpressure. A larger pressure drop would result in
more vessel liquid outflow. Sweh was slightly higher than Swel due
to the excess water in the MH deposits. The liquid water saturation
level in porous media was higher before MH dissociation.
Total liquid water outflows from the vessel at different initial
saturation levels and backpressures were calculated, as shown in
Fig. 10. The largest amount of outflow liquid water was observed
under excess water (HWS) condition. The lowest amount of liquid
water outflow was observed under excess gas (HGS) conditions.

Fig. 9. Comparison of MH dissociation for variable backpressures between low


water saturation and excess water conditions.

187

Fig. 10. The amount of outflow water during MH dissociation under 2.8, 2.6, 2.4 and
2.2 MPa for variable gas and water saturation deposits.

This indicated that the former condition had high liquid water
mobility during MH dissociation, partly due to the actual pressure
inside the vessel being larger than the set backpressure. Otherwise,
because the liquid water saturation level was high, the gas produced by MH dissociation would carry liquid water out of the vessel, i.e., methane gas output would be hampered by the liquid
water. During the excess gas MH dissociation process, some of
the liquid water flowed from the core tube. However, the amount
was very small, and most of the liquid water remained in the core
tube. For MH dissociation under low liquid saturation (LWS) conditions, because the original pressure in the vessel was kept constant
during the water injection process, the liquid water saturation
level before dissociation was slightly higher than that under the
excess gas conditions. Therefore, the amount of water flowing
out during the MH dissociation was not high, and the remaining
liquid in the core tube hindered gas output, and affected hydrate
dissociation to some extent.
4. Conclusion
MH dissociation via depressurization was investigated for MH
deposits with different water saturation levels. MRI images and
pressure measurements were obtained to analyze MH dissociation
characteristics at backpressures of 2.8 MPa, 2.6 MPa, 2.4 MPa and
2.2 MPa. The following conclusions were obtained.
(1) The dissociation patterns did not change for the different
backpressures under excess gas conditions. Instead, the dissociation was dominated by heat transfer. Based on the differences in average MH dissociation and gas production
rates under variable backpressures, a larger depressurization
range accelerated MH dissociation.
(2) For excess water deposit conditions, areas with more liquid
water undergo rapid MH dissociation similar to the pattern
observed under variable backpressures. Compared with the
other two types of deposits, the mobility of liquid water
for excess water deposit conditions was large; therefore,
methane gas output will be hindered. A larger depressurization range enhanced the average rate of MH dissociation and
gas production.
(3) By comparing the excess gas and low water saturation conditions, liquid water saturation levels significantly affected
MH dissociation within the same depressurization range.
The average dissociation rate decreased with increasing liquid water saturation levels. As liquid water saturation
increased from low levels (22.627.4%) to excess water conditions (43.950.0%) and as the depressurization range
shifted from 0.341.16 MPa to 2.213.33 MPa, the depres-

188

M. Yang et al. / Applied Energy 187 (2017) 180188

surization range was the dominant parameter affecting MH


dissociation. When the liquid water saturation level
increased from the level under excess gas conditions
(12.917.5%) to the level under excess water condition
(43.950.0%) and as the depressurization range shifted from
0.551.41 MPa to 2.213.33 MPa, the liquid water saturation
level was the dominant parameter.
(4) The liquid water saturation level and the depressurization
range are two key factors that influence MH dissociation.
An increasing liquid water saturation level reduces the average rate of MH dissociation and gas production, and a larger
depressurization range can enhance the average dissociation
rate. These two parameters have different effects on the dissociation rate. Further research would be required to identify these differences.

Acknowledgements
This project was financially supported by the National Natural
Science Foundation of China (51576025, 51436003, and
51227005), the Program for Liaoning Excellent Talents in University (LJQ2014007), the Natural Science Foundation of Liaoning Province of China (2015020628) and the Fundamental Research Funds
for the Central Universities of China.
References
[1] Sloan ED, Koh CA. Clathrate hydrates of natural gases. CRC PressI Llc; 2008.
[2] Englezos P. Clathrate hydrates. Ind Eng Chem Res 1993;32:125174.
[3] Lu S-M. A global survey of gas hydrate development and reserves: specifically
in the marine field. Renew Sustain Energy Rev 2015;41:884900.
[4] Li B, Li X-S, Li G, Chen Z-Y. Evaluation of gas production from Qilian Mountain
permafrost hydrate deposits in two-spot horizontal well system. Cold Reg Sci
Technol 2015;109:8798.
[5] Matsumoto R, Ryu B-J, Lee S-R, Lin S, Wu S, Sain K, et al. Occurrence and
exploration of gas hydrate in the marginal seas and continental margin of the
Asia and Oceania region. Mar Pet Geol 2011;28:175167.
[6] Chong ZR, Yang SHB, Babu P, Linga P, Li X-S. Review of natural gas hydrates as
an energy resource: prospects and challenges. Appl Energy 2015;162:163352.
[7] Zhao J, Zhu Z, Song Y, Liu W, Zhang Y, Wang D. Analyzing the process of gas
production for natural gas hydrate using depressurization. Appl Energy
2015;142:12534.
[8] Liu C, Meng Q, He X, Li C, Ye Y, Lu Z, et al. Comparison of the characteristics for
natural gas hydrate recovered from marine and terrestrial areas in China. J
Geochem Explor 2015;152:6774.
[9] Li G, Li X-S, Li B, Wang Y. Methane hydrate dissociation using inverted fivespot water flooding method in cubic hydrate simulator. Energy
2014;64:298306.
[10] Linga P, Haligva C, Nam SC, Ripmeester JA, Englezos P. Recovery of methane
from hydrate formed in a variable volume bed of silica sand particles. Energy
Fuels 2009;23:550816.
[11] Sloan Jr ED. Fundamental principles and applications of natural gas hydrates.
Nature 2003;426:3539.
[12] Moridis GJ. Toward production from gas hydrates: current status, assessment
of resources, and simulation-based evaluation of technology and
potential. Lawrence Berkeley National Laboratory; 2008.
[13] Falser S, Uchida S, Palmer AC, Soga K, Tan TS. Increased gas production from
hydrates by combining depressurization with heating of the wellbore. Energy
Fuels 2012;26:625967.
[14] Yang X, Sun C, Su K, Yuan Q, Li Q, Chen G. A three-dimensional study on the
formation and dissociation of methane hydrate in porous sediment by
depressurization. Energy Convers Manage 2012;56:17.
[15] Wang S, Yang M, Wang P, Zhao Y, Song Y. In situ observation of methane
hydrate dissociation under different backpressures. Energy Fuels
2015;29:32516.
[16] Tang L-G, Li X-S, Feng Z-P, Li G, Fan S-S. Control mechanisms for gas hydrate
production by depressurization in different scale hydrate reservoirs. Energy
Fuels 2006;21:22733.
[17] Haligva C, Linga P, Ripmeester JA, Englezos P. Recovery of methane from a
variable-volume bed of silica sand/hydrate by depressurization. Energy Fuels
2010;24:294755.
[18] Loh M, Too JL, Falser S, Linga P, Khoo BC, Palmer A. Gas production from
methane hydrates in a dual wellbore system. Energy Fuels 2014;29:3542.

[19] Chejara A, Kvamme B, Vafaei MT, Jemai K. Simulations of long term methane
hydrate dissociation by pressure reduction using an extended
RetrasoCodeBright simulator. Energy Convers Manage 2013;68:31323.
[20] Li X, Yang B, Zhang Y, Li G, Duan L, Wang Y, et al. Experimental investigation
into gas production from methane hydrate in sediment by depressurization in
a novel pilot-scale hydrate simulator. Appl Energy 2012;93:72232.
[21] Vinay Kumar BG, Singh AP, Ganguly S. Effect of heat diffusion in the burden on
the dissociation of methane in a hydrate bearing formation. J Nat Gas Sci Eng
2014;16:706.
[22] Ahmadi G, Ji C, Smith D. A simple model for natural gas production from
hydrate decomposition. Ann NY Acad Sci 2000;912:4207.
[23] Li B, Li X-S, Li G, Feng J-C, Wang Y. Depressurization induced gas production
from hydrate deposits with low gas saturation in a pilot-scale hydrate
simulator. Appl Energy 2014;129:27486.
[24] Reagan M, Moridis G, Johnson J, Pan L, Freeman C, Pan L, et al. Field-scale
simulation of production from oceanic gas hydrate deposits. Transp Porous
Media 2015;108:15169.
[25] Wang Y, Li X, Li G, Huang N, Feng J. Experimental study on the hydrate
dissociation in porous media by five-spot thermal huff and puff method. Fuel
2014;117(Part A):68896.
[26] Wang Y, Li X, Li G, Zhang Y, Li B, Feng J. A three-dimensional study on methane
hydrate decomposition with different methods using five-spot well. Appl
Energy 2013;112:8392.
[27] Feng J, Wang Y, Li X, Li G, Zhang Y, Chen Z. Effect of horizontal and vertical well
patterns on methane hydrate dissociation behaviors in pilot-scale hydrate
simulator. Appl Energy 2015;145:6979.
[28] Wang X, Sun C, Chen G-J, He Y, Sun Y, Wang Y, et al. Influence of gas sweep on
methane recovery from hydrate-bearing sediments. Chem Eng Sci
2015;134:72736.
[29] Veluswamy HP, Kumar R, Linga P. Hydrogen storage in clathrate hydrates:
current state of the art and future directions. Appl Energy 2014;122:11232.
[30] Zhang Y, Li X, Chen Z, Ruan X, Huang N. Methane hydrate dissociation by
depressurization in sediments with different hydrate saturations in cubic
hydrate simulator. Energy Proc 2014;61:9904.
[31] Kumar A, Maini B, Bishnoi P, Clarke M, Zatsepina O, Srinivasan S. Experimental
determination of permeability in the presence of hydrates and its effect on the
dissociation characteristics of gas hydrates in porous media. J Petrol Sci Eng
2010;70:11422.
[32] Ji C, Ahmadi G, Smith DH. Natural gas production from hydrate decomposition
by depressurization. Chem Eng Sci 2001;56:580114.
[33] Nazridoust K, Ahmadi G. Computational modeling of methane hydrate
dissociation in a sandstone core. Chem Eng Sci 2007;62:615577.
[34] Li G, Moridis GJ, Zhang K, Li X-S. Evaluation of gas production potential from
marine gas hydrate deposits in Shenhu Area of South China Sea. Energy Fuels
2010;24:601833.
[35] Moridis GJ, Kowalsky M. Depressurization-induced gas production from class
1 and class 2 hydrate deposits. Lawrence Berkeley National Laboratory; 2006.
[36] Moridis GJ, Kowalsky MB, Pruess K. Depressurization-induced gas production
from class-1 hydrate deposits. SPE Reservoir Eval Eng 2007;10:45881.
[37] Moridis G, Collett T. Strategies for gas production from hydrate accumulations
under various geologic conditions. Lawrence Berkeley National Laboratory;
2003.
[38] Fujii T, Saeki T, Kobayashi T, Inamori T, Hayashi M, Takano O, et al. Resource
assessment of methane hydrate in the Eastern Nankai Trough, Japan. In:
Offshore technology conference; 2008.
[39] Moridis GJ, Sloan ED. Gas production potential of disperse low-saturation
hydrate accumulations in oceanic sediments. Energy Convers Manage
2007;48:183449.
[40] Konno Y, Masuda Y, Hariguchi Y, Kurihara M, Ouchi H. Key factors for
depressurization-induced gas production from oceanic methane hydrates.
Energy Fuels 2010;24:173644.
[41] Song Y, Wang S, Yang M, Liu W, Zhao J, Wang S. MRI measurements of CO2-CH4
hydrate formation and dissociation in porous media. Fuel 2015;140:12635.
[42] Johnson A, Patil S, Dandekar A. Experimental investigation of gas-water
relative permeability for gas-hydrate-bearing sediments from the Mount
Elbert gas hydrate stratigraphic test well, Alaska North Slope. Mar Pet Geol
2011;28:41926.
[43] Sloan Jr ED, Koh C. Clathrate hydrates of natural gases. CRC Press; 2007.
[44] Yang M, Fu Z, Zhao Y, Jiang L, Zhao J, Song Y. Effect of depressurization pressure
on methane recovery from hydrategaswater bearing sediments. Fuel
2016;166:41926.
[45] Sloan Jr ED. Clathrate hydrates of natural gases. CRC Press; 1998 [Revised and
expanded].
[46] Kneafsey TJ, Moridis GJ. X-Ray computed tomography examination and
comparison of gas hydrate dissociation in NGHP-01 expedition (India) and
Mount Elbert (Alaska) sediment cores: experimental observations and
numerical modeling. Mar Pet Geol 2014;58:52639.
[47] Falser S, Uchida S, Palmer A, Soga K, Tan T. Increased gas production from
hydrates by combining depressurization with heating of the wellbore. Energy
Fuels 2012;26:625967.