Alejandro Cardona

Estimates of in situ effective permeability of hydrate reservoirs are tens to hundreds of times less than most laboratory measurements on natural hydrate‐bearing sediments. The measured permeabilities are high because methane‐free water creates and expands flow channels in the sample during pressure core storage and laboratory testing. We study this process in natural hydrate‐bearing sediments retrieved from the deep‐water Gulf of Mexico. During a 90‐day methane‐free water flow experiment, the effluent methane concentration decreases from the equilibrium concentration to zero, with the concentration inversely proportional to flow rate. The permeability increases 20‐fold as hydrate saturation decreases from 0.88 to 0. In a second experiment, permeability remains constant when using methane‐saturated water as the permeant, but immediately increases when switching to methane‐free water. We develop a one‐dimensional reaction‐transport model to simulate our observations. Model results indicate that flow channeling is required to explain the data. Dissolution starts at the specimen outer rim during long‐term storage and evolves into a complex geometry during the flow test, leading to high effective permeability. Our interpretation of artificially high measured permeability on natural hydrate sediments is compatible with permeability estimates based on field data, synthetic hydrate samples, and theoretical models. In turn, a more realistic low in situ permeability would limit the ability to produce methane by pressure drawdown and hinder CO ₂ injection to form hydrate.

In the summer and fall of 2023, the University of Texas (UT) Deepwater Hydrate Coring Expedition (UT-GOM2-2) drilled, cored, made downhole measurements, and analyzed samples from the seafloor to the base of the gas hydrate stability zone at Site H, in the Walker Ridge Protracted Area Block 313 (Site H, WR313), in the Terrebonne Basin, deepwater Gulf of America (Gulf of Mexico). Analyses of data and samples from the expedition will inform biological, geochemical, and geomechanical models to constrain the role of gas hydrates in the carbon cycle and the potential for gas hydrates as an energy resource.

Methods used during the University of Texas (UT) Deepwater Hydrate Coring Expedition (UT-GOM2-2) include work done onboard the Helix Q4000 in the offshore Gulf of America (Gulf of Mexico), herein “the Gulf”, “dockside” in Salt Lake City, Utah, and some shore-based work in individual laboratories. The goal of this report is two-fold: to provide enough detail on the methods so they can be repeated by others; and to provide a reference document for the team to enhance cross-disciplinary understanding and knowledge. Methods include drilling operations, depth references and depth modification, downhole tool deployment, coring tool performance assessment, core processing, lithostratigraphy, biostratigraphy, as well as physical properties, including core logging and imaging, rock magnetism, dissolved methane concentration, hydrate saturation, microbiology, and geochemistry. An extensive amount of operational work and planning was required before mobilization of the expedition to permit, build mobile labs, and test downhole tools for deepwater drilling.

Pressure and conventional cores were collected at Site H of the Walker Ridge Protracted Area Block 313 in the Terrebonne Basin, deepwater Gulf of America (Gulf of Mexico) during the University of Texas (UT) Deepwater Hydrate Coring Expedition (UT-GOM2-2). Pressure and conventional cores were collected continuously to a depth of 155.1 meters below the seafloor (mbsf). At deeper depths, cores were taken periodically from hydrate-bearing sands and their bounding muds to a total depth of 861.3 mbsf. 162.6 m of conventional core and 54.8 m of pressure core were recovered. Twelve temperature measurements were made between 27.1 and 144.5 mbsf to determine the geothermal gradient. At the seafloor, more than 4 m of sandy silt of unknown origin was encountered. Beneath this sand, to a depth of about 200 mbsf, the section was composed of interbedded mud and biogenic carbonate ooze. The ooze correlated to low density and high porosity intervals observed in the previously acquired logging while drilling (LWD) data and as measured. These ooze intervals also correspond to lighter sediment color, increased Ca content based on X-ray florescence (XRF) core scanning, and increased calcareous nannofossil abundance. Calcareous nannofossil biostratigraphy constrains the entire record to the Pleistocene (< 0.91 million years), with a pronounced increase in sedimentation rate with depth. Below 200 mbsf, the section was predominantly composed of mud with two thicker, hydrate-bearing coarse-grained intervals, which are commonly known as the Blue and Orange sands. The dissolved gas concentration was quantified from pressure cores. In the shallow section, dissolved methane concentration increased below the sulfate-methane transition zone (SMTZ) and reaches saturation (the limit of solubility for methane) at 147 mbsf. Gas expansion was very common in conventional and depressurized pressure (conventionalized) cores below the SMTZ. At deeper depths, the methane concentration within muds bounding the Blue and Orange reservoirs was generally found to be less than saturation. The dissolved and hydrate gas composition is consistent with a microbial source, containing greater than 99.99% methane and only trace concentrations of ethane, propane, and butane. The methane to ethane ratio (C1 /C2 ) and the methane to ethane plus propane (C1 /(C2 +C3 )) decrease with depth down to at least 678 mbsf, mainly driven by the increase in ethane with depth. It is unclear if this trend continues through the Orange sand interval. The δ13C isotopic signature of methane ranges between -69.9 and -78.5 ‰ relative to the Vienna Pee Dee Belemnite (VPDB) standard. Pressure core recovery in sandy intervals was poor. However, pressure core logs of the Orange sand show intervals of low density and high velocity, which are indicative of high hydrate saturation. One pressure core was degassed and the average hydrate saturation in the core was determined to be 24%. One core from within the Orange sand was composed of interbedded graded sandy silt and mud. The sandy silts from this core are composed of mainly quartz and feldspar with some lithics. Most of the recovered pressure core samples are maintained at near in-situ pressure and temperature (within the hydrate stability field) at the University of Texas Pressure Core Center awaiting analysis. In the shallow section, samples will be used to determine the flux of organic carbon through the basin system, find the rate at which that carbon was consumed, and understand the microbial population responsible for these processes. In the deeper section, samples from in and around the hydrate reservoirs will be used to determine the petrophysical properties of the reservoir and bounding seals in these systems.

Phase 5 of the Deepwater Methane Hydrate Characterization and Scientific Assessment research project (DOE Award No. DE-FE0023919) occurred from Oct. 1, 2020 to Nov. 15, 2023. Throughout Phase 5, UT performed all aspects of project management and planning according to the award, project management plan, and statement of project objectives (Task 1). UT maintained and augmented the capability to transport, store, manipulate and analyze pressure cores (Task 13). UT’s hydrate core effective stress chamber can now run tests at effective stresses up to 20 MPa. A benchmark study was conducted and confirmed that the K0 permeameter accurately estimates geomechanical and petrophysical properties of geomaterials under uniaxial strain conditions. UT continued to analyze remaining UT-GOM2-1 pressure cores from GC955 (Task 10). A long-term hydrate dissolution test indicated that laboratory-based effective permeability measurements of hydrate-bearing sediments are hundreds of times greater than the in-situ effective permeability of hydrate reservoirs. Constant-rate uniaxial strain experiments indicated that hydrate-bearing sediments behave visco-plastically when subjected to stress. Gas analyses of hydrate-bearing cores from GC955 are consistent with the occurrence of predominantly microbial methane, with few notable exceptions consistent with thermogenic gas signatures. Results from a numerical model simulating gas production from methane hydrate reservoirs was applied to a constant stress depressurization experiment, indicating that the rate of hydrate dissociation and gas production is controlled by heat supply to the hydrate dissociation front. Throughout Phase 5, project scientists presented thirty-one abstracts, talks, or posters on research results at eight conferences. Ten journal articles were published in a second AAPG special volume dedicated to the scientific results of UT-GOM2-1. The PCTB Mk. 5 was completed and underwent bench testing and a full-function land test (Task 14). The UT-GOM2-2 Operations Plan and UT-GOM2-2 Prospectus: Science and Sample Distribution Plan were completed (Task 11). UT Austin executed an agreement with Helix ESG to perform UT-GOM2-2 with the Helix Q4000 MSV (Task 12). All preparations for UT-GOM2-2 were completed, including regulatory compliance and permitting, science and operational workshops, training, procurement and mobilizing equipment and staff (Task 14). The UT-GOM2-2 Deepwater Hydrate Coring Expedition was completed between Jul. 30 and Sep. 28, 2023 in the Gulf of Mexico, WR313 (Task 16). Two wells were drilled and cored at the location of the existing JIP II H001 LWD well. A total of 560 ft of conventional core, 180 ft of pressure core, and 11 in-situ temperature measurements were taken. UT achieved the highest scientific priority of recovering pressure cores from the Orange and Blue hydrate-bearing sands and their seals. Images and measurements of velocity and density in these cores show the presence of sand with high concentrations of hydrate. Onboard the Q4000, pressure cores were logged and X-rayed. Dissolved methane was measured and 9 samples were stored in liquid nitrogen for microbiological analysis. Conventional cores underwent infra-red imaging and measurement of sediment strength. A total of 350 subsamples were taken for of microbiology, paleontology, and gas and fluid composition analysis. After UT-GOM2-2 marine demobilization all cores were shipped to land for additional sampling and analysis and/or long term storage. Conventional cores underwent CT scanning and additional sampling and measurement for geomechanical analyses, splitting, and visual characterization. All pressure cores and conventional cores were delivered to UT Austin for long term storage, analysis, and sample distribution. These cores will form the foundation of future work measuring hydrate concentration, strength, and permeability of the reservoirs and their seals (Task 17).

The fractal topography of fracture surfaces challenges the upscaling of laboratory test results to the field scale, therefore the study of rock masses often requires numerical experimentation. We generate digital fracture analogues and model invasion percolation to investigate the capillarity-saturation P_c-S_w fracture response to changes in boundary conditions. Results show that aperture is Gaussian-distributed and the coefficient of variation is scale-independent. The aperture contraction during normal stress increments causes higher capillary pressures and steeper P_c-S_w curves, while shear displacement results in invasion anisotropy. The three-parameter van Genutchen model adequately fits the fracture capillary response in all cases; the capillary entry value decreases with fracture size, yet the fracture P_c-S_w curve normalized by the entry value is size-independent. Finally, we combine the fracture and matrix response to infer the rock mass response. Fracture spacing, aperture statistics and matrix porosity determine the rock mass capillarity-saturation P_c-S_w curve. Fractures without gouge control the entry pressure whereas the matrix regulates the residual saturation at high capillary pressure P_c.

The in‐situ stress state and geomechanical properties of hydrate‐bearing sediments impact hydrate formation and gas production strategies. We explore the uniaxial strain compression and stress evolution of natural hydrate‐bearing sandy‐silts from Green Canyon Block 955 in the deep‐water Gulf of Mexico. We performed constant rate of strain uniaxial strain experiments, interrupted by periods where we held the axial stress constant, to explore the vertical deformation and the evolution of the ratio of lateral to axial effective stress (K₀) with time. The hydrate‐bearing sandy‐silt is stiffer and has a larger K₀ than the equivalent hydrate‐free sediment upon loading. During stress holds, the void ratio decreases sigmoidally with the log of time, and K₀ converges to isotropic conditions. We interpret that during loading, the hydrate bears the load and deforms. With time, the hydrate redistributes the load and K₀ increases. We used a viscoelastic model to describe the behavior. The model accurately captures deformation and K₀ trends but does not reproduce all the complex interactions of the hydrate with the porous skeleton. We anticipate that viscous effects within hydrate sediments will impact reservoir compression and stresses during production (hours to days), result in isotropic stress state over geological timescales, and explain the creeping movement in submarine landslides.