The analysis of drilling cores, sidewall cores, and fluorescence thin sections reveal that there are several types of reservoirs in the study area, including fractured, fractured-dissolved, and dissolved pore/vug reservoirs, with the fractured-dissolved reservoirs in dominance. A large number of fractures can be observed on the drilling cores. The fractures have straight surfaces, indicating that they are mainly structural fractures. The statistics of drilling cores show that fracture opening varies greatly, in a range of 0.5–8 mm. There are abundant dissolved pores/vugs along the fractures, with maximum diameter of 1.3–12 mm. Note that the dissolved pores/vugs are mainly distributed along the structural fractures (Figures 2A–C).

Numerous structural fractures can be seen on the thin section, with the characteristics of mutual cutting, indicating at least two stages of fracture initiation events (Figure 2D). Microscopically, many fractures are obverse to have been filled with calcite in the late stage (Figure 2D). The fillings are obviously different, and can be divided into silty-fine crystal filling and coarse crystal filling. There are also many unfilled fractures, along which strong dissolution characteristics can be seen, which is consistent with the observation results of drilling cores (Figures 2E, F). Asphalt can be seen in the fractures in the thin section (Figure 2G). In the fluorescence thin section, abundant oil and gas shows can be seen, which indicates that both structural fractures and dissolved pores in the study area are the main effective storage spaces (Figures 2H, I).

The physical properties of 105 sidewall cores were measured. The porosity is 0.1%–9.3%, concentrated at 1%–3%, and a lot of samples exhibit the porosity ˂ 5% (Figure 3A). The permeability is mainly ˂ 10 mD, and a lot of many samples exhibit the permeability of 0.01–1 mD. The cross plot shows that the porosity and the permeability are less correlated (correlation coefficient was 0.2208), indicating that the reservoir is mainly a fractured-porous medium (Figure 3B), which is consistent with the observation results of drilling cores and thin sections.

Structural fractures are extremely developed in the study area. Numerous structural fractures are observed on cores and thin sections. The occurrence and dip angle of the structural fractures were depicted by electric imaging logging. It is found that there are mainly two groups of fractures in the CFD area: NEE, and NWW, by tending, with the NEE-trending fractures in dominance. The structural fractures show an obvious sine dark curve on the electric imaging log, and medium-sized pores/vugs can be seen along the structural fractures. In addition, some fractures are relatively wide and may be dissolved by atmospheric water (Figures 4A, B). The fractures are quite different in dip angle: the NWW-trending fractures have the dip angles of 30°–80°, with an average of 48°; the NEE-trending fractures have the dip angles of 50°–80°, with an average of 67°. Clearly, the NEE-trending fractures show higher dip angles than the NWW-trending fractures, possibly suggesting that they were formed in different tectonic settings (Figures 4C–F).

Fracture fillings are important for revealing diagenetic fluids. There are a huge amount of calcite fillings in the structural fractures in the study area. The C and O isotopes of fillings in fractures and surrounding rocks were analyzed to determine the source of diagenetic fluids. It is found that the C isotope of surrounding rocks is generally positive and occasionally negative, mostly greater than −1‰, and the O isotope is between −18‰ and −4‰, representing the isotopic characteristics of seawater during sedimentation.

The fracture fillings exhibit significantly different C and O isotopes from the surrounding rocks. According to the crystallization degree, they can be divided into silty-fine crystal dolomite and limestone, and coarse crystal dolomite and limestone, which are different in the C and O isotopes. For the silty-fine crystal dolomite and limestone, the C isotope ranges from −2‰ to 0‰, and the O isotope from −18‰ to −8‰. For the coarse crystal dolomite and limestone, the C isotope ranges from −4‰ to −2‰, and the O isotope from −18‰ to −8‰. It can be seen that the coarse crystal fillings are equivalent to the fine crystal fillings in the O isotope, but display a lower C isotope generally, possibly indicative of different diagenetic environments (Figures 5, 6; Table 1).

The formation and distribution of structural fractures are often related to strong tectonic activities (Ye et al., 2022b; Guo et al., 2022). In the study area, the fractures were mainly formed due to the Indosinian and Yanshanian tectonic activities, with the former inducing NWW-trending fractures and the latter inducing NEE-trending fractures (Figure 4).

Based on the geological profile, the tectonic evolution process of the CFD area was restored. This area deposited a set of epicontinental sea carbonate during the early Paleozoic and a set of transitional deposits during the late Paleozoic (Figure 7A). The collision between the South China Plate and the North China Plate in the Indosinian induced a series of NWW-trending thrust faults in the study area, and also caused the Paleozoic fold deformation (Figure 7B). The NWW-trending faults stopped their activity in the Yanshanian (Figure 7C). At this time, a group of NEE-trending strike-slip faults was formed, and they were mainly featured by sinistral compression, leading to the uplift and fold deformation of the strata near the fault belt (Figure 8). Since the Himalayan, the NWW-trending faults had been reactivated and controlled the deposition of Paleogene, and the buried hill was finally buried and shaped (Figures 7D, E).

The multi-phase tectonic activities in the buried hill provided favorable external stress for the development of structural fractures. During the Indosinian compression, the CFD area was close to the thrust fault belt, and a large number of NWW-trending fractures were developed near the NWW fault belt. For example, the imaging logging data of Well CFD2-F and Well CFD2-C reveal that the NWW-trending fractures have low dip angles, possibly because they were formed in the thrust nappe setting.

During the formation of the strike-slip fault, intense fracturing deformation might occur along the fault belt, thus deriving a great number of structural fractures (Zhang et al., 2010; Liu J. S. et al., 2022; Yun and Deng, 2022). However, the distribution of fractures controlled by strike-slip faults is complex, which is related to the specific spatial shape of faults and the derived local stress field. Strike-slip faults present differences role in the distribution of fractures, in which the stress in the strong deformation zone of the strata near the strike-slip fault zone is the most concentrated and is the main development zone of fractures. The structural fractures formed by the strike-slip fault belt are mainly located inside the main fault belt, but when the trend of the strike-slip fault belt changes, the local stress field derived from it forms the “sweet spot” of fractures (Yu et al., 2014; Li et al., 2021). Field measurement shows that obvious pressure increases and pressure-releasing belts are formed at the strike-slip bending position. The pressure-increasing belt is often accompanied by terrain uplifting and the creation of numerous structural fractures, while the pressure-releasing belt mostly forms graben with few fractures (Wei, 2015). Abundant NEE-trending fractures were developed near the NEE fault belt during the Yanshanian, especially in the strong uplifting area of the buried hill caused by strike-slip faults, where the fracture density is the largest. For instance, all the fractures in Wells CFD2-D, CFD2-E, CFD2-F, and CFD2-C mainly trend in the NEE direction (Figure 8). It is worth noting that the structural fractures induced by strike-slipping have high dip angles, and they exhibit higher effectiveness than the NWW-trending fractures.

The formation of a high-quality buried hill reservoir is not only affected by external factors but also related to the physical properties of the rock itself (Hou et al., 2015; Wang et al., 2015). The compressive strength of rock determines its ability to form fractures. The lower the compressive strength, the more easily the rock forms fractures (Guo et al., 2018). Compressive strength tests on different lithologies confirm that the compressive strength is 49 MPa for dolomite, 70.74 MPa for limy dolomite, 75.4 MPa for dolomitic limestone, and 104.9 MPa for limestone; that is, with the increase of dolomite content in carbonate, the compressive strength tends to decrease gradually (Figure 9B). It is worth noting that the compressive strength of dolomite is only half of that of limestone, that is, the stress required for dolomite to break and form structural fractures is only half of that of limestone. This means that the higher the content of dolomite, the easier the rock is to break and form structural fractures.

The linear density of fractures in intervals with different dolomite contents was calculated by imaging logging data, element well-logging data, and mud logging data. The results show that, except for a small amount of limestones (low dolomite content) that exhibit high fracture density, the overall fracture density increases significantly with the increase of dolomite content (Figure 9A), which also confirms the control of lithology on fracture distribution in the study area. A lot of drilling data verify that at the structural high near the fault belt, both limestone and dolomite can form high-quality reservoirs, while at the structural low and inside the buried hill, high-quality reservoirs are mainly developed in intervals with high dolomite content. Essentially, in the interior of the buried hill and at the structural low, where the stress is weak, the limestone is difficult to form fractures, while the dolomite with lower compressive strength is broken to become reservoirs.

Structural fractures provide good pathways for fluid flow, which is the basis for the development of karst reservoirs. The negative carbon isotope anomaly is mainly related to the influence of organic carbon; The negative oxygen isotope anomaly is mainly affected by the leaching of hydrothermal solution or atmospheric water. According to the observation of fracture fillings and the analysis results of C and O isotopes, at least two stages of karstification have occurred in the study area. In the first stage, the fillings were mainly silty-fine crystal carbonate, with the C isotope ranging from −2‰ to 0‰ and the O isotope ranging from −16‰ to −8‰, which may represent the atmospheric freshwater dissolution in the supergene period. In the second stage, the fillings were mainly coarse crystal carbonate, with the C isotope ranging from −4‰ to −2‰ and the O isotope ranging from −18‰ to −8‰, indicating that the diagenetic temperature was higher than that in the first stage, which may be related to karstification during burial.

Combined with the tectonic evolution process and the above analysis, the development model of Carbonate buried hill reservoirs in the CFD area was established. In the study area, fractured-dissolved reservoirs are dominant. The early structural fractures provide flow pathways for later fluid dissolution, which is the key to controlling the development of reservoirs. The distribution of high-quality reservoirs is jointly controlled by structure, lithology, and fluid. Before Indosinian period, the tectonic setting was relatively quietly (Figure 10A). During the Indosinian, nearly NWW folds were formed in the study area due to SN compression, and intensive fracture belts were formed near the Indosinian thrust fault belt. Meanwhile, the first supergene karstification occurred (Figure 10B). Subsequently, the NEE intensive fault belt was formed near the strike-slip faults by the strike-slip transpression process during the Yanshanian. At this time, the Paleozoic buried hill was still exposed and the supergene dissolution continued (Figure 10C). Under this background, the fracture fillings showed a relatively high C isotope. In the Himalayan, due to the continuous extension, the Paleozoic strata experienced burial karstification along bedding on the slope, leading to more negative C isotope of the fillings, due to large burial depth (Figure 10D). With the continuous burial in the Cenozoic, the O isotope became generally negative, and finally the present reservoir pattern was formed (Figures 10E, F).

Based on the above analysis, the development model of fractured-dissolved reservoirs was established. The stress is most concentrated near the Indosinian and Yanshanian faults, which are the dominant areas with fractures developed. Dolomite is more brittle than limestone, and may more easily form structural fractures in the process of structural transformation, thus it is the lithology favorable for reservoir development. The preexisting structural fractures experienced two types of dissolution with different properties in the supergene stage and the burial stage, which further improved the physical properties of the reservoirs and guaranteed the development of effective high-quality reservoirs (Figure 11).