Development Characteristics of Silurian Strike-Slip Faults and Fractures and Their Effects on Drilling Leakage in Shunbei Area of Tarim Basin

In recent years, the Ordovician fault-controlled fracture-cavity reservoirs developed in the basement strike-slip fault zone in the Shunbei area of the Tarim Basin has achieved major breakthroughs. However, during the drilling process of the strike-slip fault zone in the Shunbei area, the problem of mud leakage in the frequently interbedded Silurian sandstone and mudstone strata overlying the Ordovician target layer is very significant, and it has seriously affected the normal drilling and wellbore stability. In this study, taking the Silurian of the No. A Strike-slip Fault Zone in the Shunbei area as an example, the development characteristics of strike-slip faults and fractures, the strength of the in-situ stress field, and the influence of these factors on the drilling mud leakage were systematically studied using 3D seismic, logging, drilling, logging, well log, and engineering construction data. The results show that the mud leakage in strata S1t is significantly larger than that in S1k, and the leakage amount in sandy mudstone is the largest; the strong strike-slip extension developed the negative flower-shaped normal faults and the right-order swan-type faults and caused serious stratigraphic fragmentation. The amount of mud leakage increases with the increase of fault distance. Moreover, the closer to the fault, the higher the frequency and amount of mud leakage. When the distance between the wellbore and the fault exceeds 300 m, the possibility of mud leakage decreases significantly. The Silurian S1t is dominated by high-angle and vertical tension-shear fractures with good opening; while the S1k is dominated by low-angle structural and horizontal bedding fractures. The differences in fracture type cause the mud leakage in S1t to be significantly larger than that of S1k. In addition, the fracture development intervals identified by the R/S-FD method are in good agreement with the mud leakage intervals, which further indicates that the degree of fracture development is the key factor leading to the drilling mud leakage. The study also found that the degree of fracture development and the difference in horizontal principal stresses are the dominant factors leading to high Silurian mud leakage.


INTRODUCTION
Basement strike-slip fault zones in petroliferous basins are an important type of high-yield hydrocarbon enrichment zone (Jiao, 2017;Jiao, 2018;Santosh and Feng, 2020;Qi et al., 2021). Controlled by the action of basement strike-slip structures, the deformation of strike-slip structures is usually confined to a narrow strip along the main fault strike. Furthermore, most of them are linear, belt-shaped, geese-shaped, S-shaped or reverse-S-shaped on the plane, and have a flower-like structure on the cross-section. In three-dimensional space, the main faults often show the phenomenon of "ribbon effect" and "dolphin effect", and indirectly cause the strike-slip fault zone to have obvious segmentation differences. Then, "strike-slip extension", "strikeslip overhang" and "strike-slip translation" can be identified (Deng, et al., 2019a;Li et al., 2019;Fan et al., 2020;Ding et al., 2021;Li, 2022). Moreover, they are affected by multistage tectonic movements and multiple sets of detachment layers, so that the deformation styles of strike-slip faults in different tectonic layers have obvious stratification differences Xue et al., 2021).
Strike-slip fault zones often exhibit the characteristics of "planar segmentation, vertical stratification, multi-period superposition, and variable stress" (Deng et al., 2018;Deng et al., 2019b;He et al., 2020;Hower and Groppo., 2021;Zhan Zhao et al., 2021). They are also the internal mechanism of the "reservoir and hydrocarbon-controlling effect" of strike-slip faults. The multi-stage activities of strike-slip faults have an important impact on the formation of fracture systems and the improvement of petrophysical properties in carbonate rock, shale, and tight sandstone reservoirs (Han et al., 2016;Zhao et al., 2019;Mohammed et al., 2021). In addition, fault activity can also cause the wellbore instability of drilling near the strike-slip fault zone (such as mud loss, wellbore collapse, falling blocks and sticking, etc.) (Al-Ajmi and Zimmerman., 2006;Chen et al., 2014;Yin et al., 2015;Yan and Zhao, 2018;Yin et al., 2018;Zheng et al., 2020;Chen et al., 2021;Gao, 2021).
The Paleozoic and Mesozoic Cenozoic sedimentary strata in Shunbei area of Tarim Basin are well developed, and Permian volcanic rocks and upper Ordovician intrusive dolerite bodies are also developed. Among them, the carbonate rocks of the Middle and Lower Ordovician are the target layers for oil and gas exploration, while the overlying Silurian sand-mudstone interbedded strata, Permian volcanic rocks, and Upper Ordovician diabase strata have serious wellbore instability. Especially in the Silurian sand-mudstone interbedded strata, the leakage of drilling mud is the most serious, and the phenomenon of "nine leakages in ten wells" is remarkable. Therefore, with the continuous expansion of the oil and gas exploration scope of the strike-slip fault zone in the Shunbei area, studies have found that the Silurian sand-mudstone strata overlying the Ordovician target layer have low stress bearing capacity, developed faults, and developed a large number of complex open and closed fracture systems . Therefore, the random multi-point mud leakage in the whole well section is very serious, which has seriously affected the normal drilling and wellbore stability. In-situ stress is the key factor of leakage (Li et al., 2011).
In this study, taking the Silurian of the No. A Strike-slip Fault Zone in the Shunbei area as an example, the theory of reservoir geomechanics was used to analyze the development characteristics of strike-slip faults and fractures, as well as the distribution characteristics of in-situ stress fields, and their relationship with drilling mud leakage (Yin and Wu., 2020;Zhao et al., 2020;Li and Li., 2021;Zhao et al., 2021). This study can provide scientific guidance for revealing the wellbore instability mechanism of strike-slip fault zones in similar areas (Jin et al., 1999;Chen et al., 2013;Chen et al., 2015;Gao et al., 2016;Wang et al., 2020;Guo et al., 2021).

GEOLOGICAL SETTING
The Shunbei area is located on the Shuntuoguole Uplift in the central part of the Tarim Basin ( Figure 1). The structural characteristics of this area are high in the north and low in the south, and high in the east and low in the west. It is a relatively stable ancient tectonic unit in the Tarim Basin. Affected by the Caledonian, Hercynian, Indosinian, Yanshanian and Himalayan Movements, 18 strike-slip faults in the Shunbei area and its adjacent areas developed in a nearly SN-NE trend. In addition, fault-fracture-type oil and gas reservoirs developed in ultra-deep Ordovician carbonate rocks, and it has been confirmed that these strike-slip fault belts are rich in oil and gas. The Jurassic strata in the Shunbei area are missing; the Cretaceous, Permian, Triassic and Carboniferous strata are partially missing; and other strata are normally developed. From bottom to top, the study area develops Cambrian, Ordovician, Silurian, Devonian, Carboniferous, Permian, Triassic, Cretaceous, Paleogene, Neogene, and Quaternary ( Figure 2). Among them, the Silurian is the stratum with interbedded sand and mudstone, and the layers with serious leakage of drilling mud are the Lower Silurian Kepingtag Formation and Tata Ertag Formation.

Materials and Experiments
The data in this study include the Silurian drilling mud leakage data for 19 wells near the strike-slip fault zone in the study area. The specific data include the geological layer at the time of leakage, the section of the leakage well (m), and the lithology; the latter includes

Finite Element Method
In this study, the finite element method was used to simulate the tectonic stress field (Ding., et al., 2010;Ding et al., 2012;Jiu et al., 2013;Ding et al., 2016;Liu., et al., 2018). The structural geological model of the northern segment of the Strike-Slip Fault Zone A in the Shunbei area is shown in Figure 3, which is subdivided into a series of node and unit grids, including 40,187 nodes and 20,926 grid units ( Figure 4).    According to the experimental test results of acoustic emission in-situ stress, the pressure gradient of the overlying rock of the target layer is 0.0256 MPa/m, the maximum horizontal in-situ stress gradient is 0.0234 MPa/m, and the minimum horizontal insitu stress gradient is 0.0187 MPa/m. Finally, we determined that the direction of the maximum horizontal principal stress during the main rupture formation period of the Silurian sand-mudstone stratum was NE-SW, and the direction of the minimum horizontal principal stress was NW-SE ( Figure 5).
To facilitate the calculation process, the tensile rupture rate I t and the shear rupture rate I n are introduced: where σ T is the effective tensile stress, MPa; σ t is the tensile strength of the rock, MPa. When I t ≥ 1, the rock undergoes tension fracture. I n τ n |τ| (2) where τ n is the effective shear stress, MPa, and |τ| is the shear strength of the rock, MPa. When I n ≥ 1, the rock undergoes shear fracture. The fracture mode of rock is a comprehensive reflection of tensile stress and shear stress, and a comprehensive fracture coefficient is introduced.
In the formula, a and b are the proportions of tension fractures (including tension-shear fractures) and shear fractures obtained from core and electron microscope observations, respectively. In this study a:b = 6:4. Similarly, when I z ≥ 1, the rock reaches a fractured state, and the higher the comprehensive fracture rate value, the greater the fracture degree.

R/S-FD Method
R/S analysis, also known as rescaled range analysis or variable scale analysis. This method was originally proposed by British hydrologist Hurst on the basis of long-term research on the relationship between water volume and storage capacity of reservoirs (Hurst, 1951). Through the processing of conventional logging curves by R/S analysis method, the fracture development interval can be preliminarily determined by manually identifying abnormal positions on the R/S curve (Rangarajan and Sant., 2004;Xiao et al., 2019).
In the R/S analysis method, R is the range, which is the difference between the maximum value and the minimum value, indicating the fluctuation range and complexity of the time series; S is called the standard deviation, indicating the average trend of the time series. R/S is a dimensionless time series representing fluctuation strength. The calculation methods of R and S are shown in formula 4 and formula 5 respectively: Among them, Z is a logging curve for R/S analysis; n is the number of sampling points in the logging analysis interval (one  sample is taken every 0.125 m), and each n has a corresponding R(n)/S(n) value; u is the number of sampling points increased each time since the beginning; i and j represent variables; D is the fractal dimension, dimensionless, formula (6). Technical roadmap of strike-slip faults and fractures and their influence on drilling leakage is shown in Figure 6.

Mud Leakage Characteristics
Statistics on the Silurian mud loss data of 19 wells in the Shunbei area show that in general, the mud loss in S 1t is significantly larger than that in S 1k . The leakage of sand-bearing mudstone is the largest, followed by mud-bearing sandstone and mudstone layers (Figure 7). The mud leakage of each well in different lithologic sections varies greatly.

Log Identification Results of Fractures
R(n)/S(n) is the R/S value corresponding to the nth sampling point. The R/S value and n of each interval are plotted as a logarithmic scatter plot. If the relative change of the two is observed as a densely distributed fitting curve, it indicates that   the sequence has good fractal characteristics. The slope H of the R(n)/S(n) curve is called the Hurst exponent. The fractal dimension D of a certain interval can be calculated by formula (6), and its value reflects the heterogeneity of carbonate reservoirs. When some scholars use fractal theory to study the degree of fracture development, they believe that the larger the fractal dimension D, the more developed the fractures (Qie et al., 2021;Wang and Wang., 2021;Wen et al., 2004). The existence of fractures in the formation will lead to a great increase in the complexity of the formation, which corresponds to a higher fractal position value D and a smaller Hurst exponent. The R(n)/ S(n) curve shows a decreasing slope segment, which is specifically characterized by the downward shift of the R(n)/S(n) curve. From this, it can be concluded that morphologically, the down-concave segment of the R(n)/S(n) curve can be considered as a potential fracture development segment (Figure 8).
Based on the above methods and theories, this paper comprehensively conducts R/S analysis and calculation on five conventional logging curves (CAL, AC, GR, SP and RD) of the Silurian drilling in Shunbei area. If most of the R(n)/S(n) curves show depression in a certain section at the same time, it can indicate that this section is a fracturedeveloped section. Large errors are usually introduced when manually identifying the concave section of the R(n)/S(n) curve. In order to solve the shortcomings of this method, the second derivative of the function in advanced mathematics is introduced to improve the method. Since the value of R(n)/ S(n) is discrete data, this paper uses the finite difference method (FD) to approximate the derivative operation. The logging fracture identification results show that the fracture identification accuracy based on the R/S-FD analysis method can reach 78%.

Relationship Between Fault and Mud Leakage
The three-dimensional seismic data can well reflect the fault distribution characteristics of the Silurian system in both vertical and horizontal directions. The faults appear as flower-like structures on the seismic profiles, with obvious strike-slip fault characteristics. The Silurian tectonic layer is an asymmetric negative flower-like tectonic style, with a large number and well-developed strike-slip extensional branch normal faults. The vertical fault throws of main and branch faults are small, reflecting weak extension (Figure 9). In addition, the Silurian fault system is mainly arranged in a geese-type normal fault combination pattern on the plane, with a small extension length, and some of them exist in pairs ( Figure 3). Nearly north-south extensional negative flower-like branch normal faults are well developed in the study area, which are distributed in a right-order geese pattern, and a few local high points are developed. These features indicate that the Silurian strike-slip extension is strong in the Shunbei area, and the stratum is severely fractured. According to the statistical data of 20 wells, the mud leakage increases with the increase of the fault distance, and the frequency and the leakage volume are the highest when the distance is closer to the fault. Moreover, the leakage frequency is the highest when the wellbore is within 150 m of the fault, followed by 150-300 m. When the wellbore is more than 300 m away from the fault, the likelihood of leakage is lowest (Figure 10). The development of negative  flower-like branch normal faults is the macro-controlling factor of Silurian sand-mudstone formation leakage. The deep tectonic pattern in the study area is shown on the seismic section as a large-angle or near-upright strike-slip fault breaking into the basement, which is mainly developed at the T 7 4 interface (top surface of the middle-lower Ordovician unified Yijianfang Formation) and underlying strata. These faults are formed in one stage.

Relationship Between Fracture Occurrence and Mud Leakage
In Well Shunbei Y, medium and high-angle tension-shear fractures are mainly developed in the sandstone section of the Silurian Tata Ertag Formation (S 1t ), and low-angle fractures are prone to occur at the sand-mudstone interface. The residues of the invading mud during drilling can be seen on the fracture surfaces of the opened fractures. In addition, Well Shunbei Y mainly developed low-angle fractures in the sandstone section of the Silurian Kalpintag Formation (S 1k ), and the fractures were mainly generated along weak planes such as horizontal or oblique bedding.
Statistics show that fractures of different dip angles are developed in Silurian S 1t , mainly high-angle and vertical fractures. In S 1k , low-angle and horizontal fractures are the main factors (Figure 11), which is also the main factor causing the mud loss in S 1t to be significantly larger than that in S 1k .

Relationship Between Vertical Development Characteristics of Fractures and Mud Leakage
Based on the R/S-FD analysis method, this paper uses four logging curves (AC, CAL, SP, GR) to calculate the fractures. Most of the R(n)/S(n) curves show a fracture development segment in the concave curve segment (Figure 8, Figure 12). And the study found that the fracture development section is in good agreement with the mud leakage section, indicating that the Silurian fracture development is the key factor leading to the leakage.

Tectonic Stress Field and Fracture Distribution and Their Relationship With Mud Leakage
In this study, the finite element method was used to simulate the tectonic stress field and predict the fracture distribution of the Silurian in the Shunbei No. A Fault Zone in the Shunbei area. The maximum principal stress of Shunbei No. A Fault Zone is mainly compressive stress, and its value is concentrated between 115-130 MPa, and it is greatly affected by the fault zone. The stress inside the fault zone is relatively small, and the stress is relatively large near the fault zone ( Figure 13). The shear stress value of Shunbei No. A Fault Zone is generally positive, ranging from 3.6 to 21.6 MPa. It is reflected that it is dominated by the counterclockwise left-handed shear stress field. In addition, the shear stress values along the main fault zone and its vicinity are larger, and the simulation results are consistent with the deformation characteristics of the compressive-torsional strikeslip structure in the Shunbei area. In addition, the horizontal stress difference range is relatively small (45-60 MPa) within the fault zone, while it is relatively large near the fault zone ( Figure 14).
In this study, the finite element method was used to calculate the planar distribution of the fracture development coefficient I F of the Silurian strata in the northern segment of the No. A Fault Zone in the Shunbei area ( Figure 13). The fracture development coefficient I F of the Silurian strata in this fault zone is distributed between 0.61 and 2.01. When the I F of the formation is less than 1, it can be considered that the rock is not fractured. When the I F of the formation is greater than 1, the rocks are fractured to different degrees. The larger the value of I F , the greater the probability of rock fracture. According to the distribution range of I F value of Silurian in Shunbei area, it can be divided into three grades A, B and C ( Table 1). The larger the I F value and the higher the classification, the easier the rock is to be fractured and the more developed the fractures. In the main fault zone and its vicinity, the fracture development coefficient is between 1.0 and 1.35 ( Figure 15). For the four wells drilled on the No. A fault zone in the Shunbei area, we calculated the fracture development coefficient, horizontal stress difference, and their relationship with mud leakage (Figure 16). The Silurian as a whole shows that the higher the degree of fracture development, the greater the mud leakage, and the horizontal stress difference is negatively correlated with the mud leakage. For example, Well W3 has a large leakage and a small fracture development coefficient, but the stress difference is small, 46 MPa, which is 2-12 MPa smaller than other stress differences. It shows that the leakage of this well is mainly controlled by in-situ stress. The degree of fracture development and the difference in horizontal stress in the Silurian sand-mudstone formation are the main factors leading to high mud leakage.
Through the correlation analysis between the mud leakage and the horizontal stress difference, distance from the fault, and fault distance, it is concluded that the horizontal stress difference is the most direct cause of the mud leakage.

CONCLUSION
(1) In this study, based on 3D seismic, logging, drilling, logging, well log, and engineering construction data, development characteristics of strike-slip faults and fractures, the strength of the in-situ stress field, and the influence of these factors on the drilling mud leakage were systematically studied.
(2) The mud leakage in S 1t is significantly larger than that in S 1k , and the leakage amount in sandy mudstone is the largest; the strong strike-slip extension developed the negative flowershaped normal faults and the right-order swan-type faults and caused serious stratigraphic fragmentation. The amount of mud leakage increases with the increase of fault distance.
Moreover, the closer to the fault, the higher the frequency and amount of mud leakage. When the distance between the wellbore and the fault exceeds 300 m, the possibility of leakage decreases significantly. (3) The Silurian S 1t is dominated by high-angle and vertical tension-shear fractures with good opening; while the S 1k is dominated by low-angle structural and horizontal bedding fractures. The difference in fracture type causes the mud leakage in S 1t to be significantly larger than that of S 1k . (4) The fracture development intervals identified by the R/S-FD method are in good agreement with the mud leakage intervals, which further indicates that the degree of fracture development is the key factor leading to the drilling mud leakage. (5) The degree of fracture development and the difference in horizontal principal stresses are the dominant factors leading to high Silurian mud leakage. Mud leakage has significant positive and negative correlations with the fracture development coefficient and the horizontal stress difference, respectively.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/supplementary material further inquiries can be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
HL is responsible for the idea, seismic/logging data interpretation and writing of this paper.