|Characterization of Natural Hydrate Bearing Sediments and Hydrate Dissociation Kinetics||Last Reviewed 12/15/2012|
Work conducted under this field work proposal (FWP) includes two distinct phases. Ongoing Phase 2 work is discussed directly below. Click here to review the completed, Phase 1 work, associated with this FWP.
Phase 2 Project Information
Characterization of Natural Hydrate Bearing Core Samples
The overarching goal of this project is to gain an improved understanding of the dynamic processes of gas hydrate accumulations in geologic media by combining laboratory studies, numerical simulation, and analysis of shipboard infrared imaging of hydrate core samples. This project comprises four principal components: (1) fundamental laboratory investigations, (2) numerical simulator development and verification, (3) hydrate core characterization and analysis, and (4) applied laboratory and numerical investigations.
Pacific Northwest National Laboratory (PNNL), Richland, Washington
Numerical Simulator Development and Verification: Numerical simulation provides scientists and engineers with analytical tools for investigating different strategies for producing natural gas hydrate-bearing geologic formations. A fundamental objective for this project is to develop numerical simulators with capabilities for investigating a broad spectrum of coupled processes and production technologies. Corollary to this objective is the validation of the developed numerical tools against experimental observations at the laboratory and field scales. This project has supported the development of two numerical simulators: an equilibrium-based simulator and a kinetic-based simulator. The equilibrium-based simulator (STOMP-HYD) was used during the DOE-NETL sponsored international code comparison activities. The kinetic-based simulator (STOMP-HYD-KE) will be used to model ternary hydrate systems involving a mixture of CH4, CO2, and N2 guest molecules.
Hydrate Core Characterization and Analysis: A key issue in analyzing hydrate samples retrieved from the field is pressure loss during handling and the unavoidable warming of the sampler as it rises to the surface that will induce varying amounts of hydrate decomposition. The effects of partial hydrate decomposition on the measured properties of core samples are at present very poorly understood. A much better fundamental understanding of the effects of hydrate saturation and thermal and pressure cycling on the properties of hydrate-bearing sediments is needed. Infrared data collected on hydrate core samples during the Ocean Drilling Program (ODP) Leg 204 expedition (Hydrate Ridge); the Integrated Ocean Drilling Program (IODP), Expedition 311 (Cascadia Margin); and the Indian National Gas Hydrate Research Program (NGHP), Expedition 01 (east and west coasts of the Indian subcontinent and the Andaman Islands) will be analyzed to quantitatively assess hydrate abundance and compare hydrate occurrence as determined from well logs and other hydrate proxies.
Applied Laboratory and Numerical Investigations: Experimental research has demonstrated the value of using N2 together with CO2 as the injectant in the guest-molecule-exchange technology for producing natural gas hydrates from both sub-oceanic and arctic deposits. The benefit of using N2 with CO2 comes from the preference for N2 to displace CH4 from the small cages in the sI hydrate structure,; whereas, CO2 preferentially displaces CH4 from the large cages. Pure CO2 sI hydrates naturally form with CO2 occupying both the small and large cages. However, during the guest-molecule-exchange process, the thermodynamic preference is for CH4 to remain in the small cages. The additional benefit of combining N2 and CO2 is that the N2 shifts the hydrate equilibrium curve, which provides production operators with a system control. Injecting pure CO2 into a natural gas hydrate-bearing formation can only be controlled by the injection pressure and fluid temperature, where pressure is the primary control mechanism. By including N2 in the injectant mixture, the concentration of N2 can be used as an additional controlling mechanism. KIGAM has previously conducted and is currently conducting laboratory-scale experiments involving the injection of gaseous N2-CO2 mixtures into pure CH4 hydrate-bearing cores and sand columns.
Numerical Simulator Development and Verification: This project has supported the development of two numerical simulators for the production of natural gas hydrates from geologic deposits: STOMP-HYD and STOMP-HYD-KE. The STOMP-HYD simulator assumes equilibrium conditions and has simulation capabilities for four classes of production technologies: (1) thermal stimulation, (2) depressurization, (3) inhibitor injection, and (4) guest molecule exchange. For the guest-molecule exchange technology, STOMP-HYD is limited to binary hydrate systems with CH4 and CO2 guest molecules. The STOMP-HYD-KE allows for non-equilibrium conditions with a kinetic transition from non-equilibrium to equilibrium conditions. STOMP-HYD-KE has capabilities for the four classes of gas hydrate production technologies, but extends the hydrate system to a ternary capability with CH4, CO2, and N2 guest molecules. Numerical simulators with proven capabilities for modeling the production of gas hydrates from geologic accumulations provide scientists and engineers with needed tools for assessing reservoir potentials and developing production strategies. Producing gas hydrates from geologic reservoirs involves thermal, multi-fluid hydrological, and geomechanical processes, with a potential for geochemical considerations. Numerical simulation allows for consideration of these processes in a coupled manner.
Hydrate Core Characterization and Analysis: A fundamental objective for this component of the project is to develop techniques for correlating hydrate abundance to a wide range of physical, chemical, and stratigraphic properties that will enable accurate prediction of hydrate occurrence based on expected or known values of the correlated variables. During the course of this project, characterization techniques and procedures have been developed and applied during hydrate-core-recovery expeditions at Hydrate Ridge, Northern Cascadia Margin, and the India Ocean. Sampling approaches?for which there is close spatial coupling among infrared anomalies, pore-water analyses, and grain-size analyses?are enabling the development and verification of multivariate approaches for cross-correlating all three measurements.
Applied Laboratory and Numerical Investigations: The guest-molecule-exchange production technology for natural gas hydrates conceptually involves the injection of sI hydrate formers to replace those occupying the cages of a natural gas hydrate deposit in geologic formations (primarily sub-oceanic and arctic). The technology holds promise in that there is potential for effecting natural gas production without significantly altering the existing hydrate structure, thus avoiding changes in the mechanical stress of the system. A secondary benefit is that the process yields sequestration of CO2 in stable hydrate form. The challenge of the technology is devising production approaches that can be controlled and yield high exchange efficiencies. Numerical simulation is an invaluable tool for understanding complex physical processes such as those involved in the guest-molecule-exchange (e.g., multi-fluid flow, heat transfer, phase transitions, thermodynamic equilibrium, kinetics, geologic heterogeneity). Without experimental validation, however, numerical simulation becomes somewhat speculative in that the code developer is responsible for defining and implementing the important processes.
Hydrate Core Characterization and Analysis: During Leg 204 (2002) of the Ocean Drilling Program, extensive infrared (IR) images were obtained shipboard on cores from Hydrate Ridge, off the Oregon shore. The challenge has been to extract quantitative information on hydrate abundance from the IR data and to analyze that information by correlating it with stratigraphic information, geochemistry, physical property data, and other proxies for hydrate occurrence. A major step forward occurred during the Integrated Ocean Drilling Program?s Expedition 311 (2005) when gas hydrate IR anomalies were closely linked with pore water (chlorinity and salinity) anomalies and sediment grain-size. This was possible because of the enhanced IR data collection and display system developed for Expedition 311.
A comparable approach was used in the Indian National Gas Hydrate Program (NGHP) Expedition 01 (2006) using a similar data collection and display system. Infrared imaging was performed on all cores (except pressure cores), immediately after retrieval. A research-grade, automated IR imaging system was used to collect quantitative IR images on each 20 cm of core. Cores were imaged while still in the butyrate core liner. Handheld IR images were also taken of split core faces. Gas hydrate-bearing intervals are cooler than non-hydrate bearing zones due the endothermic nature of dissociation of gas hydrate to CH4 gas and liquid H2O. Temperature data were extracted from the IR images and used to define a ΔT or the extent of cooling produced by gas hydrate dissociation. Numerous variables control the IR temperature of the core liner surface, but the ΔT is readily measured for each gas hydrate zone in the core and thus provides a comprehensive and spatially detailed record of gas hydrate occurrence in marine sediments. In conjunction with PWC, the ΔT values are converted to % pore space occupied by gas hydrate, and then the gas hydrate occurrences are summed to estimate the total amount of gas hydrate in the GHOZ at the site.
Preliminary simulations of the recovery of hydrate-bearing cores from sub-oceanic deposits from the NGHP Expedition 01, Site 17, were conducted with STOMP-HYD to assess the ability to inverse model the hydrate saturation distribution in the core from the external infrared signal and core lithology. The core sample was modeled using a two-dimensional cylindrical domain as a layered system hydrate-bearing volcanic ash (3 cm) with hydrate-free marine sediments above and below. The numerical simulations were able to predict the infrared signature on the core surface, demonstrating the potential for in situ hydrate assessments using numerical simulator reconstruction. Color-scaled images generated from the numerical simulations with STOMP-HYD show the hydrate saturation and temperature after 50 minutes.
Numerical Simulator Development and Verification: The National Energy Technology Laboratory (NETL) and the U.S. Geological Survey (USGS) are guiding a collaborative, international effort to compare methane hydrate reservoir simulators. The goals of the effort are (1) to exchange information regarding gas hydrate dissociation and physical properties, enabling improvements in reservoir modeling; (2) build confidence in all the leading simulators through exchange of ideas and cross-validation of simulator results on common datasets of escalating complexity; and (3) establish a repository of gas hydrate-related experiment/production scenarios with the associated predictions of these established simulators that can be used for comparison purposes. PNNL has contributed to these code comparison activities by helping to define test problems, executing its STOMP-HYD simulator on the test problems, reporting simulation results, and participating in the analyses of the simulation predictions. Personnel conducting the code comparison study have completed six test problems and are currently working on the final suite of hydrate production problems for a range of hydrate accumulations on the Alaska North Slope.
A new simulator has been added to the series of STOMP simulators for modeling natural gas hydrate production from geologic accumulations. The first simulator in the series, STOMP-HYD, was capable of simulating four production technologies: (1) depressurization, (2) thermal stimulation, (3) inhibitor injection and (4) CO2 exchange. This simulator assumed equilibrium conditions between the mobile and hydrate components of the hydrate formers, CH4 and CO2. Experiments conducted at the Korea Institute of Geoscience and Mineral Resources (KIGAM), however, demonstrated that guest molecule exchange was a kinetic process, with respect to the time scales for flow through geologic media. The second simulator in the series, STOMP-HYD-KE, extended the capabilities of STOMP-HYD, by solving separate conservation equations for the mobile and hydrate components of the hydrate formers, CH4 and CO2. Hydrate formers transitioned between mobile and hydrate forms via hydrate formation, dissociation, and exchange, where all three mechanisms were controlled via kinetic rates. The STOMP-HYDT-KE simulator extends the capabilities of its predecessor by including a third hydrate former, N2. As with the two other hydrate formers, CH4 and CO2, the mobile and hydrate components of N2 are solved separately. In its full capability configuration, the STOMP-HYDT-KE solves nine conservation equations at each grid cell: (1) energy, (2) water mass, (3) mobile CH4 mass, (4) hydrate CH4 mass, (5) mobile CO2 mass, (6) hydrate CO2 mass, (7) mobile N2 mass, (8) hydrate N2 mass, and (9) inhibitor mass. The modular design of the simulator allows for one or two of the hydrate formers and/or the inhibitor to be eliminated from the solution. The transition between STOMP-HYD-KE and STOMP-HYDT-KE involved two significant changes in the code: (1) equation of state module and (2) ternary hydrate equilibria.
Applied Laboratory and Numerical Investigations: Experiments conducted at KIGAM on the exchange of carbon dioxide and nitrogen gas mixtures with methane hydrate in unsaturated, sand-filled columns, showed carbon dioxide and nitrogen in the effluent stream early in the production process, indicating the guest-molecule exchange was occurring as a kinetic process. The first attempt at modeling these experiments was conducted using the STOMP-HYD simulator, an equilibrium and binary hydrate simulator. The equilibrium assumption in STOMP-HYD yields full exchange of the injected carbon dioxide with the methane in the gas hydrate, yielding no carbon dioxide in the effluent stream, as observed. The binary-hydrate limitation in the simulator required ignoring the injected nitrogen. To resolve these limitations in the STOMP-HYD simulator, two joint projects were created: a kinetic-exchange project funded by KIGAM and a ternary-hydrate project funded by NETL. The kinetic-exchange project was completed with the development and application of the STOMP-HYD-KE project. The ternary-hydrate code is currently under development. In addition to the application of numerical simulation to guest-molecule exchange experiments, a more general short course on the application of the STOMP simulator to problems involving the coupled subsurface processes of multi-fluid flow and transport, heat transfer, geochemistry, and geomechanics was held at the International School for Geoscience Resources (IS-Geo) at KIGAM.
Current Status (December 2012)
A new simulator in the series of STOMP simulators for the production of natural gas hydrates from geologic accumulations has been developed and is currently being benchmarked. This new simulator, STOMP-HYDT-KE, has capabilities for modeling the production of natural gas hydrates via depressurization, thermal stimulation, inhibitor injection, and guest molecule exchange. The ?HYDT? in the simulator name indicates that it considers the ternary hydrate system CH4-CO2-N2, and the ?KE? in the simulator name indicates that the exchange of guest molecules and formation/dissociation of gas hydrates is modeled as a kinetic process. One of the biggest obstacles to the development of STOMP-HYDT-KE was devising an equation of state that was reasonably accurate, computationally efficient, and free of convergence failures. The CH4-CO2-N2 system is particularly difficult to resolve in the hydrate stability region, as the mixture is often near its critical point. A hybrid tabular-cubic equation of state was developed that overcame convergence issues associated with pure cubic equations of state near the mixture critical point. The hybrid scheme uses tabular data and an innovative interpolation algorithm to establish the existing phases, gas molar fractions, and phase compositions. The cubic equation of state is then used to calculate phase densities and fugacity coefficients. STOMP-HYDT-KE is currently being benchmarked and will be applied to an experiment conducted at KIGAM involving the injection of a CO2-N2 mixture into a CH4 hydrate bearing, unsaturated column of sand. Simulator benchmarking should be completed by the end of December 2012.
Project Start: September 1, 2001
Project End: December 31, 2012
Project Cost Information:
All DOE Funding *
FY01 - DOE Share - $100,000
FY02 - DOE Share - $40,000
FY03 - DOE Share - $200,000
FY04 - DOE Share - $210,000
FY05 - DOE Share - $228,000
FY06 - DOE Share - $450,000
FY07 - DOE Share - $250,000
FY08 - DOE Share - $94,000
FY09 - DOE Share - $260,000
FY10 - DOE Share - $130,000
Total Funding to Date - $1,962,000
* Includes Phase I Funding
NETL ? John Terneus (email@example.com or 304-285-4254)
PNNL – Mark White (firstname.lastname@example.org or 509-372-6070)
In addition to the information provided here, a full listing of project related publications and presentations as well as a listing of funded students can be found in the Methane Hydrate Program Bibliography [PDF].
Mt Elbert HXRD Letter Report [PDF-230KB] July 2009
Topical Report NGHP-01 Hydrate Samples High-Pressure X-Ray Diffraction Results [PDF-407KB]
2008 Hydrate Peer Review [PDF-2.32MB] - "Characterization of Natural Hydrate Bearing Core Samples"
2008 Hydrate Peer Review [PDF-1.55MB]- "Comparative Assessment of Advanced Gas Hydrate Production Methods" and "International Hydrate Code Comparison"
Phase 1 Project Information
Characterization of Methane Hydrate Bearing Sediments and Hydrate Dissociation Kinetics
In this project, new methods were developed to conduct quantitative measurements of flow and transport properties of hydrate-bearing sediments and methane hydrate dissociation rates. Efforts were mainly associated with equipment development in the work completed to date, but the results demonstrate proof-of-principle.
Direct measurements of gas permeability in hydrate-bearing sediments are rare in the literature, principally because of the difficulties involved in stabilizing the hydrate under high pressure. In this project, the approach for conducting permeability measurements with unconsolidated sediments was to pack a clay-lined pressure cell with artificial 20/30 sand in a temperature-controlled apparatus. After charging the sand pack with a residual water saturation and methane, the system was cooled to allow the formation of methane hydrates. Evidence of both ice and hydrate formation was observed in the temperature profiles. Gas flow rate was then recorded across a range of pressure differentials.
The data show very little dependence of gas flow rate on hydrate saturation until pore saturations exceed 50%. These results are consistent with the wetting behavior expected in unsaturated sediments where water preferentially coats the grains of the porous medium and occupy the smaller pores. Since, gas flow principally occurs through the larger pores, it is only after the larger pores begin to fill at hydrate saturations greater than 50% that a marked drop in permeability occurs. Existing models for estimating relative gas permeability did not match the change in gas permeability observed at hydrate saturations below 50%. Alternative models are presently being explored in an attempt to explain the results.
Basic parameters lacking from studies of gas hydrates are the translational diffusion characteristics of the hydrate forming species in sediments as a function of the hydrate saturation. Nuclear magnetic resonance (NMR) spectroscopy can be useful in studying the diffusion of molecular species in porous media, since it allows the observation of molecular transport in the pores without interfering with the diffusion process. Pulsed field gradient (PFG) NMR allows for precise determination of self-diffusion coefficients in porous media systems. However, high-pressure applications of PFG-NMR are extremely rare.
Proof-of-principle PFG-NMR measurements were performed at ambient pressure with mixtures of deuterated tetrahydrafuran (THF-d8) and water in the presence of quartz sand. THF-d8 was used in this test case to avoid interference with the water signal. PFG-NMR measurements were carried out over a range of THF/water concentrations. To induce hydrate formation, the sand/solution was frozen in the NMR tube before running the PFG-NMR experiments with visual confirmation of the solid hydrate in the tube.
Values for the self-diffusion coefficient of water at 25°C and 2°C were comparable to values reported in the literature. THF/water solutions without hydrate showed a slower diffusion coefficient for water than for the pure water case, as would be expected as the water activity declines with increasing THF-d8 concentration.
A comparison of the diffusion coefficient data with and without quartz sand present (no hydrate) shows a reduction in diffusion coefficient on average of 37% in the quartz sand; identical to the porosity of the sand pack. Surprisingly, a comparison of the water diffusion coefficient with and without hydrate present showed essentially no change. Free water is present in the 10% and 15% THF solutions and the signal from the PFG-NMR measurement is apparently dominated by this free water. The 20% THF solution, however, should have completed converted to hydrate and thus water diffusion coefficients should have dropped dramatically. Additional measurements are in progress to determine if incomplete conversion to hydrate occurred in these experiments.
To conduct measurements of hydrate dissociation kinetics in porous media as function of pressure and temperature, especially at P-T conditions near hydrate equilibrium conditions, new methods of analysis and customized laboratory equipment needed to be developed. The approach was to form gas hydrate with deuterated methane (CD4) and then monitor hydrate dissociation by tracking the mass of CD4 released as a function of time under a back pressure of CH4. Evolution of CD4 was followed by monitoring the CD4 concentration with Raman spectroscopy or with precision gas chromatography. Because CD4 can be easily distinguished from CH4 with either of these analytical methods, it was expected that hydrate dissociation rates could be monitored even at CH4 pressures near the hydrate equilibrium pressure. The thermodynamic properties of CH4 and CD4 are so similar that any isotope effects on the dissociation rate can be neglected.
Initial work involved the development and testing of a scanning laser Raman system that could directly interrogate the composition of gas hydrate formed in porous media within a cell capable of forming gas hydrate at relatively low pressure (<70 psi). Considerable effort was focused on the development of software for controlling the position of a fiber optic probe head and for automatic collection and storage of Raman spectra. After successful proof-of- principle testing of the low pressure system, a high-pressure capable dissociation kinetics cell was designed and built. Dissociation kinetics measurements using the pressure cell are in progress.