Total Pore–Throat Size Distribution Characteristics and Oiliness Differences Analysis of Different Oil-Bearing Tight Sandstone Reservoirs—A Case Study of Chang6 Reservoir in Xiasiwan Oilfield, Ordos Basin

Abstract

In the observation of tight sandstone cores, the variations in the hydrocarbon charging usually can be observed in the same geological age reservoirs, which manifest as differential oil staining on the core surface. In order to clarify the micro total pore–throat size distribution (TPSD) characteristics and oil content differences of different oil-bearing tight reservoirs, we drilled two types of oil-bearing cores in the Chang6 formation of Xiasiwan Oilfield, conducted casting thin section (CTS), scanning electron microscopy (SEM), and X-ray diffraction (XRD) to qualitatively and quantitatively analyze petrological and pore–throat characteristics. The TPSD of different oil-bearing cores were quantitatively characterized and compared by combining high-pressure mercury injection (HPMI) and constant rate mercury injection (CRMI). Meanwhile, we quantitatively evaluated the complexity of the pore–throat structure based on fractal theory. Our results reveal significant difference in the clay mineral contents between the two types of cores, despite both being classified as arkose. Due to higher contents of illite, calcite, and chlorite, the pores of oil-smelling sandstone are obviously affected by cementation. The result of TPSD characteristics shows that the oil-appearing sandstone samples exhibit well-developed big pores and throats, displaying bimodal distribution, and three-stage fractal characteristics in the TPSD curves. Conversely, oil-smelling sandstone samples manifesting a left-skewed bimodal, pore space contribution of the samples is more dependent on pores and throats smaller than 0.12 μm. The TPSD curves exhibit three-stage and four-stage fractal characteristics. Therefore, the differences in oil-bearing properties between the two types of cores are attributed to variations in mineral composition, diagenesis, clay mineral content, pore types, pore–throat size distribution (PSD), and pore–throat complexity. Our results provide crucial guidance for subsequent reservoir quality assessment in this study area and the development of tight sandstone reservoirs with similar geological characteristics.

1. Introduction

The ongoing paradigm shift in the global energy landscape, driven by the persistent decline of conventional hydrocarbon resources, has profoundly reshaped global strategies for energy exploration and development worldwide. The tight sandstone hydrocarbon resources as important unconventional energy sources, are progressively becoming pivotal components of the global energy supply [1,2,3,4]. Significant advancements have been achieved in the exploration and development of unconventional tight sandstone hydrocarbon resources within the major sedimentary basins of China, including the Ordos Basin, Junggar Basin, Songliao Basin, and Bohai Bay Basin. Great progress has taken place in the developing technology of unconventional tight sandstone hydrocarbon resources, which plays a significant role in enhancing domestic oil and gas production capacity [4,5,6,7,8]. Tight sandstone reservoirs are typically characterized by low to ultra-low porosity (less than 15%), low to ultra-low-permeability (less than 1 × 10⁻³ μm²), inherently low formation energy, pronounced heterogeneity, and complex pore–throat structures [9,10,11,12]. Such reservoirs usually demonstrate brief natural production periods, necessitating rapid transition to waterflooding development. In the process of waterflooding, the pore–throat structure of the reservoir directly impacts the water flooding effect of injected water [13,14,15]. Therefore, this study on the micropore–throat structure of reservoirs constitutes a pivotal step in guiding the later development of such oil and gas fields. It can provide critical insights into oil–water displacement patterns during the development period and offer theoretical foundations for residual oil characterization in the late development stage [16,17,18,19].

The continuous technological evolution in reservoir characterization has driven rapid innovation in methods for analyzing pore-throat structures [20,21]. The early-stage research of reservoirs, usually using through CTS and SEM, only enabled the qualitative assessment of pore–throat structures, which has disadvantages in quantitative research [22,23,24]. Then, quantitative and semi-quantitative experimental approaches, including mercury injection capillary pressure (MICP), nuclear magnetic resonance (NMR), micro-nano computed tomography (CT), and nitrogen adsorption (NA), have been increasingly integrated into the characterization of micropore–throat structures [25,26,27]. At the present stage, the traditional single method for qualitative and quantitative characterization exhibits inherent limitations in terms of characterization scope and analytical precision. Thus, integrated experimental methodologies need be provided to achieve more refined characterization of pore–throat structures [28,29,30,31]. In recent studies, Song, L. et al. (2017) [27] and Zhang, Q. et al. (2020) [32] established complete PSD curves of reservoir rocks through the integrated application of HPMI and CRMI, achieving the characterization of TPSD structures. Xiao, D. et al. (2016) [33] and Qu, Y. et al. (2020) [34] demonstrated that CRMI exhibits distinct advantages in characterizing macropores to comprehensively characterized PSD of reservoir rocks and employed fractal theory to quantitatively assess the complexity of pore–throat structures. Zhang, Q. et al. (2022) [35] and He, T. et al. (2024) [36] employed an integrated characterization approach combining NA, HPMI, and NMR techniques to achieve total pore–throat characterization of rock samples and integrated fractal theory with experimental data and established a quantitative evaluation method for both reservoir heterogeneity and the structural complexity of pores and throats. Fractal theory has emerged as a well-established mathematical approach and has been extensively applied across multiple disciplines and research fields. Peta, K. et al. (2025) [37] integrated fractal theory to conduct a mathematical comparison of surface texture measurements between hardened clay samples and pottery fragments and further described the surface topography of archeological clay with wear marks using fractal shapes. Xia, B. et al. (2023) [38] and Cai, J. et al. (2023) [39] applied fractal theory to characterize the development characteristics of fracture networks in coal seams and further established a fractal permeability model that incorporates complex curved fracture attributes to predict coal seam permeability. Guo, R. et al. (2020) [40] and Zhang, H. et al. (2024) [41] developed a mathematical model for permeability prediction in tight sandstone reservoirs by integrating fractal theory. This model incorporates pore–throat structure complexity via fractal dimensions during permeability estimation. Yang, H. et al. (2025) [42] integrated fractal theory with rock mechanics to develop the fractal model of the rock cutting mechanical properties, which provides a foundation for the optimization of the fracturing construction.

In this study, we investigated the tight reservoir characteristics of the Chang6 formation in the Xiasiwan Oilfield through systematic analysis of oil-bearing sandstone core samples. The collected samples were classified into two distinct categories based on their oil manifestation characteristics. The first group exhibited a higher degree of oil charging intensity, demonstrating visible oil immersion on sandstone core surfaces. The samples from the first group were designated as oil-appearing sandstone. The other type could be classified as oil-smelling sandstone because of the lack of macroscopic oil traces and a pronounced petroleum odor upon surface exposure. According to the results of the integrated application of CTS analysis and XRD techniques, we began with systematic petrographic characterization of two types of oil-bearing sandstone cores, specifically, the detailed assessment of cement/interstitial material types and content. Subsequently, SEM and CTS analytical techniques were employed to conduct the qualitative identification of mineralogical compositions, pore–throat category, and cement characteristics in two distinct categories of oil-bearing sandstone cores. By integrating the experimental results from HPMI and CRMI techniques, a comprehensive quantitative characterization of the total pore–throat structural characteristics was systematically conducted for two types of oil-bearing sandstone cores. Leveraging fractal theory, this study conducts comprehensive research on the total pore–throat fractal characteristics of sandstone cores. Fractal dimensions are used to quantitatively characterize the geometric similarity and structural complexity of micropore–throat structures within the target reservoirs. The distinction between two categories of oil-bearing sandstone cores was identified in petrological/mineralogical studies, which is mainly reflected in interstitial materials, pore types, pore–throat structure, and pore–throat complexity. We also elucidate the micro structural difference between two types of oil-bearing sandstone cores through various testing experiments, to interpret the phenomenon of oiliness difference in the tight sandstone reservoir from the Micro perspective. Our results can provide guiding suggestions for later reservoir quality evaluation and tight reservoir exploitation.

2. Geological Setting and Samples

The Ordos Basin, geographically situated in central-western China (Figure 1a), represents a large-scale continental sedimentary basin formed through Mesozoic multi-stage superimposed sedimentation, tectonically characterized by a regional monoclinal structure with overall eastern uplift and western depression [43,44,45]. The Ordos Basin exhibits subdued internal structures characterized by an absence of large-scale faults and anticlines, while demonstrating pronounced structural features along its margins. Based on contemporary structural characterization, this cratonic basin can be tectonically partitioned into six first-order tectonic units: the Yimeng uplift, Weibei uplift, Jinxi fault-fold Belt, Yishan slope, Tianhuan depression, and Western thrust belt (Figure 1b).

The Chang7 periods of the Triassic Yanchang formation represent the peak period of organic matter deposition in the basin’s evolutionary history, whereas the studied Chang6 periods are situated in the early phase of depositional decline. This stratigraphic layer is characterized by braided river delta front deposits, where hydrocarbon migration and accumulation are predominantly controlled by sandstone developed in distributary channels [46,47,48,49,50]. The vertical lithostratigraphic profile demonstrates alternating sequences of gray fine-grained sandstone, grayish-black sandy mudstone, and black mudstone, with the Chang7 Member developing organic-rich black oil shale as the principal source rock, while the sandstone intervals serve as effective hydrocarbon reservoirs within the study area (Figure 1d) [51]. Core samples of oil-bearing sandstone from the Chang6 Member (Figure 1c) were collected from six cored wells in the study area, all exhibiting varying content of oil charging. These sampling intervals were log-interpreted as the oil–water layer or poor oil layer. The oil-appearing sandstone intervals demonstrate obvious oil stains, as evidenced by well-logging data revealing distinct oil stains and visible oil show indications; oil-smelling sandstone intervals show limited charging degrees, lacking visible oil stains but exhibiting a detectable oil odor in the cores (Figure 2a). In HPMI testing, tested porosity values range from 6.33% to 13.98%, with a mean of 9.06%, while permeability measurements span from 0.003 × 10⁻³ μm² to 1.154 × 10⁻³ μm², averaging 0.33 × 10⁻³ μm². These petrophysical parameters conclusively identify the formation as a typical ultra-low porosity and ultra-low-permeability tight sandstone reservoir.

Figure 1. (a) Location of the Ordos Basin in China; (b) location of the study area in the Ordos Basin (red rectangle); (c) Location of the sampled wells; (d) comprehensive stratigraphic characteristics of Southeast Ordos Basin (modified from reference [51]).

Figure 1. (a) Location of the Ordos Basin in China; (b) location of the study area in the Ordos Basin (red rectangle); (c) Location of the sampled wells; (d) comprehensive stratigraphic characteristics of Southeast Ordos Basin (modified from reference [51]).

Figure 2. (a) Schematic diagram of samples in different oil-bearing sandstone; (b) schematic diagram of experimental design.

Figure 2. (a) Schematic diagram of samples in different oil-bearing sandstone; (b) schematic diagram of experimental design.

3. Experimental Methods and Methodology

Cylindrical sandstone samples (10–15 cm in length × 2.5 cm in diameter) were drilled parallel to bedding planes from two obvious oil-bearing lithofacies within the Chang6 reservoir in the Xiasiwan field. These samples were obtained across six representative wells, targeting the oil-bearing strata of the Chang6 Member. The study focused on two types of oil-bearing lithofacies, where a total of 36 tight sandstone samples (including cylindrical sandstone cores and flake-shaped sandstone cores) were subjected to experimental analysis. Prior to experimental procedures, all selected samples underwent solvent extraction using a benzene–methanol mixture to remove indigenous oil, asphaltenes, and other organic matter from the rock pores. This preparatory treatment effectively eliminated residual petroleum components that might interfere with subsequent experimental measurements, thereby ensuring data reliability during the characterization of reservoir petrophysical properties and fluid–rock interactions [52]. After undergoing sectioning and solvent extraction treatments, experimental design including CTS, SEM, XRD, HPMI, and CRMI were correspondingly conducted under sufficient sample availability conditions (Figure 2b). The experimental design employed in this study which the parallel samples from one specimen were used in each experiment, effectively minimized the impact of inherent heterogeneity on the results.

3.1. Experimental Methods

3.1.1. X-Ray Diffraction

XRD analysis in this study comprises two distinct components: whole-rock analysis and clay mineral characterization. As any crystalline substance has a specific crystal structure, different crystal structures will form different diffraction characteristics, which can quantitatively explain the composition and content of rock minerals and clay minerals. Depending on the requirements of different experiments, samples were processed through differential grinding protocols using an agate mortar. For whole-rock analysis, approximately 3 g of desiccated sample material was pulverized to achieve a particle size of 300 mesh (corresponding to 48 μm particle diameter). In clay mineral characterization procedures, larger aliquots (10 g dry weight) were subjected to extended grinding cycles to attain finer particulates below 200 mesh (<74 μm particle diameter). Analyze the diffraction pattern of rocks by performing X-ray diffraction on the SmartLab X-ray diffractometer manufactured by Rigaku corporation in Tokyo, Japan. Because the intensity in the diffraction pattern is related to the content of the mineral in the sample, comparing the measured pattern with the characteristic spectrum of the standard mineral can determine the relative mass content of each mineral in the samples.

3.1.2. High-Pressure Mercury Injection

The HPMI samples use a cylindrical sample with a diameter of 2.5 cm and a length of about 5 cm. Before testing, the oil in the sample is treated with a mixture of benzene and methanol and its physical properties are measured. The capillary pressure curves of the samples were determined using the AutoPore IV 9505 automatic mercury porosimeter manufactured by Micromeritics Instrument Corporation in the Norcross, GA, USA, in accordance with the relevant specifications [53]”. This advanced porosimetry device enables the quantitative characterization of PSD spanning 0.001 to 630 μm through controlled mercury intrusion under precisely regulated pressure conditions, with maximum injection pressure capacity reaching 200 MPa. The testing process adopts segmented injection of non-wetting phase mercury. During this process, the injection pressure progressively overcomes capillary resistance within the rock matrix. When dynamic equilibrium is established between the applied pressure and capillary forces, the corresponding mercury intrusion volume at that specific pressure level is measured. This quantified mercury volume represents the cumulative pore–throat volume governed by the prevailing pressure conditions. Following the acquisition of mercury intrusion data across multiple pressure increments, the Washburn equation [54] is employed to mathematically convert the pressure-dependent intrusion measurements into equivalent pore–throat radius values. The fundamental relationship derives from the capillary pressure equation:

$$r={\displaystyle \frac{-2\sigma \mathit{cos}\theta }{P_c}}$$

(1)

where r is the pore–throat radius, μm; P_c is the mercury pressure, MPa; σ is the interfacial tension between mercury and air, 485 mN/m; and θ is the wetting contact angle, 140°.

3.1.3. Constant Rate Mercury Injection

The CRMI experiment also utilized cylindrical sandstone core samples with a diameter of 2.5 cm. Following solvent extraction, the prepared core sample was saturated with mercury through ultra-low injection rate displacement (10⁻⁶ mL/s). As a non-wetting phase relative to rock surfaces, mercury preferentially invades pores and throats during injection, with the advancing front exhibiting concave morphology while capillary pressure progressively increases. Upon breakthrough into interconnected pores and throats, the mercury volume undergoes redistribution under quasi-static conditions, accompanied by an abrupt decline in capillary pressure. Subsequent injection induces gradual pressure recovery until maximum values are attained at full mercury saturation within the pore space. This pressure maximum is followed by another abrupt pressure drop when the non-wetting phase penetrates subsequent throats. Through repeated cycling of this pressure-dependent invasion pattern, mercury progressively infiltrates smaller pores and throats until ultimate saturation occurs throughout the rock specimen. The pore–throat structure was characterized by continuously monitoring mercury injection pressure and intrusion volume during experiments, enabling the differentiation of pore and throat volumes and testing of the structural parameters. The ASPE-730 constant rate mercury injection instrument, manufactured by Coretest Corporation in the San Jose, CA, USA, was used for the CRMI experiments, with the maximum mercury injection pressure of 900 psi (approximately 6.2 MPa), and the minimum radius of the samples measured was 0.12 μm.

3.2. Fractal Theory

Fractal theory, initially proposed by French mathematician Benoit Mandelbrot, provides a mathematical method to characterize irregular natural objects exhibiting self-similar structural patterns across multiple scales [55,56]. This theory and method have been extensively applied in petroleum geology research, particularly in addressing critical challenges such as fracture network prediction in reservoirs and the quantitative evaluation of pore–throat structure heterogeneity micro-complexity in rock cores [57,58]. Fractal dimension (D) serves as a critical parameter for quantitatively characterizing the complexity and heterogeneity of micropore–throat structures in reservoirs. Previous investigations have demonstrated that the D of reservoir pore–throat systems typically ranges between 2 and 3, where the closer the D value is to 2, the stronger the reservoir homogeneity, and the closer the D value is to 3, the stronger the heterogeneity and complexity [59]. Guided by the fundamental fractal principle of power-law scaling between object quantity and the measurement scale [55,56], the relationship between pore–throat count and pore–throat radius in reservoir rock systems can be mathematically described as follows:

$$N\left(r\right)=a{r}^{-D}$$

(2)

N(r) is the number of pores–throats with radii r; a is the fractal coefficient; and D is the fractal dimension.

The widely used Brooks–Corey model [60] approximates the pore–throat system within reservoir rocks as an assemblage of capillary tubes with varying radii. Within this assumption, the cumulative pore–throat volume corresponding to a specific radius r can be mathematically represented as follows:

$$V\left(r\right)=N\left(r\right)·\pi r^{2}l=a\pi r^{2-D}l$$

(3)

l is the length of the capillary tube, μm.

The self-similarity of the pore–throats determines that the ratio of the radius to the length of the capillary is a constant; thus, Equation (3) can be rewritten as

$$V\left(r\right)=a\pi r^{3-D}$$

(4)

Differentiate Equation (4) and calculate the derivative:

$${\displaystyle \frac{dV(r)}{dr}}=a(3-D)\pi r^{2-D}$$

(5)

Within the pore–throat radius interval r_min < r < r_max, the cumulative volume percentage V(r) corresponding to pore–throats with radii smaller than r can be mathematically expressed as follows:

$$V\left(r\right)={\displaystyle \frac{V(<r)}{V(>r)}}={\displaystyle \frac{{\int }_{r_{min}}^{r}a(3-D)\pi r^{2-D}dr}{{\int }_{r_{min}}^{r_{max}}a(3-D)\pi r^{2-D}dr}}={\displaystyle \frac{r^{3-D}-r_{min}^{3-D}}{r_{max}^{3-D}-r_{min}^{3-D}}}$$

(6)

r_max is the maximum value of the pore–throat radius, and r_min is the minimum value of the pore–throat radius, μm. For any capillary tube bundle model, r_max >> r_min exists, so Equation (6) can be rewritten as

$$V(r)\approx {({\displaystyle \frac{r}{r_{max}}})}^{3-D}$$

(7)

By taking the logarithm of both sides of Equation (7), we obtain

$$\mathit{lg}V\left(r\right)=(3-D)(\mathit{lg}r-\mathit{lg}{r}_{max})$$

(8)

As demonstrated by Equation (8), the presence of a linear correlation between logV(r) and logr in a double logarithmic coordinate system indicates that the PSD in reservoir rocks conforms to fractal theory. The fractal dimension D, which quantitatively characterizes the heterogeneity of pore–throat structures, can be determined through linear regression analysis by calculating the slope of the fitted trend line.

4. Results

4.1. Petrological and Pore–Throat Characteristics

The mineralogical composition and pore–throat characteristics of two types of different oil-bearing samples were investigated through an integrated analytical approach employing CTS, SEM, and XRD. Based on the analytical results from XRD presented in Figure 3, the terrigenous clastic mineral composition in both oil-appearing sandstone and oil-smelling sandstone demonstrates a distinct mineralogical composition dominated by quartz and feldspar constituents, with interstitial material exhibiting a substantial volumetric contribution. The oil-appearing sandstone demonstrates a quartz content ranging from 17.9% to 36.5% (average 26.1%) and feldspar content varying between 41.8% and 68.1% (mean 56.07%). Clay mineral constituents constitute 10.1–17.1% of the lithology (average 13.6%), with the clay minerals primarily composed of illite, chlorite, and illite/smectite mixed-layer minerals (I/S). The oil-smelling sandstone exhibits the following mineralogical composition: quartz content ranges from 22.7% to 44% (average 31.2%), while feldspar content varies between 40.6% and 54.7% (average 49.5%). Clay mineral constituents constitute 14.3–19.7% of the bulk composition (mean: 16.3%), predominantly comprising illite and chlorite. Notably, sample W200-01 demonstrates obvious development of I/S, indicative of specific diagenetic conditions. The average calcite content in oil-appearing sandstone and oil-smelling sandstone is 2.5% and 2.9%, respectively. Calcite cementation is frequently observed in the microscope of CTS. Based on the Folk’s sandstone nomenclature diagram constructed from XRD analytical results for sandstone classification (Figure 3b), the two types of oil-bearing core samples demonstrate similar lithological characteristics, with both predominantly classified as arkose.

Based on the CTS and SEM results, the study conducted qualitative analyses of pore–throat characteristics, detrital compositions, and clay mineral types in core samples with different oil-bearing properties. Analytical results demonstrate that primary pores in two types of oil-bearing cores experienced differential destruction during diagenetic evolution, with obvious pore volume reduction primarily attributable to compaction and clay mineral cementation. Oil-appearing sandstone primarily contains intergranular pores as its dominant pore type. However, the original intergranular pores were substantially destroyed by ferriferous calcite cementation (Figure 4a), and chlorite-infilled residual intergranular pores are obvious (Figure 4b). Under the action of dissolution, the feldspar develops secondary dissolution pores of limited dimensions (Figure 4c). Acidic fluids preferentially infiltrate along cleavage planes of feldspar minerals, generating secondary dissolution pores and dissolution fractures within feldspar grains (Figure 4d,e). Although these micro fractures exhibit restricted spatial extension and make relatively minor contributions to overall porosity and permeability compared to other pore types, they serve as critical migration pathways for fluid transport. This enhanced fluid mobility facilitates localized dissolution phenomena through accelerated acidic fluid migration (Figure 4d). In oil-smelling sandstone samples, obvious cementation is observed between detrital minerals and clay minerals. The CTS analysis of the tested samples reveals substantial occlusion of primary intergranular pores by ferriferous calcite cementation (Figure 4f), with localized preservation of residual intergranular pores (Figure 4g). The clay minerals predominantly included chlorite and illite. Petrographic observations reveal two distinct morphological types of chlorite cementation: (1) leaf lamellar chlorite exhibiting pore-filling cementation within intergranular pores (Figure 4h), and (2) acicular illite adhering to detrital grain surfaces (Figure 4i). These authigenic clay minerals collectively contribute to substantial pore reduction, resulting in the serious occlusion of primary intergranular pores. Compared to oil-appearing sandstone, oil-smelling sandstone exhibits more obviously destructive diagenetic alteration, characterized by obviously higher depletion of primary pores and substantially reduced dimensions of primary pores.

4.2. Characteristics of Pore–Throat Structure Determined by HPMI

HPMI experiments were conducted to obtain capillary pressure curves and PSD curves for two types of oil-bearing core samples. The characteristic pore–throat structural parameters derived from these experiments are systematically presented in

Table 1. Pore–throat structure parameters obtained from HPMI and CRMI.

HPMI													CRMI
Sample Types	Sample ID	Well	Depth/m	K (10−3 μm2)	Φ (%)	Pcd (MPa)	Rmax (μm)	Ra (μm)	R50 (μm)	Sp	Smax (%)	We (%)	Pcd (MPa)	Rt (μm)	Rp (μm)	Smax (%)
Oil-appearing sandstone	W201-01	W201	891.12	1.154	13.978	0.138	5.331	0.974	0.636	2.841	98.561	38.599	0.232	2.128	138.07	75.39
	B7-01	B7	824.3	0.023	8.556	0.674	1.091	0.244	0.151	2.383	96.164	37.008	0.219	2.149	146.85	71.97
	B5-01	B5	766.32	0.652	7.904	1.367	0.538	0.129	0.099	1.813	95.296	22.123	0.372	1.206	125.19	51.93
	Average			0.610	10.146	0.726	2.320	0.449	0.295	2.346	96.674	32.577	0.274	1.828	136.70	66.43
Oil-smelling sandstone	B1-01	B1	774.86	0.008	6.338	5.498	0.134	0.033	0.025	1.544	93.223	31.087	3.684	7.5	205	15.68
	B3-03	B3	768.35	0.097	8.314	4.129	0.178	0.047	0.038	1.479	95.746	26.095	3.613	9.412	157.65	20.38
	W200-01	W200	959.09	0.003	6.650	2.052	0.358	0.078	0.018	2.234	89.321	19.558	2.963	0.184	115.03	25.25
	Average			0.036	6.101	3.893	0.223	0.053	0.027	1.752	92.763	25.580	3.420	5.699	159.23	20.44

Note: K: Permeability; Φ: Porosity; P_cd: Threshold Pressure; R_max: Maximum Pore–Throat Radius; R_a: Average Pore–Throat Radius; R₅₀: Median Pore–Throat Radius; S_p: Sorting Coefficient; S_max: Maximum Mercury Saturation; W_e: Mercury Removal Efficiency; R_t: Throat Radius; R_p: Pore Radius.

. The experimental results demonstrate that the average pore–throat radius (R_a) of the oil-bearing samples ranges from 0.033 μm to 0.974 μm (mean 0.251 μm), with the maximum radius of fine pores reaching 5.331 μm. The maximum mercury injection saturation exhibits values between 89.0% and 98.6%. It predominantly exhibits well-developed fine-sized and micropores and throats. Notably, post-diagenetic preservation has maintained partial pore–throat connectivity within these reservoir pores despite distinct diagenetic alteration.

Comparative analysis of two types of cores’ experimental results revealed distinct petrophysical characteristics in oil-bearing sandstone samples. The oil-appearing sandstone samples demonstrated a mean porosity of 10.146% (arithmetic average) coupled with an average permeability of 0.61 × 10⁻³ μm², exhibiting typical low-permeability reservoir properties. The capillary pressure curve is characterized by an extended smooth segment that remains entirely below a mercury injection pressure of 10 MPa (Figure 5a). Analytical results demonstrate that the threshold pressure (P_cd) of the samples ranges from 0.138 to 1.367 MPa, yielding an average value of 0.726 MPa. Notably, the maximum mercury saturation (S_max) attains values exceeding 95% across all samples. The PSD curves predominantly exhibit right-skewed unimodal and bimodal patterns (left-high and right-low), as illustrated in Figure 5b. Mercury intrusion analysis reveals that the total pore volume is contributed by pores and throats within the 0.09~1.56 μm range, indicating a relatively concentrated distribution. The mean pore–throat radius (R_a) demonstrates obvious variability across samples, ranging from 0.129 μm to 0.974 μm (mean 0.449 μm). The type of cores exhibits predominant development of residual intergranular pores, feldspar dissolution pores, and feldspar dissolution fractures, demonstrating relatively low cementation intensity within the pore–throat system (Figure 5c).

The analysis of oil-smelling sandstone samples revealed average petrophysical properties characterized by a porosity of 6.101% and permeability measuring 0.036 × 10⁻³ μm². This tight reservoir exhibits both low porosity and ultra-low-permeability characteristics. The capillary pressure curves exhibit a prolonged smooth segment, demonstrating higher smooth intervals compared to oil-appearing sandstone samples. Notably, portions of these curves extend beyond mercury injection pressures of 10 MPa (Figure 5d). The threshold pressure (P_cd) ranges from 2.052 to 5.498 MPa, with an average value of 3.893 MPa. This higher-pressure range indicates obviously greater resistance to mercury intrusion in oil-smelling sandstone samples. The analyzed samples exhibited lower average values in both maximum mercury intrusion saturation (S_max) (92.76%) and mercury withdrawal efficiency (W_e) (25.58%) when compared to oil-appearing sandstone counterparts. The PSD curves of samples B3-03 and B1-01 predominantly exhibit left-skewed unimodal patterns, whereas sample W200-01 demonstrates a multimodal distribution. Mercury intrusion analysis reveals that the majority of pore volume (Figure 5e) is concentrated in fine pores with diameters below 0.1 μm. Quantitative measurements indicate an average pore–throat radius ranging from 0.033 μm to 0.078 μm across the samples, yielding a mean value of 0.053 μm. Under micro examination, representative samples exhibit obvious cementation, with the predominant pore types being feldspar dissolution micropores and intercrystalline pores. Additionally, compaction has obviously modified the primary pores of the samples (Figure 5f).

Integrated analysis reveals obvious differences in pore–throat structure characteristics between the two oil-bearing sandstone cores, despite both exhibiting well-developed micro-scale pores and throats. The oil-appearing sandstone demonstrates superior reservoir quality compared to the oil-smelling sandstone, manifested through enhanced physical properties (including porosity and permeability), more favorable pore type distributions with larger average pore sizes, and better preserved pore–throat connectivity. When reservoirs contain both types of sandstone, oil as non-wetting phase, preferentially migrates through and accumulates within the oil-appearing sandstone during petroleum migration and accumulation processes.

4.3. Characteristics of Pore–Throat Structure Determined by CRMI

The CRMI is a petrophysical characterization technique that employs quasi-static constant rate injection of non-wetting phase mercury into rock core samples. Through the continuous monitoring of capillary pressure evolution during the experimental process, this method enables distinct differentiation between pores and throats [62,63]. Based on CRMI analysis, the threshold pressure (P_cd) of oil-appearing sandstone samples ranged from 1.206 MPa to 2.149 MPa, with a mean value of 0.372 MPa. Maximum mercury intrusion saturation (S_max) demonstrated variation between 51.93% and 75.39%, averaging 66.43%. Pore–throat structure analysis yielded a mean throat radius (R_t) of 1.828 μm and an average pore radius (R_p) of 136.7 μm (Figure 6a–c and Table 1). The mercury intrusion curves for pores and throats exhibited long profiles in the three oil-appearing sandstone samples, with sample B5-01 demonstrating a particularly notable feature. A prolonged overlap was observed between the pore mercury intrusion curve and the total mercury intrusion curve in sample B5-01, suggesting a higher prevalence of well-developed larger pores. Analysis of the PSD curves (Figure 7a) reveals a quasi-normal distribution pattern with distinct morphological characteristics. Specifically, samples W201-01 and B7-01 exhibit bimodal distributions characterized by left-skewed asymmetry (higher left peak and lower right peak), while sample B5-01 demonstrates a unimodal distribution with leftward skewness. The pore–throat radius within the oil-appearing sandstone samples exhibits a multimodal distribution pattern. This indicates uniform pore–throat sizing in specific intervals, demonstrating favorable pore–throat sorting characteristics.

The threshold pressure of oil-smelling sandstone samples ranges from 2.963 MPa to 3.684 MPa, with a mean value of 3.420 MPa. The total mercury intrusion saturation varies between 15.68% and 25.25%, averaging 20.44%. These samples exhibit an average throat radius (R_t) of 5.699 μm and an average pore radius (R_p) of 159.23 μm. Distinct differences in both morphological characteristics and parametric features of capillary pressure curves are observed when compared with oil-appearing sandstone samples (Figure 6d–f). Notably, the throat-dominated mercury intrusion curves exhibit near-complete overlap with the total mercury intrusion profiles, with negligible contributions from pore-controlled intrusion phases. The elevated mean throat and pore radius values relative to oil-appearing sandstone samples are attributed to limited pore–throat development and poor connectivity within the rock matrix. The PSD across all samples demonstrates a dispersed pattern, characterized by bimodal distribution curves with peaks at both extremities (Figure 7c). Mercury intrusion volume reveals that the principal contribution to total pore volume originates from a limited proportion of larger pores and throats coupled with substantial fine-scale pores and throats. This PSD curve manifests a pore–throat structure exhibiting pronounced heterogeneity and suboptimal sorting characteristics.

In summary, although both types of oil-bearing core samples exhibit well-developed larger pores, distinct differences are observed in three critical aspects: the configuration of capillary pressure curves, the pattern of PSD curves, and the quantitative parameters of pore–throat structures. These differential characteristics demonstrate that the oil-smelling sandstone cores possess more complex pore–throat structures and exhibit a higher pore–throat heterogeneity compared to oil-appearing sandstone samples.

4.4. TPSD Characteristics

The pores and throats in tight sandstone reservoirs are typically characterized by ultra-fine dimensions and poor sorting characteristics, exhibiting a heterogeneous distribution pattern across multiple scales ranging from nanometer to micrometer magnitudes [64,65]. Therefore, when conducting quantitative characterization of pore–throat structures, reliance on a singular analytical methodology typically proves insufficient to comprehensively delineate the PSD within rock samples. The experimental procedures of both HPMI and CRMI involve the forced intrusion of mercury as a wetting phase into pores and throats under external pressure. Through the precise measurement and theoretical conversion of mercury intrusion volumes, these techniques enable the quantitative assessment of pore–throat structure characteristics, including their size distribution and connectivity [59]. HPMI experiments demonstrate superior capability in characterizing a broad spectrum of PSD, particularly effective in resolving fine-scale pore–throats within low-porosity and low-permeability rock samples. However, this methodology presents two critical limitations. Firstly, the elevated injection pressures may induce the structural deformation of ductile mineral components within the rock matrix, potentially compromising the integrity of the original pore–throat structure. Secondly, the rapid mercury intrusion process under high-pressure conditions induces the “Haines jump” phenomenon (a non-equilibrium displacement mechanism characteristic of multiphase flow in porous media). This dynamic process manifests as mercury preferentially advancing through interconnected pores and throats via discontinuous jumps rather than achieving complete saturation equilibrium in larger pores. Consequently, such non-wetting phase behavior leads to the loss of larger pore quantification in the experiment [66]. CRMI conducted at an injection rate of 5 × 10⁻⁵ mL/min enables mercury intrusion to be characterized as a uniform quasi-static process. This methodology demonstrates enhanced accuracy in characterizing macropores, with reduced measurement errors for larger pores and throats. However, the technique exhibits inherent limitations in micropore detection due to its testing constraints of pore–throats with radii below 0.12 μm remaining undetected under the relatively low injection pressures employed in this experiment [27,67]. Analysis of pore–throat structural parameters derived from mercury injection experiments (Table 1) reveals obvious discrepancies between the average throat radius (R_a) measured by HPMI and the average throat radius (R_t) along with the average pore radius (R_p) obtained through CRMI. This deviation can be attributed to two reasons: (1) The two mercury injection experiments were not performed on identical samples, but rather on adjacent segments from the same core samples (Figure 2b). This methodological approach introduces potential discrepancies due to inherent heterogeneity within the cores, which may influence experimental outcomes. (2) The results from both experiment methodologies demonstrate the phenomena of partial pore–throat being lost within the measurement range, thereby inducing discrepancies between the statistically averaged values of pores and throats. Based on a comparison of their respective technical merits and limitations, the author combined the experimental results of HPMI and CRMI: HPMI results were preferentially adopted for pore–throats smaller than 0.12 μm where its superior resolution in nano-scale domains proves advantageous, while CRMI measurements were selectively utilized for those exceeding 0.12 μm to capitalize on its enhanced accuracy in larger micro-scale features. This synergistic approach effectively reconciles the measurement discrepancies between the two techniques, yielding an optimized reconstruction of the total pore–throat size that demonstrates improved representativeness across multiple scales compared to a single method.

According to the combined results of HPMI and CRMI, the range of sample pore–throat size characterization exceeds 0.001 μm. Oil-appearing sandstone samples have well-developed macropores and throats, and the TPSD curves show a bimodal shape, characterized by a pore–throat radius between 0.001 μm and 28.870 μm. Both pore–throats larger than 0.12 μm and smaller than 0.12 μm provide large pore space in this type of sample (Figure 7b). The TPSD curves of oil-smelling sandstone samples show a bimodal shape with a high left and a low right. The development of the pores and throats is relatively small, characterized by a pore–throat radius between 0.001 μm and 25.589 μm. The main contribution of pore space is concentrated in the pore–throat radius range of 0.01 μm to 0.12 μm (Figure 7d). The two sample types exhibit distinct differences in the pore–throat radius ranges contributing to their primary pore spaces. In contrast, oil-smelling sandstone samples predominantly rely on small pores with radii less than 0.12 μm, exhibiting structural characteristics that are relatively unfavorable for oil migration and accumulation.

5. Discussion

5.1. Fractal Characteristics of TPSD

By analyzing the total pore–throat fractal characteristics of two types of oil-bearing samples through LgV(r)–Lg(r) scatter plots (Figure 8a–f), it is evident that the fractal characteristic curves of both oil-bearing sample types exhibit a distinct multi-interval pattern rather than conforming to a single linear trend across the total pore size spectrum. This suggests that the samples exhibit self-similarity across distinct pore–throat intervals, while demonstrating complex fractal characteristics in their entirety. The fractal dimensions (D) of individual intervals from two types of oil-bearing core samples were statistically analyzed (

Table 2. Statistics of fractal dimension of the total pore–throat size.

Lithofacies	Samples ID	D1	R2	S1	D2	R2	S2	D3	R2	S3	D4	R2	S4	DT
Oil-appearing sandstone samples	W201-1	2.130	0.982	7.678	2.510	0.998	82.985	2.960	0.913	9.336	-	-	-	2.523
	B5-01	2.395	0.993	14.891	2.521	0.984	75.456	2.957	0.828	9.653	-	-	-	2.544
	B7-01	2.099	0.989	13.621	2.617	0.998	77.284	2.953	0.948	9.095	-	-	-	2.577
	Average	2.208	0.988	12.064	2.549	0.994	78.575	2.957	0.896	9.361	-	-	-	2.548
Oil-smelling sandstone samples	W200-01	2.239	0.942	41.246	2.806	0.995	30.218	2.661	0.930	25.983	2.989	0.930	2.553	2.539
	W200-01 (weighted mean)	2.479	-	-	2.690	-	-	2.989	-	-	-	-	-	-
	B1-01	2.350	0.969	79.112	2.983	0.925	9.936	2.932	0.936	10.952	-	-	-	2.477
	B3-03	2.304	0.983	79.872	2.978	0.898	8.234	2.919	0.981	11.894	-	-	-	2.433
	Average	2.378	0.964	66.743	2.884	0.939	16.129	2.946	0.947	16.277	-	-	-	2.483

Note: D_i: the i-th fractal dimension; S_i: the percentage of mercury injection volume in the i-th interval of the pores and throats to the total mercury injection volume, %. R²: correlation coefficient. D_T: the total fractal dimension of the sample.

). Considering the differential distribution characteristics of pore–throat development and their varying contributions to pore space across segments, direct arithmetic averaging would inadequately represent the sample’s comprehensive structural complexity. To address this limitation, the author implemented a weighted averaging method using mercury intrusion volume percentage (S) (each interval of mercury intrusion volume relative to the total intrusion volume). This approach yields a comprehensive fractal dimension (D_T) that quantitatively characterizes the overall pore–throat structural complexity throughout the full pore size spectrum. The calculation formula is expressed as follows:

$$D_T=\sum _{i=1}^n{\displaystyle \frac{D_i{S}_i}{S_i}}$$

(9)

In the formula, D_T represents the total fractal dimension of the sample; i is the i-th paragraph; n is the number of fractal intervals in the sample; D_i is the i-th fractal dimension; and S_i is the percentage of mercury injection volume in the i-th interval of the pores and throats to the total mercury injection volume, %.

Based on the fractal analysis results, the fractal characteristic curves of oil-appearing sandstone samples exhibit a three-interval pattern (Figure 8a–c). All intervals demonstrate strong linear correlations, with coefficients of determination (R²) exceeding 0.8 in each interval. Based on previous studies in correlation analysis, we summarize that an R² value greater than 0.7 is considered to indicate a strong correlation between two datasets, while values ranging from 0.4 to 0.7 and 0.1 to 0.4 suggest moderate and weak correlations, respectively, where an R² value below 0.1 implies no correlation [13,14,15,18,19,34,35,36,63,64,65,66,68]. The analysis integrating fractal characteristic curves with TPSD curves reveals that the fractal inflection point across the samples does not correspond to a specific pore–throat radius value. However, the fractal intervals corresponding to each pore–throat radius ranges exhibit close proximity. As the pore–throat radius increases, the fractal dimension shows and increasing trend. This observation indicates that pore–throat structures with smaller radii display higher structural homogeneity, lower heterogeneity, and roughness of the pore–throat surface. Conversely, as the pore–throat radius increases in a sample, both the structural complexity and heterogeneity of the pore–throats demonstrate obvious intensification. The fractal dimensions of different intervals in the samples ranged from 2.099 to 2.960, with mean values for D₁, D₂, and D₃ measuring 2.208, 2.549, and 2.957, respectively. The weighted average of the composite fractal dimension (D_T) was calculated as 2.548. Notably, the fractal dimension corresponding to small pore segments exhibited values closer to 2. The mercury intrusion volume is predominantly concentrated within the fractal second-stage pore–throat interval, with an average S₂ value of 78.575%.

The fractal characteristic curves of the oil-smelling sandstone samples reveal distinct structural variations. Sample W200-01 exhibits four-interval fractal patterns (Figure 8d), while the other two samples demonstrate three-interval patterns (Figure 8e,f). The four-interval fractal pattern observed in sample W200-01 indicates obviously higher complexity in its pore–throat structure compared to the other two samples. In this study, the authors adopted a weighted averaging approach to determine the fractal dimensions for pore–throat characterization. Specifically, D₁ and D₂ values were integrated through Equation (9) to derive the composite fractal dimension D₁ for the first pore–throat interval. Subsequent fractal dimensions were systematically designated as follows: D₃ was assigned as D₂, representing the second interval, while D₄ was assigned as D₃ for the third interval (Figure 8d). This processing method ensures analytical consistency across samples by maintaining comparable ranges of the characterized PSD. The fractal dimensions of oil-smelling sandstone samples range from 2.304 to 2.989, with average values of 2.378 (D₁), 2.884 (D₂), and 2.946 (D₃) for distinct fractal intervals. The weighted average composite fractal dimension (D_T) was calculated as 2.483. The predominant mercury intrusion volume (average S₁ = 66.743%) was concentrated within the pore–throat size range corresponding to the first fractal interval (D₁), indicating that this interval dominates the pore–throat structure. The sample W200-01 displayed a distinct hierarchical relationship: D₁ < D₃ < D₂ < D₄. It is consistently observed that the fractal dimension increases with larger pore–throat sizes in a sample, indicating more complex pore–throat structures.

Collectively, the overall feature in a sample is that the larger the pores and throats, the higher the fractal dimension and the more complex the pore–throat structure. Due to differences in PSD, the oil-appearing sandstone samples exhibit slightly higher total fractal dimensions (D_T). However, when comparing the dominant pore–throat distribution ranges (D₁ and D₂ intervals in Table 2), the oil-smelling sandstone samples demonstrate more complex pore–throat structures.

5.2. Relationship Between D and Mineral Composition

Fractal dimension enables the piecewise quantitative characterization of the structural complexity in pore–throat structures within tight sandstone reservoirs, serving as a critical bridge connecting micropore–throat structures to macro reservoir heterogeneity [59,69,70]. Reservoir heterogeneity is macroscopically manifested through variations in reservoir physical properties. Porosity is predominantly governed by intrinsic rock characteristics. In contrast, the Influencing factors of permeability are more complex and have their own variability [70]. Scatter plots were generated to analyze the relationship between fractal dimension D and various parameters. However, due to the distinct fractal characteristics exhibited by different intervals, the weighted average total fractal dimension D_T demonstrated no statistically obvious correlation with any parameters. The D₁ interval and D₂ interval serve as primary contributors to pore space in both sample types. D₁ and D₂ exhibit statistically weak negative correlations with core porosity, as shown in Figure 9a. This inverse relationship implies that the enhanced uniformity and regularity of dominant pores and throats within the study area correspond to reduced reservoir heterogeneity and improved petrophysical properties. The D₃ interval, representing relatively larger pore–throat features with limited development, therefore demonstrates inapparent correlation between D₃ and porosity.

The core lithology in the study area is classified as arkose (Figure 3), exhibiting a mineral assemblage dominated by quartz, feldspar, and clay minerals. Detailed mineralogical analysis reveals that the clay minerals primarily consist of illite and chlorite, with their content documented in Figure 3a. A weak positive correlation is observed between quartz content and fractal dimension D₂ in the D₂ interval (Figure 9b). Quartz, characterized by weak plasticity, exerts a supportive framework function in rock matrices. Its abundance obviously influences the development of intergranular pores, thereby enhancing the positive correlation with fractal dimension. CTS analysis reveals that the intergranular pores in the studied samples have undergone multiple phases of diagenetic alteration, including interstitial material filling, cementation, and dissolution processes. These diagenetic alterations resulted in the reduction in primary intergranular porosity, which consequently accounts for the observed weak positive correlation between D₂ and quartz content. D₁, D₂, and D₃ exhibit weak negative correlations or show no obvious correlation with feldspar content. Among these, D₂ demonstrates the most pronounced correlation (albeit still weak), with a coefficient of determination R² = 0.1042 (Figure 9c). Feldspar usually undergoes dissolution and produces dissolution pores in the later stage. On the one hand, the more feldspar there is, the easier it is to produce uniform dissolution pores, which means that the small pores tend to be uniform and reduce the fractal dimension (D); on the other hand, the surface of the small pores produced is rough and irregular, which will increase the heterogeneity of the pores and throats. Therefore, the fractal dimension D of tight sandstone samples with small pores mainly developed in the study area did not show an obvious correlation with feldspar content. The content of clay minerals exhibits a weakly positive correlation with fractal dimension D, particularly within D₂ (Figure 9d). Clay minerals primarily function as interstitial fillings occupying primary intergranular pores, thereby reducing pore–throat homogeneity. Subsequent incomplete dissolution processes result in surface roughening and irregular pore geometries. This diagenetic modification predominantly governs the PSD in the D₂ interval, which accounts for the enhanced correlation observed between D₂ and clay mineral contents. The clay mineral assemblage is characterized by relatively high proportions of illite and chlorite. Illite demonstrates an obvious positive correlation with D₁ and a weak positive association with D₃ (Figure 9e). It occupies interparticle pore spaces (Figure 9i), resulting in irregular and roughened pore geometries that substantially enhance reservoir heterogeneity. Chlorite exhibits obvious positive correlation within D₂ (Figure 9f), manifesting as authigenic chlorite rims coating quartz grain surfaces (Figure 4h). These chlorite films exacerbate the roughness and irregularity of pores, further amplifying the heterogeneity characteristics of the pores and throats.

The fractal dimension (D) of pore–throat structures in the study area exhibits correlations with multiple reservoir factors, with both lithological composition and clay minerals exerting obvious influences on pore–throat structure homogeneity and complexity. The development of quartz minerals induces a slight increase in the fractal dimension, primarily attributed to the surface roughening and geometrical irregularity of intergranular pores during late-stage diagenetic processes. Feldspar dissolution exhibits two-sided effects on fractal dimension (D) variations through the generation of secondary dissolution pores, though no statistical correlation was observed between feldspar and the fractal dimension (D) in the samples. Particularly noteworthy is the critical role of illite and chlorite in clay minerals, as their pore-filling characteristics during early diagenesis followed by differential dissolution in subsequent stages collectively contribute to enhanced surface asperity and complexity of pore–throat structures, thereby resulting in elevated fractal dimension values.

5.3. Relationship Between D and Pore–Throat Structure Parameters

The parameters of pore–throat structures obtained from HPMI and CRMI experiments serve as critical indicators for characterizing pore–throat dimensions, distribution characteristics, and connectivity. A correlation was established between these parameters and the fractal dimension (D). Comparative analysis through parametric plotting reveals that HPMI-derived pore–throat parameters demonstrate relatively stronger correlations with fractal dimension D compared to CRMI-derived counterparts. This discrepancy primarily stems from the dominant contribution of small-scale pores in the study area samples. The pore range of CRMI testing tends to have macropores and loses pores smaller than 0.12 μm, and the limited number of macropores in the sample results in a larger average pore radius.

Based on the analysis of the scatter plot between average pore–throat radius (R_a) and fractal dimension (D) derived from HPMI measurements (Figure 9g), distinct correlations emerge within different PSD intervals. A weak negative correlation is observed between fractal dimensions and the dominant PSD intervals D₁ (R² = 0.4181) and D₂ (R² = 0.3538). However, no obvious correlation is identified in the D₃ interval. Notably, within self-similar PSD intervals sharing identical fractal characteristics, a higher average pore–throat radius consistently corresponds to lower fractal dimension values. The HPMI-derived sorting coefficient (S_p) exhibits a weak negative correlation with fractal dimensions D₁ and D₂, while showing no obvious correlation with D₃ (Figure 9h). A higher SP indicates improved pore–throat sorting and enhanced uniformity in PSD, which consequently reduces fractal dimension values. Maximum mercury saturation (S_max) and mercury withdrawal efficiency (W_e), as critical indicators of pore–throat connectivity, demonstrate weak negative correlations with fractal dimensions D₁ and D₂ (Figure 9i,j). This indicates that the larger S_max and We, the higher the mercury injection efficiency, the better the pore–throat connectivity, and the lower the complexity of the pores and throats, with a fractal dimension close to 2. The threshold pressure P_cd reflects the difficulty of mercury entering the pores. The P_cd measured by HPMI and CRMI is positively correlated in both D₁ and D₂ intervals, with a larger pore size in the D₃ interval and no obvious correlation (Figure 9k,l). The higher the threshold pressure, the more difficult it is for mercury to enter the pore space, resulting in a more complex pore–throat structure, stronger heterogeneity, and larger fractal dimension.

5.4. Analysis of Causes of Oiliness Differences

Previous studies have identified critical factors influencing the differential accumulation of oil and gas, including the hydrocarbon generation capacity of source rocks, hydrocarbon migration dynamics, migration pathways, sedimentary microfacies, reservoir physical properties, pore–throat structure, lithological associations, and diagenesis [71,72,73,74,75]. Focusing on the microscopic characteristics of reservoirs, this study employs experimental analyses integrated with fractal theory to elucidate the disparities in oil-bearing properties between two types of oil-bearing reservoirs in the study area.

Based on previous research, the two types of oil-bearing core samples exhibit distinct differences in mineral characteristics, pore–throat structures, and the complexity of pore–throats. Oil-smelling sandstone predominantly develops within finer-grained sandstone intervals deposited in channels. These intervals correspond to either the upper section of fining-upward sequences or the lower section of coarsening-upward sequences. Core observations reveal distinct calcareous cementation within these sandstones (Figure 10a). The original intergranular pores were occluded by cements including ferriferous calcite, illite, and chlorite. The present pores predominantly consist of dissolution pores formed by subsequent feldspar dissolution (Figure 10b). The TPSD curves exhibit a left-skewed bimodal pattern, dominated by micropores with diameters less than 0.12 μm. The sample exhibits an average mercury injection saturation of 51.59%, while the saturation for pores larger than 0.12 μm averaging 20.44% (Figure 10c). The fractal characteristics display either a three-interval or four-interval pattern (Figure 10d).

Oil-appearing sandstone predominantly develops within the coarser-grained sandstone intervals of channel deposits. These intervals correspond to the lower sections of fining-upward sequences or the upper sections of coarsening-upward sequences. Core surfaces exhibit distinct oil immersion (Figure 10e). Compared to oil-smelling sandstone, the degree of pore destruction for oil-appearing sandstone samples is relatively lower. The developed pore types include primary intergranular pores, residual intergranular pores, and dissolution pores (Figure 10f). The samples exhibit well-developed pores larger than 0.12 μm, with an average mercury injection saturation of 66.43%. In contrast, pore–throats below 0.12 μm demonstrate a mercury saturation of 19.41%. The TPSD curves display a symmetric bimodal morphology (Figure 10g), and the fractal characteristic curves exhibit tripartite segmentation (Figure 10h). For the dominant pore–throat distribution intervals D₁ and D₂, the fractal dimensions (D₁ and D₂) are obviously lower than those observed in oil-smelling sandstone samples (Table 2). Comparative analysis reveals obvious differences in the hydrocarbon potential between the two types of cores in the study area. These differences are attributed to distinct variations in their mineralogical composition, diagenesis, clay mineral content, pore types, PSD, and complexity of pore–throat structures. Oil-appearing sandstone is characterized by low clay mineral content, minimal destructive diagenesis, well-developed pores, and low pore–throat complexity. These properties collectively enhance hydrocarbon migration and accumulation within such rock reservoirs.

The results of this study reveal the micro reasons for the differences in different oil-bearing tight sandstone reservoirs, providing guidance for the explanation of the differences in different oil-bearing reservoirs, reservoir classification evaluation, and selection of later development intervals. It also provides insights for the exploration and development of other similar tight sandstone oil reservoirs.

6. Conclusions

This study adopts a parallel-sample design and integrates fractal theory to conduct qualitative and quantitative comparative analyses from multiple perspectives, including petrological characteristics, pore types, diagenesis, TPSD characteristics, and pore–throat structure complexity. These analyses elucidate the differences in oil-bearing properties between two types of reservoir units in the Xiasiwan Oilfield, leading to the following conclusions:

(1)

Although the two types of oil-bearing cores share a similar lithology (primarily composed of arkose), an obvious distinction lies in clay mineral composition. Compared with oil-appearing sandstone, the oil-smelling sandstone has poorer pore development along with a higher content of illite and chlorite, while exhibiting more pronounced impacts of destructive diagenetic processes on the pore–throat structure.

(2)

Concerning the results of the TPSD curves, the oil-appearing sandstone samples exhibit a symmetrical bimodal distribution and three-stage fractal characteristics. While the oil-smelling sandstone samples display a bimodal distribution characterized by a higher left peak and a lower right peak, they exhibit three-stage and four-stage fractal patterns. The pore space contribution further demonstrates that the oil-smelling sandstone samples have a greater reliance than the oil-appearing sandstone on small pores and throats below 0.12 μm. Under the influence of clastic mineral composition and content, clay mineral content, pore–throat size, and pore–throat sorting, the oil-smelling sandstone samples exhibit higher complexity compared with oil-appearing sandstone samples.

(3)

The oil-appearing sandstone is characterized by low clay mineral content, minimal destructive diagenesis, well-developed pores, and low pore–throat complexity, which is conducive to the hydrocarbon migration and accumulation within such reservoirs. These findings provide crucial guidance for reservoir evaluation.