Mingjun Yang,Zhanquan Ma,Yi Gao,Lanlan Jiang*

Key Laboratory of Ocean Energy Utilization and Energy Conservation of Ministry of Education,Dalian University of Technology,Dalian 116024,China

Keywords:Methane hydrate Depressurization Thermal stimulation Dissociation characteristics

ABSTRACT Methane hydrate is considered as a potential energy source in the future due to its abundant reserves and high energy density.To investigate the influence of initial hydrate saturation,production pressure,and the temperature of thermal stimulation on gas production rate and cumulative gas production percentage,we conducted the methane hydrate dissociation experiments using depressurization,thermal stimulation and a combination of two methods in this study.It is found that when the gas production pressures are the same,the higher the hydrate initial saturation,the greater change in hydrate reservoir temperature.Therefore,it is easier to appear the phenomenon of icing and hydrate reformation when the hydrate saturation is higher.For example,the reservoir temperature dropped to below zero in depressurization process when the hydrate saturation was about 37%.However,the same phenomenon didn't appear as the saturation was about 12%.This may be due to more free gas in the reservoir with hydrate saturated of 37%.We also find that the temperature variation of reservoir can be reduced effectively by combination of depressurization and thermal stimulation method.And the average gas production rate is highest with combined method in the experiments.When the pressure of gas production is 2 MPa,compared with depressurization,the average of gas production can increase 54%when the combined method is used.The efficiency of gas production is very low when thermal stimulation was used alone.When the temperature of thermal stimulation is 11°C,the average rate of gas production in the experiment of thermal stimulation is less than 1/3 of that in the experiment of the combined method.

1.Introduction

Natural gas hydrate(NGH)is widely distributed in permafrost regions and offshore areas where the conditions of steady state hydrate are satisfied(high pressure,low temperature)[1,2].NGH may exist in the depth of between 130 and 1100 m in permafrost regions and between 800 and 4000 m in oceanic continental margin.The amount of methane stored in marine sediments containing NGH is huge.Up to now,more than 230 potential deposits of NGH have been identified worldwide,and the estimated global NGH reserves are estimated from 3114 to 7.634×106TCM[3,4].And it is estimated that there are 2.0×1014to 1.2×1017m3(STP)CH4in the gas hydrate reservoir all over the world[5].And it is widely believed that the total carbon content in NGH is more than twice of all conventional fossil fuels[6,7].If 1 m3of methane hydrate decomposes at standard pressure and temperature,it can produce 164 m3of natural gas and 0.8 m3of water[1].Natural gas hydrate is generally considered as a potential energy sources for future due to its high energy density and abundant reserves[8-10].Although natural gas hydrates contain huge energy resource,there are still great challenges in commercial exploitation.

Now there is an innovative technology for natural gas storage and transportation.It is the use of gas hydrates because of its huge gas storage capacity.This technology is also known as natural gas solidification technology.It is a safe and environmentally friendly natural storage model by storing large amounts of natural gas in compact hydrate crystals[11].And the technology of natural gas storage and transportation can also be used to store hydrogen[12].In addition,hydrate technology has been widely used in carbon dioxide storage,pipeline flow safety,seawater desalination,gas separation,and so on[13-16].

At present,commercial exploitation technologies for natural gas hydrate are still under development.Common methods for hydrate exploitation are:(1)Depressurization,where the hydrate reservoir pressure falls below the phase equilibrium pressure to decompose the hydrate[17].(2)Thermal stimulation,where the hydrate reservoir is heated above the phase equilibrium temperature to decompose the hydrate[18,19].(3)Chemical inhibitor injection,where a strong hydrogen bond chemical(such as methanol or ethylene glycol)is injected into the reservoir to change the hydrate phase equilibrium conditions and induce hydrate dissociation [20,21].(4)Carbon dioxide replacement,where carbon dioxide is injected into the hydrate reservoir to replace methane[22].Considering the feasibility and economic problems of exploit different types of hydrate reservoirs,depressurization or depressurization combined with other methods is considered as the most efficient exploitation method[23].Chen et al.[24]used a visual method to analyze the effect of water migration on hydrate dissociation from the viewpoint of chemical potential difference,and proposed a new method to improve hydrate exploitation—water flow erosion.

At present,depressurization is the most widely used in methane hydrate on-site mining and laboratory simulation.In 2013,Japan conducted the world's first sea hydrate test mining in the Eastern Nankai Trough,Japan.The production method used in the test mining was the depressurization method.And the depressurization method was also used in at the second offshore gas production test site in Japan in 2017.In the same year,hydrate test mining method in Shenhu area of the South China Sea is depressurization[25,26].Wang et al.[27]used the depressurization method to study the dissociation of hydrate in three kinds of high-pressure reactors with different internal volumes.And the scaling criteria for decomposing hydrates by depressurization method were also developed,and the experimental results were scaled by this modified scaling standard to predict the production of natural gas from off-site hydrate reservoirs.In addition,they also generate hydrate samples in cubic hydrate simulator(CHS)and small-scale hydrate simulator(PHS)based on the condition of South China Sea,and carried out a single-level well hydrate dissociation experiment and established the entropy production model.It is found that the changes of pressure,temperature,gas production and water production in the hydrate dissociation process of different scale hydrate reservoirs are similar,and the entropy production of the mixed gas release stage is the largest.During the dissociation gas release phase,the entropy production ratio is significantly reduced with the increase of hydrate reservoir size[28].Zhao et al.[23]used three different gas production pressures to study the behavior of methane hydrate dissociation in three porous media.The natural gas production process can be mainly divided into three stages:free gas release,hydrate dissociation maintained by reservoir sensible heat and hydrate dissociation driven by environmental heat transfer.They also analyzed the hydrate reformation and ice formation during hydrate dissociation,and also analyzed the Stefan(Ste)number and dissociation rate constant.The hydrate reformation seriously affected the efficiency of methane production.For this,Wang et al.[29]compared formed/reformed modes in hydrate exploitation and analyzed the effects of residual water,residual methane hydrate and methane flow rate on the hydrate reformation.Yang et al.[30]simulated two types of NGH deposits by forming methane hydrate in a porous medium with gas excess or water excess.They used magnetic resonance imaging(MRI)to monitor the distribution of liquid water and quantified the amount of methane hydrate during the methane hydrate dissociated by depressurization method.Zhang et al.[31]performed methane hydrate dissociation at different depressurization ranges and depressurization rates,and directly observed the hydrate dissociation process using magnetic resonance imaging(MRI).It was found from experiments that for a given range of depressurization,the average rate of dissociation increases with the increase of depressurization rate.For a given depressurization rate,the average rate of dissociation increases with the increase of decompression range.The effects of ice formation and hydrate reformation on exploitation are also analyzed.Zhang et al.[32]study the gas and water production characteristics using depressurization method during the hydrate dissociation process in the hydrate sediment temperature at 281.5 K.The study found that higher bottom-hole pressure (corresponding to lower dissociation driving forces)leads to slower gas production and lower water production.Moridis et al.[33]emphasize that the relative permeability of hydrate deposits is a complex process that plays a key role in the production of gas from depressurization-induced class 1 and class 2 hydrate reservoirs.Yang et al.[34]conducted the formation and decompression experiments of methane hydrates in class 1 and class 2 sediments using magnetic resonance imaging,and studied the effects of liquid water and heat transfer on hydrate dissociation.In addition,Gao et al.[35]used the method of depressurization to conduct the hydrate dissociation experiments of class 2 and class 3 methane hydrate reservoirs in the case of excess water,and analyzed the dissociation characteristics of the hydrates in the two types of reservoirs under different back pressures.

Although the mining efficiency is higher and the cost is lower for depressurization alone method,the phenomenon of icing and hydrate reformation tends to occur and affects the production efficiency.At present,researchers at home and abroad put the focus of hydrate mining research on the depressurization combined with other mining methods.Wang et al.[36]used the three methods of depressurization,thermal stimulation and a combination of two methods in PHS to study the dissociation and gas recovery of hydrate in accumulation layer of gas excess or water excess.Besides,the fluid flow mechanism and heat transfer characteristics of hydrate dissociation in different hydrated reservoirs were studied.Nair et al.[37]used the methods of depressurization,thermal stimulation and inhibitor injection to study the recovery of methane in two different types of clay deposits,thus providing insights into the exploitation of hydrate in clay reservoir.Feng et al.[1]conducted experiments on warm water stimulation to promote hydrate dissociation in CHS and PHS with dual horizontal wells,and studied the effect of reservoir size on hydrate dissociation.The results show that the hydrate simulators of different scales have little effect on temperature variation during the depressurization stage,but the effects during the constant pressure phase are significant.Wang et al.[38]studied the effects of thermal stimulation on gas production characteristics in depressurization experiments by means of environmental heat transfer and hot water injection.The gas production process is divided into five processes in detail so as to study the heat consumption in the hydrate dissociation process,and then they explained the importance of external thermal stimulation during the dissociation of hydrate.At present,the research on the dissociation of methane hydrate focuses on the depressurization method.But there are few comparative studies on the gas production characteristics in depressurization,thermal stimulation and a combined of two methods.There are also few studies on dissociation characteristic in different hydrate saturation reservoirs and in different height reservoirs.In this study,we compared the gas production characteristics of hydrate dissociation at hydrate saturation of 12%and 37%,respectively.We also conducted experiments at 1.0 MPa,1.5 MPa and 2.0 MPa,and 11°C,16°C.In this study,we analyzed dissociation characteristic with different methods under different conditions.

2.Experimental

2.1.Experimental apparatus and materials

Fig.1.Schematic of the experimental system.

The hydrate formation and dissociation experimental system in this study is shown in Fig.1.The experimental system mainly consists of the following components:hydrate formation system,hydrate dissociation system,temperature control system and data acquisition system.The hydrate formation system is mainly composed of a high-pressure reactor,a vacuum pump,a methane injection pump,and a water injection pump.The high-pressure reactor is made of 316 stainless steel and has a maximum sustain pressure up to 15 MPa.The internal height and diameter are 120 mm and 103 mm respectively.The effective volume of this reactor is 1.0 L,and there is a cooling jacket outside the reactor connected with a cooling water bath.According to the scaling model stated by Wang et al.[39,40],we can simulate real-site mining using this reactor.The methane injection pump is a high precision piston pump(ISCO 500D,volume:507.38 ml,pressure range:0.06895 to 25.86 MPa,ambient temperature range:5 to 40°C).The hydrate dissociation system is mainly composed of a PID(proportion integral derivative)controller,a back-pressure control device,a gas-liquid separation tank,and a buffer gas tank.The PID controller is manufactured by OMEGA to set the back pressure of back pressure control device.The back-pressure control device is produced by EMERSON,model 51111S,for precise control of reactor back pressure.The gas-liquid separation tank and the buffer gas tank are also made of 316 stainless steel.The gas-liquid separation tank is filled with dry silica gel particles for absorbing the moisture contained in the dissociation gas,where the dissociation gas enters from the bottom of the gas-liquid separation tank and goes out from the top,then entering the buffer gas tank.It is designed to ensure that the moisture in the dissociation gas can be completely left in the gas-liquid separation tank.The temperature control system consists of a circulating cooling water bath and matching pipelines.The water bath is a temperature-controlled water bath of the model AP20R-3C-A12Y produced by Poly-Science,and the temperature control accuracy is 0.1°C.The data acquisition system consists of thermocouples,pressure gauges (ROSEMOUT company),a balance(Changzhou Wantai Tianping company,CN8DPT-305),and data acquisition modules(Beijing Advantech Electronics Technology company).The thermocouples are T-type 6-point thermocouples produced by Dalian Keda Instrument Co.,Ltd.,and three thermocouples are evenly distributed in the reactor.Each thermocouple has six temperature measuring points.There are 18 temperature measurement points totally,which divide the reactor into 7 layers.The type of quartz sand used in the experiment was BZ-01(AS-ONE；diameter 0.1 mm；porosity 0.37).The methane used in the experiment was produced by Dalian Special Gas Co.,Ltd.,and the purity was 99.999%.

Methane gas is injected from the top of the reactor,and the dissociated gas and water leave the reactor to the gas-liquid separation tank from the top side.The injection and the outlet tube are equipped with a filter to effectively prevent the porous medium from entering the pump and the back-pressure control device.The reactor is equipped with a vertical well extending to the middle of the rector for injecting warm water.

2.2.Experimental procedures

2.2.1.Hydrate formation process

The experiment is mainly divided into hydrate formation process and hydrate dissociation process.The experimental parameters are shown in Table 1.

Table 1 Parameters of hydrate formation and dissociation process

P0is the pressure of the reactor before the hydrate formation；Pendand Tendis the pressure and the average temperature of the reactor after the formation of hydrate respectively；Vais the amount of methane gas consumed for hydrate formation under standard conditions.Where Swi,Shiis the initial saturation of water and hydrate,respectively；P is the pressure of gas production；Tbis the temperature of water bath；R is the average rate of gas production under standard conditions；Vbis the amount of gas production.

The experimental steps are as follows:(1)The reactor is cleaned and the glass sand is thoroughly mixed with deionized water.(2)The reactor is filled with the mixture of quartz sand and deionized water,and the mixture was uniformly compacted in the reaction vessel.The water content of the mixture is 40%,so that the initial water saturation after the hydrate formation is kept constant.(3)Connect the devices and check the air tightness of the system.(4)Vacuum the vacuum pump for 30 min.(5)Inject methane to the specified pressure and then turn on the water bath to cool the rector down.A detailed description can be found in the article by Song et al.[41].The ambient temperature during the experiment was 25°C.We use two methods to form hydrate:constant volume and constant pressure.In the constant volume method,methane is injected into the reactor up to 6.2 MPa,then close all valves,open the water bath,set the water bath temperature to 2.5°C.When the hydrate is formed,the temperature in the reactor rise sharply,and the pressure drops sharply.When the temperature pressure tends to be steady,the hydrate formation is considered to finish.In the constant pressure method,methane is injected under 6.2 MPa.Turn off all valves except valve-5 and then turned on the water bath.When the hydrate is formed,we close the valve-5,and the amount of methane reduction in the ISCO pump is the amount of hydrate formation.The ISCO pump was connected with computer to record its date in real time,such as remaining volume,flow rate and so on.The amount of methane gas injected can be calculated from the remaining volume of the ISCO pump.And the temperature of the injected methane is 2.5°C.It has been found that the hydrate initial saturations formed by the constant pressure method are higher than those of the constant volume method,because the constant pressure method has a larger driving force during the hydrate formation process.

2.2.2.Hydrate dissociation process

Three hydrate dissociation methods were used in the experiments:depressurization,thermal stimulation and a combination of them.

The temperature of the water bath kept constant when the hydrate was dissociated by the depressurization method.The gas production pressures were 2.0 MPa,1.5 MPa and 1.0 MPa,respectively,and both pressures were lower than the phase equilibrium pressure at 3.6°C.The back pressure of the reactor is controlled by adjusting the backpressure control device.When we reduce the pressure of rector,the gas-liquid mixture first enters the gas-liquid separation tank for separation.Then the dry gas enters the buffer gas tank,and the amount of gas production was studied by the temperature and pressure parameters of the buffer tank.

In the thermal stimulation experiments,water bath temperature increase into 11°C and 16°C.The specific experimental method is to set the back-pressure control device to 6.0 MPa,then open the valve to release part of the free gas.This pressure is higher than the phase equilibrium pressure at the corresponding temperature without hydrate dissociation.Later,we adjust water bath temperature at specified temperature,such as 11°C or 16°C.When the pressure in the buffer tank tends to be stable,the dissociation process can be considered as completed.

When the hydrate is dissociated by depressurization combined with thermal stimulation,the operation method is almost the same as that of thermal stimulation method.The difference is that the pressure of the hydrate reservoir is reduced to 2 MPa which is lower than the equilibrium pressure of the hydrate phase.Then we increase the water bath temperature to 11°C and 16°C.

3.Results and Discussion

The hydrate dissociation experiments with different gas production pressures and different water bath temperatures were conducted in this study.The temperature and pressure variation through the hydrate reservoir,the cumulative amount of gas production and consumption,and the rate of gas production are analyzed during the hydrate dissociated.Three methods of depressurization,thermal stimulation and combined method are compared in the aspect of reservoir temperature variation,amount of gas production and average rate of gas production.And the hydrate dissociation characteristics in different locations can be studied through the 6-point thermocouple in the reservoir.From this study,we compared the impact of gas production pressure and thermal injection temperature on methane hydrate dissociation.Run1-Run9 studies the hydration decomposing characteristics of hydrates in the case of excess gas,and almost no water is produced during dissociation.

3.1.Hydrate dissociation by depressurization

In the hydrate formation experiment,the initial pressure of the reservoir was about 6.2 MPa after the gas injection was completed.The constant volume method and the constant pressure method were used to promote the formation of hydrate.We can get different hydrate initial saturations through these two different methods to study the gas production behavior at different production pressures and different initial saturations of the hydrate.The initial saturation of the hydrate was about 12%when we used the constant volume method in the experiment.However,the initial saturation of the hydrate formed by the constant pressure method was up to 37%.It may be due to the greater driving force in the process of hydrate by constant pressure method.When the hydrate is formed by the constant volume method,the pressure in the reactor is reduced,which leads to the need for a lower temperature to form hydrate.Therefore,the hydrate initial saturation formed by constant pressure method is higher than that formed by constant volume method.

Fig.2.Temperature variation of hydrate reservoir in depressurization process.

Fig.3.The cumulative amount of gas production in depressurization process.

The temperature variation during dissociation in the reservoir at different hydrate saturations and different gas production pressures is shown in Fig.2.It can be seen from the figure that the temperature drops sharply during the initial stage of depressurization,it is mainly because the free gas passes through the porous media and induce the Joule Thompson effect.On the other hand,it may be caused by part of hydrate dissociation,where the hydrate dissociation is an endothermic reaction process.Dissociation of hydrates during the depressurization process consumes the sensible heat in the reservoir,causing a sudden drop of reservoir temperature.Then the heat from the boundary is transferred to the middle of the reservoir for hydrate dissociation and causes the reservoir temperature rises slowly [42,43].It can also be seen from Fig.2 that the greater the pressure drop,the greater the range of temperature change of the hydrate reservoir.When the gas production pressures are the same,the higher the initial saturation of hydrate,the greater the change in hydrate reservoir temperature.When the initial saturation of the hydrate is about 12%and the gas production pressure is 1.0 MPa and 2.0 MPa,the reactor initial temperature decreased to 0.2°C and 0.4°C,respectively.While when the initial saturation of the hydrate is about 37% and the gas production pressure is 1 MPa,1.5 MPa and 2.0 MPa,the reservoir initial temperature decreased to-1.7°C,-1.8°C and-1.7°C,respectively.It can be seen from Figs.3 and 4 that when the gas production pressure is the same,the lower the hydrate initial saturation,the shorter the time taken for the complete dissociation of the hydrate.And when the gas production pressure is the same,the higher hydrate initial saturation caused the higher cumulative amount of gas production.In the experiments of Run1 and Run3,the gas production pressures are both 2.0 MPa.The times taken for gas production are about 50 min and 150 min with the hydrate initial saturations of 12%and 37%in Run1 and Run3.And the cumulative amounts of gas production are 13 and 36 L respectively under the standard conditions.When the hydrate initial saturation is the same,the lower gas production pressure is corresponding to the shorter time for the complete dissociation of hydrate.And when the hydrate initial saturation is the same,the lower the gas production pressure,the higher the cumulative amounts of gas production.Comparing the experiments of Run3,4 and 5,the times taken for gas production are about 150 min,75 min and 50 min with the gas production pressure of 2.0 MPa,1.5 MPa and 1 MPa,respectively.It can be seen from Fig.5 that the dissociation rate of hydrate is fast in the first 10 min of the production process,and the maximum gas production rate of the Shi=37%is higher than that of the Shi=12%.Because there is more free methane in the reservoir when the hydrate initial saturation is 37%.As shown in Fig.2,when the gas production pressure is 2 MPa and the hydrate saturation is 36.13%(Run3),the phenomenon of ice formation is obviously observed with the reservoir temperature rising sharply after 5 min of gas production.Wang et al.[44,45]proposed that when the hydrate is dissociated,the water in reservoir will freeze,which will accelerate the dissociation of hydrate,and the lower the gas production pressure,the faster the ice generation rate and the higher the hydrate decomposition rate.

Fig.4.The volume of gas production consumption and the recovery rate of gas in depressurization process.

By comparing Figs.2-4,it can be found that the process of methane production can be divided into three stages:free gas release,hydrate dissociation and constant pressure [23].The free gas release stage is the time before the pressure drop to the hydrate equilibrium pressure,and it is very short.This stage is accompanied by a sharp drop in pressure and temperature,a high gas production rate,and a steep cumulative amount of gas production curve.In the hydrate dissociation stage,the hydrate is dissociated along the phase equilibrium curve.Compared with the first stage,the temperature and pressure change slightly in this stage.This stage is an endothermic reaction,and the sensible heat of the reservoir is heat from water bath.There may be the phenomenon of ice and hydrate reformation in this stage.At last,pressure and temperature tend to be stable without gas production.

3.2.Hydrate dissociation by thermal stimulation and combined method

In Run6-9,methane hydrate was formed under the constant pressure.And the initial saturations of hydrate are 36.98%,37.92%,37.16%,and 33.43%,respectively.Hydrate was dissociated by thermal stimulation in Run6 and 7,depressurization combined with thermal stimulation in Run8 and 9.And it can be seen from Fig.6 that the reservoir temperature goes faster at the beginning than the later in Run6 and 7.That is because the large initial temperature difference between reservoir and water bath.Another reason is that a majority of hydrate dissociate when the reservoir temperatures reach up to about 8°C.

Fig.5.Rate of gas production by depressurization.

Fig.6.Variation of temperature and pressure in hydrate reservoir by thermal stimulation and combined method.

While in Run8 and 9 most of the hydrate immediately dissociate after depressurization stage.It can be seen from Fig.6 for Run6 and 7,little hydrate begins to dissociate when the temperature increase to about 4.5°C,meanwhile the reservoir pressure suddenly rises at the beginning of dissociation.The change of temperature and pressure in Run8 and 9 is similar to Run3 during the hydrate dissociation.The cumulative gas production amounts of Run8 and Run9 are identical before 14 min of hydrate dissociation.That indicates that the heat of thermal is not used to dissociate the hydrate but to melt the ice[45].For hydrate production,the pressure of gas production has a great relationship with the gas recovery rate(the ratio of produced gas to experimental gas consumption).The higher the gas production pressure,the lower the gas recovery rate,which can be seen from Fig.7.It can be seen from Fig.8 that the higher the thermal stimulation temperature,the higher the hydrate maximum dissociation rate when the hydrate is dissociated by simple thermal stimulation.When the hydrate is dissociated using different methods,the hydrate dissociation rate rises to the maximum in the first few minutes and then gradually decreases to the final production rate in the remaining time[46].

Fig.7.The volume of gas production and consumption and the recovery of rate of gas by thermal stimulation and combined method.

It can also be seen from the Fig.9 that the hydrate dissociation time of Run6-9 are about 200,125,115 and 90 min,respectively dissociation.And the initial hydrate saturation of Run7 is higher than that of Run6.This indicates that the higher the hydrate reservoir temperature,the higher the average rate of hydrate.The average rates of the gas production for Run6-9 are 0.11,0.18,0.37 and 0.43 L·min-1under the standard conditions with the reservoir temperature of 11°C,16°C,11°C,16°C.It can be seen from the Figs.6 and 9 that the reservoir temperature,pressure change and gas production rate are very similar for Run8 and Run9.It proves that depressurization plays a vital role in the experiment of depressurization combined with thermal stimulation.Due to the higher hydrate saturation of Run8,the gas production rate is slightly higher than that of Run9 within 14 min after depressurization and thermal stimulation.But the dissociation rate of Run9 is higher than Run8 after 14 min of depressurization and thermal stimulation.That is because that reservoir temperature of Run9 is higher than Run8.It proves that thermal stimulation plays a vital role in late stage of the experiment of depressurization combined with thermal stimulation.In summary,the effect of depressurization is more obvious on the gas production during the whole process of depressurization combined with thermal stimulation.

3.3.Dissociation characteristics with different methods

When methane hydrate dissociates,the gas production rate of the reservoir has a great relationship not only with the hydrate saturation and gas production pressure,but also with the dissociation method.As shown in Fig.10,the hydrate dissociated using depressurization,thermal stimulation and a combined of two methods in Run3,6 and 8,respectively.The gas production pressures of Run3,6 and 8 are 2 MPa,6 MPa and 2 MPa,respectively.And the thermal stimulation temperature used for Run6 and Run8 is 11°C.It can be seen from Fig.11 that the average rate of gas production and the cumulative amount of gas production are the highest for the combined method (Run8).When we use the thermal stimulation method (Run6)alone,the average rate of gas production and the cumulative amount of gas production are far below those of other methods.However,comparing Run3 with Run8 in the aspect of the average gas production rate and the cumulative amount of gas production,it can be found that the difference between them is small.

However,in addition to considering the average gas production rate and the cumulative amount of gas production,we should also take into account other factors in the evaluation of mining method,such as ice formation and hydrate reformation.It can be seen from Fig.10 that the Joule Thompson effect caused by the release of free gas at the initial time(0-5 min)causes the reservoir temperature dropped sharply.Although the gas production pressures are the same,the temperature variation of simply using depressurization(Run3)is slightly larger than that of using the combined method (Run8).Besides,when a large amount of hydrate is dissociation(5-50 min),compared with depressurization,the hydrate reservoir temperature is significantly increased when the hydrate dissociated by the combined method.The reason is that the sensible heat of the surrounding reservoir is the only source of heat when the hydrate dissociated by depressurization method alone.However,in addition to the sensible heat of the reservoir,the external thermal stimulation is also the source of heat when using the combined method.Therefore,ice generation and hydrate reformation can be effectively reduced when we mine the hydrate using the combined method.

3.4.Temperature variation of different heights in the reservoir

Fig.8.Rate of gas production by thermal stimulation and combined method.

Temperature variation of different heights is studied by three 6-point thermocouples,and the distribution of the thermocouples in the reservoir is shown in Fig.12.The temperature change of the T2 thermocouple during the hydrate formation process in the Run8 were shown in Fig.13.The temperature rise of 650 min-800 min is due to the formation of hydrate.It can be seen from this figure that the temperature difference between the six temperature measurement points is small.And the temperature of T21 and T22 is slightly higher,which may be due to the poor cooling effect at the top and bottom of the reservoir.While it can be seen from the figure that the temperature rise of the temperature measurement points T23,T24 and T25 in the middle of the reservoir is larger when the hydrate is formed,indicating that the amount of hydrate formation in the middle of the reservoir is larger.The temperature variation curves of different heights in the reservoir are shown in Fig.13 during the hydrate dissociation.It can be seen from Fig.13 that the temperature at the locations of T13,T23 and their vicinity rise relatively later in the experiments,which due to the hydrate dissociation here.Therefore,we can speculate that hydrate tends to grow in the middle of the reservoir,and the amount of hydrate formed at the top and bottom of the reservoir is less.The temperature of the hydrate reservoir at different heights of the thermocouple T2 rises a bit later than T1,which indicates that the hydrate saturation at the thermocouple T2 is higher.This may be due to the fact that there is a vertical well extended into the middle of the reactor,and the ends of the vertical well is wrapped with gauze for sand prevention.Therefore,hydrate formation in the rector is unevenness,and the gauze can promote the growth of the hydrate around it.The higher the heat stimulation temperature,the faster the hydrate decomposes,which can also be seen from Fig.14.When the thermal stimulation temperature is 11°C,the hydrate reservoir temperatures begin to rise gradually after 50 min.While the temperatures of the hydrate reservoir rise 10 min early,and the slopes of the temperature rise are larger when the thermal stimulation temperature is 16°C.

Fig.9.The cumulative amount of gas production by thermal stimulation and combined method.

4.Conclusions

Fig.10.Temperature variation of hydrate reservoir by different dissociation methods.

Fig.11.Comparison of gas production amount and average gas production rate by different dissociation methods.

Fig.12.The distribution of the thermocouples in the reactor of(A)front view；(B)top view；(C)three-dimensional figure.

Fig.13.Temperature variation of T2 thermocouple during hydrate formation by combined method.

In this work,the dissociation characteristics of methane hydrate by depressurization combined with thermal stimulation were studied.Both constant volume method and constant pressure method was used to form the hydrate sample with different initial saturations,and the hydrate dissociation character was studied at the gas production pressure of 2 MPa,1.5 MPa,1 MPa and thermal stimulation temperature of 11°C,16°C.The following conclusions can be drawn from this study.

(1)It is found that the hydrate saturation formed by the constant pressure method is much higher than that formed by the constant volume method when the initial conditions for hydrate formation are the same.The reason is that the constant pressure method has a larger driving force.And it is more suitable for the formation of hydrates in the actual conditions of the ocean sediment.

(2)When the initial hydrate saturation is the same,the lower the gas production pressure,the faster the gas production rate of the hydrate reservoir,and the higher the hydrate reservoir temperature,the faster the average hydrate dissociation rate.When the pressure of gas production is 2 MPa,compared with depressurization simply,the average of gas production can increase 54%when the combined method is used.The efficiency of gas production is very low when thermal stimulation was used alone.When the temperature of thermal stimulation is 11°C,the average rate of gas production in the experiment of thermal stimulation is less than 1/3 of that in the experiment of the combined method.

Fig.14.Temperature variation at different reservoir locations when the hydrate is dissociated with combined methods for(A)P=2 MPa,Tb=11°C,Thermocouple:T1；(B)P=2 MPa,Tb=11°C,Thermocouple:T2；(C)P=2 MPa,Tb=16°C,Thermocouple:T3；(D)P=2 MPa,Tb=16°C,Thermocouple:T2；

(3)When the hydrate dissociated by depressurization combined with thermal stimulation,pressure reduction is the main effect at the initial stage,while temperature of thermal stimulation has a greater influence at the late stage of the gas production.

(4)When the gas production pressures are the same,the higher the initial saturation of hydrate,the greater the change in hydrate reservoir temperature.Therefore,it is easier to appear the phenomenon of icing and hydrate reformation when the hydrate saturation is higher.This may be due to more free gas in the reservoir with hydrate saturated of 37%.