Oil-based mud electric imaging logging crack opening degree calculation method

Abstract

An oil-based mud electric imaging logging fracture opening degree calculation method comprises the steps of firstly determining response characteristics of oil-based mud electric imaging logging to various fractured stratums, then constructing a stratum multi-parameter calculation model, further constructing a fractured stratum multi-parameter image, then establishing a fracture recognition and image extraction processing method, and finally calculating the fracture opening degree of the oil-based mud electric imaging logging fracture opening degree. Finally, a fracture opening degree calculation method and model are established, the fracture opening degree can be calculated and evaluated based on electric imaging logging in the oil-based mud environment, stratum model evaluation and fracture stratum model evaluation are combined together, the respective functions of the stratum model evaluation and the fracture stratum model evaluation are highlighted, a mean value stratum model solves various necessary unknown parameters, and the evaluation accuracy is improved. A foundation is laid for crack opening degree calculation; in a homogeneous formation model and a fractured formation model, all possible multi-parameter combination measurement and calculation can be carried out, the application range is wide, and fracture parameter evaluation of various reservoirs is facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of geophysical well logging, and in particular to a method for calculating fracture opening in oil-based mud electrical imaging well logging. Background Art

[0002] The petroleum exploration and development industry plays a vital role in the development of the national economy and energy. Oil and gas reservoir evaluation provides advantageous technical support for oil exploration and development, with well logging technology and interpretation and evaluation playing a crucial role. Well logging, also known as geophysical logging, falls within the field of applied geophysics. It utilizes electromagnetic, acoustic, nuclear magnetic, radioactive, and electrochemical instruments to measure formation parameters such as resistivity, dielectric constant, and density, as well as their distribution, within underground boreholes. It is widely used in oil and gas exploration and development.

[0003] Electrical imaging logging is an electrical logging method, also known as microresistivity scanning imaging logging. It uses a microresistivity scanning array composed of densely distributed array button electrodes on the logging instrument plate. Combined with supporting shielding electrodes, return electrodes, adjustment and acquisition circuits and other components, it can simultaneously measure dozens to hundreds of resistivity curves. After data processing, a high-definition resistivity distribution image of the well wall formation is obtained, which can be used for formation fracture and hole identification, thin layer analysis, geological structure interpretation, etc.

[0004] During field logging operations, the wellbore is filled with mud, which serves to maintain wellbore pressure and lubricate the wellbore. The most commonly used mud type is water-based mud, which consists of a continuous water phase and other components in a dispersed phase. It has low resistivity. Initially, microresistivity scanning imaging logging tools were designed for low-resistivity water-based mud. However, in many cases, water-based mud fails to meet the requirements of field operations, such as in highly deviated and horizontal wells, shale formations, and deep-sea reservoirs. Oil-based mud, with its excellent lubricity, high-temperature and high-pressure resistance, wellbore stability, and improved operational efficiency, has gained widespread application in these specialized environments. However, oil-based mud, with its continuous oil phase, has a high resistivity, often hundreds or even tens of thousands of times that of water-based mud. This limits the application of existing microresistivity scanning imaging logging tools for water-based mud, and consequently, the ability to perform formation interpretation and evaluation work such as fracture and vug identification and calculation using electrical imaging logging in oil-based mud environments.

[0005] In recent years, electrical imaging logging instruments and formation parameter calculation methods have emerged in oil-based mud. Patent applications with application numbers 201910294886.8 and 201910124532.9, as well as a paper titled "Four-Parameter Calculation Method for Electrical Imaging Logging of Oil-Based Mud in Low-Resistivity Formations Based on Depressed Electrode Pairs" (Acta Petrolei Sinica, Vol. 41, No. 8, 2020), focus on core parameters represented by formation resistivity and provide equivalent models for their respective measurement processes and calculation methods for parameters such as formation resistivity. Patent application number 202011019657.4 provides a method for identifying and determining wellbore wall cracks and holes using electrical imaging logging of oil-based mud based on depressed electrode pairs. The paper titled "Numerical Simulation of Electrical Imaging and Instrument Response to Fractures in Oil-Based Drilling Fluid Environments" (Journal of China University of Petroleum (Natural Science Edition), Vol. 38, No. 5, 2014) uses numerical simulation methods to study the response characteristics of a long-loop oil-based mud electrical imaging logging instrument to a single fracture. The paper titled "Analysis and Quantitative Inversion of Electrical Imaging Logging Responses in Oil-Based Drilling Fluid Environments" (Journal of China University of Petroleum (Natural Science Edition), Vol. 42, No. 3, 2018) analyzes the response characteristics of oil-based mud microresistivity scanning imaging logging and uses an inversion calculation model to study formation resistivity and the mud cake thickness (interval) between the electrode plate and the wellbore wall. The above summary of existing technologies reveals that there is currently no research on fracture aperture calculation methods based on electrical imaging logging in oil-based mud environments. Summary of the Invention

[0006] In order to obtain a fracture aperture evaluation effect similar to that in a water-based mud environment in an oil-based mud environment, the present invention aims to provide a fracture aperture calculation method for oil-based mud electrical imaging logging. First, the response characteristics of oil-based mud electrical imaging logging to various fractured formations are determined. Then, a multi-parameter formation calculation model is constructed, and then a multi-parameter image of the fractured formation is constructed. Subsequently, a fracture identification and extraction image processing method is established. Finally, a fracture aperture calculation method and model are established. This method can calculate and evaluate fracture aperture based on electrical imaging logging in an oil-based mud environment, and carry out high-quality oil and gas formation interpretation and evaluation.

[0007] In order to achieve the above object, the technical solution of the present invention is:

[0008] A method for calculating fracture opening by oil-based mud electrical imaging logging comprises the following steps:

[0009] Step A, establishing a forward model of basic response of oil-based mud electrical imaging logging;

[0010] Based on the existing logging system, the logging system includes a logging instrument 4 suspended in a wellbore 1, and the logging instrument 4 is a micro-resistivity scanning imaging logging instrument. The imaging plate 10 at the end of the logging instrument 4 contacts the well wall, and the well wall formation is measured at intervals of a sampling interval. A basic response forward model of oil-based mud electrical imaging logging of a homogeneous formation without cracks is established using a numerical simulation method. The numerical simulation method includes a three-dimensional finite element method and a finite difference method. In the basic response forward model, the geometric and material properties of the formation 2, the logging instrument 4, and the imaging plate 10 are set. Megahertz electric field signal excitations of different frequencies are added, and the measurement signals of the imaging plate 10 are collected through grid division and solution calculation. Combined with the instrument constants, the impedance ratio, impedance amplitude, and impedance phase information are obtained. In the process of measuring the well wall formation, the parameter of the formation 2 is the formation resistivity R t and the formation dielectric constant ε t ; The parameter of mud / mud cake 3 attached to the well wall is the mud resistivity R m , mud dielectric constant ε m , plate gap (mud / mud cake thickness) d m The measurement response expression of the imaging plate 10 is:

[0011] Y=F(fr,G,R m ,ε m ,d m ,R t ,ε t ) (1)

[0012] Where Y is the measurement response vector, including information such as the measured impedance ratio, impedance amplitude, and impedance phase at different frequencies, and f r To measure frequency, G is the comprehensive response value of the structure and material properties of the logging tool 4 and the imaging pad 10. When the structure and material properties are determined, G is a constant value and F is the response function.

[0013] Step B, establishing a multi-parameter forward modeling database for oil-based mud electrical imaging logging;

[0014] Using the basic response forward model of the imaging plate 10, a forward database under the influence of multiple parameters is established in batches, and the multiple parameters include: measurement frequency (f r1 、f r2 ,…,f rn ), oil-based mud resistivity R m , oil-based mud dielectric constant ε m , plate gap (mud / mud cake thickness) d m , formation resistivity R t and the formation dielectric constant ε fThe forward modeling database also includes the measurement responses under the influence of multiple parameters, namely, the measured impedance ratio, impedance amplitude, impedance phase, etc. and the corresponding multi-parameter data. The measurement responses under the influence of multiple parameters correspond one to one with the multi-parameter combination.

[0015] Step C, establishing a multi-parameter cascade inversion model for oil-based mud electrical imaging logging;

[0016] The cascade means that the parameter inversion process has a sequence, and the established parameter inversion sequence is: mud resistivity, mud dielectric constant, formation resistivity, plate gap (mud / mud cake thickness), and formation dielectric constant. The specific process is: first, the data corresponding to the low-resistance formation is screened out from the forward modeling database, and the low-resistance formation is determined according to a pre-set threshold; then, a mud parameter calculation model is established in the low-resistance formation data using a machine learning algorithm or a Gauss-Newton algorithm, and the mud resistivity and dielectric constant can be determined based on the model; then, all the data corresponding to the determined mud parameters are sequentially screened out from the forward modeling database; finally, in each group of data corresponding to the determined mud parameters, a three-parameter inversion calculation model is established using a machine learning algorithm or a Gauss-Newton algorithm, and the measured impedance amplitude and measured impedance phase at each frequency are used as input (known quantities), and the three parameters (output quantities) are calculated in sequence, namely, formation resistivity, plate gap (mud / mud cake thickness), and formation dielectric constant (modifications have been made, no problem, they are corresponding).

[0017] Step D, establishing a fractured formation model for oil-based mud electrical imaging logging;

[0018] A fracture 11 is inserted into the formation 2. The fracture 11 has three morphological parameters, namely, the fracture inclination θ (the angle between the fracture plane and the horizontal direction), the fracture length l (the extension length along the fracture plane), and the fracture opening d (the fracture width perpendicular to the fracture plane). Based on the three morphological parameters, an oil-based mud electrical imaging logging fracture-type formation model is constructed; the imaging plate 10 moves upward or downward close to the well wall and measures the formation 2. When passing near the fracture 11, within the detection range (detection depth and resolution) of the imaging plate 10, the measurement signal of the imaging plate 10 will contain the contribution of the fracture 11.

[0019] Step E, classifying the crack types;

[0020] According to the inclination state and filling properties of the cracks, the cracks are divided into six types, namely, low-angle open cracks, with an inclination angle θ≤15°; medium-angle open cracks, with an inclination angle of 15<θ≤60°; high-angle open cracks, with an inclination angle θ>60°; low-angle closed cracks; medium-angle closed cracks; and high-angle closed cracks.

[0021] Step F, establishing a numerical simulation model for fractured formation response in oil-based mud electrical imaging logging;

[0022] In the numerical simulation model, the geometric and material properties of the formation 2, fracture 11, logging instrument 4, and imaging plate 10 are set, and megahertz electric field signals of different frequencies are added for excitation. Through meshing and solution calculation, the measurement signal of the imaging plate 10 is collected and combined with the instrument constants to obtain the impedance ratio, impedance amplitude, and impedance phase information. In this process, the key parameter of the formation 2 is the formation resistivity R t and the formation dielectric constant ε t , as the background parameter information of the crack 11; the key parameter of the crack 11 is the resistivity R of the crack filling material f , the dielectric constant ε of the filling material f , crack opening d f , fracture length l, fracture inclination θ; In addition, the key parameters of the wellbore wall attached mud / mud cake 3 are the mud resistivity R m , mud dielectric constant ε m , plate gap (mud / mud cake thickness) d m At this time, the measurement response expression of the imaging plate 10 is

[0023] Z=F(fr,G,R m ,ε m ,d m ,R t ,ε t ,R f ,ε f ,d f ,l,θ) (2)

[0024] Among them, Z is the measurement response vector of fractured formation, which mainly includes information such as measured impedance rate, impedance amplitude and impedance phase at different frequencies, and f r For measuring frequency, G is the comprehensive response value of the structure and material properties of the instrument 4 and the plate 10. When the structure and material properties are determined, G is a constant value. F is the crack response function.

[0025] Step G, electrical imaging logging response and sensitivity analysis of oil-based mud in fractured formations;

[0026] Based on the response characteristics, the crack opening response sensitive area and insensitive area are divided. The sensitive area is where the measurement signal has obvious changes in response to the change of crack opening. The parameter range corresponding to each area is determined to guide the subsequent crack opening calculation process and effect analysis.

[0027] Step H, establishing a physical model of oil-based mud electrical imaging logging in fractured formations;

[0028] Under laboratory or field logging conditions, the rock is first drilled to resemble a wellbore. Each core block is then evenly divided into two sections. Grooves of varying inclinations, apertures, and extensions are then carved into the inner wall of the drilled hole, representing fractures. Some fractures are filled with quartz and feldspar minerals, representing high-resistance closed fractures, while others are filled with clay and mud, representing low-resistance closed fractures. The two sections of rock are then assembled into a single block, with the gaps between the rocks cemented. Multiple fractured rocks are then constructed using the same method, and all the rocks are assembled together, penetrating the wellbore, to create a fractured formation model. Finally, oil-based mud is injected into the wellbore of the formation model and held stationary for a period of time, allowing the mud to penetrate the open fractures and complete the construction of the fractured formation. Throughout this process, the resistivity and dielectric constant of the rock blocks are measured, and the inclination, aperture, extension, depth, and filling type of the drilled fractures, as well as the resistivity and dielectric constant of the oil-based mud, are recorded and preserved.

[0029] Step I, fracture type formation physical model oil-based mud electrical imaging logging response and correction;

[0030] Place the imaging plate 10 into the well of the fractured formation physical model constructed in step H, carry out measurements at multiple frequencies, record and save information such as the plate position, measured impedance amplitude, and impedance amplitude. Combined with the formation model information recorded in step H, construct the measurement response expression of the fractured formation physical model:

[0031] Z * =F(fr,G,R m ,ε m ,d m ,R t ,ε t ,R f ,ε f ,d f ,l,θ) (3)

[0032] In the above formula, F, (fr, G, R m ,ε m ,d m ,R t ,ε t ,R f ,ε f ,d f ,l,θ) has the same meaning as expression (2). * It represents the response data in the physical model of fractured formation. When the parameters of mud, formation and fracture are the same, the response value Z of numerical simulation is different from the response value Z of physical model. * There are gain and offset phenomena between them, so we further establish a correction function from the physical model response value to the numerical simulation response value, that is,

[0033] Z * =P(Z) (4)

[0034] In the above formula, the function P represents the response value Z of the numerical simulation and the response value Z of the physical model. * There are gain, offset, or more complex correction operations between them.

[0035] Step J, a database of multi-parameter responses of oil-based mud electrical imaging logging in formations containing open fractures;

[0036] Based on the numerical model of oil-based mud electrical imaging logging in fractured formations in step F, the physical model in step H, and the response analysis in step I, the aperture range of the opening fractures is first determined in batches, and the aperture numerical points are determined using a linear or logarithmic partitioning method; the inclination angle range of the opening fractures is determined, and the inclination angle numerical points are determined using a partitioning method every 5 degrees; the length of the opening fracture is determined, and the fracture length numerical points are determined using a linear partitioning method; then, the formation resistivity range is determined in batches, and the formation resistivity numerical points are determined using a logarithmic partitioning method; the formation dielectric constant range is determined, and the formation dielectric constant range is determined using a linear or logarithmic partitioning method; then, the plate gap (mud / mud cake thickness) range is determined, and the plate gap (mud / mud cake thickness) is determined using a linear partitioning method. Finally, under the action of voltage signals of more than 2 megahertz, batch calculations are performed to collect the corresponding measured impedance ratios, impedance amplitudes, and impedance phase angles for each parameter combination, thereby establishing a multi-parameter response database for oil-based mud electrical imaging logging in formations containing opening fractures.

[0037] Step K, establishing a fracture opening calculation model for oil-based mud electrical imaging logging in formations containing open fractures;

[0038] First, the mud resistivity, mud dielectric constant, resistivity and dielectric constant of the background formation where the fracture is located, the plate gap (mud / mud cake thickness) opposite the fracture opening surface, and the parameters of the fracture opening dip and extension length are determined in sequence. Then, the response data corresponding to the determined parameters are extracted from the multi-parameter response database of oil-based mud electrical imaging logging in formations containing fractures. Finally, the measured impedance amplitude and phase at multiple frequencies are used as input (known quantities) and the fracture aperture is used as the output (to be solved). A fracture aperture calculation model for oil-based mud electrical imaging logging in formations containing fractures is established using a machine learning algorithm or a Gauss-Newton algorithm. The above steps are repeated to achieve a one-to-one mapping of the measured response data corresponding to all combinations of mud parameters, formation parameters, plate gap (mud / mud cake thickness) parameters, dip parameters, and extension length to the fracture aperture.

[0039] Step L, obtaining a three-parameter image of a fractured formation;

[0040] The oil-based mud electrical imaging logging cascade inversion model is used to process the oil-based mud electrical imaging logging data of fractured formations in the order of first calculating mud parameters, then calculating formation resistivity, plate gap (mud / mud cake thickness), and formation dielectric constant. Thus, data such as fractured formation resistivity, plate gap (mud / mud cake thickness), and formation dielectric constant can be obtained, and then a three-parameter image of the fractured formation can be obtained using an image display method.

[0041] Step M, identifying open fractures from actual well logging data;

[0042] After obtaining a three-parameter image of a fractured formation, closed fractures must be eliminated from all fractures to identify open fractures. Based on the resistivity image of the fractured formation and the plate gap (mud / mud cake thickness) image, a resistivity threshold and a distance threshold are first established. All fractures are then traversed, and if the resistivity at a fracture location is greater than or equal to the resistivity threshold, the fracture is annotated. Subsequently, within the annotated fractures, if the plate gap (mud / mud cake thickness) at the fracture location is greater than or equal to the distance threshold, the fracture is annotated again. Finally, these twice-annotated fractures are numbered according to their depth, completing the identification of open fractures.

[0043] Step N, calculating the open fracture dip angle from actual well logging data;

[0044] For the identified open fractures and their resistivity images, a binary image of the open fractures is obtained, and the noise influence of the fracture background stratum is removed. The fracture pixels are marked as 0 and the background stratum is marked as 1. The fracture morphology is drawn using a sine curve. The height between the peak and valley points of the sine curve is recorded as h. Combined with the wellbore diameter d, the fracture inclination is calculated, which is:

[0045]

[0046] Step O, extracting the resistivity and dielectric constant of the open fracture background formation from the actual well logging data;

[0047] First, a sliding window is used to ensure that the open fracture can be completely placed in the sliding window. Then, the background formation resistivity and dielectric constant of the open fracture are filtered within the sliding window. Then, the inverse distance weighted algorithm is used to extract the background formation parameters, which can be expressed as:

[0048]

[0049] Among them, R tb , ε fb are the calculated values ​​of the background resistivity and dielectric constant of the extracted open fractures, m and n are the number of rows and columns of image pixels in the sliding window, R tinv,i,j , ε finv,i,jare the inversion calculated values ​​of the formation resistivity and dielectric constant of each pixel in the slip image, α i,j is the weight of each pixel point in the image within each sliding window, which satisfies the inverse distance weighted algorithm according to the distance between each pixel point and the crack.

[0050] Step P, calculating the open fracture opening of actual logging data;

[0051] A fracture aperture calculation model for oil-based mud electrical imaging logging in formations containing open fractures is called. First, a corresponding fracture aperture calculation model is selected based on a combination of mud resistivity, mud dielectric constant, formation resistivity, formation dielectric constant, plate gap (mud / mud cake thickness), fracture dip, and fracture length parameters from the actual logging data. Then, based on the calibration relationship between the physical model and the numerical model in step I, the measured impedance amplitude and phase responses at the fracture locations in the actual data are calibrated to the numerical simulation model responses. Subsequently, the calibrated values ​​of the measured impedance amplitude and phase responses are input into the fracture aperture calculation model, and the fracture aperture is calculated. Finally, all open fractures in the logging data are traversed, and the aperture calculations for all open fractures are completed in sequence.

[0052] Compared with the prior art, the advantages of the present invention are:

[0053] (1) The present invention combines step D and step J, adopts a numerical simulation model and a physical model, and can obtain the response data of oil-based mud electrical imaging logging in fractured formations, realize the scale conversion from numerical simulation data to measured data, and can obtain the response data under all possible parameter combination conditions in batches, thereby improving the applicability of the method proposed in the present invention and better serving the interpretation of actual logging data.

[0054] (2) In step K, a machine learning algorithm is proposed to establish a fracture opening calculation model, which can realize the calculation of fracture opening under the influence of many parameters, especially considering the influence of fracture inclination. This has not been mentioned in the existing fracture opening calculation method, and has great advantages over the simple experimental fitting calculation method used in traditional water-based mud electrical imaging logging.

[0055] (3) Steps M to P, for the first time, proposed a method and process for fracture identification and aperture calculation based on actual data of fractured formations, which can distinguish between open fractures and closed fractures. A method for extracting background formation parameters of open fractures was also proposed, and the aperture of open fractures was calculated, thus solving the problem of fracture parameter evaluation in oil-based mud environment.

[0056] In summary, the present invention develops a method for calculating the aperture of open fractures using a machine learning algorithm after first obtaining data such as mud parameters, plate gap (mud / mud cake thickness), fracture background formation parameters, and fracture inclination. This method solves the problem of quantitative fracture parameter evaluation that has plagued oil-based mud electrical imaging logging, and can better serve formation imaging and interpretation evaluation. It combines homogeneous formation model evaluation with fracture formation model evaluation, highlighting their respective roles. The mean formation model solves for multiple necessary unknown parameters, laying the foundation for fracture aperture calculation. In both homogeneous formation models and fracture formation models, all possible multi-parameter combination measurements and calculations can be performed, which has a wide range of applications and is beneficial for fracture parameter evaluation in various reservoirs. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is an overall schematic diagram of the implementation of the present invention.

[0058] Figure 2 This is a flow chart for calculating fracture opening degree in oil-based mud electrical imaging logging according to the present invention.

[0059] Figure 3 This is a flow chart of the multi-parameter forward modeling database of oil-based mud electrical imaging logging according to the present invention.

[0060] Figure 4 It is a schematic diagram of electrical imaging logging of oil-based mud in fractured formations according to the present invention.

[0061] Figure 5 It is a schematic diagram of the process of building a physical model of a fractured formation according to the present invention.

[0062] Figure 6 is the electrical imaging logging response curve of oil-based mud in fractured formation of the present invention, wherein, Figure 6 (a) and (b) are the measured impedance amplitude and phase response when the fracture resistivity is 2Ωm. The horizontal axis is the background formation resistivity, and the legend is the fracture opening size. Figure 6 (c) and (d) are the measured impedance amplitude and phase responses when the crack resistivity is 2000Ωm.

[0063] Figure 7 is the fracture extraction result of the electrical imaging logging image of the fractured oil-based mud of the present invention, wherein, Figure 7 (a) is the formation resistivity image obtained through inversion imaging; Figure 7 (b) is the grayscale image obtained after processing; Figure 7 (c) in the figure is the binary image obtained by thresholding. Figure 7 (d) in the figure is the image after thinning; Figure 7 (e) in the figure is the image obtained after image filtering.

[0064] Figure 8 is the crack opening calculation result based on machine learning in the present invention, where Figure 8 (a) to (f) represent the calculation results under different determination coefficients.

[0065] Figure 9 This is a schematic diagram of the effect of the crack inclination on the crack opening calculation of the present invention, wherein: Figure 9 (a)-(c) in the figure give the calculation results of crack opening when the crack inclination angle is 20°, 50°, and 80° respectively. Figure 9 (d) in the figure is the result of calculating the crack inclination angle when all data with crack inclination angles of 20°, 50°, and 80° are put into a machine learning model. DETAILED DESCRIPTION

[0066] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings.

[0067] The present invention is actually applied in well logging situations, such as Figure 1 shown. Figure 1 In the figure, a wellbore 1 passes through a formation 2 filled with oil-based mud. The formation 2 contains multiple layers of different reservoir types. Due to the pressure differential between the wellbore and the formation, a layer of mud / mud cake 3 of uneven thickness adheres to the wellbore wall. A logging instrument 4, in this case a microresistivity scanning imaging logging instrument, is suspended in the wellbore 1. The logging instrument 4 is connected to a derrick 6 on the surface via an armored cable 5. The other end of the armored cable 5 is connected to a winch 8 on a logging vehicle 7. The logging vehicle 7 is equipped with a microcomputer control system (not shown) to control the movement of the downhole logging instrument 4. The logging instrument 4 is connected to an imaging plate 10 via a support arm 9. During operation, the support arm 9 pushes against the imaging plate 10, bringing it into close contact with the wellbore wall. As the logging instrument 4 moves within the wellbore, the imaging plate 10 measures the wellbore wall and formation at regular sampling intervals.

[0068] Combined with the above Figure 1 The overall schematic diagram of the work implementation in the present invention is a method for calculating fracture parameters based on the measurement results of the imaging plate 10, that is, a method for calculating fracture opening based on oil-based mud electrical imaging logging. The entire calculation process is as follows: Figure 2 As shown, it includes a homogeneous formation model part, a fractured formation model part and a fractured formation measured data processing part, including the following steps:

[0069] Step A, establish a forward model of the basic response of oil-based mud electrical imaging logging. Here, the "basic response" is to distinguish the oil-based mud electrical imaging logging response of the fractured formation in the following steps. Use numerical simulation methods, such as three-dimensional finite element and finite difference methods, to establish a numerical simulation model of oil-based mud electrical imaging logging in homogeneous formations (without fractures). In the numerical simulation model, set the geometric and material properties of the formation 2, instrument 4, and plate 10, add megahertz electric field signal excitations of different frequencies, collect the measurement signal of the imaging plate 10 through grid division and solution calculation, and combine the instrument constants to obtain key information such as impedance ratio, impedance amplitude, and impedance phase. In this process, the key parameter of formation 2 is the formation resistivity R t and the formation dielectric constant ε t In addition, the key parameter of the wellbore wall adhered mud / mud cake is the mud resistivity R m , mud dielectric constant ε m , plate gap (mud / mud cake thickness) d m At this time, the measurement response of the imaging plate 10 can be expressed as

[0070] Y=F(fr,G,R m ,ε m ,d m ,R t ,ε t ) (1)

[0071] Among them, Y is the measurement response vector, which mainly includes information such as the measured impedance ratio, impedance amplitude and impedance phase at different frequencies, and f r For frequency measurement, G is the combined response of the structure and material properties of instrument 4 and electrode plate 10. When the structure and material properties are fixed, G is a constant. F is the response function, a complex, nonlinear, multivariate response function whose analytical expression is difficult to directly determine.

[0072] Step B: Establish a multi-parameter forward modeling database for oil-based mud electrical imaging logging. Figure 3 As shown, the forward model of the imaging plate 10 is used to batch build a forward database under the influence of multiple parameters. Here, the multiple parameters include: measurement frequency (f r1 、f r2 ,…,f rn ), oil-based mud resistivity R m , Dielectric constant of oil-based mud ε m , Plate gap (mud / mud cake thickness) d m , formation resistivity R t and the formation dielectric constant ε fThe forward modeling database includes the measured responses under the influence of multiple parameters, namely, the measured impedance ratio, impedance amplitude, impedance phase, etc., as well as the corresponding multi-parameter data. The measured responses correspond one-to-one to the multi-parameter combination.

[0073] Step C, establish a multi-parameter cascade inversion model for oil-based mud electrical imaging logging. Here, "cascade" means that the parameter inversion process has a sequence. The parameter inversion order is formulated as: mud resistivity, mud dielectric constant, formation resistivity, plate gap (mud / mud cake thickness), and formation dielectric constant. The specific process is: First, the data corresponding to the low-resistance formation is screened out in the forward modeling database. The low-resistance formation can be determined as a low-resistance formation based on a pre-set threshold, such as less than 10Ωm. Then, in the low-resistance formation data, a mud parameter calculation model is established using a machine learning algorithm or a Gauss-Newton algorithm, and the mud resistivity and dielectric constant can be determined based on this model. Subsequently, a set of data corresponding to the determined mud parameters is screened out in the forward modeling database, such as a mud resistivity of 1200Ωm and a mud dielectric constant of 8. Finally, among the screened data corresponding to the mud parameters, a three-parameter inversion calculation model is established using the machine learning algorithm or the Gauss-Newton algorithm. The measured impedance amplitude and measured impedance phase at each frequency are used as input (known quantities), and the three parameters (output quantities) are calculated in sequence, namely, formation resistivity, electrode gap (mud / mud cake thickness), and formation dielectric constant.

[0074] Step D: Establish an electrical imaging logging model for oil-based mud in fractured formations. Figure 1 Based on the overall schematic diagram of the logging process shown in FIG, a fracture 11 is inserted into the formation 1, as shown in FIG. Figure 4 As shown in the figure, fracture 11 has three key morphological parameters: fracture inclination θ (the angle between the fracture plane and the horizontal), fracture length l (the extension length along the fracture plane), and fracture opening d (the fracture width perpendicular to the fracture plane). The imaging plate 10, driven by the instrument 4 and the push arm 9, moves up and down against the wellbore wall. When it passes near fracture 11, within the detection range (detection depth and resolution) of the plate 10, the measurement signal of the plate 10 will contain the response signal of the fracture 11.

[0075] Step E, classify the types of cracks. According to the natural filling properties of the cracks 11, they are divided into open cracks and closed cracks. Under the pressure of the wellbore 1, mud will invade the inside of the open cracks to form mud-filled cracks. Closed cracks are generally filled with various minerals, such as feldspar, quartz, clay, etc. According to the size of the crack inclination angle θ, the cracks are divided into low-angle cracks (θ≤15°), medium-angle cracks (15<θ≤60°) and high-angle cracks (θ>60°). Therefore, according to the inclination state and filling properties of the cracks, the cracks can be divided into six types, namely low-angle open cracks, medium-angle open cracks, high-angle open cracks, low-angle closed cracks, medium-angle closed cracks, high-angle closed cracks, etc. It should be noted that the open and closed states of the same crack at the contact points in different orientations may not be unique, such as in Figure 4 In the figure, at point 11a, the crack may be in an open state, while at point 11b, the crack may be in a closed state. Therefore, the same crack may also be defined as a partially open or closed crack.

[0076] Step F, establish a numerical simulation model of the electrical imaging logging response of oil-based mud in fractured formations. Use numerical simulation methods, such as three-dimensional finite element and finite difference methods, to establish a numerical simulation model of the electrical imaging logging response of oil-based mud in fractured formations. In the numerical simulation model, set the geometric and material properties of the formation 2, fracture 11, instrument 4, and plate 10, add megahertz electric field signal excitations of different frequencies, collect the measurement signal of the imaging plate 10 through grid division and solution calculation, and combine the instrument constants to obtain key information such as impedance ratio, impedance amplitude, and impedance phase. In this process, the key parameter of formation 2 is the formation resistivity R t and the formation dielectric constant ε t , as the background parameter information of crack 11. The key parameter of crack 11 is the resistivity R of the crack filling material. f , the dielectric constant ε of the filling material f , crack opening d f , fracture length l, fracture inclination θ. In addition, the key parameters of the mud / mud cake 3 attached to the well wall are the mud resistivity R m , mud dielectric constant ε m , plate gap (mud / mud cake thickness) d m At this time, the measurement response of the imaging plate 10 can be expressed as

[0077] Z=F(fr,G,R m ,ε m ,d m ,R t ,ε t ,R f ,ε f ,d f ,l,θ) (2)

[0078] Where Z is the measurement response vector, which mainly includes information such as the measured impedance ratio, impedance amplitude and impedance phase at different frequencies, and f r For frequency measurement, G is the combined response of the structure and material properties of instrument 4 and plate 10. When the structure and material properties are fixed, G is a constant. F is the crack response function, a complex, nonlinear, and multivariate response function whose analytical expression is difficult to directly determine.

[0079] Step G, electrical imaging logging response and sensitivity analysis of oil-based mud in fractured formations. Using the electrical imaging logging response model of oil-based mud in fractured formations, the logging response characteristics of different fracture openings are studied under different fracture filling materials (oil-based mud, feldspar, clay, etc.), different fracture inclinations, different background formation parameters, etc. According to the response characteristics, the fracture opening response sensitive area and non-sensitive area are divided. The sensitive area is where the measurement signal has a significant change in response to the change in fracture opening; the non-sensitive area is where the measurement signal has no significant change in response to the change in fracture opening. Determine the parameter range corresponding to each area, so as to guide the subsequent fracture opening calculation process and effect analysis.

[0080] Step H: Establish a physical model of oil-based mud electrical imaging logging in fractured formations. Figure 5 As shown, under laboratory or logging field conditions, the first step is to use rocks that are easy to cut and shape [ Figure 5 (a)], such as 1m×1m granite, quartzite, etc., use rock drilling mechanical device to penetrate the rock to form a hole, which is equivalent to a wellbore [ Figure 5 (b)]. Then, each core block is evenly divided into two parts using a core cutting mechanical device. Figure 5 (c)]. Subsequently, a carving device was used to carve grooves of different inclination angles, different openings, and different extension lengths on the inner wall of the borehole, which were equivalent to cracks. Some of the cracks were filled with minerals such as quartz and feldspar, which were equivalent to high-resistance closed cracks, and some of the cracks were filled with clay and mud, which were equivalent to low-resistance closed cracks. Figure 5 (d)]. Next, the two rock blocks are assembled into a whole rock, and the gaps between the rocks are cemented. Figure 5 (e)]. Subsequently, multiple fractured rocks were constructed in the same way, and all the rocks were assembled together in a way that they penetrated the wellbore, thus building a fractured formation model. Figure 5 (f)]. Finally, oil-based mud is injected into the wellbore of the formation model and held stationary for a period of time to allow the mud to penetrate the open fractures, completing the construction of the fractured formation. Throughout this process, the resistivity and dielectric constant of the rock mass are measured, along with the inclination, opening, extension, depth, and filling type of the fractures, as well as the resistivity and dielectric constant of the oil-based mud, and these data are recorded and stored.

[0081] Step I: Response and correction of oil-based mud electrical imaging logging for a fractured formation physical model. Place the imaging pad 10 into the well of the fractured formation physical model. Perform measurements at multiple frequencies, recording and saving pad position, measured impedance amplitude, and other information. Combined with the formation model information recorded in Step H, the measured response expression for the fractured formation physical model can be constructed:

[0082] Z * =F(fr,G,R m ,ε m ,d m ,R t ,ε t ,R f ,ε f ,d f ,l,θ) (3)

[0083] In the above formula, F, (fr, G, R m ,ε m ,d m ,R t ,ε t ,R f ,ε f ,d f ,l,θ) has the same meaning as expression (2). * Indicates the response data in the physical model of fractured formations. When the parameters of mud, formation, and fracture are the same, the response value Z of the numerical simulation is the same as the response value Z of the physical model. * There are gain, offset and other phenomena between them, so we further establish a correction function from the physical model response value to the numerical simulation response value, that is,

[0084] Z * =P(Z) (4)

[0085] In the above formula, the function P represents the response value Z of the numerical simulation and the response value Z of the physical model. * There are gain, offset or more complex correction operations between them. In this way, the numerical simulation results can be used to process actual logging data.

[0086] Step J, a multi-parameter response database of electrical imaging logging of oil-based mud in fractured formations. Based on the numerical model, physical model and response analysis of electrical imaging logging of oil-based mud in fractured formations, first determine the opening range of the fracture in batches, such as the range of 10μm to 1cm, and determine the opening numerical points according to linear or logarithmic division; determine the inclination angle range of the fracture, such as the range of 0° to 90°, and determine the inclination angle numerical points according to the division method of every 5°; determine the length of the fracture, such as using the radial detection depth of the electrode 10 as the threshold, and determine the fracture length numerical points according to the linear division method in the range of 0mm to the radial detection depth and the range of the radial detection depth to more than 10cm. Then determine the formation resistivity range in batches, such as 0.2Ωm to 20000Ωm, and determine the formation resistivity numerical points according to the logarithmic division method; determine the formation dielectric constant range, such as 2 to 200 (frequency f r1 ), 1.8 to 120 (frequency f r2 ), determining the range of the formation dielectric constant using a linear or logarithmic partitioning method. Subsequently, the plate gap (mud / mud cake thickness) range is determined, for example, from 0.2 mm to 20 mm, using a linear partitioning method. Finally, under the action of voltage signals of two or more megahertz, batch calculations are performed to collect the corresponding measured impedance ratios, impedance amplitudes, and impedance phase angles for each parameter combination, thereby establishing a multi-parameter response database for oil-based mud electrical imaging logging in formations containing open fractures.

[0087] Step K: Calculate the fracture aperture of oil-based mud electrical imaging logging in formations containing open fractures. The fracture aperture calculation process for formations containing open fractures is relatively complex. Similar to the cascade inversion process for oil-based mud electrical imaging logging, three sets of parameters must be determined sequentially: first, the mud resistivity and dielectric constant must be determined; then, the resistivity and dielectric constant of the background formation in which the fracture is located, and the plate gap (mud / mud cake thickness) directly opposite the fracture opening surface must be determined; and finally, the inclination and extension length of the open fracture must be determined. Then, corresponding response data are extracted from the multi-parameter response database for oil-based mud electrical imaging logging in formations containing open fractures in this order. Finally, a machine learning algorithm or Gauss-Newton algorithm is used to establish a fracture aperture calculation model, using the measured impedance amplitude and phase at multiple frequencies as inputs (known quantities) and the fracture aperture as output (to be solved). Repeat these steps to achieve a one-to-one mapping of the measured response data corresponding to all combinations of mud parameters, formation parameters, gap parameters, inclination parameters, and extension length to fracture aperture.

[0088] Step L: Inversion calculation and imaging of three parameters of oil-based mud electrical imaging logging in fractured formations. Using the oil-based mud electrical imaging logging cascade inversion model, the oil-based mud electrical imaging logging data from fractured formations are processed in the order of first calculating mud parameters, then formation resistivity, plate gap (mud / mud cake thickness), and formation dielectric constant. This yields data such as fractured formation resistivity, mud / mud cake thickness (plate gap), and formation dielectric constant. Using an image display method, a three-parameter image of the fractured formation is obtained, laying the foundation for subsequent fracture extraction and analysis.

[0089] Step M, identification of open fractures based on actual logging data. After obtaining the three-parameter image of the fractured formation, it is necessary to eliminate closed fractures from all fractures and identify open fractures. Based on the resistivity image of the fractured formation and the plate gap (mud / mud cake thickness) image, first formulate a resistivity threshold and a distance threshold, such as 1000Ωm and 1mm. Then traverse all fractures, and when the resistivity at the fracture position is greater than or equal to the resistivity threshold, mark the fracture; then, in the marked fractures, when the plate gap (mud / mud cake thickness) at the fracture position is greater than or equal to the distance threshold, mark the fracture again; finally, number the fractures that have been marked twice according to the change in formation depth to complete the identification of open fractures.

[0090] Step N, the calculation method of the open fracture inclination of the actual logging data. For the identified open fractures and their resistivity images, first use image processing methods, such as image filtering, grayscale, binarization, thinning and other image operations, to obtain a binary image of the open fracture, and remove the noise influence of the fracture background stratum. The fracture pixels are marked as 0 and the background stratum is marked as 1. Then use curve fitting methods, such as the Hough transform method, to use a sine curve to draw the fracture morphology. Subsequently, record the height value between the peak point and the valley point of the sine curve, recorded as h. Finally, combined with the wellbore diameter d, the fracture inclination can be calculated, that is

[0091]

[0092] Step O, a method for extracting the background formation resistivity and dielectric constant of open fractures from actual logging data. The background formation resistivity and dielectric constant near a certain open fracture are generally not a single value, so it is difficult to directly use a certain resistivity value and dielectric constant value as the background formation resistivity and dielectric constant near the open fracture. In order to extract the background formation resistivity and dielectric constant near the open fracture, first use a sliding window, such as a sliding window with a window length L of 5cm or 20cm, or satisfy L=5h, to ensure that the open fracture can be completely placed in the sliding window. Then, within the sliding window, the background formation resistivity and dielectric constant of the open fracture are filtered, such as median filtering, to reduce the influence of background noise. Subsequently, the inverse distance weighted algorithm is used to extract the background formation parameters, that is, the closer the background formation parameters are to the open fracture, the greater the proportion, and the farther the background formation parameters are from the open fracture, the smaller the proportion, which can be expressed by the formula:

[0093]

[0094] Among them, R tb , ε fb are the calculated values ​​of the background resistivity and dielectric constant of the extracted open fractures, m and n are the number of rows and columns of image pixels in the sliding window, R tinv,i,j , ε finv,i,j are the inversion calculated values ​​of the formation resistivity and dielectric constant of each pixel in the slip image, α i,j is the weight of each pixel point in the image within each sliding window, which satisfies the inverse distance weighted algorithm according to the distance between each pixel point and the crack.

[0095] Step P: Calculate the open fracture aperture of actual well logging data. After obtaining the open fracture dip angle, background formation resistivity, and background formation dielectric constant, a further assumption is made: the extended length of the open fracture is greater than the radial detection depth of the electrode plate 10. Therefore, a fracture aperture calculation model for oil-based mud electrical imaging logging in formations containing open fractures is called. First, based on a combination of parameters such as mud resistivity, mud dielectric constant, formation resistivity, formation dielectric constant, electrode gap (mud / mud cake thickness), fracture dip angle, and fracture length from the actual well logging data, a corresponding fracture aperture calculation model is selected. Then, based on the calibration relationship between the physical model and the numerical model in step I, the measured impedance amplitude and phase responses at the fracture locations in the actual data are calibrated to the numerical simulation model responses. Subsequently, the calibrated values ​​of the measured impedance amplitude and phase responses are input into the fracture aperture calculation model, and after calculation, the fracture aperture is obtained. Finally, all open fractures in the well logging data are traversed, and the aperture calculations for all open fractures are completed in sequence.

[0096] In order to illustrate and verify the effects and benefits of the present invention, the following implementation cases were studied.

[0097] Example 1

[0098] In order to illustrate the technical content of the invention, the electrical imaging logging model of oil-based mud in fractured formations is used to study the measurement response law under the influence of multiple parameters, such as Figure 6 As shown. The horizontal axis is the background formation resistivity, and the legend is the fracture opening size. Figure 6 (a) and (b) are the measured impedance amplitude and phase response when the fracture resistivity is 2Ωm. The horizontal axis is the background formation resistivity, and the legend is the fracture opening size. Figure 6 (c) and (d) show the measured impedance amplitude and phase responses when the fracture resistivity is 2000Ωm. It can be clearly seen that when there is a significant difference between the fracture resistivity and the background formation resistivity, the measured impedance amplitude and phase angle can reflect changes in fracture opening. For example, when the fracture resistivity is 2Ωm, within the range of background resistivity greater than 10Ωm, the impedance amplitude curve and phase curve show a clear separation, indicating that the impedance amplitude and phase can be used to characterize changes in fracture opening. When the fracture resistivity is 2000Ωm, within the range of background resistivity less than 200Ωm, the impedance amplitude curve and phase curve also show a clear separation, and the impedance amplitude and phase can also be used to characterize changes in fracture opening.

[0099] Example 2

[0100] In order to illustrate the technical content of the present invention, fracture identification and extraction operations are performed on the resistivity image of oil-based mud electrical imaging logging in fractured formations. Figure 7 (a) is the formation resistivity image obtained through inversion imaging. It can be seen that there are three inclined fractures distributed in this image. Figure 7 (b) is the grayscale image obtained after processing, and the color changes from black to gray and then to white. Figure 7 (c) is a binary image obtained using the threshold method. The crack trajectory is marked in black and the background stratum is marked in white. However, the width of the cracks is different and it also contains background noise. Figure 7 (d) is the image after thinning. The pixels corresponding to the crack width remain consistent and are used to extract cracks, but still contain background noise. Figure 7 (e) is the image obtained after image filtering, which removes the background noise. The image only contains the crack trajectory and no longer contains the background formation noise. Figure 7 (f) is the sine curve obtained by Hough transform. Therefore, the amplitude and phase information of the sine function can be obtained according to the sine function, and the crack inclination can be calculated using formula (5).

[0101] Example 3

[0102] In order to illustrate the technical content of the invention, a machine learning algorithm is used to establish an open crack opening calculation model, and the degree of consistency between the actual opening value and the calculated value is compared. Figure 8 Comparative results are presented for different fracture resistivities (different fracture filling materials) and plate gaps (mud / mud cake thickness). Fractures are horizontal, and fracture apertures range from 0.01 to 100 mm. The abscissa represents the number of fracture aperture samples, and the ordinate represents the fracture aperture. Each graph contains two sets of data, with the actual fracture aperture values ​​represented by lines and the calculated values ​​represented by asterisks. Fracture apertures span five orders of magnitude, and the calculated and actual values ​​in each graph are highly consistent, with coefficients of determination (R²) exceeding 98%. This demonstrates that, given known fracture resistivity and plate gap (mud / mud cake thickness), a machine learning algorithm can be used to establish a relationship between measured response and fracture aperture. This validates the logic and feasibility of calculating fracture aperture after obtaining the formation resistivity distribution (which already has the fracture resistivity) and plate gap (mud / mud cake thickness) using a cascaded multi-parameter inversion model for electrical imaging logging of oil-based mud in homogeneous formations.

[0103] Example 4

[0104] In order to illustrate the technical content of the present invention, a machine learning algorithm is used to establish the influence of different crack inclination angles on the calculation of crack opening. Figure 9 (a)-(c) respectively give the calculation results of the crack opening when the crack inclination angle is 20°, 50°, and 80°. It can be found that when the crack inclination angle is determined (whether it is low or high), the corresponding machine learning model can be used to calculate the crack opening, and the determination coefficient R2 can reach more than 99%. However, when the crack inclination angle is unknown, such as Figure 9 As shown in (d), when all data with crack inclination angles of 20°, 50°, and 80° are put into a machine learning model, the determination coefficient of the crack inclination angle calculation result is 95.8%, which is lower than the calculation result of a single known crack inclination angle, indicating that the crack inclination angle has an impact on the calculation of crack opening, verifying the logic and feasibility of calculating the crack opening after obtaining the crack inclination angle in the content of the present invention.

Claims

1. A method for calculating fracture opening of oil-based mud electrical imaging logging, characterized in that: The following steps are involved: Step A, establishing a basic response forward model of oil-based mud electrical imaging logging; Step B, establishing a multi-parameter forward modeling database for oil-based mud electrical imaging logging; Step C, establishing a multi-parameter cascade inversion model for oil-based mud electrical imaging logging; The cascade means that the parameter inversion process has a sequence, and the parameter inversion sequence is as follows: mud resistivity, mud dielectric constant, formation resistivity, plate gap i.e. mud / mud cake thickness, formation dielectric constant; The specific process is as follows: first, the data corresponding to the low-resistance formation is screened out in the forward modeling database, and the low-resistance formation is determined according to a preset threshold; then, in the low-resistance formation data, a mud parameter calculation model is established using a machine learning algorithm or a Gauss-Newton algorithm, and the mud resistivity and dielectric constant can be determined based on the model; then, all the data corresponding to the determined mud parameters are screened out in the forward modeling database in turn; finally, in each group of data corresponding to the determined mud parameters, a three-parameter inversion calculation model is established using a machine learning algorithm or a Gauss-Newton algorithm, and the measured impedance amplitude and measured impedance phase at each frequency are used as input quantities (known quantities), and three parameters (output quantities) are calculated in sequence, namely, formation resistivity, electrode gap (mud / mud cake thickness), and formation dielectric constant; Step D, establishing a fractured formation model for oil-based mud electrical imaging logging; A fracture (11) is inserted into the formation (2), and the fracture (11) has three morphological parameters, namely, fracture inclination angle θ (the angle between the fracture plane and the horizontal direction), fracture length l (the extension length along the fracture plane direction), and fracture opening d (the fracture width in the direction perpendicular to the fracture plane direction). An oil-based mud electrical imaging logging fracture formation model is constructed based on the three morphological parameters; the imaging plate (10) moves upward or downward close to the well wall and measures the formation (2). When passing near the fracture (11), within the detection range (detection depth and resolution) of the imaging plate (10), the measurement signal of the imaging plate (10) will contain the contribution of the fracture (11); Step E: Classify crack types According to the inclination state and filling property of the cracks, the cracks are divided into six types, namely, low-angle open cracks, with an inclination angle θ≤15°; medium-angle open cracks, with an inclination angle of 15<θ≤60°; high-angle open cracks, with an inclination angle θ>60°; low-angle closed cracks; medium-angle closed cracks; high-angle closed cracks; Step F: Establishing a numerical simulation model for fracture-type formation response in oil-based mud electrical imaging logging In the numerical simulation model, the geometric and material properties of the formation (2), the fracture (11), the logging instrument (4), and the imaging plate (10) are set, and megahertz electric field signal excitations of different frequencies are added. Through grid generation and solution calculation, the measurement signal of the imaging plate (10) is collected, and the impedance ratio, impedance amplitude, and impedance phase information are obtained by combining the instrument constants. In this process, the key parameter of the formation (2) is the formation resistivity R t and the formation dielectric constant ε t , as the background parameter information of the crack (11); the key parameter of the crack (11) is the resistivity R of the crack filling material f , the dielectric constant of the filling material ε f , crack opening d f , fracture length l, fracture inclination θ; in addition, the key parameters of the mud / mud cake attached to the well wall 3 are the mud resistivity R m 、Mud dielectric constant ε m , plate gap (mud / mud cake thickness) d m At this time, the measurement response expression of the imaging plate 10 is Z=F(fr,G,R m ,ε m ,d m ,R t ,ε t ,R f ,ε f ,d f ,l,θ) (2) Among them, Z is the measurement response vector of fractured formation, which mainly includes information such as measured impedance ratio, impedance amplitude and impedance phase at different frequencies, and f r is the measurement frequency, G is the comprehensive response value of the structure and material properties of the instrument 4 and the plate 10, when the structure and material properties are determined, G is a constant value; F is the crack response function; Step G: Electrical imaging logging response and sensitivity analysis of oil-based mud in fractured formations According to the response characteristics, the crack opening response sensitive area and insensitive area are divided; the sensitive area is where the measurement signal has obvious changes in the crack opening change; determine the parameter range corresponding to each area to guide the subsequent crack opening calculation process and effect analysis; Step H, establishing a physical model of oil-based mud electrical imaging logging in fractured formations; Under laboratory or logging site conditions, firstly, the rock pores are equivalent to wellbores, and then each core block is evenly divided into two parts; then, on the inner wall of the pores, grooves with different inclinations, different openings, and different extension lengths are carved, which are equivalent to cracks, of which some cracks are filled with quartz and feldspar minerals, which are equivalent to high-resistance closed cracks, and some cracks are filled with clay mud, which are equivalent to low-resistance closed cracks; the two parts of the rock blocks are assembled into a whole rock, and the gaps between the rocks are cemented with cement; then, multiple fractured rocks are constructed in the same way, and all the rocks are assembled together in a way of penetrating the wellbore, so as to build a fractured formation model; finally, oil-based mud is poured into the wellbore of the formation model and kept still for a period of time, so that the oil-based mud can penetrate into the open cracks to complete the construction of the fractured formation. During the whole process, the resistivity and dielectric constant of the rock blocks need to be measured, the inclination, opening, extension length, depth position, filling type of the fractures, and the resistivity and dielectric constant of the oil-based mud need to be recorded and preserved; Step I, fracture type formation physical model oil-based mud electrical imaging logging response and correction; The imaging plate 10 is placed in the well of the fracture-type formation physical model constructed in step H, and measurements are carried out at multiple frequencies. The plate position, measured impedance amplitude, impedance amplitude and other information are recorded and saved. Combined with the formation model information recorded in step H, the measurement response expression of the fracture-type formation physical model is constructed: Z * =F(fr,G,R m ,ε m ,d m ,R t ,ε t ,R f ,ε f ,d f ,l,θ) (3) In the above formula, F, (fr, G, R m ,ε m ,d m ,R t ,ε t ,R f ,ε f ,d f ,l,θ) has the same meaning as expression (2); Z * It represents the response data in the physical model of fractured formation. When the parameters of mud, formation and fracture are the same, the response value Z of numerical simulation is different from the response value Z of physical model. * There are gain and offset phenomena between them, so we further establish a correction function from the physical model response value to the numerical simulation response value, that is, WITH * =P(Z) (4) In the above formula, the function P represents the response value Z of the numerical simulation and the response value Z of the physical model. * There are gain, offset, or more complex correction operations between them; Step J, a database of multi-parameter responses of electrical imaging logging of oil-based mud in formations containing open fractures; Based on the numerical model of oil-based mud electrical imaging logging in fractured formations in step F, the physical model in step H and the response analysis in step I, firstly, the range of the opening of the opening fractures is determined in batches, and the numerical points of the opening are determined in a linear or logarithmic division mode; the range of the inclination angle of the opening fractures is determined, and the numerical points of the inclination angle are determined in a division mode of every 5°; the length of the opening fractures is determined, and the numerical points of the fracture length are determined in a linear division mode; then the range of the formation resistivity is determined in batches, and the numerical points of the formation resistivity are determined in a logarithmic division mode; the range of the formation dielectric constant is determined, and the range of the formation dielectric constant is determined in a linear or logarithmic division mode; then the range of the plate gap (mud / mud cake thickness) is determined, and the plate gap (mud / mud cake thickness) is determined in a linear division mode; finally, under the action of more than 2 megahertz voltage signals, the corresponding measured impedance ratio, impedance amplitude and impedance phase angle under the combination of each parameter are collected through batch calculation, thereby establishing a multi-parameter response database of oil-based mud electrical imaging logging in formations containing opening fractures; Step K, establishing a fracture opening calculation model for oil-based mud electrical imaging logging in formations containing open fractures; Step L, obtaining a three-parameter image of a fractured formation; Step M, identifying open fractures from actual well logging data; Step N, calculating the open fracture dip angle of actual well logging data; Step O, extracting the resistivity and dielectric constant of the open fracture background formation from the actual well logging data; Step P, calculating the open fracture opening of the actual logging data.

2. The method for calculating fracture opening of oil-based mud electrical imaging logging according to claim 1, characterized in that: The following steps are involved: The step A, establishing the forward model of the basic response of oil-based mud electrical imaging logging is specifically as follows: Based on the existing logging system, the logging system includes a logging instrument 4 suspended in a wellbore 1, the logging instrument 4 is a micro-resistivity scanning imaging logging instrument, the imaging electrode 10 at the end of the logging instrument 4 is in contact with the well wall, and the well wall formation is measured at intervals of a sampling interval, and a basic response forward model of oil-based mud electrical imaging logging in a homogeneous formation without cracks is established by a numerical simulation method, and the numerical simulation method includes a three-dimensional finite element method and a finite difference method; In the basic response forward model, the geometric and material properties of the formation 2, the logging instrument 4, and the imaging plate 10 are set; the megahertz electric field signal excitation of different frequencies is added, and the measurement signal of the imaging plate 10 is collected through meshing and solution calculation, and the impedance ratio, impedance amplitude, and impedance phase information are obtained in combination with the instrument constant; in the process of measuring the wellbore formation, the parameter of the formation 2 is the formation resistivity R t and the formation dielectric constant ε t ; The parameter of mud / mud cake 3 attached to the well wall is the mud resistivity R m 、Mud dielectric constant ε m , plate gap (mud / mud cake thickness) d m ; The measurement response expression of the imaging plate 10 is Y=F(fr,G,R m ,ε m ,d m ,R t ,ε t ) (1) Where Y is the measurement response vector, including information such as the measured impedance ratio, impedance amplitude and impedance phase at different frequencies, and f r is the measurement frequency, G is the comprehensive response value of the structure and material properties of the logging tool 4 and the imaging plate 10. When the structure and material properties are determined, G is a constant value, and F is the response function.

3. The method for calculating fracture opening of oil-based mud electrical imaging logging according to claim 1, characterized in that: The step B, establishing a multi-parameter forward modeling database for oil-based mud electrical imaging logging, is specifically as follows: Using the basic response forward model of the imaging plate 10, a batch forward database under the influence of multiple parameters is established, wherein the multiple parameters include: measurement frequency (f r1 、f r2 ,…,f rn ), oil-based mud resistivity R m 、Oil-based mud dielectric constant ε m , Plate gap (mud / mud cake thickness) d m , formation resistivity R t and the formation dielectric constant ε f The forward modeling database also includes the measurement responses under the influence of multiple parameters, namely, the measured impedance ratio, impedance amplitude, impedance phase, etc. and the corresponding multi-parameter data. The measurement responses under the influence of multiple parameters correspond one to one with the multi-parameter combination.

4. The method for calculating fracture opening of oil-based mud electrical imaging logging according to claim 1, characterized in that: The step K, establishing a fracture opening calculation model for oil-based mud electrical imaging logging in formations containing open fractures, is specifically as follows: Firstly, the resistivity and dielectric constant of mud, the resistivity and dielectric constant of the background stratum where the fracture is located, the plate gap (mud / mud cake thickness) opposite to the fracture opening surface, the inclination angle and extension length parameters of the fracture opening are determined in sequence; then, the response data corresponding to the determined parameters are extracted from the multi-parameter response database of oil-based mud electrical imaging logging in formations containing open fractures; finally, the measured impedance amplitude and phase at multiple frequencies are used as input quantities (known quantities), and the fracture opening is used as the output quantity (quantity to be solved); a fracture opening calculation model is established using a machine learning algorithm or a Gauss-Newton algorithm; and the above operations are repeated to achieve a one-to-one mapping of the measured response data corresponding to the combination of all mud parameters, formation parameters, plate gap (mud / mud cake thickness) parameters, inclination parameters and extension length to the fracture opening.

5. The method for calculating fracture opening of oil-based mud electrical imaging logging according to claim 1, characterized in that: The step L of obtaining the three-parameter image of the fractured formation is specifically as follows: The oil-based mud electrical imaging logging cascade inversion model is used to process the oil-based mud electrical imaging logging data of fractured formations in the order of first calculating mud parameters, then calculating formation resistivity, plate gap (mud / mud cake thickness), and formation dielectric constant. Thus, data such as fractured formation resistivity, plate gap (mud / mud cake thickness), and formation dielectric constant can be obtained, and then a three-parameter image of the fractured formation can be obtained using an image display method.

6. The method for calculating fracture opening of oil-based mud electrical imaging logging according to claim 1, characterized in that: The step M, identifying the open cracks in the actual logging data, is specifically as follows: after obtaining the three-parameter image of the fractured formation, it is necessary to eliminate the closed cracks from all the cracks and identify the open cracks, and firstly formulate a resistivity threshold and a distance threshold according to the resistivity image of the fractured formation and the plate gap (mud / mud cake thickness) image; then traverse all the cracks, and when the resistivity at the crack position is greater than or equal to the resistivity threshold, mark the crack; Subsequently, when annotating fractures, when the plate gap (mud / mud cake thickness) at the fracture location is greater than or equal to the distance threshold, the fracture is annotated again; finally, the fractures annotated twice are numbered according to the change in formation depth to complete the identification of open fractures.

7. The method for calculating fracture opening of oil-based mud electrical imaging logging according to claim 1, characterized in that: The step N of calculating the open fracture dip angle of the actual logging data is specifically as follows: For the identified open fractures and their resistivity images, a binary image of the open fractures is obtained, and the noise influence of the fracture background strata is removed. The fracture pixels are marked as 0 and the background strata are marked as 1. A sine curve is used to draw the fracture morphology. The height value between the peak point and the valley point of the sine curve is recorded as h. Combined with the wellbore diameter d, the fracture inclination is calculated, that is, 8. The method for calculating fracture opening of oil-based mud electrical imaging logging according to claim 1, characterized in that: In step O, the method for extracting the resistivity and dielectric constant of the open fracture background formation from the actual well logging data is specifically as follows: First, a sliding window is used to ensure that the open fracture can be completely placed in the sliding window. Then, the background formation resistivity and dielectric constant of the open fracture are filtered in the sliding window. Then, the inverse distance weighted algorithm is used to extract the background formation parameters, which can be expressed as: Among them, R tb , ε fb are the extracted background formation resistivity and dielectric constant values ​​of the open fractures, m and n are the number of rows and columns of image pixels in the sliding window, respectively. tinv,i,j , ε finv,i,j are the inversion calculated values ​​of the formation resistivity and dielectric constant of each pixel in the slip image, α i,j is the weight of each pixel point in the image within each sliding window, and satisfies the inverse distance weighted algorithm according to the distance between each pixel point and the crack.

9. The method for calculating fracture opening of oil-based mud electrical imaging logging according to claim 1, characterized in that: The step P, calculating the open fracture opening of the actual logging data is specifically as follows: The fracture opening calculation model of the oil-based mud electrical imaging logging in the formation containing open fractures is called. First, according to the combination of mud resistivity, mud dielectric constant, formation resistivity, formation dielectric constant, plate gap (mud / mud cake thickness), fracture inclination and fracture length parameters of the actual logging data, the corresponding fracture opening calculation model is selected. Then, according to the correction relationship between the physical model and the numerical model in step I, the measured impedance amplitude and phase and other responses of the fracture position of the actual data are corrected to the numerical simulation model response; then, the correction values ​​of the measured impedance amplitude and phase response are input into the fracture opening calculation model, and the fracture opening can be obtained after calculation; finally, all the open fractures in the logging data are traversed, and the calculation of the opening of all the open fractures is completed in sequence.

Images