The TOC contents of the 16 samples are listed in Table 1. Samples from Dalong Formation in Well XY1 mainly contained Type-Ⅱ OM with TOC ranging from 2.72% to 6.79%. Samples from Longtan Formation in Well XY1 contained mainly Type-Ⅲ OM with TOC ranging from 1.72% to 10.29% (Bao et al., 2016), TOC content of the coal sample was 38.41%. Samples from Dalong Formation in Well EJ1 contained Type-Ⅱ OM with TOC ranging from 5.52% to 11.25% (Qiu et al., 2019). The TOC of samples close to the coal seam increased in Longtan Formation (Fig. 2). The thermal maturity results showed that the Ro of Well XY1 samples ranged from 1.22% to 1.43%, which was assigned to early generation stage of the condensate oil and wet gas. The Ro of Well EJ1 samples ranged from 2.62% to 2.97%, which was assigned to the over-mature stage and generation of dry gas.
Sample Depth (m) Formation Lithofacies TOC (%) Ro (%) Kerogen type Mineral composition (wt.%) Quartz Feldspar Dolomite Calcite Clay Pyrite Other XY-1 600.22 Dalong Black calcareous shale 3.69 1.34 Ⅱ 60 6 2 10 20 2 XY-2 602.67 Dalong 2.72 / Ⅱ 63 3 2 23 8 1 XY-3 608.08 Dalong 4.96 1.22 Ⅱ 45 7 4 19 23 2 XY-4 612.76 Dalong 5.09 1.37 Ⅱ 49 7 1 22 18 3 XY-5 624.87 Dalong 5.07 / Ⅱ 33 7 2 19 36 3 XY-6 629.59 Dalong 3.13 / Ⅱ 55 4 6 20 13 2 XY-7 631.18 Dalong 6.79 / Ⅱ 49 5 / 10 33 3 XY-8 641.78 Dalong 3.00 1.27 Ⅱ 29 8 / 38 21 4 XY-9 685.07 Longtan Grey-black silt mudstone 1.72 / Ⅲ 53 16 / / 31 / XY-10 689.10 Longtan Black-mudstone 7.17 / Ⅲ 14 5 / 2 57 / 22 (Siderite) XY-11 692.66 Longtan Black mudstone 10.29 1.41 Ⅲ 32 3 5 / 60 / XY-12 697.81 Longtan Coal 38.41 1.43 Ⅲ 40 / / / 48 12 EJ-1 827.87 Dalong Black shale 5.52 2.62 Ⅱ 12 30 / 40 14 2 2 (Gypsum) EJ-2 845.00 Dalong 6.14 2.84 Ⅱ 70 7 1 8 11 3 EJ-3 850.75 Dalong 11.25 2.97 Ⅱ 76 7 3 9 4 1 EJ-4 856.46 Dalong 7.62 / Ⅱ 52 10 1 8 26 3
Table 1. Basic properties of sixteen samples used in this work. TOC. total organic carbon (%), /. not available
The mineral compositions of Dalong Formation in the wells XY1 and EJ1s mainly consisted of quartz, carbonate minerals, feldspar, and clay minerals. The clay fraction was the dominant mineral species for Longtan Formation in Well XY1, with almost no carbonate minerals present (Table 1).
According to the classification of isotherms proposed by IUPAC (Thommes et al., 2015), the CO2 adsorption isotherms of all 16 samples showed Type-Ⅰ curve characteristics (Figs. 3a-3c). Dalong Formation samples in Well XY1 displayed the least amount of adsorption, while Longtan Formation samples and Dalong Formation samples in Well EJ1 showed a much higher gas adsorption capacity, indicating that these samples had more micropores. Micropore volumes calculated from CO2 adsorption are listed in Table 2. Micropore volumes of Dalong Formation and Longtan Formation in Well XY1 were in the range of 0.002-0.005 and 0.004-0.010 cm3/g, respectively, while Dalong Formation in Well EJ1 had the highest micropore volumes ranging from 0.007 to 0.011 cm3/g.
Figure 3. CO2 adsorption (a)-(c) and N2 adsorption and desorption isotherms (d)-(f) of Longtan and Dalong samples determined in wells XY1 and EJ1.
Sample ID Formation TOC (wt.%) DFT micropore volume (cm3/g) BJH mesopore volume (cm3/g) BJH macropore volume (cm3/g) Hg porosity (%) XY-1 Dalong 3.69 0.004 0.003 0.004 / XY-2 Dalong 2.72 0.003 0.003 0.004 / XY-3 Dalong 4.96 0.005 0.004 0.005 0.53 XY-4 Dalong 5.09 0.004 0.003 0.007 0.31 XY-5 Dalong 5.07 0.003 0.002 0.003 0.18 XY-6 Dalong 3.13 0.002 0.003 0.005 0.26 XY-7 Dalong 6.79 0.004 0.003 0.004 / XY-8 Dalong 3.00 0.003 0.004 0.004 / XY-9 Longtan 1.72 0.004 0.007 0.007 1.21 XY-10 Longtan 7.17 0.010 0.012 0.004 / XY-11 Longtan 10.29 0.009 0.013 0.006 1.53 XY-12 Longtan 38.41 0.010 0.002 0.004 / EJ-1 Dalong 5.52 0.007 0.013 0.008 0.85 EJ-2 Dalong 6.14 0.008 0.013 0.006 2.12 EJ-3 Dalong 11.25 0.011 0.019 0.013 5.42 EJ-4 Dalong 7.62 0.010 0.016 0.007 1.63
Table 2. N2, CO2 adsorption and MICP results for Longtan and Dalong shales samples. DFT. density functional theory; BJH. Barrett Joyner Halenda method; micropore volume is obtained from CO2 adsorption by DFT; mesopore and macropore volume are obtained from N2 adsorption by DJH; Hg porosity is calculated by MICP
Figures. 3d-3f show the N2 sorption-desorption isotherms of all 16 samples. According to the classification of isotherms proposed by IUPAC (Thommes et al., 2015), the isotherms of all 16 samples were Type-Ⅱ and Type-IV(a), indicating multi-layer adsorption. Dalong Formation samples in Well EJ1 had the highest amount of adsorption, which was approximately five times higher than that of Dalong Formation samples in Well XY1 and twice as much as that of Longtan Formation samples. From this we inferred that there were more pores present in Dalong Formation of Well EJ1. According to the classification of hysteresis loops proposed by IUPAC (Thommes et al., 2015), the hysteresis loops of Longtan samples in Well XY1 and Dalong samples in Well EJ1 were the mixture of H2(b) and H3 types, while the hysteresis loops of Dalong samples in Well XY1 were assigned to H4 (Fig. 3). All isotherms showed a hysteresis loop, indicating that capillary condensation had occurred in the mesopores. It was obvious that the opening of the hysteresis loops of Dalong samples in Well XY1 were narrow, reflecting relatively small pore size (Fig. 3d). The mesopore and macropore volumes were the highest in Dalong Formation samples from Well EJ1 and lowest in Dalong Formation samples from Well XY1 (Fig. 4).
Figure 4. The micropore, mesopore, macropore volume for Longtan and Dalong samples from wells XY1 and EJ1 (calculated from CO2 and N2 adsorption data).
According to the classification of isotherms proposed by IUPAC (Thommes et al., 2015), the absolute CH4 adsorption isotherms of all the samples show the characteristics of Type I curve (Fig. 5) with varying Langmuir adsorption capacity.
Figure 6 shows the curves of the mercury intrusion and extrusion at pressures ranging from 0.034 to 413 MPa. The mercury intrusion volumes of Dalong Formation samples from Well EJ1 were the highest (6.65-8.84 μL/g), which was approximately four times higher than that of Dalong Formation samples from Well XY1 (1.23-2.16 μL/g), and approximately twice as much as that of Longtan Formation samples (4.68-5.65 μL/g). As the mercury intrusion pressure ranged from 0.1 to 1 MPa, the XY-11 sample from Longtan Formation in Well XY1 took a larger volume of mercury than the other samples from Dalong Formation in both wells EJ1 and XY1s (Fig. 6). Therefore, it could be inferred that Longtan sample had more micron-sized pore throats, whereas Dalong sample had more nanoscale pore throats. When intrusion pressure reached 100 MPa, the volumes of mercury injected into Dalong samples of the EJ1 increased significantly (Fig. 6c), reflecting the good connectivity of the nanoscale pore throats. In contrast, the curves determined for Dalong samples from Well XY1 showed the least amount of injected mercury (Fig. 6a), indicating the worst pore development and connectivity. According to the distribution of pore throat histogram, 3-20 nm size pores account for a large proportion of the total pore volumes in Dalong Formation (Figs. 7a, 7b, 7e, 7f), whereas 20-200 nm size pores accounted for a large proportion of the total pore volumes in the Longtan Formation (Figs. 7 b, 7c). Porosity of Dalong samples in Well EJ1 (0.85%-5.41%) was much higher than that of Dalong samples in Well XY1 (0.18%-0.53%), and the porosity development of Longtan samples was between (1.21%-1.53%).
Figure 6. Cumulative mercury intrusion and extrusion curves (a)-(c) for Longtan and Dalong samples from wells XY1 and EJ1.
Many oil inclusions with blue fluorescence were observed in the calcite veins developed in Dalong shales of Well XY1 (Figs. 8b, 8d). Under transmission light, oil inclusions were densely distributed and transparent where the coating wall was thin and clear (Figs. 8a, 8c). Gas inclusions were relatively dark under transmission light and sparsely distributed (Figs. 8a, 8e). Under UV light, a certain proportion of gas inclusions was covered with a thin layer of liquid hydrocarbon inside and emitted weak yellow-green fluorescence (Figs. 8b, 8d). A small number of pure gas inclusions did not emit fluorescence (Figs. 8e, 8f). The laser Raman spectrogram for the gas inclusions in Well XY1 showed a significant CH4 peak at wavelength 2 909.473 9 cm-1, indicating this was a pure CH4 inclusion (Fig. 9a). Except for the CH4 peak, some gas inclusions also presented weak C2H6 peaks (2 879.657 3 and 2 944.351 2 cm-1) on both sides of the CH4 peak (Fig. 9b), indicating that the gas inside was a mixture of gaseous hydrocarbons (CH4 and C2H6). A large number of experimental analyses have confirmed that fluorescence color can be used as an indicator of OM maturity (George et al., 2001; Stasiuk and Snowdon, 1997; Videtich and Roger, 1988). The evolution of OM from low mature to high mature will lead to a change of inclusion type from heavy oil inclusion to condensate oil inclusion. Meanwhile, the fluorescence color emitted will change consistently from red, to yellow, to yellow-green, and to blue (Zhao et al., 2019; Zhao and Chen, 2008). Based on the observation that all oil inclusions show blue fluorescence accompanied by a small amount of gas inclusion, we inferred that there was light oil generated and expelled from the adjacent Dalong Formation shales, which was accompanied by a small quantity of gas due to gas-prone maceral degradation in the kerogen. The characteristics of the hydrocarbon inclusions were also consistent with the measured Ro (1.22%-1.34%) of Dalong samples in Well XY1 (Table 1) which was in the early stage of the condensate oil and wet gas window.
Figure 8. Microscopic characteristics of fluid inclusions in calcite veins. (a), (b) The number of oil inclusions is predominated with blue fluorescence, Well XY1, 602 m; (c), (d) oil inclusions with blue fluorescence and gas inclusions with weak yellow-green fluorescence, Well XY1, 602 m; (e), (f) pure gas inclusions with no fluorescence, Well XY1, 602 m; (g), (h) CH4 inclusions Well XY1, 850.75 m.
All hydrocarbon inclusions observed in calcite veins from Well EJ1 were gas inclusions without fluorescence (Figs. 8g, 8h). The laser Raman spectrogram for all gas inclusions showed a characteristic CH4 peak only (Fig. 9c). The measured Ro of the adjacent Dalong shales ranged between 2.62% and 2.97%, which was in the late stage of the dry gas window. It was inferred that the CH4 inclusions were the product of secondary cracking from the adjacent Dalong shales.
Based on observations of FE-SEM images, the interior of certain OM from Dalong shales of Well XY1 was not homogeneous and could be divided into dark and bright parts with distinct boundaries (Figs. 10a-10d). It is noticeable that spot-like and irregular pores were prone to develop in the light part, while dark part rarely developed observable pores. In addition, the boundary of bright part was irregular and serrated, whereas that of dark part was a regular and smooth curve (Figs. 10a-10d). Some OM partly filled the internal body cavity of the pyritized fossil (Fig. 10g). Many clay mineral flakes were entirely wrapped by OM (Figs. 10e, 10f). A small amount of OM which occurred as discrete particles with sharply defined edges and a relatively homogeneous internal texture can be observed within the mineral matrix (Figs. 10h, 10i).
Figure 10. FE-SEM images of OM from Dalong Formation in Well XY1. (a) OM divided into nonporous part (bitumen) and porous part (kerogen) with mesopores, Dalong Formation, Well XY1, 608.08 m; (b) bitumen having spot-like micropores and kerogen having irregular mesopores, Dalong Formation, Well XY1, 608.08 m; (c) OM divided into nonporous part (bitumen) and porous part (kerogen) with micropores, Well XY1, 624.87 m; (d) OM divided into nonporous part (bitumen) and porous part (kerogen) with micropores, Well XY1, 624.87 m; (e) kerogen enclosing plenty of clay mineral flakes, Well XY1, 608.08 m; (f) an enlargement of the rectangle area marked in Fig. 10e, showing a small amount of spot-like micropores, Well XY1, 624.87 m; (g) bitumen filling part of internal body cavity space of pyritized fossil, Well XY1, 608.08 m; (h) terrestrial woody debris with rounded shape and gas-prone kerogen with few micropores, Dalong Formation, XY1, 624.87 m; (i) an enlargement of the rectangle area marked in Fig. 10h, showing filling of bitumen in macropores in terrestrial woody debris (white arrow), Well XY1, 624.87 m.
Milliken et al. (2013) found a similar phenomenon in Pennsylvania Marcellus shales and interpreted this as wood fragments. Loucks et al. (2017) observed similar OM which was micron-size and nonporous from Yanchang Formation in the Ordos Basin, and interpreted this to be Type Ⅲ woody maceral. In this study, we considered that this kind of OM was terrestrial woody debris derived from the original plant tissue. There was a distinct boundary between two different OM (Figs. 10h, 10i).
The morphology of OM in Longtan shales of Well XY1 was significantly different from that in Dalong shales. As shown in Fig. 11, the OM in Longtan shales of Well XY1 had a distinct texture with a long strip shape. The boundary of the OM was smooth and regular. Due to compaction, some OM were highly curved and obviously oriented (Figs. 11e, 11f). The pore development in different OM varied considerably. Some OM were nonporous, while the adjacent one may possess a large number of pores, which were filled with quartz and illite (Figs. 11a-11d). Liu et al. (2017) found that vitrinite was nonporous, whereas the interior of inertinite generally had a certain number of circular cellular pores, some of which were filled with authigenic quartz. The sedimentary environment of Longtan Formation was lagoon-swamp facies and almost all OM was from terrestrial inputs. Sample XY-11 was very close to the coal seam at the base of Longtan Formation. Therefore, it was likely that the OM in Longtan Formation samples were all terrestrial woody debris.
Figure 11. FE-SEM images of organic matter from Well XY1, Longtan Formation. (a) Terrestrial woody debris with rounded pores, Well XY1, 692.66 m; (b) terrestrial woody debris with rounded pores and some nonporous kerogen debris, Well XY1, 692.66 m; (c) abundant rounded and nano-scale pores, Well XY1, 692.66 m; (d) curved and distorted kerogen debris and long-strip kerogen debris, Well XY1, 692.66 m; (e) an enlargement of the rectangle area marked in Fig. 11d, showing curved and distorted kerogen debris with rounded pores, Well XY1, 692.66 m; (f) kerogen debris with rounded pores, part of which are filled by quartz, Well XY1, 692.66 m.
The morphology of the OM in Dalong samples of Well EJ1 all showed the characteristics of filling interparticle pores and appeared pervasive throughout the mineral matrix (Fig. 12). Moreover, they had an amorphous shape as well as significant flow structure (Figs. 12e, 12h, 12i). Some OM enclosed mineral grains and clay flakes (Figs. 12c-12e). There were a small number of terrestrial woody debris whose morphology was substantially different from that of the adjacent amorphous OM (Figs. 12g, 12h). There were some differences between the OM in Dalong shales in the two wells. Unlike the OM in Well XY1, which had a granular texture and develop only a small number of pores inside, the interior of the OM in Well EJ1 was almost filled with sponge-like nanopores (Figs. 12a-12c, 12f). As Dalong shales from the two wells had the same type of OM, thermal maturity may have been the determining factor that controled the morphology and inner structure of the OM.
Figure 12. FE-SEM images of OM in Well EJ1 Dalong Formation. (a) Pyrobitumen filling pore space within mineral grains, Well EJ1, 827.87 m; (b) an enlargement of the rectangle area marked in Fig. 11a, showing plenty of alveolate, spongy and irregular pores, Well EJ1, 827.87 m; (c) pyrobitumen enclosing clay mineral flakes with abundant nano-scale pores, Dalong Formation, Well EJ1, 856.46 m; (d) pyrobitumen enclosing mineral grains with abundant nano-scale organic pores, Dalong Formation, Well EJ1, 856.46 m; (e) pyrobitumen with flow structure, Well EJ1, 856.46 m; (f) pyrobitumen filling pore space within mineral grains with bubble-like, alveolate and spongy organic pores, 856.46 m; (g) terrestrial woody debris with long-strip shape, Well EJ1, 856.46 m; (h) terrestrial woody debris with rounded pores and pyrobitumen with flow structure, 856.46 m; (i) an enlargement of the rectangle area marked in Fig. 11h, showing distinct boundaries of pyrobitumen and terrestrial woody debris, 856.46 m.
FE-SEM images showed OM pores in shale samples from different formations were different in terms of shape, size, and distribution density (Figs. 10-12). The OM pores in Dalong shales of Well XY1 were mainly micropores, sparsely distributed in the OM (Fig. 10). The morphology of the micropores was spot-like, while most mesopores had a irregular shape. A small number of round macropores could be observed in the terrestrial woody debris (Figs. 10h, 10i). Considering the thermal maturity of the samples (Ro=1.22%-1.34%), it could be inferred that the tiny spot-like and irregular OM pores were caused by the generation of gaseous hydrocarbons in gas-prone maceral in kerogen, which was also consistent with the occurrence of a small amount of CH4 inclusion in calcite veins.
The OM pores in Longtan shales of Well XY1 mainly consisted of mesopores and macropores, and had a relatively rounded shape with well defined-boundaries (Fig. 11). The distribution density of OM pores in different types of debris varied considerably. Some debris developed abundant pores with different sizes, ranging from several nanometers to hundreds of nanometers and even a few micrometers (Figs. 11a-11c), while other debris had only sporadic pores or no pores at all (Figs. 11b, 11d). Xu and Sonnenberg (2017) observed a similar phenomenon in the immature Bakken Shales, indicating that these rounded pores may not be the outcome of thermal evolution. The shape of the strip debris was extremely distorted and twisted, while the internal OM pores were still rounded and equant, implying that this kind of debris was hard enough to resist compaction (Figs. 11d-11e).
FE-SEM images showed that the irregular and spongy-like pores with diameter ranging from several to hundreds of nanometers were densely developed in the OM from Dalong shales of Well EJ1 (Fig. 12). There was a small number of terrestrial woody debris that had no pores or only a few rounded pores (Figs. 12g, 12i). The higher degree of thermal evolution gave rise to more densely distributed pores in the OM, while the quantitative distribution of pores in the terrestrial woody debris was almost the same as that from the samples from Well XY1, indicating that these rounded pores in the woody debris were not the product of thermal evolution.