New insights into the dissociation of mixed CH4/CO2 hydrates for CH4 production and CO2 storage

Recent but limited studies have shown that multistep slow depressurization based on mixed CH 4 /CO 2 hydrate dissociation can enhance CH 4 recovery and increases CO 2 storage after CO 2 injection into CH 4 hydrate [1,2]. For the first time, the resistivity variation and gas recovery and storage variation was investigated to study the change in hydrate saturation and production/storage yield. Lab-scale CH 4 and CO 2 rich mixed hydrates were synthesized to mimic the production and injection well scenario. The mixed hydrates were synthesized in sandstone with moderate to high water saturation using two different CH 4/ CO 2 gas mixtures. Furthermore, mixed CH 4 /CO 2 hydrates were dissociated three to six steps based on cyclic depressurization. Pressure, resistivity and gas chromatography data were collected. The presence of two thermodynamic stability zones provided an opportunity for additional CH 4 recovery and CO 2 storage during mixed hydrate dissociation. Gas and water migration between the injection and production well caused CO 2 hydrate reformation, improvement in CO 2 sweep area and movement of the CO 2 hydrate front toward the production well. Multiple peaks in CH 4 recovery and CO 2 storage suggest major dissociation and reformation. Peak values were independent of mixed hydrate type. Peaks values of CH 4 rich hydrates occurred at high pressure than peak values of CO 2 rich hydrates. The slight change in resistance during depressurization below pure CH 4 hydrate stability pressure confirms the loss of CH 4 hydrate mass recovered by the formation of CO 2 hydrate mass. This study discusses the correlation between the change in resistivity and type of guest molecule and its concentration and initial water saturation. The results of this study will be useful to explore the application of slow depressurization for the dissociation of CH 4 /CO 2 mixed hydrates to improve CH 4 recovery and CO 2 storage.


Introduction
Gas hydrates are ice-like compounds of water and certain gases under high pressure and low temperature. The ice-like crystalline compounds are formed by hydrogen bonding and van der Walls forces [3,4]. Naturally occurring gas hydrate deposits contain large amounts of natural gas and are considered a potential gas supply source [5][6][7]. Attempts are being made to extract gas from these deposits, both in permafrost and in deep oceans. The depressurization based production method is the most discussed and tested [8]. There are many variations of depressurization to optimise gas recovery. These variations include constant rate decompression [9], cyclic decompression [10,11], and slow multistage decompression [12][13][14][15][16]. To overcome the challenges associated with depressurization, hybrid techniques combined with other production techniques have also been proposed, including hybrid with gas injection [17,18] or hybrid with thermal stimulation. The main purpose of the hybrid technique is to avoid geo-mechanical instability, increase heat input to prevent ice or secondary hydrate formation, and enhance CH 4 recovery without excessive water or sand production.
Among hybrid techniques proposed, depressurization combined with CO 2 injection is very promising due to the potential to store CO 2 while producing CH 4 from CH 4 hydrate dissociation. CO 2 rich gas injection into CH 4 hydrate itself is a promising technique that offers CO 2 storage and assists CH 4 gas in hydrate reservoir [19][20][21][22][23][24]. Pilot tests [25,26] showed that the method leads to lower CH 4 recovery and store less CO 2 that did not justify the capital investment in CO 2 transportation, infrastructure in remote locations etc. The main reason for lower yield is the mass transfer barrier created by CO 2 hydrate film [27][28][29] formation at the injection well that reduces gas injectivity and CO 2 sweep area. Gas molecule transport and CO 2 concentration at CH 4 hydrate surface also reduced due to decreased relative gas permeability caused by hydrate film formation. In laboratory-scale experiments, CO 2 injection into CH 4 hydrate would lead to inlet plugging due to CO 2 hydrate formation, resulting in low CH 4 /CO 2 hydrate saturation and heterogeneous hydrate distribution.
The depressurization based hybrid method could be a solution to solve this mass transport problem and improve CH 4 recovery and storage yield [18,27,28]. In a recent preliminary study, CO 2 gas was injected into depressurized methane hydrate [29], while in another study, depressurization was performed after CO 2 injection [30]. Rapid depressurization before CO 2 injection could be risky as it could trigger significant hydrate dissociation that could not be prevented by CO 2 injection later. Morphology study based on CO 2 injection into depressurized CH 4 hydrate shows the water production was probabilistic after pure CO 2 injection, and probability of water production and presence of hydrates depends on CO 2 concentration in injected gas, injection gas pressure and degree of depressurization [29]. In another study, depressurization was carried out before CO 2 injection and depressurization only contributed 10% in overall CH 4 recovery [31]. On the other hand, mixed CH 4 /CO 2 hydrate, when depressurized using stepwise dissociation, showed multiple dissociations and reformation events between CH 4 and CO 2 stability pressure. This phenomenon was exciting as it led to free water consumption in creating CO 2 hydrate, and the risk of water production was reduced [1].
The efficiency of the hybrid depressurization combined with the CO 2 injection technique is dependent on two key factors. First, higher CO 2 injectivity caused a large CO 2 sweep area, and the second is the appropriate depressurization technique. Higher CO 2 injectivity can be improved if CO 2 hydrate film formation can be delayed. A hydrate inhibitor such as low dosage MeOH can improve CO 2 injectivity by delaying CO 2 hydrate film formation at an injection well. MeOH is known to be a thermodynamic hydrate inhibitor when used at concentrations above 30 wt%; however, when used at low or ultralow concentration, it behaves as a promoter [32,33] or an anti-agglomerate additive [34]. Delayed CO 2 hydrate film formation would allow a larger gas-liquid contact area, and a higher sweep area can be achieved. In our recent studies, experimental studies showed a longer induction time during CH 4 hydrate formation in the presence of MeOH compared to water or other promoters [30]. The selection of appropriate depressurization techniques is also another critical variable. It has been shown that the slow depressurization method is more advantageous compared to direct rapid depressurization. Studies show that slow depressurization [12][13][14], including cyclic and stepwise, reduces water and sand production risk, maintains the reservoir temperature [33][34][35], and avoids ice and secondary hydrate production when applied on pure CH 4 hydrate. When cyclic depressurization was applied on CH 4 hydrate, the CH 4 recovery factor improved from 42% to 71% [11]. Moreover, the average production rate increased 17 fold compared to rapid dissociation [10].
In a recent study, we studied cyclic depressurization based on CH 4 / CO 2 hydrate dissociation in bulk and unconsolidated sand, with and without additives [2]. It was observed that when cyclic depressurization was implemented between CH 4 and CO 2 hydrate stability pressures, improvement in CH 4 mole fraction and reduction in CO 2 mole fraction was observed. CH 4 recovery rates were higher for higher CH 4 hydrate saturation and T > 0℃.
At the field scale, CO 2 injection into CH 4 hydrate would initiate the CH 4 /CO 2 mixed hydrate synthesis with low CH 4 recovery and CO 2 storage yield [29]. CO 2 injection into CH 4 hydrate introduces two gas hydrate stability boundaries (pure CH 4 hydrate and pure CO 2 hydrates) and many intermediate stability zones due to the different compositions of CH 4 /CO 2 mixed hydrates. The presence of undissociated CH 4 hydrate and a CO 2 -rich vapour phase presents an opportunity that could be further exploited, such that undissociated CH 4 hydrate melt and released water are mixed with the CO 2 vapour phase triggers CO 2 hydrate reforming. Therefore, further insight into the dissociation of CH 4 /CO 2 mixed hydrates below the CH 4 hydrate stability is needed. In particular, this is needed to confirm CH 4 -hydrate dissociation and CO 2 -hydrate reforming from free pore water, which could improve CH 4 recovery and CO 2 storage without loss of hydrate saturation and sediment strength.
Resistivity variation can be used to study hydrate saturation and hydrate stability. Electrical resistivity logs in conjunction with other logs such as electromagnetic measured are typically used to identify and estimate gas hydrate concentration, distribution, and quantification [35][36][37][41][42][43]. As hydrates are poor conductors of electricity compared to the formation water, hydrate formation would indicate an increase in  3 electrical resistance [38] due to the conversion of formation water into hydrate. Studies show that electrical resistivity is dependent on hydrate saturation, pore water salinity, and hydrate morphology [37,39]. Hydrate conductivity is lower than water [38]. Therefore electrical resistance could be used to identify formation and dissociation and increase in hydrate saturation that suggests a change in physical properties of hydrate-bearing sediments [40][41][42][43]. Studies show that sediment stiffness and electrical resistance follow a similar trend; therefore, increased electrical resistivity would increase hydrate strength and stiffness [44]. Another area of research is the optimization of gas recovery and storage. Optimization is essential to minimize the production time and increase the gas recovery rate. No such previous study is available yet.
This study assumes that replacement has already occurred and the CH 4 /CO 2 mixed hydrate system has been formed. Such systems will be present after injection of CO 2 into CH 4 hydrates as the distribution between injection and production well varies depending on CO 2 sweep efficiency. We have tried to find an answer to the following questions: (1) How do the different CH 4 /CO 2 mixed hydrate systems behave during slow depressurization? (2) How does the occurrence of reformations below the CH 4 hydrate stability pressure affect the resistivity? A change in resistivity would indicate a change in hydrate volume that affects the geo-mechanical stability of hydrate-bearing sediments during pressure release. The two different mixed CH 4 /CO 2 hydrates are formed from two different CH 4 /CO 2 gas mixtures. The slow depressurization technique was chosen to exploit two different stability zones such that the CO 2 gas phase forms CO 2 hydrates while CH 4 hydrate dissociated and released water support CO 2 hydrate formation. Mass balance analysis based on gas chromatography was performed to calculate the recovery and storage yield.

Experimental setup and material
Two different CH 4 /CO 2 gas mixtures were obtained from Air Liquide (Denmark). Sandstone cores were used for the analysis. The sandstone cores have a length of 7.83 cm and a diameter of 2.55 cm. The average porosity of the sandstone used is 0.3, and the average permeability is 50 mD. The details of the gasses during each experiment is in Table 1 in Appendix. A.
The schematic of the experimental setup is shown in Fig. 1. The core holder used in this study is 316 stainless steel, and the maximum working pressure was 200 bar. Sandstone with different initial water saturation was used for the experiment. Before inserting the core sample into the core holder, the core sample was wrapped with heat shrink tubing and a rubber sleeve. An ISCO pump was used to maintain cuff pressure around the core samples. The system was gradually pressurized first with the cuff pressure and later with the gas pressure so that the cuff pressure was 20-30 bar above the gas pressure. The pressure was gradually increased so that the cuff pressure increased from 40 to 60-80-110 bar while the injection pressure increased from 20 to 40-60-80 bar. The pressure at the inlet and outlet was measured using two differential pressure transmitters. A platinum resistance thermometer (Pt 100) was placed at the inlet side to monitor the temperature. A cooling jacket was placed around the core holder connected to the cooling bath, and the temperature was controlled manually. The core holder was modified by inserting the insulator at the outlet to develop resistance along the length of the core holder. The electrical resistance was recorded with two crocodile clips placed before and after the inlet and outlet caps of the high-pressure core holder (see Fig. 1). The electrical resistance was recorded with two alligator clips placed before and after the inlet and outlet caps of the high-pressure core holder (see Fig. 1). The resistance was measured using a data acquisition device. The voltage was 1 V to minimize electrode polarization and electrochemical reactions at the electrode during the measurement. The variation of electrical resistance during formation and the MCD method was recorded. Pressure (P), Temperature (T) and Resistance (R) were recorded using an Agilent data logger. Gas samples were taken at the outlet to study the variation of gas molecules CH 4 , CO 2 in the residual vapour phase using a gas chromatograph (GC7880A).

Experimental procedure & analysis
Mixed gases (CH 4 /CO 2 gas mixture) was used to prepare artificial CH 4 /CO 2 mixed hydrate. For the laboratory-scale experiment, we did not inject CO 2 into the CH 4 hydrate for two reasons: (1) to avoid blockage of the inlet due to the formation of CO 2 hydrate. (2) To obtain two different CH 4 /CO 2 mixed hydrates by using two different gas mixtures.
For the field scale conversion, the CH 4 /CO 2 mixed hydrate would be produced from the CH 4 hydrate system after CO 2 injection. For fieldscale implementation, using a kinetic inhibitor/anti-agglomerate compound is recommended to avoid CO 2 plugging at the injection site. Such chemicals can retard CO 2 hydrate formation around the injection well and improve CO 2 -injectivity and improve the synthesis of CH 4 /CO 2 hydrates [30,45]. After injection of CO 2 from the injection well, CO 2 migrates away from the injection well, and CO 2 migration depends on several factors, such as the kinetics of CO 2 hydrate formation, the relative permeability of CO 2 , and the permeability of the hydrate-bearing sediment. During CO 2 migration, both the CO 2 -water interface and the CO 2 -hydrate interface develop. Both cause the formation of hydrate (pure and mixed). As the distance from the wellbore increases, the CO 2 concentration decreases, resulting in a decrease in the quantity and quality of hydrate formed. Therefore, the hydrate distribution between injection and production wells is expected to be heterogeneous, such that CO 2 -rich hydrates are concentrated in the surrounding closer to the injection well, while CH 4 -rich hydrates are mostly concentrated closer to the production well. Therefore, in our study, we have investigated CO 2 rich and CH 4 rich mixed hydrate dissociation.
Five experiments were conducted to investigate hydrate saturation and optimal storage and recovery yield from CH 4 /CO 2 mixed hydrate dissociation using the MCD technique. The methodology of the experiments is explained in Fig. 2. The core flow experiment was conducted assuming one injection well and one production well.
The results and discussion are divided into two sections. Section 3.1 deals with CH 4 /CO 2 hydrate synthesis using two different gas mixtures, while section 3.2 focuses on mixed CH 4 /CO 2 hydrate dissociation using Multi-step cyclic depressurization (MCD) in a porous medium. Resistivity, P-T and GC analysis data were collected.
In the first stage, mixed CH 4 /CO 2 hydrates were synthesized using two different feed gasses (30% CO 2 and 70% CO 2 ) in sandstone with different initial water saturations. The main advantage of using a CH 4 / CO 2 gas mixture to synthesise mixed CH 4 /CO 2 hydrates is that we can produce different CH 4 /CO 2 mixed hydrate types and improve the overall hydrate saturation and distribution along the core length. In the lab scale study, we did not inject pure CO 2 or use kinetic inhibitors/antiagglomeration agents. We synthesized CH 4 /CO 2 mixed hydrate directly from CH 4 /CO 2 gas mixture and circumvented the problems with CO 2 injectivity.
To achieve a higher CH 4 and CO 2 molar number in the hydrate and a uniform distribution in the pore space between inlet and outlet, heating and cooling cycles between 1 and 25℃ were applied. Using a CH 4 /CO 2 gas mixture to achieve uniform hydrate distribution was better than CO 2 injection into CH 4 hydrates. It was a common observation that pressure differential between inlet and outlet was negligible before dissociation, confirming uniform distribution. Pressure and resistance variations were used to investigate whether water and gas migration occurred during the heating and cooling cycle, leading to dissociation and new formation [46]. Hydrate formation was considered complete when pressure remained constant over time, and no differential improvement in hydrate saturation was observed. At this point, a GC sample was taken to examine the change in moles of CH 4 and CO 2 in the vapour phase, which in turn was used to calculate the moles of CH 4 and CO 2 stored in the hydrate (see Appendix A).
During the study, the resistivity variation was also measured. In each experiment, the resistivity was normalized twice. First, during CH 4 /CO 2 hydrate synthesis, the resistivity value was normalized concerning the initial resistivity value (before formation). The different experiments had different initial resistance values due to the different gas/water distribution caused by pressure, initial water saturation, temperature and confining pressure. Therefore, first normalization (NR1) allowed the comparison of the resistance variations between the different experiments.
The second resistance normalization (NR2) was performed in the MCD method. The resistance change was normalized with respect to the resistance value at the beginning of the MCD method. This was done to measure the increase/decrease in resistivity, which could provide information about the change in hydrate volume and overall hydrate saturation during the MCD method. To apply the MCD method, the outlet valve was opened and closed rapidly so that the pressure fell in the range of 2-6 bar. Before closing the valve, a gas sample was taken to record the mole fraction in the gas phase for the mass balance analysis. The shut-in took 10-20 h to allow gas-water mixing, phase change, and fluid redistribution. Previous studies on the slow, gradual depressurization technique have been limited to pure CH 4 hydrate dissociation. Three types of pressure history curves have been described in the literature when cyclic depressurization was performed on CH 4 hydrates [16]. A stable curve above the CH 4 hydrate stability pressure, a rapid pressure rise curve below CH 4 hydrate stability pressure and a later flatter curve. To the best of the author's knowledge, the pressure response curves during CH 4 /CO 2 mixed hydrate dissociation using the slow depressurization method have not been are studied previously. Previous studies show that parameters such as shut-in period, pore water chemistry and degree of hydrate saturation also affected the pressure response curve [14] and are therefore studied in the current study.
The presence of CH 4 and CO 2 in the hydrate create two thermodynamic stability boundaries (Pure CH 4 hydrate and Pure CO 2 hydrate) within the CH 4 /CO 2 mixed hydrate system. These zones' boundaries depend on the reservoir properties, including pore size distribution, pore water chemistry, reservoir temperature, hydrate morphology, etc. The CH 4 hydrate stability pressure remains above the CO 2 hydrate stability pressure at the hydrate formation temperature (temperature below 280 K). Between the thermodynamic stability limits of CH 4 and CO 2 hydrate, there may be other temporary hydrate stability boundaries controlled by the composition of the CH 4 /CO 2 mixed hydrates. Information on pure stability zones is required to design the slow depressurization scheme. It is advantageous to depressurize the mixed hydrate system so that the pressure is below the CH 4 hydrate phase but above the CO 2 pure hydrate phase. For simplicity, we used the CSMGem-based hydrate stability zones for the bulk-only phase; however, in a porous medium, the stability zones would be different.
There are many variations of the slow depressurization technique, e. g., cyclic, constant pressure, constant rate, etc. Most of these studies are limited to pure CH 4 hydrate dissociation. The application of slow depressurization on CH 4 /CO 2 mixed hydrates is very new, and there are still many unknowns. There is no study comparing the best available option between different variants, and this should be the subject of further research. Our previous publication was based on cyclic depressurization, which involved a controlled pressure drop and an extended shutdown period (20 h) [2]. The cyclic pressure release was chosen to allow sufficient time for the reformation to be triggered. There are many unknowns, such as pressure drop, the number of pressure steps, and the total shut-in period. During the shut-in phase, there is continuous mixing of the liquid gas in the pore space. This mixing triggers CO 2 -rich hydrate reforming below the CH 4 hydrate stability pressure.
During MCD, the system pressure moves through different stability pressure zones. When the system pressure is above the CH 4 hydrate stability boundary, only the free gas phase is produced, and all hydrate phases remain stable, including pure CH 4 , pure CO 2 and mixed CH 4 /CO 2 hydrate. Below the CH 4 hydrate stability pressure, pure CH 4 hydrate would dissociate into water and gas. The mixing of released gas and water in the pore space would trigger the formation of CH 4 /CO 2 mixed hydrate and CO 2 hydrate [1,2]. Therefore, below the CH 4 hydrate stability zone, the produced gas phase would be rich in CH 4 . As the pressure moves further away from the CH 4 hydrate stability boundary and approaches the CO 2 -hydrate stability pressure, the unstable mixed hydrate would dissociate, and the CO 2 content in the mixed hydrate would increase.

CH 4 /CO 2 mixed hydrate synthesis
Artificial CH 4 /CO 2 hydrates were formed in sandstone with pure water and two different gasses (Gas A, Gas B). Gas A with 30%CO 2 and gas B with 70% CO 2 resulted in different synthesized CH 4 /CO 2 mixed hydrates (refer to Table 2 in Appendix A). Hydrate formation and distribution affect hydrate-containing sediments' physical properties, including porosity, permeability, tensile strength, stiffness, and resistivity. Fundamental properties such as initial water saturation, initial injection pressure, formation temperature, and CO 2 mole content in the CH 4 /CO 2 gas mixture affect the moles of CH 4 /CO 2 stored in the hydrates.
In general, it was observed that no pressure difference developed between inlet and outlet during hydrate formation. In Exp.3, where pressure difference was developed, there was no correlation between pressure difference (measured at inlet and outlet) and normalized resistance (refer to Fig. 3C). It can be said that the pressure variation was more of a localized phenomenon that may have affected the hydrate distribution in the core but not the resistivity variation. Due to the high water saturation, the hydrate morphology was considered to be porefilling [47]. Using the methodology in Appendix B, the moles of CO 2 and CH 4 stored in the hydrates were calculated. It was found that the ratio of moles of CH 4 /CO 2 stored in the hydrate varied between 1.8 and 2.0, indicating that CH 4 /CO 2 hydrates are CH 4 -rich and are referred to here as type A hydrates.
The NR1 variation in Exp.2 and Exp.3 was different. Exp.2 and Exp.3 had similar pressure drops at the end of the formation (39 bar in Exp.2 and 38 bar in Exp.3). However, in Exp.2, the CO 2 concentration stored in hydrates was higher due to the higher initial injection pressure (91 bar in Exp.2 and 79 bar in Exp.3, see Table A2 in Appendix A). A recent study showed that the conductivity of hydrate crystals depends on the guest molecules. The study shows that pure CO 2 hydrates have higher conductivity (lower resistivity) than pure CH 4 hydrates [48]. Therefore, in Exp.2, the CO 2 hydrate concentration increased during the subsequent cooling and heating cycles, resulting in a lower NR1. (NR1 was 1.1 In Exp.4-5, gas B (70 mol% CO 2 ) was used to form CH 4 /CO 2 hydrates. S wi varied between 54% and 89%, injection pressure varied P inj = 54-70 bar, and formation temperature varied T f = 0.8-− 1.7 ℃. Fig. 4 shows the variation of pressure (inlet and outlet), T (inlet) and NR1 in Exp.4-5. During the experiment, 2-3 heating and cooling cycles were performed. During the last cooling, the temperature was gradually reduced from 4-5℃ to 0-1℃ to promote hydrate formation in the pore space.
Due to the different initial operating pressure, S wi values and water distribution in the pore space, NR1 varied in Exp.4 and 5. The NR1 value was higher in Exp.4 than in Exp.5 due to the higher total moles and CO 2 moles in the hydrates in Exp.4 (see Table A2 in Appendix A). In Exp.4, NR1 value further decreased during the subsequent heating and cooling cycles, which could be attributed to the preferential formation of CO 2rich hydrates over CH 4 -rich hydrates. This could be due to the higher CO 2 solubility in the water phase and the more significant thermodynamic driving force (Higher injection pressure in Exp.4). The discussion of NR1 and its correlation with hydrate mass volume indicate that the resistivity of CH 4 /CO 2 mixed hydrates may be influenced by the nature of the guest molecule and the total hydrate volume. Normalized resistance variation during hydrate formation showed that NR1 increased by 10-14% (in Type A) and 14-19% (in Type B) compared to the initial conditions. The increase in NR1 confirms the presence of hydrate due to the decrease in conductivity caused by the increase in hydrate volume and decrease in pore water volume. The increase in hydrate saturation suggests increased stiffness as resistivity and stiffness follow a similar trend during hydrate formation. Fig. 5 summarises the main results, including the moles (total, CH 4 , CO 2 ) in the hydrates and the resistivity (NR1) in Exp.1-5. For Exp.1-3, the total moles stored in the hydrates were higher than Exp.4-5 due to the higher initial injection pressure. The higher water saturation in Exp.4-5 provides a larger gas-liquid contact area, which accelerated   Fig. 6. shows the variation in P, T, NR2 and XCH 4 / XCO 2 in Exp.1-3 (Type A). A) shows the variation of P, T, NR2 and XCH 4 /XCO 2 in Exp.1, CH 4 stability pressure PCH 4 = 31.7 bar, CO 2 stability pressure PCO 2 = 15.8 bar at T = 2.0℃. CD stage is divided into seven pressure steps (step A to step G). Starting pressure P start = 37.3 bar, and final pressure P finish = 22.1 bar, B) shows the variation of P, T, NR2 and XCH 4 /XCO 2 in Exp.2 CH 4 stability pressure PCH 4 = 30.8 bar, CO 2 stability pressure PCO 2 = 15.2 bar at T = 1.7℃. CD stage is divided into six pressure steps (step A to step F). Starting pressure P start = 52.5 bar, and final pressure P finish = 21.2 bar. C) shows the variation of P, T, NR2 and XCH 4 /XCO 2 in Exp.3 CH 4 stability pressure PCH 4 = 28.4 bar, CO 2 stability pressure PCO 2 = 10.3 bar at T = 0.9℃. CD stage is divided into 5 pressure steps (step A to step E). Starting pressure P start = 41.1 bar, and final pressure P finish = 19.4 bar. higher CO 2 mole stored than CH 4 mole when CO 2 rich gas (Gas B) was used. The average NR1 values for Exp.4-5 were higher than NR1 values in Exp.1-3. It can be said that NR1 variations correlate with S wi values. A similar correlation was observed in our previous study [49]. The correlation between S wi and NR1 also shows that the hydrate distribution is heterogeneous and non-uniform along the core length. The resistance variation in the CH 4 /CO 2 mixed hydrate system depends on the CH 4 and CO 2 moles stored in the hydrates and the initial water saturation.
Type A and type B hydrates are also expected to differ in hydrate morphology. CH 4 -rich type A hydrates are expected to be dominated by hydrate films encapsulating the gas phase, whereas CO 2 -rich type B hydrates are formed in the liquid phase due to CO 2 solubility in the water phase. The correlation between NR1 and total hydrate volume is shown in Table A2 in Appendix A. It can be seen that for type A, an increase in hydrate mass volume correlates with a decrease in NR1, while for type B, an increase in hydrate mass volume correlates with an increase in NR1.

CH 4 /CO 2 hydrate dissociation
Synthesized Type A and Type B hydrates were dissociated using Multi-step cyclic depressurization (MCD) to produce additional CH 4 stored in hydrate while storing CO 2 into hydrates. Typically, MCD included 5-7 steps, including rapid degassing followed by a 10-20 h long shut-in period sequence. (Please refer to Appendix B). Tables 3-5 shows the details of the mixed hydrate composition, the MCD scheme applied, and the gas yield and storage yield in Exp.1-5. Data processing concerning this section is supplied in Appendix. C

Cyclic depressurization of type A hydrates
Exp.1-3 typed A hydrates (rich in CH 4 hydrates) were dissociated by the MCD. (Refer to Tables 4 and 5 in Appendix B). Fig. 6 shows the variation of P, T, NR2 and XCO 2 /CH 4 during the multistep cyclic depressurization based dissociation.
In Exp.1, MCD included 7 pressure drops and shut-in sequences steps (step A to step G). Of the 7 pressure drops, the pressure drop in the last 6 steps varied between 3 and 5 bar (step B to step G). The first pressure drop at the start of step A was 18 bar, which brought the system pressure into a zone between CH 4 and CO 2 stability pressure. During this rapid pressure drop, a pressure differential was created between the inlet and outlet. During the rapid degassing, the Joule-Thomson cooling may have resulted in ice and secondary hydrate formation at the outlet, which may have caused a pressure differential between the inlet and outlet. Later, due to the long shut-in period, sensible heat from the reservoir was sufficient to dissociate all accumulated hydrates and the pressure difference between inlet and outlet decreased. The pressure differential caused fluid movement in the pore space and additional gas-water mixing. No significant change in the CH 4 mole fraction was observed during step A. CH 4 mole fraction at the end of step A was 75.3%. Thereafter, degassing was not rapid between Step B and Step G, and a pressure drop of 2-4 bar on average was achieved before shut-in. During this time, there was no pressure difference between the inlet and outlet. After degassing in each step, the pressure rebounded and remained stable between CH 4 and CO 2 hydrate stability line. At the end of step G, the CH 4 mole fraction was 70.1%, and the CO 2 mole fraction was 29.9%. No NR2 variation was recorded during Exp.1. Fig. 6A show mole fraction variation during Exp.1. The CH 4 mole fraction varied from 70% to 77.6%, with a single peak at the end of Step C (at P = 24.0 bar). RCH 4 varied from 73% to 92.7%, with the first peak of 92.7% at the end of step B (at P = 24.1 bar) and the second peak of 78.2% at the end of step G. The peak CH 4 mole fraction and RCH 4 did not occur at the same pressure steps. It is possible that the gas mole fractions were dependent on the gas/water mixture and formation kinetics, while the recovery factors were related to the molar CH 4 concentration in the hydrates.
During the analysis, SCO 2 increases, reaching a single peak of 83.2% at the end of step D (at P = 22.4 bar). This indicates that the CO 2 concentration in the hydrates was highest at 22.4 bar and further pressure drops attenuated the driving force required for CO 2 hydrate reforming. The SCO 2 peak occurred after the peak in RCH 4 , suggesting that CH 4 was involved in CO 2 -rich hydrate formation after the peak in RCH 4 optimization. Most of the hydrate reforming was a CO 2 -rich hydrate with a lower concentration of CH 4 hydrates. The hydrate mass volume calculations show that the hydrate mass volume loss is most significant when the pressure is closer to the CH 4 hydrate stability pressure. When the pressure falls below the CH 4 hydrate stability pressure and is in the zone between CH 4 and CO 2 hydrate stability pressures, the loss of CH 4 hydrate volume is compensated by the formation of CO 2 -rich hydrate volume. For example, the loss of hydrate volume was equal to − 1.159, which was highest at the end of step A (at P = 24.3 bar). ΔV began to decrease as the pressure dropped below 24.3 bar. At the end of step C, at P = 24 bar, the volume loss was reduced to − 0.694. Thus, while the pressure remained largely stable between step A and stepped C, dissociation and sequential shutdown led to rapid gas/ water mixing, resulting in mixed hydrate reforming. Fig. 6B and C shows the mole fraction variation in Exp.2-3 during the MCD method.  3. In Exp.2, outlet pressure was quickly rebounded after degassing, and the pressure differential between inlet and outlet was minimized during the shut-in period. The NR2 value increased during the pressure rise, indicating either the formation of CH 4 -rich mixed hydrates or an increase in gas saturation in the pore space.
In Exp.3, outlet pressure continued to drop after degassing, indicating rapid hydrate formation at the outlet. During the shut-in period, the pressure differential remained intact (see C in Fig. 6). The pressure curve at the outlet (Exp.3) indicates rapid reformation and dissociation along the core length, leading to an initial decrease in NR2 followed by an increasing trend. This suggests that CO 2 -rich hydrate reformation causes a decrease in NR2. Rebound in pressure curve indicates the partial dissociation of CO 2 -rich hydrate and subsequent increase in NR2. The difference in water saturation (In Exp.2 and 3) could have triggered the different responses during the shut-in period. Another factor could be the degree of degassing at the start of step B.
In Exp.2, XCH 4 , RCH 4 and SCO 2 had two peaks. The CH 4 mole fraction varied from 72% to 78%, with the first peak XCH 4 = 78.1% at the end of step B (at P = 26.2 bar) and the second peak XCH 4 = 77.9% at the end of step D (at P = 22.0 bar). RCH 4 varied from 55% to 77.2%, with the first peak RCH 4 = 76.7% at the end of step C (at P = 21.9 bar) and the second peak RCH 4 = 77.2% at the end of step F. Another peak suggests increased CH 4 mole fractions in the vapour phase due to dissociation of CO 2 -rich hydrates or dissociation of CH 4 hydrates that were screened by CO 2 hydrates [1]. Similarly, SCO 2 varied from 59% to 84.1%, with the first peak SCO 2 = 84.1% at the end of step D (at P = 22.0 bar) and the second peak SCO 2 = 84% at the end of step F (at P = 21.2 bar). The NR2 value decreased by 6%, and the value changed from 1 to 0.94 at the end of the experiment.
In Exp.3, the variation of CH 4 mole fraction and SCO 2 showed only one peak, while RCH 4 showed two peaks. For CH 4 mole fraction, peak XCH 4 = 78.3% was recorded at the end of step D (at P = 20.3 bar). For SCO 2 , the peak SCO 2 = 82.9% was recorded at the end of step D (at P = 20.3 bar). For RCH 4 , the first peak, RCH 4 = 74.3%, was recorded at the end of step B (at P = 21), and the second peak, RCH 4 = 75.2%, was recorded at the end of step E (at P = 22.1 bar). NR2 continued to increase, and an overall improvement of 9% was recorded at the end of the experiment. Degassing at the lower pressure values did not cause any significant change in the CH 4 mole fraction. In general, after degassing, the pressure increased again within a few minutes and stabilized at a slightly lower pressure than the previous steps. After several steps (5-7), the pressure value stabilized in the 19-21 bar range.
The change in hydrate volume (ΔV) remained negative (in Exp.1-3), indicating the loss of hydrate mass volume during MCD-based dissociation of type A hydrates. The degree of loss in hydrate volume was different across the experiments. For example, the loss of hydrate volume was less in Exp.3 than in Exp.2, which may be due to the higher water saturation in Exp.3 supporting CO 2 hydrate reformation. Interestingly, the maximum loss in Exp.3 was observed at the end of step A, and an increase in hydrate volume was observed after that. This increase was correlated with an increase in NR2. In contrast, in Exp.2, a loss of hydrate mass was observed in the subsequent steps after step A, which was also correlated with the loss of NR2.

Cyclic depressurization of type B hydrates
Based on the assumptions and calculations described in Appendix B, parameters associated with type B hydrate dissociation is provided in Table 6 in Appendix B. Fig. 7 shows the variation of P, T, XCO 2 /XCH 4 , NR2 during Exp.4-5. Exp.4-5 contained type B hydrate (rich in CO 2 ), so the ratio of moles of CO 2 /CH 4 stored in the hydrate varied between 2.3 and 2.6. S wi before hydrate formation varied between 54% and 84%, resulting in a pore-filling hydrate morphology.
In Exp.4, MCD was applied in 3 steps (step A to step C). The CH 4 mole fraction (XCH 4 ) varied between 33.8% and 37%. XCH 4 had a single peak XCH 4 = 36.9% at the end of step A (at P = 15.1 bar). RCH 4 varied between 70% and 73.6%, with a peak of only 73.6% at the end of step C (at P = 14.3 bar). Similarly, SCO 2 varied between 77.8% and 78.4%, with two peaks at step A and C's end, respectively. The improvement of SCO 2 was marginal compared to the improvement of RCH 4 . The variation of NR2 shows a loss of resistance, and the total value decreased from 1 to 0.97. This decrease in resistance could be due to either improvement in CO 2 hydrate concentration or decreased gas saturation. Fig. 7A shows the response curve in Exp.4. The initial pressure before MCD was well below the CH 4 hydrate stability pressure of 30.8 bar. A degassing of 12 bar was performed before step A, which resulted in a pressure drop below the CO 2 hydrate pressure. In subsequent pressure drops, the pressure did not rebound and remained closer to the CO 2 hydrate stability pressure line. Due to the lower CH 4 stored in the hydrate, the CH 4 mole fraction increased slightly from 33.8% to 36.9%.
In contrast, the CO 2 mole fraction decreased from 66.25 to 65.4%. This suggest that dissociation closer to CO 2 stability pressure did not lead to any significant dissociation. The NR2 variation shows a loss of resistivity due to dissociation closer to CO 2 hydrate stability pressure. Extend of CO 2 hydrate formation was limited in this case; thus, loss of resistivity was observed. The NR2 value decreased below 1 and varied between 0.98 and 0.99. The decrease in NR2 value could also be due to decreased hydrate saturation (bulk mass).
In Exp.5, MCD was applied in 6 steps (step A to step F); the pressure decreased from 37.8 bar to 14.6 bar in 6 steps. The CH 4 mole fraction in the residual vapour phase varied from 28.9% to 45.7%, with two intervening peaks. The first peak XCH 4 = 45.7% at the end of step C and the next peak, XCH 4 = 43.3% at the end of step F. RCH 4 varied from 56.8% − 81.5% with two intervening peaks. The first peak RCH 4 = 81.5% at the end of step B, while the second peak RCH 4 = 77.7% at step E. Similarly, SCO 2 varied between 54.4% − 87.4% with two peaks. The peak SCO 2 = 87.4% was recorded at the end of step C and step F. Fig. 7B shows the pressure response curve in Exp.5. Before the first step A, the initial pressure was above the CH 4 hydrate stability pressure. At the end of step A, no significant change in XCH 4 was observed. During step B, just after the degassing, a massive pressure drop at the outlet was observed, indicating a significant hydrate-reforming event. This could be trigger due to the migration of CO 2 saturated pore water. The pressure difference developed between the inlet and outlet after the degassing gradually decreased during the shut-in phase, and system pressure stabilized near the CO 2 hydrate stability pressure. The CH4 mole fraction increased to 35.2% at the end of step B, confirming the reforming. Reformation continued in step C, and the CH 4 mole fraction increased to 45.7% at P = 15.2 bar. It is possible that during step C, the slight pressure drops were not sufficient to change the gas/water mixing that occurred during step B. The increase in CH 4 mole fraction was complemented by the increase in pressure value at the end of step C. Then, the CH 4 mole fraction decreased to 41.1% at the end of step D.
The CH 4 mole fraction improved slightly during steps E and F, and the mole fraction reached a second peak at the end of step F when the pressure reached 14.6 bar. This pressure was much lower than the CH 4 stability pressure PCH 4 = 27.8 bar and was slightly higher than PCO 2 = 12.5 bar. The NR2 variation in Exp.5 was different from Exp.4. NR2 (in Exp.5) increased above 1 when the system was continuously depressurized near the CO 2 stability pressure. NR2 improved from 1 to 1.02 in Step B; a sudden drop in NR2 was observed at the end of Step B and Step E, which later recovered during the shut-in phase. The rapid degassing in the zone between CH 4 and CO 2 hydrate stability pressures may have triggered hydrate reformation due to pore water freshening and the local temperature drop caused by the Joule Thomson cooling effect [50,51].
The change in hydrate mass volume (ΔV) concerning the initial condition was also measured. It was found that ΔV remained positive during the stepwise dissociation of type B hydrates. It indicates the evolution of additional mass into the formation at the end of a given pressure step. Exp.4 had a higher mass-volume addition (0.30 in Exp.4) at any given pressure than Exp.5 (0.18 in Exp.5). Although Exp.4-5 had volume addition during MCD, NR2 decreased in Exp.4 while NR2 increased in Exp.5. It is possible that due to the high water saturation in Exp.5, the hydrate mass volume was more evenly distributed. Comparing NR2 variation with hydrate mass volume in Exp.4-5, it can be said that a higher increase in hydrate mass volume resulted from CO 2 incorporation into the hydrate, leading to a decrease in NR2. In contrast, the minor increase in mass volume, but well distributed due to higher water saturation, caused an increase in NR2 in Exp.5.
Multiple peaks (XCH 4 ) in the vapour phase indicate a primary dissociation and reforming event. Below pure CH 4 hydrate stability, reforming involves the incorporation of CO 2 gas molecules into the hydrate, increasing the CH 4 mole fraction in the vapour phase. Another reason could be the dissociation of the CH 4 -rich mixed hydrate phase, increasing the CH 4 mole fraction. Peaks in CH 4 mole fraction closer to CH 4 hydrate suggest reformation, while peaks in CH 4 hydrate closer to CO 2 hydrate indicate dissociation of CH 4 hydrate (which was previously shielded by CO 2 hydrate). Fig. 8 summarize the results collected in Exp.1-3 (type A) and Exp.4-5 (type B). From Exp.2-3 (Type A), it was found that peak RCH 4 varied between 74% and 77% in the pressure range P = 20-23 bar. Peak SCO 2 varied between 82% and 84% obtained in the pressure window of  20 bar to 22 bar. In Exp.4-5 (Type B), peak RCH 4 varied between 74% and 77% obtained in the pressure range P = 19-22 bar. The peak value of SCO 2 varied between 83% and 84%, obtained in the pressure window of 20-21 bar. The peak values of RCH 4 and SCO 2 do not coincide but can occur within a specific pressure window. This pressure window was dependent on the type of hydrates. (20-23 bar for type A and 19-22 bar for type B). The peak values were not dependent on the type of mixed hydrates. The RCH 4 and SCO 2 values were near similar for type A and type B. It was frequently observed that the RCH 4 value had two peaks. Multiple peaks in RCH 4 confirmed primary dissociation but weaker reformation events. The second peak in RCH 4 was closer to the CO 2 hydrate stability pressure, suggesting that CH 4 gas remains stable in hydrates, either in the form of CO 2 -rich mixed hydrates and/or protected by a CO 2 hydrate layer. Fig. 9 shows the present status of research focused on depressurization based gas production from CH 4 hydrates. Previous research studies on depressurization include rapid depressurization, which raises many concerns, including significant dissociation, ice formation, sand/water production, and wellbore damage. The primary concern is geomechanical instability due to hydrate dissociation. Currently, no longterm field-scale observations could improve our understanding of what happens when pressure is below the thermodynamic stability pressure. To address the concerns associated with rapid degassing, research is currently focused on implementing slow incremental degassing to optimize production and to address concerns associated with rapid degassing. Slow pressure reduction (degassing) prevents icing, maintains reservoir temperature, prevents rapid dissociation, and avoids downhole damage. As the hydrate saturation across the hydrate reservoir is heterogeneous, hydrate stability pressure varies due to the pore size distribution that causes uneven hydrate dissociation and gas production. Thus, even though slow depressurization (degassing) prevent any significant dissociation, concerns associated with geomechanical stability below CH 4 stability pressure remains there.

Practical implications on field-scale trial
In our proposed method, CO 2 injection into the partially depressurized CH 4 hydrate reservoir is suggested. The partial depressurization is achieved by a slow depressurization technique, which allows optimal CH 4 production without destabilization of the in situ hydrates. CO 2 injectivity into partially depressurized CH 4 would be higher due to improved permeability and availability of migration pathways. To further improve CO 2 injectivity and achieve a high sweep range, kinetic inhibitors or anti-agglomeration agents are required to avoid forming a mass transfer barrier (CO 2 hydrate film) at the injection well [30,45].
A high CO 2 injection volume (in partially depressurized CH 4 and in the presence of kinetic inhibitors/anti-agglomeration agents) would ensure a high CO 2 concentration in the vapour phase, high CO 2 solubility in the residual pore water, and enhanced synthesis of CH 4 /CO 2 mixed hydrates. The CO 2 itself is stored in the form of pure CO 2 hydrates. The next step in our proposed method is to dissociate CH 4 /CO 2 mixed hydrates by slow depressurization (stepwise/cyclic) to enhance CH4 recovery further and store additional CO 2 .
The quantity and quality of mixed CH 4 /CO 2 hydrate formation in the porous medium varies due to several factors. In the hydrate production scheme with injection and production well and the presence of good CO 2 injectivity and a large sweep region, the hydrate distribution between injection and production well can be divided into three main zones. Type B (CO 2 rich) mixed hydrate would be available closer to the injector well, while type A (CH 4 rich) hydrate would be available further away from the injector well. Pure CH 4 hydrate would be available closer to the production well. The interfaces between the different zones and CH 4 /CO 2 mixed hydrate distribution depend on the total amount of CO 2 injection, permeability and porosity of hydrate-bearing sediments, relative CO 2 permeability, pore size distribution, initial water saturation, and reservoir temperature.
The differences in hydrate distribution indicate heterogeneity in the system. The presence of heterogeneity in the hydrate system due to CO 2 injection would affect the gas production behaviour during depressurization. The gas production technique would also affect the production behaviour. Recent studies indicate that it is beneficial to use a slow stepwise or cyclic method of depressurization [2,16]. In the MCD method, the reservoir pressure initially moves towards the pure CH 4 hydrate stability boundary. This would cause pure CH 4 hydrate to dissociate at the production well first. The sensible reservoir heat recovers the temperature loss at the production well during the shut-in period, and unwanted hydrate/ice formed at the production well would dissociate. During the MCD technique (or other slow depressurization techniques), when the pressure value falls below the CH 4 hydrate stability pressure, mixed hydrate dissociates according to their stability pressure decided by CO 2 and CH 4 gas concentration inside the hydrates. If the pressure drops further below the CH 4 hydrate stability pressure, type A mixed hydrates would dissociate first as their stability pressure is reached. During the shut-in period, gas/water mixing may occur, and any temperature loss would be recovered. Mixed hydrates, when dissociate, releases water and gas. Fig. 10 shows the hydrate distribution after CO 2 injection (Fig. 10 A) and during the MCD (Fig. 10B). The released gas, containing both CH 4 and CO 2 , mixes with water and accelerates the formation of CO 2 -rich mixed hydrate. The likelihood of CO 2 -rich mixed hydrate formation would be higher because a higher partial pressure-based driving force is available for CO 2 hydrate than Fig. 10. Overview of proposed hydrate distribution before and after cyclic depressurization to enhance CH 4 recovery from CH 4 hydrate reservoirs after CO 2 injection. A) shows the hydrate distribution between the injection well and production well after CO 2 injection. Hydrate distribution includes type A, type B, and pure CH 4 hydrates. B) shows the hydrate distribution between the injection well and the production well after cyclic depressurization. The hydrate distribution of type B increases while that of type A decreases.

CH 4 hydrates.
When CH 4 -rich hydrate (both pure and type A) begins to dissociate at reservoir pressures below the CH 4 hydrate stability pressure, the CH 4 mole fraction in the production well will increase. Due to hydrate dissociation, hydrate saturation in the pore space would decrease, which would improve relative gas permeability and gas migration ability. Additional CO 2 from the injection well side starts to move towards the production well, and gas and water mix happen closer to the production well during the shut-in period.
During the shut-in period, CO 2 saturated pore water may also transport in the pore space, which could trigger CO 2 -rich hydrates at the gas-liquid interface. This would improve the CH 4 mole fraction in the residual vapour phase. CH 4 mole fraction is also affected by the CH 4 hydrate saturation remaining in the hydrate and the pressure difference between type A and type B hydrates. When the CO 2 hydrate reforming is triggered near the production well, the CO 2 -rich type B hydrate front moves toward the production well and the type A hydrate would decrease. (see Fig. 10B). The advancement of the type B front toward the production well indicates that the CO 2 sweep area increases and additional CO 2 gas molecules may interact with the surface of the CH 4 hydrates, leading to CH 4 production by hydrate exchange. CO 2 reformation kinetics below CH 4 hydrate stability pressure is essential in optimizing CH 4 recovery below CH 4 hydrate stability pressure without destabilizing hydrate-bearing sediments.
In this study, no significant loss of resistivity was observed. The increase in resistivity value and improvement in hydrate mass volume at high water saturation is attributed to CO 2 rich hydrate reformation enhanced by uniformly dispersed CO 2 saturated liquid phase. This reduced the risk of water production or loss of hydrate mechanical strength. Further experiments on mechanical strength during cyclic depressurization are needed to compare the strength of CO 2 -rich hydrate versus CH 4 -rich hydrate.
Key factors that can trigger and control reforming include pore water chemistry, gas/water contact area, formation kinetics, gas/water migration, capillary forces, and pore water freshening. Although an extended shut-in period is beneficial because it provides sufficient time for reformation and hydrate redistribution and storage, valuable production time is lost. Therefore, it is necessary to minimize the shut-in period to optimize gas production while maintaining reservoir temperature, avoiding ice formation and hydrate strength, and allowing hydrate reformation [11]. Chemical compounds such as kinetic inhibitors/ agglomeration inhibitors would affect the reforming kinetics at the gas/ liquid interface and affect the shut-in period. The shut-in period may either prolong or shorten due to pore water chemistry alteration. Therefore, further studies on the slow depressurization of CH 4 /CO 2 mixed hydrate in the presence of different chemicals are needed.
The speed and magnitude of the pressure drop control the driving force of hydrate dissociation. Large and rapid pressure drops could result in a significant temperature drop that may trigger ice formation during production, requiring a longer shutdown time. A study showed that smaller pressure drops improve the dissociation behaviour than more significant pressure drops for a pure CH 4 hydrate system [12]. Similar studies are required for the mixed CH 4 /CO 2 hydrate system

Conclusions
CO 2 injection into a CH 4 hydrate reservoir produces CH 4 /CO 2 mixed hydrates with CO 2 -rich mixed hydrates closer to the injection well and CH 4 -rich mixed hydrates away from the injection wells. The presence of two distinct hydrate stability regions allowed us to produce CH 4 gas while storing additional CO 2 in hydrates using the Multistep cyclic depressurization relief (MCD) technique [2]. This study investigated the resistivity variation and CH 4 production/CO 2 storage behaviour for CH 4 -rich and CO 2 -rich mixed hydrates synthesized in the sandstone pores. The main conclusions are given below • The ratio of CH 4 /CO 2 stored in hydrates depends on the CH 4 /CO 2 composition in the feed gas. CH 4 -rich feed gas produces a CH 4 -rich mixed hydrate system, while CO 2 -rich feed gas produces a CO 2 -rich mixed hydrate system. The relative content of CH 4 /CO 2 in the feed gas also controls the hydrate morphology. CH 4 -rich hydrate morphology would be dominated by hydrate films encapsulating the gas phase and high residual water saturation, while CO 2 -rich hydrate formation involves crystallization of gas-saturated water and low residual water saturation. • In the synthesis of CH 4 /CO 2 mixed hydrates, the resistance variation shows 10-14% improvement for CH 4 -rich mixed hydrates and 14-19% for CO 2 -rich hydrates. The resistance variation depends on the initial water saturation (S wi ), the type of guest molecule and its concentration, the total hydrate volume. The resistance variation correlated with the pressure variation but could not account for the heterogeneity of hydrate saturation over the core length.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.    Table 3 Experimental details of mixed hydrates system in sandstone. T SD is the temperature during cyclic depressurization, PCH 4 = CH 4 hydrate stability pressure at T SD , PCO 2 = CO 2 hydrate stability pressure at T SD ; nCO 2 and nCH 4 are the moles of CO 2 and CH 4 stored in the hydrates. P start = pressure before the first pressure drop, P finish = pressure at the end of the experiment.

C.1 Artificial CH 4 /CO 2 hydrate
Initial moles of gas (n CH4,i , n CO2,i ) injected into the core flooding setup is given by Where P i is the initial operating pressure recorded after gas mixture was injected, V 1 is the available gas volume inside the core, T i is the recorded experimental temperature at inlet corresponding to P i , R is the universal gas constant, and z i is the compressibility factor at the given pressure and temperature, calculated from the Benedict-Webb-Rubin-Starling equation of state and corresponding mole fraction of each gas component respectively. y CH4,i y co2,i are the gas mole fractions of CH 4 and CO 2 in the injected gas. For constant volume processes, assuming, hydrate formation inside the core holder is batch process with constant gas volume, number of moles of each gas (n CH4,f , n CO2,f ) in vapor phase after mixed CH 4 /CO 2 hydrate formed is determined by P f is the stabilized pressure recorded at the end of the cooling cycle with the stabilized T f for hydrate formation, and z f is the compressibility factor for P f and T f conditions and corresponding mole fraction of each gas component respectively. Thus, the moles number of CH 4 trapped in the solid hydrate crystal after hydrate formationΔn CH4 ,H is given by

C.2. Experimental data processing during the multistep cyclic depressurization
To calculate the CH 4 recovery and CO 2 storage during MCD, simplified mass balance equations were used. Basic assumption during mass balance are following.  Mass balance on moles of CO 2 Key assumption included, 1-Δn CH4,step is the number of moles of CH 4 remaining in the hydrate at the end of the cyclic depressurization. During multistep cyclic depressurization, when the pressure in the system was below the CH 4 hydrate stability pressure, it can be assumed that all moles of CH 4 are recovered and no moles of CH 4 are left stored in the hydrates Δn CH4,Step = 0. 2-For simplicity, it was assumed that all gas escaping during MCD is rich in CH 4 as CH 4 hydrate dissociates and CO 2 gas go into hydrate during CO 2 hydrates reformation. Total CO 2 moles escaped during venting are negligible Δn CO2,Re = 0.
Before multistep cyclic depressurization technique implementation, moles of CH 4 and CO 2 in the vapor phase (n CH4,f , n CO2,f ) and hydrate phase (Δn CH4,H , Δn CO2,H ) were known. During the controlled multistep cyclic depressurization stage, the pressure at the outlet was rapidly dropped to P step1 and thereafter valve was shutin. After approximately 10-12 h of shut-in period, pressure changed to P step,3 inside the core holder due to multi phase flow on account of dissociation and/or reformation. Change in CH 4 or CO 2 mole fraction in the cell indicating a change in molar concentration of CH 4 and CO 2 in vapor phase due to dissociation and reformation within CH 4 /CO 2 hydrate system. Assuming that this was last step no further pressure drops were applied. At this point, GC analysis was performed to know the mole fractions of CH 4 and CO 2 in gas phase. The number of moles at this stage was given by n CH4,v1 = y CH4,step2 P step2 V 1 z step2 RT step2 (7) n CO2,v1 = y CO2,step2 P step2 V 1 z step2 RT step2 (8) Wherey CH4,step2 y CO2,step2 are the molar composition of the gas for pressure P step , 2 collected and determined by gas chromatography. Total CH 4 recoveryR CH4 (%)during multistep cyclic depressurization technique from mixed hydrate can be given by R CH4 (%) = Δn CH4,Re Δn CH4,H + n CH4,f × 100 Δn CH4,Re is the total CH 4 moles released during the venting. It can be calculated based on the mass balance equation discussed previously. Total CO 2 storage S CO2 (%) during multistep cyclic depressurization technique is given by Δn CO2,step is the total CO 2 moles stored in the hydrate. It can be calculated based on the mass balance equation discussed previously. Total volume of hydrate mass was calculated using net moles produced or stored in each step. Positive ΔV indicate gain in hydrate mass volume and negative ΔV indicates loss in hydrate mass volume with respect to starting condition.
Appendix D Table 7 shows the ratio (mole of CH 4 /mole of CO 2 ) in the experimental studies. Previous studies show that in the synthesis of CH 4 /CO 2 mixed hydrates, CO 2 goes into large cages 5 12 6 2 , and CH 4 molecules go into small cages 5 12 because the ratio of the cage to molecular diameter is optimized [34,35]. The results (see Table 7) show that the CH 4 cage occupancy in large and small cages depends on the CH 4 concentration in the feed gas. Based on CSMGem calculation, in Exp. 1-3, the CH 4 hydrate cage occupancy in the large cage was higher than that in the small cage when 70% CH 4 gas was used in the feed gas. When the CO 2 mole fraction was 70%, CH 4 cage occupancy was higher in the small cages than in the large cages. Exp. 1-3 (type A) the ratio was above 1, while in Exp. 4-5 (type B), the ratio was less than 1. Table 7 Provides the details of cage occupancy and total moles stored in hydrates for the CH 4 /CO 2 gas mixture. R(CH 4 /CO 2 ) is the ratio of CH 4 /CO 2 moles stored in hydrates. Exp