A method and system for identifying natural fractures in tight oil reservoirs based on R / S analysis and finite difference method

Abstract

The present invention discloses a method and system for identifying natural fractures in tight oil reservoirs based on R / S analysis and finite difference method, the system comprising a logging curve fracture sensitivity analysis module, an R / S analysis module, a finite difference method analysis module, a fracture type discrimination module, and a line density measured data supervision iteration module. The present invention introduces a discrete data finite difference method, and achieves the purpose of accurately judging the concave section of the R / S curve by obtaining the secondary derivative of the R / S curve; achieves the purpose of distinguishing bedding fractures from high-angle fractures by judging whether the GR curve passes the R / S independent analysis; and approaches the identified natural fracture line density data to the measured data through an adaptive iterative cycle. The present invention can not only more accurately indicate the development locations of different types of natural fractures and output fracture line density data, but also greatly reduce the cost of natural fracture identification and prediction work in oil fields, and has a high practical value for the prediction of natural fractures in tight oil reservoirs.

Description

Technical Field

[0001] The present invention belongs to the technical field of unconventional oil and gas field exploration and development, and in particular relates to a method and system for identifying natural fractures in tight oil reservoirs based on R / S analysis and a finite difference method. Background Art

[0002] The identification and prediction of natural fractures in reservoirs has always been a problem that must be solved in the field of oil and gas geological exploration and engineering development. Accurate and efficient identification of the development location of reservoir fractures not only helps to improve the accuracy of favorable area prediction, but also allows the formulation of more detailed fracturing construction plans. Conventional logging is low-cost and has a huge number of wells, which can be used as basic data for identifying natural fractures in reservoirs. The reservoir fracture identification method based on R / S analysis is based on conventional logging data. By performing R / S analysis on multiple conventional logging curves, the concave jitter segment of the R / S curve is identified to determine the location of fracture development ( Figure 1 ).

[0003] The traditional natural fracture identification method for tight oil reservoirs based on R / S analysis and finite difference method has three important defects:

[0004] 1. The identification of the concave section of the R / S curve is inaccurate and imprecise, and the workload is large

[0005] The traditional method for identifying natural fractures in tight oil reservoirs based on R / S analysis and finite difference method requires the naked eye to judge the concave section of the R / S curve. The R / S curve is a curve in the double logarithmic space domain, and its data points are distributed loosely at the beginning and densely at the end. Therefore, in the process of using the naked eye to identify the concave section of the R / S curve, a large number of human factors are inevitably added. The entire identification process is not only labor-intensive, but also causes a large number of abnormal points to be missed, affecting the final fracture identification results;

[0006] 2. The identification results do not distinguish between different types of natural fractures

[0007] Due to the different characteristics of fracture development in different lithologies, especially in tight oil reservoirs, mudstone is mainly dominated by bedding fractures, while sandstone is mainly dominated by high-angle shear fractures. Therefore, in the process of identifying natural fractures using conventional logging data, the identification methods cannot be generalized and must be discussed in categories. However, in the traditional reservoir natural fracture identification technology based on "R / S" analysis, the results of fracture identification are expressed in the form of fracture line density, without considering the different types of fracture line densities.

[0008] 3. The determination of optimal parameters is too subjective

[0009] The traditional method of identifying natural fractures in tight oil reservoirs based on R / S analysis requires visually judging the concave section of the R / S curve. However, there is no basis for defining the degree of concave. It is also not certain whether the degree of concave is related to the degree of fracture development. Summary of the invention

[0010] In view of the above problems, the present invention proposes a method and system for identifying natural fractures in tight oil reservoirs based on R / S analysis and finite difference method. First, in view of the problems of inaccurate and imprecise identification of the concave section of the R / S curve and large workload, the present invention introduces the discrete data finite difference method, and achieves the purpose of accurately judging the concave section of the R / S curve by obtaining the secondary derivative of the R / S curve. Secondly, in view of the problem that the identification results do not distinguish different types of natural fractures, based on the observation conclusions of the field and core of the tight oil reservoir, the R / S independent analysis step of the natural gamma logging curve (GR) is added, and the purpose of distinguishing bedding fractures from high-angle fractures is achieved by judging whether the GR curve passes the R / S independent analysis. Finally, in view of the problem that the determination of the optimal parameters is too subjective, an adaptive iterative cycle step using measured data supervision is added. The natural fracture line density data obtained by actual measurement means such as imaging logging and core observation data is added to supervise the identification results, and the identified natural fracture line density data is approached to the measured data through an adaptive iterative cycle, and the optimal parameters that can be promoted and applied are finally obtained.

[0011] In order to achieve the above object, the present invention adopts the following technical solutions:

[0012] On one hand, the present invention proposes a natural fracture identification system for tight oil reservoirs based on R / S analysis and finite difference method, comprising a logging curve fracture sensitivity analysis module, an R / S analysis module, a finite difference method analysis module, a fracture type discrimination module, and a line density measured data supervision iteration module;

[0013] The logging curve fracture sensitivity analysis module is used to select the logging curve data of the fracture section and the non-fracture section, and to identify and analyze the logging curve that is more sensitive to the natural fracture information by drawing the intersection diagram;

[0014] The R / S analysis module is used to perform R / S analysis on the logging curves obtained by the logging curve fracture sensitivity analysis module to obtain the R / S value of each logging curve;

[0015] The finite difference analysis module is used to perform secondary derivative of the R / S value of each logging curve by using the discrete data derivative rule, remove all values ​​less than 0 in the secondary derivative of the R / S value of each logging curve, and calculate the fracture development index;

[0016] The fracture type identification module is used to obtain the natural fracture type by using the fracture development index to determine the natural fracture development segment, and to perform R / S analysis on the GR curve of the natural fracture development segment;

[0017] The line density measured data supervision iteration module is used to set a screening threshold. When the crack development index is greater than or equal to the screening threshold, the crack line density is calculated, and the crack identification error rate is calculated based on the obtained crack line density. The crack identification error rate is preset. When the crack identification error rate is less than the preset crack identification error rate, the size of the screening threshold is gradually increased to achieve a cyclic iteration of the identification result, and finally the crack line density is approximated and identified.

[0018] Furthermore, the crack type identification module is specifically used for:

[0019] The natural fracture development section is obtained through the fracture development index. The layer section with a fracture development index that is not zero is the natural fracture development section. The GR curve of the natural fracture development section is subjected to R / S analysis. When the GR curve passes the R / S analysis, the natural fracture development section belongs to the developed bedding fracture; when the GR curve does not pass the R / S analysis, the natural fracture development section belongs to the developed high-angle fracture.

[0020] Furthermore, the line density measured data supervision iteration module is specifically used for:

[0021] Obtain line density data of different types of natural fractures; set a screening threshold, when the fracture development index is greater than or equal to the screening threshold, calculate the actual fracture line density, and calculate the fracture identification error rate based on the line density data of different types of natural fractures and the obtained actual fracture line density; preset the fracture identification error rate, when the actual fracture identification error rate is less than the preset fracture identification error rate, increase the size of the screening threshold until the actual fracture identification error rate reaches the preset fracture identification error rate, so as to achieve a cyclic iteration of the identification results and finally approximate the identification of the fracture line density.

[0022] Another aspect of the present invention provides a method for identifying natural fractures in tight oil reservoirs based on R / S analysis and finite difference method, comprising:

[0023] Step 1: Select the logging curve data of the fracture section and the non-fracture section through the logging curve fracture sensitivity analysis module, and identify and analyze the logging curves that are more sensitive to natural fracture information by drawing cross-plots;

[0024] Step 2: Perform R / S analysis on the logging curve obtained by the logging curve fracture sensitivity analysis module through the R / S analysis module to obtain the R / S value of each logging curve;

[0025] Step 3: Using the finite difference analysis module, the R / S value of each logging curve is secondarily derived by the discrete data derivation rule, all values ​​less than 0 in the second derivative of the R / S value of each logging curve are eliminated, and the fracture development index is calculated;

[0026] Step 4: Using the fracture type identification module, the natural fracture development section is obtained through the fracture development index, and the GR curve of the natural fracture development section is subjected to R / S analysis to obtain the natural fracture type;

[0027] Step 5: Set the screening threshold through the line density measured data supervision iteration module. When the crack development index is greater than or equal to the screening threshold, calculate the crack line density, and calculate the crack identification error rate based on the obtained crack line density. Preset the crack identification error rate. When the crack identification error rate is less than the preset crack identification error rate, gradually increase the size of the screening threshold to achieve a cyclic iteration of the identification results, and finally approximate the identification of the crack line density.

[0028] Furthermore, the step 4 comprises:

[0029] The fracture type discrimination module is used to obtain the natural fracture development section through the fracture development index. The layer section with a fracture development index that is not zero is the natural fracture development section. The GR curve of the natural fracture development section is subjected to R / S analysis. When the GR curve passes the R / S analysis, the natural fracture development section belongs to the developed bedding fracture; when the GR curve does not pass the R / S analysis, the natural fracture development section belongs to the developed high-angle fracture.

[0030] Furthermore, the step 5 comprises:

[0031] The linear density data of different types of natural fractures are obtained through the linear density measured data supervision iteration module; a screening threshold is set, and when the fracture development index is greater than or equal to the screening threshold, the actual fracture linear density is calculated, and the fracture identification error rate is calculated based on the linear density data of different types of natural fractures and the obtained actual fracture linear density; the fracture identification error rate is preset, and when the actual fracture identification error rate is less than the preset fracture identification error rate, the size of the screening threshold is increased until the actual fracture identification error rate reaches the preset fracture identification error rate, so as to achieve a cyclic iteration of the identification results and finally approach the identification of the fracture linear density.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] The present invention introduces finite differences into the traditional reservoir natural fracture identification method based on R / S analysis, and at the same time adds an independent R / S analysis of the natural gamma logging curve (GR curve) to distinguish different types of natural fractures. Finally, the line density data of different types of natural fractures are used as measured supervision data, and the natural fracture identification result that best meets the fracture development conditions of the work area is obtained through iterative cycles.

[0034] The present invention can not only relatively accurately indicate the development locations of different types of natural fractures and output fracture line density data, but also greatly reduce the cost of natural fracture identification and prediction in oil fields, and has high practical value for the prediction of natural fractures in tight oil reservoirs. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 The schematic diagram of the fracture identification method based on conventional well logging curves of "R / S" analysis and finite difference method;

[0036] Figure 2 This is a schematic diagram of the architecture of a natural fracture identification system for tight oil reservoirs based on R / S analysis and finite difference method according to an embodiment of the present invention;

[0037] Figure 3 This is a data processing flow chart of a crack type identification module according to an embodiment of the present invention;

[0038] Figure 4 This is a data processing flow chart of the line density measured data supervision iteration module according to an embodiment of the present invention;

[0039] Figure 5 This is an example diagram of the intersection diagram of logging information of the non-fracture section and the fracture section according to an embodiment of the present invention;

[0040] Figure 6 This is a distribution diagram of the values ​​of f" (R(n) / S(n)) of each logging curve of the vertical-X well in an embodiment of the present invention;

[0041] Figure 7 This is a comparison chart of the vertical-X well fracture identification result and the logging identification result in an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments:

[0043] like Figure 1 As shown in the figure, a natural fracture identification system for tight oil reservoirs based on R / S analysis and finite difference method adopts modular design idea, inputs logging curve, and outputs fracture development location (i.e. natural fracture development section), fracture type and linear density data of different types of fractures ( Figure 2The system specifically includes a logging curve fracture sensitivity analysis module, an R / S analysis module, a finite difference method analysis module, a fracture type discrimination module, and a line density measured data supervision iteration module.

[0044] The following is an introduction to each module:

[0045] 1. Logging curve fracture sensitivity analysis module

[0046] Combined with the data of core and imaging logging in the work area, which can directly reflect the development of fractures, the natural fracture development position is calibrated on the well as the measured fracture development data. Then the logging curve data of the fracture (development) section and the non-fracture (development) section are selected, and the logging curves that are more sensitive to natural fracture information are identified and analyzed by drawing cross-plots, and these logging curves are used as the data basis for natural fracture identification.

[0047] 2. R / S analysis module

[0048] In the R / S analysis method, R is called the range, which is the difference between the maximum cumulative deviation and the minimum cumulative deviation, representing the complexity of the time series; S is called the standard deviation, which is the square root of the variation and represents the average trend of the time series. The ratio of the two, R / S, represents the relative volatility intensity of the dimensionless time series. The specific expressions of the full-layer segment difference and standard deviation of the process series are shown in Equation 1 and Equation 2 respectively.

[0049]

[0050]

[0051] Where: Z(i) represents the selected logging curve data;

[0052] n represents the total number of well logging sampling points in the point-by-point analysis layer section;

[0053] u represents the number of sample points increasing from 0 to n starting from the endpoint;

[0054] i, j represent the variables of the number of sample points;

[0055] R(n) represents the range of the whole layer of the process sequence;

[0056] S(n) represents the standard deviation of the whole layer section of the process sequence;

[0057] The R / S analysis module is used to calculate R(n) / S(n), which is the ratio of the range to the standard deviation corresponding to the nth sample point. When n changes from 3 (the first two points cannot be calculated due to the limitation of mathematical formula) to the total number of logging sampling points, each value of n has a corresponding R(n) / S(n) value.

[0058] 3. Finite difference analysis module

[0059] The R / S of multiple logging curves was obtained by "R / S" analysis. k After the well logging curve, enter the finite difference analysis module. This module uses the discrete data derivation law to analyze the R / S k By determining R / S k The second derivative of the value (f″(R / S) k ) to determine the positive or negative of R / S k The convexity and concavity of a curve.

[0060]

[0061]

[0062] Where: h is the calculation step length;

[0063] Eliminate f″(R / S) k After all values ​​less than 0, the crack development index F is calculated by formula 5. In the formula, C is the amplification parameter, which is only used to amplify the value for drawing and is selected according to the actual situation.

[0064] F=f″(R / S)1×f″(R / S)2×f″(R / S)3×...×f″(R / S) k ×C (5)

[0065] 4. Crack type identification module

[0066] By observing the crack development characteristics of tight oil reservoirs in the field and in the core, it is found that natural cracks in shale are mainly bedding cracks, and natural cracks in sandstone are mainly high-angle cracks. Based on this understanding, the present invention achieves the purpose of distinguishing crack types by performing "R / S" analysis on the GR curve of the crack development section.

[0067] The natural fracture development section is obtained from the previous module. Specifically, the F value is the fracture development index. The layer section with an F value of non-zero is a natural fracture development section. Perform "R / S" analysis on the GR curve of the fracture development section. If the GR curve does not pass the "R / S" analysis (i.e., the fracture development index is zero), it means that the lithology of the formation has changed little, and high-angle fractures have developed; if it passes the "R / S" analysis (i.e., the fracture development index is not zero), it means that the lithology has changed greatly, and bedding fractures have developed ( Figure 3 ).

[0068] 5. Line density measured data supervision iteration module

[0069] The first four modules can be used to obtain the natural fracture identification line density (CFD). Specifically, the natural fracture development segment is obtained through the F value in the fracture type discrimination module. In a group of formations, the ratio of the number of natural fracture development segments to the thickness of the formation is the natural fracture identification line density. The line density data (MFD) of different types of natural fractures can be obtained through lithology, imaging logging and other data that can intuitively reflect the development characteristics of natural fractures. The fracture identification error rate A is obtained by formula 6: C .

[0070] A C =((|MFD-CFD| / CFD))*100% (6)

[0071] The linear density data (MFD) of different types of natural fractures are used as supervision data, and a fracture identification error rate A is preset. P , by gradually increasing the screening threshold T i The size of the crack line density CFD ( Figure 4 ).

[0072] As an implementable method, in order to verify the effect of the present invention, the following specific examples are performed:

[0073] The research object is the tight oil reservoir of the Changchang 7th section of the Huachi block of the Changqing Oilfield in the Ordos Basin. The basic data are a vertical well (Li-X well) and a horizontal well (Horizontal-X well), both of which contain imaging logging data. First, the logging curve type used for natural fracture identification is selected through the logging curve fracture sensitivity analysis module.

[0074] 1. Logging curve fracture sensitivity analysis module

[0075] The conventional logging curves such as natural gamma (GR) curve, acoustic time difference (AC) curve, density (DEN) curve, wellbore (CAL) curve, natural potential (SP) curve, shallow lateral resistivity (LL8) curve were selected for interactive analysis ( Figure 5 ).

[0076] After analyzing the four cross-plots, it was found that the AC curve and the DEN curve had a strong ability to distinguish cracks, followed by the CAL curve and the LL8 curve, and the SP curve and the GR curve had the least ability to distinguish cracks. Therefore, this study selected the AC curve, the DEN curve, the CAL curve, and the LL8 curve as the data basis for crack identification.

[0077] 2. R / S analysis module Finite difference analysis module

[0078] Well logging curves f″(R / S) of Li-X well kValue distribution diagram (AC curve, LL8 curve, CAL curve, GR curve and DEN curve) Figure 6 For the convenience of display, logarithmic coordinates are used for drawing. Figure 6 The F value is the single well fracture development index, which is calculated by the following formula:

[0079] F=K AC *K LL8 *K DEN *....*K CAL *C

[0080] Among them, C is the magnification parameter, which is only used to magnify the value for drawing. It is selected according to the actual situation. This time it is set to 10 10 . K AC , K LL8 , K DEN , ..., K CAL They are the second derivatives of the R / S values ​​corresponding to the AC, LL8, DEN, ..., CAL logging curves.

[0081] 3. Crack type identification module and line density measured data supervision iteration module

[0082] After counting the imaging logging data of the horizontal-X well, it was found that the natural fractures in the Chang 7 section of the study area showed two types of inclination angles. One is the bedding fracture with an inclination angle less than 20°, and the other is the shear fracture with an inclination angle greater than 50°. Among them, the number of bedding fractures accounts for more than 60% of all fracture types. Therefore, there are two types of fractures in the study area, namely high-angle fractures and bedding fractures.

[0083] Table 1 Statistics of cracks in the 7th section of the work area

[0084]

[0085] The vertical well trajectory is nearly perpendicular to the bedding fractures, which can be used as the measured bedding fracture line density. The horizontal well trajectory is nearly perpendicular to the high-angle fractures, which can be used as the measured high-angle fracture line density. The fracture line density obtained by interpreting the two imaging logging data is used as the measured data in the fracture identification work to supervise the cyclic iteration in the identification work.

[0086] Set the preset recognition error rate A P The initial screening threshold T is 20%. i is 0.0001, and the screening step length K is 0.0001. Finally, the identification is completed after 10621 iterations, and the actual crack identification error rate A is obtained. C The screening threshold is 19.62%. iThe final density of bedding sutures is 0.165 / m, and the final density of high-angle sutures is 0.059 / m. The different types of fracture development locations finally identified are also highly consistent with the actual fracture development locations ( Figure 7 ).

[0087] On the basis of the above embodiments, the present invention further proposes a method for identifying natural fractures in tight oil reservoirs based on R / S analysis and finite difference method, comprising:

[0088] Step 1: Select the logging curve data of the fracture section and the non-fracture section through the logging curve fracture sensitivity analysis module, and identify and analyze the logging curves that are more sensitive to natural fracture information by drawing cross-plots;

[0089] Step 2: Perform R / S analysis on the logging curve obtained by the logging curve fracture sensitivity analysis module through the R / S analysis module to obtain the R / S value of each logging curve;

[0090] Step 3: Using the finite difference analysis module, the R / S value of each logging curve is secondarily derived by the discrete data derivation rule, all values ​​less than 0 in the second derivative of the R / S value of each logging curve are eliminated, and the fracture development index is calculated;

[0091] Step 4: Using the fracture type identification module, the natural fracture development section is obtained through the fracture development index, and the GR curve of the natural fracture development section is subjected to R / S analysis to obtain the natural fracture type;

[0092] Step 5: Set the screening threshold through the line density measured data supervision iteration module. When the crack development index is greater than or equal to the screening threshold, calculate the crack line density, and calculate the crack identification error rate based on the obtained crack line density. Preset the crack identification error rate. When the crack identification error rate is less than the preset crack identification error rate, gradually increase the size of the screening threshold to achieve a cyclic iteration of the identification results, and finally approach the line density of the crack to be identified.

[0093] Furthermore, the step 4 comprises:

[0094] The fracture type discrimination module is used to obtain the natural fracture development section through the fracture development index. The layer section with a fracture development index that is not zero is the natural fracture development section. The GR curve of the natural fracture development section is subjected to R / S analysis. When the GR curve passes the R / S analysis, the natural fracture development section belongs to the developed bedding fracture; when the GR curve does not pass the R / S analysis, the natural fracture development section belongs to the developed high-angle fracture.

[0095] Furthermore, the step 5 comprises:

[0096] The linear density data of different types of natural fractures are obtained through the linear density measured data supervision iteration module; a screening threshold is set, and when the fracture development index is greater than or equal to the screening threshold, the actual fracture linear density is calculated, and the fracture identification error rate is calculated based on the linear density data of different types of natural fractures and the obtained actual fracture linear density; the fracture identification error rate is preset, and when the actual fracture identification error rate is less than the preset fracture identification error rate, the size of the screening threshold is increased until the actual fracture identification error rate reaches the preset fracture identification error rate, so as to achieve a cyclic iteration of the identification results and finally approach the linear density of the fracture to be identified.

[0097] In summary, the present invention introduces finite differences into the traditional reservoir natural fracture identification method based on R / S analysis, and at the same time adds an independent R / S analysis of the natural gamma logging curve (GR curve) to distinguish different types of natural fractures. Finally, the line density data of different types of natural fractures are used as measured supervision data, and the natural fracture identification result that best meets the fracture development conditions of the work area is obtained through iterative cycles.

[0098] The present invention can not only relatively accurately indicate the development locations of different types of natural fractures and output fracture line density data, but also greatly reduce the cost of natural fracture identification and prediction in oil fields, and has high practical value for the prediction of natural fractures in tight oil reservoirs.

[0099] The above is only a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A natural fracture identification system for tight oil reservoirs based on R / S analysis and finite difference method, characterized in that: It includes logging curve fracture sensitivity analysis module, R / S analysis module, finite difference method analysis module, fracture type identification module, and line density measured data supervision iteration module; The logging curve fracture sensitivity analysis module is used to select the logging curve data of the fracture section and the non-fracture section, and to identify and analyze the logging curve that is more sensitive to the natural fracture information by drawing the intersection diagram; The R / S analysis module is used to perform R / S analysis on the logging curves obtained by the logging curve fracture sensitivity analysis module to obtain the R / S value of each logging curve; The finite difference analysis module is used to perform secondary derivative of the R / S value of each logging curve by using the discrete data derivation rule, remove all values ​​less than 0 in the secondary derivative of the R / S value of each logging curve, and calculate the fracture development index; The fracture type discrimination module is used to obtain the natural fracture development section through the fracture development index, perform R / S analysis on the GR curve of the natural fracture development section, and obtain the natural fracture type; the fracture type discrimination module is specifically used to: obtain the natural fracture development section through the fracture development index, the layer section with a fracture development index not equal to zero is the natural fracture development section, perform R / S analysis on the GR curve of the natural fracture development section, when the GR curve passes the R / S analysis, the natural fracture development section belongs to the development of bedding fractures; when the GR curve does not pass the R / S analysis, the natural fracture development section belongs to the development of high-angle fractures; The line density measured data supervision iteration module is used to set a screening threshold. When the crack development index is greater than or equal to the screening threshold, the crack line density is calculated, and the crack identification error rate is calculated based on the obtained crack line density. The crack identification error rate is preset. When the crack identification error rate is less than the preset crack identification error rate, the size of the screening threshold is gradually increased to achieve a cyclic iteration of the identification result, and finally the crack line density is approximated and identified.

2. A natural fracture identification system for tight oil reservoirs based on R / S analysis and finite difference method according to claim 1, characterized in that: The line density measured data supervision iteration module is specifically used for: Obtain line density data of different types of natural fractures; set a screening threshold, and when the fracture development index is greater than or equal to the screening threshold, calculate the actual fracture line density, and calculate the fracture identification error rate based on the line density data of different types of natural fractures and the obtained actual fracture line density; The crack recognition error rate is preset. When the actual crack recognition error rate is less than the preset crack recognition error rate, the size of the screening threshold is increased until the actual crack recognition error rate reaches the preset crack recognition error rate, so as to achieve a cyclic iteration of the recognition results and finally approximate the recognition crack line density.

3. A method for identifying natural fractures in tight oil reservoirs based on R / S analysis and finite difference method, characterized in that: include: Step 1: Select the logging curve data of the fracture section and the non-fracture section through the logging curve fracture sensitivity analysis module, and identify and analyze the logging curves that are more sensitive to natural fracture information by drawing cross-plots; Step 2: Perform R / S analysis on the logging curve obtained by the logging curve fracture sensitivity analysis module through the R / S analysis module to obtain the R / S value of each logging curve; Step 3: Using the finite difference analysis module, the R / S value of each logging curve is secondarily derived by the discrete data derivation rule, all values ​​less than 0 in the second derivative of the R / S value of each logging curve are eliminated, and the fracture development index is calculated; Step 4: using a fracture type discrimination module, obtaining a natural fracture development section through a fracture development index, performing R / S analysis on the GR curve of the natural fracture development section, and obtaining the natural fracture type; said step 4 comprises: using a fracture type discrimination module, obtaining a natural fracture development section through a fracture development index, a layer section whose fracture development index is not zero is a natural fracture development section, performing R / S analysis on the GR curve of the natural fracture development section, when the GR curve passes the R / S analysis, the natural fracture development section belongs to a developed bedding fracture; when the GR curve does not pass the R / S analysis, the natural fracture development section belongs to a developed high-angle fracture; Step 5: Set the screening threshold through the line density measured data supervision iteration module. When the crack development index is greater than or equal to the screening threshold, calculate the crack line density, and calculate the crack identification error rate based on the obtained crack line density. Preset the crack identification error rate. When the crack identification error rate is less than the preset crack identification error rate, gradually increase the size of the screening threshold to achieve a cyclic iteration of the identification results, and finally approximate the identification of the crack line density.

4. The method for identifying natural fractures in tight oil reservoirs based on R / S analysis and finite difference method according to claim 3, characterized in that: The step 5 comprises: The linear density data of different types of natural fractures are obtained through the linear density measured data supervision iteration module; a screening threshold is set, and when the fracture development index is greater than or equal to the screening threshold, the actual fracture linear density is calculated, and the fracture identification error rate is calculated based on the linear density data of different types of natural fractures and the obtained actual fracture linear density; the fracture identification error rate is preset, and when the actual fracture identification error rate is less than the preset fracture identification error rate, the size of the screening threshold is increased until the actual fracture identification error rate reaches the preset fracture identification error rate, so as to achieve a cyclic iteration of the identification results and finally approach the identification of the fracture linear density.

Images