Depositional Environment and Diagenetic History of Gas Hydrate-Bearing Sediments of the Alaska North Slope (HYDRATE 02 Geo Data Well)

Abstract

We characterized the lithological, solid-phase geochemical, and rock magnetic properties of sediment cores from subpermafrost strata on the Alaska North Slope collected as part of the U.S. Department of Energy (DOE), the Japan Organization for Metals and Energy Security (JOGMEC), and the U.S. Geological Survey (USGS) Collaborative Gas Hydrate Research and Development (R&D) Project in Alaska. Pressure cores were recovered from the HYDRATE 02 Geo Data Well (GDW) at the Kuparuk State 7-11-12 site through two hydrate-bearing sand reservoirs, the B1 and D1 sands, and their finer-grained bounding sediments. Both the B1 and D1 sands are quartz-rich, chert-bearing, very fine litharenite sands. These submature to immature sands are poorly to moderately sorted and angular with local fine-scale cross-stratification. The B1 sand has larger grains and a lower clay content as compared to the D1 sand. These sands are the upper portions of upward coarsening progradational sequences. The clay- and silt-rich bounding muds exhibit fine-scale wavy and lenticular laminations, suggesting episodic high energy deposition events. Overall, these sediments were deposited in a shoreface to nearshore marine environment during the Paleogene, modulated by changing depositional energy associated with sea level variation. Organic matter δ¹³C and C/N composition of these sediments indicate a source mixture of land plants and freshwater plankton, with possible minor presence of oil within the B1 sand. These sediments contain high and variable total sulfur (TS) content, suggesting porewater conditions influenced by variable salinity over time. Rock magnetic analyses indicate precipitation of greigite in the mud beneath the D1 sand, in the basal C shale overlying the B1 sand, and the mud beneath the B1 sand, likely due to intervals of sulfate limitation during periods of lower Arctic Ocean salinity. Based on pyrite abundance, TS, and δ³⁴S-TS composition, the D1 sand and bounding sediments experienced more substantial sulfide precipitation. The unit E basal shale overlying the D1 sand contains a 0.25 m thick pyrite-dominant sand, indicating local bottom water anoxic conditions. X-ray fluorescence scanning and discrete measurements provide additional supporting interpretation of the depositional and diagenetic processes that have occurred at this site and track the relative amount of drilling fluid contamination with barium abundance in the core samples.

1. Introduction

Gas hydrates have been widely observed in permafrost environments. − In particular, gas hydrates hosted in sand deposits represent a potential energy resource, − and subpermafrost gas hydrate reservoirs have been a major focus of energy research for several decades through analysis of industry well logs, coring projects, and field experiments. −

The lithology of hydrate-bearing sediments has an important impact on hydrate formation processes and dissociation behavior during the production of gas hydrates. Sediment properties related to depositional processes, such as grain size and bedding, can directly influence porosity, permeability and heat transfer, enhancing or limiting fluid flow and hydrate formation. − In addition, postdepositional diagenetic precipitates can fill pore space or preserve porosity with depth. , As such, the composition of sediments, along with the structure style, exerts substantial control on the hydrate petroleum system. , Collecting and analyzing sediment cores from the production site allows us to characterize the lithological composition of reservoir and bounding sediments and interpret the geological evolution of the sequence as well as ground-truth well log measurements.

Depressurization of a gas hydrate-bearing reservoir was tested in 2023–2024 on the Alaska North Slope (ANS) through an extended gas hydrate production test as part of a collaboration between the U.S. Department of Energy (DOE), the Japan Organization for Metals and Energy Security (JOGMEC), and the U.S. Geological Survey (USGS). Over nine months of production activities focused on producing gas from a high saturation subpermafrost, sand-rich gas hydrate reservoir, the effects were monitored across multiple boreholes. , Prior to production operations, pressure cores were collected across two gas hydrate-bearing sand reservoirs, the D1 and B1 sands, and their bounding sediments. The shallowest cored interval sampled from within the basal E shale, through the D1 sand, and into fine-grained sediments below the D1 sand. The deeper cored interval sampled from basal C shale, through the B1 sand, and into finer-grained sediments below the B1 sand. We present a comprehensive characterization of core lithology, bulk elemental and isotopic composition, and rock magnetic properties of pressure cores collected from the Geo Data Well (GDW), HYDRATE-02 at the Kuparuk 7-11-12 site to better understand the depositional environment and diagenetic processes across these hydrate-bearing intervals associated with the extended-duration production test as part of the JOGMEC–DOE–USGS Collaborative Gas Hydrate R&D Project in Alaska.

2. Geologic Background

The Alaska North Slope is underlain by the Colville foreland basin that initially formed due to rifting of the Arctic Composite Terrane, followed by compressional orogeny of the Brooks Range in the Jurassic and Cretaceous. The basin is bounded by the Brooks Range to the south and the Barrow Arch to the north near the modern Arctic Ocean coastline. The uplifted basement of the Barrow Arch separates the Colville basin from the Arctic Ocean Basin. Sediment thickness of the Colville Basin increases toward the Brooks Range. The oldest stratigraphic sequence of the basin is the Ellesmerian sequence, which consists of passive margin clastic sediments and carbonates deposited from the Mississippian to Triassic prior to rifting. The Beaufortian sequence was deposited during the Jurassic and Early Cretaceous and consists of syn-rift marine sediments. The youngest sequence in the basin is the Brookian sequence, which consists of terrestrial and shallow marine sediments sourced from the Brooks Range. − The Brooks Range orogeny caused folding and thrust faulting to provide structural traps, while foreland sedimentation buried source rocks into the temperature window for petroleum generation. ,

The Sagavanirktok Formation within the Brookian sequence contains nonmarine and marine sediments deposited between the Paleocene and Pliocene. , The Sagavanirktok Formation hosts multiple sand-bearing units that host subpermafrost gas hydrate accumulations identified from well logs across the Prudhoe Bay-Kuparuk River areas. ,, These accumulations are known as the Eileen Gas Hydrate Trend and consist of six mappable units that contain gas hydrate-bearing sands known as the A (oldest) through F (youngest) units. The C and D units, as well as the uppermost interval of the B unit, were cored during the Mount Elbert Gas Hydrate Stratigraphic Test Well in 2007. These cores revealed the presence of very fine to fine sands within the C and D units, bounded by silty clay and silt-dominant sediments. Sidewall cores collected during the drilling of the HYDRATE 01 Stratigraphic Test Well (STW) in 2018 as part of the JOGMEC–DOE–USGS Collaborative Gas Hydrate R&D Project in Alaska recovered quartz-rich sand to clayey silt from the B and D units. As part of this project, a refined informal lithostratigraphic terminology was developed that recognized that several of these previously defined units contained multiple mappable sands as well as regionally extensive shale-rich units typically at the base (Boswell et al., 2021). The key zones identified are the hydrate-bearing sand at the top of unit D (the D1 sand) and another at the top of unit B (the B1 sand). The overlying clay-rich sediments (“seal”) above the D1 sand lie within the base of unit E and are called the “basal E shale”. Similarly, the “seal” above the B1 sand lies at the base of unit C and is called the “basal C shale”.

Based on biostratigraphic markers within palynological assemblages in cores collected at the Mount Elbert Well, the B1 sand was deposited during the Paleocene and the D1 sand was deposited during the Early Eocene. The Paleocene–Eocene Thermal Maximum (PETM) likely occurred during deposition of the basal C shale, although the PETM carbon isotope excursion has yet to be identified in these sediments.

3. HYDRATE-02 Geo Data Well Pressure Coring

The Geo Data Well (GDW) (API# 50029237280000) was drilled at the Kuparuk 7-11-12 State well pad, located at 70° 19′ 1.84032″ N, 149° 12′ 0.95289″ W at 16 m (52 ft) above mean sea level (Figure ). The total measured depth (MD) of the GDW was 1084 m (3558 ft). Between 365 m (1199 ft) and the bottom of the hole, the well was drilled with an average inclination of 12.9° with a maximum of 19.3° (Figure ). Drilling operations occurred over 42 days between October 9 and November 19, 2022.

1.

Location map of the Alaska North Slope with the location of GDW at the Kuparuk 7-11-12 location. Locations of the gas hydrate research project are shown as red stars. Production pads are shown as blue boxes. Green stars show gas hydrate occurrences outside existing production pads. Modified from Boswell et al. (2022). Copyright 2022 American Chemical Society.

2.

(a) Hydrate 02 GDW overview showing the vertical depth and horizontal offset of the well. The pressure cored intervals are shown in orange, and the base of ice-bearing permafrost and the base of gas hydrate stability are shown by dashed lines. (b) Various horizons are indicated by black triangles. Details of pressure coring operations can be found in ref . From ref , Copyright 2026 American Chemical Society.

All cores were collected using the HPTC IV (high pressure temperature corer) owned by JOGMEC, which was a modified version of HPTC III. Coring operations occurred between October 22 and November 6, 2022. In the upper cored interval, 11 pressure cores were collected from the D1 sand and bounding shale sediments between 722.07 m (2369.0 ft) and 755.78 m (2479.6 ft) measured depth (Figure ). In the lower cored interval, 12 pressure cores were recovered in the B1 sand and the bounding shale sediments between 875.69 m (2873.0 ft) and 915.31 m (3003.0 ft) measured depth (Figure ). A total of 66.9 m (219.56 ft) of core was recovered (91.2% recovery). All pressure cores returned to the rig floor at pressure.

3.

Summary of core recovery through the D1 and B1 sands and bounding units in the GDW. (a,c) P-wave velocity (V _p) of pressurized cores, as measured by the Geotek Pressure Core Analysis and Transfer System (PCATS). (b,d) The recovery of pressure cores in the D and B intervals, respectively. From ref , Copyright 2026 American Chemical Society.

In this paper, we present results in depth scale of “measured depth” (MD) in meters below the rig floor, which is based on the length of the drill pipe deployed down the hole. This allows for direct comparison to other measurements along the primary measurement scale. We also present results with a scale of total vertical depth (TVD) below sea surface (ss) in millimeters, allowing for true vertical depths and comparisons to other holes with different borehole inclinations and surface elevations. After recovery at the Kuparuk 7-11-12 site, the core and liner were transferred from the HPTC autoclave to the Geotek Pressure Core Analysis and Transfer System (PCATS). The whole round core was initially scanned by using linear X-ray imaging, P-wave velocity, and γ ray attenuation density measurements. A second scan for 3-D X-ray computed tomography (CT) was done after the initial scan. Based on these initial scans, select samples were transferred into pressurized storage chambers for quantitative degassing for transport from the ANS to either Geotek Coring Salt Lake City (SLC), UT, United States, or National Institute of Advanced Industrial Science and Technology (AIST) in Sapporo, Japan, for further analysis. Depressurized core sections were frozen after degassing and transported to SLC. Additional core subsections were quantitatively degassed and frozen in SLC. The frozen conventionalized core sections were split into archive and working halves with a saw.

4. Core Analysis Methods

4.1. Core Imaging and Logging

Archive-half core sections were imaged and logged with the Geotek BoxScan automated multisensor core logger. Core photos were taken with a Geotek Geoscan linescan camera. Volume-normalized magnetic susceptibility (κ) was measured using a Bartington MS3 meter and MS2E point sensor. κ was measured at a 1 or 0.2 cm resolution, integrated over 5 s. X-ray fluorescence (XRF) was also measured at a 1 or 0.2 cm resolution using an Evident Vanta XRF spectrometer in Geochemistry mode with a 30 s measurement interval. We measured a total of 34 elements: Mg, Al, Si, P, S, K, Ca, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Se, Rb, Sr, Y, Zr, Nb, Mo, Cd, Sn, Sb, Ba, W, Hg, Pb, Bi, Th, and U. The elemental abundances are reported as parts per million (ppm) but are uncalibrated during analysis. We use the Si/Al ratio as a quartz proxy and the Zr/Rb ratio as a grain size/heavy mineral proxy. ,

4.2. Visual Core Description

Archive-half core sections were visually described in an approach like those used during scientific ocean drilling programs (e.g., Ocean Drilling Program, International Ocean Discovery Program). , Sediments were classified based on their textural and compositional characteristics. Sedimentary structures were noted. The thickness of beds and lamination was noted. A sand gauge was used to aid in grain size classifications. Sediment color was documented by using charts from the Munsell Soil Color Book (Munsell Color). Visual descriptions were summarized with core photos and other core log data using Strater software (Golden Software LLC). Detailed visual descriptions, core photos, X-ray images, and select physical properties are presented for each core in the Supporting Information (Figures S1–S24).

4.3. Core Sub-sampling

Discrete samples were selected from working half core sections to represent major and minor lithologies and used to analyze lithologic characteristics, elemental and isotopic compositions, and rock magnetic properties. At each sampling depth, an 8 cm³ (2 × 2 × 2 cm) magnetic cube sample was collected from a 2 cm interval. The remaining half round of sediment from the 2 cm interval was sampled and later split into petrographic samples and dried and powdered samples. During initial sampling of the half round, the outer rind (several millimeters) was scraped off in an attempt to minimize drilling fluid contamination. 83 samples were collected from the D1 sand and bounding intervals (average sampling resolution of 33 cm in recovered core). 139 samples were collected from the B1 sand and bounding intervals (average sampling resolution of 28 cm).

4.4. Petrographic Analysis

A very small subsample was collected with a toothpick to create smear slides. The sediment was dispersed on a glass microscope slide (25 mm × 75 mm) and dried, and the slides were covered in a glass coverslip (30 mm × 22 mm) adhered with Norland Optical Cement No. 61 (Edmund Optics). Smear slides were described and imaged using a Leica DM750 petrographic microscope. Approximately 1 g of sediment was selected for coarse fraction analysis and dispersed in a 500 mL polycarbonate Erlenmeyer flask with 40 g/L sodium hexametaphosphate, shaken for 5 min, left overnight, and shaken for another 5 min. The sediment suspension was sieved through 63 μm (no. 230), 250 μm (no. 60), and 2 mm (no. 10) sieves, representing the fine-to-very fine sand, medium-to-very coarse sand, and pebble fraction, respectively. Each coarse fraction was dried and observed under a Leica Ivesta3 stereomicroscope or a Leica DMS1000 digital microscope. A select number of samples shared with Stratum Reservoir were impregnated with epoxy and used to create 30 μm-thick thin sections. These thin sections were stained for more accurate identification of plagioclase and potassium feldspars and then imaged and analyzed with point counts.

4.5. Grain Size and Mineralogy

Sixty-six samples were measured for laser diffraction particle size analysis at the Stratum Reservoir after cleaning with toluene and chloro-methanol azeotrope. Twenty-eight samples were measured at the Stratum Reservoir for bulk and clay mineralogy using quantitative X-ray diffraction (XRD). Details of the grain size and XRD measurement approaches have been previously published as part of analysis of Mount Elbert conventional core and STW sidewall core sediments. , Oils from drilling mud were extracted from the sediments using chloroform–methanol azeotrope in a Soxhlet extractor and the cleaned sediments were measured using a Malvern Mastersizer 2000 particle size analyzer. The full grain size distribution data allow us to interpret sorting and skewness statistics that provide insight on depositional environment and reservoir conditions. We also use grain size data measured at a higher down-core resolution (296 samples) using laser diffraction particle size analysis at AIST. These grain size samples were measured using a Microtrac MT3300 EX II particle size analyzer after oil from drilling muds were extracted using toluene and organic matter digested with hydrogen peroxide. Results of laser diffraction grain size measurements can vary between analytical devices and optical parameters, especially for clay-rich sediments. Sample pretreatments, even with centrifugation of the samples, may result in small variations in the fine-grained fraction. In this case, each grain size data set adds its own advantages for characterizing these sediments (full distribution statistics available versus high down-core resolution).

An additional four samples (duplicates of those measured at the Stratum) were measured for bulk and clay mineral abundance by XRD at the James Hutton Institute. For bulk (whole sample) quantitative analysis, samples were wet ground in a McCrone mill and spray dried to produce random powder specimens. XRD patterns were measured over a range of 65° 2θ or more using Cu radiation. Three of the samples were measured using a normalized full-pattern reference intensity ratio method. , For the nontypical composition of a high pyrite sample, quantitative analysis was made by the Rietveld method. For semiquantitative analysis of clay minerals, clay fractions of <2 μm were separated by timed sedimentation, prepared as oriented mounts using the filter peel transfer method and scanned from 2 or 3° 2θ to 45° 2θ using Cu radiation. Three scans are made, one in the air-dried state, one after glycolation, and one after heating to 300 °C for 1 h. Clay minerals identified were identified and their abundance quantified using a mineral intensity factor approach based on calculated XRD patterns.

4.6. Bulk Elemental and Isotopic Analysis

The remaining sediment was dried for 48 h in an oven at 40 °C and powdered using a mortar and pestle. Sediment powders were run for total carbon (TC), total nitrogen (TN), and total sulfur (TS) using an Elementar Unicube CHNS Elemental Analyzer at the University of New Hampshire. A ∼10 mg subsample was weighed out and folded untreated in 5 mm × 9 mm tin capsules. Replicate samples were run every 11 samples. Additional splits of the sediment powders were weighed out (∼10 mg) in tin capsules, moistened with 50 μL of deionized water, and placed in a desiccator containing concentrated hydrochloric acid (HCl) to remove inorganic carbon. Replicate samples were run every 11 samples. These samples were analyzed for total organic carbon (TOC) and δ¹³C-TOC at the Washington State University Stable Isotope Core Laboratory using a Costech Analytical ECS 4010 elemental analyzer and ThermoFinnigan Delta Plus XP isotope ratio mass spectrometer. δ¹³C-TOC is reported as ‰ Vienna Peedee belemnite (VPDB). Replicate samples for TOC were within 0.05 wt %, and replicates for δ¹³C-TOC were within 0.37‰ VPDB. Inorganic carbon (IC) was calculated by subtracting TOC from TC. Additional powdered samples were measured for δ³⁴S-TS at the University of California Berkeley Center for Stable Isotope Biogeochemistry using a Eurovector EuroEA3028-HT Elemental Analyzer interfaced with a GV Instruments IsoPrime isotope ratio mass spectrometer. δ³⁴S-TS is reported as ‰ Vienna Canyon Diablo Troilite (VCDT). Duplicate samples were within 3.5‰ VCDT.

Element concentrations in discrete sediment powder samples were determined by using a Bruker CTX 800C benchtop XRF spectrometer in the USGS Woods Hole Coastal and Marine Science Center Sediment Analysis Laboratory. Measurements were acquired by using the Bruker GeoExploration Oxide3Phase calibration package, which consists of three sequential analytical phases. Each phase was run for 100 s for samples and 20 s for standards. Each sample was analyzed three times, and replicate analyses were averaged to obtain the reported concentrations. Mean values and standard deviations were calculated by using in-house spreadsheets to quantify analytical precision based on repeat measurements. Instrument performance and analytical accuracy were monitored at the beginning of each analytical session through repeated measurements of the Bruker CS-M2 calibration standard using the same calibration as samples. We analyzed four major oxides: Al₂O₃, SiO₂, MgO, and K₂O, and 42 elements: Ag, As, Au, Ba, Bi, Ca, Cd, Ce, Cl, Co, Cr, Cu, Fe, Ga, Hf, Hg, La, Mn, Mo, Nb, Ni, P, Pb, Pd, Pt, Rb, Rh, S, and Sb. Se, Sn, Sr, Ta, Th, Ti, Tl, U, V, W, Y, Zn, and Zr. In this work, we directly discuss these elements in our interpretation: Si, Al, K, Fe, S, Zr, Rb, Ba, Mo, and Cu.

4.7. Rock Magnetic Analysis

An 8 cm³ plastic paleomagnetic cube was collected at the same sampling depth as the elemental and isotopic analyses described above and measured for mass-normalized magnetic susceptibility (χ) using a Bartington MS3 magnetic susceptibility meter with an MS2B dual-frequency sensor. A calibration stone measured before each measurement set matched the known value within 0.5%. Each sample was measured at 465 Hz (low frequency, LF) and 4650 Hz (high frequency, HF), and the frequency dependence % of χ (χ_fd) is calculated as χ_fd = (χ_LF – χ_HF)/χ_LF × 100. A subset of samples was analyzed at the University of Minnesota Institute for Rock Magnetism. We measured anhysteretic remanent magnetization (ARM) on sediment cubes using a 2G Enterprises 760-R SQUID magnetometer with a track system and inline ARM. ARM was acquired in a peak alternating field (AF) of 150 mT in the presence of a DC bias field of 0.1 mT. The samples were then subjected to AF demagnetization to 170 mT in steps of 5 mT (5 to 30 mT), 10 mT (40 to 60 mT), and 20–30 mT (80 to 170 mT). Each sample was packed in a gel capsule and measured for magnetic hysteresis and backfield remanence at room temperature using a Princeton Measurements MicroMag Vibrating Sample Magnetometer (VSM). For the hysteresis loops, the maximum applied field was 1.2 T measured in 5 mT increments, with each step averaged over 1 s. Backfield remanence was acquired from 100 μT to 1.2 T over 100 steps, with each step averaged over 3 s. The VSM hysteresis loops and backfield curves were used to calculate the saturation magnetization (M _s), saturation remanence (M _r), coercivity (B _c), and coercivity of remanence (B _cr). M _r is also known as saturation isothermal remanent magnetization (SIRM). These parameters were used to further calculate the S-ratio, the ratio of the isothermal remanent magnetization (IRM) remaining after a reversed field of 300 mT is applied to the SIRM: S ₃₀₀ = −IRM_–300/SIRM. These parameters were also used to measure the hard coercivity IRM (HIRM) component: HIRM = (IRM_–300 + SIRM)/2. Hysteresis parameters were used to calculate the D _JH ratio for identifying magnetic iron sulfides: D _JH = (M _rs/M _s)/(B _cr/B _c). This ratio is often an indicator of a magnetite-dominant (<15 kA/m), mixed magnetite and magnetic iron sulfide (15 to 30 kA/m), or magnetic iron sulfide-dominant (>30 kA/m) magnetic mineral assemblages. ,

A subset of the VSM-measured samples was characterized for high temperature κ using a high-temperature Geofyzika KLY-2 KappaBridge AC Susceptibility Bridge. Powdered sediment samples were measured over a range from 40 to 700 °C at ∼3 °C intervals. Temperature ranges with a rapid decrease in κ represent Curie or unblocking temperatures that can be used to identify magnetic minerals present in the sample. For instance, magnetite and hematite lose their magnetization at 580 and 680 °C, respectively. The magnetization of greigite decreases substantially between 270 and 350 °C during heating. , A subset was also analyzed using a Quantum Designs Magnetic Property Measurement System (MPMS) 5s cryogenic magnetometer at temperatures between 20 and 300 K. Samples were subjected to field-cooling (FC) from 300 to 20 K in a 2.5 T field. Samples were then heated to 300 K with remanence measurements during warming. This sequence was repeated with zero-field cooling (ZFC) to 20 K, followed by acquisition of a low-temperature saturation isothermal remanent magnetization (LTISRM), and then heating to 300 K with measurement of ZFC remanence. Lastly, a room-temperature saturation isothermal remanent magnetization (RTSIRM) was acquired in a 2.5 T field, followed by cooling to 20 K, and heating to 300 K with RTSIRM remanence measurements during heating and cooling. These FC–ZFC–LTSIRM–RTSIRM sweeps result in remanence curves that can be used to identify low temperature transitions that are distinctive for several magnetic minerals.

5. Results

5.1. Lithostratigraphy

In both the B1 and D1 sands, visual observation of depressurized and split cores reveals relatively massive intervals of sand with very faint structure in the visual description of depressurized and split cores (Figures , S4–S6, and S16–S18). The sands are primarily very fine sands, with some fine sand and coarse silt grains. PCATS CT scanning of pressurized hydrate-bearing cores allows for visualization of sedimentary structures that would be disturbed due to hydrate dissociation during conventional core recovery. Orthogonal slab images of core CT data show millimeter-scale cross-laminations in centimeter-scale packages with low-angle bounding surfaces. Although the width of the core limits interpretation of larger structures, these packages are consistent with hummocky or swaley cross-stratification. Wavy mud drapes were visible in X-ray CT as dark (denser) laminations (Figure ). Higher angle dark laminations present in this sand lithofacies may be diagenetic bands, noted as cross-cutting stratification and persisting as relatively consistent angles within cores. This lithology was observed between 885.63 and 895.33 m MD (B1 sand) and from 731.57 to 739.26 m MD (D1 sand) (Figures h and h).

4.

Linescan and X-ray CT slab images of B1 sand reservoir sediments. The very fine reservoir sands are fairly massive with cross-stratification visible in X-ray constrained by low angle bounding surfaces. (a,b) Core section 16P-2. (c,d) Core section 17P-4. (e,f) Zoomed-in examples of darker laminations in the X-ray. Undulating laminations are likely mud drapes, while higher angle bands may be caused by diagenetic precipitates.

5.

Primary lithostratigraphic results for the B1 sand and bounding intervals (including the C unit basal shale). (a) Median grain size (D ₅₀) (green diamonds measured by Stratum Reservoir and black circles measured by AIST) and Zr/Rb from X-ray fluorescence (XRF) core scanning (brown lines) and discrete measurements (orange triangles). (b) % sand size fraction (purple squares measured by Stratum Reservoir and brown upside-down triangles measured by AIST). (c) % clay size fraction (red triangles measured by Stratum Reservoir and yellow-green circles measured by AIST) with the sum of aluminum, potassium, and magnesium from XRF core scanning (black lines) and discrete measurements (light blue circles). (d) Sorting index (orange squares) and skewness (blue circles). (e) Si/Al from XRF core scanning (blue line), discrete measurements (tan circles), and quartz from X-ray diffraction (XRD) (black diamonds). (f) Total tectosilicates (brown circles measured by the Stratum Reservoir and cyan triangles measured by the James Hutton Institute) and phyllosilicates (green squares measured by the Stratum Reservoir and upside-down purple triangles measured by the James Hutton Institute) from XRD. (g) K/Al from XRF core scanning (orange lines) and discrete measurements (brown diamonds) and the ratio of mixed layer illite–smectite clays (Mx I–S) to illite clays (I) from XRD (dark red triangles measured by the Stratum Reservoir and blue circles measured by the James Hutton Institute). (h) Lithology interpreted from the visual core description, microscopic description, and core scans. The top of the B1 sand is noted by a dashed line.

6.

Primary lithostratigraphic results for the D1 sand and bounding intervals (including the C unit basal shale). (a) Median grain size (D ₅₀) (green diamonds measured by the Stratum Reservoir and black circles measured by AIST) and Zr/Rb from X-ray fluorescence (XRF) core scanning (brown lines) and discrete measurements (orange triangles). (b) % sand size fraction (purple squares measured by the Stratum Reservoir and brown upside-down triangles measured by AIST). (c) % clay size fraction (red triangles measured by the Stratum Reservoir and yellow-green circles measured by AIST) with the sum of aluminum, potassium, and magnesium from XRF core scanning (black lines) and discrete measurements (light blue circles). (d) Sorting index (orange squares) and skewness (blue circles). (e) Si/Al from XRF core scanning (blue line) and discrete measurements (tan circles) and quartz from X-ray diffraction (XRD) (black diamonds). (f) Total tectosilicates (brown circles measured by the Stratum Reservoir and cyan triangles measured by the James Hutton Institute) and phyllosilicates (green squares measured by the Stratum Reservoir and upside-down purple triangles measured by the James Hutton Institute) from XRD. (g) K/Al from XRF core scanning (orange lines) and discrete measurements (brown diamonds) and the ratio of mixed layer illite–smectite clays (Mx I–S) to illite clays (I) from XRD (dark red triangles measured by the Stratum Reservoir and blue circles measured by the James Hutton Institute). (h) Lithology interpreted from the visual core description, microscopic description, and core scans. The top of the D1 sand is noted by a dashed line.

Particle size analyses from both sets of measurements indicate that the B1 sand (886.66–895.17 m MD) is a silty very fine sand with a median grain size (D ₅₀) ranging from 76 to 103 μm (mean: 89 μm) (AIST) (Figure a). Grain size analyses from Stratum over this same interval show a similar range from 88 to 104 μm (mean: 95 μm). The percent sand in the B1 sand is >66% in both sets of analyses (Figure b). Clay-sized content ranges from 4 to 9% (Stratum) or 1 to 4% (AIST) (Figure c). Based on the Folk and Ward sorting index, most sediments from the B1 sand and bounding units are poorly sorted (1.02 to 1.91), but one sample within the B1 sand is moderately sorted (0.95) (Figure d). The sorting index trends toward better sorting in the B1 sand relative to the transition and bounding units. The grain size distribution is strongly fine-skewed (0.38 to 0.58 Φ) in the B1 sand and lower transition and approaches fine-skewed to symmetrical (0.0 to 0.16 Φ) in the bounding muds (Figure d).

7.

Photomicrograph images of reservoir sands. Images of sieved coarse fractions (63 to 250 μm) from the D1 sand (a,c) and B1 sand (b,d). Thin section images from the D1 sand (e) and B1 sand (f). Petrographic analysis shows both the B1 and D1 sands to be quartz-dominant chert-bearing very fine sand with angular grains.

In the D1 sand (732.17–737.39 m MD), D ₅₀ ranges from 46 to 61 μm (mean: 54 μm) in AIST measurements, and D ₅₀ ranges from 54 to 78 μm (mean: 67 μm) in Stratum measurements (Figure a). The percent sand ranges from 35 to 48% (mean: 42%) in AIST measurements and 43 to 63% (mean: 54%) in Stratum measurements (Figure b). The clay-sized fraction ranges from 11 to 16% in AIST measurements and from 4 to 7% in Stratum measurements (Figure c). The larger discrepancy between the two data sets in the D1 sand is likely driven by a higher clay-sized fraction compared to the B1 sand. Clay minerals within the clay-sized fraction are subject to higher analytical uncertainty and possible loss during pretreatments. All samples from the D1 sands and bounding units are poorly sorted to very poorly sorted (1.48 to 2.12 Φ), and unlike the B1 sand, there is no trend toward moderate sorting within the D1 sand (Figure d). There is, however, a trend toward fine skewness in the D1 sand relative to the bounding muds, similar to the B1 sand. Skewness is strongly fine-skewed in the D1 sand (0.52 to 0.57 Φ) and is coarse-skewed, symmetrical, and fine-skewed in the C basal shale (−0.10 to 0.22 Φ). Skewness is fine-skewed to strongly fine-skewed in the lower transition and bound of the D1 sand (0.16 to 0.58 Φ).

Microscopic description of coarse fraction, smear slide, and thin section samples from the B1 and D1 reservoir sands reveals quartz-rich sands bearing lithic fragments (Figure ). The lithic fragments are dominantly chert with minor igneous, metamorphic, and shale fragments. Quartz sand grains are angular to very angular, while chert fragments are subangular to subrounded, suggesting earlier reworking (Figure ). There are minor occurrences of plagioclase, potassium feldspar, and muscovite mica. Heavy minerals such as iron oxides and zircon are very rare or absent. However, the Zr/Rb ratio shows a clear increase within the B1 and D1 sands that suggests that zircon abundance is higher within the sands (Figures a and a). Minor occurrences of round clusters of iron sulfides are present in the sands. Trace siderite was observed in smear slides from the B1 sand, but not in the D1 sand. These compositions and textures are relatively consistent across all reservoir samples. XRD measurements indicate a drastic increase in quartz and total tectosilicates within the B1 and D1 sands and a distinct decrease in total phyllosilicates (Figures e,f and e,f). The Si/Al ratio from XRF closely follows the pattern in XRD quartz (Figures e and e).

The seal sediments overlying the B1 sand (basal C shale, 875.69 to 885.53 m MD) are composed of interbedded silty clay, clayey silt, and silt on the scale of cm to tens of cm. Millimeter-scale silt laminations are common within silty clay intervals, while bioturbation is observed in silty clay and clayey silt intervals (Figures S12–S15). The clay-sized fraction of grain size measurements measured at a high resolution (AIST) shows a high degree of variation due to the interbedding of silt with silty clay and clay (Figure c). Wavy and lenticular bedding was present intermittently throughout this seal interval (Figures c,d and e). There is a lag deposit at 876.39 m MD composed of well-rounded chert granules and pebbles (Figure c). A scoured erosional surface, overlain by an irregular very fine sand, is observed at 880.38 to 880.42 m MD. Dark-colored sulfide bands, patches, and mottling are common throughout the cores recovered from the basal C shale. Microscopic analysis of smear slides and coarse fractions from within the basal C shale reveals a lithogenic fraction composed mainly of quartz grains with minor chert. Quartz grains are angular to subangular in silt laminations and thin beds but are angular to subrounded in clay-rich intervals. Iron sulfide precipitates, sulfide-cemented sediment clusters, and pyritized wood fragments were commonly observed in coarse fractions.

8.

Linescan and X-ray CT images of fine-grained bounding interval structures. (a,b) Silty clay with silt laminations and wavy bedding and sulfide banding from core section 2P-1 in the basal E shale. (c,d) Clay with wavy, iron sulfide-rich silt laminations and pebble lag interval from core section 12P-1 in the basal C shale. (e,f) Silty clay and clayey silt, with wavy and lenticular silt beds, from core section 11P-1 in the lower bounding mud beneath the D1 sand. There are numerous sulfide-filled burrows within this section. (g,h) Silty clay with silt laminations from core section 24P-2 in the lower bounding mud beneath the B1 sand. Wavy and lenticular bedding and iron sulfide-filled burrows and silt laminations.

9.

Photomicrograph images of bounding sediments. (a) Rounded pyrite sand grains from the basal E shale. (b) Pyrite-cemented clusters and chert from the coarse fraction (250 μm to 2 mm) from the mud below the D1 sand. (c) Rounded chert pebbles and granules from the mud below the D1 sand. (d) Pyrite-cemented clusters and wood fragment from the mud below the B1 sand. (e) Thin section from the basal E shale with thin laminations of mud and silt offset by a fault. (f) Thin section from the basal C shale with clay-filled burrows within a silt.

The seal sediments overlying the D1 sand (basal E shale, 722.07 to 731.52 m MD) are primarily silty clay to clayey silt (Figures a,b and S1–S3). Within this fine-grained interval, there is a 25 cm-thick bed containing fine-to-medium sand composed primarily of rounded iron sulfide grains between 725.17 and 725.42 m MD (Figure a). The dark sulfide sand is more massive near the center of this 25 cm interval and is interbedded with clayey silt at the 1 to 3 cm scale in its upper and lower transition. Clay- and silt-dominated sediments from the basal E shale commonly contain iron sulfide precipitates.

The lower transition of the B1 sand is characterized by a shift from very fine sand to silty very fine sand at 895.33 to 897.18 m MD (Figures and S19–S24). Between 897.18 and 903.37 m MD clayey silt and silt are interbedded generally on the scale of tens of cm. Some clayey silt intervals contain millimeter-scale silt laminations in sets up to 7 cm thick. Silt occasionally occurs in the lenses in this interval. Sulfide patches and mottling are common throughout this interval. Below 903.37 m MD, the sediments are primarily silty clay and clayey silt with common silt laminations and lenses, including several 4 to 8 cm silt beds. Wavy and lenticular bedding is common in this interval (Figure g,h). Sulfide precipitation is more intense than above, with sulfide patches and mottling as well as dark sulfide bands. Iron sulfide precipitates, sulfide-cemented sediment clusters, and pyritized wood fragments were commonly observed in sieved coarse fractions (Figure d).

Similar to the B1 sand, the lower transition of the D1 sand is characterized by a shift from very fine sand to silty very fine sand to silt, between 739.26 and 747.11 m MD. Silty clay, clayey silt, and clay are the dominant lithologies between 747.11 and 755.42 m MD (to the bottom of the interval) (Figures and S7–S11). Silt beds between 1 and 14 cm thick are present in this fine-grained interval. Elongate, dark, and well-rounded chert granules and pebbles are present as a lag deposit at 749.66 to 750.59 m MD (Figure c). Bioturbation intensity is high between 750.71 and 752.16 m MD with common sulfide-filled burrows. Bioturbation is less common between 752.17 and 755.42 m MD, and silt laminations are more common. Iron sulfide precipitates, sulfide-cemented sediment clusters, and pyritized wood fragments were commonly observed in coarse fractions (Figure b).

5.2. Bulk Elemental and Isotopic Composition

Across the B1 sand and bounding sediments, TOC ranges from 0.25 to 4.27 wt % (mean: 1.09 wt %), and nearly all samples are below 2.5 wt %. TOC, along with TC, is generally higher in the bounding intervals and decreases in the B1 sand and lower transition, except for the upper part of the B1 sand between 886.63 m and 892.22 MD where TOC and TC remain relatively high, 0.84 to 1.54 wt %. TN generally increases with TOC and TC, except for this interval of the upper B1 sand, which also includes an anomalous increase in TOC/TN (44.5 to 129.7) and decrease in δ¹³C-TOC (−29.6 to −29.1‰ VPDB; mean: −29.5‰ VPDB). For comparison, the range of TOC/TN and δ¹³C-TOC range for the remainder of the cored B unit interval is 3.7 to 36.1 (mean: 16.9) and −29.1 to −27.3‰ VPDB (mean: −27.3‰ VPDB) for TOC/TN and δ¹³C-TOC, respectively. IC is relatively low across the entire interval (mean: 0.46; maximum: 2.64), indicating only minor carbonate content.

TS is variable across the B1 sand and bounding intervals, ranging from 0.05 to 3.0 wt % (mean: 0.56 wt %). TS is slightly elevated in the basal C shale (mean: 0.30 wt %) with sharp increases at 876.4 and 867.99 m MD. TS is very low within the upper B1 sand (mean: 0.08 wt %; maximum: 0.17 wt %) between 887.83 and 890.83 m MD. TS then steadily increases with depth from 0.29 wt % at 892.22 m MD to 1.71 wt % and 898.20 m MD. TS is high between 898.46 and 905.00 m MD (mean: 1.40 wt %) with multiple spikes to 2–3 wt %. TS is lower (mean: 0.15 wt %; maximum: 0.46 wt %) between 905.24 and 912.14 m MD. TS is elevated again (mean: 0.93 wt %) at the bottom of the interval from 912.37 to 915.52 m MD. δ³⁴S-TS ranges from −21.2 to +35.4‰ VCDT (mean: 1.5‰ VCDT). Sulfur from XRF core scanning is highly variable at the centimeter scale within intervals of elevated TS. There are multiple positive δ³⁴S-TS excursions, several of which correspond to intervals with high TS.

Across the D1 sand and bounding sediments, TOC ranges from 0.21 to 5.5 wt % with a mean of 1.2 wt % (Figure ). Except for the highest sample, the TOC in most samples is less than 2.5 wt %. TC, TN, and TOC are generally higher in the bounding intervals than in the D1 sand, except for spikes at 737.51–738.98 and 747.46 m MD. TC, and especially TN, drops significantly between the basal E shale and the upper D1 sand. Within the anomalous increases in TOC, TOC/TN increases to 52 (Figure ). δ¹³C-TOC ranges from −25.7 to −30.4‰ VPDB (mean: 27.8‰ VPDB).

11.

Carbon, nitrogen, and sulfur elemental and isotopic composition of the D1 sand and E unit basal shale. (a) Total nitrogen (TN) (cyan circles) and total carbon (TC) (purple triangles). (b) Total organic carbon (TOC) (green diamonds) and inorganic carbon (IC) (orange squares). (c) δ¹³C of TOC as ‰ relative to Vienna Peedee belemnite (VPDB) (dark blue diamonds) and the atomic ratio of TOC to TN (magenta triangles). (d) Total sulfur (TS) from CHNS elemental analysis (red circles), sulfur from X-ray fluorescence (XRF) (uncalibrated ppm) (orange lines), and pyrite abundance from X-ray diffraction (XRD). XRD measurement of pyrite was done at two different laboratories (cyan upside-down triangles and green squares). (e) δ³⁴S-TS in ‰ relative to the Vienna Canyon Diablo Troilite (VCDT). (f) Lithology interpreted from visual core description, microscopic description, and core scans. The top of the D1 sand is noted by a dashed line.

Within the pyrite sand interval in basal E shale, TS ranges from 19 to 41 wt %. In the rest of the D1 sand, TS ranges from 0.08 to 5.4 wt % with a mean of 1.67 wt %. TS is generally lower in the D1 sand (1.15 to 1.87 wt %, mean 1.51 wt %) compared to the basal E shale (1.81 to 3.62 wt %, not including the pyrite sand) and the lower transition and bounding mud (1.55 to 5.36 wt %, mean: 1.55 wt %). δ³⁴S-TS ranges from −39.3 to 25.1‰ VCDT with a mean of −11.0‰ VCDT.

Across 21 replicate samples, reproducibility of TC, TN, and TS (as 2 standard deviations) was 0.18, 0.01, and 0.11 wt %, respectively. Two shale standards, Boquillas Shale (Sh-BOQ-1) and Cody Shale (SCo-2), were run 8 times across the measurement sequence. Sh-BOQ-1 matches the reference values for TC and TS to within 0.7 and 0.02 wt %, respectively, with a TC, TN, and TS reproducibility (2 standard deviations) of 0.19, 0.03, and 0.10 wt %. Reproducibility (2 standard deviations) of TC, TN, and TS in SCo-2 was 0.05, 0.03, and 0.08 wt %, respectively.

Si measured by XRF increases within the B1 and D1 sands, reaching ∼50 wt % SiO₂ from a background of 40 to 45% in the bounding muds (Figures S24 and S25). Si/Al tracks quartz abundance by normalizing to aluminosilicate Si, matching trends in quartz from XRD, median grain size, and sand abundance (Figures and ). Zr/Rb also follows a similar trend, tracking the heavy mineral components. Al₂O₃, K₂O, and MgO are at their highest concentrations in the bounding muds (∼9, 1.5, and 0.8 wt % respectively) and decrease to about half that concentration in the B1 and D1 sands (Figures , , S24, and S25). This pattern in Al, Mg, and K closely follows the clay mineral abundance in both cored intervals. Fe is generally at a higher concentration in the bounding muds (∼3 to 5 wt %) compared to the sands (1 to 3 wt %) (Figures S24 and S25). Fe is 9 to 18 wt % in the pyrite sands in the E unit basal shale. Fe generally matches the S pattern in many intervals due to the abundance of iron sulfides driving the content of these elements. Ca is generally low in abundance (mean: 0.6 wt %) but does occur in spikes up to 3.8 wt %, likely associated with the sparse authigenic carbonate precipitation. Cross plots comparing the XRF core scanning and discrete measurements (averages of the 2 mm core scan data over the 2 cm sampled for geochemical characterization) show generally good agreement between the calibrated discrete measurements and the uncalibrated scans (Figure S27), with some systematic differences in Si, Mg, and Ti.

5.3. Rock Magnetism

Magnetic susceptibility, both κ from core scanning and χ from discrete samples, increases as a series of anomalous spikes in the basal C shale and in the bounding mud beneath the B1 sand. κ and χ are variably elevated in silts and muds of the basal C shale between 876.21 and 885.39 m MD (Figure a). There is an overall good correlation between κ and χ across the two cored intervals (Figure S27). Magnetic susceptibility is low throughout the B1 sand and much of the lower transition. In the underlying bounding mud, spikes in κ and χ occur between 902.56 m MD and the bottom of the cored interval at 915.36 m MD. From high resolution κ scanning, many of these spikes in κ are only several centimeters thick and do not correspond to a change in lithology. M _s and M _r, along with ARM, increase with χ in the bounding muds (Figure b). The ratio of M _r to χ also increases in the bounding sediments (Figure b), surpassing a value of 30 kA/m that is indicative of a magnetic mineral assemblage dominated by magnetic iron sulfides like greigite within the bounding intervals. Lower M _r/χ within the B1 sand and lower transitions indicate that the magnetic mineral assemblage is primarily magnetite. An elevated D _JH ratio in the bounding sediments , also indicates that magnetic iron sulfides are the dominant magnetic mineral. M _r/χ is low in sediments with χ less than 20 × 10^–8 m³/kg but >30 kA/m in high χ intervals (Figure S27g). S ₃₀₀ is nearly 1.00 for all samples, indicating the magnetic mineral assemblage is dominantly low-coercivity, consistent with both greigite and magnetite. There is a very slight decrease to 0.98 in S ₃₀₀ along the lower transition of the B1 sand to 900 m MD. χ_fd is <5% in the high χ sediments and increases to 5–10% in the B1 sand and the lower χ interval around 910 m MD. Overall, χ_fd is >5% only in sediments with χ < 60 × 10^–8 m³/kg (Figure S27e).

12.

Rock magnetic properties of the B1 sand and basal C shale. (a) Volume-normalized magnetic susceptibility (κ) (dark blue lines) and mass-normalized low-frequency (LF) magnetic susceptibility (χ) (green circles). (b) Ratio of saturation remanence (M _r) to χ (magenta diamonds). Vertical dashed lines represent ranges for magnetic iron sulfides (mis), magnetite (mag), and mixed mineral assemblages (mix). Anhysteretic remanent magnetization (ARM) (tan squares). (c) Ratio of hysteresis parameters (D _JH). The fraction of κ loss between 270 and 350 °C during stepwise heating experiments. (d) The ratio of backfield isothermal remanent magnetization (IRM) (−0.3 T) to saturation IRM (1.2 T) (S ₃₀₀). The hard (high coercivity) component of IRM. (e) Frequency dependence of magnetic susceptibility (χ_fd). (f) Lithology interpreted from visual core description, microscopic description, and core scans. The top of the B1 sand is noted by a dashed line.

High temperature measurements confirm the magnetic mineralogy based on the room temperature properties. In many samples from the bounding sediments, χ decreases dramatically on warming between 275 and 350 °C, consistent with the unblocking temperature of greigite or pyrrhotite (Figure ). Remaining χ decreases to zero by 580 °C, consistent with the Curie temperature of magnetite. In the lower χ reservoir sands, χ decreases to 580 °C on heating, with no indication of magnetic iron sulfides. Plotting the magnitude of decrease (% loss) in χ between 275 and 350 °C with depth shows a pattern consistent with D _JH and M _r/χ. This loss is consistently >20% in sediments with χ greater than 40 × 10⁸ m³/kg (Figure S27h). Low temperature measurements from MPMS are also consistent with the presence of greigite and magnetite. In both sands, the Verwey phase transition at 120 K is distinct in low temperature remanence curves (Figure ). A very slight Verwey transition is observable in samples from the high χ bounding sediments, but otherwise, the curves show relatively continuous change with temperature, consistent with the lack of major low-temperature phase changes for magnetic iron sulfides. Greigite has no low temperature transitions between 20 and 300 K.

13.

High temperature magnetic susceptibility (χ) curves for the basal C shale (a,b), the B1 sand (c,d), and the lower bounding mud beneath the B1 sand (e,f). The upper panel shows the warming (red) and cooling (blue) trends in χ. The lower panel shows the change in χ per °C (dχ/dT). Vertical dashed lines show the unblocking or Curie temperature for magnetic iron sulfides (mis), magnetite (mag), and hematite (hem). The fine-grained bounding muds (a,b,e,f) show a distinct drop in χ at around 350 °C associated with magnetic iron sulfides, while the B1 sand shows a decrease to 580 °C associated with magnetite (c,d).

14.

Low temperature magnetic susceptibility results. These results show the change in field-cooled (FC) remanence (red), zero field-cooled (ZFC) remanence (blue), room temperature (RT) remanence during cooling (gray), and RT remanence during warming (green). Examples from core section 5P-3 in the D1 sand (a,b), core section 17P-6 in the B1 sand (c,d), core section 11P-2 in the lower bounding mud of the D1 sand (e,f), and core section 24P-2 in the lower bounding mud of the B1 sand (g,h). Samples from the D1 (a,b) and B1 (c,d) sands show a clear change in remanence at the Verwey transition (VT) in magnetite at about 120 K.

In intervals of high χ within bounding sediments, acquisition of ARM followed by AF demagnetization reveals a slight gyroremanent magnetization (GRM), in which the ARM intensity (J _ARM) shows an increase in the higher alternating fields applied to the sample (120 to 170 mT) after an initial decrease (Figure a,c). GRM acquisition during AF demagnetization is a characteristic property of greigite. − Within the B1 and D1 sands, the AF demagnetization curves show a continuous decrease in J _ARM consistent with magnetite (Figure b).

15.

Example alternating field (AF) demagnetization curves after anhysteretic remanent magnetization (ARM) acquisition from the basal C shale (a), the B1 sand (b), and the fine-grained lower bound of the B1 sand (c). Each plot shows the change in ARM intensity (J _ARM) with increasing AF. Samples from the bounding muds (a,c) show a slight increase between 150 and 170 mT AF that suggests a slight gyroremanent magnetization (GRM).

κ and χ are low within the basal E shale through the D1 sand (Figure a). κ and χ are variably elevated between 749.29 and 754.72 m MD, with minor increases starting below 742 m MD. These increases in magnetic susceptibility are associated with an increase in the levels of M _s, M _r, and ARM (Figure b). Higher χ samples in the lower bounding mud also have an increase in M _r/χ consistent with magnetic iron sulfides, while the overlying sediments have M _r/χ consistent with magnetite (Figure b). The D _JH index and the amount of χ loss between 275 and 350 °C also increase with κ and χ below the D1 sand. S ₃₀₀ is consistently between 0.98 and 1.00 except a drop to 0.84 below the D1 sand. χ_fd mostly decreases with depth and has an abrupt decrease associated with the pyrite sand within the basal E shale.

16.

Rock magnetic properties of the D1 sand and basal E shale. (a) Volume-normalized magnetic susceptibility (κ) (dark blue lines) and mass-normalized low-frequency (LF) magnetic susceptibility (χ) (green circles). (b) Ratio of saturation remanence (M _r) to χ (magenta diamonds). Vertical dashed lines represent ranges for magnetic iron sulfides (mis), magnetite (mag), and mixed mineral assemblages (mix). Anhysteretic remanent magnetization (ARM) (tan squares). (c) Ratio of hysteresis parameters (D _JH). The fraction of κ loss between 270 and 350 °C during stepwise heating experiments. (d) The ratio of backfield isothermal remanent magnetization (IRM) (−0.3 T) to saturation IRM (1.2 T) (S ₃₀₀). The hard (high coercivity) component of IRM. (e) Frequency dependence of magnetic susceptibility (χ_fd). (f) Lithology interpreted from visual core description, microscopic description, and core scans. The top of the D1 sand is noted by a dashed line.

6. Discussion

6.1. Depositional Environment

The very fine to fine-grained sand lithology in the B1 and D1 sands (Figures and ) and (Figures and ) is consistent with deposition in a relatively high energy environment. The presence of cross-stratification (most apparent in X-ray CT scans of the B1 sand (Figure ) in sets consistent with hummocky or swaley cross-stratification) indicates deposition was most likely within the influence of the wave base in the neritic zone. These characteristics are consistent with a lower shoreface environment. In both the B1 and D1 sands, poor-to-moderate sorting, angular grains, and clay content of at least 2–5% classify these as immature to submature sands. , This suggests relatively rapid subaqueous transport from weathering and erosion of the Brooks Range to the coastal environment.

17.

Grain size distributions of seal sediments from (a) the basal C and E shales and (b) the B1 and D1 sands. Sand–silt (62.5 μm) and clay–silt (3.9 μm) grain size class boundaries shown as vertical dashed lines.

18.

Sediment classification based on sand–silt–clay composition. (a) Cored interval including the B1 sand and bounding units. (b) Cored interval, including the D1 sand and bounding units.

The shape of the grain size distribution (only available in Stratum particle size measurements) is consistent between the B1 and D1 sands with a major peak in the sand range and a skewed tail toward fine sediments with additional modes in the silt and clay-sized range (Figure ). The B1 sand has a more prominent sand peak with a higher median and maximum grain size (D ₅₀: 104 μm, D ₉₀: 170 μm) than the D1 sand (D ₅₀: 76 μm, D ₉₀: 140 μm). The D1 sand has greater clay-sized content and clay mineral content (average 5.8 and 10.8% respectively) than the B1 sand (2.4 and 8.0%, respectively). In the ternary sand–silt–clay space, both sands show a relatively consistent range in grain sizes from silty sand to clayey silt, with the D1 sand spanning from sand to silty sand (Figure ). Plots of the coarsest 1% size versus D ₅₀ (CM plot) suggest a trend in deposition from graded suspension to uniform suspension to pelagic suspension (Figure ). This trend is consistent with deposition of sands in a wave-dominated environment into deposition of finer sediments from quiet water deposition during shoaling of the wave base. The consistency of grain size distributions and trends, along with consistent detrital mineralogy across the two reservoirs, suggests a common sediment source for the two cored intervals but varying energy of the system as the shoreline and wave base evolve with sea level.

19.

Coarsest 1% versus median grain size for all cored intervals (CM) plot with interpreted transport dynamics.

The fine-grained clay-rich intervals surrounding the B1 and D1 sands, both the upper seal and lower bounding muds, exhibit abundant thin silt beds and laminations (Figure ). Often, these are present as wavy and lenticular bedding, suggesting varying depositional energy or oscillating flows, possibly associated with storm events, tidal currents, or surface wave oscillations that occasionally induce seafloor energy, resulting in symmetric ripples. Storm-driven runoff events can result in silt and fine sand deposition as wavy and lenticular bedforms on the continental shelf. Tides can generate tidal currents that influence sediment deposition on the continental shelf, beyond the areas within tidal range. Arctic sea ice can exert significant tidal dampening and the lack of evidence for perennial sea ice in the Arctic Ocean prior to ∼41 Ma suggests that tidal ranges higher than today were possible during deposition of units B through E, which could influence a larger area of the shelf. Thin intervals of pebbles and granules within clay and silty clay intervals in cores 10P and 12P (Figures S10 and S12) suggest local channel deposition or transgressive lag winnowing. Channels such as these on the shelf could focus tidal energy. The basal C and E shales represent the most landward extent of the transgressive flooding surface during sea level rise after the deposition of the B1 and D1 sands. This shallowest shelf deposition of the marine shales was likely above the storm wave base and more likely to experience tidal energy.

Overall, the lithostratigraphy of the sediments cored in the GDW represents progradational parasequences in which fine-grained marine sedimentation gives way to an upward coarsening sequence of shoreface sedimentation, followed by rapid sea-level rise producing a transgressive flooding surface and return to fine-grained sedimentation. These cycles are consistent with the interpretation of the Early Eocene C and D units at Mount Elbert as well as those interpreted throughout the Sagavanirktok Formation. Global sea levels varied by ∼15 to ∼40 m on orbital time scales during the Paleogene, driven by astronomical forcing that could drive these depositional cycles on 19 to 100 kyr time scales.

Based on thin section point counts and quartz–feldspar–lithic (QFL) plots, B1 and D1 sand samples classify as litharenites (38–42% quartz, 18–10% lithic fragments, 2–4% feldspar) (Figure ). Silt and silty clay samples from the seal sediments (basal E shale and basal C shale) have lower lithic fragment abundance relative to quartz (23–42% quartz, 3–10% lithic fragments, 2–5% feldspar), and the coarse fraction can be best described as sublitharenite in composition. The composition of the B1 and D1 sands is consistent with sandstones of recycled orogen material. , The dominant lithic material is chert, with only minor metamorphic or igneous fragments. K/Al ratios from XRF and the smectite illite ratio (Mx I–S/I) suggest higher rates of weathering and clay formation in the bounding muds (Figures and ). Chert may be less abundant relative to quartz during sea level high stands due to density driven sorting. We measured the density of chert fragments from core 10P, which ranged from 2.41 to 2.61 g cm^–3 (mean: 2.55 cm^–3). With increased weathering and transport, the denser quartz grains may be more separated from the lighter chert grains of similar size. In addition, chert likely starts out as larger fragments during weathering, while quartz grains can be liberated as sand and silt-sized grains directly from the weathering of sandstones. Slightly lower angularity in the quartz grains within the bounding intervals compared to the B1 and D1 sands suggests longer exposure of these grains to transport and weathering.

20.

Quartz–feldspar–lithic (QFL) plot from point counts for all cored intervals with the general classification sandstone classification.

The values of δ¹³C-TOC and TOC/TN are mainly consistent with a mixture of terrestrial and freshwater plankton , (Figure ). This is consistent with the nearshore environment interpreted for the sediments recovered from the B1 and D1 sands and the shelf environment for the bounding intervals. Pollen analysis from pelagic Paleogene Arctic Ocean sediments, recovered during Integrated Ocean Drilling Program (IODP) Expedition 302 at Lomonosov Ridge, suggests widespread subtropical coastal mangrove-like vegetation and upland ecosystems ranging from broadleaf to pine forests and significant export of terrestrial organic matter to the marine environment. − The nearshore environment of the ANS would have been even more dominated by terrestrial organic matter than the deepwater IODP site.

21.

δ¹³C of total organic carbon (TOC) and the atomic TOC to total nitrogen ratio (TOC/TN) for sediment samples, along with ranges of potential organic carbon sources. , The measured δ¹³C range of mineral oil-based mud used while drilling the HY02 well (−28.2 to −27.15‰ VPDB) is shown in the yellow band. The δ¹³C range of naturally occurring Barrow–Prudhoe type oils in Alaska North Slope reservoirs (−30.3 to −29.8‰ VPDB) is shown in the purple band.

Anomalous increases in TOC within the B1 sand (886 to 892 m MD) correspond to exceptionally high TOC/TN (Figure ) that are consistent with oil rather than particulate organic matter or plants with expected ratios <70 (Figure ). Measured δ¹³C-TOC in this interval ranges from −29.3 to −29.6‰ VPDB in this interval, which is distinctly different than the range in δ¹³C measured in the mineral oil-based mud samples used during drilling of the GDW (−28.4 to −27.5‰ VPDB). The δ¹³C-TOC from this portion of the B1 sand is similar in composition to Barrow–Prudhoe type oils (−30.3 to −29.8‰) (Figure ). While mineral oil-based muds are present in some core samples, this composition is consistent with the potential for migrated oil within the upper part of the B1 sand.

10.

Carbon, nitrogen, and sulfur elemental and isotopic composition of the B1 sand and C unit basal shale. (a) Total nitrogen (TN) (cyan circles) and total carbon (TC) (purple triangles). (b) Total organic carbon (TOC) (green diamonds) and inorganic carbon (IC) (orange squares). (c) δ¹³C of TOC as ‰ relative to Vienna Peedee belemnite (VPDB) (dark blue diamonds) and the atomic ratio of TOC to TN (magenta triangles). (d) Total sulfur (TS) from CHNS elemental analysis (red circles), sulfur from X-ray fluorescence (XRF) (uncalibrated ppm) (orange lines), and pyrite abundance from X-ray diffraction (XRD). XRD measurement of pyrite was done at two different laboratories (cyan upside-down triangles and green squares). (e) δ³⁴S-TS in ‰ relative to the Vienna Canyon Diablo Troilite (VCDT). (f) Lithology interpreted from visual core description, microscopic description, and core scans. The top of the B1 sand is noted by a dashed line.

6.2. Early Diagenetic Processes

Iron sulfide precipitation is variable across both of the sampled intervals, as shown by variation in total sulfur content, pyrite mineral abundance, and magnetic indicators of greigite. TS content is significantly higher overall in the D1 sand and bounding units (mean: 1.67 wt %, excluding the pyrite sand) compared to the B1 sand and bounding interval (mean: 0.56 wt %). The average pyrite abundance in the D1 sand and bounding sediments (3%, excluding the pyrite sand) is substantially higher than in the B1 sand and bounding units (mean: 1%) (Figure a,b). This comparison, along with the presence of the pyrite sand (>60% pyrite, 31% mean TS) in the basal C shale, indicates a stronger diagenetic influence of iron sulfide precipitation on these sediments. Based on the correspondence of TS with finer-grained bounding intervals, especially the increase of TS with decreasing grain size through the lower transition of the B1 sand (Figures and ), sulfide precipitation very likely occurred during shallow burial due to the reduction of marine sulfate.

22.

Total organic carbon (TOC) versus total sulfur (TS) plots. (a,b) The two different cored intervals. The 1:3 line of TS/TOC represents normal marine sediments experiencing sulfate reduction. The plot in (b) shows a zoomed-in scale on a more limited TS range due to the high TS within the pyrite sand. The D1 reservoir and basal E shale have generally higher TS than in B1 and in the basal C shale. The plots in (c,d) show the same TOC and TC data but binned by magnetic susceptibility (χ) ranges. Higher χ generally occurs in lower-to-moderate TS sediments.

Due to relatively high TS/TOC ratios, we suggest that most of the hydrogen sulfide produced in these sediments, ultimately sequestered as iron sulfides, was generated through anaerobic oxidation of methane (AoM) with relatively minor contribution from organoclastic sulfate reduction (OSR) (Figure ). Most ratios distinctly exceed the 1:3 ratio of TS/TOC expected due to OSR. The more intense sulfidization in the upper interval may be due to a higher sulfate availability in porewaters during early burial or lower sedimentation rates.

δ³⁴S-TS is more depleted in the D1 sand and bounding sediments (mean: −11.0 VCDT) compared to the B1 sand and bounding sediments (mean: 1.5‰ VCDT). Sulfur undergoes substantial fractionation during sulfate reduction with selective uptake of ³⁴S-depleted sulfate, leaving behind a pool of heavier ³⁴S-enriched sulfate. , This is often expressed as depletion with depth through the sulfate reduction zone and hydrogen sulfide reflects this trend of increasing δ³⁴S with depth to the sulfate methane transition zone (SMTZ). δ³⁴S of pyrite precipitated because of hydrogen sulfide generated from OSR or AoM reflects this increase in δ³⁴S with substate depletion. , The lighter δ³⁴S in the D1 sands and bounding sediments could be a result of higher sulfate availability due to variable freshwater input and changes in Arctic Ocean salinity during the Eocene.

The overall smaller grain sizes and increased clay content of the D1 sand and bounding intervals (Figures and ) also suggest that depositional energy and possibly sedimentation rates in the D1 sand were lower than during the deposition of the B1 sand. Lower sedimentation rates allow for sediments to spend more time within the sulfate reduction zone, allowing more time for sulfate diffusion into the sediments, hydrogen sulfide production via OSR/AoM, and iron sulfide precipitation. Association of higher TS with slower sedimentation rates and finer lithologies has been observed in offshore sediments with migrating SMTZs. ,

Greigite forms as a precursor for pyrite within the sulfate reduction zone as hydrogen sulfide reacts with reactive iron. , Although greigite has previously been considered a metastable mineral, recent work indicates that it is stable over geologic time once no longer within the presence of hydrogen sulfide. − Greigite can be a paleoenvironmental indicator of sulfur limitation at marine–terrestrial transitions due to salinity variation, − form under nonsteady state diagenesis in marine sediments due to variable sedimentation rates, − or associated with gas hydrate accumulations in marine sediments. ,,− At the GDW, we see a clear association of greigite and other pyrite formation with the depositional environment, indicating shallow formation within the sulfate reduction zone during early burial. Greigite occurrence is not associated with the gas hydrate-bearing sands. Large increases in κ only occur within the lower and upper bounding muds of the B1 sand and the lower bounding muds of the D1 sand (Figures and ). Direct correlation of κ and high-resolution S and Fe from XRF in these intervals confirms an iron sulfide driver of variation in κ (Figure ). The lack of correlation between κ and Zr/Rb that would otherwise be expected if κ variation was driven by detrital ferrimagnetic minerals like magnetite , also indicates that κ is of a diagenetic origin. Quartz (Si/Al) and heavy mineral (Zr/Rb) proxies associated with silt beds in the bounding muds generally increase in intervals between and adjacent to the spikes in κ, Fe, and S (Figure ). This suggests that greigite precipitation occurred in finer-grained sediments that were likely deposited more slowly, and the intermittent high-energy deposition can drive variable sedimentation rates and nonsteady state diagenesis.

23.

Detailed profile of the sediments underlying the B1 sand (Cores 20P to 24P, 900 to 916 m MD) with key rock magnetic, elemental, and sulfur isotopic measurements. (a) Volume-normalized magnetic susceptibility (κ) (dark blue lines) and mass-normalized low-frequency (LF) magnetic susceptibility (χ) (green circles). (b) Total sulfur (TS) from CHNS elemental analysis (red circles), sulfur from X-ray fluorescence (XRF) core scanning (orange lines), and pyrite from X-ray diffraction (XRD). (c) Fe/Al from XRF core scanning (green lines) and discrete XRF measurements (purple triangles). (d) Si/Al from XRF core scanning (black lines) and discrete XRF measurements (yellow squares). (e) Zr/Rb from XRF core scanning (brown lines) and discrete XRF measurements (upside down cyan triangles). (f) δ³⁴S-TS in ‰ relative to the Vienna Canyon Diablo Troilite (VCDT). (g) Frequency dependence of magnetic susceptibility (χ_fd). (e through l) A higher resolution presentation of the data shown in panels a through e between 902 and 904.5 m MD. (m through q) A higher resolution presentation of the data shown in panels a through e between 912.5 and 914.5 m MD.

Intervals with high κ from greigite in the bounding muds of the B1 and D1 sands are more likely to have been precipitated from beneath the continental shelf, subject to variable sedimentation rates and variable salinity waters with the potential for brackish pore waters. Greigite occurring between 902 and 905 m MD and 912 and 915 m MD corresponds to enriched δ³⁴S-TS, suggesting sulfur limitation during sulfate reduction. Greigite present between 908 and 912 m MD corresponds to more depleted δ³⁴S-TS, suggesting higher availability of sulfur for microbial processes but enough limitation of sulfide production to promote greigite formation over pyrite. This S limitation resulting in greigite formation is observed within the fine-grained, deeper water marine sediments, possibly as the result of brackish Arctic Ocean salinity and high rates of precipitation in the Arctic during the Paleogene ,, due to an enhanced hydrologic cycle amid global warmhouse/hothouse conditions. − The precipitation of greigite in bounding muds, although deposited in deeper water and more marine conditions than the B1 and D1 sands, suggests the presence of lower salinities on the shelf during sea level high stands, likely linked to higher precipitation during warmer conditions. ,,

Also in the North American Arctic, greigite has been observed and detected via rock magnetic techniques in gas hydrate-bearing intervals of the Mackenzie Delta (Mallik 5L-38 well). At the Mallik location, greigite was associated with sands and interpreted to have formed due to gas hydrate-related processes. Greigite of both early diagenetic and deeper burial origins has been observed in Cretaceous rocks of the Brookian sequence in the ANS.

The pyrite sand in the basal E shale (core section HY02-02P-4) likely represents a period of likely anoxic conditions on the continental shelf that allowed for the reworking/deposition of rounded pyrite grains. This resulted in deposition of a very dense layer (bulk γ density > 2.5 g cm^–3 and grain density > 3 g cm^–3) that is very distinct in X-ray CT (Figures a and S2). This layer also shows anomalies in trace elements measured by XRF that associate with pyrite (As, Cu, and Mo) , as well as trace elements associated with anoxic conditions (U and Mo) , (Figure d,e). These trace elements correspond to the Fe and S pattern that represents pyrite abundance at high resolution (Figure b). The pyrite sand is associated with a negative δ³⁴S-TS signature that suggests a high sulfate availability during sulfate reduction (Figure c). Anoxic conditions were present in the bottom waters of the Arctic Ocean during the Eocene, resulting in deepwater black shale deposition. − This pyrite sand suggests that anoxic conditions were present at least periodically on the continental shelf. Rounded pyrite grains similar to those in the GDW were observed in Eocene sediments of the lower C unit in the Mount Elbert well ∼20 km to the northwest, suggesting these shallow anoxic conditions were locally extensive on the ANS.

24.

Pyrite sand in HY02-2P-4. (a) High-resolution γ density scans from the Pressure Core Analysis and Transfer System (PCATS) and grain density (ρ_g). (b) Iron and sulfur from X-ray fluorescence (XRF) core scanning (gray line Fe and blue line S) and discrete measurements (gray diamonds Fe). Total sulfur (TS) from CHNS elemental analysis (red circles). (c) δ³⁴S-TS in ‰ relative to the Vienna Canyon Diablo Troilite (VCDT) (green diamonds). (d) Molybdenum and arsenic from XRF core scanning (orange line Mo and brown line As) and discrete measurements (red squares Mo and upside-down black triangles As). (e) Uranium and copper from XRF core scanning (green line Cu and dark red line U) and discrete measurements (dark red triangles Cu). (f) Si/Al and Zr/Rb ratios from XRF core scanning (blue line Si/Al and purple line Zr/Rb) and with sand percent from grain size measurements (yellow circles). (g) X-ray computed tomography slab image of the pressurized whole round core from PCATS. (h) Linescan image of the split archived half.

This pyrite sand corresponds to a distinct negative anomaly in the downhole resistivity log in the GDW reaching <1 Ω m, compared to a background of 10 to 20 Ω m in nonhydrate-bearing sediments at the GDW. This negative resistivity anomaly is also observed in both Production Test Wells (PTW-1 and PTW-2) at the Kuparuk State 7-11-12 Site. The high abundance (>60% from XRD) of semiconducting pyrite (typical resistivity <1 Ω m) in this interval explains the overall low resistivity in this interval. Low resistivity anomalies could be used to find other potential pyrite-rich intervals on the ANS, if such deposits were recurring during the Paleogene.

6.3. Tracking Drilling Fluid Contamination

Ba measured by XRF is a potential tracer of the relative amount of drilling fluid contamination in sediment cores from the GDW. Barite (BaSO₄) was used during drilling operations to increase the density of mineral oil-based drilling mud used in the main hole due to its very high grain density (4.48 g cm^–3). Ba concentration is generally low (<1000 ppm) in a global distribution of continental crustal rocks and shale standards, e.g., − so any measured Ba concentrations greater than 1000 ppm (0.1 wt %) in sediments from GDW cores are a likely indicator of drilling fluid contamination in the recovered cores. Sediment Ba concentration from XRF ranges from 0.05 to 2.4 wt % with a mean of 0.24 wt % across all 117 core samples measured (Figure ). 88 of these samples (75%) have Ba content greater than 0.1 wt %, indicating a high likelihood of some level of drilling fluid contamination. Ba content is highest in core sections 6.1 and 06P-2 (1.8 to 2.4 wt %) associated with clearly observed disturbance of the core with mixing of sediments and drilling mud. The liner of core 06P cracked during recovery and became stuck in PCATS, requiring the core to be depressurized rapidly within PCATS. This depressurization resulted in a high level of disturbance to sediments with high hydrate saturation in core sections 6.1 and 06P-2, resulting in a disaggregated mixture of sediment and drilling fluid. This contaminated interval has elevated Ca and Sr corresponding to the spike in Ba and a decrease in lithogenic elements including Si, Al, Mg, K, and Ti (Figure S25) due to dilution by the drilling fluid.

25.

Barium measured by X-ray fluorescence (XRF) core scanning (purple lines) and discrete XRF measurements (black circles) for the B1 sand and unit C basal shale (a) and the D1 sand and unit E basal shale (c). The vertical red dashed lines represent 0.1 wt % Ba, which is an upper end for ranges of average continental crust and for shale. − (b,d) Total sulfur (blue triangles) and predicted barite-derived sulfur from drilling fluid contamination.

Overall, there is little association between Ba content and lithology or degassing style, except for in core 06P. Elevated Ba is observed in some coarse-grained intervals of the B1 sand (core sections 16P-1a,b,c, and 17P-5, ranging from 0.15 to 0.53 wt %) but not others [core sections 16P-2, 17P-1, 17P-4, and 17P-6 (0.05 to 0.1 wt %)]. Moderately elevated Ba in core section 16P-1 is present at 887 m MD within only the upper part of the high TOC, potentially oil-bearing interval of the B1 sand (886 to 892 m MD). Drilling fluid contamination is not a clear explanation for the high TOC in this sand, unlike in core 06P from the D1 sand. Within the fine-grained bounding intervals of the B1 and D1 sands, Ba is periodically elevated in thin millimeter- to centimeter-scale spikes, associated with drilling fluid that is present in fractures parallel to bedding.

Barite in the drilling mud is also a potential source of sulfur contamination in the cores. We used the Ba content of the core samples and the stoichiometric ratio of Ba to S in barite to estimate the amount of barite-derived S present in these samples. These calculations suggest that GDW core samples could contain between 0.01 and 0.56 wt % S from barite. Estimated barite-sourced S content is between 0.41 and 0.56 wt % in the heavily disturbed 06P-1 and 06P-2 core sections, but ranges between 0.01 and 0.17 wt % (mean: 0.04 wt %) in samples from all other core sections. Because TS is relatively high in these cores (up to 41.1 wt %) with a mean and median of 2.34 and 0.95 wt %, respectively, the impact of barite S is somewhat limited, and the trends in TS and δ³⁴S-TS are representative of natural sulfur phases in the formation. Intervals with lower TS abundance and Ba >0.1 wt %, like between 878 and 882 m MD in the unit C basal shale, between 908 and 911 m MD at the lower bound of the B1 sand, and between 754 and 755 in the lower bound of the D1 sand, may have a more significant impact on the bulk sulfur content.

7. Conclusions

We provide a fundamental, integrated geologic, solid-phase geochemical, and rock magnetic characterization of the depositional and postdepositional processes that shaped the state of the D1 and B1 sands and bounding sediments. These properties influence the potential production characteristics as well as the petroleum system that promotes gas hydrate accumulation in these reservoirs. Our results confirm a sand lithology in the B1 and D1 sands that is consistent with the higher porosity and permeability of the hydrate reservoir. , Our results also confirm the fine-grained, interbedded lithologies in the basal C and E shales that are consistent with the low permeabilities − needed to isolate the reservoir during production. Description of these cores at a centimeter-scale resolution with quantitative sedimentological, geochemical, and rock magnetic measurements at a tens of centimeters resolution, along with millimeter-scale core scanning, provides a detailed characterization of these hydrate-bearing sands and their bounding sediments that can enhance the constraints on geologic and production models of the Kuparuk State 7-11-12 location.

We observe transitions from fine-grained clay-rich sediments in the bounding muds to quartz-rich, chert-bearing immature, poorly to moderately sorted sands in both the B1 and D1 sands. The lithological variation we observe here is driven by a continuum of depositional energy from wave-influenced shore face environments in the B1 and D1 sands to lower energy marine shelf environments, regulated by sea level variation. The larger grain sizes of these sands create the higher porosity needed to host high saturations of gas hydrate. , Although the silty very fine sand to sandy silt size of B1 and D1 sands are on the fine end of the spectrum for sand reservoirs, this composition provides sufficient permeability to host high saturation hydrate accumulations. ,, The larger median grain size, higher sand content, and lower clay abundance of the B1 sand compared to the D1 sand likely explain the higher hydrate saturation (>80% of the pore space) in the B1 sand. The basal E shale (seal of the D1 sand) is finer with less silt content compared to the basal C shale (seal of the B1 sand). Despite numerous silt laminations and thin beds in the seal of the B1 sand, net permeability is still low across this interval. Although the seal of the B1 sand is relatively heterogeneous, the quality of the reservoir is higher, resulting in enough permeability contrast in the trap to promote high saturation of methane. Within the Eileen Gas Hydrate Trend, higher depositional energy parasequences are likely the best reservoirs for hydrate accumulation, even where there are coarser intervals within the overlying seal sediments.

We observe moderate organic carbon contents (>1 wt %) in bounding mud sediments that are typical for continental shelf settings. The primarily terrestrial source of this organic matter is consistent with the mangrove-like coastal ecosystems that were known to be present in the warm conditions of the Paleogene. The amount of organic carbon in these sediments was likely sufficient to drive organoclastic sulfate reduction, microbial methanogenesis, and anaerobic methane oxidation during shallow burial to drive hydrogen sulfide production and iron sulfide precipitation. High organic carbon within the upper part of the B1 sand is consistent with the composition of Barrow–Prudhoe type oils, rather than the composition of particulate organic carbon like that present in the bounding muds or the oil-based drilling mud.

We observe variable sulfur content, pyrite abundance, and magnetic iron sulfide occurrence that represent variable sulfate availability during early diagenesis. Sulfate limitation during early burial could be driven by local freshwater runoff or basin-wide freshening in the Eocene. Sulfide precipitation is an important influence on composite matrix density and grain density measurements, which were used to calculate porosity and hydrate saturation in GDW sediments. , Although iron sulfide precipitation is considered a less impactful diagenetic process on reservoir quality than carbonate or silica precipitation (not observed in GDW sediments), the high concentration of sulfur in these sediments raises the possibility that pyrite precipitation further reduces the pore throat size of these fine-grained sediments, influencing permeability. Sulfide precipitation within silt laminations of the seal of the B1 sand likely further reduces the available pore space and permeability. The pyrite sand in the basal E shale is a rare depositional event that likely represents anoxic conditions on the ANS continental shelf during the Early Eocene.

Pressure coring facilitated the recovery of high-quality cores from the B1 and D1 sands and their bounding units, which provide insight into the shoreface to continental shelf depositional environments and associated early diagenetic processes that shaped the sediments that now host subpermafrost gas hydrates. This characterization provides insight potentially applicable to other hydrate systems in sediments deposited in nearshore parasequences, in which sea level controls depositional energy, and ultimately the porosity and permeability contrast needed for hydrate-bearing reservoirs and effective seals.

Glossary

Abbreviations and Symbols

2D

two-dimensional

3D

three-dimensional

AF

alternating field

ANS

Alaska North Slope

ARM

anhysteretic remanent magnetization

B _c

coercivity

B _cr

coercivity of remanence

CT

computed tomography

D ₅₀

median grain size

D _JH

ratio of saturation and coercivity parameters

DOE

Department of Energy

FC

field-cooled

GDW

Geo Data Well

HIRM

hard isothermal remanent magnetization

HF

high-frequency

HPTC

high pressure temperature corer

IC

inorganic carbon

IRM

isothermal remanent magnetization

JOGMEC

Japan Organization for Metals and Energy Security

LF

low frequency

LTSIRM

low temperature saturation isothermal remanent magnetization

MD

measured depth

MPMS

magnetic property measurement system

M _s

saturation magnetization

M _r

saturation remanence

Mx I–S

mixed layer illite–smectite clays

PCATS

pressure core analysis and transfer system

R&D

research and development

RTSIRM

room temperature saturation isothermal remanent magnetization

S ₃₀₀

S-ratio at 300 mT

SIRM

saturation isothermal remanent magnetization

SQUID

superconducting quantum interference device

STW

stratigraphic test well

TOC

total organic carbon

TN

total nitrogen

TS

total sulfur

TVD

total vertical depth

USGS

U.S. Geological Survey

VCDT

Vienna Canyon Diablo Troilite

VPDB

Vienna Peedee belemnite

VSM

vibrating sample magnetometer

XRD

X-ray diffraction

XRF

X-ray fluorescence

κ

volume-normalized magnetic susceptibility

χ

mass-normalized magnetic susceptibility

χ_fd

frequency dependence of magnetic susceptibility

XRF, magnetic susceptibility, and γ density core scan data, along with core photos, are available at 10.5281/zenodo.16369768 in the Zenodo online database. Grain size, XRD, and thin section point counts/photos are available at 10.5281/zenodo.16416235 in the Zenodo online database. Sediment description, bulk elemental and isotopic, and rock magnetic data are available at https://doi.org/10.5281/zenodo.16421337 in the Zenodo online database.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.energyfuels.5c05174.

• A set of sediment barrel sheet summary figures for each core (1P to 24P) with major lithology, visual core descriptions, X-ray images, core section curation, linescan images, magnetic susceptibility, and Si/Al from XRF, and figures with additional rock magnetic property and XRF data and correlations (PDF)

The manuscript was written through contributions of all authors by generating data, contributing to the interpretation of data, and writing the manuscript. All authors have given approval to the final version of the manuscript.

This research was directly supported by the U.S. Department of Energy through DOE–USGS Interagency Agreement 89243320SFE000013.

The authors declare no competing financial interest.

Published as part of Energy & Fuels special issue “Recent Advances on Gas Hydrates Scientific Drilling in AlaskaAn Update from Production Testing Operations during 2023/2024 (Part I)”.