The study area is located in Block A in the west of the Ordos Basin (Figure 1). From the structural background, the study area is located in the southern part of the Tianhuan Depression (Figure 1). Affected by the thrust and nappe of the western margin of the Ordos Basin, the tectonic activity in the study area was relatively strong, and a large number of small-scale faults are developed (Liang et al., 2019). The extension length of the fault is mostly distributed in 1–15 km, and the fault was mainly formed at the end of the Yanshanian Movement (Li et al., 2017). Low-amplitude uplift traps are developed in the local strata of the Yan’an Formation, and the traps are mainly distributed between 1–6 km².

The stratigraphic units of the Jurassic strata in this area include the Yan’an, Zhiluo and Anding Formations, among which, the Yan’an Formation has a thickness between 220–350 m. Thick river-lake delta deposits are developed in the Middle Jurassic Yan’an Formation period. In addition, coal seams are developed in the Yan’an Formation, and the Jurassic oil reservoirs in the study area are mainly distributed in the Lower system, which is also the main research target layer of this study.

The Middle Jurassic Yan’an Formation, as the main coal and oil-bearing strata in the basin, is relatively developed in the whole basin. The lithology is mainly gray and dark gray mudstone intercalated with coal seam, yellow-gray medium and fine-grained sandstone and siltstone. Because the sandstone at the bottom of the Zhiluo Formation has a strong scouring effect on the top strata of the underlying Yan’an Formation during the deposition process, the thickness of the Yan’an Formation in some areas becomes thinner. Figure 2 shows the stratigraphic unit division of the Yan’an Formation and the vertical lithologic features of the Yan’an Formation revealed by the Well Weisha 2 (Ji et al., 2014).

The experiments used in this study include thin section, scanning electron microscopy, nuclear magnetic resonance, and relative permeability tests. Thin sections and scanning electron microscopy were used to study the pore structures and the distribution of different mineral components in the sandstone. Furthermore, these experiments were used to better understand the lithology and physical properties of the Yan’an Formation sandstone.

The analysis and test of the movable fluid characteristics in this paper include nuclear magnetic resonance and relative permeability experiments. The NMR test instrument is a MagneT2000 tester. The NMR test curve includes the T₂ time spectrum curve under the condition of 100% saturated formation water (before centrifugation) and the T₂ time spectrum curve after centrifugation at 450 psi. In the nuclear magnetic resonance test, the technical indicators are: the maximum working temperature is 155°C/0.5 h, the maximum working pressure is 137.9 MPa, the minimum resistivity of the drilling fluid is 0.02 Ω m, the electronic circuit AC voltage is 180 V, the frequency is 60 HZ, and the current is 250 mA. The probe DC voltage is 600 V and 800 mA/pulse. In addition, the “unsteady-state” experimental method was adopted to carry out the oil-water relative permeability test experiments on the cores. Through this test, the oil-water relative permeability curve and its characteristic parameters of the experimental samples can be obtained.

The lithologies of the Jurassic sandstones in the study area include lithic quartz sandstone, lithic feldspar sandstone and feldspar lithic sandstone. The sand bodies have poor lateral continuity, and most of the cross-sections are in the form of lenses with a flat top and a convex bottom. In addition, the sand bodies develop various types of cross-bedding, which are vertically bell-shaped and box-shaped. According to the difference of throats, the pore types of the Yan’an Formation sandstones mainly include four types (Figure 3) (Bhatti et al., 2020; Zhong et al., 2021). The first type has constricted throats with little difference in the size of pores and throats (Figure 3A); the second type has a thin-necked throat, and the size of the throats is significantly lower than the pores (Figure 3B); the third type has deformed sheet-like throats, which are associated with mineral compaction (Figure 3C); the fourth type has tubular throats, which are usually developed in the intercrystalline pores of interstitials, and the pores are mostly filled by kaolinite and illite (Figure 3D).

The observation results of casting thin section and electron microscope scanning show that there are various types of reservoir spaces in the target sandstone. The target layer mainly develops intergranular pores (Figure 4A), followed by dissolution pores and a small amount of intragranular pores. However, the dissolution pores are dominated by feldspar dissolution pores. The dissolution pores include intergranular dissolution pores, feldspar dissolution pores, detrital dissolution pores and a small amount of carbonate dissolution pores (Figure 4B).

The surface porosity of the target sandstone mainly ranges from 7.90 to 10.79%, with an average value of 8.78%. The content of residual intergranular pores in the Yan’an Formation is relatively high. Residual primary intergranular pores developed in strong hydrodynamic environments and were common in thin sections (Figure 4A). Intercrystalline pores are mainly developed in clay minerals such as kaolinite and illite. Intercrystalline pores refer to pores developed inside authigenic minerals. The larger the volume of kaolinite crystals, the better the development of intercrystalline pores (Schmitt et al., 2015; Ren et al., 2020; Zhao et al., 2020). This is due to the fact that kaolinite crystals are stacked together in a book-like form and are not sensitive to water. Therefore, a large number of intercrystalline pores between the crystal particles are retained.

Statistics show that feldspar has the highest degree of development of dissolved pores, and its content ranges from 0.77 to 1.55%, with an average of 1.17%. Followed by intergranular dissolved and debris dissolved pores, the content of which is distributed between 0.75 and 1.70%, with an average of 1.02%. Overall, the corrosion effect of the target layer is not strong. In addition, some micro-fractures are developed in the target layer, which can significantly improve the permeability of the reservoir.

The relationship between the porosity and gas permeability of the 10 groups of NMR samples tested is shown in Figure 5. It can be seen that there is a very good positive correlation between porosity and permeability. The porosity of the samples is mainly distributed in 9–17%, and the gas permeability is mainly distributed in 1 mD∼35 mD. The target sandstone belongs to the medium-porosity and medium-permeability sandstone reservoir. The good correlation between porosity and permeability indicates that the target layer is a pore-type reservoir. Pores are mainly developed in the target layer, while fractures are less developed. The low degree of fracture development is related to the inactive tectonic activity in the Ordos Basin.

The target sandstone has experienced strong cementation. Cementation refers to the precipitation of supersaturated dissolved components in pore water under the conditions of elevated temperature and pressure, and then the clastic sediments are cemented into rocks (Wang et al., 2017; Zhong et al., 2021). The cementation experienced by the target layer mainly includes clay, siliceous, calcareous and feldspar cementations.

Statistics show that the clay cement content in the target layer ranges from 3.31 to 10.75%, with an average content of 5.65%. The authigenic clay minerals grow vertically from the pore surface to the pore center, and therefore, play a damaging role in reservoir porosity. The authigenic clay minerals in the study area are mainly kaolinite, illite, and a small amount of illite/smectite mixed layer and chlorite (Figure 6A).

Dissolution mainly occurs in the mature stage of organic diagenetic evolution, which is often consistent with the rapid transformation of smectite in the smectite/smectite clay minerals. It is also the peak period of organic acid production, so secondary pores are mostly formed in the middle diagenetic A stage. Intervals with strong dissolution are often secondary pore development segments, and are more likely to form favorable reservoirs. The dissolution of various components (clasts, matrix, cement, etc.) in the sandstone promotes the formation of a large number of secondary pores (Figure 6B). The development of dissolution pores has an extremely important influence on the physical properties of the reservoir.

The movable fluid saturation of the sandstone samples of the target layer in the study area is distributed between 40 and 72%, with an average value of 54%. The different peak characteristics of the saturation component NMR curves before centrifugation can reflect the proportion of pores of different sizes in the rock. In addition, the T₂ relaxation time spectra of the NMR component curves before centrifugation can also reflect the pore size. When T₂ time is less than 1 ms, it indicates micropores (pore diameter <0.1 μm); when T₂ time is between 1 and 10 ms, it indicates small pores (pore diameter is between 0.1 and 0.5 μm); when T₂ time is between 10 and 100 ms, it indicates mesopores; when T₂ time is between 100 and 1000 ms, it indicates macropores (pore diameter is between 2.5 and 10 μm); when T₂ time is greater than 1000 ms, it indicates karst caves (pore diameter >10 μm). Generally, the types of NMR curves of sandstone reservoirs can be divided into three types: unimodal, bimodal and trimodal. Through analysis, it is considered that the NMR curve of the Yan’an Formation sandstone reservoir in the study area is relatively single in peak shape, with double peaks developed (Figure 7).

According to the different pore sizes reflected by the saturation component NMR curves before centrifugation, the pore types of the target sandstone reservoirs are divided into three categories: namely, micropore-small-pore-mesopore type (the left peak is significantly larger than the right peak), small pore-mesopore type (the left peak is significantly smaller than the right peak), micropore-small-pore-mesopore uniform distribution type (the left peak and the right peak have little difference).

The micropore-small-pore-mesoporous type is represented by the sample No. X-1 (Figure 7A). The saturation component curve before centrifugation shows that the T₂ relaxation time of the sample X-1 is mainly distributed in the range of 1–100 ms; the T₂ time distribution of the left peak is wider, which means that micropores and small pores dominate the rock sample. In addition, according to the coverage of the T₂ spectrum, the proportions of micropores and small pores are roughly the same. The small-mesopore type is represented by the sample No. X-2 (Figure 7B). The saturation component curve before centrifugation shows that the T₂ relaxation time of the X-2 sample is mainly distributed in the range of 10–100 ms; the T₂ time distribution of the right peak is wider, which means that the small pores and mesopores dominate the rock sample. The uniform distribution pattern of micropore-small-pore-mesopore is represented by the sample X-3 (Figure 7C). The saturation component curve before centrifugation shows that the T₂ relaxation time of the X-3 sample is mainly distributed in the range of 0.1–100 ms; and the T₂ time distribution range of the left and right peaks is wide, which means that the micropore, small pores and mesopores are well developed, and the proportion of small pores is higher on the whole.

For all samples, there is a T₂ relaxation time distribution in the interval of 100–1000 ms, but the proportion is small. It represents that a few large pores are also developed in the rock sample. In addition, there is an interval with T₂ relaxation time greater than 1000 ms in the micropore-small-pore-mesoporous samples. It shows that a few karst caves may be developed in this kind of rock samples, while no karst caves are developed in the other two types of samples.

The internal pores and throat spaces of sandstone reservoirs are narrow, and there is a certain distribution of irreducible water and movable water. Therefore, the T₂ time spectrum of NMR can be used to identify fluids with different occurrence states. The T₂ cut-off value can be determined according to the projected point value of the accumulated NMR signal intensity of the sample after centrifugation before the centrifugation (100% saturated water) on the T₂ accumulated signal intensity curve of the sample. The area covered by the post-centrifugation curve above the T₂ cut-off represents movable water, and correspondingly, the area covered by the post-centrifugation curve below the T₂ cut-off represents bound water.

The T₂ cut-off values of the sandstone samples were mainly distributed between 1 and 25 ms, with an average value of 8.24 ms. The correlation between T₂ cut-off value and porosity is not obvious. It is related to the complex pore structures of the studied sandstone and the narrow variation range of T₂ cut-off value distribution. In addition, there is a certain negative correlation between the T₂ cut-off value and the percentage of movable fluids in the rock samples. When the T₂ cut-off value is higher, the irreducible water saturation in the rock increases, and the corresponding percentage of movable fluid decreases. There is a very good negative correlation between mobile fluid percentage and irreducible water saturation.

Generally, the amount of movable fluid is mainly affected by the permeability of the reservoir, that is, it is mainly controlled by the throat in the reservoir space. Therefore, there is a certain positive correlation between the movable fluid content and the physical parameters. The experimental results show that the reservoir of the Yan’an Formation has high saturation of movable fluids. It was mainly distributed in 43.17–71.24%, with an average of 56.90% (Figure 8A). At the same time, the fractured samples have higher movable fluid saturations. This is because the existence of fractures greatly reduces the capillary pressure inside the tight rock, and some of the bound fluids are transformed into movable fluids. In addition, the movable fluid saturation has a significant correlation with permeability (Figure 8B). The results of well test productivity comparison show that when the movable fluid saturation is greater than 50%, the oil-producing well basically does not produce water. The tight sandstone of the Yan’an Formation has the characteristics of water wettability, and the pore surface is covered by irreducible water. The oil molecules mostly exist in a free state. Thus, high movable fluid saturation represents high oil saturation. Then, in the early stage of reservoir development, this type of reservoir only produces oil and basically does not produce water.

For different samples, when the water saturation (S_w) changes in the same range, the oil and water relative permeability ratio K_o/K_w will change by different times, that is, the slope of the curve will be different. The smaller the slope, the smaller the change in the relative permeability ratio, and the more stable the oil and gas development effect is (Figure 9). Therefore, for the original low natural energy water-flooding reservoir, the water saturation of the tight reservoir will continue to increase with the increase of the injected water. Correspondingly, the oil-water ratio production index value will continue to decrease.

Generally, the relative permeability curve is divided into three stages: the oil phase seepage section in the state of irreducible water, the oil-water two-phase co-permeability area, and the water phase seepage in the residual oil state. Typical oil-water relative permeability curve types include Type I (concave K_w-K_o Line), Type II (descending K_o line - apex of the sloping K_w line), Type III (slowly rising K_w line - steeply falling K_o line). The target layer has the characteristics of Types I and II (Figure 10). In the oil-water two-phase seepage interval, the oil-level relative permeability of the Yan’an Formation reservoir is low. And its average residual oil saturation is 31%, and the oil-water two-phase co-seepage zone has a wide range (34% on average).

The results of water flooding experiments show that the average irreducible water saturation of the Yan’an Formation sandstone reservoir is 35.14%, and the average final oil displacement efficiency is 51.14% (Figure 11). There is a good positive correlation between the oil displacement efficiency and the co-permeability zone (Figure 11A). As the co-permeability zone range increased from 15 to 55%, the oil displacement efficiency increased from 30 to 65%. There is a significant negative correlation between oil displacement efficiency and irreducible water saturation. That is, as the irreducible water saturation increases from 20 to 55%, the oil displacement efficiency decreases from 65 to 30%. This shows that high irreducible water content is unfavorable for oil displacement. In addition, there is a certain negative correlation between oil displacement efficiency and residual oil saturation (Figure 11C). The higher the residual oil content, the lower the oil displacement efficiency. There is a good power exponential positive correlation between oil displacement efficiency and average oil recovery rate (Figure 11D). However, if the oil production rate is too high, it will cause interlayer interference and pore pressure to drop too fast, which is not conducive to stable oil and gas production. Therefore, the oil recovery rate should be kept within a relatively reasonable range.

This study found that when the cores have fractures, they have characteristics of high permeability, high oil recovery rate, high oil displacement efficiency in anhydrous period, low irreducible water saturation and low residual oil saturation (Figure 11). The initial productivity of the fractured samples is high, but the excessively high oil production rate can cause premature water breakthrough and reduce the productivity rapidly. At the same time, water channeling caused by fractures can form a high-speed seepage channel, which affects the displacement effect and recovery factor.