Fuel 94 (2012) 486494

Contents lists available at SciVerse ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Experimental study on gas production from methane hydrate in porous media

by huff and puff method in Pilot-Scale Hydrate Simulator
Xiao-Sen Li , Bo Yang, Gang Li, Bo Li, Yu Zhang, Zhao-Yang Chen
Key Laboratory of Renewable Energy and Gas Hydrate, Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences, Guangzhou 510640, China

a r t i c l e i n f o a b s t r a c t

Article history: Pilot-Scale Hydrate Simulator (PHS), a novel three-dimensional 117.8-L pressure vessel was developed to
Received 23 May 2011 investigate the gas production from methane hydrate in porous media using the huff and puff method. In
Received in revised form 31 October 2011 situ methane hydrate was synthesized in the pressure vessel with methane gas and deionized water in
Accepted 7 November 2011
quartz sand with grain sizes between 300 and 450 lm. In the PHS, a 9-spot distribution of vertical wells,
Available online 18 November 2011
a single horizontal well and 49-spot distributions of the thermometers and resistance ports are respec-
tively placed in three horizontal layers, which equally divide the vessel into four parts. During the exper-
Keywords:
iment, the huff and puff stages, including the injection, the soaking and the production, are carried out for
Hydrate dissociation
Porous media
hydrate dissociation. A vertical well at the axis of the PHS was used as the injection and production well.
Three-dimension The initial hydrate saturation before dissociation is 33.66% in volume, and the percentage of the hydrate
Huff and puff dissociated could be approximately 94% after 15 huff and puff cycles. The experimental results indicate
Vertical well that with the constant hot water injection rate, the range of the thermal diffusion is restricted around the
well, and depressurization rather than thermal stimulation is dominant for gas production. The decline of
the cumulative gas produced during each cycle and the diminishing uptrend of the percentage of the
hydrate dissociated indicate that the hydrate dissociation rate decreases over time. The gas production
efciency is improved by prolonging the hot water injection time, while this enhancement is limited
by the stronger pressurization effect.
2011 Elsevier Ltd. All rights reserved.

1. Introduction pressuretemperature equilibrium conditions, and (4) a combina-

Natural gas hydrates (NGHs) are solid, non-stoichiometric com-
pounds formed by host water molecules with small guest mole-
cules, such as CH4, C2H6, C3H8, CO2, and H2S. [1]. Natural gas
hydrate deposits involve mainly CH4, and occur in the permafrost
and in deep ocean sediment, where the necessary conditions of
low temperature and high pressure exist for hydrate stability. Re-
cent seismic explorations and geological researches show that nat-
ural gas hydrates existing in the sediment constitute a large
natural gas reservoir that invites consideration as a potential stra-
tegic energy resource [24].
There are several methods for gas production from the hydrate
deposit: (1) depressurization [58], to decrease the deposit pres-
sure below the hydrate dissociation pressure at a specied temper-
ature; (2) thermal stimulation [912], involving raising the deposit
temperature above the dissociation temperature at a given pres-
sure of the hydrate with the hot water, hot brine or steam injec-
tion; (3) thermodynamic inhibitor effects [1315], based on the
use of chemicals, such as salts and alcohols to shift the hydrate
Corresponding author. Tel./fax: +86 20 87057037.
E-mail address: lixs@ms.giec.ac.cn (X.-S. Li).
tion of these methods [1618]. Of the above single methods for hy-
drate dissociation for gas production, the depressurization appears
to be the most efcient one and the thermal stimulation could be
the auxiliary method [4,17,1921].
Numerical studies have shown that thermal stimulation by hot
water, hot brine or steam injection enhances gas production when
used in conjunction with depressurization with the vertical well
[21,22], but tend to be ineffective when used as the main dissoci-
ation strategies [21]. The huff and puff method, also known as the
cyclic steam stimulation (CSS), was accidentally discovered by
Shell Oil Company in 1960 during a Venezuela recovery project,
and is widely used in the oil industry to enhance oil recovery
[2325]. The hot water, hot brine or steam huff and puff method
is a special form of the combination of depressurization and ther-
mal stimulation methods for gas production from the hydrate de-
posit. A numerical simulation of gas production from marine gas
hydrate deposit using the huff and puff method has been used to
investigate into the gas production strategy in the Shenhu Area
of the South China Sea [18].
In this work, the Pilot-Scale Hydrate Simulator (PHS), a novel
three-dimensional 117.8-L pressure vessel has been developed
for the production of the gas hydrate. The gas production behavior
of methane hydrate in the porous media using the huff and puff

0016-2361/$ - see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.fuel.2011.11.011
 X.-S. Li et al. / Fuel 94 (2012) 486494
Nomenclature
Minj cumulative volume of the injected water (L)
MW cumulative volume of the produced water (L)
P pressure (MPa)
Qavg average gas production rate (L/min)
Qinj hot water injection rate (L/min)
RGW cumulative gas to water production ratio (ST m3 of CH4/
m3 of H2O)
SA initial aqueous saturation, volume
method was investigated in the PHS. A vertical well at the axis of
the PHS is used as the injection and the production well.
2. Experimental
2.1. Apparatus
The schematic of the experimental apparatus used in this work is
shown in Fig. 1. The PHS, a novel 117.8-L pressure vessel, is made of
stainless steel. The PHS can withstand pressures of up to 30 MPa. The
low temperatures required for hydrate formation and dissociation
are achieved by placing the whole apparatus inside of a walk-in cold
room, and there is water jacket around the PHS. Fig. 2 shows the
schematic of the layers and the well design of the PHS. The PHS is
a cylindrical pressure vessel (0.60 m length, 0.50 m diameter) with
the inner volume of 117.8 L. As shown in Figs. 1 and 2, there are three
horizontal layers named Layers AA, BB and CC inside the vessel,
which equally divide the cubic vessel into four regions. The distance
between the Layer AA and Layer BB is 0.15 m, a quarter of the
internal length of the PHS, which is the same with that between
the Layer BB and Layer CC, while the Layer BB is in the middle
of the PHS. As shown in Fig. 2, a 9-spot distribution of the vertical
wells is xed in the top ange (the Top Surface) of the PHS, and there
are three vertical wells at each spot (V1, V2, . . . , V9), which extend
into the vessel to the Layers AA, BB and CC, respectively. As a typ-
ical example, it is shown in Fig. 2 that the Wells V5A, V5B and V5C are
Fig. 1. Schematic of the experimental apparatus.
487
SH initial hydrate saturation, volume
T temperature (C)
T time (min)
tinj duration of the hot water injection (min)
Tinj initial temperature of the hot water (C)
VP cumulative volume of the produced CH4 (L)
P driving force of depressurization (MPa)
all placed on the spot V5, and the bottoms of these wells are on the
Layers AA, BB and CC, respectively. The Wells V5A, V5B and V5C
are at the axis of the PHS. In general, a total of 27 vertical wells are
distributed in the PHS, and the wells on the spots V1, V3, V7 and
V9 are all immediately close to the inside edge of the PHS, and the
bottom of each well is right on the corresponding layer.
As shown in Fig. 2, in the Right Surface of the PHS, 3 horizontal
wells, the Wells HA, HB and HC, are inserted into the spots H1, H2
and H3 on the Layers AA, BB and CC, respectively. Each hori-
zontal well is extended to the inside surface of the Left Surface of
the PHS. Fig. 3 shows the schematic of the distribution of the ther-
mometers (the temperature measuring spots) in the PHS. There are
49 thermometers evenly distributed on each layer, with a total of
147 spots in the PHS. In other words, on each layer (Layers AA,
BB and CC), it is a 49-spot distribution of the thermometers
(T1T49), with T25 at the center and T1, T7, T43 and T49 at the cor-
ner. The thermometers at the same spots are distinguished by the
different layers, for example, as shown in Fig. 3, the 43st thermom-
eter on Layer AA is called T43A, and those on Layer BB and Layer
CC are T43B and T43C, respectively. The distribution of the resis-
tance ports is the same with that of the thermometers, with the
corresponding name of R1R49. As shown in Fig. 3, the 43st
resistance port on Layer AA is called R43A, and those on Layer
BB and Layer CC are R43B and R43C, respectively.
A pressure transducer at the center of the bottom of the PHS
(the Inlet Pressure) is used to measure the pressure of the base
488 X.-S. Li et al. / Fuel 94 (2012) 486494

Fig. 2. Schematic of the layers and the well design of the PHS.

Fig. 3. Schematic of the distributions of the thermometers and the resistance ports in the PHS.

of the Hydrate-Bearing Layer (HBL). And another pressure trans- The thermometers are Pt100 with the range of 20 to 200 C,
ducer (the Outlet Pressure) is placed at the production well. A 0.1 C. The pressure transducers are TRAGAG NAT 8251.84.2517,
sampling port is placed at on each layer. A safety valve and a vac- 040 MPa, 0.1%. Two gas ow meters, which are used to measure
uum pump were connected to the PHS. the cumulative gas injected into the PHS, the gas production rate
 X.-S. Li et al. / Fuel 94 (2012) 486494 489

and the cumulative gas produced from the vessel, are both of D07- at the mouth of the well decreases to a certain value Pmim, this
11CM, 010 L/min, 2% from Seven Star Company. The thermom- stage ceases and the injection stage of the next cycle begins.
eters, pressure transducers, gas ow meters, were calibrated using a
mercury thermometer with the tolerance of 0.01 C, a pressure
2.3. Procedure
test gauge with the error of 0.05%, and a wet gas meter with the
accuracy of 10 mL/min, respectively. Resistance is measured by
During the experiment, the raw dry quartz sand with a size
TH2810B resistivity meter from Tonghui Company. A metering
range between 300 and 450 lm was tightly packed in the vessel,
pump Beijing Chuangxintongheng HPLC P6000 with the range
and then the vessel was evacuated twice to remove air in it with
of 250 mL/min can withstand pressures of up to 15 MPa. An inlet li-
a vacuum pump. The quartz sand in the vessel was wetted to satu-
quid container with the inner volume of 10 L is used to contain the
ration with distilled water at atmospheric pressure using a meter-
deionized water used in the experiments. In order to protect the
ing pump. The sand sediment as the porous media was
metering pump from corrosion by the hot brine or the chemicals,
considered as saturation when the amount of water produced from
3 middle containers are used for the solution injection, and the in-
the vessel was equal to the amount of water injected. Using the
ner volume of each container is 4 L. A back-pressure regulator (the
fugacity model of Li et al. [26], the equilibrium hydrate dissociation
pressure range of 030 MPa, 0.02 MPa) connected to the outlet of
pressures in the sediment at the working temperature of 7.3 C
the PHS is used to control the pressure of the production well. A gas
were calculated, and the calculated value is 5.3 MPa. Then the
cylinder is used to provide the driving force of the back-pressure
methane gas was injected into the vessel until the pressure in the
regulator. A balance, used to measure the mass of liquid produced
vessel reached approximately 20 MPa, which was much higher than
from the PHS, is Sartorius BS 2202S, 02200 g, 0.01 g. The data
the equilibrium hydrate dissociation pressure at the working tem-
acquisition system records the temperature, the pressure, the
perature. After that, the vessel was closed as an isochoric system.
amount of the cumulative gas produced from the vessel, the gas
The temperature was gradually decreased to form the hydrate by
production rate and the liquid production rate. The liquid injection
changing the water bath temperature. The hydrate formation pro-
rate can be controlled by the metering pump, which is connected to
cess lasted for 3045 days, and then the pressure in the vessel de-
the data acquisition system. In this work, the methane gas with its
creased to 9.0 MPa.
purity of 99.99% is used.
The following procedure was used to investigate into the disso-
ciation characteristics of methane hydrate using the huff and puff
method. Firstly, the back-pressure regulator was used to release
2.2. Well design
the gas from the vessel and the pressure of the production well
(Well V5) gradually decreased to 5.4 MPa, which is slightly higher
Fig. 2 shows the design of the vertical and horizontal wells in the
than the equilibrium pressure at the working temperature. Sec-
PHS. As shown in Fig. 2, there are four grooves evenly distributed
ondly, the vessel was closed for approximately 30 min, and the
along the circumference of each vertical well, thus as Well V5A, Well
pressure and the temperature in the vessel maintained stable dur-
V5B and Well V5C. The grooves on Well V5A extended from the
ing this period, which indicated that there was little hydrate disso-
internal surface of the Top Surface of the PHS to the Layer AA, while
ciated. Then, the huff and puff cycles including the injection, the
those on Well V5B is from the Layer AA to the Layer
soaking and the production stages were carried out for hydrate dis-
BB, and the top and the bottom of the grooves on Well V5C are
sociation. The injection step begins immediately at the end of the
the Layer BB and the Layer CC, respectively. As shown in Fig. 2,
production step of the former cycle except the rst cycle. In this
the grooves on each horizontal well are from the Right Surface to
work, the hot water injected wells and the produced wells of gas
that of the Left Surface. The hot water or steam injection into the
and water are same. At rst, the hot water is injected into the ver-
PHS and the gas and water production from the PHS are both
tical central wells of V5A, V5B and V5C, for a period of time to heat
through the grooves shown in Fig. 2. Previous numerical study has
the hydrate reservoir around the wells. The initial temperature of
shown a similar well design [18], in which there are 8 grooves evenly
the hot water from the heater Tinj, the injection rate Qinj and the
distributed along the circumference of the single horizontal well.
duration of the hot water injection tinj maintain 160 C, 200
Comparing with using the single well with hot brine injection and
mL/min and 20 min, respectively. Then, the soaking step begins.
gas production simultaneously [19] and the well with hot water cir-
In this step, the wells are closed for 30 min to allow the heat to
culates inside the wellbore [16,17], this well design is much simpler
soak into the reservoir. At last, the gas and water in the vessel is
and practically feasible. Vertical wells at the axis of the PHS (Well
produced from the wells of V5A, V5B and V5C. In this step, the
V5A, Well V5B and Well V5C) are used simultaneously as Well V5,
back-pressure regulator pressure is set to 4.7 MPa and the pressure
which is the injection and production well during gas production
in the vessel is depressurized to the back-pressure regulator pres-
using the huff and puff method in this work. In other words, the in-
sure. Once the pressure in the system decreases to the back-pres-
jected hot water and the produced gas and water from the vessel
sure regulator pressure, next cycle of injection, soaking and
ow in the same well at the injection and the production stages.
production begins. The process is repeated until the pressure stops
The typical huff and puff method, also known as the cyclic
rising in the soaking step. Finally, after more than 15 cycles of the
steam stimulation (CSS), consists of 3 stages: injection, soaking
huff and puff process, the system was closed for more than 2 h and
and production. Firstly, hot water, hot brine or steam is injected
then the system pressure was released to 1 atm gradually.
into a well for some time to heat the hydrate deposit around the
During the hydrate dissociation, the temperatures and pressure
well. Then, the well is allowed to sit for a period to allow heat soaks
in the vessel, the gas production rate, the water injection and pro-
into the deposit. Finally, the uid is pumped out of the well for
duction rates were recorded at 20 s intervals.
some time. Once the production rate falls off, the well is put
through another cycle of injection, soaking and production. This
process is repeated until the cost of injecting water becomes higher 3. Results and discussion
than the money made from producing gas.
In this study, the hot water injection rate and the duration of 3.1. Dissociation conditions and process
the hot water injection is constant during the injection stage, and
reasonable driving force of depressurization is used during the pro- Fig. 4 shows the inlet and outlet pressures and average temper-
duction stage. During the production stage, as soon as the pressure ature of 147 thermometers in the vessel during hydrate dissociation
490 X.-S. Li et al. / Fuel 94 (2012) 486494
in the PHS. The inlet and outlet pressures change simultaneously
due to the high porosity and permeability of the sediment, which
are taken as the system pressure of the PHS. After the hydrate for-
mation, the system pressure is approximately 9.0 MPa (047 min,
Fig. 4), the initial hydrate saturation (volume) SH = 0.310, and the
initial aqueous saturation (volume) SA = 0.407 in the vessel. After
the free gas release from the vessel, the system pressure decreases
to approximately 5.4 MPa (4757 min), and the vessel is closed for
approximately 30 min (5787 min). Point A (87 min) and Point B
(1028 min) indicate the beginning and the end of the huff and puff
process including a total of 15 injection-soaking-production cycles.
Point C (390 min) and Point D (470 min) indicate the beginning and
the end of the partial enlarged plots of these gures. The huff and
puff process of the gas stage is between 87 and 1028 min. As shown
in Fig. 4, the average temperature in the vessel raises like zigzag
from 7.2 C to 16.8 C in the huff and puff process. In each soaking
step, the increase of the system pressure is smaller with the huff
and puff cycle proceeds.
As a typical example, the pressure and average temperature
changes with time of the 6th huff and puff cycle, which is between
401.5 and 462.5 min, are shown in the partial enlarged plot of
Fig. 4. As shown, in the hot water injection stage (401.5
422.5 min), the system pressure increases to approximately
5.81 MPa due to (i) the mass (water) injection and (ii) the gas re-
leased from hydrate dissociation under thermal stimulation in
the region of the hot water injection well. The average temperature
in the vessel increases continuously in this stage. Within the soak-
ing stage (422.5452 min), the system pressure in the PHS contin-
ues increasing without any mass injection, which conrms the
thermal stimulation effect on the hydrate dissociation. The heat
from the injected hot water soaks into the vessel gradually, mean-
while, the system pressure increases quite slowly, which is caused
by the limited hydrate dissociation rate after ceasing the hot water
injection process. The average temperature in the vessel decreases
gradually because of the hydrate dissociation. During the produc-
tion stage (452462.5 min), the PHS is depressurized using the
back-pressure regulator, and both gas and water are produced
from the production well (Wells V5 in Fig. 2). The average temper-
ature drops obviously due to the warm water produced from the
wells and the hydrate dissociation caused by depressurization.
When the pressure in the PHS decreases to approximately
4.7 MPa, the production well (Wells V5) is closed, and the next huff
and puff cycle begins at the same time.
Fig. 4. Inlet and outlet pressures and the average temperature in the vessel during
hydrate dissociation in the PHS.
3.2. Temperature proles and spatial distributions
Fig. 5 shows the temperature proles of T25AT25C, T17C, T9C,
T1C and T4C during hydrate dissociation in the PHS. Because of
symmetry in the cylindrical vessel, T1 approximately represents
T7, T43 and T49, and T17, T19, T31 and T33 have the similar pro-
les, and so on, while the injection and production well is at the
axis of the PHS. As shown in Fig. 3, the thermometers T25A, T25B
and T25C at spot T25 are also at the axis of the PHS and on Layers
AA, BB and CC, respectively, and T25C is at the bottom of the
grooves on the injection and production well (Well V5).
As shown in the partial enlarged plot of Fig. 5, the temperatures
of T25A, T25B and T25C at the injection well (Well V5) continue
decreasing at the rst minute after ceasing the former (5th) pro-
duction stage, i.e. from 401.5 to 402.5 min of the injection stage
of the 6th huff and puff cycle, while the system pressure already
begins to increase at 401.5 min after the hot water injected into
the vessel (Fig. 4). This indicates the duration of the heat transfer
from the injected hot water to the hydrate deposit. From 402.5
to 422.5 min of the 6th injection stage, the temperatures of T25A,
T25B and T25C at the injection well (Well V5) increases rapidly
and continuously, and the temperature of T17C in the vicinity of
the injection well also increases obviously, while the hot water
injection has a limited effect on the temperatures of T9C, T1C
and T4C, which are close to the inside edge of the PHS. The temper-
ature of T25B and T25C is obviously higher than that of T25A, due
to the gravity effect of the injected hot water. From 422.5 to
452 min of the 6th soaking stage, with the heat of the injected
hot water soaking into the PHS, the temperatures of T25AT25C
and T17C decrease consistently, and the temperatures of T9C,
T1C and T4C continue slowly increasing before the end of the soak-
ing stage. These temperaturetime proles are affected by the dis-
tribution of the thermometers, the distance away from the
injection well and the boundary conditions of the PHS in this work.
During the 6th production stage (452462.5 min), with the de-
crease of the pressure in the PHS from 6.1 to 4.7 MPa (Fig. 4), the
hydrate in the PHS dissociates and meanwhile the gas and water
are produced from the production well (Well V5). The temperature
of T25A on Layer AA at the production well (Well V5) decreases
after a sudden increase from 452 min to 454.5 min, which is a re-
sult of the uid ow of the relative warmer water from the bottom
of the production well (Well V5) and the cooling effect caused by
the endothermic hydrate dissociation, the latter is also the main
reason for the decrease of the other temperatures in Fig. 5. The
next huff and puff cycle begins at 462.5 min and the production
well is switched to the hot water injection well.
Fig. 5. Temperature proles during hydrate dissociation in the PHS.
 X.-S. Li et al. / Fuel 94 (2012) 486494 491

Fig. 7. Cumulative volumes of gas and water produced, and water injected during
hydrate dissociation in the PHS.

of the injected water stays in the vessel and the water production
is manageable during using the huff and puff method as the hy-
drate dissociation strategy. The cumulative gas produced during
each cycle gradually decreases over time, which indicates that
the hydrate dissociation rate decreases over time during using
the huff and puff method with constant injection rates Qinj and
the same driving force of depressurization P. The reason for this
Fig. 6. Spatial distributions of temperature over time during hydrate dissociation in is that with the constant hot water injection rate, the range of
the PHS. the thermal diffusion is restricted around the well (Fig. 6), and
depressurization rather than thermal stimulation is dominant for
Fig. 6 shows the temperature spatial distribution over time dur-
gas production, especially at the later cycles of the huff and puff
ing hydrate dissociation in the PHS. Fig. 6a is the beginning of the
process. The similar results have been concluded in an earlier
6th huff and puff cycle. The high temperature zones (dened as
numerical simulation [18]. As a criterion of gas production perfor-
the zone with T P 30 C) in Fig. 6b and c occur in the vicinity of
mance, the average gas production rate Qavg = VP/t during the entire
the injection well (Well V5 in Fig. 2, T25 in Fig. 3), and the highest
huff and puff process (871028 min) is approximately 2.87 L/min.
temperature in the PHS (more than 80 C) is observed around T25B
Point A to D in Figs. 8 and 9 represents the same time point with
and T25C (the vicinity of the injection well) at the end of the injec-
those in the above gures, and Point A (0 min), Point B
tion stage (422.5 min), as shown in Fig. 6c. It can be seen from
(102887 min = 941 min), Point C (303 min) and Point D
Fig. 6cf (422.5452 min) that the temperatures in the PHS declines
(383 min) are shown in Figs. 8 and 9. The beginning of the injection
with the injected heat soaking into the PHS during the soaking
stage of the 6th huff and puff cycle changed corresponding to
stage, and the highest temperatures in the PHS are lower than
401.587 min = 314.5 min, as shown in the partial enlarged plot
45 C at 452 min. During the production stage (452462.5 min,
of Figs. 8 and 9.
Fig. 6fh), the cooling effect of the hydrate dissociation during gas
Fig. 8 shows the evolution of the gas-to-water ratios RGW = VP/
and water production is observed in the PHS. Fig. 6h shows the
MW during the huff and puff cycles. RGW provides a measure of
end of the 6th huff and puff cycle (the beginning of the 7th cycle,
i.e. the start point of the injection stage of that cycle, which is sim-
ilar to Fig. 6a).

3.3. Gas and water production rate

Fig. 7 shows the cumulative volume of the produced gas (VP)

and water (MW), and the water (Minj) injected during the hydrate
dissociation in the PHS. As already discussed above (Fig. 4), the huff
and puff process begins at 87 min (Point A). The hot water is in-
jected during the injection stage, and the gas and water are only
produced during the production stage of each cycle. For example,
as shown in the partial enlarged plot of Fig. 7, during the 6th huff
and puff cycle, the hot water is injected during the injection stage
(401.5422.5 min), and the gas and water are produced during the
whole production stage (452462.5 min). With the constant hot
water injection rate and time (Qinj = 0.20 L/min and tinj = 20 min),
4.0 L of the hot water (Tinj = 160 C) is injected into the PHS during
each cycle. During the entire huff and puff process (Point APoint
B), the average water produced during each cycle (3.8 L) is lower Fig. 8. Effect of tinj on the cumulative gas-to-water ratio RGW during the huff and
than that injected into the PHS (4.0 L), which indicates that some puff cycles.
492 X.-S. Li et al. / Fuel 94 (2012) 486494
Fig. 9. Effect of tinj on the percentage of the hydrate dissociated during hydrate
dissociation in the PHS (tinj = 20 min in the reference case).
the effectiveness of hydrate dissociation as a gas-producing
method and a relative criterion of gas production performance.
Both VP and MW maintain constant in each cycle during the
injection and soaking stage and increase only during the produc-
tion stage, as shown in the partial enlarged plot of Figs. 7 and 8
(reference case). As shown in Fig. 8, RGW declines rapidly to
approximately 55105 ST m3 of CH4/m3 of H2O after the rst 5 huff
and puff cycles (0314.5 min), which is not prohibitively low even
considering the energy cost during the hot water injection pro-
cess, mainly the cost of the power used to pump the uid into
the well.
3.4. Sensitivity to tinj
In this work, we investigate into the sensitivities of the hydrate
dissociation to the hot water injection time tinj during the injection
stage of the huff and puff cycle. Figs. 8 and 9 show the dependence
of RGW and the percentage of the hydrate dissociated during hy-
drate dissociation in the PHS on the injection time of the hot water
tinj, respectively.
In the reference case discussed above, with the constant hot
water injection rate (Qinj = 0.20 L/min) and injection time (tinj = 20 -
min), 4.0 L of the hot water (Tinj = 160 C) is injected into the PHS
during each cycle. As shown in the partial enlarged plot of Fig. 9,
during the injection stage of the 6th cycle (314.5335.5 min), the
percentage of the hydrate dissociated decreases continuously from
45.5% to 43.4% over time, which means that the amount of hydrate
in the vessel increases. The reason for this is the hydrate formation
due to the pressurization effect caused by the mass (hot water)
injection, although there is hydrate dissociation under thermal
stimulation in the vicinity of the injection well (Fig. 6). During the
soaking stage of the 6th cycle (335.5365 min), after ceasing the
hot water injection process, the percentage of the hydrate dissoci-
ated increases to 45.5%, and there is hydrate dissociating and gas
releasing from hydrate. During the production stage of the 6th cycle
(365375.5 min), the percentage of the hydrate dissociated in-
creases continuously to 52.8%, and there is hydrate dissociating
and gas releasing from hydrate. In general, the percentage of the hy-
drate dissociated shows uptrend over time during the whole gas
production process. The diminishing uptrend of the percentage of
the hydrate dissociated indicates that the hydrate dissociation rate
decreases over time, and the reason for this is that the injected hot
water is restricted in the vicinity of the well, and the hydrate disso-
ciation mainly occurs in the area far away from the well, especially
at the later cycles of the huff and puff process. In the reference case,
the percentage of the hydrate dissociated is approximately 94%
after 15 huff and puff cycles, which means that almost all the hy-
drate in the PHS dissociated after the huff and puff process.
As shown in Fig. 8, the gas-to-water ratio RGW increases with tinj
from approximately 60400 min, which indicates that the gas pro-
duction efciency is improved by prolonging the hot water injec-
tion time. However, after about 6 huff and puff cycles, the
duration of the hot water injection tinj shows limited effect on
the long term RGW. The reason for this is that the larger the hot
water injection time tinj, the more water is produced during the
production stage, although more gas is produced at the same time
from the vessel during these cycles.
Reducing tinj from 20 to 10 min results in a signicant deterio-
ration of gas production performance, and the percentage of the
hydrate dissociated after 17 huff and puff cycles in the case with
tinj = 10 min is lower than 60% (Fig. 9). In the case with larger tinj,
more hot water is injected into the PHS during the injection stage,
and the entire vessel is heated more efciently, which caused the
expansion of the thermal effect range (the high temperature zone)
and more hydrate dissociation in the area far away from the well.
On the other hand, the percentage of the hydrate dissociated is
approximately 95% after 13 huff and puff cycles in the case with
tinj = 30 min, and almost all the hydrate in the PHS dissociated after
the huff and puff process. The enhancement of hydrate dissociation
is limited by increasing tinj from 20 to 30 min, due to the stronger
pressurization effect of the larger tinj.
3.5. Resistance ratio
The resistance can be used to characterize the change of the gas
hydrate reservoir [27,28]. Since the resistivity of gas hydrate is
greater than water, generally the resistance decreases as the hy-
drate decomposes. However, experimentally the resistance is also
affected by the temperature and water/gas ow in porous media.
In the work, we use the ratio between the real-time resistance dur-
ing the experiment and the resistance before the experiment (re-
ferred to as the resistance ratio) as a characterization parameter.
Fig. 10 shows the resistance ratio and temperature changes of
measuring point 45C during gas production using the huff and puff
method. It can be seen that the resistance ratio changes periodi-
cally with the huff and puff cycle. As a typical example, the resis-
tance ratio change of the 6th huff and puff cycle is given in the
in the partial enlarged plot of Fig. 10. As shown, the resistance ratio
change can be divided into four stages in 1 huff and puff cycle, of
Fig. 10. The resistance ratio, temperature and hydrate equilibrium dissociation
temperature proles at point 45C during gas production using the huff and puff
method.
 X.-S. Li et al. / Fuel 94 (2012) 486494
which the 1st and 2nd stages are within the injection stage of the
huff and puff cycle, and the 3rd and the 4th stages are the soaking
and the production stages, respectively. In the 1st stage, the resis-
tance ratio decreases obviously. In this stage, the hot water is in-
jected into the reactor from the well, the resistance ratio
decreases due to the fact the water saturation increase and the
resistance of water is lower than those of the hydrate and gas.
The 2nd stage of the resistance ratio change is also in the hot
water injection process. With the continuous hot water injection
and the thermal diffusion, the hydrate in the reactor begins to dis-
sociate and the gas content increases, which results in the resis-
tance ratio increase at point 45C. The 3rd stage of the resistance
ratio change is corresponding to the soaking stage of the huff
and puff cycle. In this stage, the hydrate dissociation rate is slow
and the gas and water saturation has no obvious change at point
45C, therefore, the resistance ratio keeps constant mostly. In the
4th stage, the production of water and gas begins, the water satu-
ration decreases and the resistance ratio increases obviously. In the
whole hydrate production process, the resistance ratio has no sig-
nicantly change in the rst 8 huff and puff cycles, as shown in
Fig. 10. As comparison, we calculated the hydrate equilibrium dis-
sociation temperature at the pressure in the reactor using the
fugacity equation given by Li et al. [26], which is also are given
in Fig. 10. It can be seen that the equilibrium dissociation temper-
ature begins to be lower than the temperature at point 45C from
the 9th huff and puff cycle (Point E), which means that the hydrate
at point 45C begins to dissociate from the ninth cycle, resulting in
the resistance ratio decrease.
Fig. 11 shows the resistance ratio, temperature and the hydrate
equilibrium dissociation temperature changes of measuring point
26B during gas production using the huff and puff method, which
is close to the hot water injection well. It can be seen that the resis-
tance ratio at point 26B only decreases obviously in the rst huff
and puff cycle and has no obvious change overall in the latter cy-
cles. It also can be seen that temperature begins to be signicantly
higher than the hydrate equilibrium dissociation pressure after the
rst huff and puff cycle, which means that the hydrate at point 26B
has dissociated over in the rst cycle. Therefore, the resistance
ratio only changes with the water injection and production in the
215th cycle.
The resistances of other points show similar behaviors. How-
ever, the time when the resistance ratio decreases depends on
the time when the hydrate begins to dissociate, and resistance ra-
Fig. 11. The resistance ratio, temperature and hydrate equilibrium dissociation
temperature proles at point 26B during gas production using the huff and puff
method.
493
tio at the point in the near-well region decrease earlier than that in
the far-well region. Before the hydrate dissociation at the point, the
resistance changes with the water, gas and hydrate saturation
changes. After the hydrate dissociation, the resistance only changes
with the water injection and the water and gas production in one
cycle, and the resistance ratio in different huff and puff cycle keeps
constant during the whole huff and puff process.
4. Conclusions
In this work, we developed a novel Pilot-Scale Hydrate Simula-
tor (PHS), a 117.8-L three-dimensional apparatus, to investigate
into the gas production behavior of methane hydrate in the porous
media using the huff and puff method. The huff and puff method,
also known as cyclic steam stimulation (CSS), includes the injec-
tion, soaking and production stages. In the PHS, a 9-spot distribu-
tion of the vertical wells, a single horizontal well and a 49-spot
distribution of the thermometers are respectively placed in three
horizontal layers, which equally divided the vessel into four parts.
A vertical well at the axis of the PHS was used as the injection and
production well. The following conclusions are drawn from the
experimental results.
In the reference case (Tinj = 160 C, Qinj = 0.20 L/min and tinj =
20 min), the evolution of the temperature and system pressure,
the spatial distribution of temperature, the cumulative volumes
of gas and water produced from the PHS during the injection-soak-
ing-production stages are shown in this work. The high tempera-
ture zones occur in the vicinity of the injection well, and the
highest temperature in the PHS (more than 80 C) is observed at
the injection well at the end of the injection stage. With the con-
stant hot water injection rate, the range of the thermal diffusion
is restricted around the well, and depressurization rather than
thermal stimulation is dominant for gas production.
The decline of the cumulative gas produced during each cycle
and the diminishing uptrend of the percentage of the hydrate dis-
sociated indicate that the hydrate dissociation rate decreases over
time. As the absolute criterion of gas production performance, the
average gas production rate Qavg = VP/t during the entire huff and
puff process is approximately 2.87 L/min. During the entire huff
and puff process, the average water produced during each cycle
is lower than that injected into the PHS, and the water production
is manageable during using the huff and puff method. The long
term gas-to-water ratio RGW is more than 55 ST m3 of CH4/m3 of
H2O, so the huff and puff method is considered to be economically
protable from the relative criterion (RGW) point of view.
The gas production efciency is improved by prolonging the hot
water injection time, while this enhancement of hydrate dissocia-
tion is limited due to the stronger pressurization effect of the larger
tinj.
Acknowledgments
This work was supported by National Natural Science Founda-
tion of China (Grants 51076155 and 51004089), CAS Knowledge
Innovation Program (Grant KGCX2-YW-3X6) and Science & Tech-
nology Program of Guangdong Province (Grant 2009B050600006),
which are gratefully acknowledged.
References
[1] Sloan ED, Koh CA. Clathrate hydrates of natural gases. 3rd ed. Boca Raton,
FL: CRC Press; 2008.
[2] Collett TS. Gas hydrates as a future energy resource. Geotimes 2004;49:247.
[3] Klauda JB, Sandler SI. Global distribution of methane hydrate in ocean
sediment. Energy Fuels 2005;19:45970.
[4] Moridis GJ, Collett TS, Boswell R, Kurihara M, Reagan MT, Koh C, et al. Toward
production from gas hydrates: current status, assessment of resources, and
494 X.-S. Li et al. / Fuel 94 (2012) 486494
simulation-based evaluation of technology and potential. SPE Reservoir Eval
Eng 2009;12:74571.
[5] Kono HO, Narasimhan S, Song F, Smith DH. Synthesis of methane gas hydrate
in porous sediments and its dissociation by depressurizing. Powder Technol
2002;122:23946.
[6] Ahmadi G, Ji CA, Smith DH. Production of natural gas from methane hydrate by
a constant downhole pressure well. Energy Convers Manage 2007;48:205368.
[7] Moridis GJ, Kowalsky MB, Pruess K. Depressurization-induced gas production
from class 1 hydrate deposits. SPE Reservoir Eval Eng 2007;10:45881.
[8] Li G, Li XS, Chen Q, Chen ZY. Numerical simulation of gas production from gas
hydrate zone in Shenhu area, South China Sea. Acta Chim Sin 2010;68:108392.
[9] Li G, Tang LG, Huang C, Feng Z, Fan SS. Thermodynamic evaluation of hot brine
stimulation for natural gas hydrate dissociation. J Chem Ind Eng (China)
2006;57:20338.
[10] Kawamura T, Ohtake M, Sakamoto Y, Yamamoto Y, Haneda H, Komai T, et al.
Experimental study on steam injection method using methane hydrate core
samples. In: Proceedings of the seventh (2007) ISOPE ocean mining
symposium. Lisbon, Portugal; 2007. p. 836.
[11] Li XS, Wan LH, Li G, Li QP, Chen ZY, Yan KF. Experimental investigation into the
production behavior of methane hydrate in porous sediment with hot brine
stimulation. Ind Eng Chem Res 2008;47:9696702.
[12] Li G, Li XS, Tang LG, Li QP. Control mechanisms for methane hydrate
production by thermal stimulation. In: The 6th international conference on
gas hydrate. Vancouver, British Columbia, Canada; 2008. p. 5783.
[13] Kawamura T, Yamamoto Y, Ohtake M, Sakamoto Y, Komai T, Haneda H.
Experimental study on dissociation of hydrate core sample accelerated by
thermodynamic inhibitors for gas recovery from natural gas hydrate. In:
Proceedings of the 5th international conference on gas hydrates (ICGH 2005).
Trondheim, Norway; 2005. p. 30238.
[14] Li G, Li XS, Tang LG, Zhang Y, Feng ZP, Fan SS. Experimental investigation of
production behavior of methane hydrate under ethylene glycol injection in
unconsolidated sediment. Energy Fuels 2007;21:338893.
[15] Najibi H, Chapoy A, Haghighi H, Tohidi B. Experimental determination and
prediction of methane hydrate stability in alcohols and electrolyte solutions.
Fluid Phase Equilibr 2009;275:12731.
[16] Moridis GJ, Reagan MT, Boyle KL, Zhang K. Evaluation of the gas production
potential of challenging hydrate deposits. In: TOUGH symposium, Lawrence
Berkeley National Laboratory, Berkeley, CA, USA; 2009.
[17] Li G, Moridis GJ, Zhang K, Li XS. Evaluation of gas production potential from
Marine gas hydrate deposits in Shenhu Area of the South China Sea. Energy
Fuels 2010;24:601833.
[18] Li G, Moridis GJ, Zhang K, Li XS. The use of huff and puff method in a single
horizontal well in gas production from marine gas hydrate deposits in the
Shenhu Area of the South China Sea. J Petrol Sci Eng 2011;77:4968.
[19] Moridis GJ, Reagan MT, Kim SJ, Seol Y, Zhang K. Evaluation of the gas
production potential of marine hydrate deposits in the Ulleung Basin of the
Korean East Sea. SPE J 2009;14:75981.
[20] Kurihara M, Funatsu K, Ouchi H, Masuda Y, Narita H. Investigation on
applicability of methane hydrate production methods to reservoirs with
diverse characteristics. In: Proceedings of the 5th international conference on
gas hydrates (ICGH 2005). Trondheim, Norway; 2005. p. 71425.
[21] Moridis GJ, Reagan MT. Strategies for gas production from oceanic class 3
hydrate accumulations. In: 2007 offshore technology conference. Houston,
Texas, USA; 2007. p. 18865.
[22] Moridis GJ, Reagan MT. Gas Production from oceanic class 2 hydrate
accumulations. In: 2007 offshore technology conference. Houston, Texas,
USA; 2007. p. 18866.
[23] Sayegh SG, Maini BB. Laboratory evaluation of the CO2 huff-n-puff process for
heavy oil-reservoirs. J Can Pet Technol 1984;23:2936.
[24] Vittoratos E. Flow regimes during cyclic steam stimulation at cold lake. J Can
Pet Technol 1991;30:826.
[25] Leaute RP, Carey BS. Liquid addition to steam for enhancing recovery (LASER)
of bitumen with CSS: results from the rst pilot cycle. J Can Pet Technol
2007;46:2230.
[26] Li XS, Zhang Y, Li G, Chen ZY, Yan KF, Li QP. Gas hydrate equilibrium
dissociation conditions in porous media using two thermodynamic
approaches. J Chem Thermodyn 2008;40:146474.
[27] Zhou XT, Fan SS, Liang DQ, Wang DL, Huang NS. Use of electrical resistance to
detect the formation and decomposition of methane hydrate. J Nat Gas Chem
2007;16:399403.
[28] Li XS, Wang Y, Li G, Zhang Y, Chen ZY. Experimental investigation into
methane hydrate decomposition during three-dimensional thermal huff and
puff. Energy Fuels 2011;25:16508.