**1. Introduction**

Hydraulic fracturing, also known as natural hydraulic fracturing and hydraulic extensional fracturing, is a common kind of geological phenomenon which refers to fractures caused by the increase of pore fluid pressure [1–5]. It covers the hydraulic effects on intact rocks and fractured rocks, which play an important guiding role in fluid migration, oil and gas preservation, and the safe exploitation of oil and gas fields. Research on hydraulic fracturing began in the late 1960s [1]. In a study on the law of joint development, Secor proposed the mechanism of natural tensile fractures and pointed out that the fracture caused by fluid overpressure can open original fractures in the formation while also playing an important role in the migration of groundwater, oil and gas, and ore-forming fluids [6,7]. After that, Phillips formally proposed the concept of hydraulic fracturing in his study of the

**Citation:** Jia, R.; Fan, C.; Liu, B.; Fu, X.; Jin, Y. Analysis of Natural Hydraulic Fracture Risk of Mudstone Cap Rocks in XD Block of Central Depression in Yinggehai Basin, South China Sea. *Energies* **2021**, *14*, 4085. https://doi.org/10.3390/en14144085

Academic Editors: Sheng-Qi Yang, Min Wang, Qi Wang, Wen Zhang, Kun Du and Chun Zhu

Received: 18 May 2021 Accepted: 2 July 2021 Published: 6 July 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

formation mechanism of mineralized normal faults in Wales, the UK, which he described as the process of fracture expansion caused by the increase of pore fluid pressure within the fracture [2]. In 1990, Sibson analyzed the migration process of abnormally high pressure fluid near the fault, described the dynamic process of fault hydraulic rupture, and pointed out that this process is periodic, intermittent, and pulsating [8]. Subsequently, the research on hydraulic fracture was deepened, many overpressure basins around the world were analyzed and studied, and it was clearly pointed out that hydraulic fracture is a potential risk leading to oil and gas leakage [9]. For example, in the research of Timor Sea, they pointed out that the fault reactivity caused by hydraulic fracturing is the main cause of oil and gas leakage [10].

Through the study of the development of hydraulic fracture, the hydraulic fracture phenomenon is similar to the fault valve mechanism [11]. It has the characteristics of episodic eduction, especially in the overpressure basins. When the pore fluid pressure exceeds the rock fracturing pressure, rocks will begin fracturing leading to highly permeable channels for fluid to flow into which carries a large quantity of minerals through the fractures. In the process, pore fluid pressure is released and gradually reduced. The mineral cementation, precipitation, and pressolution gradually decreases the width of the fractures which finally forms a closed hydraulic fracture until the fluid pressure rises to enter the next "broken-closed" cycle.

Studies have also shown that these hydraulic fractures have specific geological characteristics both macroscopically and microscopically. For example, due to the periodic activities of hydraulic fracturing, different periods of hydraulic fracturing and fluid migration lead to changes in the formation structure and in 3D seismic imaging. Furthermore, the amplitude intensity is usually increased or decreased locally along the reflection layer and it forms the vertical continuous fuzzy zone and pipe features [12,13]. In the field of outcropping, it is mainly manifested as extensional fractures extending perpendicular to the direction of minimum principal stress and filled by quartz, calcite, and gypsum veins [14–16]. The multi-stage filling of fractures by fluid minerals can also be observed under a microscope.

How to effectively and accurately determine the critical condition of hydraulic fracturing is particularly important to quantitatively evaluate the periodicity of hydraulic fracturing. Through a lot of experience, many scholars regard 85% of the overburden lithostatic pressure as the critical pressure for hydraulic fracturing to occur [6,17–20]. Similarly, a study on the Norwegian Snorre field found that when the pore fluid pressure at the top of the reservoir reaches 82% of the static pressure, approximately equal to the minimum horizontal principal stress, hydraulic fracturing increases the permeability of mudstone and leads to the escape of oil and gas from the reservoir [21]. Some investigators propose that the critical pore fluid pressure for hydraulic fractures is the sum of the principal of minimum stress and the tensile strength of cap rock [1,22–25]. Gaarenstroom defined the difference between the minimum horizontal principal stress and pore fluid pressure as the retention capacity in his study on the risk of hydraulic fracture in the cap rock of the North Sea Basin. The larger the pore fluid pressure is, the retention capacity will be smaller, and the risk of hydraulic fracture in the cap rock will be greater. He found that when the retention capacity is lower than 7 MPa, the risk of hydraulic fracture will significantly increase [24]. Most of these evaluation methods only consider the risk of hydraulic tensile fracture of rock and ignore the tensile strength of rock. However, the actual cap rock is often heterogeneous, especially when there is strong and weak interbedding in the cap rock (sandstone and mudstone are both distributed), and the cap rock may form shear fracture or mixed-mode fracture failure. Therefore, this simple evaluation method is obviously not applicable to evaluate the hydraulic fracture risk of deep overpressure cap rock such as Yinggehai Basin. At present, we mainly use the Fault Analysis Seal Technology (FAST) for the hydraulic fracture evaluation [10,26]. The Fast evaluation method is a comprehensive method proposed by Mildren et al. in 2002 based on the principle of rock rupture. This method can be used to evaluate the conditions and types of faults or fractures formed by

rock rupture and the conditions under which the faults reactivity or maintain stability. The basic principle is whether the formation fluid pressure under the current stress and pressure conditions can cause the structure to reactivate or remain stable. The quantitative evaluation results of the FAST method are in good agreement with the actual situation, and it has been widely used in oil and gas field production and gas storage [10,26]. The data required by the FAST method mainly include the characteristics of present stress field, fluid pressure, friction strength, and cohesion of the rock. In this study, although there is a lack of large-scale fault development in the study area, a large number of fractures are still developed in the cap rock. Therefore, we can still use the FAST method to evaluate whether hydraulic fracture will be induced by fluid pressure in the cap rocks. It can be used to evaluate not only the risk of complete cap fracture, but also the risk of the occurrence of pre-existing extensional fracture and shear fracture before movement takes place, which is an effective method for studying the integrity of cap rocks.

In recent years, many scholars have done a lot of research on the effectiveness of cap rocks. We keep focusing on the study of hydraulic fracturing on the effectiveness of cap rock and the risk of oil and gas leakage and have used this method to analyze hydraulic fractures of cap rocks in several blocks in different basins [27–32] which have achieved some good results. Especially in the study area, a lot of preliminary research has actually been done on the cap rocks, such as the macro- and micro-development characteristics and sealing mechanism and sealing ability of the cap rocks. Some scholars have made preliminary studies on the migration and preservation conditions of natural gas caused by hydraulic fracturing, but most of them focus on theoretical and qualitative analysis. In order to further study the influence of hydraulic fracturing on natural gas enrichment in different structural locations, the same FAST evaluation method was used in this study. Based on the in-situ stress characteristics, mudstone mechanics and pore fluid pressure are used to analyze the potential of mudstone hydraulic fracturing in present conditions, and combine with the paleo-pressure to analyze the hydraulic fracture risk in geological history to discuss its influence on gas accumulation and provide theoretical guidance for oil and gas exploration in basins with the same geological background.

#### **2. Geological Setting**

The Yinggehai Basin is a typical Cenozoic sedimentary basin on the northwestern continental shelf of the northern South China Sea [33–38]. The northeast side of the basin adjoins the southern uplift of the Beibu Gulf Basin and the Hainan Uplift. The west side is connected to the Kunsong Uplift and it extends to the Hanoi Depression in Vietnam. This basin is 750 km long and 200 km wide, with an area of about 11.3 × 104 km2. Yinggehai Basin is located in the convergence zone of the Eurasian, Pacific, and Indo-Australian plates [38]. Under the background of the expansion of the South China Sea, it is restricted by the strike-slip activity of the Red River Fault and the compression of the plate [39]. The basin is distributed in a rhomboid shape along the NW-SE direction, and composed of four first-order structural units: The Central Depression, The Yingdong Slope Zone, The Yingxi Slope Zone, and The Lingao Uplift (Figure 1).

The basin has experienced multi-stage regional tectonic movements, which can be divided into four stages: Eocene—Early Oligocene rift period, Late Oligocene fault— Depression period, Early—Middle Miocene post-rift thermal subsidence period, and Late Miocene—Quaternary accelerated subsidence period. Drilling data shows that the strata drilled from bottom to the top of the basin are as follows: The Lingtou Formation, The Yacheng Formation, The Lingshui Formation, The Sanya Formation, The Meishan Formation, The Huangliu Formation, The Yinggehai Formation, and The Ledong Formation (Figure 2).

**Figure 1.** The structure outline map of Yinggehai Basin.

The Central Depression developed a series of nearly N–S orientations diapir structures, which is distributed in an echelon arrangement [35,37]. The research area is located in the southeast of the central depression in the Yinggehai Basin (Figure 1), which is characterized by a lack of fracture, rapid settlement, high temperature and high pressure, and it has more mudstone and less sand. The Yinggehai Basin experienced rapid subsidence in the late structural period, with a deposition rate of 500–1400/Ma, and the Cenozoic maximum sedimentary thickness is over 17 km [34]. Test data of the both the drill pipe and the cable indicates that deep formation pressures below 4000 m exceed 100 MPa, with a maximum pressure factor of 2.30. The numerical simulation of different overpressure formation mechanisms shows that due to the limited abundance of organic matter and the weak contribution of hydrocarbon generation and compression, the abnormally high pressure of the formation comes from the under-compaction caused by the rapid deposition of mudstone, and the conduction effect of the diapir structure acting on the deep overpressure [40,41].

The Ledong area has developed mainly two types of traps, tectonic lithologic trap under the diapir background and lithologic trap group developed within the background context of gravity flow deposition. The Sanya Formation and the Meishan Formation in the Miocene are shallow marine facies mudstones which is the main source rocks in study area. The mudstones in Huangliu Formation and the Yinggehai Formation in upper Miocene and Quaternary Formation are the main cap rocks for nature gas. In addition, these strata are the main reservoir-forming assemblage in the study area (Figure 2). The longitudinal enrichment layers of natural gas are quite obviously different in various structural positions. In the diapir zone, oil and gas are mainly enriched in the shallow strata (Yinggehai and Ledong Formation), such as XD-A area and XD-B area, but natural gas is generally enriched in deeper layers (the Meishan and the Huangliu Formation) and in the slope zones which

are far from the diapirs, such as XD-C, XD-D, XD-E area (Figure 3). As a whole, the farther from the diapir, the deeper the gas enrichment. Published literature indicates that there are signs of large-scale fluid migration in the Yinggehai Basin (e.g., gas chimney, pockmark, and pipe) the origins of which many scholars attribute to hydraulic fracturing by overpressure fluid [36]. We speculate that the difference of vertical distribution of natural gas may be caused by the difference in hydraulic fracture strength in different structural positions. Therefore, based on the in-situ stress in the study area, this paper will use the fluid pressure in mudstone and the mechanical parameters of mudstone to analyze the risk of hydraulic fracture in the diapir area and the slope area in the long-term geological historical course as well as the current period. Furthermore, we clarify the main reasons for the natural gas differentially accumulated in the vertical between the different structural zones.

**Figure 2.** Comprehensive strata log diagram of the Yinggehai Basin with the source-reservoir-cap assemblage and main tectonic event.

**Figure 3.** Vertical distribution of gas in diapir zone and slope zone in Yinggehai Basin.

#### **3. Methods**

#### *3.1. Principle of Cap Rock Hydraulic Fracture Evaluation*

There are two kinds of hydraulic fracture modes of cap rock under the dominance of fluid, namely, the fracture of intact cap rock and the re-opening of pre-existing fractures. The constant increase of pore fluid pressure leads to the gradual decrease of the effective stress on the cap rocks, and eventually, pressure is relieved through hydraulic fractures [4].

The fracture mode on cap rocks is controlled by two factors, the tensile strength of cap rock (T) and its differential stress (S1 − S3) [1,42,43]. When the S1 − S3 of the stratum is less than 4 times the tensile strength, extensional fractures are formed; when the S1 − S3 is more than 6 times its tensile strength, it results in compressional shear fractures; the tensile strength and shear mixing fractures occur in relation to the above two conditions when the S1 − S3 is more than 4 times and less than 6 times the tensile strength (Table 1) [26]. For the intact cap rocks, in the case of extensional fractures, extensional shear fractures, and shear fractures, the fracture pressure (P) of cap rock corresponds to A1C1, A2C2, and A3C3, respectively (see points in Figure 4) [42]. However, for cap rocks with pre-existing fractures, hydraulic fractures are always re-opening along the weak surface. We used the incoherency of envelope to characterize fracture conditions, fracture pressure (P) for extensional fractures, extensional shear fractures, and shear fractures which respectively correspond to Figure 4 B1C1, B2C2, and B3C3. Analysis shows that the existence of early fractures largely increases the risk of hydraulic cracks in the cap rocks. Therefore, in order to determine the hydraulic fracture pressure, the fracture mode of the cap rock must be determined first.

**Table 1.** The fracture mode and criterion of rocks in different differential stress [26].


Where P is fracture pressure (MPa), S1 is maximum principal stress (MPa), S3 is minimum principal stress (MPa), S1 − S3 is differential stress (MPa), Sn is normal stress (MPa), T is tensile strength of rock (MPa), C is cohesion (MPa), τ is shear stress (MPa), μ is coefficient of friction.

#### *3.2. Parameters of Cap Rock Hydraulic Fracture Evaluation*

Based on the above description, the development of hydraulic fracture depends on the pore fluid pressure, the characteristics of the in-situ stress field, and the mechanical properties of the cap rock. Among them, pore fluid pressure and mechanical parameters of cap rocks can be obtained through formation tests and laboratory tests. Therefore, in order to be able to quantitatively evaluate the hydraulic fracture risk of cap rock, it is necessary to define the in-situ stress characteristics of the study area.

**Figure 4.** Hydraulic fracture pressure for the cap rock in different stress and mechanical properties. The Mohr circles, from large to small, indicate the critical conditions of shear, extension-shear and extension fracture of intact cap rocks and fractured cap rocks, and the corresponding maximum sustainable fluid pressures are A1C1, A2C2, A3C3 and B1C1, B2C2, B3C3(modified by Sibson(1996)).

3.2.1. Vertical Principal Stress

The vertical principal stress mainly comes from the rock gravity of the overburden strata. Therefore, it can be obtained by integrating density log data. Because the Yinggehai Basin is in the South China Sea, it is necessary to consider the effect of seawater on the vertical principal stress [44] (Equation (1)).

$$S\_{\mathcal{V}} = \rho\_{\mathcal{W}} g h\_{\mathcal{W}} + \int\_{h\_{\mathcal{W}}}^{h} \rho\_{\mathcal{C}}(h) g dh,\tag{1}$$

where *Sv* is the vertical principal stress (MPa), *ρ<sup>w</sup>* is the density of water(1.03 g/cm3 for Yinggehai Basin), *hw* is the water depth(average depth is 65 m in the study area), *ρ<sup>c</sup>* is the density of overburden layers (g/cm3), *g* is the gravitational acceleration constant (N/kg), *h* is the burial depth below sea level (m). However, there are always some problems when calculating vertical principal stress. For example, the density logging usually does not start from the sea bottom or surface. The missing density value can be converted from the check-shot velocity data to the density data through the Nafe-Drake transformation formula [45].

#### 3.2.2. Horizontal Principal Stress

Horizontal principal stress is usually the most difficult to interpret. At present, a relatively reliable method is extracted from hydraulic pressure tests [46]. However, these tests have not been widely used on drilling. In this study, we have chosen to use pumping pressure tests data to analyze horizontal principal stress which can be classified as FITs (Formation Integrity Tests), LOTs (Leak-Off Tests), and XLOTs (Extended Leak-Off Tests) [47]. They are curves of bottom hole pressure and pumped mud volume/time, and can be distinguished by the difference between the number of pumping cycles and the point at which pumping is ceased [47] (Figure 5a). LOTs are performed in the formation below the casing shoe, and its primary purpose is to determine the maximum mud weight

that could be allowed without drilling risk (formation loss) [24]. Whereas LOTs are not pressurized after the loss pressure is reached to avoid further formation damage, XLOTs are pressurized after the loss pressure is reached, even with multiple pressure cycles performed to obtain more information. Engineering practice shows that a classic single-cycle XLOT curve includes Leak-off Pressure (LOP), Formation Break-down Pressure (FBP), Fracture Propagation pressure (FPP), Instantaneous Shut in Pressure (ISIP), Fracture Closure Pressure (FCP) (Figure 5b–d). Of these points on the curve, LOP is the first point that deviates from the linear relationship between pressure and cumulative mud volume, which can be considered to be the point where the fluid begins to penetrate into the formation, and the rock begins to deform inelastically. Thus, the lower envelope of all LOP data can be used to represent horizontal minimum principal stress [24].

**Figure 5.** (**a**) Classic XLOT curve in the drilling [24]; (**b**–**d**) The different types XLOT curves in study area.

During the XLOT test, the failure pattern of the wellbore wall is very similar to the hydraulic fracturing in cap rock. The difference is that hydraulic fracturing detects the wellbore conditions through technology and then selects the target interval without fractures. The presence of fractures causes the formation break-down pressure to be close to the fracture reopening pressure, and the tensile strength of wellbore wall is approximately zero [48,49]. Thus, the Hubbert–Willis equation [48] can be simplified to

$$S\_{Hmax} = \Im S\_{hmin} - FRP - P\_{p\prime} \tag{2}$$

where *SHmax* and *Shmin* are the maximum and minimum horizontal principal stress, respectively (MPa); *FRP* is the fracture reopening pressure (MPa); and *Pp* is the pore pressure (MPa).

#### 3.2.3. Pore Fluid Pressure and Mechanical Parameters of Cap Rocks

The current pore fluid pressure is derived from formation Drill Stem Test (DST) and Modular Formation Dynamics Test (MDT) data and drilling fluid weight. The palaeo-pressure data was derived from CNOOC (China National Offshore Oil Corp., Beijing, China).

The mechanical parameters of cap rocks can be tested by direct tensile test and the Brazilian tensile test. However, these two methods have high requirements for sample preparation and low success rate. Considering the limited core samples, conventional triaxial compression test was chosen for this paper. Specifically, we used the GCTS Rapid Triaxial Rock Test System, and the instrument model is RTR-2000. Through tests, we measured the stress-strain curves of rock under different confining pressures and assumed that the tensile strength envelope in the tensile region conforms to the Griffith fracture criterion; the tensile strength of rock can be estimated by its relation to the cohesion, which is usually approximately twice the tensile strength.

#### **4. Results**
