The pore characteristics are studied in the overmatured marine-continental transitional (MCT) shale and simulated shale under different thermal maturity conditions, based on transitional and simulated shale samples in the eastern margin of Ordos Basin. The work uses high-pressure mercury intrusion (MICP), field emission scanning electron microscopy (FESEM), helium-mercury method, X-ray diffraction of whole-rock minerals, and hydrocarbon-generating thermal simulation to quantitatively analyze pore characteristics and main controlling factors of pore development. The results show that the shallow bay and lake facies (SBLF) shale has great exploration potential, while the delta facies (DF) shale has poor exploration potential. The SBLF shale is mainly characterized by pie shale, high quartz and carbonate, low clay, high porosity, and pore volume. The DF shale mainly develops dot shale with low quartz and carbonate content, high clay content, low porosity, and pore volume. Kaolinite has the strongest inhibition on MTC shale pore development. The pore volume of MTC shale decreases first and then increases with maturity. The pie shale is more conducive to the increase of pore volume than the dot shale. The effect of doubled TOC on porosity is greater than that of maturity in the dot shale. The effect of doubled TOC on porosity is less than that of maturity in the pie shale. Organic matter (OM) has the greatest impact on pore development, controlled by the OM content, sedimentary facies, and maturity. OM content, sedimentary facies, and maturity can be used to jointly characterize the MTC shale pore development, providing guidance for multiparameter quantitative characterization of pore development and determining the enrichment area of shale gas.

Transitional shale gas exploration is still in the preliminary stage in China, in which no transitional shale gas field has been discovered. However, large-scale marine shale gas has been exploited in the Lower Silurian Longmaxi and the Lower Cambrian Niutitang formations in China [1, 2]. Moreover, the number of shale gas fields in China is limited, compared with that in North America Song et al. [3]. However, significant progress has been made in the exploration and development of transitional shale gas in China [4, 5]. The transitional shale gas resources reach 0.49 and 2.7 trillion cubic meters, respectively, in the Carboniferous and Permian formations of the Qinshui Basin and Ordos Basin, reflecting the huge exploration potential of transitional shale gas in China Jiang [6].

Shale has complicated mineral composition and pore structure. Many successful quantitative and qualitative methods have been used to characterize the pore structure of MTC shale. Quantitative methods of pore structure include MICP, gas adsorption, small-angle neutron scattering, ultrasmall-angle neutron scattering, and nuclear magnetic resonance [713]. Commonly used qualitative methods include computed tomography, atomic force microscopy, and FESEM [1417]. Many scholars have studied the OM characteristics and pore structure of marine shale [11, 1821]. However, research on the MTC shale is still in its infancy. The studies of MTC shale mainly focus on the OM source, characteristics, and paleoenvironment in recent years [22, 23]. Different scholars have also done some research on the pore structure characteristics. Li et al. [24] studied the MTC shale pore structure in the Taiyuan formation by gas adsorption. Kuang et al. [4] qualitatively described the MTC shale pore classification. Wu et al. [25] and Gu et al. [26] mainly analyzed the mesoporous and microporous characteristics of MTC shale in the different lithofacies. However, previous studies lack the quantitative pore development of MTC shale with different OM spatial distribution and do not quantitatively clarify the main controlling factors of MTC shale pore development.

Shale pore development is controlled by geological factors such as OM characteristics, sedimentary environment, tectonic deformation, mineral composition, and thermal maturity [2, 2730]. The influence of maturity on pore development has received much attention in recent years. The scholars first used SEM to observe the shale pore structure from low-matured to overmatured thermal evolution stage and pointed out that the pore volume would decrease in the matured shale, which was not conducive to shale pore preservation [31, 32]. However, SEM observation cannot quantitatively characterize the pore volume and pore size distribution (PSD). Therefore, scholars continue to use gas adsorption and mercury injection to quantitatively analyze the pore volume and PSD [33, 34]. However, the previous pore structure studies mainly focused on marine shales. The MTC shale is emphasized in the later stage of the shale study. Xi et al. [35] studied the transitional shale samples at different maturity and suggested that maturity plays a major role in controlling the MTC shale pore structure. Wu et al. [36] studied the pore characteristics of immature MTC shale at different maturity by thermal simulation and pointed out that porosity first decreases and then increases with maturity. However, scholars have not paid attention to the influence of OM spatial distribution on the MTC shale pore structure and its relationship with maturity. Two main problems occur: (1) Whether the OM spatial distribution affects the MTC shale pore development with the increase of maturity. (2) Whether the different OM spatial distributions change differently with the increase of OM content.

Therefore, the relationship between OM spatial distribution and OM content still needs to be clarified in the MTC shale pore characteristics. Moreover, the controlling factors of pore development are not quantitatively pointed out in the MTC shale. The work uses gas adsorption, MICP, FESEM, and hydrocarbon-generating thermal simulation to quantitatively characterize shale pore characteristics and discusses the impact of OM characteristics on pore development, guiding quantitative evaluation of shale pore development area and subsequent MTC shale gas exploration and development.

The study area is in Shanxi Province and borders the Yellow River to the west, which lies in the Jinxi flexure belt in the Ordos Basin’s eastern margin with monoclinic strata (Figure 1) [29]. The Ordos Basin is one of the typical craton basins in China. The top of the Ordovician was eroded in the Late Paleozoic period of the Ordos Basin. Therefore, the Upper Carboniferous, Lower Permian Shanxi, and Taiyuan formations were developed in the study area [30, 37]. The Shanxi formation was distributed in a long and narrow belt and characterized by a delta and coastal environment, in which the typical MTC shale was developed and controlled by regional tectonic processes [27]. Multiple mudstones were deposited with a cumulative thickness of 43.5 m-187.5 m in the period [4, 29, 38]. The Shan 1 Member mainly developed lake-delta facies, mostly terrestrial sediments. In contrast, the Shan 2 Member developed delta-coastal facies with shale, coal lines, visible pyrite, siderite, and biological disturbance structures [4].

The focus of the study lies in the Shan-23 submember (Figure 1). The typical MTC shale was divided into the SBLF and DF shales according to sedimentary facies of MTC shale done by Kuang et al. [4]. The field MTC shale was in the Paluogou area (Figure 1). The field MTC shale was also divided into the SBLF and DF shales [4]. The SBLF shale in the lower section developed organic-rich shale containing bioclastics in a semireduced and reduced environment; the DF shale in the upper section developed bio-disturbed structures and coal lines [4].

This study used shale samples of Well Daji 51 collected by the team in the core bank of the Daji area and the field outcrop of the Paluogou area. The 30 black samples were sampled in the target interval of the Shanxi formation (Figure 1). The mineral composition and OM of MTC shale varied greatly in different facies and depths according to the classification of sedimentary rocks by Kuang et al. [4], so 15 samples were taken every half meter to one meter in the DF core section. 7 shale samples were taken every zero to one meter in the SBLF shale owing to the strong heterogeneity of carbonatite. Moreover, the hydrochloric acid and grain size are used to distinguish field samples of SBLF and DF shales according to different mineral compositions.

The MICP, porosity measurement, and FESEM experiments were carried out for shale samples before conducting TOC, Ro, mineral composition, and gas adsorption experiments. The MICP experiment was made by the PoreMaster60 automatic mercury porosimeter with the measurement accuracy of pore size (0.0036 μm-950 μm). The shale porosity was then measured by the QKY-ZN porosity analyzer in a cylindrical core with a height of 3 cm and a diameter of 2.5 cm at a confining pressure of 1000 psi by the helium method. The shale pore structure was then obtained by XL30 FESEM. The sample’s surface (1cm×1cm) was first polished with 600 grit sandpaper into a mirror surface. The sample surface was cut by the LEICA EM TIC 3X three ion beam cutting machine at a three-degree angle. Moreover, LEICA EM ACE 200 coating instrument was used to spray gold on the mirror surface to increase the sample surface’s conductivity. The processed sample was placed on a QUANTA FEG 650 FESM to observe the MTC shale PSD, with an energy dispersive spectrometer to analyze the mineral elements.

15 samples were collected in the SBLF shale at 7 m and 15 samples in the DF shale at 16 m, as fine as possible. However, it was impossible to test at the centimeter level in terms of the research cost owing to the rapidly changeable lithology in the MTC shale. The mineral composition was measured by the X’Pert MPD PRO X-ray diffractometer according to [39]. The shale Ro was determined by a German Leica DM 2500P polarizing microscope. The TOC content was analyzed by the CS230SH organic carbon and sulfur analyzer. Moreover, the shale samples were ground to 40-60 mesh in the gas adsorption. The ISOSORP-HP isothermal measuring adsorption system was used to measure nitrogen adsorption and carbon dioxide adsorption. The pores were divided into macropores (>50 nm), mesopores (2 nm-50 nm), and micropores (<2 nm) according to the pore classification standard. The DFT (density functional theory) model was also employed to calculate the pore volume and SSA of micropores IUPAC (International Union of Pure and Applied Chemistry) [40]. The gas adsorption adopted the Barrett-Joyner-Halenda (BJH) model to determine the mesopore volume and the Brunauer–Emmett–Teller (BET) model to calculate the mesopore SSA [24]. The PSDs of micropores and mesopores were characterized by high accuracy by the DFT model [13].

The kerogen in field shale samples was separated to form the simulated shale in the pore evolution experiment. The 400 g kerogen was extracted from SBLF (2 kg) and DF (2 kg) field shales. The solid sample was filled with 100 g in the hydrocarbon-generating thermal simulator. Moreover, the costs were high in kerogen extraction and 1 g of pure kerogen also produced a large amount of organic carbon Tissot and Durand [41]. Therefore, the quartz powder was chosen to mix with the kerogen because quartz had stable physical and chemical properties with a negligible impact on gas adsorption capacity [42]. The mixing ratios were presented in Table 1 in the simulated shale. Subsequently, the simulated samples were heated to 400°C, 500°C, and 560°C. One-tenth of heated simulated samples were used for the gas adsorption, TOC, and maturity measurement, respectively. The experimental method of the simulated shale is similar to that of the realistic shale.

4.1. Mineral Composition

TOC and minerals have different proportions in the SBLF and DF shales. The average TOC content reaches 6.93% and more than 85% of TOC content exceeds 4.5% in the SBLF shale (Table 2). The average content of quartz and clay is 56.87% and 31.33%, respectively. The carbonatite content is 2%-31% (average=8.93%) and pyrite is also developed with contents ranging from 0% to 5% (Table 2). The feldspar and siderite are hardly developed (Table 2). Moreover, the kaolinite content (average=18.19%) is the highest in the clay minerals (Table 2). In contrast, the average content of illite and chlorite are 8.2% and 4.21%, respectively, and the mixed layer illite/smectiteillite (I/S) only accounts for 0.73% (Table 2). As shown in Table 2, the average TOC value (1.96%) in the DF shale is less than that in the SBLF shale. Moreover, the average quartz content is less than 35%, while the average clay value is the highest at 64% (Table 2). The carbonate and pyrite are almost not developed and the average content of feldspar and siderite is less than 3% in the brittle minerals (Table 2). The kaolinite value ranges from 22.1% to 61.88% (average=34.95%), followed by the average illite content (17.7%) (Table 2). The average chlorite and I/S values only account for 6.44% and 4.91%, respectively (Table 2). In short, the SBLF shale has high OM content and rich brittle minerals and develops carbonatite and pyrite. The DF shale has low OM content and high clay minerals almost without carbonate minerals and pyrite. The content of kaolinite in the SBLF shale is higher than that of illite, chlorite, and I/S in the DF shale.

4.2. OM Spatial Distribution

The kerogen (OM) spatial distribution is heterogeneous in the transitional shale, controlled by the sedimentary characteristics, and tends to concentrate on some points or planes, which are the basis for shale pore development. Furthermore, the physical and chemical changes related to pore evolution are also carried out at these points or planes first [43]. Therefore, the OM spatial distribution is critical for pore development.

The OM spatial distribution is mainly studied by SEM and microscope. The 80 slices are observed in the transitional shale of cores and outcrops. The microscopic characteristics of field MTC shale are similar in those of downhole MTC shale (Figure 2). The results show the SBLF and DF shales have obvious differences in the OM spatial distribution (Figure 2). The OM in the SBLF shale tends to be layered and enriched with a certain sedimentary rhythm (Figures 2(a)–2(c)). The banded OM length varies greatly from 30 μm to 3000 μm. The longer banded OM tends to be larger in the width (Figure 3(a)). In contrast, the DF shale is featured by the dot OM without sedimentary rhythm (Figure 2(e)–2(g)). The small and dispersed OM shapes range from 10 μm to 30 μm (Figure 3(b)). The OM sizes in the SBLF and DF shales were statistically analyzed in Figure 3. The banded OM lengths (30 μm-3000 μm) account for more than 80% of the OM lengths in the SBLF shale, and few banded OM lengths exceed 3000 μm. The OM diameters (<50 μm) account for more than 90%of the OM lengths in the DF shale. However, the OM diameters (<5 μm) cannot be effectively identified and counted owing to the naked eye conditions. Therefore, the data presented in Figure 3 may have some deviation from the realistic data. However, the overall OM distribution results are just like the statistical results.

In short, the OM spatial distribution is different in the SBLF and DF shales. The OM in the SBLF shale is enriched in layers under the microscope, which is similar to the pie in space. When the OM is converted to hydrocarbon, organic pores are formed in the shale (Figure 2(d)). This pie space is a “basic unit” in the transitional shale. The OM in the DF shale is relatively scattered, small, independent from each other, and difficult to connect. Therefore, each OM distribution point is a “basic unit” (Figure 2(h)). Based on the above analysis, the SBLF shale corresponds to the “pie model” (pie shale) and the DF shale corresponds to the “dot model” (dot shale) (Figure 2).

4.3. Pore Characterization

4.3.1. Porosity

The degree of porosity development is different in the SBLF and DF shales. The average porosity is 0.96% in the MTC shale. As shown in Figure 4, the SBLF shale’s porosity is between 0% and 3.5% (average=1.8%). The porosity (<1%) accounts for 17.64%, porosity (1%-1.5%) accounts for 17.65%, porosity (1.5%-2%) reaches 41.18%, and porosity (>2%) is 23.53% (Figure 4). The DF shale has a porosity between 0% and 1.5% (average=0.7%) (Figure 4). The porosity (<0.5%) is 22.22% and porosity (1%-1.5%) is 22.22% (Figure 4). The main porosity (0.5%-1%) accounts for 55.56%. However, no porosity exceeds 1.5% (Figure 4). In short, the average porosity of SBLF shale is greater than that of DF shale.

4.3.2. Genetic Types of Microscopic Pores

The genetic types of micropores are relatively single due to the relatively single lithology and stable sedimentary environment of marine or terrestrial shale [5, 44]. The sedimentary environment changes rapidly in the MTC shale, leading to the complex lithology and genetic types of microscopic pores. The pores are all micro-nanopores in the SBLF and DF shales, including OM pores, intergranular pores, intragranular pores, and microfractures.

The pores developed in the OM include OM pores, clay-containing OM pores, and carbonate and clay-containing OM pores. These pores have different occurrences and sizes. The OM pores are smaller than 30 nm, round, regular, and independent (Figures 5(f) and 6(b)). The carbonate and clay-containing OM pores are irregular in shape and interconnected (6 nm-500 nm) (Figure 5(d)). The clay-containing OM pores develop mesopores (Figure 5(h)). Moreover, the intergranular pores are mainly interlayer pores, developed between clay minerals or between clay and OM (Figures 5(g) and 6(c)). Almost all the intergranular pores related to OM show slit-like openings along the grain boundary. Similar slit-like openings in large clay particles may also be formed due to drying shrinkage during the experiments’ preparation [45]. However, similar slit-shaped pores do not exist under high pressure in the MTC shale. Moreover, it cannot be the diagenetic fractures because the transformation from illite to montmorillonite forms the fractures to be mainly curved and partly connected [46].

The intragranular pores are mainly distributed in the carbonatite and pyrite. The carbonatite usually exists in the dissolved pores with irregular shapes and different sizes (Figure 5(e)); the pyrite mainly develops the triangular intercrystalline pores (Figures 5(h) and 6(d)). Slatt and O'Brien [47] believed the card house structure between the rigid mineral particles prevents pore from being destroyed by particle extrusion. The microfractures are mainly developed in the carbonate particles (Figure 5(d)), which are beneficial to free gas discharge in the shale [48, 49].

4.3.3. Gas Adsorption and MICP

The nitrogen (N2) and carbon dioxide (CO2) adsorption-desorption (AD) curves are used to determine the type of pore structure, the pore volume, and the SSA of mesopores and micropores (Figure 7; Table 3). The N2 AD curves are similar in the SBLF and DF shales, which are identical to the IV adsorption isotherm curve of IUPAC (International Union of Pure and Applied Chemistry) [40]. The CO2 AD curves are similar in the SBLF and DF shales. They are identical to the type I adsorption isotherm curve, meaning the presence of many micropores. When the relative pressure (P/P0) is less than 0.5 in the N2 adsorption, the gas is mainly adsorbed in a single gas layer; when the P/P0 is 0.5, the nitrogen curve has a turning point and the gas begins to adsorb in multiple layers; when the P/P0 exceeds 0.5, the adsorption curve is mainly slightly convex and the adsorption capacity gradually increases, meaning the presence of mesopores and macropores; when the pressure continues to rise at the end, the adsorption has not reached saturation, meaning the presence of large pores [50]. The CO2 AD curves also indicate the pore characteristics. The CO2 has not yet reached equilibrium when the P/P0 reaches 0.01 in the CO2 adsorption, indicating the existence of many mesopores. When the relative pressure is low, the adsorption capacity increases at a higher rate, whereas the increase rate gets slower under high pressure, meaning the existence of many microporous and mesoporous structures. Moreover, the N2 adsorption is performed under the cryogenic liquid conditions, leading to the gap AD curve because of capillary condensation. However, CO2 adsorption is carried out at a normal temperature, leading to the overlapped AD curves because of no capillary condensation.

When the AD curve has obvious hysteresis loops, the desorption curve falls quickly at a relative pressure (0.4-0.5), similar to the H2 type of IUPAC, corresponding to the ink bottle-shaped holes with large pores and narrow pore throats, most of which are intergranular pores. The AD curve with a certain width corresponds to the parallel plate-shaped pores or cylindrical pores with open ends, similar to the H3 type, mostly corresponding to pores and microfractures between layers of clay minerals. The minimal AD curve width corresponds to the slit-shaped pores with small apertures and poor connectivity, similar to the H2 curve. The AD curves of the SBLF shale have a certain width when P/P0 is smaller than 0.5, corresponding to the ink bottle-shaped holes, interlayer pores, and narrow-shaped pores with openings at both ends, which are related to the interlayer pores, microfractures, and OM pores with good connectivity (Figure 7). When the P/P0 is lower than 0.5 in the DF shale, the AD curves overlap, corresponding to the ink bottle-shaped hole and the slit-shaped hole with one end open, which is related to the poorly connected OM pores, the intragranular pores, and intergranular pores (Figure 7).

As shown in Figure 8 and Table 3, the MICP is used to determine the pore type, the pore volume, and the SSA in the macropore. The macropore volume is 0.0006 mL/g-0.0047 mL/g (average=0.0024 mL/g) in the MTC shale, almost equivalent to the micropore volume (Table 3). However, the average SSA is only 0.021 m2/g (Table 3). Moreover, the dominant pore size is featured with the mesopores and macropores (3 nm-100 nm) and microfractures (>30 μm) (Figure 8). The mercury injection volume is large at 3 nm-100 nm, whereas the mercury injection volume is small at 0.1 μm-30 μm (Figure 8) because the pores are mainly mesoporous and microporous in the SBLF and DF shales. Some developed microfractures lead to the increase of mercury injection volume at 30 μm-80 μm. The fracture development is determined by brittle minerals. The mercury injection volume in the SBLF shale is greater than that of the DF shale at more than 50 nm (Figure 8) because the macropore volume in the SBLF shale is higher than that in the DF shale, as determined by the pore development of SBLF and DF shales (Figures 5 and 6). As shown in Figure 9, the DF shale is mainly composed of the micropores and mesopores, so the mercury injection volume is large at 3 nm-50 nm and some developed microfractures lead to the increase of mercury injection volume at more than 30 μm. Overall, the mercury injection capacity in the SBLF shale is more than that in the DF shale. The pore development under the FESM also confirms that the porosity in the SBLF shale is higher than that in the DF shale (Figures 5 and 6).

4.4. Simulated Shale Characteristics with Different Maturity

The dot and pie kerogens were extracted from the SBLF and DF field shales, respectively, to form the simulated dot and pie shales. The more OM is decomposed to generate the TOC in the simulated pie shales. As is shown in Figure 10 and Table 4, the TOCs in the simulated dot shales decrease from 4.1% to 1.2% and from 8.2% to 2.4% with the increased thermal simulation temperature. The TOCs in the simulated pie shales decline even more from 4.1% to 1.1% and from 8.2% to 1.9% with the increased simulated temperature. Moreover, when the TOC is 4.1% in the simulated dot shale, the micropore volume decreases from 0.00101 mL/g to 0.00072 mL/g and then increases to 0.00120 mL/g with the increased temperature (Figure 10). The SSA declined from 1.1 m2/g to 0.7 m2/g and then increased to 4.23 m2/g. The mesopores show similar changes in the SSA (3.89 m2/g-1.2 m2/g-2.16 m2/g) and volume (0.01322 mL/g-0.00403 mL/g-0.00532 mL/g) (Table 4). Furthermore, when TOC values double in the simulated dot shale, the volumes and SSAs of micropore and mesopore also show a similar trend (Figure 10; Table 4). In contrast, the micropore volume also declines from 0.00039 mL/g to 0.00027 mL/g and then increases to 0.00052 mL/g when the TOC is 4.1% in the simulated pie shale (Figure 10). The SSA decreased from 0.52 m2/g to 0.22 m2/g and increased to 2.03 m2/g. The mesopores also present similar changes in the volume (0.00125 mL/g-0.00032 mL/g-0.00074 mL/g) and SSA (0.35 m2/g-0.11 m2/g-0.28 m2/g) (Table 4). However, the volumes and SSAs of mesopores and micropores in the simulated dot shale are smaller than those in the simulated pie shale. When TOC content doubles, a similar trend is presented in the pore volume (Figure 10).

5.1. The Overall Pore Characteristics of MTC Shale

The MTC shale is characterized by the micropores and mesopores as the main storage space, accounting for 79.83% of the total pore volume and 99.82% of the total SSA, respectively, (Table 3 and Figure 9). The micropores, mesopores, and macropores account for 20.36%, 59.47%, and 20.17% of the total pore volume (Table 3). The average SSA of the micropores and mesopores accounts for 77.51% and 22.32%, macropores’ SSA accounting only for 1.7% (Table 3). The average microporous, mesoporous, and macroporous volumes (0.0041 mL/g, 0.0078 mL/g, and 0.004 mL/g) in the SBLF shale are larger than those in the DF shale (0.0014 mL/g, 0.0066 mL/g, and 0.0014 mL/g) (Table 3). The MTC shale’s total pore volume ranges from 0.0046 mL/g to 0.023 mL/g (average=0.0119 mL/g) (Table 3). The total SSA is in the range of 2.04 m2/g to 25.07 m2/g (average=11.99 m2/g) (Table 3). Moreover, the micropore volume has a positive correlation with its SSA. However, the volumes of the total pore, mesopores, and macropores have a relatively poor correlation with their SSAs (Figure 11), which may be related to the large aperture in the shale [51, 52].

The PSD and porosities are different in the transitional shale. The shale mesopores and micropores’ PSDs are presented in Figure 12. The microporous and mesoporous main peaks are at 0.4-0.7 nm and 3-5 nm in the DF and SBLF shales (Figure 12). Moreover, the PSD in the SBLF shale is better than that in the DF shale, meaning many developed isolated pores in the DF shale. The average pore volumes are 0.0159 mL/g and 0.0094 mL/g, respectively, in the SBLF and DF shales (Table 3). The average pore sizes are 4.57 nm and 3.71 nm in the SBLF and DF shales (Table 3). This shows that the SBLF shale has larger pore volume, pore connectivity, and pore size than the DF shale. Furthermore, the PSD is good in the marine or terrestrial shale in which the main peaks are characterized by the mesopores and macropores [5355]. However, the main peaks are dominated by micropores in the MTC shale (Figure 12). Moreover, the porosities range from 3% to 15% and from 1% to 6%, respectively, in the marine and continental shales [7, 52, 5658]. The average porosity of MTC shale is smaller than that of marine or terrestrial shale.

The OM pores account for more than 60% of the transitional shale pores, regarded as the main transitional shale pores (Tables 3 and 4). The OM pore proportion is obtained by dividing the TOC homogenized volume of simulated shale by the average TOC homogenized volume of transitional shale in the transitional shale pores. It is worth noting that the quartz pore volume of simulated shale is ignored in this calculation process because the small micropore and mesoporous volumes can be ignored in the quartz [42]. The OM microporous and mesoporous volumes account for 78.13% and 63.33% of the DF shale microporous and mesoporous pores, respectively, in the simulated dot shale. The OM microporous and mesoporous volumes account for 87.54% and 81.41% of the SBLF shale microporous and mesoporous pores, respectively, in the simulated pie shale. It can be deduced that the pie OM contributes greatly to the pore volume of SBLF shale.

In short, the pore size and mesoporous and macroporous volumes in the marine and terrestrial shale are larger than those in the MTC shale, although the pore volumes and SSAs are similar in the marine, terrestrial, and MTC shales [56, 59]. However, the SBLF shale has better pore characteristics with great exploration potential. The DF shale has a relatively poor pore structure with low exploration potential. Moreover, the contribution ratio of pie OM to SBLF shale pore volume is higher than that of dot OM to DF shale pore volume.

5.2. Controlling Factors of Pore Development

The simulated shale has eliminated the impact of clay minerals and other brittle minerals on organic pores owing to its composition (quartz and kerogen), which would not be elaborated on the following quantitative analysis. This work primarily quantitatively studies the impact of organic matter content, OM spatial distribution, and maturity on pore volume and SSA.

5.2.1. TOC and OM Spatial Distribution

The simulated shale is used to quantitatively analyze the impact of OM spatial distribution on pore development. When the simulated dot shale reaches the overmatured thermal stage (560°C) and its TOC content doubles, the microporous and mesoporous volumes increase by 45% and 44%, respectively (Table 4). In contrast, when the TOC content doubles in the simulated overmatured pie shale, the microporous and mesoporous volumes increase by 75% and 66%, respectively (Table 4). The following result excludes the impact of maturity and OM spatial distribution on OM pore volume in the shale with the same type and similar maturity. The higher OM content is more likely to produce micropores and mesopores because more hydrocarbon is discharged from the OM (Table 4). Moreover, the increased pore volume in the simulated pie shale is higher than that in the simulated dot shale when the maturity increases by the same degree at the same TOC content in the simulated dot and pie shales (Table 4). This indicates that the pie OM is more favorable for discharging more hydrocarbons to increase the pore volume. Su et al. [43] also suggested that the dot OM is not conducive to hydrocarbon expulsion to increase the pore volume, while the pie OM is easy to expulse hydrocarbons to form OM pores.

The TOC content in the SBLF shale is higher than that in the DF shale in the realistic transitional shale. The TOC content is positively correlated with the volumes of the micropores, mesopores, macropores, and total pores in the SBLF and DF shales (Figure 13). Moreover, when the TOC value is less than 4%, the slope is large in the linear fitting equation between TOC and micropores, mesopores, and macropores. When the TOC value is more than 4%, the slope decreases. Furthermore, the OM is carbonized and converted into graphite because the OM has reached the late maturity stage in the study area. The pore structure has collapsed to a certain extent owing to the confining pressure. The degree of increased porosity decreases as the TOC content increases [11, 18]. However, the study does not conduct corresponding simulation experiments to verify the effect of higher TOC content on porosity in the simulated shale owing to the huge consumption of kerogen samples. More samples will be collected in the subsequent work to deepen the relevant TOC experiment.

The OM pore development has connections with the OM spatial distribution in the MTC shale. The OM in the SBLF shale is mainly distributed in the bands with few dots. The OM associated with carbonate is easy to dissolve to form macropores and mesopores in the SBLF shale. Moreover, the connectivity is good in the pores. Simultaneously, the carbonate particles are prone to cracks due to stress (Figure 5(d)) and are easily dissolved to form macropores and mesopores (Figure 5(c)). The banded OM coexists with pyrite and clay minerals in the SBLF shale. The intergranular pores, developed by pyrite and transformed between clay minerals, increase the development of mesopores and macropores and the different sizable pore connectivity in the OM (Figure 5(h)). In contrast, the OM pores (6 nm-30 nm) are independent of each other with poor pore connectivity in the DF shale (Figure 5(f)). The OM is mainly distributed in the dots (Figure 5(a)). The particles’ overall orientation is relatively strong with small carbonate particles and quartz particles owing to the relatively strong hydrodynamic force of the delta sedimentary environment. The dot OM mesopores are sporadically and poorly developed (Figure 6(b)). There are some interlayer shrinkage pores in clay, carbonate dissolved pores, and pyrite intergranular pores (Figures 6(c) and 6(d)). It can be deducted that the OM pores are developed in the SBLF shale, intergranular pores are relatively poorly developed, and the different sizable pores are well connected. Moreover, the developed microfractures are conducive to the discharge of free natural gas. However, the DF shale has poor OM pores, intergranular pores, and microfractures. The pore and microfractures in the SBLF shale are more developed and connected than those in the DF shale, which is a favorable exploration facies for the MTC shale gas. However, whether kerogen orientation affects OM hydrocarbon expulsion cannot be verified by the simulated shale experiment because of the loose structure of the simulated shale itself, although kerogen in the realistic shale has orientation under SEM. This scientific problem would be solved in the subsequent studies.

In short, the banded OM is more favorable to the pore development than the dot OM in the simulated and realistic MTC shales. The increased TOC also increases the pore volume in the quantitative and qualitative analysis results.

5.2.2. Maturity

The pore volume decreases first and then increases with maturity in the simulated shale (Figure 10). This is consistent with the results of transitional shale simulation experiments done by Wu et al. [36]. The two parameters (Mj and Nj) are also introduced to quantitatively represent the pore volume with the increased maturity. The specific formula is as follows:
(1)Mj=VojVmjVoj,(2)Nj=VnjVmjVnj,
where Mj is the reduced ratio of pore volume, Nj represents the increased ratio of pore volume, Voj represents the pore volume of simulated shale when the maturity ranges from 0.81 to 0.83, Vmj represents the pore volume of simulated shale when the maturity ranges from 1 to 1.2, Vnj represents the pore volume of simulated shale when the maturity ranges from 2.5 to 2.62, j=1 represents micropore, j=2 represents mesopore, and j=3 represents the sum of mesoporous and microporous volumes.

When the TOC content is 4.1% in the simulated dot shale, the N1 and M1 are 47.52% and 28.71% in the micropore volume, the N2 and M2 are 9.76% and 69.52% in the mesopore volume, and the N3 is 37% in the total pore volume (Table 5 and Figure 14). When the TOC content doubles, the N1, N2, and N3 are 66.67%, 32.5%, and 37.71%, and the M1 and M2 are 33.92% and 69.7% (Table 5). The impact of organic matter content and OM spatial distribution on pore volume was excluded by fixing the TOC content and OM spatial distribution. When the fixed TOC is 4.1%, the M1 and M2 decrease (Table 5). This is consistent with the SEM observations [36] because some oil discharged from early source rocks preferentially blocks mesoporous pores. However, the simulated shale cannot be observed by SEM. The work adopts the previous experimental results done by Wu et al. [36] because the massive shale with a fixed structure is required for the SEM observation. Moreover, the N1 and N2 increase. However, the N1 is larger than the N2 because the dispersed simulated dot shale has little hydrocarbon expulsion to form the micropores (Table 5; Figure 14). In contrast, when the OM content doubles in the simulated dot shale, the M1 and M2 further decrease owing to the increase in hydrocarbon production. Moreover, the increased N1 (19.15%) is less than the increased N2 (22.74%) because the increased hydrocarbon expulsion is more conducive to the expansion of mesopore volumes (Table 5; Figure 14). Furthermore, the similar N3 indicates that the pore development of the dot OM is less affected by the TOC content because the dispersed dot OM in the minerals leads to the difficult hydrocarbon expulsion owing to the less contact with other minerals in which larger pores cannot be formed [60, 61].

When the fixed TOC is 4.1% in the simulated pie shale, the N1, N2, and N3 are 92.59%, 131.25%, and 113.56%, and the M1 and M2 are 30.77% and 74.4% (Table 5 and Figure 14). When TOC content doubles, the N1 and M1 are 136.36% and 38.89% in the micropore volume, and the N2 and M2 are 184.88% and 79.81% in the mesopore volume (Table 5 and Figure 14). The N3 is 161.96% in the total pore volume (Table 5). The M1 and M2 in the simulated pie shale are larger than those in the simulated dot shale because the pie shale is easy to expulse hydrocarbons to decrease the pore volume (Table 5). The increased N1 (43.77%) is also less than the increased N2 (53.63%). However, the N3 in the simulated pie shale is far higher than that in the simulated dot shale because the higher pie OM content is more favorable for the formation of micropores and mesopores [60, 61].

In short, maturity has little effect on the N3 in the simulated dot shale (37%-38%) (Table 4; Table 5). However, the N3 is primarily influenced by the TOC content in the simulated pie shale, reaching 114%-162% (Tables 4 and 5). It can be deduced that the impact of doubled TOC on porosity is greater than that of maturity in the simulated dot shale (Table 4). However, the impact of doubled TOC on porosity is less than that of maturity in the simulated pie shale (Table 4). However, the work does not consider that the pore structure will collapse when the maturity exceeds 3.5 [33], which should be focused on the later gas exploration. Furthermore, this work only analyzed the effect of maturity on organic pores, not on inorganic pores. However, organic pores are dominant in transitional shale pores, so quantitative work is also valuable.

5.2.3. Other Minerals

This work mainly focuses on the quantitative OM pore development because the MTC shale pores are mainly featured by organic pores. The influence of brittle minerals and clay minerals on pore development is analyzed by the qualitative method. However, brittle minerals and clay minerals certainly have a quantitative influence on the shale pores, which will be further studied in the following study.

The content of brittle minerals is closely related to hydraulic fracturing. The quartz content in the SBLF shale is much greater than that in DF (Table 1). The quartz content has a negative correlation with the TOC content in the SBLF and DF shales (Figure 13). The quartz and TOC are positively correlated in marine shale, indicating that the quartz is of biogenic origin Wang et al. [52]. However, the quartz has a negative correlation with TOC, indicating that the quartz is derived from territorial sources [62]. It can be concluded that the quartz in DF and SBLF shales are derived from territorial sources.

The porosity is positively correlated with the TOC in the SBLF and DF shales. Nevertheless, the porosity of DF shale is less than that of SBLF shale at the same TOC value (Figure 13). It shows that the clay and brittle minerals besides TOC have a great impact on the shale porosity. The TOC-normalized (TN) pore volume does not correlate with the quartz content (Figure 15). The quartz from terrestrial sources has a low contribution to the porosity. Moreover, the carbonatite and pyrite are developed in the MTC shale with the developed and connected pores in the MTC shale. The carbonatite and pyrite are positively correlated with the porosity and TN total pore volume (Figure 15), contributing greatly to pore development.

The clay minerals are negatively correlated with the porosity and TN total pore volume (Figure 15). The clay minerals have a positive correlation with the TN micropore volume in the DF shale (Figure 15). However, the clay minerals have a negative trend with the TN macropore volume in the SBLF shale (Figure 15). The content of chlorite and I/S has no correlation with the TN total pore volume in the clay minerals, while illite has a positive correlation with the TN total pore volume (Figure 16). The kaolinite content is negatively correlated with the TN total pore volume (Figure 16). Moreover, kaolinite has the highest content in the clay minerals (Figure 16; Table 2). It can be deduced that kaolinite inhibits pore development. However, illite contributes to pore development to a certain extent. Furthermore, the high kaolinite content means that the acidic water of the digenesis is conducive to the conversion of other clay minerals into the kaolinite [63, 64]. The humic OM with the low phytoplankton and algae aerobic organisms leads to a high degree of hypoxia and acidity in the water, leading to the higher kaolinite content [28, 41, 65, 66]. Moreover, the clay minerals in marine and continental shales promote pore development in which the illite contributes to pore development [67]. Furthermore, the higher kaolinite content in the DF shale largely inhabits the pore development because the hydrodynamic alternation is higher with high oxygen content in the DF shale. Interestingly, the kaolinite does not exist in the marine shale, because its kerogen is mainly provided by the plankton and algae. The alkaline water causes the kaolinite to be easily transformed into other clay minerals owing to many aerobic organisms [2, 37, 68]. The kaolinite content is lower than 10% in the continental shale because lower algae content is lower [67, 69].

5.3. Pore Development Area of MTC Shale Reservoir

The two key insights can be obtained by analyzing the pore characteristics of simulated and realistic shales: (1) SBLF shale mainly develops pie shale and DF shale primarily develops dot shale. The SBLF shale has good exploration potential, with high TOC content, maturity, pore volume, and PSD. In contrast, the DF shale has less exploration potential. (2) The impact of doubled TOC on porosity is greater than that of maturity in the simulated dot shale. The impact of doubled TOC on porosity is less than that of maturity in the simulated pie shale. When OM content doubles, the pore volumes increase by more than 44% and 69%, respectively, in the simulated overmatured dot and pie shales.

The two insights are vital to our exploration. The OM has the greatest influence on pore development in the MTC shale, mainly controlled by TOC, sedimentary facies, and maturity. The maturity distribution is clarified in the eastern margin of Ordos Basin, as proposed by Kuang et al. [4]. The maturity increases from NE to SW in the range of 0.8%-2.8%. Furthermore, the maturity ranges from 0.8% to 1.2% from Baode to Xingxian to Jiaxian, in which the pore volume decreases and the transitional shale is not at the peak of hydrocarbon generation. The maturity ranges from 1.2% (Liulin) to 2.8% (Yichuan), where the transitional shale is at the peak of hydrocarbon generation with the larger pore volume. The maturity exceeds 2.8% in the fuxian area. In short, it is likely to find the best area for shale pore development according to the maturity distribution. However, a discrepancy occurs in the predictable and actual geological conditions. The prediction does not consider the distribution of sedimentary facies and TOC. The combination of seismic facies and well logging first needs to be intensified to characterize the distribution of SBLF and DF facies. The TOC distribution can then be described by the relationship between logging, seismic, and TOC.

The simulated shale experiment also has academic value. The quantitative analysis of simulated shale can pave the way for multiparameter quantitative characterization of pore development areas and determining the most favorable shale gas enrichment areas. The work describes the quantitative impacts of TOC and maturity on shale pore development in different sedimentary facies. At present, the study of simulated shale is in the preliminary stage of exploration. The work studies the quantitative increased ratio of pore volume in the different simulated shales with the different TOC (1%-6%) from the early stage of high maturity to the overmaturity stage, which has a certain applicable value. Moreover, differences exist in realistic and simulated shales. The OM in transitional shale has orientation, while the simulated shale does not have orientation. Does the OM orientation have an impact on pore development? The pore development of transitional shale is also affected by carbonate, pyrite, and clay minerals. What is the quantitative proportion of these effects? At present, these problems have not been solved by simulated shale. However, OM pores in the simulated shale are the main pores of transitional shale, controlled by OM content, maturity, and OM spatial distribution, which increase the OM pore volume by more than 37%. The simulated shale study is also important from the perspective. This work also suggests that the OM spatial distribution has an important effect on the pore development of the MTC shale.

This paper clarifies the controlling factors of the pore development area; quantitatively analyzes the influence of OM content, maturity, and OM spatial distribution on pore development; and guides multiparameter quantitative characterization of pore development area through the analysis of pore characteristics of simulated and realistic shales.

  • (1)

    The transitional shale is dominated by organic pores, which account for more than 60%. The average pore size and porosity in the SBLF shale are larger than those in the DF shale. The micropores, mesopores, and macropores in the SBLF shale are more developed than those in the DF shale. The pyrite and carbonatite contribute significantly to pore development. The kaolinite has the strongest inhibition on MTC pore development

  • (2)

    The SBLF shale has great exploration potential, while the DF shale has poor exploration potential. The SBLF shale is mainly characterized by the pie shale with high quartz and carbonate, low clay, high porosity, and pore volume. The DF shale mainly develops dot shale with low quartz and carbonate content, high clay content, low porosity, and pore volume

  • (3)

    The main controlling factors of OM pore development include OM content, maturity, and OM spatial distribution. The pore volume decreases firstly and then increases with the increase of maturity in the MTC shale. The simulated pie shale is more conducive to the increase of pore volume than the simulated dot shale. The effect of doubled TOC on porosity is greater than that of maturity in the simulated dot shale. The effect of doubled TOC on porosity is less than that of maturity in the simulated pie shale

The next work is to perform the pore quantification of the fine maturity change and the quantitative characterization of shale pore development by clay minerals and brittle minerals. The quantitative multiparameter combination is then used to characterize shale pore volume.

     
  • MTC:

    Marine-terrestrial transitional

  •  
  • SBLF:

    Shallow bay and lacustrine facies

  •  
  • DF:

    Delta facies

  •  
  • TOC:

    Total organic carbon

  •  
  • OM:

    Organic matter

  •  
  • SSA:

    Specific surface area

  •  
  • PSD:

    Pore size distribution

  •  
  • FESEM:

    Field emission scanning electron microscopy

  •  
  • DFT:

    Density functional theory

  •  
  • AD:

    Adsorption-desorption

  •  
  • TN:

    TOC-normalized.

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy restrictions.

We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work, and there is no professional or other personal interest of any nature or kind in any product, service, and/or company that could be construed as influencing the position presented in, or the review of, the manuscript entitled “Controlling factors and quantitative characterization of pore development in marine-continental transitional shale”.

Kun Xu did the writing and original draft preparation, data curation methodology, software, and experiment. Shijia Chen did the conceptualization. Ziqiang Tao did the data processing. Jungang Lu performed the investigation. Qingbo He acquired the software and did the validation. Chen Li did the writing and reviewing and editing.

We would like to express our gratitude to the State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation and the Sichuan Natural Gas Geology Key Laboratories for providing the experimental analysis support. This study was funded by the Science and Technology Cooperation Project of the CNPC-SWPU Innovation Alliance (Nos. 2020CX030000 and 2020CX050000), National Natural Science Foundation of China (Nos. 41872165, 42072185, and 42002176), and Research and Innovation Fund of Southwest Petroleum University (2021CXYB57).

Exclusive Licensee GeoScienceWorld. Distributed under a Creative Commons Attribution License (CC BY 4.0).