Pore Preservation and Failure Mechanism of Sinian Dengying Formation Carbonate Reservoirs: A Case Study of Two Ultradeep Wells in the Sichuan Basin, Western China

Despite being one of the most important factors in deep oil and gas exploration, the preservation mechanisms of ultradeep carbonate reservoirs remain poorly understood. This study performed thin-section, geochemistry, field emission scanning electron microscopy, fluid inclusion, and basin model analysis of samples from two boreholes over 8,000m deep in the Sichuan Basin to determine the pore features and preservation mechanism of the Sinian (Ediacaran) Dengying Formation carbonate reservoirs. The reservoir of CS well #1 is characterised by pore diameters larger than a centimetre (average porosity 7.48%; permeability 0.8562mD), and the pores are mainly filled with dolomite or bitumen. In contrast, the reservoir of MS well #1 is predominantly composed of micron-scale residual pores (average porosity 1.74%; permeability 0.0072mD), and the pores are typically filled with dolomite, bitumen, and multistage quartz. The burial thermal histories suggest that both reservoirs were subjected to high pressure (i.e., pressure coefficient > 1:5) before the Late Cretaceous. However, the pressure coefficient of the reservoir of MS well #1 has decreased to less than 1.0 owing to strong structural adjustment this well since the Late Cretaceous, which allowed other ore-forming fluids to enter and fill the pores, resulting in further compaction of the pores. In contrast, the pressure coefficient of CS well #1 is 1.1–1.2, which effectively prevented other ore-forming fluids from entering and filling the pores. The findings show that the dynamic adjustment of the Dengying Formation palaeo-gas reservoir indirectly affects the preservation or failure of the reservoir. The occurrence and geometry of bitumen in the Dengying reservoir exhibit good consistency with the pressure changes in both boreholes. In particular, bitumen with an annular shape and contraction joints in reservoir pores is widespread in CS well #1, which is attributed to the continuous preservation of palaeo-gas fields. Conversely, bitumen with a broken particle shape is located among the epigenetic minerals widespread in MS well #1, which is attributed to failure and depletion of the palaeo-gas fields.

For ultradeep exploration, the most important factor is a high-quality reservoir; this is because of the substantial reduction in the porosity and permeability of rocks with depth [10]. Moreover, strata below 4,000-5,000 m exhibit heterogeneous porosity [10][11][12]. As a result, the preservation and destruction mechanisms of ultradeep carbonate reservoirs, for example, burial dissolution associated with hydrocarbon generation [13,14], are poorly known [15][16][17]. The acidic fluid in carbonate reservoirs formed by H 2 S and CO 2 can lead to thermochemical sulphate reduction (TSR), which can improve or preserve reservoir quality [13,14,18,19]. Furthermore, deep hydrothermal fluids related to normal fault activity can facilitate the formation of hydrothermal dolomite reservoirs, for example, in North America and the Middle East [20,21]. However, some researchers argue that deep to ultradeep reservoirs are characterised by a closed system of fluid flow (e.g., pore water flow) and geochemistry, indicating that the total porosity remains unchanged [15,22,23]. It should be noted that acidic or hydrothermal fluids will also lead to a large amount of cementation [15,23,24]. Therefore, research and discussion on the preservation mechanism of ultradeep carbonate reservoir have theoretical and practical significance.
The Sichuan Basin, located in the South China continental block, is one of the largest sedimentary basins in China and is rich in petroleum resources ( Figure 1). A major hydrocarbon deposit in the Sinian Dengying Formation was discovered in the Anyue gas field at a depth of approximately 6,000 m, with a volume of 4,403:8 × 10 8 m 3 (15.6 tcf) [25,26]. The nearby PT well #1, with a depth of~7,000 m, records production of 130 × 10 4 m 3 per day. The Dengying Formation is a typical ultradeep reservoir (Figure 1) [27][28][29] that provides an excellent case study for analysing the pore preservation and failure mechanisms in ultradeep high-quality dolomite reservoirs. However, there is lack of study on ultradeep carbonate reservoirs (depth > 8000 m) in current research, which makes the preservation mechanism of carbonate reservoirs in ultradeep extreme conditions remains unknown. Recently, two boreholes were drilled in the northern Sichuan Basin to target the Dengying Formation, that is, MS well #1 and CS well #1, with depths of 8,044 m and 8,146 m, respectively. In this study, we investigate the Sinian Dengying carbonate reservoirs using samples from these two boreholes, as well as analyses based on thin sections, geochemistry, field emission scanning electron microscopy, fluid inclusions, basin models, etc. The findings suggest that (1) the dynamic adjustment of palaeo-gas controls the preservation or failure of ultradeep carbonate reservoirs, and (2) the occurrence and geometry of bitumen in the Dengying reservoir show good consistency with pressure changes in both boreholes.

Geological Setting
The Sichuan Basin is located along the northwest margin of the South China Block, which is a part of the Precambrian Rodinia supercontinent [30,31]. The Sinian/Ediacaran sedimentary rocks represent the early sedimentary history of the

Geofluids
Sichuan Basin, of which the Denying Formation comprises shallow-water dolomitic carbonates. Because of the Tongwan tectonic movement during the Late Edicarian, the Dengying dolomites were exposed to freshwater karstification, resulting in widespread development of caverns and dissolution porosity, especially along unconformities on the top of the Dengying Formation ( Figure 1) [19,32]. These dolomites became crucial for petroleum accumulation in the Proterozoic Dengying Formation (Figure 1), as demonstrated by widespread bitumen and proven gas reserves of 408:61 × 108 m 3 (1.5 tcf) and 4,403:8 × 108 m 3 (15.6 tcf) in the Weiyuan and Anyue gas fields, respectively [25]. During the Mesozoic to Cenozoic periods, this area experienced several phases of tectonic movement, including the Caledonian, Indosinian, and Yashanian movements [33,34]. The Caledonian tectonic phase in this region resulted in erosion of most of the upper Silurian, Devonian, and Carboniferous strata across the South China Block. East-west striking Leshan-Longnvshi palaeo-uplift formed in the basin centre during the Caledonian movement. Palaeo-uplift had an important influence on hydrocarbon accumulation during the Triassic to Jurassic across the Sichuan Basin [35][36][37]. The Indosinian and Yashanian movements resulted in the formation of complex foreland basins from the Late Triassic to Cretaceous, such as the western Sichuan and northern Sichuan basins [38,39]. Both CS well #1 and MS well #1 are located in the northern Sichuan Basin (Figure 1(a)). The total drilling depth of CS well #1 is 8,448 m, with 302 m cutting of the Dengying Formation, whereas that of MS well #1 is 8,418 m, with 374 m cutting of the Dengying Formation. In general, the Dengying Formation can be divided into four members from bottom to top: Deng-1 (Z 2 dn 1 ), Deng-2 (Z 2 dn 2 ), Deng-3 (Z 2 dn 3 ), and Deng-4 (Z 2 dn 4 ) (Figure 1(b)) based on lithology and algal abundance [27,[40][41][42]. Despite the lack of large framework biota, the main building groups of the biological community in the Z 2 dn 2 and Z 2 dn 4 members are microorganisms, bacteria, and algae [26,43]. The Z 2 dn 1 member is characterised by micrite dolomite, whereas the Z 2 dn 3 member, which is deposited in semi-open shore facies, is characterised by bluish-grey argillaceous siltstone and black shale intercalated with thin-layer dolomite [7,44]. In contrast, the Z 2 dn 2 and Z 2 dn 4 members are mainly composed of algal microbial dolomite (Figure 1(b)). Algal dolomites with botryoidal and stromatolite dominate Z 2 dn 2 , whereas light-grey to white massive algal dolomite and silt-crystal dolomite dominate Z 2 dn 4 [45,46]. 3.2. Methods. High-pressure mercury injection (HPMI) can provide excellent porosity, permeability, median pressure, migration pressure, skewness coefficient, and maximum mercury saturation estimates [3]. HPMI was performed on nine samples (i.e., six samples in CS well #1 and three samples in MS well #1) using the AutoPore 9510 IV mercury injection apparatus at the State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation. Prior to HPMI, the dolomite samples were cleaned with ethanol and dried under vacuum at 150°C for at least 12 h. The maximum injection pressure was 340.528 MPa.

Samples and Experimental Methods
FESEM-focused ion beam analysis was performed at the State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation using the field emission scanning electron microscope Quanta 250 FEG (US FEI) and energy spectrum analyser Oxford InCAX-MAX20 (UK Oxford) on 17 samples. First, the samples were prepared in a square column of approximately 10 × 10 × 5 mm, and the surface was coated with a thin gold layer using an EM SCD500 ion sputtering device (Leica, Wetzlar, Germany). Second, the minerals were observed and analysed using a Quanta 250 FEG scanning electron microscope at 23°C and 35% humidity according to the National GBBT 17361-2013 national standard. Third, the observed minerals were analysed using an Oxford InCAX-MAX20 energy spectrum analyser for qualitative or semiquantitative microregion compositions.
Fluid inclusion observations can provide extensive information on the diagenetic mineralisation and infilling conditions of palaeo-fluids [47][48][49], particularly the palaeopressure of reservoirs and gas fields [50][51][52][53]. In this study, 12 thin sections were observed. The homogenisation and ice temperatures of the fluid inclusions were measured on a linkam-TH600 hot and cold platform. The homogenisation temperatures of the two-phase saline inclusions (containing hydrocarbons) were also determined. The initial temperature rise rate of the uniform temperature measurement was 10°C/min, which was then reduced to 1°C/min; the accuracy of the uniform temperature measurement was ±1°C.  The cores of MS well #1 are characterised by needle tiplike micropores (Figure 3(b)), rather than centimetre-or millimetre-scale pores. In the thin section, there are dark irregular filamentous or bubble-like aggregates, along with a large number of destroyed and transformed algal microbial lattice pores (Figure 3(e)). In general, the latticed holes are strongly damaged and transformed, and the original pore structure is difficult to identify. However, the morphology of the early developed millimetre-scale corrosion holes before partial filling or destruction is still visible, as shown in Figures 3(c) and 3(d). The maximum original pore size is approximately 0:9 × 1:7 mm, and the pores are fully filled by multiple minerals.      (Figure 6(a)). Most pores are filled with bitumen (Figure 2(d)). Dry cracks and postthermal cracking shrinkage cracks of the bitumen are clearly observed (Figure 2(e) and Figures 4(a), 4(c), and 4(f)), indicating that the bitumen was retained in situ after thermal cracking. Overall, the multistage mineral filling sequence of the CS well #1 Z 2 dn 4 member pores is as follows: firstgeneration recrystallised fine dolomite, second-generation bitumen, and/or third-generation dolomite secondary outgrowth filling (Figure 2(e) and Figures 4(a) and 4(b)).

Results
The pores of the MS well #1 Z 2 dn 4 member are strongly reformed, and the primary pore structure is substantially reformed. The filling minerals in the pores are predominantly silicified dolomite, bitumen, recrystallised dolomite, and authigenic quartz. Some early pore structures are also observed (Figures 3(c) and 3(d) and Figure 5(a)), indicating high-quality primary pores in MS well #1. As shown in Figures 5(a)-5(c), authigenic quartz with three-stage growth is observed in ultralarge solution pores, with a crystal diameter of approximately 2.3 mm and a regular crystal shape. Furthermore, multistage growth of bitumen and quartz is also found in the pores (Figure 3(c)). This suggests that the original pore space is large enough to allow selfcrystallisation and precipitation of fluid minerals in later stages. The characteristics of the growth filling sequence of multistage quartz in the pores are as follows: first-stage dentate authigenic quartz, second-stage secondary quartz, and third-stage quartz enlargement edge (Figures 3(c) and  3(d)). The contact boundary between some authigenic quartz particles and other minerals (such as bitumen or    (Figure 7(a)). The average median radius of the pore throat of the CS well #1 Z 2 dn 4 member sample is 0.33-0.86 μm, with a median pressure of 1.01-6.55 MPa (Table 1). Almost no platform section is observed in the mercury injection curve; however, the mercury intake rate is good, with a maximum exceeding 90% (Figure 7(a)). Among the pore throats, 64.08% have a radius of 0.16-1.6 μm, indicating a large porethroat size range and high connectivity (Figure 7(b)). The average pore-throat radius of the MS well #1 Z 2 dn 4 member sample is 0.03-0.09 μm, and the median pressure is 23.61-44.65 MPa ( Table 1). The maximum mercury saturation is only 55% (Figure 7(a)), and the median pressure is 44.25 MPa. The pore-throat radius range of 0-0.1 μm accounts for 83.83% of all pore throats, indicating a micropore-throat size range and low connectivity (Figure 7(b)). In general, CS well #1 exhibits a better average median radius and lower median pressure and more connectivity than MS well #1.

Fluid Inclusions.
Three types of fluid inclusions are observed in the samples: gas-liquid two-phase inclusions, pure liquid inclusions, and pure gas inclusions, with gas-liquid two-phase inclusions being dominant. Gas and gas-liquid two-phase inclusions are found in local mineral crystals, indicating that they were trapped at the same time. Inclusions are banded or grouped and distributed in recrystallised dolomite, secondary outgrowth of dolomite, and authigenic quartz (Figures 6(a) and 6(c)), with elliptical (round), strip, approximately square, and irregular shapes. In general, the inclusions exhibit a small gas/liquid ratio (less than 20%), and the inclusion diameter is typically 10-25 μm (Figures 6(b)-6(f)). It should be noted that hydrocarbon inclusions can be subdivided into liquid-oil inclusions and gas-phase methane inclusions according to their phase states and compositions. Liquid-oil inclusions are mainly distributed in the secondary outgrowth of dolomite and appear dark brown and yellow brown under transmission single polarised light (Figures 6(b) and 6(d)). Gas-phase methane inclusions occur predominantly in dolomite secondary outgrowths and authigenic quartz. Under a polarising microscope, they appear mostly grey in colour, with a bright line observed in the middle. These inclusions are generally 9-20 μm in size and often star-scattered. These inclusions are mostly symbiotic with brine inclusions (Figures 6(e) and 6(f)).
The trapping pressure of fluid inclusions in Z 2 dn 4 can be calculated by the functional relationship between the homogenisation temperature of the inclusions, freezing point tem-perature, and salinity [50][51][52][53]. The palaeo-fluid pressure coefficient (Table 2) can be further calculated by combining the palaeo-burial depth (see following section for details). The palaeo-pressure coefficient of the CS well #1 Z 2 dn 4 member is 1.10-1.46, with a median of 1.1-1.2, indicating weak overpressure in the Z 2 dn 4 member. The palaeo-pressure coefficient of MS well #1 Z 2 dn 4 member formation was 0.88-1.43, with a median of 0.9-1.1, indicating normal pressure.

Burial Thermal History and Palaeo-Pressure of the Dengying Formation
BasinMod-1D software was used to reconstruct the burial and thermal histories of CS well #1 and MS well #1 (Figure 9) by integrating the stratigraphic thickness, lithology, absolute age, measured borehole temperature, and vitrinite reflectance ( Table 3). The total organic content (TOC) in the core samples was further measured using a LECO CS-200 analyser.  6.04%, respectively. The modelling process was calibrated using the thermal maturity of the formation and borehole temperatures, combined with the Easy% Ro method [54] and the NULL chemical kinetic model [55]. According to the burial histories of CS well #1 and MS well #1 (Figure 9), the palaeo-pressure coefficient was further calculated using the fluid inclusion data, and the palaeo-pressure evolution characteristics of the two wells were obtained by combining the data of fluid trapping times to reconstruct key events ( Figure 10).
The burial depth thermal history of CS well #1 can be divided into three stages (Figures 9(a) and 10(a)).
(1) From the early Permian to early Triassic palaeoreservoir formation period, the overlying strata of the Dengying Formation gradually thickened, the Lower Cambrian source rocks entered the oil generation stage, crude oil migrated to the reservoir through the top unconformity surface of the Dengying Formation (Figure 10(a)), and the homogenisation temperature of the reservoir trapping inclusions ranged from 113.4°C to 123.7°C. Furthermore, the palaeo-pressure coefficient was 1.1-1.2, which indicates normal pressure and weak overpressure dry gas, with uplift erosion since the Late Cretaceous. The homogenisation temperature of reservoir trapping inclusions ranged from 201.5°C to 220.1°C, and the palaeo-pressure coefficient was 1.1-1.2, denoting weak pressure adjustment and release characteristics The thermal history process and accumulation evolution of the MS well #1 Z 2 dn 4 member can also be divided into three periods (Figures 9(b) and 10(b)).
(1) The Late Silurian to the Middle Permian saw the formation and overpressure enrichment stage of the palaeo-oil reservoir. With the opening of the hydrocarbon generation threshold and the entry of crude oil into the pores, the homogenisation temperature range of fluid inclusions trapped in the Dengying Formation reservoir was 104.5-143.6°C, and the pressure coefficient was 1.20-1.29, which is characteristic of weak overpressure (Figure 10(a)). From the Late Triassic to Late Cretaceous, with continued deepening of burial strata, the palaeo-burial depth of the Dengying Formation increased to more than 6,000 m, and crude oil in the palaeo-oil reservoir began to crack to form natural gas. The inclusion temperature range was 153.0-177.1°C, and the abnormal pore pressure increased rapidly. The pressure coefficient reached 1.49, indicating an overpressurised system (Figure 10     14 Geofluids ( Figure 11); this resulted in better original porosity in the early reservoirs of CS well #1 and MS well #1. In particular, the unconformity at the top of the CS well #1 in the Dengying Formation at a depth of 8,146 m resulted in the development of solution pores. Therefore, early meteoric dissolution in the upper Yangtze platform caused the formation of many karst cavities in both MS well #1 and CS well #1 Z 2 dn 4 members, resulting in high-quality porosity ( Figure 11).

Geofluids
With the deep burial of dolomite, compaction and early cementation strongly reduced the porosity of the dolomite reservoir. Late Silurian-Triassic crude oil from the Lower Cambrian source rocks migrated into the reservoir pores of the Dengying Formation [29,56]. The Permian period was affected by the Emei movement, with hydrothermal events occurring in the dolomites of the Dengying Formation in the Sichuan Basin, resulting in slight modification of the secondary solution pores [57,58]. Since the Jurassic, as the burial depth and geothermal temperature increased, the crude oil in the pores of the Dengying Formation cracked into gas, forming a palaeo-gas reservoir. To date, the pores of the two drilling wells in the Dengying Formation dolomite reservoirs experienced similar diagenesis. Until the Late Cretaceous, as a result of the Himalayan movement in the Sichuan Basin, both wells of the Dengying Formation experienced changes in the dynamic preservation of oil and gas reservoirs, which indirectly led to differences in their reservoir pore preservation conditions.   Figure 11: Diagenetic sequence and porosity development in the Sinian Z 2 dn 4 member during the formation of dolomite reservoirs. Note that differential porosity evolution occurred in the Cenozoic between MS well #1 and CS well #1. 15 Geofluids and scattered in or between the crystals of various pore epigenetic minerals (Figures 3(c) and 3(d) and Figures 5(a)-5(d)). It should be noted that the bitumen particles are embedded in the crystal boundary with distinct ruptures (Figures 5(b), 5(d), 5(e), 5(g), and 5(j)).

Bitumen Mechanics of Bitumen with Different
Geometry. CS well #1 and MS well #1 underwent deep burial processes, resulting in widespread pyrolysis of hydrocarbon substances in the reservoir to form carbonaceous bitumen. Therefore, we suggest that two genetic processes led to the  Figure 12: Differential evolution of pores in the Z 2 dn 4 member in the two wells (modified from [25]). Annular-shaped bitumen with contraction joints in pores of the Z 2 dn 4 reservoir is widespread in CS well #1, which is attributed to continuous preservation of the CS well #1 palaeo-gas reservoir. In contrast, broken granular bitumen is scattered in pores with authigenic quartz in MS well #1, indicating significant destruction and dissipation of the palaeo-gas field in MS well #1. 16 Geofluids formation of different bitumen geometries in MS well #1. (1) Tectonic movement resulted in deformation of the host strata and bitumen formation in pores, as well as the aggregation of broken bitumen particles at the bottom of pores.
(2) Late fluid intrusion into the pores of the reservoir resulted in the breakage of bitumen on the cave wall, accompanied by the precipitation and growth of minerals in the fluid. Thus, the bitumen particles are either incorporated into the new mineral crystals or gradually pushed to the intercrystalline phase during mineral growth. According to our thin-section and FESEM observations, we prefer the latter process to account for the observed geometries of bitumen in the CS well #1 and MS well #1.

Correlation between Bitumen Geometry and Pore
Evolution. The two wells experienced the same events: crude oil filling and thermal cracking into gas from the Triassic and Early Jurassic-Early Cretaceous. Since the Late Cretaceous, the palaeo-pressure and related preservation conditions of the CS well #1 and MS well #1 Z 2 dn 4 member were different. CS well #1 is located on the low and gentle northern slopes of central Sichuan. Owing to the lack of strong deformation, the early oil and gas reservoirs exhibited good preservation conditions with overpressure. This effectively prevented other fluids from entering the pores and is consistent with the annular bitumen shape and shrinkage joints in CS well #1 (Figure 2(e) and Figures 4(a), 4(c), and 4(f)). Thus, we infer that the crude oil was not disturbed or transformed by other fluids after thermal cracking into carbonaceous bitumen (Figure 12(c)). Conversely, MS well #1 is located in the foreland of the Micang Mountain, dominated by strong deformation and rapid erosion since the Late Cretaceous, which destroyed the preservation and depletion of gas in the MS well #1 Dengying Formation. The depleted natural gas provided space for other fluids in the pores. Large-scale fluid flowed rapidly into the pores (the homogenisation temperature of inclusions was 247.6-282°C), resulting in the disturbance and fragmentation of bitumen in the pores (Figure 12(d)). Overall, the fluid minerals in the pores were filled, grown, and enclosed in broken bitumen particles. At the same time, with the growth of mineral crystals, bitumen particles were pushed to the crystal boundary ( Figures 5(a)-5(f) and Figures 12(e)).
The palaeo-pressure coefficient of CS well #1 was 1.1-1.5, indicating overpressure in the Dengying Formation ( Figure 10(b)). In the Cenozoic, the palaeo-pressure coefficient remained at approximately 1.2, indicating that this abnormal pore pressure could still support the overlying strata and maintain the pores (Figure 2(e) and Figure 12(c)). However, the abnormal pore pressure of MS well #1 decreased rapidly from a pressure coefficient of 1.5 in the Late Cretaceous, to a weak normal pressure of approximately 1.0 in the Cenozoic. Under such a weak normal pressure state, compaction of the overlying strata would affect the reservoir pores, resulting in densification of the reservoir (Figure 3(f) and Figures 5(d)-5(f)). Therefore, compaction during the late-stage diagenesis of MS well #1 led to extrusion fracture of the boundary between the crys-tals of some epigenetic minerals in the pores (Figures 5(d)-5(g)), with even crushed quartz particles becoming embedded in the bitumen (Figure 5(j) and Figure 12(f)).
In summary, the Z 2 dn 4 member reservoirs sampled in CS well #1 and MS well #1 experienced similar diagenesis prior to the Late Cretaceous, in the form of microbial mat construction, dolomitisation, karst weathering, crude oil filling, and dissolution during the burial period. However, the diagenesis stages of the two wells differed in the natural gas preservation stage ( Figure 11). The long-term effective preservation of the CS well #1 natural gas reservoir resulted in intact pores (Figure 12(c)). Since the Late Cretaceous, MS well #1 experienced rapid gas depletion with water entering the pores and mineralisation, resulting in pore densification (Figure 12(e)). As a result, the palaeo-pressure decreased rapidly with the depletion of natural gas, which led to further compaction of pores in the reservoirs (Figure 12(f)).

Conclusions
(1) The carbonate reservoirs in the Z 2 dn 4 member of CS well #1 and MS well #1 are dominated by microbial algal dolomite. The Z 2 dn 4 reservoir of MS well #1 exhibits more complex pore diagenesis and poorer reservoir physical properties than that of CS well #1 (2) Dynamic adjustment of the Dengying Formation palaeo-gas reservoir indirectly affects the preservation or failure of the reservoir. The palaeo-gas and pore normal pressure (1.1-1.2) of the CS well #1 Dengying Formation were preserved, which effectively prevented other ore-forming fluids from entering and filling the pores. Conversely, the palaeo-gas of MS well #1 was destroyed and depleted, which allowed other ore-forming fluids to enter and fill the pores; thus, the abnormal pore pressure decreased rapidly from a pressure coefficient of 1.5 to a weak normal pressure (slightly less than 1.0), resulting in further compaction of the pores (3) The pore morphology and geometry of reservoir bitumen can be used as effective indicators of the level of preservation or destruction of the reservoir pores. Annular-shaped bitumen with contraction joints in reservoir pores is widespread in CS well #1, which is consistent with the continuous preservation of palaeo-gas fields. In contrast, broken particles of bitumen scattered among the epigenetic minerals are widespread in MS well #1, which is consistent with multistage minerals and a reduced pressure coefficient in the reservoirs, indicating failure and depletion of palaeo-gas fields

Data Availability
State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation, Chengdu University of Technology, Chengdu, Sichuan 610059, China.