Development Mechanism and Pattern of the Microscopic Pore Structure in Deep Tight Sandstone Reservoirs: Xihu Depression, East China Sea Basin

Abstract

Deep tight sandstone reservoirs are characterized by strong microscopic pore structure heterogeneity and commonly exhibit a high-porosity, low-permeability profile, posing significant challenges for effective reservoir evaluation and “sweet spot” prediction. The microscopic pore structure of 209 tight sandstone samples from the deeply buried Huagang Formation in the Xihu Depression, East China Sea Basin, was systematically characterized by integrating multiple analytical techniques, including casting thin sections, scanning electron microscopy (SEM), X-ray diffraction (XRD), nuclear magnetic resonance (NMR), and high-pressure mercury injection (HPMI). The results indicate that the reservoir space is dominated by mesopores (55.48%) and transition pores (32.39%), with macropores (2.09%) and micropores (10.04%) being relatively underdeveloped. A significant vertical heterogeneity in reservoir quality is observed. The H4 member exhibits superior properties, characterized by a higher average movable fluid saturation (averaging 46%) and better pore connectivity. In contrast, the H5 member is more compact, with a notably higher proportion of bound fluid (averaging 47%). The differences in reservoir quality are controlled by a sedimentary–diagenetic coupling mechanism. High-energy, coarse-grained facies underwent a constructive pathway involving chlorite coating protection and dissolution enhancement, forming high-quality pore networks. In contrast, low-energy, fine-grained facies experienced a destructive pathway dominated by intense compaction and cementation, leading to the deterioration of pore structure. The petrophysical properties of the deep reservoirs are primarily governed by the three-dimensional connectivity and spatial distribution of effective “pore-throat assemblages” composed of dominant throats. Accordingly, a “binary” pore structure development pattern is established for the deep tight sandstone reservoirs in the study area. This pattern posits that the reservoir space is heterogeneously composed of a minority of connected “effective percolation assemblages” and a majority of isolated “ineffective assemblages”.

1. Introduction

The transformation of the global energy structure is driving the exploration and development of unconventional oil and gas resources [1,2,3,4,5,6,7]. Among these, deep to ultra-deep tight sandstone gas has become a crucial replacement resource [8,9,10]. However, tight sandstone reservoirs commonly exhibit extremely strong heterogeneity and complex microscopic pore structures. These characteristics directly control the gas storage capacity, seepage mechanisms, and ultimate recoverability, thereby severely constraining their economic viability [11,12,13]. Compared to conventional reservoirs, deep tight sandstones have typically undergone intense diagenetic alteration [14,15,16,17,18], resulting in highly complex pore systems and seepage networks. Macroscopically, they often display strong heterogeneous features characterized by high porosity (φ) but low permeability (K) [19,20,21]. The complexity of φ and K relationship poses significant challenges for evaluating reservoir effectiveness and predicting high-productivity sweet spots. The root of this challenge lies in an insufficient understanding of the mechanisms controlling the formation and evolution of the microscopic pore structure.

Currently, it is widely acknowledged in academia that tight sandstone reservoirs are dually controlled by both sedimentation and diagenesis [22,23,24,25,26,27,28,29]. For instance, the reservoir quality of the shoreface deltaic sandstones of the Brent Group in the North Sea Basin, UK, under conditions of moderate burial depth and temperature (approximately 9000 ft and 95 °C) [27], is primarily governed by the original detrital composition and grain size, which are controlled by sedimentary facies. High-energy, clean sands exhibit greater resistance to compaction due to a lower content of ductile grains and are less conducive to the extensive development of early diagenetic cements, thereby preserving relatively good porosity. In contrast, low-energy, clay-rich sands are prone to intense compaction and cementation, leading to a sharp decline in reservoir quality. In the marginal marine sandstones of the Jauf Formation in Saudi Arabia [30], differences in sedimentary facies serve as the primary controlling factor for variations in reservoir properties. Subsequent diagenetic processes further exacerbate reservoir heterogeneity on this basis. Within deep to ultra-deep burial settings, diagenetic processes become particularly complex due to high temperatures, high pressures, and multi-stage fluid–rock interactions, including organic–inorganic reactions [31]. On one hand, intense mechanical compaction and silica/carbonate cementation can cause rapid reservoir densification, which is particularly evident in the deep-water fan sandstones of the Butano Formation in California, USA [24], which experienced deep burial and potential hydrothermal activity. On the other hand, fluids such as organic acids expelled from hydrocarbon source rocks during hydrocarbon generation can induce constructive dissolution, creating secondary porosity. Furthermore, specific diagenetic products, such as early chlorite coatings, can effectively inhibit the development of destructive quartz overgrowths, thereby protecting porosity [28].

However, significant gaps remain in the research on the genesis of the microscopic pore structure of tight sandstone reservoirs within specific geological contexts characterized by multi-phase tectonic inversion, rapid burial, and abnormal overpressure [22,32,33,34]. Firstly, the internal mechanism by which sedimentation and diagenesis collaboratively control the differential evolution of the pore structure is insufficiently explained. Secondly, under deep, intense, and multi-phase diagenetic alteration, the dominant factors and patterns controlling pore structure evolution have not been systematically dissected. This, in turn, leads to low prediction accuracy for effective reservoirs in deep tight sandstones.

The Huagang Formation in the Xihu Depression of the East China Sea Shelf Basin harbors abundant tight sandstone gas resources, making it a key target interval for current offshore natural gas exploration. This reservoir suite is generally buried deeper than 3500 m, having undergone high temperatures and intense diagenetic alteration, resulting in extremely strong reservoir heterogeneity and typically exhibiting a high φ and low K profile. Therefore, this study systematically characterized the microscopic pore structure of 209 tight sandstone samples from the deeply buried Huagang Formation in the Xihu Depression, East China Sea Basin, by integrating multiple analytical techniques, including petrographic thin section analysis, scanning electron microscopy (SEM), X-ray diffraction (XRD), nuclear magnetic resonance (NMR), and high-pressure mercury injection (HPMI). The study aims to achieve the following specific objectives: (1) to quantitatively characterize the differences in microscopic pore structure parameters between the H4 and H5 reservoir intervals; (2) to reveal the coupled controlling mechanism of sedimentary facies and key diagenetic processes on the pore structure; (3) to identify the key pore structure parameters controlling reservoir flow capacity; and (4) ultimately, to establish a applicable “binary” microscopic pore structure development model for deep tight sandstone reservoirs in the study area, providing a micro-mechanism basis for sweet spot prediction.

2. Materials and Methods

2.1. Geological Setting and Samples

The East China Sea Shelf Basin is the largest Mesozoic-Cenozoic petroliferous basin offshore China (Figure 1a), with a relatively low level of hydrocarbon exploration, still in the early stages of exploration and development. Bounded by the Diaoyu Islands Paleo-Uplift and the Okinawa Trough [35,36], the basin can be divided from west to east into three roughly parallel first-order tectonic units: the Western Depression Belt, the Central Uplift Belt, and the Eastern Depression Belt. Currently, exploration and development efforts are focused on the Xihu Depression, located within the Eastern Depression Belt. The sag extends in a NNW-SSE direction, covering an area of approximately 5.9 × 10⁴ km². It is further subdivided into five secondary structural units: the Western Slope Belt, the Western Sub-sag, the Central Inversion Anticline Belt, the Eastern Sub-sag, and the Eastern Fault Terrace Belt (Figure 1b) [17,37]. The study area, the Z gas field, is situated within the Central Inversion Anticline Belt, where four exploration wells have been deployed with a well spacing of approximately 2.3 to 5.9 km (Figure 1c).

Figure 1. (a) Location of the East China Sea Basin. (b) Location of the study area (modified from [38]). (c) Schematic cross-section of the sampling wells in the study area.

The Xihu Depression is primarily filled with Cenozoic clastic deposits, reaching a considerable total thickness of 9 to 15 km. During the Cenozoic, the sag experienced multiple tectonic events, including the Oujiang, Yuquan, Huagang (dominantly rift-related subsidence), Longjing (representing tectonic inversion), and Okinawa Trough movements (Figure 2). These events collectively controlled the subsidence, sedimentation, and stratigraphic framework of the sag. The Huagang Formation was deposited during a period of rapid subsidence, leading to the accumulation of thick fluvial–lacustrine clastics [14,15,16]. The subsequent Longjing inversion uplift shaped the present structural framework, including the Central Inversion Anticline Belt where the study area is located, and influenced the burial/thermal history of the reservoirs [15]. Therefore, the tectonic evolution set the macroscopic stage for sedimentation and subsequent diagenetic processes [9,15,27].

Figure 2. Comprehensive stratigraphic and sedimentary evolution column of the Xihu Depression (modified from [14]).

From bottom to top, the stratigraphic sequence of the Xihu Depression comprises the Eocene Pinghu Formation, the Oligocene Huagang Formation, the Miocene Longjing, Yuquan, and Liulang formations, followed by the Pliocene Santan Formation and the Quaternary Donghai Group. Among these, the Oligocene Huagang Formation is a crucial hydrocarbon reservoir, characterized by a widely distributed and continuous sandstone succession deposited in a shallow-water braided river delta system [14,15,38,39]. The Huagang Formation is further subdivided into upper and lower members. The upper member includes sandstone units H1–H5, while the lower member comprises units H6–H12. The focus of this study is on the H4 and H5 sandstone units within the upper member of the Huagang Formation (Figure 2). This sandstone interval is deeply buried (3800–4150 m), subjected to high formation temperatures (135–155 °C), has undergone intense diagenetic alteration, and exhibits strong reservoir heterogeneity [14,15,35].

Samples were collected from the H4 and H5 members of the upper Huagang Formation (3800–4500 m) in the Z gas field, Xihu Depression, East China Sea Basin (Figure 2). A total of 209 tight sandstone samples and 1716 conventional core test data points were obtained. Of the 209 acquired core samples, 113 were subjected to cast thin section petrography analysis for quantitative mineralogical and porosity assessment. SEM imaging was performed 107 times to elucidate diagenetic textures and microscale pore features, and 31 XRD analyses were conducted to determine bulk mineralogy and clay mineral composition. For in-depth pore structure characterization, 76 representative cylindrical samples (2.54 cm in diameter and 2.5 cm in length) were selected, undergoing 75 HPMI tests and 23 NMR tests. A total of 10 sandstone samples were prepared as cathodoluminescence thin sections for the analysis of authigenic minerals. To systematically compare the microscopic pore structure characteristics of samples formed under different sedimentary–diagenetic conditions, the selected suite encompasses a range of lithologies, diagenetic types, grain sizes, hydrodynamic conditions, physical properties, and mineral compositions, ensuring their representativeness and comparability. While the study is necessarily based on discrete well points, these samples are considered representative of the key reservoir intervals and their variability within the study area for the purpose of establishing the microscopic genetic mechanisms explored herein.

2.2. Experiment Methods

Samples for petrographic analysis were impregnated with blue epoxy resin and prepared as cast thin sections with a thickness of approximately 30 µm. The preparation procedure followed the Chinese petroleum industry standard SY/T 5913-2021 [40]. For pore identification and carbonate mineral discrimination, selected thin sections were stained with Alizarin Red S and potassium ferricyanide, imparting characteristic colours to carbonate minerals such as calcite and dolomite under optical microscopy. A systematic point-counting analysis was performed on the sample thin sections using a Zeiss Axio Imager A2m optical microscope (Carl Zeiss AG, Oberkochen, Germany). A minimum of 500 points were counted for each thin section on a uniformly spaced 20 × 25 grid to quantitatively determine the types and abundances of detrital components, authigenic cements, and visible pores. Digital images of the thin sections were acquired using an optical microscope and subsequently analyzed with the ImageJ software (ImageJ 1.54e) to assist in systematic point-counting and morphological measurements. The standard error of the point-counting results was controlled within approximately 5% [41]. Furthermore, cathodoluminescence (CL) imaging was conducted on selected, thicker polished thin sections using a Cambridge CL8200 MK5-2 instrument (Cambridge Image Technology Ltd., Hatfield, UK) attached to the Zeiss microscope, operating with an accelerating voltage of 17 kV and a beam current of 600 µA.

The scanning electron microscopy (SEM) analysis was conducted following the Chinese petroleum industry standard SY/T 5162-2021 [42]. A Quanta FEG 450 field-emission scanning electron microscope (FEI, Hillsboro, OR, USA) was employed, with an accelerating voltage ranging from 0.1 to 20 kV and a best resolution of 5 nm for backscattered electron (BSE) imaging. Analyses were performed at resolutions of 0.5 or 0.83 µm to obtain clear morphological information. Sample preparation included oil washing, sample selection, mounting, drying, dust removal, and gold coating. The preparation process involved polishing, high-energy argon ion beam milling, and surface coating with a conductive carbon film. By acquiring high-resolution secondary electron and backscattered electron images, the pore types, pore-throat morphology, cement morphology, and mineral distribution characteristics of the samples were systematically characterized.

Samples for whole-rock and clay mineral (<2 µm) X-ray diffraction (XRD) analysis were prepared and analyzed following the Chinese petroleum industry standard SY/T 5163-2018 [43]. The measurements were performed using an Ultima IV X-ray diffractometer (Rigaku Corporation, Tokyo, Japan). The relative content of clay minerals (e.g., illite, kaolinite, chlorite) and their crystalline characteristics were thereby determined. The obtained diffraction patterns were processed, and mineral identification and semi-quantitative analysis were performed using the JADE software (MDI JADE 6.5). Porosity and permeability were measured in accordance with the Chinese National Standard GB/T 34533-2023 [44].

Nuclear magnetic resonance (NMR) is a sophisticated, nondestructive, and quantitative tool for examining the pore structure [45,46]. Experiments were conducted following the Chinese petroleum industry standard SY/T 6490-2023 [47]. A Geospace2 NMR analyzer (Oxford Instruments, Abingdon, UK) was used to perform transverse relaxation time (T₂) spectrum measurements on plug samples. The NMR spectrometer operated at a resonance frequency of 2 MHz. The key parameters for the Carr–Purcell–Meiboom–Gill (CPMG) pulse sequence were set as follows: a polarization wait time (Tw) of 10,000 ms, an echo spacing (TE) of 0.3 ms, 4096 echoes acquired, and 128 scan repetitions. All rock samples were processed according to the standard core analysis workflow, sequentially undergoing oil washing, drying, vacuum evacuation, and saturation with simulated formation brine to achieve a 100% water-saturated state. First, NMR measurements were performed on the samples in the fully water-saturated state. Subsequently, the samples were centrifuged at 10,000 rpm until the water saturation reached a stable state, thereby achieving an ideal irreducible water condition. NMR measurements were then conducted again on the samples in this irreducible water state. The raw NMR data were inverted to obtain the T₂ spectrum using the dedicated software suite (GeoSpec system software, https://nmr.oxinst.com/geospec, accessed on 2 January 2026) provided by the instrument manufacturer. The amplitude of the T₂ spectrum (Y-axis) is directly correlated with the sample porosity, and the T₂ time (X-axis) is proportional to the pore size [48,49]. The T₂ cutoff value, was determined based on the inflection point in the cumulative T₂ spectrum of samples at irreducible water saturation after centrifugation. Based on the T₂ spectrum and calibrated T₂ cutoff value, parameters such as movable fluid saturation and bound fluid saturation were calculated.

The high-pressure mercury intrusion (HPMI) is an approach that is typically used for assessing pore properties, notably the distribution of pore throat sizes in porous media [13,19]. Experiments were performed in accordance with the Chinese National Standard GB/T 29171-2023 [50] using an AutoPore IV-9520 porosimeter (Micromeritics Instrument Corporation, Norcross, GA, USA). Prior to the experiments, the core samples underwent pretreatment, including oil washing and drying, and their fundamental petrophysical properties, such as φ, K, and density, were measured. The cylindrical plug samples used for testing had a diameter of approximately 2.54 cm and a length of about 4.5 cm. Experiments were conducted under constant temperature (25 °C) and humidity (35%–50% relative humidity) conditions. During testing, non-wetting mercury was injected into the rock pores. The capillary pressure curve was obtained by recording the cumulative mercury intrusion volume at different injection pressures. The maximum intrusion pressure was 116.67 MPa, corresponding to a minimum measurable pore-throat radius of 0.006 µm. The measurement accuracy for mercury intrusion/withdrawal volume was better than 0.1 µL. Finally, based on the conversion relationship between capillary pressure and pore-throat radius (Equation (1)), the capillary pressure curves were transformed into pore-throat radius distribution curves for evaluating the pore structure characteristics of the tight sandstones:

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

(1)

3. Results

3.1. Reservoir Characteristics

Thin-section analysis indicates that the deep tight sandstones of the Huagang Formation in the study area are primarily composed of quartz, feldspar, and lithic fragments. Quartz content ranges from 51.6%–87.1% (average 69.8%). Feldspar content ranges from 1.4%–33.4% (average 16.3%). Lithic fragment content ranges from 7.2%–21.4% (average 14.0%). According to the sandstone classification scheme of Folk (1974) [51], the tight sandstones in the Huagang Formation of the study area are predominantly feldspathic litharenites, with minor amounts of lithic arkoses and litharenites (Figure 3). Whole-rock and clay mineral X-ray diffraction data show that clay mineral content ranges from 2.6%–9.0% (average 5.1%). Clay minerals are dominated by chlorite, illite, and mixed-layer illite-smectite, with very little kaolinite. Carbonate cement content is generally less than 4%, primarily consisting of calcite and ferroan calcite.

Figure 3. Lithological classification of the tight sandstones from the Huagang Formation, Xihu Depression, East China Sea Basin (Q: quartz, F: feldspar, L: lithic fragments).

Core-based grain-size analysis reveals that the reservoir in the study area is dominated by medium-grained sandstone, followed by fine-grained sandstone, with the content of medium-grained sandstone showing an increasing trend with depth (Figure 4). Specifically, within the H4b sublayer, the grain size in the shallow section (H4b1–H4b2) is generally finer than that in the middle to deep section (H4b3–H4b5). Concurrently, the grain-size distribution in H4b1–H4b2 is relatively wide and poorly sorted, whereas it is more uniform in the other sublayers. The H5a sublayer exhibits two distinct upward-coarsening cycles: the upper cycle (H5a1–H5a2) and the lower cycle (H5a3–H5a5) both show a similar upward-coarsening trend. Pebbly sandstone is more developed in H5a2 and H5a5 sublayers. In terms of lithological composition, H5a1 sublayer is primarily composed of fine-grained sandstone with minor medium-grained sandstone, while H5a3–H5a5 consists mainly of medium-grained sandstone, interbedded with coarse-grained or fine-grained sandstone.

Figure 4. Grain size distribution characteristics of tight sandstones in different sub-layers of the Huagang Formation, study area: (a) H4 member and (b) H5 member.

Porosity and permeability measurements (Figure 5) show that the reservoir φ in the study area ranges from 6.05%–7.91% (average 7.16%), and the K ranges from 0.131 to 1.240 mD (average 0.374 mD). Overall, the reservoir exhibits medium to low φ and ultra-low K characteristics, with strong vertical heterogeneity. Specifically, the physical properties of the middle section of the H4b member are relatively good, while those of the shallow (H4b1–H4b2) and deep sections are significantly poorer. Within the H5a member, the H5a2 and H5a3 sublayers are relatively favorable reservoir intervals. This member overall displays two upward-coarsening cycles, with the better-quality reservoirs mainly developed in the coarser-grained lithologies at the base of each cycle.

Figure 5. Vertical distribution characteristics of petrophysical properties of the Huagang Formation reservoirs, study area: (a) H4 member and (b) H5 member.

Furthermore, a weak positive correlation is observed between φ and K (Figure 6), and the differences in petrophysical properties among sandstones of different grain sizes are not significant. Only pebbly sandstone shows slightly better properties, while siltstone generally exhibits poor properties. This preliminarily indicates that in deep reservoirs, K is not primarily controlled by φ but is more influenced by pore-throat size, structure, and connectivity. In addition, intense compaction and cementation are key factors leading to reservoir densification [15,18,19,23]. Coarse-grained sandstones, due to their stronger resistance to compaction, are more conducive to preserving effective flow throats. Consequently, relatively better petrophysical properties are retained in local pebbly to coarse sandstone intervals.

Figure 6. Correlation between φ and K for different lithologies.

3.2. Diagenetic Characteristics

The tight sandstone reservoirs of the Huagang Formation in the study area, with burial depths generally exceeding 3500 m, have undergone significant mechanical compaction. Petrographic observations indicate that the overall compaction intensity is strong (Figure 7a–d). Detrital grains are predominantly in line to concavo-convex contact, with clearly visible pressure solution sutures. Ductile grains, such as mica and argillaceous lithic fragments, commonly show deformation and pseudomatrix formation. Rigid grains, including quartz, feldspar, and quartzose lithic fragments, exhibit fracturing.

Figure 7. Compaction and dissolution phenomena observed in typical microscopic images. (a) Intense compaction, deformation of ductile grains, and fracturing of rigid grains, 4295.80 m. (b) Line concavo-convex contacts, deformation of ductile grains, fracturing of rigid grains, and pressure solution features, 4288.00 m. (c) Intense compaction, deformation of ductile grains, and fracturing of rigid grains, 4245.00 m. (d) Point-line concavo-convex contacts and pressure solution features, 4322.80 m. (e) Dissolution of K-feldspar and volcanic lithic fragments, 3701.00 m. (f) Dissolution of K-feldspar grains, 4295.20 m.

Dissolution is an important constructive diagenetic process in the study area. Unstable components, such as feldspar and volcanic lithic fragments, are widely dissolved, forming intragranular dissolution pores, moldic pores, and enlarged intergranular dissolution pores (Figure 7e,f). The intensity of dissolution shows vertical variation: the H4 reservoir has relatively well-developed secondary dissolution pores, with higher visible φ and better pore connectivity. In contrast, the H5 reservoir contains relatively fewer secondary dissolution pores, which are mostly isolated and poorly connected. Additionally, localized dissolution of quartz grain edges or overgrowths is observed at certain depths.

Silica cementation is extremely prevalent in the study area and occurs primarily in three forms: (1) euhedral to subhedral quartz overgrowths surrounding detrital quartz grains (Figure 8a); (2) euhedral quartz microcrystals within intergranular or intragranular dissolution pores (Figure 8b,c); (3) anhedral microcrystalline quartz filling residual pores (Figure 8b,c). The development intensity and morphology of silica cementation exhibit significant vertical heterogeneity. In the H4 reservoir, silica cementation is relatively weak, mostly occurring locally as incomplete overgrowths or euhedral microcrystals within pores (Figure 8l). In contrast, silica cementation is significantly enhanced in the H5 reservoir, often manifested as continuous, complete quartz overgrowths or abundant pore-filling microcrystalline quartz associated with illite (Figure 8i).

Figure 8. Cement phenomena observed in typical microscopic images. (a) Chlorite coating development; quartz cement initiating as authigenic microcrystals from areas where the coating is absent, 3820.00 m. (b) Generally thin chlorite coatings; quartz cement occurring as overgrowths, 4048.30 m. (c) Dissolution of K-feldspar, ferroan calcite cement, and quartz overgrowth, 4193.60 m. (d) Local pore-filling ferroan calcite cement and replacement of feldspar grains, 3904.00 m. (e) Local pore-filling ferroan calcite cement, 4106.50 m. (f) Early-stage polkilotopic (poikilotopic) ferroan calcite cement, appearing orange-red in CL; middle-stage patchy ferroan calcite cement, appearing orange-yellow in CL; late-stage dispersed granular to locally polkilotopic ferroan calcite cement, occupying primary intergranular and secondary dissolution pores, appearing dark red in CL, 3915.70 m. (g) Coating chlorite, mixed-layer illite/smectite, and microcrystalline quartz cement, 4193.26 m. (h) Dissolution of K-feldspar and authigenic illite, 4288.00 m. (i) Intergranular pores, authigenic illite, and quartz grains, 4197.11 m. (j) Pores compartmentalized by authigenic illite, 4290.60 m. (k) Pores compartmentalized by mixed-layer illite/smectite, 4297.60 m. (l) Flaky/dispersed chlorite and quartz overgrowth, 4294.00 m.

Carbonate cements are dominated by calcite and ferroan calcite. Based on their occurrence, replacement relationships, and CL characteristics, at least three phases of carbonate cementation can be identified: Early-stage calcite commonly occurs as polkilotopic cement, exhibits an orange-red CL color, and has a relatively high iron content (Figure 8d,f). Intermediate-stage calcite typically appears as dispersed granular crystals, replacing earlier minerals or filling pores, shows an orange-yellow CL color, and has a lower iron content (Figure 8d,f). Late-stage (ferroan) calcite often forms patchy or polkilotopic cement filling secondary dissolution pores, displays a dark red CL color, and has the highest iron content (Figure 8e,f). Furthermore, calcite cement shows minor variation with depth in the study area and is generally present in low abundances. In terms of spatial distribution, calcite cement with higher abundance is mostly early-stage calcite and is most developed at sandstone–mudstone interfaces.

Authigenic clay minerals primarily include chlorite, illite, and mixed-layer illite/smectite (I/S). Chlorite exhibits two main habits: In the H4 reservoir, it primarily occurs as pore-linings, forming chlorite coatings on detrital grain surfaces with an average thickness of 2–4 μm and relatively high coverage (Figure 8a,g). In contrast, in the H5 and deeper reservoir, such coatings are barely developed, typically being <2 μm thick, and pore-filling chlorite is more common (Figure 8l). The chlorite coatings generally grow perpendicular to grain surfaces, exhibiting an isopachous, bladed habit. Illite and mixed-layer I/S mainly occur in platy, filamentous, or hairy forms (Figure 8g–i), bridging or filling pore throats (Figure 8j,k). Furthermore, the relative content of illite increases with burial depth, and the degree of illite cementation is higher in the H5 reservoir compared to the H4 reservoir.

3.3. Micro-Pore Structure Characteristics

Cast thin section observations indicate that the reservoirs of the Huagang Formation in the study area primarily develop residual intergranular pores, intergranular dissolution pores, intragranular dissolution pores, moldic pores, and minor fractures (Figure 9). Among these, residual intergranular pores are relatively low in content (average 0.59%), typically exhibit straight-edged boundaries and are relatively large in size (30–150 μm), and show good connectivity (Figure 9a–c). Intergranular dissolution pores are the most common pore type, with a content generally <10% (average 1.5%). They are mainly formed by partial dissolution of feldspar and lithic fragment edges, are relatively large in size (50–180 μm), and exhibit good connectivity (Figure 9d,e). Both intragranular dissolution pores and moldic pores are formed by the dissolution of unstable components such as feldspar and lithic fragments, and both types are present in low abundances. The content of intragranular dissolution pores is generally <5% (average 0.79%), with smaller pore sizes (5–20 μm) and poorer connectivity (Figure 9d,f). Moldic pore content is generally <4% (average 0.7%), with larger pore sizes (100–200 μm) but also relatively poor connectivity (Figure 9d,g).

Figure 9. Micro-pore structure characteristics observed in typical microscopic images. (a) Intergranular pores preserved after compaction and cementation, 3720.00 m. (b) Pore preservation supported by a rigid grain framework, 3723.50 m. (c) Pore preservation by chlorite coatings, 3721.60 m. (d) Intragranular and intergranular dissolution pores, 3984.20 m. (e) Dissolution pores between mineral grains, 3984.20 m. (f) Intragranular dissolution pores within feldspar, 3710.80 m. (g) Moldic pores formed by feldspar dissolution, 3966.30 m. (h) Intercrystalline pores within clay minerals, 4280.20 m. (i) Mica cleavage pores, 3994.60 m.

HPMI testing can effectively characterize the pore-throat structure and pore-throat radius distribution of tight sandstone reservoirs [22,31,52,53]. Results from selected representative tight sandstone samples (Table 1) show that the maximum mercury saturation ranges from 64.25%–99.24% (average 81.19%), the relative sorting coefficient ranges from 0.17 to 0.58 (average 0.27), and the skewness ranges from 0.29 to 2.01 (average 1.37). The mean pore-throat radius is 0.14–0.74 µm (average 0.37 µm), the displacement pressure is 0.01–2.00 MPa (average 0.70 MPa), and the withdrawal efficiency is 20.08%–47.60% (average 37.04%).

Table 1. Results of high-pressure mercury intrusion for representative tight sandstone samples from the Study Area.

Sample No.	Maximum Mercury Saturation (%)	Relative Sorting Coefficient	Skewness	Mean Pore-Throat Radius (μm)	Displacement Pressure (MPa)	Mercury Withdrawal Efficiency (%)
#1	81.04	0.29	1.83	0.32	0.80	31.31
#2	78.25	0.18	1.51	0.61	0.03	38.08
#3	90.24	0.21	1.92	0.67	0.20	43.37
#4	87.71	0.23	2.01	0.64	0.15	40.08
#5	85.30	0.17	1.43	0.50	0.70	40.23
#6	82.24	0.27	1.73	0.28	0.70	39.23
#7	76.48	0.27	0.74	0.38	0.67	31.47
#8	79.69	0.22	1.34	0.50	0.53	32.53
#9	77.61	0.33	0.37	0.18	0.32	31.83
#10	73.86	0.29	0.46	0.41	0.33	31.47
#11	80.72	0.28	1.67	0.35	1.01	43.04
#12	79.68	0.25	0.71	0.33	0.67	30.90
#13	79.75	0.15	1.66	0.69	0.20	34.50
#14	77.55	0.32	1.67	0.33	0.60	37.22
#15	84.20	0.25	1.94	0.30	0.80	40.15
#16	74.79	0.33	0.53	0.21	0.50	40.21
#17	74.43	0.33	1.55	0.24	2.01	43.52
#18	79.41	0.27	1.56	0.38	1.50	47.60
#19	87.99	0.19	1.68	0.41	1.50	43.34
#20	86.80	0.22	2.01	0.33	0.70	30.86
#21	90.27	0.20	1.85	0.34	1.00	36.01
#22	88.33	0.21	1.93	0.33	0.70	41.79
#23	77.79	0.31	1.63	0.40	1.00	42.10
#24	68.73	0.31	0.49	0.15	0.50	28.37
#25	85.28	0.25	1.51	0.29	0.78	45.59
#26	84.90	0.29	1.42	0.32	0.40	38.12

Note: The relative sorting coefficient and skewness are dimensionless parameters.

The HPMI capillary pressure curves overall exhibit a two-stage characteristic (Figure 10a). During the initial low mercury saturation stage, the curve is relatively gentle, indicating the presence of a certain proportion of dominant flow channels within the reservoir. As mercury saturation increases, the curve rises sharply, revealing the pore space dominated by fine pores and throats. The overall curve morphology, characterized by a short gentle segment, a long steeply rising segment, and the absence of a distinct plateau, is typical of tight reservoirs with fine pore-throats and strong heterogeneity.

Figure 10. Pore structure characteristics revealed by (a) HPMI capillary pressure curves and (b) pore-throat radius distribution for different tight sandstone samples.

The pore-throat radius distribution curves (Figure 10b) show that the pore-throat radii are primarily distributed between 0.004 and 63.0 μm, presenting a unimodal pattern. According to the pore classification scheme of Xoaoth [54,55] for tight rocks, the pore structure is categorized as: macropores (radius > 1.0 μm), mesopores (radius 0.1–1.0 μm), transition pores (radius 0.01–0.1 μm), and micropores (radius < 0.01 μm). The pore structure of the tight sandstone reservoirs in the study area is dominated by mesopores (volume percentage 55.48%) and transition pores (volume percentage 32.39%), which together contribute approximately 87.87% of the total pore space. Micropores (volume percentage 10.04%) are also relatively well developed, whereas macropores (volume percentage 2.09%) are extremely poorly developed.

4. Discussion

4.1. Pore Geometry and Connectivity from Thin-Section Analysis

Pore structure is a key parameter determining reservoir properties. Its variability is governed by the combined effects of sedimentary and diagenetic processes, leading to different pore structure characteristics under varying lithofacies or diagenetic facies conditions. The primary macroscopic petrophysical parameters are φ and K, while pore structure encompasses diverse parameters, which can be categorized mainly into pore-throat types and abundances, pore-throat size and sorting, and pore-throat fluid characteristics.

Systematic observation of cast thin sections allows for the acquisition of microscopic structural parameters, including pore types, throat types, and pore and throat radii. The results show that the average specific surface area exhibits a negative correlation with both φ and K (Figure 11a,b). A high specific surface area primarily originates from the contribution of abundant micropores and nano-scale pore throats, reflecting well-developed internal matrix surfaces, complex pore-throat structures, and tortuous flow paths. This typically leads to a significant increase in fluid flow resistance and is accompanied by a considerable volume of ineffective storage space. Consequently, a lower specific surface area is one of the microscopic indicators of high-quality reservoirs, signifying a relatively simple pore structure dominated by large and medium-sized pores and throats.

Figure 11. Relationships between microscopic pore structure parameters observed from cast thin sections and reservoir properties under different lithofacies (The fewer data points for the H5 member reflect its currently lesser degree of exploration and core sampling in the study area). (a) Correlation diagram between porosity and specific surface area. (b) Correlation diagram between permeability and specific surface area. (c) Correlation diagram between porosity and mean pore diameter. (d) Correlation diagram between permeability and mean pore diameter. (e) Correlation diagram between porosity and pore-throat homogeneity coefficient. (f) Correlation diagram between permeability and sorting coefficient. (g) Correlation diagram between porosity and average throat width. (h) Correlation diagram between permeability and average throat width.

The mean pore diameter shows a good positive correlation with both φ and K (Figure 11c,d). This parameter directly characterizes the scale of the pore space; a larger mean pore diameter generally reflects a larger storage capacity and is often associated with a configuration of larger pores matched with coarser throats. The homogeneity coefficient exhibits a negative correlation with φ (Figure 11e), while the sorting coefficient shows a positive correlation with K (Figure 11f). Under deep, tight conditions, a higher homogeneity coefficient (more uniform pore-throat sizes) and a lower sorting coefficient (better sorting) paradoxically correlate with a declining trend in reservoir φ and K. This indicates that in an intense compaction and cementation diagenetic environment, a highly uniform pore-throat distribution often signifies the absence of a well-connected network of larger, effective flow pathways, with the overall pore system trending towards finer sizes. Therefore, moderate heterogeneity (the presence of a certain number of dominant, effective flow throats) is a necessary condition for forming relatively high K reservoirs.

The relationship between the average throat width and petrophysical properties shows significant complexity (Figure 11g,h). Average throat width exhibits a weak positive correlation with K but shows no clear correlation with φ. This suggests that the flow capacity in deep tight sandstone reservoirs is likely controlled primarily by throat connectivity, pore-throat configuration, or the tortuosity of the pore structure. Even if the average throat width is relatively large, K can remain low if the throats are highly isolated or the flow paths are extremely tortuous, reflecting the stronger heterogeneity and more complex pore structure characteristic of deep reservoirs.

4.2. Fluid Mobility and Pore Effectiveness from NMR Analysis

The occurrence state and distribution characteristics of fluids within pores directly control reservoir development efficiency, with pore structure being the core factor influencing fluid flow. NMR results indicate that the H4 reservoir overall possesses superior fluid mobility potential compared to the H5 reservoir. In the H4 reservoir, the movable fluid saturation exhibits a wide distribution range (16%–89%) with an average of 46%; the bound water saturation is relatively low (35%–86%, median 39%). This suggests that the H4 reservoir contains a relatively high proportion of effective pores and exhibits significant diversity in pore structure, with reservoir quality spanning from poor to good. In contrast, the movable fluid saturation in the H5 reservoir shows a narrow and concentrated distribution (36%–47%), averaging 44%, which is slightly lower than that of the H4 reservoir. Its bound water saturation is significantly higher (50%–96%, average 47%). This indicates that the H5 reservoir is overall denser, with strong pore structure heterogeneity dominated by micro-scale to nano-scale ineffective pores, resulting in limited movable fluid space.

Correlation results (Figure 12) indicate that lithofacies exert a primary control on fluid mobility. Porosity shows a negative correlation only with bound water saturation (Figure 12a,c), whereas K exhibits a significant positive correlation with movable fluid saturation and a negative correlation with residual water saturation (Figure 12b,d). Coarse-grained lithofacies (pebbly sandstone, coarse-grained sandstone) possess higher movable fluid saturation and K, reflecting their well-connected pore-throat networks. In contrast, fine-grained lithofacies (fine-grained sandstone, siltstone) demonstrate higher residual water saturation and lower K, resulting from their poorly connected micro-pore systems. This further confirms that the depositional facies control the initial pore framework, which, after diagenetic modification, ultimately governs the mobility of fluids within the reservoir.

Figure 12. Relationships between microscopic pore structure parameters obtained from NMR experiments and reservoir properties under different lithofacies (The fewer data points for the H5 member reflect its currently lesser degree of exploration and core sampling in the study area). (a) Correlation diagram between porosity and movable fluid saturation. (b) Correlation diagram between permeability and movable fluid saturation. (c) Correlation diagram between porosity and irreducible water saturation. (d) Correlation diagram between permeability and irreducible water saturation.

Furthermore, NMR T₂ spectra can intuitively reflect the pore radius distribution (Figure 13). Under the same diagenetic intensity background (Figure 13a), from coarse-grained to fine-grained sandstone, the T₂ spectral shape shows a regular variation: the T₂ cutoff value gradually shifts leftward (decreases), and the spectral peak shape changes from broad and gentle to sharp. This indicates that finer grain size leads to an overall trend of smaller pores and throat systems. A decrease in the T₂ cutoff value means the boundary between movable and bound fluid shifts towards smaller pore radii, resulting in a decrease in movable fluid saturation. This demonstrates that under identical diagenetic conditions, coarse-grained lithology provides the material basis for obtaining high fluid mobility. Under the same lithological grain size background (Figure 13b), as the pore-reducing effect of diagenesis intensifies from weak to strong, the T₂ cutoff value shifts significantly rightward (increases). This indicates that some originally medium-sized pores are compressed or plugged, transforming them into bound pores. Concurrently, the decreasing bimodal amplitude difference in the T₂ spectra suggests a sharp decline in the proportion of large pores, with the pore size distribution becoming increasingly concentrated in the smaller pore range. The direct consequence is a substantial reduction in movable fluid saturation and a deterioration in reservoir quality.

Figure 13. Variation characteristics of NMR T₂ spectra under identical (a) diagenetic intensity and (b) grain-size conditions.

4.3. Pore-Throat Network Characteristics from High-Pressure Mercury Injection Analysis

HPMI experiments characterize the pore-throat system controlling fluid flow by measuring the intrusion and extrusion processes of the non-wetting phase (mercury) under high pressure. Key parameters derived from the mercury intrusion curves include: the mean pore-throat radius, representing pore-throat size; the pore-throat homogeneity coefficient and sorting coefficient, representing pore-throat sorting (i.e., the concentration of the size distribution); and the skewness and kurtosis, representing the distribution shape. Among these, better throat sorting indicates a more concentrated distribution and weaker heterogeneity.

The maximum mercury saturation shows a positive correlation with both φ and K (Figure 14a,b). This parameter represents the total volume of the interconnected pore network filled with mercury at high pressure, reflecting the scale of potentially effective storage space in the reservoir. In contrast, the withdrawal efficiency exhibits a negative correlation with φ and K (Figure 14c,d). A low withdrawal efficiency indicates that a significant volume of mercury remains trapped in the rock after pressure release, corresponding to inefficient or ineffective pore space controlled by extremely fine or bottleneck throats. Therefore, a combination of high maximum mercury saturation and high withdrawal efficiency is indicative of a high-quality reservoir, signifying a well-connected and highly effective pore network. Conversely, high maximum mercury saturation coupled with low withdrawal efficiency is a typical characteristic of strongly heterogeneous reservoirs, often manifesting macroscopically as high φ and low K. Compared to maximum mercury saturation, withdrawal efficiency is a more critical indicator for screening effective reservoirs.

Figure 14. Relationships between microscopic pore structure parameters obtained from HPMI experiments and reservoir properties under different lithofacies (The fewer data points for the H5 member reflect its currently lesser degree of exploration and core sampling in the study area). (a) Correlation diagram between porosity and maximum mercury saturation. (b) Correlation diagram between permeability and maximum mercury saturation. (c) Correlation diagram between porosity and withdrawal efficiency. (d) Correlation diagram between permeability and withdrawal efficiency. (e) Correlation diagram between porosity and relative sorting coefficient. (f) Correlation diagram between permeability and relative sorting coefficient. (g) Correlation diagram between porosity and mean pore-throat radius. (h) Correlation diagram between permeability and mean pore-throat radius.

The relationship between the relative sorting coefficient and petrophysical properties exhibits a distinct threshold value of 0.22 (Figure 14e,f). When the coefficient is less than 0.22, it shows a positive correlation with properties, indicating that good sorting against a background of larger pore-throats is beneficial for seepage. When the coefficient exceeds 0.22, the correlation becomes negative. In this case, a higher sorting coefficient signifies extremely non-uniform pore-throat sizes, primarily caused by a substantial influx of newly formed, extremely fine throats, which severely degrades flow capacity. Consequently, reservoir evaluation requires a combined assessment that considers the absolute size of the pore-throats: good sorting within a large-throat background indicates a high-quality reservoir, whereas poor sorting within a fine-throat background signifies a poor reservoir.

The relationship between the mean pore-throat radius and K displays complexity (Figure 14g,h). A positive correlation exists only within the range where the mean pore-throat radius is less than 0.4 µm. Beyond this value, the correlation weakens significantly or even disappears. This indicates that in deep reservoirs, when throats are generally fine, their average size is the primary controlling factor. However, once the throats reach a certain scale, the dominant control on K shifts to the three-dimensional spatial distribution and connectivity of the dominant flow throats, rather than their average size. Structures such as “isolated larger throats with a network of finer throats” or “poor pore-throat configuration” may exist, leading to low K despite a large average radius.

4.4. Differential Impact of Sedimentary–Diagenetic Coupling on the Microscopic Pore Structure of H4 and H5 Reservoirs

Synthesizing the multi-scale experimental data, the microscopic pore structure of the deep tight sandstone reservoirs in the Huagang Formation, Xihu Depression, is the result of the spatio-temporal dynamic coupling between sedimentary and diagenetic processes. The coarse-grained, well-sorted, quartz-rich sandstones deposited in high-energy channels possess a triple advantage. Firstly, their rigid, grain-supported framework provides strong resistance to compaction, aiding in the preservation of primary pore space (Figure 9b). Secondly, their clean grain surfaces are conducive to the early formation of chlorite coatings (Figure 9c) [56,57,58]. These coatings effectively inhibit the later destructive cementation by quartz overgrowths during early diagenesis (Figure 8a,g), thereby preserving the foundation for constructive diagenesis.

In contrast, the fine-grained, matrix-rich sandstones deposited in low-energy channels are predisposed to a disadvantageous foundation of being susceptible to intense compaction and cementation from the onset of burial. Firstly, the high content of ductile components subjects the rock to extremely strong mechanical compaction during early burial [16,17,32]. The deformation of ductile grains forms pseudomatrix, leading to the rapid reduction in primary pore space (Figure 7b,d and Figure 9a). Secondly, the pores that survive compaction are predominantly isolated and extremely fine (Figure 7a,c), and their surfaces are extensively coated by clay matrix (Figure 8c–f). This dense and isolated pore system severely hinders the effective migration of formation fluids, making it difficult for later potentially corrosive fluids to penetrate. Consequently, dissolution is weak, and any secondary pores created tend to be isolated, failing to improve overall connectivity [23,31]. Simultaneously, the abundant clay mineral surfaces and the confined pore spaces provide ample nucleation sites for the precipitation of silica and carbonate cements [52,53], promoting the widespread development of destructive cementation.

Building upon these different sedimentary “presets”, subsequent diagenesis led to significant divergence in pore structure. For the coarse-grained sandstones in high-energy facies, diagenesis followed a constructive pathway. The primary pore network preserved by chlorite coatings [14,19,59] provided efficient pathways for the migration of organic acids [14,19,26], causing intense dissolution of feldspar and lithic fragments (Figure 9d–g) and generating abundant secondary dissolution pores that significantly improved pore connectivity. Both NMR and HPMI data confirm that such reservoirs possess high movable fluid saturation, larger mean pore-throat radius, and high mercury withdrawal efficiency (Figure 12, Figure 13 and Figure 14). For the fine-grained sandstones in low-energy facies, diagenesis was dominated by a destructive pathway. Intense mechanical compaction drastically reduced pore space [27,60,61]. The lack of protective coatings allowed pervasive silica and carbonate cementation. Dissolution was weak and produced mostly isolated pores, ultimately resulting in a dense system dominated by micropores and sheet-like/tubular throats with poor connectivity (Figure 8i–k), which manifests macroscopically as high φ and low K [62,63].

To address the interplay of multiple variables (burial depth, grain size, diagenesis), we distinguish their influences as follows: First, within the same reservoir (H4) member at similar burial depths, reservoir quality varies significantly with lithofacies. Coarse-grained, high-energy facies exhibit superior properties compared to fine-grained, low-energy facies (Figure 12), underscoring the primary control of sedimentary facies on the initial pore framework. Second, the key difference between H4 and H5 is not merely ~300 m of additional burial, but the facies-dependent diagenetic overprint that the increased burial depth and temperature facilitated. The H5 member, with an overall finer-grained depositional background, experienced a more pervasive destructive diagenetic pathway (intense compaction and cementation) under deeper burial, whereas the coarser-grained H4 facies, even at shallower depth, were predisposed to a constructive pathway (chlorite coating and dissolution). This coupling of sedimentation and diagenesis at different burial depths ultimately created the strong microscopic heterogeneity of the reservoirs and fundamentally controlled the comprehensive petrophysical differences between the H4 and H5 reservoirs (Figure 5). Therefore, we interpret that burial depth acted as an intensifier of diagenetic processes, but the specific diagenetic pathway was fundamentally predetermined by the sedimentary facies (grain size, composition and water energy).

It is noteworthy that the overall finer-grained character and stronger diagenetic overprint of the H5 member are consistent with regional observations and its current status as a reservoir with relatively low gas production capacity, which also explains the relatively smaller core dataset available for H5 in this study. The spatiotemporal coupling of these processes, leading to divergent reservoir outcomes, is conceptually summarized in Figure 15.

Figure 15. Conceptual model illustrating the divergent diagenetic pathways controlled by sedimentary facies in the deep tight sandstones of the Huagang Formation. (a) The constructive pathway, dominant in high-energy, coarse-grained facies, is characterized by chlorite coating preservation and subsequent dissolution, leading to well-connected pore networks. (b) The destructive pathway, prevalent in low-energy, fine-grained facies, is dominated by intense compaction and cementation, resulting in isolated microporous systems. This divergence underpins the “binary” pore structure and the differences in reservoir quality between the H4 and H5 members.

4.5. A “Binary” Microscopic Pore Structure Development Pattern for Deep Tight Sandstone Reservoir in the Xihu Depression, East China Sea Basin

Observations from different perspectives—including image observation (SEM), fluid behaviour (NMR) and pore-throat access (HPMI)—have collectively revealed that K is not merely controlled by the microscopic pores and throats themselves but, more critically, by the heterogeneous three-dimensional distribution of structural units composed of pores and throats with distinct functions, termed “pore-throat assemblages” (Figure 16). Effective pore-throat assemblages refer to clustered or banded flow units formed by multiple pores effectively connected by relatively coarse and well-connected throats, constituting the dominant seepage network within the reservoir. These assemblages typically correspond to coarse-grained units from high-energy deposition, intense dissolution zones, or microfracture-developed areas. The movable fluids detected by NMR primarily reside and flow within these assemblages (Figure 12a,b). In contrast, ineffective (or isolated) pore-throat assemblages consist of pore clusters severely blocked by fine throats, sheet-like throats, or cements, exhibiting very poor or no connectivity to the dominant flow network, thereby constituting the ineffective storage space of the reservoir. These ineffective assemblages typically develop in lithofacies characterized by high original matrix content and subjected to intense compaction [19,30,33]. The pseudomatrix formed during compaction and the subsequently formed authigenic clay minerals are the primary causes of the development of sheet-like throats and pore blockage (Figure 8j,k). They compartmentalize and isolate larger pores, which, although contributing to porosity, are almost entirely excluded from effective fluid flow. NMR analysis indicates that the bound fluid primarily resides within this type of ineffective storage space (Figure 12c,d). The “pore-throat assemblage heterogeneity” framework provides a unified and deeper explanation for the differences between the H4 and H5 reservoirs. The movable fluid parameters in the H4 reservoir correlate with multiple factors, indicating that the effective assemblages within it are diverse in type and complex in origin (e.g., influenced jointly by dissolution, grain size, and fractures), and their spatial distribution exhibits a certain degree of randomness. This results in strong reservoir heterogeneity but allows for the presence of local sweet spots. In contrast, the movable fluid saturation in the H5 reservoir shows a strong correlation only with K, indicating that the reservoir is overall dense, with effective assemblages being extremely scarce and associated only with a few effective throats or microfractures. The reservoir quality exhibits a “binary” distribution characteristic: a small number of connected fracture dissolution type assemblages coexist with a large number of isolated compaction cementation type assemblages.

Figure 16. A “Binary” microscopic pore structure development pattern for deep tight sandstone reservoir in the Xihu Depression, East China Sea Basin.

This understanding also reasonably explains the previously mentioned seemingly paradoxical phenomena. In the H5 reservoir, the lack of correlation between average throat width and petrophysical properties (Figure 11g) occurs because the coarse throats within the few effective assemblages are averaged out by the vast number of extremely fine throats within the abundant ineffective assemblages, rendering the statistical mean unrepresentative. Permeability is actually determined solely by the former. Similarly, the negative correlation between the homogeneity coefficient and petrophysical properties (Figure 11e) indicates that a highly uniform pore-throat distribution precisely signifies a complete lack of effective assemblages within the rock, which is instead composed entirely of uniformly fine pores and throats, leading to very low K. Therefore, a moderate degree of macroscopic heterogeneity is a necessary condition for the development of local, dominant flow pathways.

Therefore, a “binary” microscopic pore structure development pattern for the deep tight sandstone reservoirs in the Xihu Depression, East China Sea Basin, can be established (Figure 16). The pattern posits that the reservoir space is a highly heterogeneous, three-dimensional mosaic composed of a minority of effective percolation assemblages that control fluid flow and a majority of ineffective/isolated assemblages that contribute minimally. The ultimate flow capacity of the reservoir is determined by the spatial abundance, connectivity, and three-dimensional configuration of the former.

The “binary” pore structure model proposed herein distinguishes itself from conventional pore network or percolation-based models by its emphasis on the three-dimensional, heterogeneous distribution of functional “pore-throat assemblages” as the fundamental units controlling flow, rather than focusing solely on the properties of individual pores or throats. While existing models often use threshold parameters (e.g., throat radius) to statistically define effective pores, our model conceptualizes that flow is dominated by a minority of spatially connected clusters or pathways (effective pore-throat assemblages) that arise from specific sedimentary-diagenetic histories. This perspective aligns with percolation theory but is grounded in and interpreted through our geological observations. To provide a quantitative foundation, the “effective pore-throat assemblages” can be indirectly defined and characterized using integrated experimental data: (1) The loss of correlation between permeability and average throat radius beyond ~0.4 μm (Figure 14g,h) suggests that throats larger than this scale, if poorly connected, do not guarantee flow capacity. Therefore, throat radius alone is an insufficient descriptor. (2) Parameters such as high movable fluid saturation (Figure 12b) serve as robust proxies for the abundance and connectivity of these effective assemblages, as demonstrated by their strong positive correlation with permeability.

5. Conclusions

This integrated multi-scale study on the deep tight sandstones of the Huagang Formation in the Xihu Depression reveals that the reservoir quality, characterized by typical high-porosity and low-permeability, is fundamentally controlled by a sedimentary–diagenetic coupling mechanism. Petrographic and petrophysical analyses confirm systematic differences between the H4 and H5 members: the H4 member exhibits superior quality with higher movable fluid saturation and better pore connectivity, dominated by mesopores (volume percentage 55.48%) and transition pores (volume percentage 32.39%), whereas the H5 member is denser with a higher proportion of bound fluid. This divergence stems from facies-dependent diagenetic pathways. High-energy, coarse-grained, quartz-rich facies underwent a constructive pathway where early chlorite coatings inhibited destructive quartz overgrowths and facilitated intense dissolution, enhancing secondary porosity and connectivity. In contrast, low-energy, fine-grained, matrix-rich facies followed a destructive pathway dominated by intense compaction and pervasive cementation, resulting in isolated micropore systems. The permeability of these deep reservoirs is not primarily governed by porosity or average throat size but by the three-dimensional connectivity and spatial distribution of a minority of well-connected “effective pore-throat assemblages,” as evidenced by the strong correlations between NMR-derived movable fluid saturation, HPMI withdrawal efficiency, and permeability. Accordingly, a “binary” pore structure development model is established, conceptualizing the reservoir as a heterogeneous mosaic of connected effective assemblages and isolated ineffective ones. The validity of this model and the proposed coupling mechanism is cross-validated by the consistency among data from cast thin sections, SEM, XRD, NMR, and HPMI. For exploration, this study recommends that sweet spot prediction should focus on identifying intervals where high-energy sedimentary facies coincide with the diagenetic overprint of chlorite coating preservation and subsequent dissolution. Future research could quantify the 3D geometry of these effective storage assemblages through digital rock physics to further refine predictive models.