In this study, the cores of deep shale gas reservoirs in 5 shale gas wells in Luzhou area of Sichuan Basin were selected for experimental test and analysis. These five wells are respectively L211, L217, Y101H3-8, Y101H91-4, and Y101H65-5, whose locations are displayed in Figure 1b. In order to enhance the comparability between samples, the samples in this study were taken from the high GR section at the bottom of Longmaxi formation in five wells. Among them, two core samples were selected from each well for experimental testing. The mineral composition, TOC content, and porosity of these samples are shown in Table 1. The samples have a TOC content of over 4%, with quartz content exceeding 50% and clay content less than 20%, indicating good gas generation potential and fracturing conditions. However, there are inter-well differences in porosity values, ranging from 3.52% to 5.81%, with an average of 4.88%. Among them, the porosity of wells L211 and Y101H91-4 is relatively low, falling below the average level.

As the mercury injection pressure increases from 0.035 MPa to 413.6 MPa, the cumulative mercury injection volume of 10 shale samples gradually increases (Figure 2a). For some samples (such as L217-1/Y101H91-4-1/Y101H65-5-1/Y101H65-5-2, etc.), when the pressure reaches its maximum value (413.6 MPa), it begins to decrease, but the cumulative mercury injection volume continues to increase before gradually decreasing. This “over topping” phenomenon mainly occurs in low-porosity and low-permeability shale samples, which is due to the recovery of mercury volume expansion during the pressure reduction process (Sun et al., 2020). The MICP porosity values of the samples, calculated based on the cumulative mercury intrusion volume under maximum pressure and the bulk density measured under initial pressure, are presented in Table 2. The samples show significant variation in MICP porosity values, ranging from 0.42% to 3.16%, with an average of 1.56%. Well L217 exhibits the highest average MICP porosity (2.68%), while Well L211 shows the lowest (0.56%). According to the Washburn equation, the connected pore volume related to the corresponding pore throat size was calculated. The pore throat size distribution curve shows that more than half of the total pore volume is connected by pore throats smaller than 20 nm, and the pores and micro-fractures larger than 150 nm to several micrometers are also developed (Figure 2). In this study, the porosity of overmature marine shale samples with pore throats ranging from 3 to 20 nm accounts for 50.34%∼82.56% of the MICP porosity, providing substantial storage space for shale gas.

To gain a clear understanding of the phenomenon of mercury molecules injecting into pores and micro-fractures during MICP process, we designed WM immersion experiments under different pressure conditions, and observed the locations of WM filling into pores and micro-fractures under FE-SEM. Sample L217-1 was subjected to WM impregnation at pressures of 500, 1700, 8,000, and 30,000 psi, respectively.

At a low pressure of 500 Psi, FE-SEM observations show that WM mainly occupied micro-fractures with large width and long length parallel to the lamination (Figure 3a). However, micro-fractures with both narrow width and short length remain unfilled by WM, particularly those diagenetic contraction fractures that exhibit almost no WM infiltration (Figure 3b). As the pressure increases to 1700 Psi, corresponding to a pore size of 105 nm, WM filled micro-fractures and a small number of pores, but organic pores and smaller pore size inorganic pores are basically not filled (Figure 3c). As the pressure increases to 8,000 Psi, corresponding to a pore size of 25 nm, WM successfully filled small organic pores, small-diameter inorganic pores, and diagenesis-related micro-fractures in sample L217-1 (Figures 3d,e). As the pressure increases to 30,000 psi, the pore size accessible to WM decreases to 7 nm. Under a low magnification, it is clearly visible that the pores and micro-fractures are filled with a large amount of WM (Figure 3f). Upon continuous magnification and observation of micro-fractures and pores, it was found that some interconnected micro-fractures with both narrow width and short length exhibited a partial filling with WM, which may be related to the volume shrinkage of the WM after cooling (Figure 3g). The pores (organic and inorganic pores) connected to the micro-fractures were fully filled with WM (Figures 3g,h). However, due to poor connectivity, a large number of intragranular pores did not exhibit WM filling (Figure 3i). It can be seen that with the increase of pressure, WM is regularly immersed in the pores and micro-fractures of shale. WM typically penetrates into larger-scale micro-fractures, then fills smaller-scale micro-fractures and inorganic pores, subsequently entering inorganic and organic pores with smaller diameters until most of the connected pores and micro-fractures are occupied by WM.

The pore-fracture network in shale consists of a fracture system connected to the matrix pores (Davudov et al., 2020). Regarding the study of pore connectivity, scholars have employed methods such as spontaneous imbibition, CM-SANS, and FIB-SEM 3D reconstruction to analyze the pore connectivity within shale reservoirs, and these studies indicate that the organic pore connectivity in high-overmature marine shale gas reservoirs is relatively good (Sun et al., 2020; Sun et al., 2023; Hu et al., 2015; Hu et al., 2017), but the connectivity between micro-fractures and pores remains poorly understood. To address this knowledge gap, this study quantitatively analyses the types of pore-fracture combinations through FE-SEM observations after WM immersion experiments and MICP.

A large number of nano pores and micro nano scale micro-fractures are developed in shale gas reservoir. Numerous studies have shown that in the overmature Longmaxi Formation shale reservoirs, the pore diameters of inorganic pores are larger than those of organic pores, but both are smaller than those of micro-fractures (Yu et al., 2019; Yang et al., 2019). Among the inorganic pores, there are numerous isolated intragranular pores, and the pore diameters and quantities of organic pores in kerogen are significantly smaller than those in pyrobitumen (Teng et al., 2022; Liu et al., 2022; Guo X. W. et al., 2019). The MICP and WM immersion experiments involve injecting liquid metal fluid into the shale pore-fracture system under external force. This process follows the Washburn equation, which indicates that the mercury first enters fractures and pores with larger pore sizes, and then moves into pores with relatively smaller pore sizes. Therefore, by taking the partial derivative of the cumulative mercury injection volume with respect to the injection pressure, we can effectively distinguish the rate of liquid mercury filling into the pore-fracture system, which reflects the connectivity of pore-fracture system in shale samples. Figure 4 shows that the mercury injection rate (dV/dP) of the shale reservoir in the study area gradually decreases with pressure increases. The pressure corresponding to the first minimum value after the highest peak of dV/dP is used to distinguish micro-fractures, and the pressure corresponding to the minimum value of dV/dP tending to 0 for the first time is used to distinguish pores. Taking L217 sample as an example, according to the value of dV/dP, mercury injection process can be roughly divided into three intervals: high mercury injection rate (<1.2 MPa), moderate mercury injection rate (1.2∼16 MPa), and low mercury injection rate (>16 MPa) (Figure 4). It is worth noting that due to differences in the pore structure among shale samples, the pressure corresponding to the mercury injection rate range at different stages also varies. For example, the two samples from well Y101H65-5 have the lowest mercury injection rate (Figure 4) and also can be divided into three intervals: high mercury injection rate (<3 MPa), moderate mercury injection rate (3∼16 MPa), and low mercury injection rate (>16 MPa), indicating poor connectivity of pore-fracture system in the shale reservoir of this well. On the contrary, the dV/dP of shale samples from Well L217 and Well Y101H3-8 is relatively high, indicating good connectivity of pore-fracture system in the shale reservoir.

To better distinguish the stages of mercury molecules entering pores and micro-fractures during the MICP experimental process, the L217-1 sample that had completed the WM immersion experiment was selected for detailed analysis. Based on the change in dV/dP value, the mercury injection stage of the MICP is divided into three stages (Figure 5). The dV/dP value corresponding to point A is the first minimum value after the highest peak, indicating a turning point where the mercury injection rate changes from low to high and then back to low, and the pressure corresponding to point A is 1.2 MPa. Point B represents the minimum value of dV/dP tending to 0 for the first time, revealing the process where the mercury injection rate changes from high to low and then slight increase. The pressure value corresponding to point B is 16 MPa. So, when the mercury injection pressure is less than 1.2 MPa, the main space where mercury molecules penetrate is micro-fractures, with little penetration into pores and smaller-scale diagenetic contraction fractures (Figures 3a,b). When the injection pressure ranges from 1.2 to 16 MPa, micro-fractures are completely filled with WM, and then some inorganic pores begin to be filled with WM (Figure 3c). As pressure gradually increases, WM fills a large number of intergranular pores and organic pores, but intragranular pores and organic pores with poor connectivity are not filled (Figures 3g–i). Therefore, the mercury injection curve is categorized into three stages: the micro-fracture filling stage with mercury (<1.2 MPa), the stage of mercury filling pores and micro-fractures (1.2∼16 MPa), and the pore filling stage with mercury (>16 MPa). This sequential penetration pattern, where mercury infiltrates micro-fractures first followed by pores, is consistent with previous studies (Sun et al., 2023; Yu et al., 2019). The same analytical approach can be applied to other samples for pore-scale classification during mercury injection processes.

Some scholars have classified the types of micro pore-fracture combinations in shale gas reservoirs, mainly through qualitative understanding of the contact relationships between micro-fractures and pores based on electron microscopy observations (Fu et al., 2021), but quantitative characterization has not been achieved. In the previous section, the mercury injection curve was divided into three stages, revealing the process of mercury entering pores and fractures, and achieving quantitative division of shale reservoir pores and fractures using the mercury injection curve, which lays the foundation for quantitative analysis of pore-fracture combination type.

The pore-fracture combination types in shale reservoirs aim to understand the connectivity between pores and micro-fractures, and further enhance the understanding of the difficulty of shale gas enrichment and extraction. MICP can characterize pores ranging from 3 nm to several micrometers, providing significant advantages in understanding pore-fracture structure (Yu et al., 2019). MICP involves injecting fluid into the pores and micro-fractures of shale reservoirs under external force. When characterizing pore connectivity using MICP, the main focus is on the mercury injection volume and rate. Based on this, we have defined the micro pore-fracture combination parameter (Q_P-F), which can be calculated using the following formula: Q P − F = C × Q P Q F (1)

Where, Q_P is pore connectivity parameter; Q_F is fracture connectivity parameter; C is total connectivity parameter. C, Q_P and Q_F can express as Equations 2–4, respectively. C = Φ C Φ T (2) Q P = Φ P Φ T · V P V T · K P K P + K F (3) Q F = Φ F Φ T · V F V T · K F K P + K F (4)

Where, Φ_T is total porosity of shale core, %; Φ_C is connective porosity calculated by MICP, %; V_P is mercury injection volume after pressure exceeds point A, mL/g; V_T is accumulated mercury injection volume, mL/g; V_F is mercury injection volume at pressure less than point A, mL/g; K₁ is average mercury injection rate after pressure exceeds point A, (mL/g)/MPa; K₂ is average mercury injection rate at pressure less than point A, (mL/g)/MPa.

From Equation 1, it can be understood that Q_P-F represents the pore volume connected by the unit micro-fracture volume. The larger the value of Q_P-F, the better the connectivity of the pore-fracture system. Relevant calculation parameters were listed in Table 2. Our calculations reveal that when the Q_P value exceeds the Q_F value by an order of magnitude, the Q_P-F value is greater than 1; conversely, the Q_P-F value is less than 1. From the geological perspective of shale gas, it is believed that shale reservoirs are characterized by a large number of organic pores and a relatively small number of micro-fractures (Li et al., 2025; Chen et al., 2019). However, these micro-fractures serve as connecting pores, enhancing the migration of shale gas (Li et al., 2024; Liu et al., 2024). Studies also indicate that in areas with fault development, shale reservoirs with a high degree of natural fractures and micro-fractures exhibit very poor pore development, often resulting in poor reservoir performance and gas-bearing properties (Hu et al., 2024; Fu et al., 2021; Fan et al., 2020). It is evident that a higher degree of micro-fractures development hinders pore development. This serves as indirect evidence that our calculation results align with the geological features of shale gas reservoirs. Therefore, greater attention should be directed towards the connectivity between micro-fractures and pores, particularly in scenarios of limited micro-fractures development.

The Q_P-F values of all shale samples are distributed between 0.01 and 9.67, with an average value of 1.82, indicating that the degree of pore development in the shale reservoir is higher than that of micro-fractures. Among them, the Q_P-F values of the four samples from Well L217 and Well Y101H3-8 are greater than 1, with an average value of 4.02, indicating that the shale reservoirs of these two wells exhibit good pore-fracture connectivity. The Q_P-F value of Well L211 is the smallest, with two samples being 0.20 and 0.10 respectively, indicating that the shale reservoirs of this well exhibit poor pore-fracture connectivity. Based on the magnitude of the Q_P-F value, the pore-fracture combination types in shale reservoirs are categorized into three types: strongly connected pore-fracture type (SCPFT) (Q_P-F>1), moderately connected pore-fracture type (MCPFT) (0.5<Q_P-F<1), and weakly connected pore-fracture type (WCPFT) (Q_P-F<0.5). The Q_P value of SCPFT shale sample is much greater than Q_F value (Table 2), indicating that the degree of pore development is much better than that of micro-fractures. The Q_P value of WCPFT shale sample is equal to or slightly greater than the Q_F value (Table 2), reflecting that the pores of shale reservoir are developed, but the connectivity between pores and micro-fractures is poor. MCPFT shale reservoir is between the first two types. It is worth noting that the Q_P-F value of the shale reservoir of well L217 is much higher than that of other shale gas wells, especially the L217-2 sample, because the well is located far away from other wells and is relatively flat (Figure 1b), with weak structural deformation, high reservoir overpressure and high degree of pore development. Therefore, Q_P-F can indirectly distinguish the quality of shale reservoir and guide the optimization of favorable areas for shale gas exploration.

Micro pores and micro-fractures in shale reservoirs are important reservoir spaces and seepage channels of shale gas, which affect the enrichment and single well recovery of shale gas. Therefore, the pore-fracture combination type has an important impact on the exploration and development of shale gas. In the previous section, according to the Q_P-F value, the pore-fracture combination types of shale reservoir are divided into three categories. The following mainly analyzes the reservoir capacity and gas bearing property of shale reservoirs with different pore-fracture combination types.

Figure 6a shows that the porosity of all SCPFT shale samples is high, with an average of 5.28%; but the porosity of MCPFT and WCPFT shale samples varies greatly. It shows that SCPFT shale reservoir has strong reservoir capacity, while MCPFT and WCPFT shale reservoirs have strong heterogeneity of reservoir capacity. The corresponding shale gas reservoir has a similar variation law that the gas content of all SCPFT shale samples is greater than 7.5 m³/t, with an average of 8.28 m³/t, and the gas content of MCPFT and WCPFT shale samples is great difference (Figure 6b). Shale gas reservoirs exhibit typical characteristics of being both source and reservoir, and shale gas does not need to be migrated or migrated for a short distance in situ (Curtis et al., 2012; Liu et al., 2016; Hu et al., 2024). This will cause some disconnected pores to still contain gas in MCPFT and WCPFT shale reservoirs. However, it is difficult to exploit shale gas in these pores during the exploitation of shale gas. SCPFT shale gas reservoirs have high porosity (5.28%), high gas content (8.28 m³/t), and good pore-fracture connectivity which is conducive to shale gas exploitation. At the same time, this kind of reservoir has high quartz content, which is conducive to fracturing operation and forms complex fracture network structure. Therefore, SCPFT shale gas reservoir is the most important target in the process of shale gas exploration and development.

In order to further analyze the differences of shale gas enrichment and percolation in different pore-fracture combination types, the correlation between pore-fracture combination type parameters and gas bearing property and permeability was drawn in this study. The results show that there is a good logarithmic relationship between Q_P and gas content (Figure 7a), there is no correlation between Q_F and gas content (Figure 7b), and there is a positive correlation between Q_P-F and gas content to some extent (Figure 7c), suggesting that shale gas in shale reservoirs is predominantly stored within pores, with fractures exhibiting a relatively weaker storage capacity. A shale reservoir characterized by more developed pores exhibits superior productivity (Fu et al., 2021; Liu et al., 2024), aligning with our established understanding. The correlation between Q_P-F and gas content is not as good as Q_P (Figures 7a,c). Q_P-F merely reflects the degree of pore development to a certain extent, but more crucially, it embodies the connectivity of pore-fracture system. So there is a good positive correlation between Q_P-F and shale permeability (Figure 7d). Generally, the higher the permeability of a shale reservoir, the greater its flow capacity becomes, resulting in better productivity for a single well (Davudov et al., 2020). Figure 8 demonstrates a positive correlation between the Q_P-F value and the estimated ultimate recovery (EUR) of a single well, indicating that shale gas wells with superior pore-fracture connectivity tend to exhibit higher gas production rates.

Among the five shale gas wells selected for this study, Wells L217 and Y101H3-8 exhibit Q_P-F values for their shale reservoirs exceeding 1, classifying them as SCPFT. This kind of shale reservoir has good development of pores and micro-fractures, and the contact degree between pores and micro-fractures is high (Figure 9a), which is conducive to the enrichment and exploitation of shale gas. Well L211, on the other hand, has a Q_P-F value for its shale reservoir below 0.5, categorizing it as WCPFT. The development degree of pores and micro-fractures in this kind of shale reservoir is poor, and the connectivity of pores-fractures is poor (Figure 9c), which is not conducive to the enrichment and exploitation of shale gas. For Wells Y101H91-4 and Y101H65-5, their shale reservoir Q_P-F values are within the range of 0.16–0.97 but above 1 (note: this should be a clarification or correction as per context, possibly meaning “above 0.16 but some values are less than 1″), with average values of 0.4 and 0.64, respectively. Consequently, Well Y101H91-4 falls into the WCPFT category, while Well Y101H65-5 is classified as MCPFT, which shows good development characteristics of pores and micro-fractures, but the connectivity is relatively weak (Figure 9b). This elucidates the reason why Wells Y101H91-4 and Y101H65-5 exhibit high porosity and gas content; however, despite undergoing the same fracturing process, they failed to yield satisfactory gas production results. Therefore, in shale reservoirs where Q_P-F is less than 1, the pores-fracture connectivity is inadequate. To address this, the fracturing operation method must be optimized, including measures such as increasing the number of perforation clusters, enhancing the intensity of fracturing fluid, and augmenting the sand concentration. These adjustments aim to achieve a more intricate fracture network structure, thereby unlocking more pores with limited connectivity and ultimately enhancing shale gas recovery.