Elsevier

Applied Energy

Volume 236, 15 February 2019, Pages 825-836
Applied Energy

Efficiency enhancements in methane recovery from natural gas hydrates using injection of CO2/N2 gas mixture simulating in-situ combustion

https://doi.org/10.1016/j.apenergy.2018.12.023Get rights and content

Highlights

  • Inclusion of 15% N2 in CO2 stream increases CH4 recovery by at least 25%.

  • CH4 is first replaced by CO2 in large cages followed by N2 in small cages.

  • N2 is selectively captured in hydrate cages below 12 °C.

  • Higher sequestration potential observed at lower heating rates.

Abstract

Thermal stimulation was combined with an injection of a mixture of CO2 (85%) + N2(15%) to investigate efficiency enhancements from pure thermal stimulation and thermal stimulation with CO2 injection approaches. Tests were performed at initial hydrate saturation of 10% and 300 ml/min CO2 + N2 injection rate with three different heating rates of 20, 50 and 100 W. The results indicate that thermal stimulation with CO2 + N2 injection is the most efficient method available for methane gas recovery. At 10% Hydrate Saturation (SH) and 100 W heating rate, the number of moles of CH4 recovered increased from 8.5 to 16 to 20 in the case of thermal stimulation, thermal stimulation with CO2 exchange and thermal stimulation with CO2 + N2 exchange respectively. The experimental results reported here are aligned with model and Raman spectroscopy predictions in terms of replacement mechanism and recovery efficiency, reported in the literature. The results obtained from CO2/N2 composition ratio show that in the exchange process, CO2 first replaces CH4 in the large cages of Structure I hydrates followed by N2 targeting CH4 in the small cages of Structure I hydrates. This replacement mechanism has been predicted in the literature by Liu et al. (2016) using Molecular Dynamics simulations. It is also found from this work that N2 is selectively captured in hydrate cages below 12 °C. The values of carbon sequestration index (defined as moles of CO2 sequestered divided by moles of CH4 recovered) were 0.32, 0.52 and 0.85 respectively for 100, 50 and 20 W heating tests. The data obtained from our work in terms of gas composition, methane recovery and CO2 sequestered is consistent with the key findings reported in the literature.

Introduction

Clathrate hydrates are comprised of water molecules and a small guest molecule, for example, methane, propane, cyclo-pentane, and carbon dioxide [1], [2], [3], [4]. Methane hydrates are formed naturally in sub-sea floors and permafrost regions under the conditions of high pressure and low temperature. In the hydrate form, a methane molecule is not chemically bound to the water molecule but is entrapped within the cages formed by a water molecule. In physical appearance, the gas hydrates are similar to ice, but when exposed to atmospheric temperature and pressure, they release methane gas due to their dissociation. The chemical structure of gas hydrates is well studied in the literature [5], [6], [7], [8], [9], [10], [11]. Three main categories of gas hydrates – Cubic Structure I, Cubic Structure II and Hexagonal Structure H have been discussed in the literature [6], [12], [13]. The small gas molecules of interest in this work (CH4, CO2) form Structure I hydrates whereas N2 forms Structure II hydrates [14], [15].

Gas hydrates are dispersed in the continental slope areas and onshore permafrost, as a result gas hydrates are globally accessible (US, Canada, Former Soviet Union, Latin America & Caribbean, Europe, Southern Africa, Middle East, China, Japan, South and Pacific Asia) [16]. Carbon dioxide is considered as one of the major contributors to global warming. The atmospheric levels of CO2 increased from 310 ppm in the 1960s to a value of 410 ppm in 2016 [17]. This increase in the CO2 levels is due to increased use of high carbon fossil fuels such as coal and petroleum products. Natural gas has the lowest carbon intensity of all the fossil fuels due to its high hydrogen to carbon ratio. Therefore it is anticipated that the continued transition to low carbon fuels (such as methane) will help reduce the carbon footprint of the energy generation sector. Another marked feature of extracting methane from the gas hydrate phase is that this method could be coupled with permanent carbon sequestration. Thus on a global scale the switch to cleaner fuels such as natural gas can be enabled through unconventional sources such as methane hydrates.

The most common methods of hydrate dissociation are – (1) Thermal Stimulation, (2) Depressurization and (3) Chemical Inhibitor Injection. Thermal stimulation technique involves increasing the temperature of the hydrate reservoir such that hydrates enter the instability region and dissociate into gas and water. Similarly, the depressurization method causes a drop in the reservoir pressure which makes hydrates unstable and dissociates into water and gas. Chemical inhibitor method makes use of an injection of a substance that disrupts the thermodynamic equilibrium, such as methanol. A combination of different techniques has also been reported in the literature [8], [10], [15]. For lab scale studies synthetic gas hydrates are used, that emulate the permafrost conditions and they are often grown in a lab using higher sub-cooling rates or hydrate promoters which accelerate the hydrate formation [18].

Several lab scale tests are reported that focus on thermal and depressurization methods [19], [20], [21], [22], [23]. Most of the depressurization tests require a small thermal input for initiating the gas recovery. The field tests made use of an injection of steam or hot water to increase the temperature of the sediment [10]. However, these methods suffer from the heat loss during transit of the fluid from the ground to the hydrate sediment location which results in lower production efficiency. We [7] proposed an idea of in situ (downhole) combustion that will utilize the methane released from the hydrate phase as a fuel for further dissociation. The work presented in this paper expands on this idea of in situ combustion by making use of a point electrical heater with CO2 and N2 injection to simulate the combustion. The simulated in situ combustion technique has been shown to provide better efficiencies due to the elimination of heat losses during hot fluid injection [8]. In past publications, we have successfully demonstrated the thermal stimulation, depressurization and CO2 sequestration capabilities of the experimental setup used in this work [7], [8], [10], [14], [15], [19], [24], [25]. We have also shown that the results obtained from this setup are capable of matching the hydrate formation behavior in the permafrost and relevant to the field scale tests [10], [19]. Leveraging the understanding from those findings, the work reported here continues to the next level of complexity in which we utilize a mixture of CO2 and N2 combined with thermal stimulation for methane recovery purposes. In the case of in-situ combustion, the reaction mixture will contain CH4 (fuel), N2 (from the use of air as an oxidant), CO2 (combustion product), and water. For experimental simplification, the presence of trace amount of unreacted O2 was neglected. The small amount of O2 present in the stream will cause a slight increase in equilibrium pressure of less than 1%. The data reported here is obtained at conditions closest to the conditions of in-situ combustion. As an example, assuming 70% conversion of CH4 to CO2 and H2O – this will generate heat at the rate of 100 W and CO2 at the rate of 240 ml/min. As a result, 100 W heating rate and total flow of 300 ml/min (240 ml/min CO2 + 60 ml/min N2) of 85% CO2 and 15% N2 has been chosen as one of the operating conditions. This coupling of CO2 injection rate with heating rate allows a realistic prediction of in-situ combustion.

Chong et al. [26] provided an extensive review of recovering methane from gas hydrates and also discussed the prospects and challenges in the process. The authors identify the key parameters in the CO2/CH4 replacement process such as particle size, diffusion barrier, fugacity of the components and the properties of the hydrate sediment including the thermodynamic conditions under which replacement process takes place. Most importantly, the authors also discuss the possibility of replacing pure CO2 with a flue gas mixture. Our work in this paper expands on the idea of sequestering a mixture of CO2 and N2. Wang et al. [27], in a notable gas hydrate research program in China, performed pilot scale studies to recover methane gas using a reactor of volume 117.8 L. The authors investigated cyclic depressurization methods in which they compared the results of their study with regular depressurization. The authors report that the gas production rates using cyclic depressurization were 17 times higher than using regular depressurization. The authors attribute this to the enhanced heat transfer obtained by utilizing cyclic method. The higher production rates reported in their studies have the potential to reduce the cost of production of gas per unit volume. In another study, Wang et al. [28] utilized the same experimental setup to study hydrate dissociation below the quadruple point in a sandy sediment. The authors found that the hydrate dissociation process below quadruple point benefits in terms of higher dissociation rates because of ice formation. The authors note that the water released in the process forms ice immediately which helps the dissociation due to the release of latent heat in the process. The authors also report that the rate at which ice is formed is slower in water saturated environment as opposed to gas saturated reservoirs.

Park et al. [29] performed Raman spectroscopy studies to predict the effect of CO2/N2 during the hydrate dissociation process. CH4 forms Structure I hydrates which has two types of cages, small and large. Structure I hydrates are formed by small guest molecules, which have the molecular diameter of less than 6 Å such as CH4 and CO2. Table 1 summarizes the ratio of the molecular diameter to cavity diameter for CH4, CO2, and N2.

Evident from Table 1, the CO2 molecule has nearly the same dimensions as the small cavity diameter of Structure I hydrate. This limits the replacement of methane by CO2. Park et al. [29] found using Raman spectroscopy that the efficiency of replacement when using pure CO2 was 64% whereas replacement efficiency was 85% when using a mixture of CO2 + N2. Further observation of Table 1 reveals that this can be attributed to the value of the ratio of molecular diameter to cavity diameter for N2. The N2 molecule can fit and replace CH4 from the small cages of Structure I hydrate effectively, as opposed to pure CO2. The value of this ratio is 1 for CO2 whereas it is 0.804 for N2. It means that the N2 molecule is smaller in size than the cavity diameter whereas the CO2 molecule is equal in size to the cavity diameter. This implies that the replacement of CH4 in small cages is structurally more favorable by N2 rather than CO2. This geometric consideration may explain the increase in CH4 recovery by addition of N2 to the mixture. The data presented by us will quantitatively determine the efficacy of the addition of N2 to the sequestration gas mixture.

Koh et al. [30] used natural gas hydrate sediments to study the replacement process using a mixture of CO2/N2. The authors used 20 mol%/80 mol% mixture of CO2/N2. The authors report that the CO2 injected during dissociation targeted the CH4 molecules in the large cages of Structure I hydrates. They report a duration of 20 h for a complete recovery of CH4. The recovery rate reported in the CO2 exchange was nearly 64%. On the other hand, during the injection of a mixture of CO2/N2, the authors observed that the injected N2 targets the CH4 molecules in the small cages. This increased the methane recovery rate to 85% from 64% value in the case of pure CO2 injection.

Cha et al. [31] studied the kinetics of a replacement process of CH4 with a mixture of CO2/N2 using in-situ Raman spectroscopy. It took 65 h for the process to replace 42 mol% of CH4 from the hydrate cages. The authors claim this to be a replacement where no significant amount of free water was generated. They report that the amount of methane recovered from large cages was more than that recovered from small cages. The results showed that the replacement assisted in recovering methane from hydrate phase as well as it helped stabilize the resulting CO2 hydrate.

Babu et al. [5] studied the formation of mixed hydrate (CO2/N2) using a batch stirred reactor at constant temperature and pressure. An experiment conducted at 11 MPa showed that a higher number of N2 moles were consumed in the hydrate formation than CO2. This group also studied the effect of adding Tetra Hydro Furan (THF) to the CO2/N2 mixture. As expected, THF was seen to reduce the pressure at which hydrates are formed.

Koh et al. [30] attempted methane recovery from naturally occurring hydrate sediment using flue gas mixture. Their results show that CO2 replaces CH4 in the hydrate phase by swapping mechanism. They appear to be the first research group to pull out methane from clay interlayers using a CO2/N2 mixture. In another study [32] performed by the same group, they studied the replacement process of CH4 by a mixture of CO2/N2 using a one dimensional, 8 m reactor. They report that the lower the CO2/N2 injection rate, the higher is the total recovery of methane. They attribute this to the gas-solid contact time. They find that in their system, a length of 5.6 m is required before significant recovery of methane is achieved.

In a recent study by Li et al. [33] the authors utilized pure CO2, and two different compositions of CO2/N2 (1:3 and 3:1) to study the gas production and replacement process from a fracture filled hydrate reservoir. The authors identified that gas composition, pressure, and morphology were the parameters that dominated the replacement and recovery process. They indicate that the addition of nitrogen in the sequestration mixture will increase the efficiency of the methane recovery process. The work we report here demonstrates this through quantitative measurements of the gas composition during the replacement process within the sediment on a large scale setup. Seo et al. [34] studied CO2/N2 (50/50) replacement in Structure II hydrates. One of the major findings from their work is the replacement occurs by replacing CH4 in small cages by N2 and replacement of CH4 in large cages by CO2 molecules. Although Structure I hydrates are most common, the authors utilized a mixture of CH4 + C3H8 to obtain Structure II hydrates as they are prevalent in some regions. Using various analytical methods, they report a replacement efficiency of 54%. Zang et al. [35] studied ternary mixture hydrate formation that involved CH4, CO2 and N2 for hydrate based gas separation process. They report using X-Ray Diffraction that the ternary mixture hydrates formed Structure I type. The authors also conclude that the driving force between operating conditions and thermodynamic equilibrium conditions dominate the ternary hydrate formation process. Chazallon et al. [36] studied various compositions of CO2/N2 hydrates using in-situ Raman spectroscopy. They report that Structure I hydrates have higher selectivity towards CO2 capture than Structure II hydrates. The authors report that the CO2 molecules occupy large cages of Structure I hydrates for the CO2 composition of between 2 and 20%.

Liu et al. [37] performed molecular dynamics simulations to study the replacement of CH4 by CO2 and N2 on a micro scale. The authors focused on the thermodynamic and kinetic aspects of the replacement process. The authors report that the process of replacement of CH4 by CO2 and N2 has a negative Gibbs free energy which indicates that this process is thermodynamically favorable. They report that the replacement of CH4 by CO2 in small cages is structurally not favorable for the reasons already discussed earlier in Table 1. Although this replacement is structurally not favorable, the replacement takes place by dissociation of CH4 hydrate followed by formation of CO2 hydrates as the process is thermodynamically feasible. The calculations in their work show that the replacement process begins by replacing CH4 with CO2 in large cages followed by a replacement of CH4 by N2 in small cages. This replacement process is expected to produce mixed CO2-N2 hydrates. The important finding from their work is that N2 has a higher diffusion rate into the hydrate phase compared to CO2 which makes this process kinetically favorable. In summary, the authors report that the replacement process is a combination of thermodynamically driven CO2 replacement and kinetically driven N2 replacement. The work presented here provides quantitative experimental confirmation of the spectroscopic and MD simulation results of the pure systems. Importantly the experimental findings here quantify the replacement rates associated that would be observed in the field.

The gas recovery using the injection of CO2 is well established in the literature [26], [35], [36], [38], [39], [40], [41] the same using CO2/N2 mixture is a recent development and a novel technique. This work makes use of the concept of simulated down-hole combustion [7] to study the effect of the simultaneous addition of the heat and CO2/N2 on the gas recovery. The concept of downhole combustion was proposed by our research group in 2007 [7] and this work is a continuation of the work previously reported. Furthermore, a relevant-scale setup that closely approximates the aspect ratio of field operations has been used. Heat and CO2/N2 were simultaneously injected in the hydrate sediment to realistically achieve down-hole combustion that will allow the study of CH4-CO2/N2 hydrate exchange while sequestering CO2 in the sediment. The test matrix in this work has been set up to study the effect of CO2/N2 injection on gas recovery and gas exchange process, the effect of addition of the heat, simultaneously with CO2/N2 injection on gas recovery and gas exchange and the enhancement in the recovery efficiency of methane by comparison with thermal stimulation and thermal stimulation with pure CO2 exchange. The studies [7], [42], [43], [44] that used a mixture of CO2/N2 for sequestration purposes utilized a flue gas mixture (∼85% N2/15% CO2). The novelty of the work reported in this manuscript is the use of CO2 rich sequestration mixture. Recently there is a growing interest in developing technologies to capture carbon using clathrate hydrates for pre and post combustion capture. It is also shown in a notable article by Babu et al. [45] that this process has the potential to be commercialized for adaptation to the Integrated Gasification Combined Cycle (IGCC) systems. The use of flue gas for sequestration purposes is two-fold, it could be used as a sequestration gas stream during the recovery of methane for energy purposes and at the same time, it could be used to capture CO2 from IGCC systems.

Lim et al. [43] used a flue gas mixture to study the phase equilibrium of mixed hydrates formed by using CH4/CO2/N2 and the thermodynamic stability of the resulting hydrates. The authors report that the hydrate phase is enriched in CO2 whereas the ratio of N2/CO2 in the hydrate phase is independent of the composition of CH4 in the reservoir. Yang et al. [44] used a simulated flue gas to study carbon capture and gas recovery at typical gas hydrate reservoir conditions (273–284 K and 4.2–13.8 MPa) and obtained 50% CH4 composition in the gas phase. They report up to 70% conversion of CO2 from flue gas into the hydrate phase. Peter et al. [46] used concentrated flue gas mixture in a 5.3 L crystallizer and with horizontal injection tubes and a high pressure calorimeter to study the replacement process. The authors report that the replacement process is limited due to the presence of limited water and mention the need for CH4/CO2 separation unit to complete the cyclic operation. Hassanpouryouzband et al. [47] used flue gas mixture to study the effect of reservoir conditions (thermodynamic conditions) on CO2 capture capacity. The authors studied a range of temperature (273.2–283.2 K) and pressure (2.6–23.8 MPa) conditions for capture. The authors report a capture efficiency of 60% in terms of CO2, in either pure CO2 hydrate state or mixed hydrate state. Lee et al. [48] used micro-differential scanning calorimetry (DSC) to study the replacement and gas recovery process using flue gas mixture. They used DSC to study the heat of dissociation during the replacement process. The authors did not notice a significant heat flow during formation and dissociation of hydrates, as a result, they concluded that the swapping process lacks structural changes and significant hydrate dissociations.

There have been several other studies done that focus on CO2/N2 mixed hydrate phase equilibria but none of them focus on the use of this mixture in the context of methane recovery from the hydrate phase [11], [45], [47], [48]. The majority of them focus on predicting the thermodynamic equilibrium conditions using hydrate promoters for CO2/N2 mixture hydrate [49].

As discussed earlier, the work done to date uses flue gas mixture for sequestration which is richer in N2 than in CO2. Flue gas stream has on an average between 12 and 15% CO2 and balance Nitrogen [50], [51], [52]. The major distinction of this work is the use of a CO2 rich sequestration stream containing 85% CO2, simulating downhole combustion which was proposed by our research group in 2007 [7]. This CO2 rich sequestration stream can be obtained from a combination of coal-fired and natural gas-fired power plants. In power plants that use coal as the feedstock, a CO2 rich gas stream is obtained from capture processes such as Selexol and Rectisol [53], [54], whereas the post-combustion gas stream (in both coal and natural gas fired power plants) is rich in N2 (flue gas) [51]. The aforementioned gas streams are readily available from Integrated Gasification Combined Cycle (IGCC) plants and hence may be practical to obtain such high concentration CO2 stream for large-scale operations [50], [51], [55], [56].

The apparatus utilized in this work has a well-established history of more than ten years in which some unique findings were reported [19]. Specifically, the data obtained from the setup utilized in this work for gas discharge during hydrate dissociation was within an order of magnitude of that was predicted by Moridis et al. [57] for field simulations. The gas discharge value predicted from an experiment using this setup [19] was 4.5 × 106 m3 whereas the value predicted by Moridis et al. for field simulation was 4.1 × 107 m3. The test results are comparable to the numerical results from Moridis’ model, considering the scale differences between the model and the actual experiment. Therefore, good agreement with the field simulation data by Moridis indicates that the hydrate formation and dissociation procedure followed in this work and the apparatus is capable of producing useful results that provide insights about hydrate dissociation in the field if the aspect ratios are maintained.

The process needs to be optimized for large scale and economically favorable production. As a result, lab scale tests are required to provide a fundamental understanding of the hydrate dissociation process. A reactor with 59.3-liter volume has been used to explore the impacts of methane production and CO2 sequestration from simulated permafrost methane hydrates. The results from this study could be used as a baseline to select a CO2 rich composition of CO2/N2 for sequestration purposes. The results are presented in terms of gas compositions, CO2 sequestered, CH4 recovered and the heat transfer characteristics during the recovery process. This is the first large-scale lab demonstration of injection of the CO2/N2 mixture with a focus on CH4 gas recovery. It is also the first successful application and demonstration of the use of CO2 rich gas stream as opposed to flue gas.

Section snippets

Equipment and methods

A Large-Scale Hydrate Vessel (LSHV) was designed and fabricated using stainless steel for the study of methane hydrate formation, dissociation and carbon sequestration purposes. The diameter and the height of the cylindrical vessel were 0.3 m and 0.81 m respectively, this provides a total internal volume of 59.3 L. The LSHV was filled with quartz sand and de-ionized water was used to saturate this sand with water to a saturation of 50% by pore volume. Fig. 1 shows the schematic of the LSHV. The

CO2/N2 hydrate formation and dissociation baseline

It is important to independently understand the formation and dissociation behavior of CO2 + N2 hydrates in this system (in the absence of CH4) before the gas mixture could be injected into the sediment for CH4 hydrate dissociation purposes. As a result, first, the data for 10% hydrate formation test using a mixture of CO2 and N2 is presented in this section. It should be noted that no CH4 is used for the studies in this subsection. The experimental procedure involves the injection of a mixture

Conclusion

Thermal stimulation combined with CO2/N2 exchange was studied in a large-scale lab reactor simulating down-hole (in-situ) combustion. The novelty of the work is the use of a CO2 rich [CO2(85%) + N2(15%)] sequestration gas stream. This is the first demonstration of the use of CO2 rich gas composition on a large-scale lab setup that imitates field aspect ratios. Thermal stimulation combined with CO2/N2 exchange provided the highest methane recovery (at three heating rates studied) as compared to

Notes

The authors declare no competing financial interest.

Acknowledgment

The authors acknowledge the City College of New York (CCNY) and Earth and Engineering Center (EEC) at CCNY for providing startup funding that fully supported this research.

References (58)

  • Z.R. Chong et al.

    Review of natural gas hydrates as an energy resource: prospects and challenges

    Appl Energy

    (2016)
  • Y. Wang et al.

    Pilot-scale experimental evaluation of gas recovery from methane hydrate using cycling-depressurization scheme

    Energy

    (2018)
  • Y. Wang et al.

    Large scale experimental evaluation to methane hydrate dissociation below quadruple point in sandy sediment

    Appl Energy

    (2016)
  • B. Li et al.

    An experimental study on gas production from fracture-filled hydrate by CO2 and CO2/N2 replacement

    Energy Convers Manag

    (2018)
  • Y ju Seo et al.

    Isostructural and cage-specific replacement occurring in sII hydrate with external CO2/N2 gas and its implications for natural gas production and CO2 storage

    Appl Energy

    (2016)
  • B. Chazallon et al.

    Selectivity and CO2 capture efficiency in CO2-N2 clathrate hydrates investigated by in-situ Raman spectroscopy

    Chem Eng J

    (2018)
  • M. Ota et al.

    Methane recovery from methane hydrate using pressurized CO2

    Fluid Phase Equilib

    (2005)
  • O. Ors et al.

    An experimental study on the CO2-CH4 swap process between gaseous CO2 and CH4 hydrate in porous media

    J Pet Sci Eng

    (2014)
  • D. Lim et al.

    Thermodynamic stability and guest distribution of CH4/N2/CO2 mixed hydrates for methane hydrate production using N2/CO2 injection

    J Chem Thermodyn

    (2017)
  • J. Yang et al.

    Flue gas injection into gas hydrate reservoirs for methane recovery and carbon dioxide sequestration

    Energy Convers Manag

    (2017)
  • P. Babu et al.

    A review of the hydrate based gas separation (HBGS) process forcarbon dioxide pre-combustion capture

    Energy

    (2015)
  • Y. Lee et al.

    CH4 recovery and CO2 sequestration using flue gas in natural gas hydrates as revealed by a micro-differential scanning calorimeter

    Appl Energy

    (2015)
  • J.-M. Herri et al.

    Gas hydrate equilibria for CO2–N2 and CO2–CH4 gas mixtures—experimental studies and thermodynamic modelling

    Fluid Phase Equilib

    (2011)
  • T. Lockwood

    A compararitive review of next-generation carbon capture technologies for coal-fired power plant

    Energy Proc

    (2017)
  • Y. Lee et al.

    Structure identification and dissociation enthalpy measurements of the CO2+N2 hydrates for their application to CO2 capture and storage

    Chem Eng J

    (2014)
  • G. Zylyftari et al.

    Modeling oilfield emulsions: comparison of cyclopentane hydrate and ice

    Energy Fuels

    (2015)
  • P.U. Karanjkar et al.

    Rheology of cyclopentane hydrate slurry in a model oil-continuous emulsion

    Rheol Acta

    (2016)
  • A. Ahuja et al.

    Calorimetric and rheological studies on cyclopentane hydrate-forming water-in-kerosene emulsions

    J Chem Eng Data

    (2015)
  • P. Babu et al.

    Morphology of methane hydrate formation in porous media

    Energy Fuels

    (2013)
  • Cited by (0)

    View full text