Vous êtes sur la page 1sur 13

Fuel 256 (2019) 115742

Contents lists available at ScienceDirect

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

Full Length Article

Interaction relationship analysis between heat transfer and hydrate T


decomposition for optimization exploitation
⁎ ⁎
Shu Liua,b, Yunpei Lianga,b, , Bo Lia,b, , Qingcui Wana,b, Xiao Hana,b
a
State Key Laboratory of Coal Mine Disaster Dynamics and Control, Chongqing University, Chongqing 400044, China
b
School of Resources and Safety Engineering, Chongqing University, Chongqing 400044, China

A R T I C LE I N FO A B S T R A C T

Keywords: Heat transfer influences hydrate decomposition. Hydrate dissociation produces water. Then, water remained in
Natural gas hydrate the hydrate sediment shows impact on heat transfer. Six runs of experiment with different exploitation methods
Heat transfer and well arrangements were conducted to study the interaction relationship between heat transfer and hydrate
Hydrate decomposition decomposition. For water production, results show that by combination method, the wellbore heating can
Water production
greatly improve the water produced outside the reactor in production stage (WPII) until the hydrate decom-
Depressurization
position ceases. While in depressurization regime, the WPII is more easily influenced by hydrate distribution and
Wellbore heating
the water produced outside the reactor in the free and mix gas stage (WPI). For heat transfer, the overall heat
transfer trend is not affected by less WPII. So the decomposition was hardly impacted. On the other hand, the heat
transfer promotes heat absorption by hydrates, which improves hydrate decomposition. Subsequently, the hy-
drate decomposition consumes heat, but it cannot prevent the rise of temperature even in depressurization. So
the heat from the surrounding environment is enough for hydrate decomposition. Then, the heat absorption by
hydrate decomposition (Qd) and the heat absorption rate (Rd) are quantitatively studied. Specially, the heat
absorption ratio (λ) is firstly defined to show the ratio of Qd to the wellbore heating consumption (Ec). It provides
a fresh attitude for optimizing exploitation according to the heat transfer. In order to save more heat and im-
prove λ, we need to utilize the heat transfer from the ambient environment and make progress in reducing
wellbore heating consumption.

1. Introduction MH decomposition is an endothermic reaction, so it is controlled by


heat transfer. The hydrate mass also shows vital effect on hydrate de-
Natural gas hydrate (methane hydrate, MH), whose major formation composition. Hydrate dissociation consumes heat and produces gas and
component is CH4, is known to be one of the most potential energy water, which further absorbs heat. The heat transfer is thus weakened.
resources on account of the large distribution in the ocean and per- Then, the hydrate decomposition is impacted, so do the gas and water
mafrost (about 2.0 × 1016 m3) and its cleanness (decomposed only to production. This is a complex interaction, and the analysis of this
produce water and methane gas) [1,2]. The formation of it requires process contributes to the understanding of hydrate production prop-
abundant CH4 gas, high pressure and low temperature. Based on the erties. So some studies are about heat transfer and hydrate dissociation.
existence conditions, at present it is impossible to exploit MH as a kind Zhao and Song et al. [19–21] adopted numerical simulation to study the
of solid substance. Thus, the MH exploitation should be focused on the effect of heat transfer on MH decomposition. The results showed that
ways to dissociate MH in sediments and then collect the produced CH4 with depressurization, the sensible heat absorbed by MH decomposition
gas. By now, the feasible methods are pure depressurization, heat in- was significantly influenced by the properties of porous media and
jection, combination method, replacement method, and inhibitor in- water remained in it. And then by adding extra heat, they found that the
jection method, which have been adopted by some countries in the field gas production rate was reduced by the increase of the specific heat
exploitation [3–18]. However, because of the layered distribution of capacity in the sediment. But the water remained in the sediment and
hydrate deposits in most situations and the special formation condition, the heat convection effect of water and gas showed limited influence on
although the exploitation of MH can be achieved, it is far from ap- hydrate recovery. Moreover, the increase of the thermal conductivity
proaching the efficiency and ensuring the safety. showed a promotion effect on MH decomposition only in the later


Corresponding authors at: State Key Laboratory of Coal Mine Disaster Dynamics and Control, Chongqing University, Chongqing 400044, China.
E-mail addresses: liangyunpei@126.com (Y. Liang), libo86@cqu.edu.cn (B. Li).

https://doi.org/10.1016/j.fuel.2019.115742
Received 15 May 2019; Received in revised form 21 June 2019; Accepted 29 June 2019
Available online 31 July 2019
0016-2361/ © 2019 Published by Elsevier Ltd.
S. Liu, et al. Fuel 256 (2019) 115742

production period. Also using numerical simulation method, Liu et al. [42] conducted experiments in which they changed the temperature of
[22] analyzed heat and mass transfer with different surrounding tem- the injected water and injection time to study the exploitation effects by
peratures and production pressures by depressurization. The results this method. The analysis of the results indicated that the rises of
indicated that MH decomposition was controlled by pressure driving temperature and injection time could result in more volumes of the
force and the heat conductivity showed greater effect on decomposi- produced CH4 gas, but they were unable to make contribution to higher
tion. However, with the increase of the temperature in surrounding energy efficiency. And the injection water and duration with 130 °C and
environment, the heat transfer and gas production were enhanced, and 5 min respectively suggested the ideal production effect.
the effect of pressure difference was weakened. For the well configuration, Li et al. [43] utilized dual horizontal
Different exploitation methods show different production effects wells in the upper and lower layers of the pilot-scale reactor respec-
due to different heat transfer rules and mass changes [23]. In addition tively for the recovery of MH. With the use of hot water injection
to exploitation ways, the well arrangement plays a vital role in hydrate combined with depressurization, both the numerical simulation and
exploitation as the well location and type will not only influence heat experimental results indicated this kind of well arrangement was pro-
and mass transfer and flow of gas and water but also the stability of the mising for commercial exploitation as the high gas-to-water ratio and
hydrate layer [2,24]. Given the above situations, the production net energy gain were achieved with it. In order to study the decom-
methods of MH and the comparison of different well patterns have at- position behaviors of MH with multiple vertical well patterns, Wang
tracted many researching focuses for finding out economic and safe et al. [44] carried out experiments in the five-spot wells (5-wells) and
ways to realize the use of this alternative energy mode. the dual wells (D-wells) with hot water injection. And they found that
Depressurization is regarded as the most energy-saving method although the longer well distance and larger water injection rate could
since the MH is decomposed by absorbing the sensible heat conducted improve gas production, they had opposite effects on energy con-
from sediments and sea water, thus no extra energy is consumed in the sumption. Furthermore, one could see from the comparison of hor-
production process [3,25–31]. Li et al. [32] carried out the hydrate izontal well and vertical well on hydrate exploitation that the former
exploitation by depressurization in a 3D cubic reactor, and they found showed the better gas recovery performance than the later pattern in
that the increasingly lower production pressure had the ability to raise terms of the energy ratio and heat transfer efficiency [2]. However,
the gas production rate. Moreover, after the comparison of the water almost all the researches about the influences of well patterns on hy-
production in the free gas production stage with that in the hydrate drate recovery have been conducted through experiments and numer-
decomposition stage, the absolutely more water production was seen in ical simulations because the field test is not economical and time con-
the previous stage. However, due to the slow heat conduction from suming [2,41,45–47]. So when it comes to the field exploitation, the
surrounding environment and limited sensible heat from the sediment, well patterns should be chosen according to the MH distribution si-
this exploitation method shows its disadvantage in recovery time [29], tuation. For example, considering the layered distribution of MH in the
the advantage of which thus appears under certain conditions. For ex- ocean environment, the distribution among the horizontal direction is
ample, Feng et al. [30] suggested adopting this method when the scale much larger than that among the longitudinal direction. In this case, if
of the hydrate sediment was relatively large. And the experiment results the horizontal well is located in the hydrate layer, the length of the well
from Yang et al. [31] showed that the higher water saturation in hy- will be as long as the sediment layer, which is hard to guarantee the
drate deposit weakened the gas production rate. Besides, Phirani et al. integrity of the wellbore after the dissociation of MH. Therefore, the use
[29] did not suggest the use of this method in unconfined hydrate de- of vertical well shows its advantages in field exploitation due to short
posits that existed in the horizontal state. length of the well.
Due to the insufficient heat transfer in depressurization, another We can see that many studies are about the effects of different
way to break the equilibrium condition of MH occurrence is to add heat production methods, and some papers analyze the heat transfer prop-
to the hydrate layer, so the heat injection is an available method for erties in numerical simulation. Few papers have focused on the char-
hydrate exploitation. Li et al. [33] studied the recovery effects with hot acteristics of heat transfer by different exploitation ways and well ar-
water injection by carrying out experiments with different hydrate sa- rangements in experiments. More specially, conventional heat injection
turations, and the results indicated that the increase of hydrate sa- is accompanied with mass injection, such as hot water or steam. Few
turation had limited impact in average gas production rate and this researchers study the heat transfer using exploitation method without
method was not promising for the exploitation of MH with the sa- mass injection. Although the exploitation method with mass injection
turation above 64.00%. Apart from this, Wang et al. [34] summarized shows its advantage in both heat conduction and heat convection in the
that the thermal stimulation was not an effective way for the hydrate sediment, which improves the heat transfer, it has its obvious shortage
dissociation as the lowest hydrate decomposition ratio was shown from in causing a lot of energy loss when transported to the hydrate sediment
their experiments. And some other researches [23,29,35,36] also in- by wells. Besides, the numerical simulation can describe the hydrate
dicated that although it increased the heat transfer rate and may cost production rule but it does not consider the deposition characteristics of
less time in gas production than depressurization in some situations, the porous media. Therefore, with experiments, this paper focuses on
there was residual hydrate that was unable to be fully decomposed and the study of heat transfer effects on hydrate decomposition by de-
a low gas recovery rate was seen by this method. pressurization and the combined method with in-situ wellbore heating.
The separated exploitation method like pure depressurization and Furthermore, with a larger specific heat capacity, water flow is sus-
pure heat stimulation can only provide single decomposition driving pected to have a converse impact on heat transfer, which may in turn
force for MH dissociation, and the gas production effect then will be influence hydrate recovery. In addition, the heat absorption by MH is
weakened to a large scale. Thus, many studies have got access to the analyzed, and the heat absorption ratio (λ) is firstly defined. It is a fresh
combination method. After conducting experiments to simulate gas aspect for analyzing the production effects of different exploitation
production by depressurization and combination method with electrical conditions.
heating, Hideki et al. [37] compared the production effects of two
methods and they thought that the use of electrical heating could im- 2. Experimental system and operating procedures
prove the hydrate recovery and the gas production rate. The same re-
sults were seen in the views of Li et al. [38]. Huff and puff method is Fig. 1 shows the experiment system, with which the hydrate can be
regarded to be a promising combination method in hydrate extraction formed in the cubic reactor (withstanding pressure range: 0–25 MPa)
due to its special exploitation steps (i.e., injection, soaking and pro- with the volume of 1.5 L (length × width × height:
duction) that can make the heat diffused more sufficiently in the hy- 100 × 100 × 150 mm). The water bath around the reactor can remain
drate sediment and reduce energy consumption [39–41]. So, Li et al. in a constant temperature, and this provides for the simulation of the

2
S. Liu, et al. Fuel 256 (2019) 115742

Fig. 1. Schematic of the experimental system.

Fig. 2. Reactor internal structure and well design.

Table 1
The experimental system.
Parts Temperature part Gas supply Fluid supply Gas-liquid separation

Unit Water bath PID valve Gas storage tank Back pressure valve Flow pump Gas flow meter Electronic balance

Calculation range −5 to 30 °C 0–6000 psi 0–30 MPa 0–30 MPa 0.1–50 mL/min 0–5 L/min 0–2200 g
Error range ± 0.1 ± 0.1% ± 0.05 ± 1% ± 0.01 g

hydrate formation and decomposition in marine and permafrost en- Fig. 2, nine temperature sensors (T1–T9) are distributed uniformly in
vironment. In this paper, we carry out experiments to study the hydrate the central horizontal plane of the reactor. After the formation of the
exploitation behaviors in marine sediments, which is simulated by the hydrate, three vertical wells (No. 1–3) with a resistance (73.5 Ω) in each
quartz sand (diameter: 0.27–0.38 mm) [24,38,48,49]. When the ex- well are responsible for gas production and wellbore heating. In addi-
periment starts, one can see the pressure in the reactor through the tion to the above parts, the experiment system is composed with other
pressure sensors located at the bottom and outside of the reactor re- parts including the gas supply part (i.e., a methane gas cylinder, a gas
spectively. In addition, for the temperature monitoring, as shown in booster pump and a high pressure gas storage tank), the fluid supply

3
S. Liu, et al. Fuel 256 (2019) 115742

part, the gas-liquid separation part and the data monitoring part. The hydrate formation, mol.
information on parameters of each part can be seen in Table 1, and one nm,W – The amount of methane dissolved in water, mol.
can get more detailed introduction of this system from the published mW –The mass of the generated water, g.
papers [24,38]. Vpore – The volume of total pore, mL.
For the hydrate formation, after the sand filling, the porosity of the vm – Molar volume of methane gas, mL/mol.
simulated hydrate sediment in the reactor is tested to be 44.20%. VP – The volume of the produced methane gas at the standard si-
Subsequently, the deionized water (377.85 mL) and methane gas tuation, L.
(3.30 mol) injection is the second step, at the end of which the pressure NH – Hydration number.
in the reactor becomes to be about 22.00 MPa. The decrease of the MH – Hydrate molar mass, g/mol.
reactor pressure to around 8.00 MPa at the temperature of 7.8 °C after ρH – Hydrate density, g/mL.
approximately 15 days means the end of the MH formation step. At this MW – Molar mass of water, g/mol.
moment, the ratio of MH, water and gas to total pore volume is about ρW – Water density, g/mL.
44.00%, 20.00% and 36.00%, respectively, which is namely the initial
saturation of MH (SH0), water (SW0) and CH4 gas (SG0). And MH sa- The remaining MH in the reactor can be calculated as:
turation is close to the data from drilling wells in South China Sea [49].
mH = ρH SH Vpore (5)
In this paper, we conduct six runs of experiments, which are Run
1–6 with different exploitation methods or different well arrangements. So the decomposition ratio is calculated as:
In the runs with depressurization, Run 1 and 4 adopt central well (No.2)
mH
for depressurizing, while Run 5 uses the side well (No.3) as production φd =
mH1 (6)
well. On the other hand, with the combination method, the central well
is employed for gas production and wellbore heating in Run 2 and 3,
and Run 6 utilizes No.1 well for heating and No.3 well for production. mH – mass of dissociated MH, g.
The parameters of these runs are shown in Table 2. The exploitation mH1 – MH mass after formation, g.
pressures of these runs include 4.50 MPa and 4.00 MPa, which are
below the equilibrium pressure at the temperature of 7.8 °C. So the MH where H well and P well stand for the heating well and production well,
exploitation can be divided into two stages. In the first stage, the respectively, and heating rate of the runs with wellbore heating is 25 W;
pressure decreases from the initial pressure of about 8.00 MPa to the P represents for the production pressure; tdII is the duration of the stage
exploitation pressure of 4.50 or 4.00 MPa. And the second stage keeps II; WPI and WPII stand for the water production in the Stage I and II
pressure constant at exploitation pressure. Thus, free gas and mixture respectively; VP in this situation is the total volume of the produced
gas are released in the first stage (Stage I), while only the decomposed methane gas at the standard situation in the stage II.
gas is produced in the second stage (Stage II). This paper therefore
mainly studies the hydrate exploitation behaviors in the Stage II, and 3. Results and analysis
the saturations of MH (SH1), liquid (SW1) and gas (SG1) at the beginning
of this stage are seen in Table 2. The calculation of these parameters 3.1. Gas production and hydrate decomposition
during experimental process is shown as follows [24,38,50–54]:
SH + SG + SW = 1 (1) In the experiments, Run 1, 4 and 5 adopted depressurization, while
the wellbore heating combined depressurization method was used in
SH = SH1 (1 − x ) (2) Run 2, 3 and 6. Fig. 3 shows the changes of gas production (VP) and the
(nm1 + ρH SH 1 Vpore x / MH − nm, W − VP /22.4) vm remaining MH in the reactor (mH). In six runs, the gas production
SG = showed an upward trend [24], whereas the mH decreased gradually to 0
Vpore (3) with time. This indicates the complete decomposition of MH in the
NH MW ρH SH 1 x mW reactor. Besides, the VP and mH curves in each run intersected at one
SW = SW 1 + − point, and the two curves were almost symmetrical about the line
MH ρW ρW Vpore (4)
parallel to the x-axis through the intersection. More specifically, the
same changing rate was seen in mH curve as its corresponding VP curve
SH, SG, SW – Hydrate, methane gas and water saturation, respec-
in separated run. For example, the hydrate mass in the reactor with
tively.
depressurization reduced slower than that with the combination
x – Decomposition percentage of MH.
method, which indicated the promoting effect of wellbore heating for
nm1 – The amount of residual methane gas in the reactor after

Table 2
The experimental parameters.
Runs 1 2 3 4 5 6

SH0/% 44.80 46.31 47.20 50.70 47.18 48.74


SW0/% 19.19 17.80 18.65 14.85 18.60 17.92
SG0/% 36.01 35.89 34.15 34.45 34.22 33.34
P/MPa 4.50 4.50 4.00 4.00 4.50 4.50
H well 2 2 1
P well 2 2 2 2 3 3
tdII/min 213.18 84.90 64.22 190.67 255.05 161.13
WPI/g 0.62 12.12 34.43 0.38 9.12 7.90
SH1/% 32.80 31.00 31.10 33.00 33.48 33.99
SW1/% 29.94 28.19 27.19 29.01 27.80 29.68
SG1/% 37.26 40.81 41.71 37.99 38.82 36.33
WPII/g 7.11 31.19 25.35 7.63 8.69 22.36
VP/L 35.29 32.65 33.04 36.00 36.11 35.43
Fig. 3. Gas production and hydrate decomposition in the stage II.

4
S. Liu, et al. Fuel 256 (2019) 115742

Fig. 4. Water production (WPII) and hydrate decomposition ratio (φd) in the
stage II; Point A and B indicate the φd when the water production of Run 4 and
5 begins to stabilize respectively, and Point C stands for the φd when the WPII Fig. 5. Water change in the stage I. Wd represents the water originated from
starts to rise in Run 1. hydrate decomposition.

decomposition. Furthermore, just like VP curves, mH curves of Run 1 formation. The same result was shown in the numerical study of Wan
and 5 basically coincided with each other, although Run 5 showed a et al. [55], and they drew this conclusion based on the temperature
longer decomposition period due to the side well as production well. In comparison between the experiment and the numerical simulation.
addition, by the same exploitation method, comparing Run 1 with 4 and Fig. 5 shows the water and pressure change in the stage I. Run 1 had
Run 2 with 3 respectively, a better decomposition effect was seen in the little WPI in the stage I. Apart from the effect of gravity making it hard
runs with lower production pressure. It was because the gas released for the water to be produced outside the reactor [2], this was the evi-
from the reactor was all originated from MH dissociation, and no pro- dence of the uneven distribution of the hydrate. Most of the MH for-
duced gas remained in the reactor. Therefore, gas production was mation happened away from the production well, and the hydrate
controlled by MH decomposition, which was closely related with the around these areas was decomposed in the stage I, while few MHs
heat transfer properties of the system. In the later section, we will around the central well contributed to the Wd in this stage, leaving
mainly focus on analyzing the key factors that influence MH decom- more pores around the areas close to the boundary. So the water firstly
position. occupied the pores around the reactor boundary after hydrate decom-
position, and only 0.62 g water was carried out of the reactor by the
3.2. Water production limited effect of the gas flow in the stage I, making almost all of the Wd
left in the reactor. In the stage II, MH decomposition started from the
Since the water in the reactor is not easy to be produced outside, we boundary gradually to the central areas of the reactor as with de-
define the water originated from hydrate dissociation as Wd, and the pressurization the MH dissociation was only triggered by the sensible
water produced outside the reactor as WPI and WPII (i.e., water pro- heat from the water bath [24]. As a result, there were few pores
duction in the stage I and II, respectively). available for the water flowing to the central well until the heat from
Fig. 4 illustrates the water production (solid line) and hydrate de- the water bath transferred to these areas. At this time, the MH around
composition ratio (dotted line) of six runs, in which Point A and B in- the production well decomposed to produce more pores and more
dicated the φd when the water production of Run 4 and 5 began to amount of Wd for the release of water outside the reactor. Thus, the
stabilize respectively, and Point C stood for the φd when the WPII started water in Run 1 started to be produced after 90.42 min when the cor-
to rise in Run 1. One could see that the water production of the runs by responding φd at Point C was 0.67, indicating that the hydrate de-
depressurization was far less than those with combination method. In composition interface was close to the central well.
addition, the accumulative WPII by depressurization was almost the The same situation was seen in Run 4 that the WPI was only 0.38 g,
same. We could see from the numerical study of Liu et al. [22] that the but it increased once the stage II started. From Fig. 5, we could see that
water production rose immediately once production began and then Run 4 had the most Wd during the stage I. The little WPII was caused by
showed a gradual growth until it tended to be stable. But the different the same reason as that in Run 1, which was that the formation of MH
water production behavior was seen in the experiment. might be uneven, and many MHs were formed in the areas away from
For the runs with depressurization, the production pressure of Run 1 the central well. But we could see from Run 1 and Run 4 in Figs. 3 and 4
was 4.50 MPa, which was the same as that of Run 5 with the side well that this uneven distribution of hydrates that had an effect on water
(No. 3) for gas production. But in Run 1, the central well (No. 2) was production did not affect hydrate decomposition. However, as shown in
used as the production well, and Run 4 with 4.00 MPa adopted the Table 2, a high SH0 was seen in Run 4. In addition, the production
central well as well. One could see that the WPII curves of Run 4 and 5 pressure of Run 4 (4.00 MPa) was the lowest among other runs with
showed the same changing trend, which increased with a large slope 4.50 MPa. The larger pressure driving force and higher SH0 in the stage I
within about 25.00 min before almost being stable in the later long indicated more hydrate was decomposed to form more Wd in the re-
production period. actor. Yet, although the hydrate decomposition might firstly happen
But the opposite changing trend was illustrated in Run 1. The WPII in away from the central well causing less WPI in the stage I, due to the
Run 1 stayed nearly unchanged for more than 1/3 of its production large hydrate decomposition areas with a lower P, more pores around
duration. And after 88.90 min, it started rising gradually to about 7.11 g the central well was filled with more Wd in Run 4 than that in Run 1 and
at the end, although there was no significant difference in the accu- the hydrate decomposition had developed to central spaces. Therefore,
mulative water yield among three runs with depressurization. The when the stage II began, the water was produced outside the reactor in
reason for almost no water released from the reactor in Run 1 before Run 4.
88.90 min might be the uneven distribution of MH in the reactor after For Run 1 and 5, it could be seen from Fig. 4 that the φd curves of

5
S. Liu, et al. Fuel 256 (2019) 115742

Run 1 and 5 were almost the same due to the same pressure driving decomposition rate, excepted for being influenced by the water re-
force, but they had opposite WPII changing rules. In Run 5, the reason mained in the reactor, the changing rule of WPII was almost unaffected
why WPII started to rise immediately when the stage II began was that by different exploitation pressures. With almost the same SH0, SW0 and
the side well was adopted as the production well. And the hydrate near SG0, the less accumulative WPII in Run 3 than that in Run 2 was caused
the boundary of the reactor was firstly dissociated by depressurization, by the more WPI in the former than the later.
making the water released easily from the side well. Moreover, after Although the heating injection rate and P of Run 6 were the same as
27.03 min, the WPII remained almost stable in Run 5, but the φd at Point those of Run 2, before 9.18 min the WPII of the former with the side well
B was only 0.32. While in Run 1, the φd at Point C corresponding to the as production well was higher than that of the later with the central
rising point of WPII was 0.67. So, the same pressure driving force re- well, and also higher than that of Run 3 with a larger pressure driving
sulted in the identical hydrate decomposition rule, but the hydrate re- force before 20.68 min. As shown in Fig. 5, the reason for the higher
covery was hardly affected by the different location of the well with water production rate of Run 6 at the beginning of the stage II was that
depressurization. This was because the heat transfer rule in the two more water was remained in the reactor during the stage I. However,
locations of production well was the same. In addition, the hydrate with the heat originated from wellbore heating of the side well trans-
decomposition rate showed no obvious influence on the growth of WPII, ferred to the water bath, the heating driving force was weakened, and
but WPII was influenced by the position of the production well and the water production rate in Run 6 decreased. In addition, the φd of Run
hydrate distribution. Hence in the vertical well system, the effect of 6 was the lowest with the smallest hydrate decomposition rate and
gravity on WPII overwhelmed the carrying effect of gas flow, as the role longest production duration among the runs by combination method.
of pressure difference in carrying water was limited with depressur- The higher pressure driving force had the ability to enhance the hydrate
ization. Besides, the WPII could be controlled by arranging the location decomposition and the decrease of mH reduced the hydrate recovery
of the production well according to the hydrate distribution. rate. But the heating driving force showed a dominating effect on the
However, the WPII curves of Run 4 (4.00 MPa) with the central well WPII. When the wellbore heating had the ability to offer excessive heat
and Run 5 (4.50 MPa) with the side well showed the same changing to maintain the fast hydrate decomposition, the water could be pro-
trend. But the φd of the former was larger than the later, so did the duced with a high rate until the hydrate within the decomposition in-
rising rate of φd. The WPII of the two remained stable after 29.03 min, terface was ran out. Besides, the amount of the water in the hydrate
and the corresponding φd of the Run 4 at Point A was 0.59, which was sediment at the end of the stage I would influence the WPII at the be-
nearly twice that of Run 5 at Point B (0.32). We could see that the lower ginning of the stage II. In all, the water was released out of the vertical
P led to the faster hydrate decomposition, as the lower pressure cor- well by the effect of driving force, and by combination method with
responded to a lower equilibrium temperature, which resulted in a strong gas and water flow, the fluid flowing effect was able to conquer
larger temperature gradient and accelerated the heat transfer rate from the gravity influence to some extent due to large heating driving force.
the boundary. But it had almost no effect on WPII when the water was However, in depressurization, the pressure driving force was too small
produced once the stage II began in both runs. The decrease of mH, one to produce a large amount of WPII, resulting in less WPII than that with
of the reasons for the reduction of Wd and gas production rate [2], did combination method.
not impact WPII. So, the changing rule of WPII was mainly influenced by Apart from these, we could see the φd curve of Run 4 by de-
the location of the production well and also the hydrate distribution, pressurization coincided with that of Run 2 and 6 by combination
but the above two aspects as well as the P showed no obvious effect on method before 14.35 min. Then the φd of Run 2 exceeded that of Run 4
the final cumulative water yield. When the depressurization was em- and 6 and remained this advantage until the end of the stage II.
ployed in the exploitation of the hydrate with the same SH0, SW0 and Although the φd of Run 4 was higher than that of Run 6 during
SG0, the accumulative water production was almost the same. 14.35–53.38 min, the later overtook the former after 53.38 min. At this
However, different changing trends in WPII were showed in Run 2, 3 time, the φd and WPII of Run 4 was about 0.75 and 7.00 g, respectively.
and 6. The WPII curves in these runs with combination method in- Run 4 adopted depressurization while the combination method was
creased sharply with high production rates once the stage II began, and employed in Run 6. One could see from Run 1 and 5 that the position of
the growth trend continued until they became stable. The accumulative the depressurization well showed little impact on hydrate decomposi-
WPII of these runs was far more than the runs with depressurization. In tion and gas production. It suggested that we could adopt depressur-
addition, since the production duration of Run 6 was longer than the ization with a lower production pressure at the beginning of the ex-
other two with the combination method, a longer stable water pro- ploitation. This was because the same decomposition effect was seen
duction period was seen in this run. But when the WPII of the three runs comparing the run of lower WPII with that of the combination method,
started to be stable, most of the hydrates were dissociated, especially especially when the side well was used as the heating well. The heating
the φd in Run 2 and 3 approached to be 1.0. So, the fast hydrate de- could be added when the pressure driving force could not maintain a
composition contributed to strengthening the flow of gas and water high decomposition rate. This would avoid energy waste in hydrate
when the hydrate was enough to provide high mass transfer rate. exploitation and improve net energy gain.
The water production rate of Run 3 was smaller than that of Run 2 From Figs. 4 and 5, one could see that the heating driving force
before 10.18 min, after which the WPII curves of them showed the same showed a positive impact on the water production, but the water pro-
changing trend although the later curve was higher than that of the duction was insensitive to the production pressure difference. Besides,
former during the whole production period, so did the final accumu- less mH at the end of stage I and more hydrate distributed far away from
lative WPII. As shown in Fig. 5, the slower rising rate of WPII in Run 3 at the production well would delay WPII and decrease the water produc-
the beginning of the stage II might be caused by the smaller Wd with tion rate in the early duration of stage II. When the hydrate exploitation
larger WPI in the stage I, leaving smaller water in the reactor at the is carried out in the field, since the use of depressurization leads to less
beginning of the stage II. However, the φd curves of the two runs WPII and has the same production effect in a lower pressure compared
showed the opposite changing trend, that is, the φd of Run 3 kept larger with combination method during the early period, one can think about
than that of Run 2 during the stage II. This was because with the same adding wellbore heating after the pressure driving force cannot main-
heat injection rate of 25 W by wellbore, the lower P in Run 3 resulted in tain a high decomposition rate. This will help to achieve the high en-
the higher φd than that in Run 2 as higher decomposition driving force ergy recovery, and reduce the amount of WPII. In the stage I, it is op-
was adopted by a higher heat transfer rate in Run 3. And it led to the timal for the exploitation if less water was left in the reactor. This is
shorter hydrate recovery duration in Run 3. It could be seen with the because the Wd will occupy the pores first before they are released
same heating injection rate, although the bigger driving force provided outside, making less WPII in the stage II. However, WPI is not impacted
by the lower exploitation pressure made contribution to the faster by the pressure difference. And in the stage II, the pressure driving force

6
S. Liu, et al. Fuel 256 (2019) 115742

rising rate. Yet, with the reduction of mH in the reactor, the heat
transferred from the water bath was abundant. These heat was ab-
sorbed by the sediment for the increase of the temperature in the re-
actor. So the temperature rising rates of Run 1, 4 and 5 showed sudden
increases after Point a, b and c. This could also be seen from the Point a,
b and c in Fig. 7 that the temperatures in the reactor started to rise to
the Tw of 7.8 °C at the corner areas. This meant the hydrate decom-
position was approaching to an end. Besides, one could see from the
Fig. 7(a) that the heat transfer dominated more than half of the hydrate
sediment in Run 1 at t = 100.80 min, which confirmed that the hydrate
decomposition interface was close to the central well, resulting in the
beginning of water production according to Fig. 4. Furthermore, the
analysis of WPII and φd in Fig. 4 suggested that it could reduce the WPII
and heat consumption by adding heating when the depressurization
could not ensure a high φd.
The lowest initial temperature was seen in Run 4 as the production
pressure of it was 4.00 MPa. Run 4 had a higher initial saturation, and a
Fig. 6. Average temperature changes of six runs in the reactor. Point A and B in lower pressure meant a larger pressure driving force in the stage I,
Run 2 and 3 respectively stand for the starting point of which the temperature making more hydrate decomposed than the other two runs with
accelerates to rise, and Point C (Run 6) indicates the starting point of which the 4.50 MPa. But the temperature of it was the first to accelerate at Point b
rise of the temperature decreases; Point a, b and c in Run 1, 4 and 5 respectively (157.63 min) and exceeded that of Run 1 and 5 at 168.98 min and
represent the starting point of which temperature accelerates to increase; Tw 157.68 min respectively. It could be observed that the heat transfer in
stands for the temperature of the water bath.
Run 4 was faster than that of Run 1 and 5 and the hydrate decom-
position was faster in it than the other two. So the lower production
also shows little influence on it, while the heat driving force has the pressure resulted in a faster heat transfer rate as the corresponding
ability to greatly improve the WPII, in spite that the distribution of equilibrium temperature was lower. Furthermore, we could see from
hydrate and mH shows an influence on it. On the other hand, the use of Fig. 6 that the temperatures of Run 1 and 5 showed the same increasing
side wells as production and heating well is not a promising method for trend with the same exploitation pressure, indicating the same heat
improving hydrate recovery and reducing WPII. In all, the relatively less transfer rate. The reason for the delay of Point c to Point a and then the
water is produced in the vertical well system due to the effect of gravity longer production period was that Run 5 with the side well as the
[2]. As a result, the pressure driving force is not that enough to enhance production well had a slightly higher SH and a longer gas and water
the flow of gas and water and conquer the effect of gravity, making the flowing distance than that in Run 1 with central well [24].
effect of hydrate and water distribution in the reactor on water pro- As shown in Fig. 7(a) and (c), the temperature distributions at the
duction more obvious in depressurization. But the heating can con- midpoint of Run 1 and 5 almost corresponded with each other. The
tribute to the fast decomposition and gas production, enforcing the flow Point a and c of them also corresponded with each other. The tem-
of water and gas out of the reactor. For understanding the production perature gradient decreased from the boundary to the center areas.
behavior clearly, the further analysis of heat transfer is needed. When using depressurization, the hydrate decomposition of Run 1 and
5 was only promoted by the heat from the water bath. So with the same
3.3. Heat transfer pressure driving force and initial temperature, they showed the iden-
tical heat transfer rule. In addition, as discussed in Figs. 3 and 4, Run 1
Fig. 6 illustrates the average temperature changes of Run 1–6 in the and 5 showed the same decomposition rule. So with the same heat
reactor, and the initial temperatures of them were below 7.8 °C as the transfer and mass transfer rate, the hydrate decomposition interface
hydrate absorbed the sensible heat for decomposition from the sedi- swept across the same area per minute in the two runs. This was also
ment in the stage I. They all showed an increasing trend with time. This the evidence that the hydrate decomposition was controlled by the heat
indicated the dominating effect of heat transfer on the hydrate de- transfer and hydrate mass transfer [14]. As a result, the same gas
composition. In the runs with depressurization, the temperatures in- production rule of them was seen in Fig. 3. But the temperature rises of
creased gradually with continuous decreasing rising rate before Point a, Run 5 in the areas near (x = 0 m, y = 0 m) and (x = 0.04 m,
b and c in Run 1, 4 and 5 respectively, after which the temperatures of y = 0.04 m) were slower than those in the same positions of Run 1. This
them showed sudden increases. Fig. 7 shows the temperature dis- was because the No.3 well was used as the production well, the water
tribution of the central plane in the reactor by depressurization. The flowing distance was further than that of Run 1. So the water stayed in
heat from the water bath transferred gradually from the reactor the reactor for a longer period of time, absorbing more heat. This in-
boundary to the central areas, making the hydrate located around the fluenced the heat transfer in these areas. Besides, there might be more
boundary firstly dissociated, and the decomposition interface moved hydrate around the No.1 well and the area of (x = 0.04 m, y = 0.04 m)
gradually to the central well. Specially, since the heat from the water in Run 5, reducing the heat transfer rate in these small areas. From the
bath could be transferred to the reactor from two boundaries at the above analysis, we could see that the water production and uneven
corner of the reactor, the hydrate around these corners was confronted distribution of hydrate reduced heat and mass transfer to some extent.
with larger heat transferring interface and could absorb more heat than Then the heat and mass transfer rate was reduced in Run 5. This
the hydrate around the center of the boundary in the same time. Thus, weakened the hydrate decomposition. Conversely, the amount of the
the temperature rise around the boundary center was behind that of the decomposed water was decreased. So the heat transfer became stronger
corner. In addition, at the beginning of the stage II, the heat transfer again. Although this interaction process prolonged hydrate exploitation
effect overpowered hydrate decomposition, leading to the rise of the duration in Run 5, the overall decomposition trend was unaffected.
temperature. However, with the increase of the decomposition interface For the runs by depressurization combined with wellbore heating,
as shown in Fig. 7, the heat transfer resistance rose due to the in- one could see from Fig. 6 that the temperature rises were far faster than
creasing heat transfer distance and the increasingly more heat absorbed those by depressurization, and the final temperatures were significantly
by hydrate dissociation, as well as the sediment sand. This weakened higher than the depressurization runs. This indicated that the wellbore
the heat transfer effect, leading to the decrease of the temperature heating could greatly improve the heat transfer. The lowest initial

7
S. Liu, et al. Fuel 256 (2019) 115742

Fig. 7. Temperature distribution of the central plane in the reactor by depressurization; (a), (b) and (c) represent the temperature distribution in Run 1, 4 and 5
respectively.

temperature was seen in Run 3, as it adopted 4.00 MPa for gas pro- respectively. With the same heating power, the lower production
duction. But the temperature of it increased sharply during the stage II pressure resulted in a faster heat transfer rate and also a larger hydrate
and exceeded that of Run 6 and 2 at 32.20 min and 34.87 min, decomposition rate, as shown in Fig. 4. In addition, Run 2 and 6

8
S. Liu, et al. Fuel 256 (2019) 115742

employed the same heating power and production pressure. The tem- consistent with the former research results [34,38,56]. Thus, the di-
perature rose with almost the same trend during 0–30.53 min, during rection of the driving force was in accordance with that of the flow of
which there were several temperature fluctuations. This made the gas and water in depressurization. The direction of gas and water flow
temperature of Run 6 lower than that of Run 2. However, after Point C, was the same with the heat transfer in Run 6, although the heating well
the temperature of Run 6 was rising with a lower rate, while that of Run was located in the boundary center. But in Run 2 and 3, the direction of
2 continued increasing even with a higher rate after Point A. It made the heat transfer was opposite to the gas and water flow. This might
the temperature of the former increasingly lower than that of the later. hinder the flow of gas and water. In addition, the heat from wellbore
This was because when the side well was adopted for heating in Run 6, heating dominated the hydrate decomposition soon after the beginning
the heat originated from the side well would continuously transfer to of the heating. But in Run 6, from Fig. 8(c) of t = 35.20 min, the hy-
the water bath with a relatively lower temperature, leading to a large drate around the two corners of the boundary (x = 0.04 m) absorbed
amount of heat loss. heat from water bath for decomposition. From Figs. 6–8(a) and (b), the
Before Point A and B, although the temperatures of Run 2 and 3 final temperature of runs by depressurization tended to be the same
increased, the heat transfer effect was not large enough to make the level, as well as that by combination method. It showed us that with the
temperature in the reactor rise at an increasingly higher rate. This was same heating driving force, the heat exchange between the reactor and
because the hydrate decomposition interface was increasing and the the water bath reached the same equilibrium, although the final tem-
hydrate needed a lot of heat for decomposition, which weakened the perature of the runs by depressurization is much lower than that of the
heat transfer to some extent. However, this weakening effect of heat later.
transfer showed little influence on hydrate decomposition, as shown in In general, the same production pressure resulted in almost the
Figs. 3 and 4. The temperature of Run 2 and 3 accelerated to rise after same initial temperature, and the heat transfer from the water bath in
Point A and B respectively. It showed that the effect of heating from the the stage I had limited effect on the initial temperature. The smaller
heating well overwhelmed the effect of heat absorption by hydrate initial temperature caused by the lower production pressure in the stage
decomposition. Fig. 8 illustrated temperature distribution in the reactor II contributed to the faster heat transfer rate. Besides, the location of the
by combination method. So this could also be observed from Fig. 8(a) production well had little influence on the heat transfer in depressur-
and (b) that the heat had transferred to most areas of the reactor. At ization, while the heat transfer was obviously impacted by the ar-
t = 30.53 min and t = 28.87 min in Run 2 and 3 respectively, one could rangement of the heating well in combination method. The combina-
see from Fig. 4 that the φd of Run 2 and 3 was about 0.70 and 0.85 tion method with wellbore heating wasted more heat for greatly
respectively. So the hydrate mass transfer showed a more serious im- increasing the temperature of the hydrate sediment. We can see that
pact on heat transfer, that is, the reduction of hydrate mass that wea- heat transfer improves hydrate decomposition. Then, hydrate decom-
kened the heat absorption for hydrate decomposition enhanced the heat position consumes heat, but it cannot prevent the rise of temperature
transfer. At this time, the cease of heating was suggested to avoid en- even in depressurization. This is because hydrate decomposition is an
ergy waste at this situation. During the end period of Run 2 and 3, the endothermic reaction, which absorbs heat. The heat transfer from
temperature rising rates of them decreased. It was due to that the heat boundaries is the dominating factor for hydrate decomposition in de-
had transferred to the boundary, and the water bath started to absorb pressurization. So even with depressurization, the method without
this heat, slowing down the increase of the temperature. This meant extra heat injection, part of the heat transferred from the boundaries
that the exploitation was to be end. The final temperatures of the two will have to be absorbed by the sediment for increasing its temperature.
runs approached to be the same, indicating the heat exchange equili- Then a temperature gradient is maintained for continuous heat transfer
brium between internal reactor and external water bath. It showed us from the boundary to the inner hydrate deposit. Actually, the heat from
that the larger pressure difference had the ability to improve the heat the surrounding environment is enough for hydrate decomposition.
transfer rate, but the final temperature of the system was determined by Therefore, this is not enough to assess the feasibility of hydrate ex-
the wellbore heating driving force. ploitation methods, and we need to consider the heat absorption and
In Run 6, there were several temperature fluctuations within energy consumption to find out the most valuable production way.
35.20 min. As shown in Fig. 4, each temperature fluctuation accom-
panied with the fluctuation of water production in Run 6. The water
production was controlled by hydrate decomposition. When the water 3.4. Heat absorption analysis
production rate decreased while the hydrate decomposition continued,
the decomposed water remained in the reactor increased. These water 3.4.1. Heat absorption characteristics
with a larger specific heat capacity than the hydrate deposit absorbed In this paper, Qd (KJ) is the heat consumed by MH decomposition,
the heat which should have been absorbed by the hydrate deposits for and it is calculated as Qd = ΔH ·mH , where ΔH = 62.8 KJ/mol is the
the increase of the temperature. So with a large amount of the Wd re- dissociation heat of MH [57]. The Rd (KJ/min) is used to show the heat
dQ
mained in the reactor by combination method, it showed some hin- absorption rate, which is thus defined as Rd = dtd . λ is the heat ab-
drance to heat transfer. But the impact on the overall heat transfer was sorption ratio, and it is adopted to show the proportion of heat ab-
limited. Thus, as shown in Figs. 3 and 4, the hydrate decomposition in sorption by MH decomposition to the energy added for the decom-
Q
Run 6 was hardly influenced by water production, as well as the gas position. So we define it as λ = Ed . But since no extra heat is provided in
c
production of it. In all, by combination method, the larger heat transfer depressurization, this paper compared the λ in Run 2, 3 and 6 by
rate resulted in the faster hydrate decomposition. Although the water combination method.
production showed some influence on heat transfer, the overall hydrate Fig. 9 illustrates the heat absorption characteristics by MH decom-
decomposition was not impacted. Therefore, the gas production was position, and the parameters are shown in Table 2. The heat absorption
determined by hydrate decomposition in combination method. used for hydrate dissociation showed an increasing trend in six runs,
For the comparison of the heat transfer between depressurization and this is the evidence of endothermic reaction of MH dissociation.
and the combination method, we could see from Figs. 7 and 8 that apart The slightly smaller accumulative Qd in Run 2 and 3 was due to the
from the heat transfer rate, the heat transfer direction of the former slightly lower SH1 of them. In Fig. 9(b), the Rd curves showed decrease
method differed from the later. By depressurization, no matter where trends approaching to 0 during the exploitation except Run 6. It in-
the production well was, the latent heat used for hydrate decomposition creased to its peak at 3.88 KJ/min at start and then decreased. This
was transferred from the boundary to the center. While with combi- verified the analysis from Fig. 8(c) that at the beginning of the pro-
nation method, the dominating heat was diffused from the heating well duction before the heating diffused fully, the hydrate could absorb heat
in the central area or around the boundary center, and these were from heating well and water bath as the Run 6 adopted No.1 and 3 well

9
S. Liu, et al. Fuel 256 (2019) 115742

Fig. 8. Temperature distribution of the central plane in the reactor by combination method; (a)–(c) represent the temperature distribution in Run 2, 3 and 6
respectively.

as heating well and production well respectively. Thus the Rd of Run 6 absorption rule and rate between Run 1 and 5. So the same heat and
was larger than that of Run 2 with the same heating rate. However, mass transfer rules resulted in the identical heat absorption rule. Thus,
with the heat transferred to the water bath and due to the larger heat the hydrate decomposition and gas production of them were nearly the
loss, the Rd of Run 6 fell behind that of Run 2. The exploitation of Run 6 same. This could also be seen from the comparison of the runs by de-
tended to be ended after 53.22 min. So the reduction of mH led to the pressurization with combination method that the faster heat absorption
stop of heat absorption by MH. From Fig. 9, we could see the same heat in combination method was caused by the better heat and mass

10
S. Liu, et al. Fuel 256 (2019) 115742

Fig. 10. Heat absorption ratio of runs with combination method. λ = 1.0 in-
dicates the balance of heat absorption by MH decomposition and heat con-
sumption by wellbore heating; Q = 25 W.

run that the extra heat injection was not excessive. However, λ > 1.00
could be regarded reasonable considering that the heat from wellbore
heating was absorbed for MH decomposition. As discussed in Fig. 6, the
sediment absorbed excessive heat making the temperature of it rise
above the surrounding water bath. This was because the pressure dif-
ference also provided decomposition driving force, and MH could also
absorb the heat from the water bath with a higher temperature for
decomposition. Furthermore, under the same heat injection rate, Run 2
with a larger P (4.50 MPa) showed obviously less λ than that of Run 3
(4.00 MPa) during nearly the entire period of production. In addition,
before 16.68 min, there was a period during which the λ of Run 6 was
higher than that of Run 2. As shown in Fig. 9(b), the Rd of Run 6 was
larger than that of Run 2. So with a larger pressure driving force in Run
3, the hydrate decomposition could be enhanced by absorbing heat
Fig. 9. Heat absorption characteristics of MH decomposition. (a) Heat absorp-
tion by MH decomposition; (b) Heat absorption rate. more quickly from the heating well and the water bath. Besides, before
a large amount of heat was wasted in Run 6, the λ could be improved by
hydrate absorbing heat from water bath as shown in Fig. 8(c). With a
transfer. And this greatly shortened the heat absorption duration.
large amount of heat loss, the λ of Run 6 was the lowest, only 0.45 in
On the other hand, the lower P had the ability to improve heat
the end. Thus, in order to save more energy and improve λ, we needed
absorption as shown from the comparison of Run 1 with 4 and Run 2
to utilize the heat transfer from the ambient environment and make
with 3, respectively. This verified that the lower P could enhance the
progress in reducing wellbore heating consumption by considering
heat and mass transfer rates due to the larger pressure driving force.
lowering heating rate and P.
Then the gas production was improved. However, in Fig. 9(b), with a
larger P, the Rd of Run 1 was higher than that of Run 4 during the latter
4. Conclusion and discussion
period of production. The same situation was seen in Run 2 with a
larger P and Run 3 with a lower P. When the mH was approaching to be
In vertical well system, the effect of gravity is so obvious that the
ran out as shown in Figs. 3 and 4, the heat was mainly absorbed by
water production is limited, especially when depressurization is
sediment sand rather than hydrate decomposition. During these time,
adopted for exploitation. This is because the pressure difference is too
we can consider the cease of wellbore heating. Conversely, one could
weak to enforce the flow of gas and water, and the effects of hydrate
see from the Figs. 7 and 8, if the heating was added later, the hydrate
distribution and the remaining water in the reactor on water production
could dissociate from two directions, one of which started from the
thus become apparent. In addition, due to the high heat and mass
boundary by absorbing heat from water bath in the early depressur-
transfer enhanced by wellbore heating, the water production is much
ization period. Another direction was from the central heating well
larger with combination method than that with depressurization. In the
after the heating was added when the depressurization driving force
field exploitation, we can employ depressurization with a lower pro-
cannot ensure the fast hydrate decomposition. This was beneficial for
duction pressure in the early production stage for the same decom-
shortening the heat transfer distance and saving energy.
position rate with less water production and energy consumption
compared with combination method.
3.4.2. Heat absorption ratio For the analysis of heat transfer, a lower production pressure results
From the above analysis, one could see the obvious advantage of in a lower initial temperature in the stage II, which can lead to a faster
depressurization in λ, as it could be considered as no extra heat injec- heat transfer rate. The temperature change during the production was a
tion. Fig. 10 illustrates the heat absorption ratio of the runs with process of interaction between heat transfer and hydrate dissociation,
combination method. It provides us another way (the analysis of heat although the effect of the heat transfer dominates the hydrate decom-
utilization) to compare the exploitation effects in different exploitation position. We can also see that the heat transfer starts from the boundary
conditions. Run 3 with the lowest P of 4.00 MPa showed the shortest to the central areas in depressurization, while with combination
production duration and the largest λ (above 1.00). So it was the only method, heat diffuses from the heating well. Besides, the small amount

11
S. Liu, et al. Fuel 256 (2019) 115742

of WPII causes a lot of Wd to remain in the reactor, which leads to a 2008;25:819–29.


certain hindrance to heat transfer. But the overall heat transfer rule is [8] Boswell R, Schoderbek D, Collett TS, Ohtsuki S, White M, Anderson BJ. The Iġnik
Sikumi field experiment, alaska north slope: design, operations, and implications for
hardly impacted. CO2–CH4 exchange in gas hydrate reservoirs. Energy Fuels 2016;31(1). https://doi.
Finally, the heat absorption by MH decomposition is controlled by org/10.1021/acs.energyfuels.6b01909.
the heat and mass transfer. The same heat and mass transfer result in [9] G.J. Moridis, T. S. Collett, S. R. Dallimore, T. Inoue, T. Mroz, Analysis and inter-
pretation of the thermal test of gas hydrate dissociation in the JAPEX/JNOC/GSC
the identical hydrate decomposition and then the same VP and Wd rule. et al. Mallik 5L-38 gas hydrate production research well, (2005).
The depressurization shows great advantage in λ, but the Qd and Rd of it [10] Ryu BJ, Riedel M, Kim JH, Hyndman RD, Lee YJ, Chung BH, et al. Gas hydrates in
are too low to result in the fast heat transfer from surrounding en- the western deep-water Ulleung Basin East Sea of Korea. Marine Petrol Geol
2009;26:1483–98.
vironment. The fast heat and mass transfer is seen in combination [11] Pang WX, Xu WY, Sun CY, Zhang CL, Chen GJ. Methane hydrate dissociation ex-
method, which can greatly reduce heat absorption duration. This is periment in a middle-sized quiescent reactor using thermal method. Fuel
beneficial for the energy recovery, but the wellbore heating is so 2009;88:497–503.
[12] Gambelli AM, Rossi F. Natural gas hydrates: comparison between two different
abundant that a lot of heat from it is absorbed by the hydrate sediment
applications of thermal stimulation for performing CO2 replacement. Energy
and water bath, causing heat waste. In order to save more energy and 2019;172:423–34.
improve λ, we need to utilize the heat transfer from the ambient en- [13] Yuan Q, ChangYu WANG, XiaoHui ZENG, XinYang X, Yang B Liu. Experimental
vironment and make progress in reducing wellbore heating consump- study of gas production from hydrate dissociation with continuous injection mode
using a three-dimensional quiescent reactor. Fuel 2013;106:417–24.
tion by considering lowing heating rate and P. Thus, combination [14] Zhan L, Wang Y, Li X-S. Experimental study on characteristics of methane hydrate
method with a lower production pressure is economical for commercial formation and dissociation in porous medium with different particle sizes using
exploitation. Although the large λ is once obtained during the start of depressurization. Fuel 2018;230:37–44.
[15] Zhang L, Yang L, Wang J, Zhao J, Dong H, Yang M, et al. Enhanced CH4 recovery
production with separated side wells for exploitation, due to the heat and CO2 storage via thermal stimulation in the CH4/CO2 replacement of methane
loss weakening the heat transfer, it shows less advantages in heat ab- hydrate. Chem Eng J 2017;308:40–9.
sorption than center well for production. So it is not applicable to be [16] Wang B, Dong H, Liu Y, Lv X, Liu Y, Zhao J, et al. Evaluation of thermal stimulation
on gas production from depressurized methane hydrate deposits. Appl Energy
used in MH exploitation. 2018;227:710–8.
Due to the obvious advantage of combination method in shortening [17] Song Y, Cheng C, Zhao J, Zhu Z, Liu W, Yang M, et al. Evaluation of gas production
production duration, we need to conduct other experiments with dif- from methane hydrates using depressurization, thermal stimulation and combined
methods. Appl Energy 2015;145:265–77.
ferent wellbore heating injection rates for finding out the way to obtain [18] Zhao J, Zhu Z, Song Y, Liu W, Zhang Y, Wang D. Analyzing the process of gas
the higher λ. Besides, considering the increasingly more heat wastes production for natural gas hydrate using depressurization. Appl Energy
and smaller hydrate remained in the reactor with time, one can con- 2015;142:125–34.
[19] Zhao J, Wang J, Liu W, Song Y. Analysis of heat transfer effects on gas production
sider the end of heating when the hydrate decomposition becomes re-
from methane hydrate by thermal stimulation. Int J Heat Mass Transf
latively slow. And in the later period, the only use of depressurization 2015;87:145–50.
can be taken into account. Considering the water production rule and [20] Song Y, Wang J, Liu Y, Zhao J. Analysis of heat transfer influences on gas pro-
heat transfer from two directions in the reactor, it is better to adopt duction from methane hydrates using a combined method. Int J Heat Mass Transf
2016;92:766–73.
depressurization in the early production period and then adding well- [21] Zhao J, Liu D, Yang M, Song Y. Analysis of heat transfer effects on gas production
bore heating. This paper analyzes the heat utilized by hydrate decom- from methane hydrate by depressurization. Int J Heat Mass Transf 2014;77:529–41.
position for the judgment of the production effects with different ex- [22] Liu J, Shao Z, Wu M, Zheng Y. Heat and mass transfer analysis of depressurization-
induced hydrate decomposition with different temperatures of over-and under-
ploitation conditions. The later research can focus on the optimization burden. J Nat Gas Sci Eng 2017;44:65–76.
of the exploitation method according to the quantitative analysis of [23] Liang YP, Liu S, Wan QC, Li B, Liu H, Han X. Comparison and optimization of
heat utilized by hydrate decomposition and wasted by increasing the methane hydrate production process using different methods in a single vertical
well. Energies 2018;12:1–21.
temperature of sediments and transferring to the water bath. This [24] Liang YP, Liu S, Zhao WT, Li B, Wan QC, Li G. Effects of vertical center well and side
provides another way to find out the feasible heating rate and pro- well on hydrate exploitation by depressurization and combination method with
duction pressure, and the moment of adding or stopping heating. wellbore heating. J Nat Gas Sci Eng 2018;55:154–64.
[25] Sun YH, Su K, Guo W, Li B, Jia R. Experimental study on mechanism of de-
pressurizing dissociation of methane hydrate under saturated pore fluid. China
Acknowledgments Petrol Process Petrochem Technol 2016;18(2):43–51.
[26] Collett TS, Kuuskraa VA. Hydrates contain vast store of world gas resources. Oil Gas
J 1988;96(19):90–4.
This work is financially supported by the National Science and
[27] Yousif MH, Li PM, Selim MS, Sloan ED. Depressurization of natural gas hydrates in
Technology Major Project of China (Grant No. 2016ZX05043005), the berea sandstone cores. J Inclusion Phenom Mol Recognit Chem 1990;8:71–88.
National Natural Science Foundation of China (51876017, 51674050 [28] Tang L, Xiao R, Huang C, Feng Z, Fan S. Experimental investigation of production
and 51506016), and the State Key Research Development Program of behavior of gas hydrate under thermal stimulation in unconsolidated sediment.
Energy Fuels 2005;19:2402–7.
China (Grant No. 2016YFC0801404), which are gratefully acknowl- [29] Phirani J, Mohanty KK, Hirasaki GJ. Warm water flooding of unconfined gas hy-
edged. drate reservoirs. Chem Eng Sci 2009;64:2361–9.
[30] Feng JC, Wang Y, Li XS. Entropy generation analysis of hydrate dissociation by
depressurization with horizontal well in different scales of hydrate reservoirs.
References Energy 2017;125:62–71.
[31] Yang M, Fu Z, Jiang L, Song Y. Gas recovery from depressurized methane hydrate
[1] Song Y, Zhang L, Lv Q, Yang M, Ling Z, Zhao J. Assessment of gas production from deposits with different water saturations. Appl Energy 2017;187:180–8.
natural gas hydrate using depressurization, thermal stimulation and combined [32] Li XS, Zhang Y, Li G, Chen ZY, Wu HJ. Experimental investigation into the pro-
methods. RSC Adv 2016;6:47357–67. duction behavior of methane hydrate in porous sediment by depressurization with a
[2] Feng JC, Wang Y, Li XS, Li G, Zhang Y, Chen ZY. Effect of horizontal and vertical novel three-dimensional cubic hydrate simulator. Energy Fuels 2011;25:4497–505.
well patterns on methane hydrate dissociation behaviors in pilot-scale hydrate si- [33] Wang Y, Li XS, Li G, Huang NS, Feng JC. Experimental study on the hydrate dis-
mulator. Appl Energy 2015;145:69–79. sociation in porous media by five-spot thermal huff and puff method. Fuel
[3] Chen J, Wang YH, Lang XM, Fan SS. Energy-efficient methods for production me- 2014;117:688–96.
thane from natural gas hydrates. Energy Chem (English Ed) 2015;24:552–8. [34] Wang Y, Feng JC, Li XS, Zhang Y, Chen ZY. Fluid flow mechanisms and heat transfer
[4] Yuan Y, Xu T, Xia Y, Xin X. Effects of formation dip on gas production from un- characteristics of gas recovery from gas-saturated and water-saturated hydrate re-
confined marine hydrate-bearing sediments through depressurization. Geofluids servoirs. Int J Heat Mass Transf 2018;118:1115–27.
2018;2018:1–11. [35] Fitzgerald GC, Castaldi MJ, Zhou Y. Large scale reactor details and results for the
[5] Milkov AV. Global estimates of hydrate-bound gas in marine sediments: how much formation and decomposition of methane hydrates via thermal stimulation dis-
is really out there? Earth Sci Rev 2004;66:183–97. sociation. J Petrol Sci Eng 2012;94:19–27.
[6] Lee MW, Collett TS. Pore- and fracture-filling gas hydrate reservoirs in the Gulf of [36] Nair VC, Prasad SK, Kumar R, Sangwai JS, Yan J. Energy recovery from simulated
Mexico Gas Hydrate Joint Industry Project Leg II Green Canyon 955 H well. Mar Pet clayey gas hydrate reservoir using depressurization by constant rate gas release,
Geol 2012;34:62–71. thermal stimulation and their combinations. Appl Energy 2018;225:755–68.
[7] Ruppel C, Boswell R, Jones E. Scientific results from Gulf of Mexico Gas Hydrates [37] Minagawa H, Ito T, Kimura S, Kaneko H, Noda S, Tenma N. Depressurization and
Joint Industry Project Leg1drilling: introduction and overview. Mar Pet Geol electrical heating of methane hydrate sediment for gas production: laboratory-scale

12
S. Liu, et al. Fuel 256 (2019) 115742

experiments. J Nat Gas Sci Eng 2018;50:147–56. [48] Kraemer LM, Owen RM, Dickens GR. Lithology of the upper gas hydrate zone, Blake
[38] Li B, Liu SD, Liang YP, Liu H. The use of electrical heating for the enhancement of Outer Ridge: a link between diatoms, porosity, and gas hydrate. Proceedings of the
gas recovery from methane hydrate in porous media. Appl Energy Ocean Drilling, Program Scientific Results. Texas A&M University; 2000. p. 229–36.
2018;227:694–702. [49] Yang S, Zhang H, Wu N, Su X, Schultheiss P, Holland M, et al. High concentration
[39] Li XS, Wang Y, Duan LP, Li G, Zhang Y, Huang NS, et al. Experimental investigation hydrate in disseminated forms obtained in Shenhu Area. North Slope of South China
into methane hydrate production during three-dimensional thermal huff and puff. Sea 2008.
Appl Energy 2012;94:48–57. [50] Li B, Li XS, Li G, Feng JC, Wang Y. Depressurization induced gas production from
[40] Li G, Li XS, Wang Y, Zhang Y. Production behavior of methane hydrate in porous hydrate deposits with low gas saturation in a pilot-scale hydrate simulator. Appl
media using huff and puff method in a novel three-dimensional simulator. Energy Energy 2014;129:274–86.
2011;36:3170–8. [51] Li B, Liang YP, Li XS, Wu HJ. Numerical analysis of methane hydrate decomposition
[41] Wang Y, Li XS, Li G, Zhang Y, Li B, Feng JC. A three-dimensional study on methane experiments by depressurization around freezing point in porous media. Fuel
hydrate decomposition with different methods using five-spot well. Appl Energy 2015;159:925–34.
2013;112:83–92. [52] Feng JC, Wang Y, Li XS, Li G, Zhang Y. Three dimensional experimental and nu-
[42] Li XS, Wang Y, Li G, Zhang Y, Chen ZY. Experimental investigation into methane merical investigations into hydrate dissociation in sandy reservoir with dual hor-
hydrate decomposition during three-dimensional thermal huff and puff. Energy izontal wells. Energy 2015;90:836–45.
Fuels 2011;25:1650–8. [53] Li B, Li XS, Li G. Kinetic studies of methane hydrate formation in porous media
[43] Li B, Liang YP, Li XS, Zhou L. A pilot-scale study of gas production from hydrate based on experiments in a pilot-scale hydrate simulator and a new model. Chem Eng
deposits with two-spot horizontal well system. Appl Energy 2016;176:12–21. Sci 2014;105:220–30.
[44] Wang Y, Feng JC, Li XS, Zhang Y. Experimental investigation of optimization of well [54] Li B, Li X-S, Li G, Wang Y, Feng J-C. Kinetic behaviors of methane hydrate formation
spacing for gas recovery from methane hydrate reservoir in sandy sediment by heat in porous media in different hydrate deposits. Ind Eng Chem Res 2014;53:5464–74.
stimulation. Appl Energy 2017;207:562–72. [55] Wan QC, Si H, Li B, Li G. Heat transfer analysis of methane hydrate dissociation by
[45] Feng JC, Wang Y, Li XS, Li G, Zhang Y, Chen ZY. Production performance of gas depressurization and thermal stimulation. Int J Heat Mass Transf 2018;127:206–17.
hydrate accumulation at the GMGS2-Site 16 of the Pearl River Mouth Basin in the [56] Fan X, Zhang Q, Wang W, Li M. Experimental study on the mechanical character-
South China Sea. J Nat Gas Sci Eng 2015;27:306–20. istics of coal around airtight borehole wall in coalbed methane reservoir. Disaster
[46] Li G, Li XS, Li B, Wang Y. Methane hydrate dissociation using inverted five-spot Adv 2013;6:161–7.
water flooding method in cubic hydrate simulator. Energy 2014;64:298–306. [57] Jamaluddin A, Kalogerakis N, Bishnoi P. Modelling of decomposition of a synthetic
[47] Wang Y, Li XS, Li G, Zhang Y, Li B, Chen ZY. Experimental investigation into me- core of methane gas hydrate by coupling intrinsic kinetics with heat transfer rates.
thane hydrate production during three-dimensional thermal stimulation with five- Can J Chem Eng 1989;67:948–54.
spot well system. Appl Energy 2013;110:90–7.

13

Vous aimerez peut-être aussi