Controls on the Gas Hydrate Occurrence in Lower Slope to Basin-Floor, Northeastern Bay of Bengal

LU Yintao, LUAN Xiwu, SHI Boqing, FAN Guozhang, RAN Weimin, XU Ning, WANG Haiqiang, SHAO Dali, DING Liangbo, and WANG Xingxing

Controls on the Gas Hydrate Occurrence in Lower Slope to Basin-Floor, Northeastern Bay of Bengal

LU Yintao1), *, LUAN Xiwu2), SHI Boqing3), FAN Guozhang1), RAN Weimin4), *, XU Ning3), WANG Haiqiang3), SHAO Dali1), DING Liangbo1), and WANG Xingxing5)

1),310023,2),,266590,3),100031,4),,266071,5),,430074,

High resolution seismic data and newly acquired logging data reveal the presence of gas hydrates in the deep-water area in the northeastern Bay of Bengal. Bottom Simulating Reflectors (BSRs) appear at 430ms beneath the seafloor, with the features of typical gas hydrates reported elsewhere except for some discontinuity. The BSR distribution is closely related to the position of anticline structures and turbidite channels. Anticlines provide conduits for the upward migration of gases from deeper intervals, while the turbidite sands within channels act as reservoirs for gas hydrate accumulation. High sedimentation ratesin the Bengal Fan were ge- nerally favorable to produce a great amount of methane gas,providing favorable preconditions for the formation of gas hydrates. The discovery of gases in adjacent area indicates the contribution of the biogenic gas to the formation of gas hydrates. Meanwhile, seismic sections provide the evidence for the potential thermogenic gas kitchen in deep intervals. The anticline structures and the associated vertical fractures may constitute vertical fluid flowing conduits, which hydraulically connect the two deeper thermogenic petroleum systems (., rifting and Eocene–Oligocene source rocks) with the turbidite reservoirs and thus facilitate the fluid migration from the sources to the reservoirs, generating favorable conditions for gas-hydrate accumulation in such foreland setting.

Bay of Bengal; gas hydrate; anticline; turbidite sands

1 Introduction

The Rakhine Basin attracts broad international interests by its unique tectonic features, as well as the prolific hydrocarbon resources. It is closely related to three spectacular geologic systems: Himalayan Range as the world’slargest orogenic system; Ganges-Brahmaputra (G-B) Delta as the world’s largest fluvio-deltaic system; and Bengal Fan as the world’s largest submarine fan system. Himalayan Range is drained by the G-B Rivers to develop the G-B Delta system (the present Bengal Basin), which is consisted of riverine channels, floodplains, and delta-plain withan area of about 200000km2in total. Deep-water sedimentary systems develop further seaward and build the Bengal Deep Sea Fan, which extends southward 3000km and reaches as far as 7?S (Moore., 1974; Emmel and Curray, 1985;Curray., 2003; Lu., 2021).

The large amount of sediment deposition and hydrocarbon potential have drawn substantial interest in and study on the Bengal Basin and Bay of Bengal (Lietz and Kabir, 1982;Hiller and Elahi, 1988;Alam, 1989; Johnson andAlam, 1991;Khan,1991;Lohmann, 1995; Shamsuddin andAbdullah,1997; Uddin and Lundberg, 1998;Gani and Alam, 1999). Historically, oil was produced along the coast in the Rakhine Basin from hand-dug wells, which indicated an active petroleum system. However, petroleum discoveryattempts were unsuccessful until 2004. From 2004 to 2006, three gas fields were found: Shwe, Shwe phyu, and Mya (Yang and Kim, 2014).These gas fields are situated in the shelf and slope settings with water depths ranging from 90m to 600m. The target interval is composed of late Plio- cene submarine turbidite sandstones, 2900–3100m beneaththe seabed. Methane is the dominant component, contribu-ting up to 99% of the gas by volume. The isotope analysis of the gas samples has demonstrated that the methane is biogenic (Yang and Kim, 2014).

Basu. (2010) postulated that this biogenic system may extend into deeper waters, forming an efficient biogenicpetroleum system in deep-water region of the Rakhine Ba-sin. Although economic thermogenic petroleum deposits have not been discovered thus far, the combination of the shallower biogenic system and the deeper thermogenic pe-troleum system make the deep-water Rakhine Basin a new promising frontier for exploration (Basu., 2010).

Comparing with the conventional petroleum system, theunconventional petroleum system, such as gas hydrate, has not been reported much in this area. Even though drilling and coring activities carried out by JOIDES Resolution have verified the existence of gas hydrates in the Krishan Godavari Basin in the western part of the Bay of Bengal (Collett., 2015), and the occurrence and distribution of these gas hydrates in western Bay of Bengal are found to be largely controlled by a large-scale fault system along the NS bathymetric mound (Dewangan., 2010, 2011), their distribution and characteristics have not systemically discussed, especially those in Northeast Bay of Bengal, including the Rakhine Basin. In this study, the occurrence and geometry of gas hydrates in the Rakhine Basin, north- eastern Bay of Bengal are revealed for the first time by seismic interpretation and geophysical analysis. Also, these these results are confirmed by the welllogs of a newly ac-quired industrial well penetrating gas hydrates reservoirs.

2 Geological Setting

The Rakhine Basin, approximately 850km in length and200km in width, lies in the western coastal province ofMyanmar and eastern fringe of the Bay of Bengal. It isbounded to the east by the Indo-Burma Ophiolite Belt, which continues northward as the Chittagong Fold Belt (CFB) in Bangladesh and also appears as the Tripura-Cachar fold and Disang Flysch Belts in India. In the South, the RakhineBasin merges with Andaman-Nicobar-Sunda-Java fore deep basin system (Fig.1).

The study area is located in the northeast deep-water area of the Bay of Bengal, which is 125km west of Myanmar and has a water depth of 1000–2100m. The progressive and sustained movement of the Indian Plate towards the Eurasian Plate led to the closure of the Tethys Ocean, and the subsequent plate collision contributed to the upliftof Himalayas and Tibetan Plateau as well as the consequent initiation and development of the Bengal-Rakhine Basin (Curray and Moore, 1971; Molnar and Tapponnier, 1975; Cochran and Stow, 1990; Gibbons., 2015). The oblique subduction of the Indian plate beneath the Burmese plate to the northeast resulted in the formation of a series of N-S trending structural terrains in the eastern Bengal area, which consist of subduction zone (the Rakhine Basin),accretionary prism (the Indo-Burman Ranges), arc-induced basins (the central Burma Basin), and plateaus (Fig.1). Theuplifted Indo-Burman Ranges are composed of flysch deposits and ophiolites, 1500km long and up to 2000m high (Alam., 2003; Curray., 2003). These compressional deformations, which are among the most prominent onshore, coastal, and partially offshore structures in Myan- mar, are related to the subduction and propagated from east to west, from the Indo-Burman Ranges to the Bay of Bengal. The regional compressional stress progrades westwards and weakens progressively from on- to offshore, causing the relatively slight anticline belts in place of the significant uplifts observed in the Bay of Bengal or the denselyspaced thrust faults (Collins., 2002). Lastly, an enormous volume of sediments from erosion of the Himalayas and Indo-Burman Ranges has been transported into the Bengal Basin and the Bay of Bengal, accumulating in the Bengal Fan(Curray., 2003). This regional progression of sub-duction, collision, tectonic uplift, and, later, intraplate deformation is evidenced in the Bengal Fanstratigraphy partially revealed by DSDP 217/218 and ODP 116 (von der Borch., 1974; Curray and Munasinghe, 1989; Cochran and Stow, 1990; Curray., 2003; Fig.2).

Fig.1 Generalized tectonic map of the northeastern Indian Ocean and surrounding areas. Sub-divisions of the Bengal Fan and its channels are modeled after Curray et al. (2003) and Bastia et al. (2010); the positions of the anticline axes are from Uddin and Lundberg (2004); the 85?E and 90?E ridge trends are from Curray et al. (2003); ODP, DSDP, industrial wells, and gas hydrate discoveries are after Dewangan et al. (2010) and Yang and Kim (2014); mud volcanos are after Maurin and Rangin (2009); the 2D published seismic lines are from Rangin and Sibuet (2017); the hinge zone separating the shallow Indian platform in the northwest from the deeper Bengal fore deep in the southeast is from Gani and Alam (2003). CTFB, Chittagong-Tripura Fold Belt; B-C, Barisal-Chandpur; SONG, swatch of no ground; COB, continental ocean boundary.

The India-Asia collision is commonly thought to have occurred at 55–50Myr (Sclater and Fisher, 1974; Hodges, 2000). The deposition gap between 50Myr and 38Myr is marked by an unconformity in the foreland basin and a hiatus at the base of the Bengal Fan, and is explained by the crustal thickening process at the collision boundary (Naj- man., 2008). Since 38Myr, the G-B and other related rivers have been delivering clastic sediments to the Bengal- Rakhine Basin, Bay of Bengal, and Indo-Burman Ranges, building a thick (20km) sequence above the unconformity(Uddin and Lundberg, 2004). The early Miocene under the thrusting along the Himalayan front is well documented by the forced-regressive sedimentation patterns in the slopeand shelf systems of the Bengal-Rakhine Basin (Khin.,2014). Sequential evolution since Miocene in Bay of Bengal is characterized by prograding delta and deep water gravity systems (Khin., 2014), in which sand-rich se- quences serve as hydrocarbon reservoirs (Murphy, 1988). The distribution of Pliocene Kyauktan formation implies that both N-S and E-W sediment fairways were active during that time (Khin., 2014; Yang and Kim, 2014).

The upper Quaternary Bengal Fan formation consists of massive channel-levee complexes formed during the period of lower sea-levels. Four units, marked by A, B, C, and D from oldest to youngest, can be seismically identified within the Pleistocene sediments in the upper fan (Cur- ray., 2003). With the rising of sea level during Holocene, the sediment supply to the Bengal Fanhas declined significantly. Holocene turbidity currents that feed the activechannel in the middle of the fan are small and infrequent (Curray., 2003).

The present Bengal Fanis characterized by very gentle slopes. Emmel and Curray (1985) reported the gradient of 0.74mkm?1(0.042?) for the distal fan to the south of 6?N and the average gradient of only 0.85mkm?1(0.049?) even as far as north of 15?N.

Fig.2 Stratigraphic chart of the deep-water offshore area in the Bay of Bengal. L, M, and U denote lower, middle, and upper, respectively.

3 Data and Methods

High resolution 3D marine seismic data were acquired in 2012, which covered an area of 2070km2(Fig.3). 2D marine seismic data across both Blocks AD-8 and AD-1 were acquired in 2008 (Fig.3). The dominant frequency of 3D seismic data in the 2000–3000ms interval is 60Hz, while the dominant frequency in the 2000–6000ms is 50Hz. The dominant frequency of 2D seismic data in the 100–2000ms interval is 60Hz, while that in the 2000–4000ms interval is 45Hz; for the deeper interval, the frequency is 25Hz. The seismic sequences determined by well data on the neighboring Shwe phyu, Shwe, Mya shallow water gas fields (Figs.2 and 3) were extended into the study area in order to validate the seismic profile interpretation. Seismic responses and attributes are analyzed, providing more detailed information on the gas hydrate-bearing interval, including the lithological variation, the spatial distribution of gas hydrates, as well as the gas-bearing response.

Fig.3 Location of the regional 2D seismic lines and the 3D survey area. The 2D regional seismic lines cover the primarily PSC blocks AD-8 and AD-1; the 3D survey area covers parts of AD-8 and AD-1. The locations of exploration and appraisal wells drilled on the shallow-water gas fields are also shown in Yang and Kim (2014). Bathymetrycontours are in meters. The 2D seismic line D published by Yang and Kim (2014) is used in this paper for sequence calibration. Shwe phyu, Shwe, and Mya are neighboring gas fields.

An industrial oil and gas exploration well (well-A) was drilled in the study area in 2018 by China National Oil andGas Exploration and Development Company (CNODC), which provided more detailed information for the seismic sequences. Meanwhile, this well (well-A) provided more significant direct information for the occurrence of gas hydrates in study area. The well log sequence used in this study includes gamma ray (GR) and delta time (DT). The lithology is interpreted by GR data, combining with seismic responses, while the DT data is used to calculate seismic velocities (). The lithology and interval property are used to analyze the possibility of the existence of gas hydrates. The bottom depth of the gas hydrate stability zone (BGHSZ) was herein estimated by using the hydrate prediction program from Colorado School of Mines (CSMHYD) (Sloan, 1998). Well-A also revealed the very low geother- mal gradient, 28℃km?1, and gave parameters for simulating the phase balance of gas hydrates. The temperature of seafloor is given as 3.92℃. The calculated depth of BGH- SZ in the area with the water depth of 1500m is 436mbsf. The DT data indicates the averagein shallow interval is about 2000ms?1, thus the BSR should be at about 430ms below seafloor in seismic sections.

4 Results

4.1 Characteristics of Anticline Belts

Anticline structures are better imaged on the 2D seismic lines (Figs.4, 5). All of these structures are low amplitude in seismic images and possess simply symmetric characteristics with vertical axial planes. The most typical anticline structure shown on the easternmost of seismic line (., anticline 4 in Fig.4 and anticline 3 in Fig.5), has uplift amplitudes (given in terms of TWT) of about 400ms (or 800m given an average velocity of 4000ms?1), spanning a distance of 12km between the westernmost and easternmost of the limbs.

The deformation of the anticline structure affected the entire sedimentary sequence from the deep interval (T70) to near the seafloor, indicating the recent tectonic activity. Maurin and Rangin (2009) reported the existence of a de- collement surface below the folded sedimentary layers and stated that the basement was unaffected by this folding. In our data, however, the drift sequence beneath T70 is clearlyfolded in the same way as the overlaying reflectors (Fig.5). As such, the pre-Cenozoic and Cenozoic sedimentary se- quences are actually locked together and behave harmo- nically as a single system. In the anticline structures imaged on seismic sections (Figs.4, 5), the high-amplitudedrift se- quences show pronounced upward bends supporting the overlaying Cenozoic anticline sequences.

The anticline structures revealed by 2D seismic lines are confined largely within the northeastern corner of PSC block AD-8 and the northern corner of PSC block AD-1 (Figs.4, 5), which is consistent with previous results given by Rangin and Sibuet (2017). These anticline structures are westward migration fronts of the Indo-Burman accretionary wedge (Hutchison, 1989; Gani and Alam, 1999).

Small but discernable anticline structures, 25km (axis to axis) to the west of the large anticline structure (anticline 3), are also identified (Fig.6). A clear volcanic relief exists at the base of this small-scale anticline structure; this volcanic relief penetrated upward through the drift se- quence and clearly supported the folding of the overlying sedimentary sequence (Figs.4, 5).

4.2 Seismic Response of Deep-Water Turbidite Systems

Seismic sections along the slope show enormous turbidite channel-levee complexes (CLCs), which were built largely by frequent turbidity currents within the highstand system tract (HST) (Fig.6). The CLCs within the 3D survey area are smaller (Fig.6) than those revealed in 2D seismic sections (CLC-1 and CLC-2, Figs.4, 5), suggesting that the eastern part of the Bay of Bengal may not be the main area which the submarine turbidite channels pass through.

Fig.4 Interpreted regional seismic section, exhibiting the anticline 4–5 developed from upper slope (NE) to lower slope (SW). Three sedimentation mega-sequences can be distinguished from bottom to top: open marine (T70–T60), forced regression (T60–T40), and delta front progradation (T40–present seafloor). Channel systems are found mainly in the forced regression layer; large CLCs are found from the Pleistocene to present layers. Dotted purple lines indicate upward fluid migration channels, and the inverted black triangle indicates the highest point of the subducting flexure. The black box is zoomed in Fig.14. For the location of the seismic line, see Fig.3.

Fig.5 Interpreted regional seismic section, exhibiting the anticline 3–6 developed from upper slope (NE) to lower slope (SW). Three sedimentation mega-sequences can be distinguished from bottom to top: open marine, forced regression, and delta front progradation. Channel systems are found mainly in the forced regression layer; large CLCs are found from the Pleistocene to present layers. Dotted purple lines indicate upward fluid migration channels, and the triangle indicates the highest point of the subducting flexure. The black box is zoomed in Fig.15. For the location of the seismic line, see Fig.3.

Fig.6 Quaternary sedimentary sequences and CLCs developed on seismic section line AA’ along the slope in the 3D survey area. MFS, maximum flooding surface; LST, lowstand system tract; MTD, mass transport depositions, HAR, high amplitude reflector, SQ, sequence; BSR, bottom simulation reflection, see zoomed-in features in Figs.8, 9. For the location of line AA’, see Fig.3.

Under the open sea or pelagic sedimentary environments, slope systems in passive margins are mostly muddy. The slope channels, however, can serve as favorable conduits for transporting large volumes of sand from shelves down- dip to the muddy slope systems. Correspondingly, the seis- mic sections contain not only the drift layers, but also highamplitude reflectors (HARs) interbedded in the low-ampli- tude background above the drift layer, especially in the forced regression and delta progradation layers (Figs.4–6). These HARs arise partly from sediments, such as silt and sand materials transported by turbidite channel systems and deposited in the form of channel fill, lobes, or sheets in the muddy slope background (Fig.6). The distribution of HARs in Fig.6 is more complicated due to the development of anticline structures. From the root-mean-square (RMS) amplitude map extracted from the top surface of HARs, we could identify the features of channel complex (Figs.6–8). As clearly shown in the 3D seismic sections (Fig.6), the anticline structure almost deformed the entire Cenozoic sedimentary column, and created faults densely distributed within the axial zone and limbs of the anticline structure.

4.3 Characteristics and Distribution of BSR

HARs also include Bottom Simulation Reflectors (BSRs), which often mark the base of gas hydrate stability zones and usually occupy the sub-surface of the 500ms TWT be- low the seafloor (Lu., 2017). In the study area, along with HAR characteristics, the BSRs typically show revers- ed polarity and crosscut the normal-strata, which makes these features prominent and easy to be identified (Figs.6, 8). The BSRs in these seismic sections are not as continu- ous as those typically reported elsewhere (Lu., 2017). These BSRs appear segments over very short distances, usually extending only several hundred meters and ending suddenly instead of gradually fading away (Figs.6, 8).

Geographically, our study area is located in the transitional zone between the lower slope and the basin floor and dips southwestwards, with the water depth increasing from 1000m at the northeast corner to greater than 2000m at the southwest corner. The 1000m and 1500m isoba- ths extend roughly in the direction of northwest in parallel to the western Myanmar coastline. The 2000m isobath, how-ever, displays roughly in east-west direction and meanders considerably (Figs.3, 7). The BSRs are concentrated predominantly in the central region of the 3D survey area, though BSRs are occasionally and irregularly distributed in the eastern and western parts of the 3D survey area which is featured by low-RMS-amplitude background (denotedby purple shading in Fig.7). Generally, four N-S elongated BSR bands can be identified from east to west in the 3D survey area (Fig.7).

Fig.7 RMS amplitude map extracted from the top BSR surface. Time window: ±20ms. The high amplitude anomalies indicate the widespread deep-water turbidite channel systems. The gas hydrate bearing sands further enhance the seismic amplitude, making the anomalies stronger. For the locations of the 3D survey area and PSC blocks AD-7, AD-1, A-3, and AD-8, see Fig.3.

Fig.8 (a) The magnified seismic image of the BSRs and deep-water sediment systems. (a’) The interpreted seismic section with more sedimentary elements, including channel complex, levee, and MTD. The BSRs and BGHS are marked by green dash lines, while gas hydrate bearing and free gas bearing sand layers are marked in seismic section as well. The BSRs clearly crosscut discontinuously the normal sediment layers. See the location in Fig.7.

4.4 Well Log Responses

One petroleum industrial well in northeast of the study area (Fig.3) was drilled through the shallow strata and re- liably revealed the occurrence of gas hydrates. The HAR package penetrated by this well is investigated by several well logging methods, such as gamma ray (), and delta time ().

According to the logging results, the reflection package in the shallow intervals could be divided into 4 sections (Fig.9). They are section A: dim reflectors in shallowest section, the internal reflectors expressing continuous undulate features; B: transparent reflectors interbedded with chaotic HARs, as the boundary between HARs and the overlying section; C: HARs section, with BSR as the base; D: medium amplitude and continuous reflectors.

Four sections are characterized by differential well log responses, includingand P-wave velocity ()calculated from(Figs.9, 10). Section A: thevalue in section A stays in high domain, with a couple of higher value intervals. Meanwhile, thekeep at low value. Sec- tion B: thevalue is still high overall, but with top high impedance interval, and drop down to lower value. Section C: thelog in HARs expresses the zigzag feature, with more box-shape low value intervals. Theincrease significantly, with several peaks. Section D: thevalueincrease dramatically, with thedropping down to a constant value. An obvious boundary could be recognized in BGHS, as BSR, dividing thelog curve into two parts. The upper part has lowerlog values (occupying the left domain), while the lower part has high values (occupying the right domain), with HARs still developing in the lower part.

Fig.9 Relative acoustic impedance section across the well-A, showing HARs in the shallow interval, which are regarded as sandy intervals. The blue curve in seismic section is GR log while drilling, while the pink curve is seismic velocity (Vp) calculated by DT. The sandy intervals are interpreted as highly concentrated gas hydrate bearing layers. The predicted BGHS, which coincides with seismic trough reflectors, is marked in seismic section. See location in Fig.3.

Fig.10 Well logs of well-A penetrating the shallow intervals. The well logs are the same in Fig.9. The lower GR value is filled by yellow color, while the higher GR value is filled by blue color. See well location in Fig.3.

5 Discussion

5.1 Occurrence of Gas Hydrates

Based on the widely accepted lithological interpretation to the logging curves (., Qian., 2018), the highervalue intervals indicate the shale or mud rich intervals, while the lowervalue intervals indicate those sand richintervals (Fig.10). Seismic velocity is generally low in sandy intervals occupied by fluids. However, the sand-rich intervals that are shown as HARs in seismic sections correspond to high seismic velocity () in study area. This anomalous feature is similar to typical gas hydrate bearingsediments elsewhere, such as Gulf of Mexico, South China Sea (Boswell., 2012; Qian., 2018; Liang., 2019; Jin., 2020).

The relative impedance section could provide both litho- logical and petrophysical information. The seismic reflec- tion characters and well log facies jointly indicate various lithologic packages in shallow target intervals (Figs.9, 10). Section A: the highvalues and dim seismic reflectors indicate mud-rich deposits are dominated in this section, while the interbedded lowervalues and high impedance reflectors indicates there are limited sandy intervals in mud rich deposit background. Section B: transparent and chaotic reflectors with highvalues indicate the mass transport complex (MTC) in the mud rich deposit background. The chaotic HARs with lowervalues indicate thin sand rich intervals in the mud-rich MTC, such assandy blocks. Section C: HARs with several intervals with lowervalues indicate sandy deposits are dominated along this section. At the base of this section BSR occurred as BHGS, which is the major gas hydrate bearing interval. The highvalues and severalpeaks in this section indicate several highly concentrated gas hydrate bearing beds (Qian., 2018). In the plan view across this section (Fig.7), these HARs are featured by deep water channel complexes. Section D: Medium amplitude and conti- nuous reflectors with highvalues indicate mud rich intervals are dominated. The continuous HARs with lowervalues indicate thin interbedded sand-rich deposits. How- ever, thevalues in this section keep at a lower level, without peaks, indicating gas bearing or water bearing in- tervals (Qian., 2018) without gas hydrates. This feature coincides with the interpretation from the BSR in seis-mic sections and predicted fluid phase distribution maps. As this section is underneath the BGHS, the possible fluids in the bed are free gas and/or water (Figs.9, 10).

5.2 Controls on Gas-Hydrate Accumulation

5.2.1 Influence of anticline structures on the gas hydrate distribution

The Shwe, Shwe phyu, and Mya gas fields were disco- vered consecutively between 2004 and 2006 in the offshore area of the northwest Myanmar (Yang and Kim, 2014). In space, the Shwe phyu, Shwe, and Mya fields are elongated from the northwest to southeast and located successively along an anticline structure named anticline 1 herein (Fig.1). The spatial characteristics of anticlines 1, 2, and 3 were given by Rangin and Sibuet (2017) and confirmed by seismic data from Yang and Kim (2014). Herein, we mainly detail the spatial characteristics of anticlines 3, 4, 5, and 6 which have not been reported yet (Fig.11). The distribution of anticline 1 to anticline 6 looks harmonious; their uniform trends and nearly regular spacing are consistent with the westward migration of the accretionary wedge. The fold amplitudes and along-strike distances of the six anticlines decrease gradually from east to west; we found only a small anticline structure in the western-most region of the study area in seismic sections (Fig.4), which indicates the present location of the deformation front. Inspired by the relationship between the anticline 1 strike andthe distribution of the Shwe, Shwe phyu, and Mya gas fields, in combination with the observed BSRs confined within the anticline belts (Fig.6), it could be easily infer- red that the BSRs are distributed along the anticline strikes (Fig.11).

However, the coupling relationship between the BSR and anticline belt along anticline 4 is complicated. The BSR belt associated with anticline 4 is divided into two branches. For the west branch (Fig.11), it is clear that thereis no corresponding anticline structure. Therefore, other me- chanisms may control the distribution of the BSRs.

Fig.11 Distribution of the BSRs (yellow area) superimposed on the seafloor morphology map, obtained from 3D seismic data. Morphology is scaled in the unit of meters. The locations of anticlines 1, 2, and 3 are taken from Rangin and Sibuet (2017) and confirmed by line D from Yang and Kim (2014). The thick black lines indicate the anticlines identified in seismic sections. For the locations of the PSC blocks, 3D survey area, gas fields, and line D, see Fig.3.

5.2.2 Coupling relationship between deep-water tur- bidite sands and gas hydrate distribution

The discontinuous BSRs could be caused by different free gas supplies through vertical fluid conduits, which are generated in association with the development of anti- cline structures under the influence of westward migration of the accretionary wedge; the BSR discontinuities can also be triggered by the instability and gasification of gas hydrates due to the overpressure and high temperature through upward fluid migration (Horozal., 2017; Liu., 2019). In the study area, the BSRs may also be disturbed by the widespread CLC systems, since the gas hy- drates tend to accumulate in the discontinuous turbidite sands within the CLC systems (Figs.6, 8).

In the RMS amplitude map, the upper BSR surface shows channel complex features, which indicates the combination of BSRs and submarine slope channels (Fig.7). Due tohigh transportation capability of turbidity currents, coarse- grained turbidite sands are commonly seen in submarine channels (., Wang., 2018). In comparison to normal pelagic or overbank silty and muddy sediments, higher- porosity turbidite sand may provide more space for gas hydrate accumulation. Thus, under the given temperature and pressure conditions, turbidite sand would be more prone to accommodate more gas hydrates than normal pe- lagic or overbank muddy deposits. When the turbidite sands are filled and cemented with gas hydrates, their seismic impedance would be enhanced. As a result, the higher RMS amplitudes of upper BSR surface are seen on the seismic profile.

5.2.3 Influence of high sedimentation rate on gas hydrate accumulation

In the Bay of Bengal, the BSRs mostly existed in the Pliocene strata, which has high sedimentation rate. Taking our regional seismic section (Figs.4, 5) as an example, the thicknesses of the strata deposited during Miocene (T60–T30), Pliocene (T30–T20), and Pleistocene to present (T20–seafloor) are 1270ms, 650ms, and 950ms in TWT, respectively, in the southwest part of the line, and 1400ms, 710ms, and 1350ms in the anticline zone in the northeast part of the line. Using 2000ms?1, 1800ms?1, and 1750ms?1derived from well logs as the interval velocities for the Miocene, Pliocene, and Pleistocene sequences, respectively, the interval thickness and the corresponding accumulation rates could be estimated as well. The results show that the sedimentation rates could reach to 71mMyr?1during Mio- cene, 212mMyr?1during Pliocene, and 322mMyr?1during Pleistocene for the southwest part of the line, and 79 mMyr?1during Miocene, 232mMyr?1during Pliocene, and 457mMyr?1during Pleistocene for the northeast part of the line. In comparison to data reported elsewhere in the world (., Luan., 2009, 2010, 2011), the Pliocene and Pleistocene sedimentation rates here are fairly high. This would be favorable for the formation of gas hy- drates from the following perspectives: a) the high sedimentation rates usually corresponded to the high porosity, which was favorable for hydrate accumulation; b) the highsediment supplies increased the pressure on the gas hydrate layer, favoring the hydrate preservation; c) the high sedimentation rate facilitated the transport of organic matter to the deep water area. The sediment supplied to the Bengal Fanwas primarily composed of turbidites sourced from Himalayas, thus the deep-water sediments would be rich inboth sands and terrigenous organic matters (Imam and Hussain, 2002; Gani and Alam, 2003).

5.2.4 Gas source of gas hydrates

The discovery of the Shew gas field adjacent to the north- east of the study area confirmed that biogenic gases are very rich and they provided the potential gas sources for the gas hydrate formation in the Bay of Bengal deep water area (Yang and Kim, 2014). However, we could not rule out the possibility that the thermogenic gas migrated from the deep intervals and contributed to the formation of gas hydrates. Although no thermogenic gas was disco- vered due to the limited penetrating depth of the well in the study area yet, the seismic sections provide evidences of thermogenic gas kitchen in deep intervals. Laterally continuous HARs prominently occur in the lower part of the sediment, at the depth of 7000ms on the regional seismic sections (Figs.4, 5). These HARs with middle to low frequency are parallel with each other and grouped into a thin sequence around 200–400ms in thickness. This kind of thin sequence is widespread and well recognized in our study area. Through comparison, this study found that the thin, high reflective sequence is the same as the sequence below the ‘green’ reflector mentioned by Sibuet. (2016) and Rangin and Sibuet (2017), which was also observed on the B2-01 line passing DSDP site 217 (Moore., 1974; Rangin and Sibuet, 2017). According to DSDP site 217, we confirm that the sequence age spans from the early Campanian through the Maastrichian to the end of the Paleocene; the Maastrichtian strata is dated based on the presence ofspecies, which indicate a pe- lagic sedimentary environment (Maurin and Rangin, 2009).The sequence is composed of chert, chalk, and nanno chalk, and the consistent distribution of chert from bottom to top is similar to the porcellanite and siliceous limestonesaccreted in the southern Arakan range in Myanmar (Moore., 1974; Rangin and Sibuet, 2017). This sequence was most likely formed in a pelagic depositional environment after the breakup of eastern Gondwana and during the se- paration and northward drift of the India/Bay of Bengal with respect to Antarctica in the late Cretaceous to Paleocene. The top of the drift sequence is denoted by T70 and is characterized by a discontinuity surface separating the underlying drift phase from the overlying southward progradation phase of Ganges delta after the Cenozoic India- Asia collision (Figs.4, 5).

Thick Tertiary sediments accumulated over the drift se- quence, which contains open marine, forced regression, and delta front progradation sequences. The entire Tertiary sedimentary succession has a thickness (given as a TWT) of over 4700ms (or about 9400m given an average velo- city of 4000ms?1). The open marine sequences over T70 are dominated by parallel, continuous, and dense thin reflectors with weak reflective amplitudes. The lowest suc- cession, with a thickness of 200ms directly above T70, contains weak-amplitude area in the seismic profile, which indicate high content of fine-grained sediments, most pro- bably mud, clay, or shale deposited in open ocean and/or deep water conditions (Fig.12). The flank onlap found in this study on both sides of the subducting flexure above T70 indicates an open marine sedimentation environment far from the provenance (Fig.12). Well-laminated seismic reflectors could be clearly seen below the drift sequence (Figs.4, 5, 12), indicating the continental (as opposed to oceanic) crust characteristics as previously discussed by Rangin and Sibuet (2017). We agree that the study area remained a part of the Indian plate throughout and occupied the distal portion of the Bengal-Mahanadi Basin, and we interpreted the dipping reflectors (DRs, in fan-shaped configurations) below the drift sequence on seismic sections (in the southwest portion of the 2D survey area) as a rifting sequence (Biswas, 1999) before the breakup of east Gondwana (Figs.4 and 5). The thickness of the rifting se- quence below the drift sequence is at least 1000ms (or 2300m given an average velocity of 4600ms?1). Thus, there are abundant source rocks developed within the rift- ing sequence compared to the Mahanadi rift basin.

Based on the interpretation of regional seismic sections shown above, we deduce that thermogenic petroleum sys- tems developed within both the rifting sequence below the drift layer and the Eocene–Oligocene sequence above the drift layer. Maturity modeling shows that both the rift- ing source rock and the Eocene–Oligocene source rock are mature enough to fall within the ‘oil window’ (Zhu and Li, 2012; Duan and Huang, 2015). The northern Bay of Bengal, which is floored by the attenuated continental crust,has not been strongly deformed since the end of the Gond-wana rifting. This suggests large hydrocarbon potentials in both thermogenic petroleum systems (Rangin and Sibuet, 2017).

Fig.12 Seismic section of the deep intervals in study area, to display the features of source rock sequences. The strata are supposed to be deposited in open marine environment, with efficient gas charging quality. The fault system acts as the migration path for gases from source kitchen to reservoirs in shallower layers. The yellow arrows indicate gas migration paths. The gas chimney and other chaotic features in fault system imply the gas migration. See location in Fig.4.

5.2.5 Gas migration for gas-hydrate accumulation

Seismic profiles show that the most common fractures take the form of ‘white cracks’ in the current study. The so-called white cracks are the white lines between the un- disturbed seismic-reflection events. Where the continuous reflections stop, the amplitude is diminished to ‘white’, forming a thin crack between the continuous reflection re- gions. The white cracks can be single, double, grouped, branching, or joining together. Most of the white cracks are rooted beneath T70 and grow up through the Cenozoic clastic sequences till the end of the Holocene (Figs.4, 5).

We interpret the anticline structures, chaotic zones, and white cracks visualized on the regional seismic lines as vertical fluid flow conduits, which hydraulically connect the deeper layers with the overburden ones (Cartwright., 2007). As mentioned earlier, the pre-Cenozoic layer and the overlaying Cenozoic clastic layer were deformed harmonically; the vertical fluid conduits rooted in the pre- Cenozoic layer connect the two thermogenic petroleum systems, including the rifting source rock below the base- ment (T70), the Eocene–Oligocene source rock above the basement, and the turbidite channel systems and delta front progradation layers (Fig.13). The compression due to the accretionary wedge migration and subduction renders the environment favorable for the efficient fluid migration up-ward from the source layers (Cartwright., 2007; L?seth., 2009).

Fig.13 Conceptual sketched model for the genesis and accumulation of gas hydrate in the deep-water region of northeastern Bay of Bengal. Two types of gas sources, biogenic gas source in shallow intervals and thermogenic gas source in deep intervals, provide gases for the accumulation of gas hydrates in study area. The thermogenic gas migrates upwards through the conduits such as faults and gas chimneys. The gas tends to accumulate in the tectonic high areas, such as anticlines; mean- while, the turbidite sediment systems provide favorable accumulation spaces for gas hydrates. DRs, dipping reflections.

Due to the development of the very thick Bengal Fan, the buried source rock layers are deeper than most current offshore petroleum systems and beyond the reach of most offshore petroleum development technology. However, thevertical fluid conduits could drive fluids to relatively shal- lower reservoirs, such as channels (Figs.4, 5), which were widened within the forced regression sequence under the compression. The channels are usually filled with coarse sands and may serve as a high-quality reservoir for the ac- cumulation of fluid migrating upward from deeper source layers (Fig.13).

6 Conclusions

The integration of seismic data and well logs reveal the occurrence of gas hydrates in the shallow intervals in nor- theastern area of Bay of Bengal. The BSRs in study area exhibit discontinuous, high-amplitude and reversed-pola- rity reflectors, which crosscut the surrounding strata. The distribution of BSRs was significantly controlled by the joint effects of anticline structures and slope channel turbidity sands. The gas tends to accumulate at the tectonic high areas, such as anticlines, and the turbidite sediment systems. High sediment rates due to the sufficient terrigenous sediment supply from Himalaya also promoted the gas-hydrate accumulation in the study area. Two types of gas sources, biogenic gas source in shallow intervals and thermogenic gas source in deep intervals, provide gases for the accumulation of gas hydrates in study area. The vertical fluid flow conduits, including anticline structures and vertical fractures, hydraulically connect the two deeper thermogenic petroleum systems (., rifting and Eocene–Oligocene source rocks) with the shallow turbidite reservoirs. The thermogenic gas could migrate from deep inter- vals to shallow intervals, contributing to the gas-hydrate accumulation in the deep-water area of the northeastern Bay of Bengal.

Acknowledgements

We wish to thank the China National Oil and Gas Exploration and Development Company (CNODC) and Hangzhou Research Institute of Geology research team for providing the seismic data and approving this publication. The research is funded by the National Natural Science Foundation of China (Nos. 92055211, 42076219, 420060 67) and the Construction Project of China ASEAN Marine Seismic Data Platform and Research Center (No. 121201 0050017001).

Alam, M., 1989. Geology and depositional history of the Cenozoic sediments of the Bengal Basin of Bangladesh., 69: 125-139, DOI: 10.1016/0031-0182(89)90159-4.

Alam, M., Alam, M. M., Curray, J. R., Chowdhury, M. L. R., and Gani, M. R., 2003. An overview of the sedimentary geo- logy of the Bengal Basin in relation to the regional tectonic framework and basin-fill history., 155: 179-208, DOI: 10.1016/S0037-0738(02)00180-X.

Bastia, R., Das, S., and Radhakrishna, M., 2010. Pre- and post- collisional depositional history in the upper and middle Bengal fan and evaluation of deepwater reservoir potential along the northeast continental margin of India., 27 (9): 2051-2061, DOI: org/10.1016/j.marpet geo.2010.04.007.

Basu, P., Verma, R., Paul, R., and Viswanath, K., 2010. Deep waters of Rakhine Basin–A new frontier?,. India, 1-7.

Biswas, S. K., 1999. A review of the evolution of rift basins in India during Gondwana with special reference to western Indian basins and their hydrocarbon prospects., 65 (3): 261-283.

Boswell, R., Collett, T. S., Frye, M., Shedd, W., McConnell, D. R., and Shelander, D., 2012. Subsurface gas hydrates in the northern Gulf of Mexico., 34: 4-30.

Cartwright, J., Huuse, M., and Aplin, A., 2007. Seal bypass sys- tems., 91: 1141-1166.

Cochran, J. R., and Stow, D. A. V., 1990. Ocean drilling program.,,116: 1445.

Collett, T., Riedel, M., Cochran, J., Boswell, R., Presley, J., Ku- mar, P.,., 2015. Indian national gas hydrate program expedition 01 report., 1442pp, DOI: org/10.3133/sir2012 5054.

Collins, A., Cutler, I., Friberg, L., and Thompson, S., 2002. Hydrocarbon generation and migration. In:. Robertson Research International Limited, 76: 8630.

Curray, J. R., and Moore, D. G., 1971. Growth of the Bengal deep-sea fan and denudation in the Himalayas., 82: 563-572.

Curray, J. R., and Munasinghe, T., 1989. Timing of intraplate de- formation, northeastern Indian Ocean., 94: 71-77, DOI: 10.1016/0012-821X(89)900 84-8.

Curray, J. R., Emmel, F. J., and Moore, D. G.,2003. The Bengal Fan: Morphology, geometry, stratigraphy, history and processes., 19 (10): 1191-223.

Dewangan, P., Ramprasad, T., Ramana, M. V., Mazumdar, A., Desa, M., and Badesab, F. K., 2010. Seabed morphology and gas venting features in the continental slope region of Krishna- Godavari Basin, Bay of Bengal: Implications in gas-hydrate exploration., 27: 1628-1641, DOI: 10.1016/j.marpetgeo.2010.03.015.

Dewangan, P., Sriram, G., Ramprasad, T., Ramana, M. V., and Jaiswal, P., 2011. Fault system and thermal regime in the vicinity of site NGHP-01-10, Krishna-Godavari Basin, Bay of Bengal., 28: 1899-1914, DOI: 10.1016/j.marpetgeo.2011.03.009.

Duan, X. M., and Huang, Y. X., 2015. Oil and gas exploration potential evaluation of Arakan coastal plain in Myanmar: Taking a block in Myanmar as an example., 42 (17): 57-58.

Emmel, F. J., and Curray, J. R., 1985. Bengal Fan, Indian Ocean. In:. Bouma, A. H.,., eds., Springer, New York, 107-112.

Gani, M. R., and Alam, M. M., 1999. Trench-slope controlled deep-sea clastics in the exposed lower Surma group in the southeastern fold belt of the Bengal Basin, Bangladesh., 127: 221-236, DOI: 10.1016/S0037-0738 (99)00050-0.

Gani, M. R., and Alam, M. M., 2003. Sedimentation and basin- fill history of the Neogeneclastic succession exposed in the southeastern fold belt of the Bengal Basin, Bangladesh: A high- resolution sequence stratigraphic approach., 155: 227-270.

Gibbons, A. D., Zahirovic, S., Muller, R. D., Whittaker, J. M., andYatheesh, V., 2015. A tectonic model reconciling evidence for the collisions between India, Eurasia and intra-oceanic arcs of the central-eastern Tethys., 28 (2): 451-492, DOI: 10.1016/j.gr.2015.01.001.

Hiller, K., and Elahi, M., 1988. Structural growth and hydrocarbon entrapment in the Surma Basin, Bangladesh. In:. Wagner, H. C.,., eds., Circum-Pacific Council for Energy and Mineral Resources, Houston, 657-669.

Hodges, K. V., 2000. Tectonics of the Himalaya and southern Tibet from two perspectives., 112: 324-350, DOI: 10.1130/0016-7606(2000)112< 324:TOTHAS>2.0.CO;2.

Horozal, S., Bahk, J. J., Urgeles, R., Kim, G. Y., Cukur, D., Kim, S. P.,., 2017. Mapping gas hydrate and fluid flow indicators and modeling gas hydrate stability zone (GHSZ) in the Ulleung Basin, East (Japan) Sea: Potential linkage between the occurrence of mass failures and gas hydrate dissociation., 80: 171-191, DOI: 10.1016/j.m arpetgeo.2016.12.001.

Hutchison, C. S., 1989.Clarendon Press, London, 407pp.

Imam, M. B., and Hussain, M. A., 2002. Review of hydrocarbon habitats in Bangladesh., 25 (1): 31-52, DOI: 10.1111/j.1747-5457.2002.tb00098.x.

Jin, J. P., Wang, X. J., Guo, Y. Q., Li, J., Li, Y. P., Zhang, X.,., 2020. Geological controls on the occurrence of recently formed highly concentrated gas hydrate accumulations in the Shenhu area, South China Sea., 116: 104294.

Johnson, S. Y., and Alam, A. M. N., 1991. Sedimentation and tectonics of the Sylhet Trough, Bangladesh., 103: 1513-1527, DOI: 10.1130/001 6-7606(1991)103<1513:SATOTS>2.3.CO;2.

Khan, A. A., 1991. Tectonics of the Bengal Basin., 2: 91-101.

Khin, K., Sakai, T., and Zaw, K., 2014. Neogenesyn-tectonic se- dimentation in the eastern margin of Arakan-Bengal Basins, and its implications on for the Indian-Asian collision in western Myanmar., 26: 89-111, DOI: 10.101 6/j.gr.2013.04.012.

Liang, J. Q., Zhang, W., Lu, J. A., Wei, J. G., Kuang, Z. G., He, Y. L.,., 2019. Geological occurrence and accumulation mechanism of natural gas hydrates in the eastern Qiongdong- nan Basin of the South China Sea: Insights from site GMG S5-W9-2018., 418: 106042.

Lietz, J. K., and Kabir, J., 1982. Prospects and constraints of oil exploration in Bangladesh.. Singapore, 1-6.

Liu, J., Haeckel, M., Rutqvist, J., Wang, S., and Yan, W., 2019. The mechanism of methane gas migration through the gas hy- drate stability zone: Insights from numerical simulations., 124: 4399-4427.

Lohmann, H. H., 1995. On the tectonics of Bangladesh., 62 (140): 29-48.

L?seth, H., Gading, M., and Wensaas, L., 2009. Hydrocarbon leakage interpreted on seismic data., 26: 1304-1319, DOI: 10.1016/j.marpetgeo.2008.09.008.

Lu, Y. T., Luan, X. W., Lyu, F. L., Wang, B., Yang, Z. L., Yang, T. T.,., 2017. Seismic evidence and formation mechanism of gas hydrates in the Zhongjiannan Basin, western margin of the South China Sea., 84: 274- 288, DOI: 10.1016/j.marpetgeo.2017.04.005.

Lu, Y. T., Shi, B. Q., Maselli, V., Luan, X. W., Xu, X. Y., Shao, D. L.,., 2021. Different types of gravity-driven flow deposits and associated bedforms in the upper Bengal Fan, offshore Myanmar., 441: 106609, DOI: 10.101 6/j.margeo.2021.106609.

Luan, X. W., Liu, H., and Peng, X. C., 2011. The geophysical interpretation of a Dongsha ancient uplift on the northern margin of South China Sea., 54 (12): 3217-3232 (in Chinese with English abstract).

Luan, X. W., Peng, X. C., and Qiu, Y., 2009. Tectonic control on the formation of high deposition ratesediment drift in the nor- thern slope of the South China Sea., 23 (2): 183- 199 (in Chinese with English abstract).

Luan, X. W., Zhang, L., and Yue, B. J., 2010. Influence on gas hydrates formation produced by volcanic activity on northern South China Sea slope., 24 (3): 424-432 (in Chinese with English abstract).

Maurin, T., and Rangin, C., 2009. Impact of the 90?E ridge at the Indo-Burmese subduction zone imaged from deep seismic reflection data., 266: 143-155, DOI: 10.1016/j. margeo.2009.07.015.

Molnar, P., and Tapponnier, P., 1975. Cenozoic tectonics of Asia:Effects of a continental collision: Features of recent continental tectonics in Asia can be interpreted as results of the India- Eurasia collision., 189: 419-426.

Moore, D. G., Curray, J. R., Raitt, R. W., and Emmel, F. J., 1974. Stratigraphic-seismic section correlations and implications to Bengal Fan history. In:. von der Borch, C. C.,., eds., U.S. Government Printing Office, Washington, D. C., 403-412.

Murphy, R. W., 1988. Bangladesh enters the oil era., 86 (9): 76-82.

Najman, Y., Nickle, M., BouDagher-Fadel, M., Carter, A., Garzanti, E., Paul, M.,., 2008. The Paleogene record of Himalayan erosion: Bengal Basin, Bangladesh., 273: 1-14, DOI: 10.1016/j.epsl.2008.04. 028.

Qian, J., Wang, X. J., Collett, T. S., Guo, Y. Q., Kang, D. J., and Jin, J. P., 2018. Downhole log evidence for the coexistence of structure II gas hydrate and free gas below the bottom simulating reflector in the South China Sea., 98: 662-674, DOI: org/10.1016/j.marpetgeo.2018.09.024.

Rangin, C., and Sibuet, J. C., 2017. Structure of the northern Bay of Bengal offshore Bangladesh: Evidences from new multi- channel seismic data., 84: 64- 75, DOI: 10.1016/j.marpetgeo.2017.03.020.

Sclater, J. G., and Fisher, R. L., 1974. The evolution of the east central Indian Ocean, with emphasis on the tectonic setting of the Ninetyeast Ridge., 85: 683-702, DOI: 10.1130/0016-7606(1974)85<683:EOTECI >2.0.CO;2.

Shamsuddin, A. H. M., and Abdullah, S. K. M., 1997. Geological evolution of the Bengal Basin and its implication in hydrocarbon exploration in Bangladesh., 69 (2): 93-121.

Sibuet, J. C., Klingelhoefer, F., Huang, Y. P., Yeh, Y. C., Rangin, C., Lee, C. S.,., 2016. Thinned continental crust intruded by volcanics beneath the northern Bay of Bengal., 77: 471-486, DOI: 10.1016/j.marpetgeo. 2016.07.006.

Sloan, E. D., 1998.. Se- cond edition. Marcel Dekker, Inc., New York, 705pp.

Uddin, A., and Lundberg, N., 1998. Unroofing history of the eastern Himalayas and the Indo-Burman Ranges: Heavy mi- neral study of the Cenozoic sediments from the Bengal Basin, Bangladesh., 68 (3): 465-472, DOI: 10. 2110/jsr.68.465.

Uddin, A., and Lundberg, N., 2004. Miocene sedimentation and subsidence during continent-continent collision, Bengal Basin, Bangladesh., 164: 131-146, DOI: 10.10 16/j.sedgeo.2003.09.004.

von der Borch, C. C., Sclater, J. C., Gartner, S., Hekinian, J. R., Johnson, D. A., McGowran, B.,., 1974. Initial reports of the Deep Sea Drilling Project. Deep Sea Drilling Project. Wa- shington, D. C., 890pp.

Wang, X., Zhuo, H., Wang, Y., Mao, P., He, M., Chen, W.,., 2018. Controls of contour currents on intra-canyon mixed se- dimentary processes: Insights from the Pearl River Canyon, northern South China Sea., 406: 193-213.

Yang, S. Y., and Kim, J. W., 2014. Pliocene basin-floor fan se- dimentation on the Bay of Bengal (offshore Northwest Myanmar)., 49: 45-58, DOI: 10.10 16/j.marpetgeo.2013.09.007.

Zhu, G. H., and Li, L. T., 2012. Exploration status and major con- trolling factors of hydrocarbon accumulation in the continental margin basins of the Bengal Bay., 31 (5): 112-118 (in Chinese with Eng- lish abstract).

May 21, 2021;

January 11, 2022;

January 27, 2022

? Ocean University of China, Science Press and Springer-Verlag GmbH Germany 2022

E-mail: luyt_hz@petrochina.com.cn

E-mail: ranweim@163.com

(Edited by Chen Wenwen)