Pore structure of tight sandstones with differing permeability: The He 8 Member of the Middle Permian Lower Shihezi Formation, Gaoqiao area, Ordos Basin

Tight sandstone has strong pore heterogeneity and complex pore structure, and the pore structure of tight sandstone varies with different permeability. To study the differences in the pore structure of tight sandstone with different permeability, this study investigated the tight sandstone of the He 8 Member in the Gaoqiao area of the Ordos Basin. Factors influencing pore formation are analyzed through experiments utilizing methods, such as distinguishing casting thin sections, scanning electron microscopy, high‐pressure mercury intrusion, and low‐temperature nitrogen adsorption. The results indicate that in this area, the pores of the tight sandstone are primarily dissolution and intercrystalline pores, with occasional intergranular pores and microcracks. Type Ⅱ samples (permeability > 0.2 × 10−3 μm2) are primarily composed of dissolution and intercrystalline pores, with a few visible intergranular and microcracks. In contrast, Type Ⅰ samples (permeability < 0.2 × 10−3 μm2) mainly consist of micropores and intercrystalline pores, where microcracks are observed locally. Pore size research demonstrates that Type Ⅰ samples have pore size under 1000 nm, with peaks primarily at 4–5 nm. There are also peaks between 10–1000 nm, but without a consistent pattern. For Type Ⅱ samples with sizes smaller than 1000 nm, pore distribution is evident. Peak values when the pore size exceeds 1000 nm. Type Ⅰ samples have fewer micron‐level pores, while Type Ⅱ samples are primarily dissolution pores. The nanopores of Type Ⅰ samples consist of flat pores with good connectivity, whereas those of Type Ⅱ samples are mainly blind pores with poor connectivity. Type Ⅰ samples have larger pore fractal dimensions, more intricate pore morphology, and rougher and irregular pore throat surfaces compared to Type II samples. Hence, the sedimentary environment, diagenesis, and mineral composition affect the pore distribution in tight sandstone. The research findings highlight the variations in the pore structure of tight sandstone with varying permeability, providing crucial guidance for classifying and assessing pore structure in tight sandstone reservoirs.

][37][38][39][40][41][42] Since various characterization methods have different principles, the scales of characterizing pores differ.We can obtain the specific parameters of pores and throats and then analyze the pore characteristics quantitatively.Rationally, using various characterization methods and comprehensively analyzing pore characteristics is the key to characterizing tight sandstone pore structures.4][45][46] Note that tight sandstone reservoirs vary greatly in permeability.What are the differences in pore types, size, and other aspects of tight sandstones with different permeability?This is also of great significance for evaluating pores in tight sandstone.This study takes the tight sandstone of the He 8 Member in the Gaoqiao area of the Ordos Basin as a case.Research on the microscopic pore structure of tight sandstones with inconsistent permeability is carried out.The microscopic pore structure of tight sandstones with different permeability was investigated by cast thin section identification, scanning electron microscopy, HPMI, and low-temperature nitrogen adsorption (LTNA).The research results demonstrate the differences in the pore structure of tight sandstone with different permeability, which is significant for classifying and evaluating the pore structure in tight sandstone reservoirs.Also, the above research provides references for exploring the efficient development of tight sandstone gas reservoirs in the study area.

| GEOLOGICAL OVERVIEW
8][49] The Ordos Basin has experienced six stages of evolution, beginning from the Tertiary period.The basin is currently divided into six main units based on their structural morphology and characteristics.1][52] The Ordos Basin hosts a variety of oil and gas formations, with the Mesozoic and the Upper Paleozoic periods being particularly notable for their petroleum strata, which are known to contain natural gas reservoirs.During the Late Paleozoic period, the sedimentary environment of the basin experienced a series of evolutionary processes, including marine deposition, marine-terrestrial transitional deposition, and continental deposition.The basin's early marine deposits, marine-terrestrial transitional facies, and delta deposits serve as favorable source rocks, while the sandstone deposited in the delta area makes up suitable reservoirs.As a result, the basin is characterized by the presence of widely distributed tight sandstone gas reservoirs.
The Gaoqiao area is situated in the central basin area.Its tectonic zoning belongs to the northern Shanxi slope (Figure 1).The Late Paleozoic strata in this area include Upper Carboniferous Benxi Formation (C 3 b), Lower Permian Taiyuan Formation (P 1 t), Shanxi Formation (P 1 s), Middle Permian Lower Shihezi Formation (P 2 x), Upper Shihezi Formation (P 2 sh), and the Upper Permian Shiqianfeng Formation (P 3 s).The C 3 b is subdivided into Ben 1 and Ben 2, the P 1 s is subdivided into Shan 1 and Shan 2, the P 2 x is subdivided into He 8 Member, He 7 Member, He 6 Member, and He 5 Member from the bottom up, and the P 2 sh is subdivided into He 4 Member, He 3 Member, He 2 Member, and He 1 Member from the bottom to up.Among them, the He 8 Member is the major gasbearing strata in the Gaoqiao area, which can be divided into He 8 up and He 8 below .The He 8 Member is found at a burial depth ranging from 2775 to 3778 m with a thickness of approximately 63 m, forming a sedimentary system characterized by a braided river delta front.The lithology of the reservoir mainly consists of gray and dark gray medium coarse, medium, and fine lithic quartz sandstone, as well as quartz sandstone and lithic sandstone.Among them, lithic quartz sandstone has the highest abundance, averaging 57.55%, while sandstone and quartz sandstone are comparatively less abundant, accounting for 24.90% and 17.55%, respectively.

| Samples
We collected 50 samples from the He 8 Member in the Gaoqiao area to analyze and compare the structural | 119 variations of tight sandstone pores.The samples were subjected to physical property testing.On the basis of the test results, we determined a permeability boundary of 0.2 × 10 −3 μm 2 .Ten samples' permeability varied, with five samples having values higher than 0.2 × 10 −3 μm 2 and the other five samples having values lower than that.The samples with permeability below 0.2 × 10 −3 μm 2 were categorized as Type I samples, while those with permeability higher than 0.2 × 10 −3 μm 2 were categorized as Type II samples.To ensure research representativeness, it is essential to ensure that the 10 selected samples can cover the entire area and are collected from different strata (He 8 up and He 8 below ) as well as depths.Moreover, during the selection process, each sample should consist of a larger volume specimen, and multiple tests should be performed on the same sample.The detailed information for each sample is shown in Table 1 (except for one sample where the HPMI experiment failed).

| Casting thin sections
The thin section of rock casting can provide a direct observation of the pores' type, size, connectivity, and development degree.The process for producing rock casting thin sections follows the guidelines set by the SY/ T 5913-2004 industry standard.It involves initially cutting the rock into slices with 25 mm × 25 mm × 5 mm.Next, red-dyed epoxy resin is poured into the rock's pore space, and the casting thin sections are completed through a pasting process.After the sample preparation, the pore characteristics of 10 samples are observed using a polarizing microscope.The experimental observation aligns with the industry standard SY/T 6103-2004 of the People's Republic of China.

| Scanning electron microscopy
Scanning electron microscopy on the micro-nano-scale is a useful tool for studying the characteristics of rock pores and throats.Before observing using scanning electron microscopy, the samples need to be cut and undergo argon ion polishing treatment.To avoid any charging effects, the sample is fixed and plated with a layer of gold film. 53,54The Quanta 400 Field Emission Gun (FEG) scanning electron microscopy is then used to observe the electronic imaging and analyze the pore and throat characteristics.

| HPMI test
The HPMI test can be adopted to quantitatively analyze the tight sandstone pores' size distribution.In this test, injection of the nonwetting phase mercury into the tight sandstone requires overcoming the capillary resistance of the pores.Different pressures correspond to the capillary resistance of various aperture sizes.By measuring the amount of mercury at specific pressure ranges, the pore size distribution of tight sandstone can be calculated using the Washburn equation. 557][58] It is evident that HPMI is more accurate in measuring large pores, while the measurement of nanosized pores often has considerable errors.This study followed the industry standard SY/T 5346-2005, and utilized the Auto Pore IV 9500 high-pressure mercury piezometer to test 10 samples.the cylindrical samples were drilled, dried, and subjected to the HPMI test, with a maximum pressure of approximately 200 MPa.

| Low-temperature nitrogen adsorption
Nanoscale pore size distribution of tight sandstones and the comparative rock area can be analyzed through LTNA experiments.In addition, the pore morphology can be explored by examining the adsorption-desorption curves. 59This experiment followed the People's Republic of China standard SY/T 6154-1995 and employed the Micromeritics ASAP 2420 specific surface tester to conduct the nitrogen adsorption experiment.Before the investigation, we crushed 10 g samples and filtered out fragments measuring 40-60 mesh.We subjected the samples to evacuation at a temperature of 120°C for 3 h to achieve a vacuum degree of 1.0 × 10 −3 Pa.Next, we utilized nitrogen with a purity greater than 99.999% as the adsorbate and monitored the sample's adsorption and desorption capacity at a constant temperature of 77 K.We draw the adsorption and desorption curve with the pressure changed.On the basis of the test data, this study used the multipoint Brunauer-Emmett-Teller (BET) theory to determine the sample's surface area. 60The Barrett-Joyner-Halenda (BJH) theory was employed to calculate the sample's pore size distribution. 61| RESULTS AND DISCUSSION

| Pore morphology characteristics
On the basis of the casting of thin sections and scanning electron microscopy observation, the primary pore types found in the tight sandstone of He 8 Member in the Gaoqiao area are dissolution pores and intercrystalline pores (Figure 2A-C).Type Ⅱ samples, which have permeability greater than 0.2 × 10 −3 μm 2 , primarily exhibit dissolution and intercrystalline pores, with a few visible intergranular and microcracks.Type Ⅰ samples, on the other hand, are mainly characterized by micropores and intercrystalline pores, with some locally visible microcracks.

| Intergranular pores
Intergranular pores are considered primary pores, which are residual pores that remain between the particles after the process of compaction and cementation.The findings indicate that numerous intergranular pores in this area have suffered significant damage, resulting in a low remaining abundance of such pores.In Type Ⅰ samples, it is challenging to observe intergranular pores, whereas they can be observed in some Type Ⅱ samples.The remaining intergranular pores are enclosed by particles like quartz, and exhibit irregular polygons shapes (Figure 2D), with pore diameters range from 50 to 400 μm.Although the large intergranular pore serve as primary storage space for natural gas, their storage capacity is relatively low, and they contribute less to the overall reservoir storage and permeability.While intergranular pores with larger pore sizes are crucial for natural gas storage, the quantity of such pores in this study area is limited, resulting in a relatively low contribution to reservoir storage and permeability.

| Dissolution pores
Dissolution pores are categorized as secondary pores that result from the dissolution of feldspar particles, rock debris particles, and carbonate rock cements due to the presence of acidic fluids.In our investigation, Type I samples do not exhibit any dissolution pores, whereas dissolution pores are commonly found in Type Ⅱ samples (Figure 2E-G).In light of the different dissolution positions, the dissolution pores in this area are intragranular dissolution pores and intergranular dissolution pores (Figure 2E-G).Intragranular dissolution pores are formed mainly by the internal dissolution of rock fragments.Dissolution can also be seen in some quartz particles.Intragranular dissolution pores are often harbor-like or irregular (Figure 2E).Intergranular dissolution pores, on the other hand, emerge from the dissolution between feldspar or rock fragment particles.Soluble particles or early colloids partially dissolve, forming irregularly dissolved pores (Figure 2F).It is evident that dissolution pores are distributed unevenly, displaying different sizes and significant heterogeneity.Some of these dissolution pores are filled with clay mineral cement and carbonate rock cement that formed later, resulting in inadequate storage space (Figure 2G).However, some dissolution pores still remain, serving as viable storage and seepage areas.

| Intercrystalline pores
Intercrystalline pores are considered a significant category of pores in the study area.These pores, mainly consisting of clay intercrystalline mineral pores, are frequently observed in both Types I and II samples.The prevalent clay minerals found in this area include kaolinite, illite, chlorite, and illite-smectite mixed-layer minerals (Figure 2H-L).Observations from electron microscopy reveal that individual kaolinite crystals display a hexagonal plate-like shape, while aggregates of kaolinite appear as paper-like or worm-like structures.
Illite is predominantly filiform or blade-like in shape; while chlorite displays rose petals or needle-like morphology.The illite-smectite mixed layer exhibits a honeycomb-like structure in most cases.During the growth process of clay mineral crystals, various large pores become obstructed, leading to the formation of diverse types of small intercrystalline pores.The intercrystalline pores of clay minerals exhibit significant variation in size, showing various shapes, like, flakes, slits, and honeycombs (Figure 2H-L).The intercrystalline pores generated by kaolinite exhibit enhanced development and relatively improved connectivity.On the one hand, these intercrystalline pores have the capacity to store natural gas, and on the other hand, they can function as conduits by connecting with other pores.Consequently, they play a crucial role in natural gas storage and penetration.

| Microcracks
Microcracks are cracks that occur as a result of diagenesis and mechanical stress.Scanning electron microscopy observations reveal the presence of microcracks in Types I and II samples from the study area, with the predominant types being mineral shrinkage cracks and stress fractures.Mineral shrinkage joints are typically curved and appear mainly around mineral particles (Figure 2M,N).On the other hand, stress joints are formed due to mechanical compaction or tectonic stress.These stress joints are predominantly straight and extend over long distances, often intersecting quartz and rock fragment particles (Figure 2O).Although the small spacing of microcracks has minimal impact on the storage capacity of tight sandstones, they serve as pathways that connect different pores, thus significantly enhancing permeability in tight sandstone formations.

| Pore diameter distribution characteristics
The distribution of pore diameter directly affects reservoir porosity and permeability of reservoirs, making it a crucial aspect of studying pore structures in tight sandstone reservoirs.There are multiple methods available for characterizing pore diameter.Differential pore volume (dV/dD) and pore diameter (D) distribution curves are suitable for analyzing nanopores.On the other hand, logarithmic differential pore volume (dV/dlog D) and pore diameter (D) distribution curves are well-suited for describing larger pores.In this study, it is acknowledged that tight sandstone exhibits a wide range of pore sizes.Therefore, logarithmic differential pore volume (dV/dlog D) and pore size (D) distribution curves were chosen to analyze the characteristics of pore size distribution.The pore diameter distribution of tight sandstone reservoirs can be determined using LTNA and HPMI techniques.LTNA enables the characteristics of pore diameter ranging from 2 to 300 nm.In the HPMI experiment, the maximum pressure applied to the mercury inlet is 200 MPa, and the minimum measurable pore diameter is 7.4 nm.Incorporating the pore diameter distributions obtained from both experiments is crucial for achieving a comprehensive representation of the fullscale pore diameter distribution.As mentioned previously, excessively high pressure in the HPMI experiment can lead to the destruction of the rock structure, resulting in changes in pore characteristics and significant errors in the obtained results.In the particular area, some samples exhibit peaks in the pore diameter distribution when the diameters are less than 30 nm, which can be attributed to false peaks induced by rock failure.For precise measurements of smaller pores, the LTNA experiment is preferred.][64][65] Therefore, this study adopts a threshold of 30 nm, where pores smaller than 30 nm are characterized using the LTNA experiment based on the BJH method, while pores larger than 30 nm are characterized using the HPMI experiment.
In this study, logarithmic differential pore volume (dV/ dlog D) and diameter (D) distribution curves were utilized to evaluate the characteristics of pore size.The pore diameter characteristics obtained from the HPMI and LTNA experiments were distinguished at a pore diameter of 30 nm. Figure 3 illustrates the complete distribution of pore diameter for various sample types.It can be observed from Figure 3 that the pore diameter distribution exhibits a "multipeaks" morphology, indicating variations in the pore size distribution among permeability samples.For pore diameter below 10 nm, a majority of the samples exhibit a peak value of around 4-5 nm, and this trend is observed in both types of samples.Type Ⅰ samples display higher peak values, indicating a greater proportion of significant peaks.This suggests that Type Ⅰ samples with lower permeability are more likely to have porosity in this size range.In the size range of 10-1000 nm, the diameter distribution of samples with varying permeability does not show a clear pattern.Some samples exhibit "double peaks" or "three peaks."In this interval, the diameter distribution of tight sandstone in this range becomes more complex with no apparent regularity.For diameter values exceed 1000 nm, the ordinate value for most Type Ⅰ samples is nearly zero, indicating the absence of significant peaks.Particularly, for diameters larger than 2000 nm, all Type I samples have an ordinate value of zero.In this range, Type Ⅰ samples exhibit minimal pore distribution, while Type Ⅱ samples show partial pore distribution.

| Pore morphology and connectivity
The examination of micron-scale pores morphology in tight sandstones is primarily conducted through casting thin sections and scanning electron microscopy analysis, as mentioned earlier.However, a more precise quantitative analysis is required for nanoscale pores by LTNA experiments.The use of adsorption and desorption curves obtained from LTNA experiments can provide insights into the morphology of these pores.The International Union of Pure and Applied Chemistry (IUPAC) classifies the adsorption and desorption curves into four forms, H1-H4, each representing a different pore structure. 59The findings indicate that a majority of the adsorption and desorption curves for Type Ⅰ samples in the studied area do not overlap.These curves exhibit a hysteresis loop, with a typical example shown in Figure 4A.Referring to the IUPAC standard, the curve corresponds to H3. Below a pressure of 0.4, the two curves overlap, and according to the BJH calculation formula, the pore diameter at this point is determined to be 3.3 nm.This indicates that a significant portion of the pores smaller than 3.3 nm are blind holes with limited connectivity.The two curves diverge above a pressure of 0.4, suggesting that the pores larger than 3.3 nm exhibit a predominantly flat geometry with good connectivity.It is hypothesized that these pores are mainly attributed to intercrystalline pores of kaolinite, which could provide favorable conditions for the storage of oil and gas.The majority of Type Ⅱ samples fail to generate hysteresis loops, as depicted in Figure 4B.The absence of a hysteresis loop, indicating that the pores are mainly blind pores, which may be primarily illite pores or dissolved micropores.While these pores can adsorb gas, they are not conducive to efficient gas flow.In general, the nanopores found in low-permeability Type Ⅰ samples typically consist of flat pores with good connectivity.On the other hand, the nanopores in high-permeability Type Ⅱ sample are predominantly blind pores with limited connectivity.

| Nanoscale pore-specific surface area and volume
The LTNA experiments were conducted to evaluate the specific area and volume of nanoscale pores in the tight sandstone.Type Ⅰ samples exhibited a BET-specific surface area ranging from 0.260 to 3.371 m 2 /g, with an average of 2.163 m 2 /g, while Type Ⅱ samples showed a range of 0.535-2.459m 2 /g, with an average of 1.149 m 2 /g.Hence, it can be concluded that samples with lower permeability possess a larger specific area of nanoscale pores.The BJH pore volume of Type Ⅰ samples ranged from 0.002 to 0.013 m 2 /g, with an average of 0.008 cm 3 /g.On the other hand, Type Ⅱ samples exhibited a range of 0.003 to 0.010 cm 3 /g, with an average of 0.005 cm 3 /g.Similar to the specific area, samples with lower permeability demonstrated a larger overall volume of the nanoscale pore.Figure 5 provides a distribution relationship between the BET-specific surface area and BJH pore volume of the two sample types.The volume of both types of samples increases with an increase in the specific surface area.The considerable specific area and volume of nanoscale pores in tight sandstone significantly contribute to gas storage.Notably, the significance of nanoscale pore storage is higher in samples with low permeability.

| Pore fractal characteristics
Fractal geometry is capable of explaining irregular phenomena observed in nature.9][80][81] Due to the wide distribution of pore diameters, researchers employed different methods to calculate the fractal dimension.Nitrogen adsorption experiments captured data for pores with a diameter of less than 30 nm, while HPMI was adopted to measure the fractal dimension of pores larger than 30 nm.

| Fractal dimensions of pores less than 30 nm
The FHH model is widely regarded as a reliable approach for calculating fractal dimensions, and its expression is as formula (1) 78 : where V is the gas volume, cm 3 /g; P is the equilibrium pressure, MPa; p 0 is the adsorbed gas' saturated vapor pressure, MPa; K is a constant related to the adsorption mechanism; C is a constant.The calculation for fractal dimension is where D is the fractal dimension, and K is the constant calculated by the formula (1).On the basis of the formulas (1) and ( 2), the slope of the linear regression obtained from Ln V and Ln[Ln(p 0 / p)] can be fitted to determine the parameter K. Subsequently, the fractal dimension can be calculated based on the derived value of K.
In this study area, the slope of the linear regression was determined by fitting Ln V and Ln[Ln(p 0 /p)] for tight sandstone samples with pores smaller than 30 nm (Figure 6 and Table 2).It is observed that both types of samples exhibited a significant linear correlation between Ln V and Ln[Ln(p 0 /p)], with a fitted curve having a correlation coefficient R 2 greater than 0.9900.This indicates a strong correlation and high accuracy of the fitting.Therefore, it can be concluded that tight sandstone with a pore smaller than 30 nm exhibits notable fractal characteristics.
According to the fitting results (Table 2), the fractal dimensions of Type Ⅰ samples range from 2.5258 to 2.6373, while for Type Ⅱ samples, the range is from 2.4911 to 2.5433.It can be observed that the fractal dimension of Type Ⅰ samples is generally larger, indicating a more complex pore morphology with rough and irregular pore throat surfaces.On the other hand, the fractal dimension of Type Ⅱ samples is slightly smaller, suggesting a lower degree of irregularity in the pore throat.In this study, the fractal dimension for pores larger than 30 nm is determined by the HPMI technique.The calculation of the fractal dimension is based on the water saturation and mercury saturation. 77,82,83For this particular study, the method chosen to fit the fractal dimension is the water saturation method.
On the basis of the fractal theory, the correlation between water saturation and pore diameters can be expressed as follows 77,82 : where p c is the capillary pressure, S is the water saturation at p c , and p max is the capillary pressure for the maximum pore diameter of the reservoir rock, MPa.
Taking the logarithm of formula (3) from both sides, we can get the following: If the pores adhere to the fractal traits, the lgS and lgp c plots can be fitted to obtain a linear relationship, with a notable correlation.By analyzing the slope and intercept, we can determine the value of D and p max , which represent the fractal dimension and the maximum pore diameter of the reservoir rock, respectively.Notably, the value of D is equal to 3 plus the slope.
After considering the above discussion, the analysis focused on pores with a size exceeding 30 nm, excluding piezometric data for smaller pores.As depicted in Figure 7, the relationship between lgS and lgp c exhibits a linear pattern.Most of the correlation coefficient R 2 surpasses 0.9000, with only one sample exhibiting an R 2 value of 0.8952.This signifies that tight sandstone possesses favorable fractal characteristics in terms of pores larger than 30 nm.The majority of these samples display a fractal dimension exceeding 2.5000, indicating a relatively high overall fractal dimension.Specifically, for Type Ⅰ samples, the fractal dimension ranges from 2.5182 to 2.9371, with an average of 2.7601.Meanwhile, Type Ⅱ samples exhibit a minimum fractal dimension ranging from 2.4581 to 2.7578, averaging at 2.6377 (Table 3).Consequently, it can be concluded that the pores larger than 30 nm exhibit a higher level of complexity in terms of morphological characteristics, displaying considerable irregularity.Furthermore, samples with lower permeability demonstrate a more intricate nature of pores larger than 30 nm, presenting higher irregularity and surface roughness.Conversely, higher permeability Type Ⅱ samples show lower irregularities and surface roughness in their pores larger than 30 nm.

| Comprehensive discussion of fractal dimension features
By utilizing the FHH and the HPMI method, the analysis of pore fractal dimensions was conducted, with a specific focus on pores limited to 30 nm.The findings reveal that tight sandstone demonstrates fractal characteristics in both smaller and larger pores exceeding 30 nm.However, when considering the fractal dimension of pores across different scales in tight sandstone, no systematic pattern emerges.Some samples exhibit multiple segments, while others exhibit a singular part.Type I samples display higher fractal dimensions, along with rough pore surfaces and irregular pore structures, indicating a lack of viability for natural gas storage and penetration within these pores.Conversely, Type II samples show low fractal   dimensions and relatively regular pores, which are conductive to enhanced pore storage and permeability.Furthermore, when assessing the fractal dimension with a 30 nm limit for each sample, it becomes apparent that larger pores surpass 30 nm in terms of fractal size.This suggests that micropores within tight sandstones are more complex and irregular.Additionally, this observation underscores the significance of pores smaller than 30 nm within this particular reservoir type.

| Influencing factors of pore development
5][86][87] In this study area, the analysis identifies sedimentary environment, diagenesis, and mineral composition as the key factors impacting tight sandstone.

| The influence of sedimentary environment on pore development
The pores of tight sandstone reservoirs are significantly impacted by the sedimentary environment.Different sedimentary environments exhibit varying hydrodynamic conditions and material compositions.Hydrodynamic conditions directly affect the grain size of tight sandstone, influencing the formation of primary pore characteristics.
During the deposition period, sandstone with coarser grain size tends to have larger pores between particles, 88,89 which facilitates the formation of primary pores.This allows for easier fluid flow and dissolution in later stages.On the other hand, sandstone with finer grain size hinders the formation of primary pores and negatively affects the formation of secondary pores in the later stage.In this study area, the tight sandstone is primarily composed of finegrained sandstones, with the presence of medium-grained and medium-coarse-grained sandstones.Most Type Ⅰ samples consist of medium-coarse-grained sandstones, some with coarse-grained sandstones.Type Ⅱ samples are predominantly medium-coarse or coarse-grained.Analysis of the pore characteristics of samples with different particle sizes reveals that fine-grained or medium-grained sandstones are mostly composed of intercrystalline pores or microcracks, with fewer dissolution pores.The pore diameter in these samples is mostly less than 1000 nm.Medium-coarse-grained or coarse-grained sandstones, on the other hand, exhibit a combination of dissolution and intercrystalline pores, with some remaining intergranular pores.The pore diameter is distributed across different size scales, and better connectivity is observed in nanoscale pores.Furthermore, the particle sorting and roundness of sediment formation differ in different sedimentary environments, leading to distinct original pore characteristics. 90In this study area, Type I samples in He 8 exhibit poor sorting and roundness, resulting in limited primary pores, which hinders the corroded pore sizes.On the other hand, Type II samples demonstrate poor sorting and moderate roundness, which promotes the formation of primary pores and facilitates the appearance of secondary corrosion pores.This is one of the important factors contributing to the observation of corrosion pores in Type II samples.

| The influence of diagenesis on pore development
After burial, the sediments undergo sedimentation and enter the diagenesis stage, where a range of physical and chemical alterations occur, significantly influencing the characteristics of the pores.Diagenesis compaction various processes, most notably compaction, cementation, and dissolution. 91he stratum in the study area has been buried to a depth of 3778 m and has undergone significant compaction during the diagenesis process.This compaction leads to a tight arrangement of detritus particles, while plastic particles deform and exert pressure on the primary pores, resulting in a substantial reduction in primary pore space and distribution of pores.It is challenging to observe the original intergranular pores in both Types Ⅰ and Ⅱ samples, and compaction plays a critical role in this.Scanning electron microscopy observations reveal that under the influence of compaction, rigid particles generate a series of microcracks.These microcracks enhance the pore space of tight reservoirs, which is particularly significant for the storage and penetration of both types of samples.Microcracks serve as important pore spaces, especially for Type Ⅰ samples with fewer dissolution pores, and have a more pronounced impact on reservoir storage and permeability.Therefore, compaction not only destroys a significant number of primary pores but also generates microcracks that have a more noticeable impact on Type Ⅰ samples.
Cementation fills the pores and leads to pore damage.However, certain types of cement can generate intercrystalline pores, thereby enhancing reservoir porosity.The He 8 Member reservoir primarily consists of carbonate cement and clay mineral cement, with some siliceous cement.The results of observing thin sections scanning electron microscopy indicate that the tight sandstone in the He 8 Member fills various primary and secondary pores.It is challenging to observe intergranular and dissolution pores in Type Ⅰ samples, as cement filling is an important destructive factor.While some dissolution pores remain in Type Ⅱ samples, most are also destroyed by compaction or cement.While some cement can produce intercrystalline pores with small pore diameters, but this has a specific storage significance for tight sandstones with fewer large pores.Among them, intercrystalline pores formed by kaolinite are crucial, and beneficial intercrystalline pores can generated by chlorite and illite.As shown in Figure 8, the volume and surface area of nanopores increases with the clay mineral content.Therefore, it can be concluded that clay mineral cement like kaolinite and illite can improve the storage and permeability of pores.In summary, in this study area, cementation leads to the destruction of large pores but also generates some small pores, resulting in an overall improvement in reservoir porosity.
The He 8 Member exhibits significant dissolution characteristics, resulting in the formation of dissolution pores through the dissolution of particles, such as feldspar and rock cuttings.As diagenesis advances, some of these dissolution pores are filled with clay or carbonate minerals, making their preservation challenging.However, it is necessary for some dissolution pores to be cemented and sealed to enhance the storage and permeability of the reservoir.The observation results indicate that Type Ⅱ samples often retain certain dissolution pores, which lead to have a larger diameter.On the basis of the collected data, it can be deduced that the dissolution effect has a more pronounced impact on enhancing the Type Ⅱ samples, leading to the formation of virtual storage spaces.Conversely, Type Ⅰ samples display minimal residual dissolution pores, suggesting that dissolution has limited constructive effect on Type Ⅰ samples.Nevertheless, the presence of clay minerals within the dissolution pores of Type Ⅰ samples results in the formation of intercrystalline pores, which have a positive influence on the overall porosity.

| The influence of mineral composition on pore development (1) Rock fragments
The dissolution of rock fragments can generate secondary pores, thereby enhancing the physical properties of the reservoir.In the He 8 Member tight sandstone, the rock fragment content ranges from 12% to 55%, with higher percentages indicating a greater presence of rock fragments.Observations reveal that extensive dissolution of rock fragments has occurred, resulting in significant improvements to the reservoir's physical properties.The dissolution of rock fragments is particularly prominent in Type Ⅱ samples and represents the most significant aspect of macroporosity.However, dissolution of rock fragments in Type Ⅰ samples is challenging to observe, primarily due to the low pore content during diagenesis and the limited flow of acidic solutions.Figure 9 shows a cross plot between the content of rock fragments and the ratio of dissolved pore surface area.It can be observed that the ratio of dissolved pore surface area tends to increase as the content of rock fragments decreases.The extensive dissolution of rock fragments, which generates dissolution pores, also confines the range of cuttings.| 131 from 36% to 67%, with an average of 52.4%, whereas Type Ⅱ samples exhibit a higher quartz content ranging from 67% to 76%, with an average of 72.2%.Therefore, Type Ⅱ samples exhibit a greater quartz content compared to Type Ⅰ samples.The increased quartz content aids in resisting compaction and facilitates subsequent dissolution processes.Additionally, observations reveal the presence of fractures in rigid quartz particles in both types of samples.These microfractures serve as conduits for various pores, ultimately enhancing the storage capacity and permeability of the reservoir.
(3) Clay minerals Clay minerals have the ability to generate more nanoscale pores, which in turn enhance the performance of the reservoir.The He 8 Member primarily comprises kaolinite and illite as its clay mineral constituents.The clay mineral content of Type Ⅰ samples ranges from 0% to 13%, with an average of 4.3%, while for Type Ⅱ, it ranges from 4% to 6%, with an average of 4.6%.There is not much variation in the clay minerals content between the two sample types.As mentioned previously, an increase in clay mineral content leads to a significant increase in the specific surface area and volume of nanopores in both sample types.This indicates that clay minerals play a significant role in improving nanopore characteristics.

| CONCLUSIONS
In the Gaoqiao area, the tight sandstone pores of the He 8 Member primarily comprise dissolution pores and intercrystalline pores, occasionally accompanied by intergranular pores and microcracks.Among the samples analyzed, are corrosion and intergranular pores, with fewer intergranular pores and microcracks.Type II samples predominantly exhibit corrosion and intergranular pores, with a lesser presence of intergranular pores and microcracks.Conversely, Type I samples predominantly display micropores and intergranular pores, with occasional occurrences of microcracks in certain areas.The pore size distribution analysis of tight sandstone in this area was carried out by splicing the pore size characteristics obtained from HPMI and LTNA experiments at 30 nm.The specific surface area and volume of nanoscale pore in tight sandstone are significantly large, which is beneficial for the development of sandstone gas reservoirs.Fractal analysis reveals that Type Ⅰ samples generally exhibit a higher fractal dimension, indicating a more complex pore morphology.The distribution of pores in tight sandstone is influenced by factors, such as sedimentary environment, diagenesis, and mineral composition.Compaction processes primarily result in the destruction of primary pores and the formation of microcracks, which notably impact the characteristics of Type Ⅰ samples.Cementation processes lead to the destruction of macropores while also creating some nanoscale pores.Dissolution processes significantly enhance the porosity of Type Ⅱ samples.Additionally, rock fragments contribute to the formation of dissolution pores, while quartz exhibits resistance to compaction and generates microcracks.Clay minerals play a dual role by blocking larger pores and producing intercrystalline pores that are favorable for reservoir storage.

CHEN ET AL. | 121 F
I G U R E 2 (See caption on next page).

F
I G U R E 3 (A) Pore size distribution diagram of Type Ⅰ samples.(B) Pore size distribution diagram of Type Ⅱ samples with permeability.Remarks: One sample's high-pressure mercury intrusion test in Type Ⅰ failed.

F I G U R E 4
Adsorption and desorption curves of representative samples.(A) Typical adsorption and desorption curve for Type Ⅰ samples (S 338 well, 3598.24m).(B) Typical adsorption and desorption curve for Type Ⅱ samples (G 68-12 well, 3163.94m).F I G U R E 5 (A) Relationship between Brunauer-Emmett-Teller-specific surface area and Barrett-Joyner-Halenda pore volume distribution of Type Ⅰ samples.(B) Relationship between Brunauer-Emmett-Teller-specific surface area and Barrett-Joyner-Halenda pore volume distribution of Type Ⅱ samples.

F
I G U R E 6 (A) Fractal characteristics of pores less than 30 nm in Sample No.1.(B) Fractal characteristics of pores less than 30 nm in Sample No.2.(C) Fractal characteristics of pores less than 30 nm in Sample No.3.(D) Fractal characteristics of pores less than 30 nm in Sample No.4.(E) Fractal characteristics of pores less than 30 nm in Sample No.5.(F) Fractal characteristics of pores less than 30 nm in Sample No.6.(G) Fractal characteristics of pores less than 30 nm in Sample No.7.(H) Fractal characteristics of pores less than 30 nm in Sample No.8.(I) Fractal characteristics of pores less than 30 nm in Sample No.9.(J) Fractal characteristics of pores less than 30 nm in Sample No.10.than 30 nm

F
I G U R E 7 (A) Fractal characteristics of pores larger than 30 nm in Sample No.1.(B) Fractal characteristics of pores larger than 30 nm in Sample No.2.(C) Fractal characteristics of pores larger than 30 nm in Sample No.3.(D) Fractal characteristics of pores larger than 30 nm in Sample No.5.(E) Fractal characteristics of pores larger than 30 nm in Sample No.6.(F) Fractal characteristics of pores larger than 30 nm in Sample No.7.(G) Fractal characteristics of pores larger than 30 nm in Sample No.8.(H) Fractal characteristics of pores larger than 30 nm in Sample No.9.(I) Fractal characteristics of pores larger than 30 nm in Sample No.10.T A B L E 3 Pore fractal dimension (FD) of tight sandstone greater than 30 nm.

( 2 )
Quartz Rigid quartz particles possess the ability to withstand compaction, resulting in reduced loss of reservoir pores and facilitating the dissolution process during the later stages of diagenesis.The quartz content of Type Ⅰ samples ranges F I G U R E 8 Cross plot of clay mineral content with the specific surface (A) and pore volume (B).
CHEN ET AL.