1 INTRODUCTION

Gas hydrates are ice-like compound composed of natural gases (mainly methane) and water that occur in the permafrost and marine sediments worldwide (Kvenvolden, 1993; Sloan and Koh, 2007). Natural gas hydrate has attracted more attention as a potential energy resource, seafloor instability, climate change and a series of drilling expeditions (e.g., Zhang et al., 2007; Bahk et al., 2011; Ruppel, 2011; Boswell et al., 2012; Collett et al., 2012; Kumar et al., 2016; Handwerger et al., 2017) and production tests were conducted in the permafrost and marine sediments by several countries (e.g., Dallimore et al., 2003; Makogon and Omelchenko, 2013; Suzuki et al., 2015; Yang et al., 2015, 2017; Li et al., 2016). The elastic and physical properties provide necessary support for analyzing and evaluating the gas hydrate-bearing reservoir, the security of production test and seafloor stability. Physical characteristics analyses of gas hydrate-bearing sediments (GHBSs) are mainly based on the rock physics models to translate the elastic properties to reservoir properties, which have been proved to be a convenient and intuitive tool to analyses the reservoir parameters (Avseth and Ødegaard, 2004; Avseth et al., 2006; Yamamoto, 2015; Haines et al., 2017; Ye et al., 2018).

Physical characteristics and depositional environments of GHBSs have been delineated with well logging, core, and seismic data in marine environments, including in the Gulf of Mexico (Andersen et al., 2009; Boswell et al., 2012; Collett et al., 2012; Haines et al., 2017), Nankai Trough (Fujii et al., 2015; Suzuki et al., 2015; Yamamoto, 2015; Yoneda et al., 2015), Ulleung Basin (Bahk et al., 2011; Kim et al., 2011), Black Ridge (Lu and McMechan, 2002; Bahk et al., 2011), Krishna-Godavari Basin (Kumar et al., 2016), and South China Sea (e.g., Chow et al., 2005; Zhang et al., 2007, 2018; Wang et al., 2014a, b; Zhong et al., 2017; Kang et al., 2018; Li et al., 2019). The geophysical parameters and characteristics of the GHBSs are critical for the fully understanding of the occurrence and behavior of gas hydrates and the production technologies (e.g., Bahk et al., 2011; Ruppel, 2011; Ito et al., 2015; Zhong et al., 2017; Nanda et al., 2019). In the Nankai production test region, the methane hydrate-bearing concentrated zone is composed of turbidite-channel sediments with a thickness over 60 m, the background P-wave velocity values ranging from 1 700 to 1 800 km/s, the average background resistivity value of 1.5 ohm-m and porosity ranging from 40% to 50% (Fujii et al., 2015; Suzuki et al., 2015). The bottom simulating reflector (BSR) is about 345 meters below seafloor (mbsf) at Site AT1 of Naikai trough, which is thicker than the BSR of Site SHSC-4, South China Sea (Suzuki et al., 2015; Guo et al., 2017; Ye et al., 2018). The sand-rich reservoirs with lateral continuity for the production test have been found in the Nankai trough, Krishna-Godavari Basin, and Gulf of Mexico (Boswell et al., 2012; Collett et al., 2012; Fujii et al., 2015; Kumar et al., 2016).

The sediment layers with high gas hydrate saturation and high reservoir thickness have been identified at two adjacent ridges by Guangzhou Marine Geological Survey (GMGS3 & 4) expedition from the logging-while-drilling (LWD) data and core samples in the Shenhu area, South China Sea (Yang et al., 2015, 2017). The morphologies of gas hydrates are pore-filling types that are invisible to the naked eye from core samples in the production test region (Yang et al., 2015). Previous studies mainly focus on the distribution of gas hydrate and saturation. The areas of Sites W18, W19, W11, and W17 in the Shenhu areas are the potential first offshore production test regions in the South China Sea (Yang et al., 2015; Guo et al., 2017; Qian et al., 2018; Li et al., 2019), while the gas hydrate production test sites are located at the ridge of W11 and W17. The differences of physical properties of sediments for gas hydrate reservoir, overlying and underlying strata of the two adjacent ridges are poorly known, which is an important problem in selecting production site. In this study, we focus on the geophysical and depositional characteristics of gas hydrate-bearing sediments at two areas by combining the core, logging data and three-dimensional (3D) seismic data to compare the differences of the two potential production zones.

2 GEOLOGICAL SETTING AND PREVIOUS STUDY

The first gas hydrate production test exploration was conducted at Site SHSC-4 (Li et al., 2018) in the Baiyun Sag, the Zhujiang (Pearl) River Mouth Basin (PRMB), South China Sea (Fig. 1a). Sites W18, W19, W11, and W17 are located at the adjacent ridges near the gas hydrate production test site, which were drilled and cored by GMGS3 expedition (Fig. 1b) (e.g., Yang et al., 2015, 2017; Wang et al., 2016; Guo et al., 2017; Zhang et al., 2017; Qian et al., 2018; Li et al., 2019). High concentrated gas hydrates are characterized by heterogeneous distribution along the four ridges of Baiyun submarine canyon system (Wang et al., 2014a). Pore-filling gas hydrates with variable saturations and thicknesses at different sites have been identified from cores, downhole well logs, and seismic data by three gas hydrate drilling expeditions (GMGS1, 3, & 4) in the Shenhu area, South China Sea (e.g., Wang et al., 2014a, b; Yang et al., 2017; Zhang et al., 2017; Qian et al., 2018; Jin et al., 2020). The distributions of gas hydrate have been delineated in previous studies from interstitial water-chlorinity, LWD and 3D seismic data at Sites W18, W19, W11, and W17 (e.g., Yang et al., 2015, 2017; Jin et al., 2017, 2020; Qian et al., 2018). The thickness and saturation of GHBSs at those four sites are summarized in Table 1. The base of methane hydrate stability zone (I-BGHSZ) has been calculated, and the depths of I-BGHSZ are 172, 171, 200, and 250 mbsf at Sites W18, W19, W11, and W17, respectively (Guo et al., 2017; Yang et al., 2017; Qian et al., 2018; Jin et al., 2020).

Seventeen migrating canyons with unidirectional migration stacking patterns are located at a water depth of 500–1 300 m of the Baiyun sag (Zhu et al., 2010; Zhou et al., 2015). Based on the characteristics of seismic reflections, a series of complex architecture elements in the submarine canyons are delineated indicating an evolution of repeated erosion, infilling, offset, and subsequent re-excavation (Zhou et al., 2015). A buried trough-like sediment feature is interpreted from the high-resolution 3D seismic data, and the architecture of hydrate-bearing reservoirs in the Shenhu area has been delineated at Sites W18 and W19 (Jin et al., 2017, 2020; Zhang et al., 2020). Each cycle of deposition in the canyon is marked by a basal erosional discontinuity (BED). The coarse-grained rich units (potential foraminifera-rich layers) are inclined to occur at the bottom of canyon deposits above the BED, such as the thalweg deposits. In addition, the slope sediments mainly consist of continuous reflections with truncated by BEDs, indicating the continuous depositional process (Chen et al., 2009; Zhou et al., 2015).

3 DATA AND METHOD 3.1 Well log and seismic data

The LWD data have been acquired by Guangzhou Marine Geological Survey (GMGS3) expedition and used for the characteristics of gas hydrate distributions in the Shenhu production test area (Yang et al., 2015, 2017; Guo et al., 2017; Zhang et al., 2017, 2018, 2020; Kang et al., 2018; Li et al., 2019). In this study, the LWD data with sample interval of 0.1524 m including natural gamma (GR), resistivity, density, density porosity, and P-wave velocities (V_p) at Sites W18, W19, W11, and W17 are systematically analyzed to compare the elastic properties, physical properties, and rock physics templates. The 3D seismic data are processed with bin spacing of 12.5 m and 25 m in the in-line and cross-line directions, respectively. The dominant frequency is about 50 Hz and the sampling interval is 1 ms.

The BED characterized by continuous unconformity reflections with the occurrence of truncations below and onlaps above are identified as the base of canyon and over-bank depositions (Fig. 2). Horizon H1 is shown by continuous reflections in canyon sediments and only occur above Sites W18 and W19. Both of BED and H1 are traced through the available 3D seismic data to connect the adjacent two ridges and to show the variations of depositional environment in the study area (Fig. 2). The horizons H1 and BED are interpreted from the 3D seismic data and are used to study the variations of physical properties at different sites from the cross plot templates of well log and the inverted P-wave velocity section. The gas hydrate occurrence with high amplitude reflections are also shown with different gas hydrate distribution at two adage ridges in the seismic profile (Fig. 2).

3.2 Cross plots for rock physics analysis

The cross plots of gamma-ray, acoustic impedance, and porosity log data from LWD are used to infer lithological variations in sedimentary section at Sites W18, W19, W11, and W17 (Fig. 3). The relationship between the measured P-wave velocity (V_p) and acoustic impedance (I) are plotted to obtain the fitting equation at Sites W11, W17, and W19 (Fig. 4). The 3^rd degree polynomial regression is obtained from the crossplot with the R-squared correlation coefficient of 0.975. The fitting equation is:

where I is the acoustic impedance calculated from V_p and density of well logging data. The porosity (Φ) and acoustic impedance (I) are also plotted to show the variations of trend for the water-saturated sediments and gas hydrate-bearing sediments (Fig. 5). The 3^rd degree polynomial regression is shown in Fig. 5, and the R-squared correlation coefficient is about 0.97. The fitting equation is:

The porosity profile can be calculated from the relationship between porosity and acoustic impedance using Eq.2.

The resistivity and P-wave velocity at the depths of 140–168, 152–170, 120–193, and 208–250 mbsf are obviously increased above the I-BGHSZ, indicating the occurrence of gas hydrates at Sites W19, W18, W11, and W17, respectively (Table 1) (e.g., Yang et al., 2015, 2017; Qian et al., 2018; Zhang et al., 2018; Jin et al., 2020). However, the thickness of gas hydrate reservoir is obviously different for each site. The thickness ranges from 18–28 m at Sites W18 and W19 to 42–73 m at Sites W11 and W17, respectively (Table 1). Comparing to the water-saturated P-wave velocity (V_pw) or resistivity, the velocity differences between V_p and V_pw show the increased anomalies of P-wave velocity (ΔV_p), which are related with gas hydrate saturations (Sh). At Sites W18 and W19, the curves of gas-hydrate saturation log show an increasing trend from I-BGHSZ (Table 1) and match the shapes of the gamma-ray log curves (Jin et al., 2020).

Archie equation is used to calculated the gas hydrate saturation, and the Archie empirical constants (a and m) are listed in Table 1, which have been obtained from crossplots between formation factor and density porosity log values. The crossplots between gas hydrate saturations from resistivity and measured V_p at these four sites show the scattered distribution (Fig. 6a–b), while the crossplot between gas hydrate saturations and ΔV_p shows better linear relationship (Fig. 6c–d). The linear fitting equation with R-squared value of 0.74 is:

where ΔV_p equals to (V_p–V_pw) with km/s. The fitting equation shows that the match is poor for the low and high gas hydrate saturations.

Grain size analyses at Site SH2 in GMGS1 and Site W18 in GMGS3 indicate that higher gas saturations are almost associated with the coarse-grained sediments, which consist of the abundance of foraminifer and calcite content (Chen et al., 2013; Kang et al., 2018; Kuang et al., 2019; Li et al., 2019; Wang et al., 2021). However, the GHBSs with high saturation are also occur in the fine-grained (almost silts-rich) sediments, such as at Sites W11 and W17 (Hu et al., 2017; Wang et al., 2018; Li et al., 2019). The analyses of core and grain-size distribution indicate the lithology of slope sediments are composed of coarse silt and clay silt within a similar and low-energy sedimentary environment (Li et al., 2019). The crossplot between gamma ray values and porosities (Φ) shows the negative correction with the R-squared value of 0.84 (Fig. 7a), and the fitting equation is:

where GR is obtained from well log data with the unit of api.

4 RESULT 4.1 Reservoir properties from well log

The crossplots of gamma-ray values versus acoustic impedance (P-impedance) and porosity log data reveal two distinct sediments units (upper layer and lower layer) from the color axis of depth and wells (Fig. 3), which is divided by horizon H1 at Sites W18, W19, W11, and W17 (Fig. 3a & c). The upper layer at Sites W18 and W19 is thicker than that of Sites W11 and W17 identified from the points numbers (several thousands of points, and almost red and black data points at Fig. 3b & d). In the lower layer, the GHBSs show different gamma-ray log values. The GHBSs detected from high P-impedance anomalies have high gamma-ray values at Sites W11 and W17 and low gamma-ray values at Sites W18 and W19 (Fig. 3b). At Sites W18 and W19, the GHBSs deviated from lower layer are characterized by low gamma-ray values ranging from 25 to 40 api and high P-impedance (> 2.5×10⁶ kg/(m²·s)) indicating the occurrence of GHBSs within coarser-grained sediments (Fig. 3b & d). In addition, at Sites W11 and W17, the GHBSs identified from high P-impedance (> 3.0×10⁶ kg/(m²·s)) occur at high gamma-ray log values almost ranging from 60 to 80 api indicating the fine-grained sediments (Fig. 3b & d). The cross plot of porosity log values versus gamma-ray log values also show the similar variations (Fig. 3c–d). The lower layers have high porosity of 0.6–0.7 and low gamma-ray values at Sites W18 and W19 (Fig. 3c & d; red and black points in Fig. 3d), while the low porosity of 0.45–0.6 and high gamma-ray values at Sites W11 and W17 (Fig. 3c–d; purple and blue data in the Fig. 3d). These anomalous zones are related to GHBSs at these four sites indicating the different lithology. Therefore, the upper and lower layers show different rock-physics models from the crossplots, and are obviously divided into two zones by horizon H1.

The cross plots of porosity and P-impedance are used to show the variations of GHBSs at Sites W18, W19, W11, and W17 with color axis of depth and wells (Fig. 5). The porosity decreases with the increasing depth and P-impedance seen from the rock physics templates of porosity and P-impedance values except for the points of GHBSs at Sites W18 and W19 (Fig. 5). The scatted points (red and black) at Sites W18 and W19 with high P-impedance (> 2.5×10⁶ kg/(m²·s)) and slightly high porosity (0.6–0.7) are obviously deviated from the background trend indicating the GHBSs (Fig. 5b). The cross plots of gas hydrate saturation versus P-wave velocity (V_p) and increased anomalies of P-wave velocity (ΔV_p) show that the GHBSs have higher V_p at Sites W18 and W19 (Fig. 6a–b). In addition, higher ΔV_p values correspond to higher gas hydrate saturations at Sites W18 and W19 (red and black points) than those at Sites W11 and W17 (Fig. 6c–d). However, the porosity is higher for low gamma ray with decreasing line relationship (Fig. 7a) in the GHBSs. Relationships between ΔV_p, gas hydrate saturation, and porosity are non-linear (Fig. 7b–c), which indicates that gas hydrate saturation cannot be directly estimated from the inverted porosity. The GHBSs have higher porosities and lower gamma-ray logs, which indicates that the reservoir properties at Sites W18 and W19 are different from those of Sites W11 and W17.

4.2 Physical properties from seismic data 4.2.1 The characteristics of GHBSs from seismic profile

The BSRs have high amplitude, cross-cutting strata, and reversed polarity compared to the seafloor reflection, which are identified from the seismic profile through Sites W18, W19, W11, and W17 (Fig. 2a). BSR and BED have nearly the approximate depths below seafloor at Sites W18 and W19, while they have different depths at Sites W11 and W17 (Table 1). Muti-stratified high-moderate amplitude reflections were observed at Sites W11 and W17 between the intervals of BSR and BED, which indicates larger thickness of GHBS in the slope sediments. One stratified high-amplitude reflection is identified above the BSR or BED at Sites W18 and W19, suggesting thinner GHBSs in the canyon deposits (Fig. 2a), which coincides to the gas hydrate layer identified from well log data (Table 1).

4.2.2 Inverted P-wave velocity

The P-wave velocity profile is inverted from seismic data using constrained sparse spike inversion (CSSI). There is an obvious boundary (horizon H1) between low velocity layer and high velocity layer (Fig. 8a). The V_pw is about 1.9 km/s (sediments shown green color between H1 and BSR) from the inverted P-wave velocity section. The GHBSs can be identified from the increasing anomalous P-wave velocity ranging from 2.0 to 2.6 km/s in the velocity profile at four sites (Fig. 8a). However, the velocities of GHBSs at Sites W18 and W19 (2.4–2.6 km/s) are higher than those at Sites W11 and W17 (2.0–2.4) (Fig. 8a). The P-wave velocity is also lower below the BSR with a value of 1.4–1.8 km/s, indicating the occurrence of free gas. The P-wave velocity near Site W11 below BSR has a value of 1.85 km/s, which is higher than that of Site W17 (about 1.4 km/s) indicating the presence of free gas.

4.2.3 Inverted porosity

The porosity profile is calculated from the fitting Eq.2. The anomalous reservoir properties are also observed from the seismic section reflections indicating the occurrence of gas hydrate (high amplitude reflections and positive polarity) above BSR and free gas (enhance reflections and negative polarity) trapped beneath BSR (Figs. 2 & 9). Commonly, low porosities are observed in the GHBSs compared to the surrounding sediments at four sites, and anomalously high porosities are found below the BSR. While, the inverted porosities of GHBSs at Sites W18 and W19 are obviously higher than those of Sites W11 and W17 (Fig. 8a).

4.2.4 Inverted gas hydrate saturation

The crossplot between gas hydrate saturation estimated from resistivities versus V_p shows the scattered distribution of samples. Gas hydrate saturation profile (Fig. 10a) is calculated from the inverted P-wave velocity profile using the sand/clay rock physics model (Wang et al., 2016). The mineralogy used in this study consists of 5% sand, 75% silt, 20% clay, which is average value at Sites W11 and W17. Gas hydrate saturation profile (Fig. 10b) can also be calculated from equation 3 using ΔV_p (Fig. 6c–d). The RMS and maximum gas hydrate saturation section shows that gas hydrate saturation at Site W19 (up to 60%) is higher than that at Sites W11, W17, and W18 (20%–40%) (Fig. 10c–d). Hydrate saturation inversions in coexisting layer of gas hydrate and free gas may be inaccurate due to the lack of free gas petrophysical models at Site W17. While Site W11 has the thickest gas hydrate-bearing layers with moderate saturation (Fig. 10a–b). The gas hydrate-bearing layers show more laterally continuous characterizations along slope sediments at Sites W11 and W17. In the plane, gas hydrate saturation at Site W19 (60%) is higher than that of at Sites W11, W17, and W18 (20%–40%). According to the statistical gas hydrate occurrence (gas hydrate saturation > 20%), the area of gas hydrate at Sites W11 and W17 is about ten times that of Sites W18 and W19, and gas hydrate cover 0.65 km² at Sites W18 and W19 and 6.94 km² at Sites W11 and W17, respectively (Table 1; Fig. 10c–d).

4.3 Depositional environment

The BED characterized by the obviously laterally continuous unconformity, truncation, and onlapping reflections in the seismic section is identified at the base of canyon deposits (Fig. 2). The BED is characterized by the upward decrease of gamma-ray log values and the upward increase of porosity log data (Fig. 11), which likely indicates a coarsening-upward succession overlain by shallow canyon deposits with low density and P-wave velocity (Figs. 2 & 12). The depths of the BED are about 166, 157, 55, and 24 mbsf at Sites W18, W19, W11, and W17, respectively (Table 1; Fig. 11). Near Sites W18 and W19, the canyon deposits above BED can be divided into two layers (e.g., the upper and lower layers) by horizon H1, which overlap to the BED at Sites W11 and W17 (Fig. 6a). The upper layer above horizon H1 at Sites W18 and W19 showed similar porosity and velocity values to those identified above the BED (upper layer occurring above BED) at Sites W11 and W17 from the cross-plots (Fig. 4) indicating the similar sediments characteristics. The lower layer characterized by low gamma-ray log values is mainly composed of the gas hydrate reservoir at Sites W18 and W19 (Figs. 2 & 11). The canyon deposits have relatively lower velocities and higher porosities at the same depth than the slope sediments (Figs. 5 & 8).

The cross-plots of P-wave velocity versus density for different depths at Sites W19 and W17 are used to separate the GHBSs unit from the non-GHBSs unit by anomalously high P-wave velocity (> 1 800 m/s) (Fig. 12a, red and green points). The P-wave velocity and density of the non-GHBSs show the polynomial relationship as the following:

where V_p is P-wave velocity with m/s and D is density. The R² value is about 0.83, indicating a well correlation factor between V_p and density. The canyon deposits characterized by low density and low velocity can be separated with the slope sediments by horizon BED from the non-GHBS unit (Fig. 12b). The horizon BED marks the boundary between the canyon deposits and slope sediments, which shows different relations between the density and P-wave velocity log data at horizon BED.

Significant variations of thickness above the BED indicate different depositional environments of this surface between two adjacent ridges. The canyon deposits characterized by the paralleled, horizontal and weak-moderate amplitude are overlapped above the BED (Fig. 2). At the Site W18 and W19 area, a buried trough is imaged by the interpreted BED (Fig. 8a) (Jin et al., 2017, 2020), indicating the trough is primarily erosional features. The low gamma-ray units are identified between the horizon BED and H1 at Sites W18 and W19 (Figs. 2, 9, & 11). In contrast, the normal deposition sequence layers shown by the paralleled, inclined and weak-moderate amplitude reflections are truncated by BED at Sites W11 and W17 (Figs. 7a & 9b). The uniform gamma ray and porosity log values likely indicate uniform grain sizes in the slope sediments (Fig. 11). The largest thickness of canyon deposits exceeds 150 m at Sites W18 and W19, while the largest thickness of slope sediments develops over 100 m at Sites W11 and W17 (Table 1; Fig. 2).

5 DISCUSSION 5.1 Reservoir physical properties for gas hydrate distribution

The gas hydrate samples recovered from the Shenhu area have the morphologies of primarily pore-filling gas hydrate within unconsolidated sediments (Zhang et al., 2007; Riedel et al., 2011; Yang et al., 2015, 2017). Hydrates with high saturation in the Nankai Trough offshore, Japan (Andersen et al., 2009; Nanda et al., 2019), Krishna-Godavari Basin, India (Ruppel et al., 2008; Andersen et al., 2009; Kumar et al., 2016; Hillman et al., 2017; Li et al., 2018), Alaminos Canyon, Walker Ridge and Green Canyon, Gulf of Mexico (Ruppel et al., 2008; Boswell et al., 2012; Collett et al., 2012; Frye et al., 2012; Hillman et al., 2017), have been found in the relatively high porosity, low background resistivity and low gamma-ray log reservoirs (generally coarse-grained sand-rich sediments). The analyses of gas hydrate saturations and the sand contents document that the sediment grain size controls on gas hydrate saturation (Torres et al., 2008; Yoneda et al., 2015; Nanda et al., 2019). In the production test region, two distinct gas hydrate-bearing reservoirs (i.e., the coarse-grained and fined-grained sediments) are investigated to have differently physical properties. The maximum saturations estimated from resistivity and chlorinity are 58.9% and 72.1% of the pore space, and correspond to the 67.1% and 68.2% maximum porosity in GHBSs at Sites W18 and W19, respectively (Table 1) (Zhang et al., 2007, 2017; Yang et al., 2015, 2017; Qian et al., 2018). At W11 and W17, three GHBS intervals can be identified from saturation log curve with maximum saturations less than 45% (Table 1) (Zhang et al., 2007; Zhu et al., 2010; Yang et al., 2015, 2017). Gas hydrate saturations are higher at Sites W18 and W19 than those at Sites W11 and W17.

The resistivity log curves with the upward increasing match well with the shapes of gamma-ray log curves at Sites W19 and W18 (Jin et al., 2017; Zhang et al., 2017) (Fig. 11), and gas hydrates with relatively high saturation preferentially accumulated in coarser sediments related to channel fillings (Jin et al., 2020; Zhang et al., 2020), which document the close relationship between gas hydrate saturations and gamma-ray inferred sediment grain sizes. The average particle size are mainly 7–16 μm, and is about 8 μm at the GHBSs at Sites W11 and W17 (Hu et al., 2017). At the ridge of Sites W18 and W19, the result of the grain size analysis showed that the sediment samples are mainly coarse silt (91.43%–97.31%), followed by sand (1.09–7.57) and clay (0.23%–2.29%) at Sites SC-01 and SC-02 (Wang et al., 2021). The sediment units are likely attributed to the presence of foraminifer and elevated concentration of carbonaceous grains (Kang et al., 2018) showing the characteristics of high porosity and low gamma ray log between horizons H1 and BED (Figs. 3 & 6). The sampled foraminifera are generally 0.065–1.0 mm in size and are within the range of "sand grain size", which increase the average grain-size of the gas hydrate host sediments and provide available pore volume for gas hydrate to accumulate (Chen et al., 2009; Kuang et al., 2019; Li et al., 2019). The foraminifera-rich turbidity sand may mainly compose of the coarse grain-size sediments, showing better reservoir physical properties than the fine-grain size sediments (almost clayey silt at Sites W11 and W17) (Li et al., 2019). These relations confirm that the preferential growth of gas hydrate with relatively coarse-grained sandy sediments where lower capillary force than in silt or clay sediments (Daigle and Dugan, 2011; Suzuki et al., 2015). The better physical properties of GHBSs may interpret why the saturation of GHBSs at Sites W18 and W19 are higher than those at Sites W11 and W17. On account of high porosity and permeability of canyon sediments, while, fluids show worse closure of seal beds at Sites W18 and W19 than those at Sites W11 and W17.

5.2 Reservoir properties from depositional environment

Analyses of well log data can provide important information about the characteristics of hydrate-bearing reservoirs (Mayall et al., 2006; Alves, 2010; Riedel et al., 2011; Waite et al., 2019). The sediment transport history, hydrodynamic conditions, and depositional environments can significantly influence the reservoir physical properties (Mayall et al., 2006; Torres et al., 2008; Waite et al., 2019). Horizons H1 and BED are identified by combining the rock physical templates (Figs. 2 & 12), the well log anomalies (Fig. 11), and seismic interpretation (Figs. 2 & 9) and inversion (Fig. 8a & b). The BED in the submarine canyons of the South China Sea is inferred to have been eroded by turbidity currents (Mayall et al., 2006; Zhou et al., 2015; Jin et al., 2020). During sea-level fluctuation, the activity of canyon-channel system is primarily controlled by the coupling between the canyon head and the coarse-grained sediment supply (Hillman et al., 2017). During periods of enhanced sediment input via gravity flows, perhaps during lowstands of relative sea level, BED developed in the canyons (Chen et al., 2013; Zhou et al., 2015). These bypass channels (or trough) can carried turbidity sediments including foraminifer-rich particles that would either be deposited in levees or be carried further to downslope (Jin et al., 2020), indicating the relatively high energy environment and high stacking rate at the base of the canyon deposits. That can interpret why the low gamma-ray unit between H1 and BED at Sites W18 and W19 occurs the coarse-grained sediments. The sediments with low compaction and high porosity supply more pore volume for gas hydrate occurrence.

In contrast, gas hydrate morphologies at Sites W11 and W17 are pore-filling within silt-dominated sediments (Zhu et al., 2010; Guo et al., 2017). The deposits above BED showing similar sediment characteristics with high porosity and low P-wave velocity to the canyon deposits may be the overbank deposits, which are transported by canyon and are deposited at the ridge (Fig. 8). The sliding surfaces are identified below the BED in the slope deposits at Sites W11 and W17 (Figs. 2 & 9b). The fine-grained sediments exceeding 100 m beneath the BED are identified by parallel-bedded reflections prevail throughout the area suggesting the lower energy sedimentary environment than the canyon deposits. The fine-grained sediments can be caused by the similar shelf slope deposits and local slope sliding deposits. The pore water will be removed during the transport and the sediments will harden, and the porosity will decrease when the sliding and slumping happen in slope deposits (Alves, 2010), which can interpret why the porosity profile at Sites W11 and W17 is lower than that at Sites W18 and W19.

5.3 Gas hydrate-bearing reservoirs for selected production zone

The sediments are dominated by coarse-fine silt deposits in the gas hydrate drilling and production zone in the Shenhu area (e.g., Chen et al., 2009; Yang et al., 2015, 2017; Li et al., 2018; Wang et al., 2021). According to analyzes above, the physical properties of gas hydrate-bearing sediments and grain size analysis show relative better reservoir properties for high saturation gas hydrate occurrence and production at Sites W18 and W19 than those at Sites W11 and W17. However, two gas-hydrate production tests were conducted at the area of W11 and W17, which indicate it is more favorable hydrate production zone. We conclude four reasons for the gas-hydrate production zone at the area of Sites W11 and W17 instead of Sites W18 and W19. Firstly, the amount of gas hydrate resources is the most important factor for location of production site. According the inverted results, gas hydrate-bearing sediments are thicker (2–5 times) and larger area (10 times) at W11 and W17 than those at W18 and W19 (Table 1), which indicate more abundant total amount of gas hydrate resources. Secondly, reservoir continuity and stiffness are also the important factors for hydrate production test (Fujii et al., 2015). The gas hydrate-bearing sediments at Sites W11 and W17 are characterized by multi-strata, inclined and continuous laterally, and may even be connected to the free gas-bearing layers below the BSR (Figs. 2 & 8). Thirdly, the seal beds above GHBSs may show better closure at Sites W11 and W17 than those at Sites W18 and W19. As discussion above, the relatively low energy depositional environment (submarine canyon) occurs lower porosity and permeability sediments, which has higher closure for gas hydrate above GHBSs at Sites W11 and W17. Moreover, the strata at Sites W11 and W17 are harder than those of Sites W18 and W19 due to the older stratigraphic age, higher P-wave velocity, and lower porosity of sedimentation. The BED is near to the gas hydrate-bearing layers at Sites W18 and W19, while it is close to seafloor at Site W11 and W17. Therefore, Sites W11 and W17 are better for engineering drilling especially for horizontal wells because of the formation hardness (Ye et al., 2018). In addition, hardened sediments is also benefit to the stability of production engineering. Lastly, the integrated analyses of LWD and core data show the coexistence of gas hydrate and free gas below the BSR (Li et al., 2018; Qian et al., 2018). The low P-wave velocities can be observed below BSR at both Sites W11 and W17 and Sites W18 and W19 in the inverted velocity profile indicating the presence of free gas (Fig. 8a). However, the distributions of gas hydrate and free gas are occurred in limited area at Sites W18 and W19. In the horizon slide of coherence cube, it can be seen that the lower fault systems are more developed at Sites W11 and W17, and fluid pathways are dominated gas chimneys at Sites W18 and W19 (Jin et al., 2020). Normal faults may be more efficient fluid migration pathway, and transport deep thermogenic gas to shallow layers. The occurrence of free gas will greatly improve the efficiency of natural gas production.

6 CONCLUSION

Combined three-dimensional seismic data with log while drilling (LWD) data and core data in the production test region, we provide a physical-properties-based description of the gas hydrate reservoir associated with the canyon-ridge system, which can help to improve the understanding on reservoir properties and depositional environment at the production test region in the Shenhu area, South China Sea. The statistical thickness and saturation of gas hydrate-bearing sediments (GHBSs) indicate the heterogeneous distribution characteristics at Sites W18, W19, W11, and W17 of two adjacent ridges. The cross-plots of gamma ray, impedance, and porosity at four sites show that the sediments can be divided into the upper and lower layers with the boundaries of horizons H1 and BED. The inverted geophysical properties fit well with the reservoir properties from rock physics models of well sites to identify the gas hydrate layer and the horizons H1 and BED. The thickness of upper layer at Sites W11 and W17 is thinner than that at Sites W18 and W19. The upper layer has higher porosity, low P-wave velocity, and non-gas hydrate-bearing sediments. The low gamma-ray values occur between horizons H1 and BED at Sites W18 and W19, indicating the coarse-grained gas hydrate reservoir from high-energy depositional environment. In the contrast, the primary gas hydrate reservoir at Sites W11 and W17 is the fine-grained sediments. Higher porosity, resistivity and P-wave velocity, coarser grain size, and lower density show better reservoir quality for gas hydrate occurrence. Integrated analysis of thickness and area of gas hydrate-bearing sediments (GHBSs), favorable engineering drilling, and free gas occurring, however, the more favorable hydrate production zone is locate at Sites W11, W17 instead of Sites W18, W19.

7 DATA AVAILABILITY STATEMENT

All data generated and/or analyzed during this study are available from the corresponding author upon reasonable request.