For the preparation of the imbibition solution, four different surfactant systems were employed in the experiment: a 0.2% KCl system, an SDS system (0.2% KCl with an anionic surfactant), a SY system (0.2% KCl with a combination of anionic and non-ionic surfactants), and an HD system (0.2% KCl with a combination of anionic, non-ionic, and amphoteric surfactants) (Table 1).

The Funing 2 Member of Qintong Depression consists of deep lake and semi-deep lake shale characterized by strong vertical heterogeneity. Various types of reservoir spaces are developed within the shale, including tectonic fractures, bedding fractures and micro-pores, with porosity generally ranging from 3% to 6.29%. Specifically, cores were extracted from 1–42/49 of SD 201 well and 4–12/24 of HY 201 well. The core from SD 201 is bulk clay-rich shale, and the core from HY 201 well is laminated calcareous shale.

Mineral composition based on X-ray diffraction (XRD) testing was performed using a Rigaku D/MAX-2500pc X-ray diffractometer. Powder samples, ground to a fineness of 200–300 mesh, were placed on slides and gently pressed with a glass slide to compact and fill the sample holder. The measurements were carried out with a tube voltage of 40 kV and a current of 40 mA, utilizing a Cu target and Ka-ray radiation. Diffraction data were collected at a scan rate of 4°/min, covering a scanning angle range of 4°–50°, with a step interval of 0.02°. Data processing for qualitative and quantitative mineral analysis was conducted using Jade 6.5 software.

The porosity of core samples measured in this study was measured using the helium porosity method. Due to the low molecular weight of helium, this method allows for more accurate measurement of small pore spaces, providing a closer estimate of the true porosity of the rocks. The helium porosity and air permeability of 8 core samples from the study area were determined using a PoroPDP-200 core pressure pore and permeability meter. Each core sample was approximately 5 cm in length and 2.5 cm in diameter. The experimental results are shown in Table 2.

Before the imbibition experiment, shale samples were thoroughly washed and dried. The cores were cleaned using an extraction method in which the samples were placed in an extractor, and an oil-cleaning agent, a mixture of toluene and ethanol, was circulated via the siphon and steam pipe at the top and bottom of the extractor. The original fluid in the core was washed out until the oil-cleaning agent becomes clear. The cores were then saturated with crude oil using a vacuum saturation method.

In terms of NMR principles, when shale is saturated with a hydrogen-containing fluid, the NMR measures the relaxation signals of the fluid, rather than those of the shale matrix. The NMR T₂ relaxation time for fluid in larger pores is longer compared to fluid in smaller pores, meaning the relaxation time increases with pore size. Additionally, the amplitude of the relaxation signal corresponds to the volume of fluid in the pore, with a large amount of fluid corresponding to a higher signal amplitude. Therefore, NMR T₂ spectra can be utilized to quantitatively evaluate both the pore radius distribution and the fluid accumulation within pores during the spontaneous imbibition process in shale. The specific experimental steps are as follows: 1. The core sample was dried in an incubator at 105°C, until its mass remained constant. The T₁- T₂ spectrum and T₂ spectrum of the dry core were then measured using an NMR instrument; 2. The core was vacuum-saturated with crude oil. Field crude oil was placed in a pressurized tank, and the core was saturated under a pressure of 30 MPa for 10 days to ensure full saturation. The saturated core was then immersed in crude oil, and the T₁- T₂ and T₂ spectra were measured after saturation.3. Remove the saturated core with tweezers, surface oil was wiped off with gauze, and the core was placed into a beaker to begin the single-end spontaneous imbibition experiment; 4. At specified imbibition time (10 min, 30 min, 1 h, 1 day, 4 days, 8 days), the core was removed, the surface liquid was wiped off, the core’s mass was measured, and the T₁- T₂ and T₂ spectra were recorded. This process was repeated until the T₂ signal no longer showed significant changes.

The NanoVoxel-2702E CT scanner was employed to scan different cross-sections before and after the spontaneous imbibition of core plugs at the same depth. The X-ray source provided a maximum power of 10 W and had a voltage limit of 150 kV. The scanner was equipped with three magnification lenses (×4, ×10, and ×20), offering a maximum theoretical resolution of 0.5 μm and a field of view of 25 cm. The charge coupled device (CCD) had a maximum resolution of 2048*2048 pixel. The object platform was capable of rotating from −180° to +180°. The hydration experiment was conducted as follows: (1) Core samples were dried in an oven for 48 h and then scanned by CT in their dry state; (2) The samples were placed in a beaker filled with water for hydration periods of 24 h, 72 h and 120 h; (3) After hydration, the samples were re-dried and subsequently scanned by CT.

For the FESEM analysis, a regular block sample with a dimension of 1.5 cm × 1.5 cm × 1 cm was selected for the experiment. The sample surface was first roughened with sandpaper and polishing liquid, and then polished with a Gatan 685C argon-ion polisher. After polishing, the samples were gold-coated and examined using a Hitachi Regulus8230 high-resolution scanning electron microscope (FESEM). The FESEM experiment has the following specifications: a maximum resolution of 0.8 nm; an acceleration voltage range of 0.01 kV–30 kV; magnification from ×20 to ×1,000,000, and a scanning speed of 80 s/frame to 25 s/frame. The experimental procedure involved immersing the sample in deionized water for various durations (1, 3, 5 days) to simulate hydration. Before electron microscope observation, the samples were dried to a constant weight under vacuum conditions to prevent water vapor from contaminating the electron microscope sample chamber. The dried samples were then analyzed again using FESEM to assess the changes in mineral composition, microfractures, and micropore structures of the shale before and after hydration.

The T₂ spectra of the SD bulk clay-rich shale, measured at different imbibition times after oil saturation, reveal that the water phase signal gradually increases throughout the imbibition process and stabilizes around 8–10 days (Supplementary Figure S1). The pores primarily involved in imbibition replacement are small pores (<1 ms) and mesopores (1 ms–10 ms). The imbibition efficiency is calculated by dividing the difference between the T₂ spectral peak area at each time point and the initial peak area by the total T₂ area of the pore space. As shown in Figure 2, the imbibition process can be divided into three stages: in the early imbibition stage (<2 days), rapid imbibition replacement occurs preferentially in the connected pore throat near the core surface, resulting in a substantial volume of crude oil being produced. In the middle imbibition stage, as the imbibition liquid penetrates deeper into the pores, deep oil phase is replaced, leading to a slower increase in imbibition efficiency. In the later stage of imbibition, the imbibition efficiency remains nearly unchanged, indicating that imbibition has reached equilibrium by approximately 8 days. Among the different imbibition agents, the SY mixed agent demonstrated the highest efficiency, followed by the HD mixed agent, SDS, and finally the 0.2% KCl solution.

The interpretation of NMR results, particularly data from two-dimensional T₁-T₂ spectra, provides comprehensive insights into the saturation levels of different fluids within shale samples (Li et al., 2019; Li et al., 2020). Developed a fluid identification method using trapezoidal regions within the two-dimensional NMR T₁-T₂ spectrum (Supplementary Figure S3). By analyzing signals from different regions of the spectrum and correlating them with multi-step pyrolysis data, they created a fluid identification chart, which is particularly applicable to various types of organic-rich rocks. Unlike curve-shaped charts, the fixed boundaries of the trapezoidal regions provide greater consistency, making this approach a more reliable and universal T₁-T₂ chart for studying fluid properties in shale. In this study, the two-dimensional NMR T₁-T₂ spectrum was employed to investigate and analyze fluid occurrences in different lithologies of the Funing 2 Member. The established T₁-T₂ fluid identification spectrum (Figure 3) reveals that the original SD bulk clay-rich shale samples show signals associated with residual kerogen/bitumen and clay mineral-bound water (Figure 3A). After oil saturation, signals of adsorbed oil are observed in the samples (Figure 3B). As the imbibition process progresses, the signals for organic matter and free oil decrease, while the signal intensity for free water increases (Figures 3C, D).

By comparing the macroscopic micro-fracture distribution characteristics of shale samples before and after water phase imbibition (Figure 4), it is evident that hydration significantly enhances the connectivity between pores and fractures within the rocks. Before hydration, the shale sample exhibited only one prominent microfracture. At the initial stage of hydration, the existing fractures first propagate along the parallel bedding planes, forming several distinct foliation fractures. As hydration progresses, both the width and length of these parallel fractures expand substantially, and some fractures connected to form a new fracture network. This development further increases both the fracture volume and surface area, thereby enhancing fluid transport pathway within the shale.

Based on an understanding of the macroscopic microfracture development characteristics, this study further employs dynamic observation techniques combined with energy spectrum analysis to perform fixed-point monitoring and elemental testing at specific microstructural locations. The objective is to analyze the dynamic expansion patterns of induced microfractures and identify key controlling factors. Using multi-point in-situ scanning technology, changes in the microstructure of shale samples before and after hydration were tracked and compared. Initially, SEM was used to examine the micro-pore structure of the samples prior to hydration, with characteristic points marked and natural fractures characterized. After various imbibition periods, the samples were positioned macroscopically and observed in situ. Continuous imaging at different magnifications was conducted at the same locations to identify and analyze correlations between mineral composition and micro-pore structures (Figure 5). SEM observations revealed a distinct dissolution of rhombohedral carbonate minerals within the shale matrix during the first 24 h of hydration (Figures 5A, B, E, F, I, J), with most dissolution occurring within this period, and minor changes in pore size afterward (Figures 5B–D, J–L). Additionally, the evolution of original natural fractures during hydration was clearly observed. The fracture evolution trends observed through in-situ SEM were closely aligned with the water saturation changes detected by NMR and the fracture volume changes observed through CT scanning. Microfractures on the original shale surface initially developed into hydration fractures within the first day, after which they stopped widening and even gradually closed between days 3 and 5, demonstrating an initial widening followed by closure. This behavior is likely associated with the swelling of clay minerals, which affects their water absorption capacity. Furthermore, on the third day of hydration, oil droplets were observed accumulating around larger pores and microfractures under SEM (Figures 5G, K). By the fifth day, oil-water displacement became clearly visible, with oil droplets effectively expelled from the shale matrix (Figures 5H, L). These observations suggest that, during the imbibition process, the migration and aggregation of the oil phase are significantly influenced by the evolving pore structure and microfracture network.

The evolution of the T₂ spectra at various stages of imbibition for the oil-saturated core of the HY laminated calcareous shale reveals that the water-phase signal gradually increases during the imbibition process, stabilizing after approximately 8–10 days (Supplementary Figure S2). The pore intervals primarily involved in the displacement are micropores (<1 ms) and mesopores (1 ms–10 ms). The imbibition efficiency can be quantified by dividing the difference between the T₂ peak area and the initial peak area by the total pore T₂ area. As illustrated in Figure 6, the core’s imbibition process can be divided into three stages: 1. Early stage (<1.5 days): Rapid imbibition displacement occurs in the connected pore throats near the core surface, driven by capillary forces, resulting in the quick expulsion of a significant amount of oil. 2. Middle stage: The imbibition liquid infiltrates deeper pores, displacing oil from these deeper regions, but the rate of increase in imbibition efficiency slows down. 3. Late stage: Imbibition efficiency plateaus, with little to no change, achieving equilibrium after 6–8 days. In terms of imbibition performance, the SY surfactant and HD surfactant demonstrate the best results, followed by the single surfactant agent, while 0.2% KCl exhibits the least effect.

According to the established T₁-T₂ fluid identification map (Figure 7), the original core of the HY laminated calcareous shale primarily exhibits signals corresponding to clay mineral-bound water and weakly bound oil (Figure 7A). After oil saturation, signals corresponding to adsorbed oil become prominent (Figure 7B). As the imbibition process progresses, the signals from organic matter and free oil gradually diminish, while the intensity of the free water signal steadily increases (Figures 7C, D).

CT scanning results of HY laminated calcareous shale core samples (Figure 8) reveal significant development of microfractures during water infiltration. Hydration leads to the widening and extension of pre-existing fractures. In areas with dense pore networks, smaller cracks merge to form new, larger fractures. During the early stage of hydration, fractures predominantly form parallel to the bedding. However, as hydration progresses, intersection between the original fractures, which are perpendicular to the bedding, begin to emerge, thereby enhancing fracture connectivity. Additionally, beyond hydration-induced fractures, 3D reconstructions reveal the formation of larger pores, likely due to mineral dissolution. This dissolution process weakens the original cementation and facilitates the growth of the interconnected fracture network.

Building on the understanding of macroscopic microfracture development, fixed-point dynamic observation combined with energy spectrum testing was employed to analyze the expansion patterns of induced microfractures and identify key controlling factors. The original surface morphology of the HY laminated calcareous shale (Figure 9A) and the changes observed 1, 3, and 5 days after hydration reveal extensive dissolution of rhombohedral carbonate minerals (Figures 9B–D). This continuous dissolution facilitates the formation of microfractures, which progressively expand into larger and intersecting cracks (Figure 9D). Additionally, the accumulation of oil droplets in shaded regions illustrates the process of oil-water displacement during hydration, where oil in smaller pores is displaced into larger pores and expelled through newly formed microfractures (Figures 9B–D). Mineral dissolution also occurs along the boundaries of more stable minerals that are less susceptible to hydration, such as quartz and albite (Figures 9E, F), leading to the shedding and migration of larger mineral particles. Fractures preferentially form and propagate along boundaries with significant differences in mineral composition, especially around pyrite (Figures 9G, H). As hydration continues, the original fractures further expand (Figures 9I, J). Elemental surface scanning reveals that the fracture areas are primarily composed of carbonate minerals, while silicate minerals are more prevalent on the sides. Fracture expansion is primarily driven by the dissolution of carbonate minerals (Figures 9K, L).

This study highlights the superior performance of surfactant mixtures, particularly the combination of anionic, non-ionic, and amphoteric surfactants, in enhancing imbibition efficiency, when compared to pure anionic surfactants or water alone. Oil displacement surfactants can be classified into four main categories based on their ionization behavior and the nature of their functional groups: non-ionic, amphoteric, cationic, and anionic surfactants (Ge et al., 2015; Meng et al., 2017; Pal et al., 2018; Das et al., 2019; Massarweh and Abushaikha, 2020; Bashir et al., 2022; Isaac et al., 2022). These surfactants exhibit distinct characteristics that influence their ability to alter the wettability of reservoir rocks under varying conditions. Anionic surfactants, which ionize in aqueous solutions to carry a negative charge, are widely used in various products (Wu et al., 2017). In chemical enhanced oil recovery, they are commonly employed in sandstone reservoirs due to their low adsorption rates and good thermal stability (Kamal et al., 2017). High molecular weight anionic surfactants, particularly sulfonate, sulfate, and carboxylate types, are effective at reducing interfacial tension (IFT) to ultra-low levels and can maintain stability across a wide range of pH conditions (Yang et al., 2010). Non-ionic surfactants possess a polar head group with a strong affinity for water, but they do not ionize in solution. These surfactants are primarily influenced by hydrogen bonding and van der Waals interactions (Kamal et al., 2017). As the temperature increases, the strength of hydrogen bonding weakens, leading to reduced solubility of the surfactant in the aqueous phase (Zhao et al., 2005). Non-ionic surfactants generally exhibit a lower ability to reduce interfacial tension (IFT) compared to their ionic counterparts (Iglauer et al., 2009). Amphoteric surfactants can carry either a positive or negative charge depending on the solution’s pH (Lv et al., 2011). These surfactants have shown promising potential for enhanced oil recovery applications, especially under high-temperature and high-salinity conditions (Kamal et al., 2017; Massarweh and Abushaikha, 2020). The inclusion of amphoteric surfactants in surfactant mixtures provides additional stability and improves the fluid’s performance in complex reservoir environments (Lv et al., 2011).

The suitability of different surfactants for various types of reservoirs stems from their distinct mechanisms for altering wettability (Karambeigi et al., 2016; Jing et al., 2019; Liu and Sheng, 2019; Liu et al., 2019). In comparison to conventional surfactant systems, the mixture of anionic, non-ionic, and amphoteric surfactants in this study demonstrated superior performance in shale oil imbibition recovery. This improvement can be attributed to the synergistic effects of these surfactants in enhancing wettability, reducing IFT, and improving emulsification properties. Anionic surfactants are highly effective in altering rock wettability by desorbing organic matter and reducing interfacial tension (IFT) between oil and water, facilitating better oil recovery. Non-ionic surfactants, on the other hand, interact with rock surfaces through hydrophobic interactions, further enhancing wettability and reducing IFT, albeit to a lesser degree than anionic surfactants. Amphoteric surfactants provide additional stability and adaptability under challenging reservoir conditions, such as high salinity and temperature. When combined, these surfactants complement each other’s strengths, enhancing overall performance by improving wettability, reducing IFT, stabilizing emulsions, and increasing the effectiveness of oil displacement in complex reservoir environments. The synergistic interactions between the surfactants in a mixed system not only optimize the individual benefits of each surfactant but also contribute to more efficient oil recovery, making mixed surfactant formulations a highly effective strategy for improving imbibition efficiency in shale oil reservoirs.

This result is consistent with previous studies that supports the effectiveness of mixed surfactant systems in enhancing oil recovery in complex reservoirs such as shale (Lv et al., 2011; Alvarez and Schechter, 2017; Das et al., 2019; Qin et al., 2019; Habibi et al., 2020; Kesarwani et al., 2021; Xiao et al., 2021; Yahya et al., 2022). The results of this study emphasize the potential of mixed surfactant formulations to improve imbibition efficiency, offering valuable insights for the optimization of enhanced oil recovery methods in shale reservoirs under challenging conditions.

Apart from the type of imbibition agent, this study reveals that the imbibition efficiency in shale reservoirs is closely linked to their lithology, with clay mineral content playing a particularly pivotal role in the hydration process. Specifically, the imbibition efficiency of SD clay-rich shale was found to be higher than that of HY laminated calcareous shale. This difference in performance is not solely due to variations in initial porosity, but rather to the higher clay content in SD shale, which enhances imbibition through osmotic pressure and the formation of additional hydration-induced fractures. In contrast, the imbibition process in HY laminated calcareous shale, which has a lower clay content, primarily relies on the dissolution of carbonate minerals.

The high clay content in SD shale significantly influences its imbibition characteristics. Clay minerals, which are abundant in shale reservoirs, act as non-ideal semi-permeable membranes. These minerals create a chemical potential difference, allowing specific ions to pass through and driving water molecules into the nanopores of the clay (You et al., 2019; Hu et al., 2020; Liu et al., 2022; Yang et al., 2023). In high-salinity environments, clay minerals generate excess negative charges due to lattice substitution. To maintain electrical neutrality, water molecules and cations adsorb onto the mineral surface, forming a double electric layer (Dominijanni and Manassero, 2012; Røyne et al., 2015; Fritz et al., 2020). When exposed to low-salinity water, osmotic pressure induces water molecules to enter the clay pores, leading to cation desorption and increased repulsion within the double electric layer (Wilson and Wilson, 2014; Du et al., 2018). This osmotic process not only promotes the expansion of the clay but also generates pore pressure, which further pushes crude oil out, enhancing imbibition.

In contrast, the lower clay content in HY laminated calcareous shale leads to a less pronounced osmotic effect. Instead, imbibition in this lithology primarily occurs due to the dissolution of carbonate minerals. The high levels of pyrite and carbonate minerals in HY laminated calcareous shale make pyrite oxidation a key trigger for these dissolution processes (Ali and Hascakir, 2017). When exposed to water and oxygen, pyrite oxidizes, creating an acidic environment, which accelerates the dissolution of carbonate minerals like calcite, which unstable under acidic conditions (Zhuang et al., 2024). This process alters the mineral composition and promotes the expansion of primary fractures and formation of secondary cracks during hydration (Du et al., 2018). Electron microscopy observations further confirm these lithological differences, highlighting the significant impact of mineral composition on the hydration process. In conclusion, the differences in imbibition efficiency between the SD clay-rich shale and HY laminated calcareous shale can be attributed to the variations in their mineral composition and the mechanisms driving the imbibition process, with osmotic pressure and mineral dissolution playing distinct roles in each lithology.