Study on the Wettability and Spontaneous Imbibition Characteristics of Lacustrine Shale

,


Introduction
Wettability describes the properties of the interface between different phases. In reservoir engineering, wettability shows the tendency of reservoir rock surfaces to be preferentially wetted by a fluid phase [1,2]. Shale oil reservoir wettability refers to the tendency of oil to expand or adhere to the surface of inorganic minerals or kerogen pores in shale reservoirs [3,4]. Wettability plays a vital role in the exploration and development of shale oil by affecting the enrichment and flow of oil in shale reservoirs. e wettability of shale determines the affinity of the surface for oil and water, which affects the ratio of free shale oil in the pores of organic matter and inorganic minerals. It is an important parameter to accurately estimate the movable resources of shale oil. Shale oil development requires extensive hydraulic fracturing of shale oil reservoirs. However, the reflux efficiency of fracturing fluid in some shale oil reservoirs is too low, which dramatically increases the cost of shale oil development. erefore, the problem of shale fracturing fluid loss has attracted the attention of a large number of scholars [5][6][7][8][9][10]. e academic circles generally believe that wettability is the main factor affecting the water spontaneous imbibition of shale and the primary mechanism that causes excessive water imbibition of shale. Studying the wettability of shale oil reservoirs is extremely important for selecting suitable fracturing fluids and improving oil and gas recovery [11][12][13][14][15][16].
e current general opinion is that inorganic minerals are water wettability under the original reservoir conditions, while the organic matter is oil wettability [20][21][22][23]. e type of kerogen will affect the oil wettability of organic matter. It makes the characterization of shale wettability extremely complicated [24]. e pore composition and connectivity of shale are complex [25][26][27][28]. Generally speaking, pores composed of inorganic minerals are hydrophilic, and pores distributed in organic matter are lipophilic [29]. e higher the maturity of organic matter, the stronger the lipophilicity [30]. e simultaneous appearance of inorganic water-wet macropores and organic oil-wet small pores makes shale have double wettability, which is also where the wettability of shale oil reservoirs is different from that of conventional reservoirs [31]. e contact angle experiment and the spontaneous imbibition experiment are the most commonly used methods for characterizing wettability. Engelder et al. [32] measured the oil and water/brine contact angles of Marcellus and Haynesville shale samples under air to characterize the wettability of the shale. Mokhtari et al. [33] measured the contact angle of deionized water at ambient room temperature and pressure on the surface of Eagle Ford, Mancos, Bakken, Niobrara, and Green River shale samples. ey analyzed the correlation between the contact angle and the total organic carbon content. Borysenko et al. [34] studied the contact angle and spontaneous imbibition behavior of two shales with different wettability (Bass Basin and Continental Officer Basin), and some pure minerals or their mixtures. Liang et al. [35] studied the wettability of organicrich shale samples from the Longmaxi Formation (LF) in China through contact angle and spontaneous imbibition experiments. Takahashi and Kovscek [36] conducted spontaneous countercurrent spontaneous imbibition experiments on low-permeability siliceous shale core samples to study the influence of the PH value of different types of brine on wettability. Dehghanpour et al. [37] studied the wettability of shale in Fort Simpson (FS), Muskwa (M), and Otter Park (OP) in Horn River Basin (HRB) and measured the contact angle of oil and water in the air on the shale surface, and a spontaneous imbibition experiment was carried out using intact shale samples. However, the wettability results of shale obtained by spontaneous imbibition experiments often contradict the results obtained by contact angle experiments [37,38]. is is because the spontaneous imbibition experiment focuses on the average wettability of the entire core, while the contact angle experiment shows the wettability of a particular surface. e spontaneous imbibition experiment combined with nuclear magnetic resonance can dynamically monitor the amount of liquid imbibed by the shale, the priority order of liquid imbibition pore throats [39], and reveal the imbibition mechanism and the influencing factors of wettability of the shale [40].
Lithofacies comprehensively reflect rock characteristics such as mineral composition, organic matter, and pore structure [41]. However, the wettability of different lithofacies shale in the northern Songliao Basin is less characterized, and there is a lack of comparative studies on contact angle and spontaneous imbibition. In view of this situation, the shale of the Qingshankou Formation in the northern Songliao Basin has been classified. Six samples with different lithological characteristics were used for this experiment. e experiment uses contact angle measurement, spontaneous imbibition combined with nuclear magnetic resonance method, to characterize the wettability of shale with different lithofacies characteristics and analyze the influencing factors of the wettability of different lithofacies shale.

Geological Background and Sample Information.
e area of this study is consistent with our previous study on using contact angle to characterize the wettability of shale [42]. e Songliao Basin in Northeast China is a Cretaceous continental rift basin.
e Qingshankou Formation was deposited in the postcracking thermal subsidence stage. e thickness of the stratigraphic group is usually in the range of 200-500 m. e lithology and lithofacies vary greatly. e lithology in the Qijia-Gulong sag is a thick lacustrine dark brown mud shale, while the lithology at the edge of the sag is an interbedded shale and sandstone [43,44]. e main lithology is dark brown mud shale with a small amount of oil shale, which is a favorable layer for shale oil exploration and development. According to the change of lithofacies, the Qingshankou Formation can be divided into 3 sections. e target interval of this study is the first member of the Qingshankou Formation. During the deposition of the first member of Qingshankou Formation, the lake basin expanded to the largest extent. A set of gray-black mudstone, oil shale, and gray-white silty fine sandstone with varying thicknesses were deposited in the basin [45].
In this study, cores were taken from the different depth sections. e samples were analyzed for whole-rock minerals, TOC, and thin rock sections. in-section observations show that the Qingshankou Formation shale in the northern Songliao Basin has various structural features such as layered, laminar, and yellow massive ( Figure 1). e lithofacies type, organic matter abundance, XRD whole-rock minerals, and clay mineral analysis results of the samples are shown in Table 1. is lithofacies classification is based on Liu Bo's classification scheme for mud shale in Qingshankou Formation in the southern Songliao Basin, using a fourcomponent three-terminal classification method, fully considering the sample's macrostructure, organic matter content, and mineral composition. ere are three representative lithofacies: sandy interlayer (YY-72-2) has high siliceous mineral content, low clay mineral content, and low TOC. Low organic matter laminated siliceous shale (YJ-62-3) has a higher content of siliceous minerals, a higher content of clay minerals, and a higher TOC. High organic matter massive clay shale (YJ-27-1) has low siliceous mineral content, high clay mineral content, and high TOC.

Sample Processing
(1) Contact Angle Experiment Sample. e surface roughness will have a significant impact on the measurement of contact angle [46]; it is necessary to pretreat the sample surface to eliminate this effect. e surface of the sample was polished by 400, 800, 1500, and 2000 mesh grinding wheels in sequence. When Ra is less than 0.5 μm, the surface of the sample can be regarded as approximately smooth. Ra is the arithmetic mean of the absolute value of the distance from each point on the measured contour to the reference line within the sampling length L, and its size can be used to characterize the roughness of the sample surface [42,46].
(2) Spontaneous Imbibition NMR Experiment Sample. Use a wire cutter to cut each sample into four cylindrical samples with a length of 10 mm and a diameter of 25 mm. Two of them were cut along the direction of the vertical bedding plane, and two were cut along the direction of the parallel bedding plane ( Figure 2) [38,47]. Samples 1 and 2 have parallel bedding for oil imbibition and water imbibition, respectively; samples 3 and 4 have vertical bedding for oil imbibition and water imbibition, respectively. e original column sample contains a certain amount of colloidal asphaltene, which blocks the pore throat and affects the process of spontaneous fluid infiltration into the pore throat. It is necessary to perform an oil washing operation before the spontaneous imbibition and nuclear magnetic resonance experiment of the sample. is oil washing operation uses the DY-6 high-efficiency oil washing instrument; the washing oil solvent is a 4 : 1 alcohol-benzene mixed solvent. e heating temperature is set to 70°C, the pressure is set to 1 MPa, and the oil washing time is 72 hours. After the oil washing process was completed, the sample was taken out, and some samples were damaged. e pore volume of the damaged sample cannot be obtained by calculation, and only the complete sample is selected for drying. e complete sample is selected and put in the drying box for drying treatment, the temperature is set to 70°C, and the time is set to 72 hours. e diameter and length of the complete sample are remeasured. e porosity of the complete sample is measured using the helium porosimetry method. To ensure that only the bottom surface of the sample touches the imbibed fluid during the imbibition process, apply epoxy resin around the cylinder sample, leaving only the upper and lower bottom surfaces. e treated samples were divided into two groups: one group was used to imbibe n-dodecane, and the other group was used to imbibe deionized water ( Figure 3 and Table 2).

Experimental Process
(1) Contact Angle Experiment. e testing instrument is the JC2000D contact angle measuring instrument produced by Shanghai Zhongchen Company. e instrument includes two parts: the CCD optical system and the stage control system. Contact angle measurement range: 0-180°, contact angle measurement resolution: 0.01°.

Geofluids 3
Experimental procedure is as follows. ① rough the control software, use a microsyringe to drop a droplet on a smooth, clean, and horizontally placed mineral surface. e size of the droplet is only 2-3 microliters to eliminate the influence of gravity on the shape of the droplet. ② After the droplet contact angle is stable, use the camera to photograph the shape of the droplet on the mineral surface. ③ Use the image processing software that comes with the instrument to analyze and measure the left contact angle, right contact angle, and the average value of the contact angle of the droplet on the mineral surface. ④ Group samples were subjected to 3-4 repeated measurements, and the average value of the contact angle was obtained.
(2) Spontaneous Imbibition NMR Experiment. e nuclear magnetic resonance experiment instrument is MicroMR12-025V nuclear magnetic resonance analyzer. e experiment is divided into the n-dodecane imbibition experiment group and the deionized water imbibition experiment group. e process of the n-dodecane imbibition experiment group is as follows. ① e quality and nuclear magnetic signal of the dried sample were measured after pretreatment. ② e sample was immersed in n-dodecane containing an evaporating dish. ③ e changes in the imbibition quality of the sample and the amount of imbibition NMR signal were collected. Previous core imbibition experiments have shown that the imbibition velocity is very high at the initial stage, and then the velocity becomes smaller until the imbibition velocity drops to zero [47]. erefore, the collection time interval in the early stage is short, and the collection time interval in the later stage is appropriately extended. ④ When the imbibition quality of the sample and the nuclear magnetic signal appear unchanged or decreased for three times, it is deemed that the imbibition is completed. ⑤ e sample after the imbibition is put into the vacuum pressure saturation device for pressure saturation for 72 h. ⑥ e quality and NMR signal of the sample were measured when pressure saturation is completed. e experimental process of the deionized water imbibition experiment group is roughly the same as that of the n-dodecane imbibition experiment group, but the deionized water imbibition experiment group did not perform saturated water. e reason is that the samples selected in this experiment contain a lot of clay minerals, and when the shale is pressurized and saturated with water, the shale is easy to fragment [48].

Data Processing.
e spontaneous imbibition curve records the change law of sample imbibition over time, reflecting the pore characteristics of the rock, the    interaction mechanism between the rock and the fluid, and so on. ere is a difference between rocks with different lithofacies. In order to focus on the influence of lithofacies on the imbibition characteristics, it is necessary to remove the effect of pore volume on imbibition. erefore, the quality of imbibed liquid m l is converted into the percentage C of the volume of imbibed liquid V l in the pore volume V p of the sample. e specific calculation process is as follows: where V p is the pore volume of the column sample, ml; V l is the volume of the imbibed liquid, ml; C is the percentage of the volume of imbibed liquid in the pore volume of the column sample; C o is the percentage of oil imbibition volume in the pore volume of the column sample; C w is the percentage of imbibed water volume in the pore volume of the column; r is the radius of the bottom surface of the column, mm; h is the height of the column, mm; Φ is the porosity of the column sample (%); m l is imbibed liquid quality, mg; and ρ l is imbibed liquid density, 10 −3 g/cm 3 . In this experiment, the density of deionized water is 1 g/cm 3 and the density of n-dodecane is 0.753 g/cm 3 .

Contact Angle.
e average value of the "gas-waterrock" contact angle of the six selected samples with different lithofacies characteristics is shown in Table 3. From the table, it can be seen that the "gas-water-rock" contact angle of the low organic shale is far lower than the high organic shale, and the contact angle of the sandy interlayer (YY-72-2) is much smaller than that of other samples. e reason is that the high organic content makes the sample more lipophilic, the sandy interlayer (YY-72-2) sample has a higher permeability, and the water droplets penetrate the surface.

e Difference of Imbibition along the Vertical and Parallel Bedding Directions.
e water imbibition rate and final imbibition volume of the samples will be affected by the bedding. For sandy interlayer and shale, the vertical bedding plane's imbibition rate and final imbibition volume are greater than those of the parallel bedding plane. Figure 4(a) is the comparison of the water imbibition curves of the parallel bedding and the vertical bedding of the sandy interlayer sample (YY-72-2). e bedding structure of the sandy interlayer sample is not well developed, and the water imbibition rate of the parallel bedding plane and the vertical bedding plane is not much different, but the final imbibition volume is different. e water imbibition of the parallel bedding plane is significantly less than that of the vertical bedding plane. Figures 4(b) and 4(c) are the comparison of Geofluids 5 water imbibition curves of parallel bedding plane and vertical bedding plane of low organic matter laminated siliceous shale (YJ-62-3) and high organic matter laminated siliceous shale (YA-151-1). e sample has a very developed lamellar structure, which leads to great differences in the imbibition in different directions. It can be seen from Figures 4(b) and 4(c) that the rate of imbibition of vertical bedding plane is much greater than that of parallel bedding plane, and the  6 Geofluids final water imbibition volume of vertical bedding plane is also greater than the final water imbibition volume of parallel bedding plane.

Differences in the Imbibition Characteristics of Oil and
Water. Different lithofacies samples have differences in the oil and water imbibition rate, imbibition time, and final imbibition volume. For sandy interlayer (YY-72-2) and high organic matter laminated siliceous shale (YA-151-1), the oil imbibition rate is obviously greater than the water imbibition rate, and the final water imbibition volume is greater than oil. For low organic matter laminated siliceous shale (YJ-62-3), high organic matter massive siliceous shale (YD-22-1), and high organic matter massive clay shale (YJ-27-1), the initial oil imbibition rate is similar to the water imbibition rate, and the oil-water equilibrium time is also similar, but the final water imbibition is greater than that of the oil. e water imbibition rate of the low organic matter laminated siliceous shale (YG-616-3) is much greater than the oil imbibition rate, and the final water imbibition volume is much greater than oil. Figure 5(a) shows the comparison of the oil-water imbibition curve of the sandy interlayer sample (YY-72-2). It is observed from Figure 5(a) that the oil imbibition rate is greater than the water imbibition rate, but the water imbibition time is longer than the oil imbibition time.
e final imbibition volume of water is greater than that of oil. Figure 5(b) shows the comparison of oil-water imbibition curves of the low organic matter laminated siliceous shale (YJ-62-3). It is observed from Figure 5(b) that the water imbibition rate is similar to the oil imbibition rate at the initial stage. ere is little difference in the imbibition time, and the final imbibition volume of water is greater than that of oil. Figure 5(c) shows the comparison of oil-water imbibition curves of the high organic matter laminated siliceous shale sample (YA-151-1). It is observed from Figure 5(c) that the initial oil imbibition rate is greater than the water imbibition rate, but it is subsequently overtaken by water. e water imbibition time is longer than the oil imbibition time, and the final imbibition volume of water is greater than that of oil. Figure 5(d) shows the comparison of oil-water imbibition curves of high organic matter massive siliceous shale (YD-22-1). It is observed from Figure 5(d) that the water imbibition rate at the initial stage is similar to the oil imbibition rate, and the final imbibition volume of water is greater than that of oil. Figure 5(e) shows the comparison of oil-water imbibition curves of low organic matter laminated siliceous shale (YG-616-3). It can be observed from Figure 5(e) that the water imbibition rate is significantly greater than the oil imbibition rate, the oil imbibition time is longer than the water imbibition time, and the final imbibition volume of water is much greater than that of oil. Figure 5(f ) shows the comparison of oil-water imbibition curves of high organic matter massive clay shale (YJ-27-1). It can be observed from Figure 5(f ) that the water imbibition rate is similar to the oil imbibition rate at the initial stage, the oil imbibition time is longer than the water imbibition time, and the final imbibition volume of water is greater than that of oil. e volume of oil imbibed by the sample is less than the pore volume; that is, the sample will not be excessively imbibed with oil. C o is the percentage of the final oil imbibition volume to the pore volume, and the relationship of C o between different lithofacies is as follows: high organic matter massive clay shale > high organic matter laminated siliceous shale > sandy interlayer > low organic matter laminated siliceous shale > high organic matter massive siliceous shale. e volume of water imbibed by the sample is greater than the pore volume; that is, excessive water imbibition occurs in the sample. Capillary force is a physical action that occurs under the action of connected pores. If capillary force is the main driving force for water imbibition, the imbibition volume will be smaller than the pore volume. erefore, the hydration of clay minerals is the cause of excessive imbibition of water by the sample. C w is the percentage of the final water volume to the pore volume, and the C w relationship of different lithofacies is as follows: low organic matter laminated siliceous shale > high organic matter massive clay shale > sandy interlayer > high organic matter laminated siliceous shale > high organic matter massive siliceous shale.

NMR Imbibition. Different lithofacies have different ways of oil imbibition. Sandy interlayer is simultaneously
imbibed with large and tiny pores, and low organic matter laminated siliceous shale and high organic matter massive clay shale first infiltrate tiny pores and then large pores. e T2 spectrum peaks of the three different lithofacies samples of n-dodecane imbibed showed a relatively obvious shift to the right. In contrast, the T2 spectrum peaks of the imbibed deionized water did not appear to be significantly shifted to the right.
is shows that most of the imbibed water is adsorbed water, and only a small amount of water becomes free-flowing water. Figures 6(a) and 6(b) show the imbibition characteristics of the sandy interlayer (YY-72-2). e imbibition rate of n-dodecane is of high speed, and the imbibition equilibrium is reached in only 1314 minutes, and almost the entire pore volume can be imbibed. ere is little difference between the final oil imbibition and saturated oil. e imbibition method is simultaneous imbibition of tiny and large pores, and the T2 spectrum peak shifts to the right. e imbibition of water is slow, and the imbibition equilibrium can be reached at 6800 minutes. e amount of imbibition is large, but most of the water is adsorbed water, and only a tiny amount of water becomes free-flowing water, and the T2 spectrum peak does not appear to shift to the right. Figures 6(c) and 6(d) show the imbibition characteristics of low organic matter laminated siliceous shale (YJ-62-3). e n-dodecane imbibition rate is fast, and the imbibition equilibrium is reached at 2250 min. e oil imbibition is low, and it is not entirely imbibed, which is a certain gap between the results of the saturated oil. e imbibition method is to first imbibe tiny pores and then large pores, and the T2 spectrum peak shifts to the right. e imbibition rate of water is slow, reaching equilibrium at 4800 minutes. Almost all of the imbibed water is adsorbed water, only a  small part is free-flowing water, and the T2 spectrum peak does not appear to shift to the right. Figures 6(e) and 6(f ) show the imbibition characteristics of high organic matter massive clay shale (YJ-27-1). e imbibition rate of n-dodecane is fast, the imbibition equilibrium is reached in 2000 min, and the oil imbibition is relatively high, which is almost the same as the imbibition result of saturated oil. e imbibition method is first to imbibe tiny pores and then imbibe large pores, and the T2 spectrum peak appears to move right shifts. e speed of water imbibition is slow, reaching equilibrium after 4000 minutes. Almost all the water imbibed is bound by clay minerals and capillary tubes. Only a small part is free-flowing water, and the T2 spectrum peak does not appear to shift to the right. After the sample imbibed water, the T2 spectrum did not shift to the right; only the sandy interlayer shifted slightly to the right. e lower the degree of shift to the right, the weaker the core's ability to absorb water to produce microfractures. e water in the small pores does not flow to the large pores; that is, water lock is prone to occur, and the reservoir does not have the ability to "automatically relieve water lock." e relationship between the potentials of different lithofacies to "automatically release water lock" is as follows: Sandy interbeds > Low organic matter laminated siliceous shale > High organic matter massive clay shale. e stronger the water wetness, the greater the "potential for mitigating water lock."

Influencing Factors of Spontaneous Imbibition
3.4.1. Bedding Plane. For sandy interlayer, the difference between the imbibition of parallel bedding planes and vertical bedding planes is not significant. For shale, there is a certain difference between the imbibition of parallel bedding planes and the imbibition of vertical bedding planes. Regardless of the imbibition rate and the final imbibition amount, the imbibition along the bedding plane is greater than the imbibition along the vertical bedding plane. e reason is that ① the bedding plane has good connectivity; ② the water imbibition along the bedding plane is more likely to induce microcracks and further widen the microcracks. From this, we can conclude that, for shale intervals, the method of lateral fracturing and injecting fracturing fluid is better than that of longitudinal fracturing. . Figures 7(a) and 7(b) show the relationship between pore volume and imbibition volume. Among them, V o has a good correlation with pore volume, while V w has a poor correlation with pore volume. e reason for this phenomenon is that shale contains certain clay minerals, and excessive water imbibition occurs in different lithofacies samples. Figure 8(a) shows the relationship between C o and TOC: the higher the TOC content of shale, the higher the C o . is is because the higher   the TOC is, the more developed the lipophilic pore network will lead to the intake of oil. Figure 8(b) shows the relationship between C w and clay mineral content: the higher the content of clay minerals, the lower the C w . Clay minerals are the cause of excessive water imbibition in shale. When the clay minerals are too high, the shale will swell and block the pores due to water imbibition, which is not conducive to the further water imbibition of the shale.

Conclusions
(1) For the shale in this study area, whether n-dodecane or deionized water is imbibed, the imbibition rate and imbibition volume along the bedding plane are significantly greater than those along the vertical bedding plane. For sandy interlayers, the difference between parallel bedding plane imbibition and vertical bedding plane imbibition is not large. erefore, in the shale oil development process, the method of lateral fracturing and injecting fracturing fluid is better than that of longitudinal fracturing.
(2) e water imbibition volume of the samples in this study area is greater than the pore volume: low organic matter laminated siliceous shale > high organic matter massive clay shale > sandy interlayer > high organic matter laminated siliceous shale > high organic matter massive siliceous shale. e higher the content of clay minerals, the lower the C w . Clay minerals are the cause of excessive water imbibition in shale. When the clay minerals are too high, the shale will swell and block the pores due to water imbibition, which is not conducive to the further water imbibition of the shale.
(3) e oil imbibition volume of the samples in this study area is less than the pore volume: high organic matter massive clay shale > high organic matter laminated siliceous shale > sandy interlayer > low organic matter laminated siliceous shale > high organic matter massive siliceous shale. e higher the TOC content of shale, the higher the Co. is is because the higher the TOC, the more developed the lipophilic pore network, and the more oil will be imbibed. (4) Nuclear magnetic experiments show that most of the water absorbed by the sample is adsorbed water, and the core has a poor ability to automatically relieve water lock. e relationship between the potentials of different lithofacies to "automatically release water lock" is as follows: sandy interbeds > low organic matter laminated siliceous shale > high organic matter massive clay shale. e stronger the water wetness, the greater the "potential for mitigating water lock."

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
Haitao Xue designed the project and wrote the main manuscript. Guozhi Ding helped in drawing the figures and drafting the manuscript. Zhentao Dong and Rixin Zhao defined the statement of problem. Ce An helped in discussing the problems and revising the manuscript. Guozhi Ding and Zhentao Dong helped in discussing the main idea and drafting the manuscript. Boheng Li and Yuan Zhou helped in calculating the data and drawing the figures. Penglei Yan, Jinliang Yan, Yuxi Jin, and Chunlei Li helped in revising the figures. All authors reviewed the manuscript.

Acknowledgments
is study was funded by the Sinopec Independent Innovation Fund (20200925145915002). Table S1. Quality change before and after imbibition of the n-dodecane sample. Table S2. Quality change before and after imbibition of the deionized water sample. Figure S1.