A method for evaluating fracture permeability based on high and low frequency reflection coefficients and attenuation of stoneley waves

Abstract

The application discloses a method for evaluating fracture permeability based on high and low frequency reflection coefficients and attenuation of Stoneley waves, and comprises the following steps: pre-processing array acoustic waveforms; separating CSG data combinations of each depth point to obtain upgoing and downgoing wave CSG combinations; combining the upgoing and downgoing waves into upgoing and downgoing wave COG data respectively; separating the upgoing and downgoing wave COG data to obtain direct waves, upgoing and downgoing reflected waves and residual direct waves; calculating average values of high and low frequency reflection coefficients respectively, calculating the difference between the high and low frequency reflection coefficients and normalizing; after data preprocessing, calculating the high and low frequency attenuations of the CSG of each depth point, calculating the difference between the high and low frequency attenuations and normalizing; multiplying the normalized difference between the reflection coefficients and the difference between the attenuations to define a fracture permeation factor, which is used for evaluating the fracture permeability of a formation. The application can solve the technical problem of detecting permeable structures in a formation.

Description

Technical Field

[0001] This invention belongs to the field of borehole geophysics, specifically relating to a method for evaluating fracture permeability using the difference in low-frequency and high-frequency reflection coefficients and attenuation of Stoneley waves. Background Technology

[0002] Fractures serve as pathways for oil and gas to enter the wellbore, making reservoir fracture characterization crucial. Commonly used ultrasonic / acoustic techniques for fracture characterization include ultrasonic borehole imaging, Stoneley wave reflection (e.g., Hornby et al., 1989; Tang et al., 1993; Kostek et al., 1998), and cross-dipole anisotropy (e.g., Joyce et al., 1998; Tang and Patterson, 2001) measurements and analyses. While all these techniques have proven effective in detecting fractures, distinguishing the fracture response from other effects remains a challenging task.

[0003] Stoneley waves are low-frequency interface waves that propagate along the wellbore. They have a relatively large amplitude and a wave velocity slightly lower than the sound velocity of the fluid within the well. They are highly sensitive to changes in formation properties; variations in formation parameters caused by formation heterogeneity affect Stoneley wave propagation and are reflected in changes in the Stoneley wave reflection coefficient. When Stoneley waves encounter fractures passing through the wellbore, due to the hydraulic interaction between the wave and the fracture, part of the wave is reflected from the fracture. Fracture density, aperture, and other factors all influence Stoneley wave reflection. Therefore, the Stoneley wave reflection coefficient is often used as a means of identifying fractures and evaluating the formation.

[0004] However, the Stoneley wave reflection technique currently used to characterize fractures has difficulty distinguishing between Stoneley wave reflections caused by formation boundaries and wellbore variations. Further exploration of Stoneley wave reflection information to differentiate reflections caused by factors such as formation boundaries and wellbore variations would provide support and a basis for fracture permeability assessment. Summary of the Invention

[0005] The purpose of this invention is to provide a method for evaluating fracture permeability based on the high and low frequency reflection coefficients and attenuation of Stoneley waves, in order to solve the technical problem of detecting permeable structures in formations by calculating the reflection coefficients and attenuation of Stoneley waves in different frequency ranges.

[0006] To achieve the above method, the present invention adopts the following processing scheme:

[0007] Step 1: Use bandpass filtering to preprocess the array acoustic waveform to eliminate random noise interference.

[0008] Step 2: Perform wavefield separation based on linear prediction on the CSG data combination at each depth point to obtain the up-wave CSG combination and the down-wave CSG combination.

[0009] Step 3: Combine the uplink and downlink waves into uplink COG data and downlink COG data, respectively. Perform wavefield separation based on linear prediction on the uplink COG data to obtain the direct wave and the uplink reflected wave; perform wavefield separation based on linear prediction on the downlink COG data to obtain the downlink reflected wave and the residual direct wave.

[0010] Step 4: Calculate the average reflection coefficients of the high-frequency and low-frequency bands using the separated direct wave and downlink reflected wave, respectively, and calculate and normalize the difference between the high-frequency and low-frequency reflection coefficients.

[0011] Step 5: After completing the data preprocessing in Step 1, perform high-frequency and low-frequency attenuation calculations on the CSG at each depth point, calculate the difference between high and low frequency attenuation and normalize it.

[0012] Step 6: Multiply the difference in normalized reflection coefficients and the difference in attenuation to define the fracture permeability factor, which is used to evaluate the formation fracture permeability.

[0013] Step one specifically involves:

[0014] The noise frequency differs significantly from the Stoneley wave frequency; a bandpass filter can be used to remove the noise. The response function of the windowed bandpass filter is:

[0015]

[0016] Steps two and three are specifically as follows:

[0017] The principle of the wavefield separation method based on linear prediction is as follows:

[0018] In a wellbore direct wave exhibiting dispersion, the slowness can be expressed as S. l (ω), the direct-path wave signal from the wellbore of the nth receiver can be expressed as:

[0019]

[0020] In the formula, A l (ω) represents the spectrum of each mode, L represents the number of modes, and z represents the distance from the sound source to the receiver.

[0021] A specific type of direct-path wave from the nth receiver to the mth receiver can be represented as:

[0022] A l (ω)exp[-iω(mn)dS l (ω)] (3)

[0023] According to linear prediction theory, the matrix form of the linear equations for solving various direct wave spectra is as follows:

[0024]

[0025] Where E l =exp(-iωdS l (ω)), d is the receiver spacing, W N (ω) represents the actual data spectrum. Let the left and right sides of Equation 4 be equal, and represent it as a matrix.

[0026] G·A=W (5)

[0027] Typically, the number of direct wave types is less than the number of array receivers, i.e., L < N. Therefore, only the least squares solution of A can be calculated, which can be obtained from the following formula:

[0028]

[0029] in Let G be the complex conjugate transpose, ε be the damping factor, and I be the identity matrix. Introducing the damping factor can improve the stability of the algorithm.

[0030] Wavefield separation using the linear prediction method described above is divided into two steps: first, the CSG combination is separated, and then the COG combination is separated.

[0031] For each depth point, the data is arranged into a CSG data set according to the receiver array. The CSG data contains an up-wave and a down-wave. The up-wave contains a direct wave from the source to the receiver and a reflected wave reflected back from the lower reflector. The down-wave is a wave reflected back from the upper reflector. The two up-waves have the same time offset, which is equal to the Stoneley wave slowness "+s". The down-wave has a negative time offset, and its propagation in the array can be regarded as having an apparent slowness "-s". Therefore, in equation (4), let L = 2, that is, s1 = +s, s2 = -s, and we can perform the analysis of the two modes.

[0032] After separating the uplink and downlink waves, data from intermediate receivers at each depth are collected and combined into uplink COG and downlink COG combinations. The uplink COG combination includes a direct wave with a slowness of 0 and an uplink reflected wave with a slowness of "+2s". The downlink wave includes a residual direct wave with a slowness of 0 and a downlink reflected wave with a slowness of "-2s". The two sets of data are separated to obtain the direct wave and downlink reflected wave required for calculation.

[0033] Step four specifically involves:

[0034] For each depth point, the ratio of the downlink reflected wave spectrum to the direct wave spectrum is calculated to obtain the reflection coefficient as a function of frequency. The average reflection coefficients in the ranges of 0-1kHz and 1-2kHz are calculated respectively to obtain the low-frequency and high-frequency reflection coefficients.

[0035] Tang and Cheng (1993) derived a theory for calculating the reflection of Stoneley waves.

[0036]

[0037] In the formula, R is the reflection coefficient, z is the fracture zone thickness, the fracture zone may contain one or more fractures perpendicular or inclined to the well axis, and kz and k0 are the Stoneley wave numbers of the fracture zone and the formation, respectively.

[0038] Figure 2 The reflection coefficient of a permeable crack, calculated using the above theory, exhibits a frequency-dependent characteristic, increasing towards lower frequencies. Based on this frequency-dependent characteristic of the Stoneley wave reflection coefficient, the average reflection coefficients at low frequencies (e.g., 0-1 kHz) and high frequencies (e.g., 1-2 kHz) were calculated separately. When the low-frequency reflection coefficient is significantly higher than the high-frequency reflection coefficient, it indicates that the reflection is likely caused by a permeable crack.

[0039] Step five specifically involves:

[0040] The data from the first receiver obtained in step one is combined into a COG data combination, and the Stoneley wave attenuation is calculated using linear fitting, as follows:

[0041] Assume that the amplitude spectrum of the waveform received by a receiver at depth z satisfies Formula 8

[0042]

[0043] In the formula, S(w) and R(w) represent the spectrum of the sound source and the receiver, respectively; G(w,z) is the Green's function that controls the geometric diffusion of the sound wave from the source to the receiver; T is the propagation time; and Q is the quality factor.

[0044] The spectral ratios at different depths z1 and z2 satisfy Formula 9.

[0045]

[0046] In the formula, f is the frequency and V is the wave velocity. It is usually assumed that the geometric diffusion G depends only on the depth and is independent of the frequency. Therefore, it can be... Treating it as a frequency-independent constant, then at the same frequency, the natural logarithm of the spectral ratio can be used. A linear fit is performed between the distance (z2-z1) between the two receivers, and the slope of the fitted line α=πf / QV is defined as the attenuation factor.

[0047] For acoustic logging instruments, since they have multiple receivers, there is a linear relationship between any two receivers, satisfying Equation 10.

[0048]

[0049] in The attenuation coefficient within the span of the receiver array can be numerically determined using the least squares linear fitting method, thus representing the sound wave attenuation.

[0050] Figure 3 The theoretically calculated attenuation of permeable cracks varies with frequency, exhibiting a characteristic of increasing towards lower frequencies. Based on this frequency-dependent characteristic of Stoneley wave attenuation, the average attenuation at low frequencies (e.g., 0-1 kHz) and high frequencies (e.g., 1-2 kHz) was calculated separately. When the low-frequency attenuation is significantly higher than the high-frequency attenuation, it indicates that the attenuation is likely caused by permeable cracks.

[0051] Step six specifically involves:

[0052] The difference between the normalized reflection coefficients and the difference in attenuation are multiplied to define the crack seepage factor, defined as follows:

[0053]

[0054] A high fracture permeability factor indicates strong formation fracture permeability, while a low value indicates weak permeability.

[0055] The present invention has the following advantages:

[0056] Compared to conventional methods of identifying reflection interfaces using reflection coefficients and attenuation, this invention utilizes the differences in high and low frequency reflection coefficients and attenuation to distinguish between permeable cracks and non-permeable layer interfaces.

[0057] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0058] Figure 1 This is a flowchart for calculating the crack seepage factor.

[0059] Figure 2 This represents the theoretical reflection coefficient of a permeable crack.

[0060] Figure 3 This is the theoretical attenuation for permeable fractures.

[0061] Figure 4-1 shows the original waveform.

[0062] Figure 4-2 This is the filtered waveform.

[0063] Figure 5 This is the original waveform CSG array.

[0064] Figure 6 It is an up-wave CSG array.

[0065] Figure 7 It is a down-wave CSG array.

[0066] Figure 8 It is a direct-wave COG array.

[0067] Figure 9 It is an uplink reflected wave COG array.

[0068] Figure 10 It is a downlink reflected wave COG array.

[0069] Figure 11 The measured data shows the variation of the reflection coefficient with frequency.

[0070] Figure 12 The diagram illustrates the results of applying this method. Detailed Implementation

[0071] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions in the embodiments of this invention will be clearly and completely described below in conjunction with the embodiments of this invention. Obviously, the described embodiments are only some embodiments of this invention, not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0072] like Figure 1 As shown, this invention proposes a method for identifying permeable cracks by utilizing the differences in high- and low-frequency reflection coefficients and attenuation between high and low frequencies. The workflow is as follows:

[0073] Step 1: Preprocess the array acoustic waveform using a bandpass filter to eliminate random noise interference. The response function of the windowed bandpass filter is:

[0074]

[0075] Here, f1 = 0.3kHz, f1 = 0.5kHz, f1 = 3kHz, and f1 = 3.2kHz.

[0076] As shown in Figure 4, the left image shows the original monopole waveform. After bandpass filtering (bandpass filtering range 0.5-3kHz), the longitudinal wave, transverse wave and random noise are eliminated, leaving only the Stoneley wave, as shown in the right image of Figure 4.

[0077] Step 2: Combine the array acoustic wave data into receiver array data according to CSG. The waveform includes direct Stoneley wave, uplink reflected wave, and downlink reflected wave, such as... Figure 5 As shown. Wavefield separation based on linear prediction is performed on the CSG data combination at each depth point to obtain the up-going wave CSG combination ( Figure 6 ) and downwave CSG combination ( Figure 7The up-going wave contains both direct waves and up-going reflected waves, requiring further separation. The down-going wave contains both down-going reflected waves and residual direct waves, requiring further filtering to remove the residual direct waves.

[0078] The principle of the wavefield separation method based on linear prediction is as follows:

[0079] In a wellbore direct wave exhibiting dispersion, the slowness can be expressed as S. l (ω), the direct-path wave signal from the wellbore of the nth receiver can be expressed as:

[0080]

[0081] In the formula, A l (ω) represents the spectrum of each mode, L represents the number of modes, and z represents the distance from the sound source to the receiver.

[0082] A specific type of direct-path wave from the nth receiver to the mth receiver can be represented as:

[0083] A l (ω)exp[-iω(mn)dS l (ω)] (3)

[0084] According to linear prediction theory, the matrix form of the linear equations for solving various direct wave spectra is as follows:

[0085]

[0086] Where E l =exp(-iεdS l (ε)), d is the receiver spacing, W N (ε) represents the actual data spectrum. Let the left and right sides of Equation 4 be equal, and represent it as a matrix.

[0087] G·A=W (5)

[0088] Typically, the number of direct wave types is less than the number of array receivers, i.e., L < N. Therefore, only the least squares solution of A can be calculated, which can be obtained from the following formula:

[0089]

[0090] in Let G be the complex conjugate transpose, ε be the damping factor, and I be the identity matrix. Introducing the damping factor can improve the stability of the algorithm.

[0091] Wavefield separation using the linear prediction method described above is divided into two steps: first, the CSG combination is separated, and then the COG combination is separated.

[0092] For each depth point, the data is arranged into a CSG data set according to the receiver array. The CSG data contains an up-wave and a down-wave. The up-wave contains a direct wave from the source to the receiver and a reflected wave reflected back from the lower reflector. The down-wave is a wave reflected back from the upper reflector. The two up-waves have the same time offset, which is equal to the Stoneley wave slowness "+s". The down-wave has a negative time offset, and its propagation in the array can be regarded as having an apparent slowness "-s". Therefore, in equation (4), let L = 2, that is, s1 = +s, s2 = -s, and we can perform the analysis of the two modes.

[0093] Step 3: Combine the uplink and downlink CSG data. Take the data from the middle receiver at each depth point and combine them according to depth to form uplink COG data and downlink COG data respectively. The uplink COG combination includes a direct wave with a slowness of 0 and an uplink reflected wave with a slowness of "+2s". The downlink combination includes a residual direct wave with a slowness of 0 and a downlink reflected wave with a slowness of "-2s". Perform wavefield separation based on linear prediction on the uplink COG data to obtain the direct wave ( Figure 8 ) and upward reflected wave ( Figure 9 ); Wavefield separation based on linear prediction is performed on the downlink COG data to filter out residual direct waves and obtain the downlink reflected wave. Figure 10 ).

[0094] Step 4: Perform Fourier transform on the downlink reflected wave and the direct wave at each depth point, and calculate the ratio between the two to obtain the reflection coefficient as a function of frequency, such as... Figure 11 As shown. The low-frequency and high-frequency reflection coefficients can be obtained by calculating the average values ​​of the reflection coefficients within the ranges of 0-1kHz and 1-2kHz, respectively. The difference between the low-frequency and high-frequency reflection coefficients is then calculated and normalized by dividing by the maximum value of the difference.

[0095] Tang and Cheng (1993) derived a theory for calculating the reflection of Stoneley waves.

[0096]

[0097] In the formula, R is the reflection coefficient, z is the fracture zone thickness, the fracture zone may contain one or more fractures perpendicular or inclined to the well axis, and kz and k0 are the Stoneley wave numbers of the fracture zone and the formation, respectively.

[0098] Figure 2The reflection coefficient of a permeable crack, calculated using the above theory, exhibits a frequency-dependent characteristic, increasing towards lower frequencies. Based on this frequency-dependent characteristic of the Stoneley wave reflection coefficient, the average reflection coefficients at low frequencies (e.g., 0-1 kHz) and high frequencies (e.g., 1-2 kHz) were calculated separately. When the low-frequency reflection coefficient is significantly higher than the high-frequency reflection coefficient, it indicates that the reflection is likely caused by a permeable crack.

[0099] Step 5: After completing the data preprocessing in Step 1, perform attenuation calculations on the CSG data at each depth point. Calculate the average attenuation for low frequencies (0-1kHz) and high frequencies (1-2kHz) respectively to obtain the low-frequency and high-frequency attenuations. Calculate the difference between the low-frequency and high-frequency attenuations and normalize by dividing by the maximum value of the difference.

[0100] The attenuation calculation principle is as follows:

[0101] Assume that the amplitude spectrum of the waveform received by a receiver at depth z satisfies Formula 8

[0102]

[0103] In the formula, S(w) and R(w) represent the spectrum of the sound source and the receiver, respectively; G(w,z) is the Green's function that controls the geometric diffusion of the sound wave from the source to the receiver; T is the propagation time; and Q is the quality factor.

[0104] The spectral ratios at different depths z1 and z2 satisfy Formula 9.

[0105]

[0106] In the formula, f is the frequency and V is the wave velocity. It is usually assumed that the geometric diffusion G depends only on the depth and is independent of the frequency. Therefore, it can be... Treating it as a frequency-independent constant, then at the same frequency, the natural logarithm of the spectral ratio can be used. A linear fit is performed between the distance (z2-z1) between the two receivers, and the slope of the fitted line α=πf / QV is defined as the attenuation factor.

[0107] For acoustic logging instruments, since they have multiple receivers, there is a linear relationship between any two receivers, satisfying Equation 10.

[0108]

[0109] in The attenuation coefficient within the span of the receiver array can be numerically determined using the least squares linear fitting method, thus representing the sound wave attenuation.

[0110] Figure 3The theoretically calculated attenuation of permeable cracks varies with frequency, exhibiting a characteristic of increasing towards lower frequencies. Based on this frequency-dependent characteristic of Stoneley wave attenuation, the average attenuation at low frequencies (e.g., 0-1 kHz) and high frequencies (e.g., 1-2 kHz) was calculated separately. When the low-frequency attenuation is significantly higher than the high-frequency attenuation, it indicates that the attenuation is likely caused by permeable cracks.

[0111] Step 6: Multiply the difference in normalized reflection coefficients and the difference in attenuation, and define the crack seepage factor as follows:

[0112] G = (Ref 低频 -Ref 高频 ) 归一化 ×(Att 低频 -Att 高频 ) 归一化 (11)

[0113] A high fracture permeability factor indicates strong formation fracture permeability, while a low value indicates weak permeability.

[0114] Figure 12 The image shows the interpretation results using this method. From the 7th monopole wave train, it can be seen that the 4155–4160m section exhibits strong attenuation of both reflected and direct waves, and the fracture permeability factor also shows a high response in this section, indicating the development of permeable fractures. Combined with the 1st and 2nd electrical imaging data, it is evident that this section contains a large number of fractures, and the two data show good consistency, proving the effectiveness of this method.

Claims

1. A method for evaluating crack permeability based on Stoneley wave high and low frequency reflection coefficients and attenuation, comprising the following steps: Step 1: Preprocessing the array acoustic waveform using bandpass filtering to eliminate random noise interference; Step 2: Perform wavefield separation based on linear prediction on the CSG data combination at each depth point to obtain the up-wave CSG combination and the down-wave CSG combination; Step 3: Combine the uplink and downlink waves into uplink COG data and downlink COG data respectively; perform wavefield separation based on linear prediction on the uplink COG data to obtain the direct wave and the uplink reflected wave; perform wavefield separation based on linear prediction on the downlink COG data to obtain the downlink reflected wave and the residual direct wave. Step 4: Calculate the average reflection coefficient of the high-frequency band and the average reflection coefficient of the low-frequency band using the separated direct wave and downlink reflected wave, respectively. frequency The high-frequency reflection coefficient and low-frequency reflection coefficient are calculated, and the difference between the high-frequency reflection coefficient and the low-frequency reflection coefficient is normalized. Step 5: After completing the data preprocessing in Step 1, calculate the average attenuation of the high-frequency band and the average attenuation of the low-frequency band for the CSG data at each depth point, calculate the difference between the average attenuation of the high-frequency band and the average attenuation of the low-frequency band, and normalize it. Step Six: Normalize the high frequency The difference between the reflection coefficient and the low-frequency reflection coefficient, multiplied by the difference between the average attenuation in the high-frequency band and the average attenuation in the low-frequency band, is defined as the fracture permeability factor, which is used to evaluate the permeability of formation fractures. Steps two and three are specifically as follows: The principle of the wavefield separation method based on linear prediction is as follows: In a wellbore direct wave exhibiting dispersion, the slowness can be expressed as: The direct-path wave signal from the wellbore of the nth receiver can be expressed as: (2) In the formula, Let L be the frequency spectrum of each vibration mode, L be the number of vibration modes, and z be the distance from the sound source to the receiver. The direct wave from the wellbore to the m-th receiver is represented as... (3) According to linear prediction theory, the matrix form of the linear equations for solving various direct wave spectra is as follows: (4) in d is the receiver spacing. For the actual data spectrum; let the left and right sides of Equation 4 be equal, and represent it as a matrix. G.A = W (5) Typically, the number of direct wave types is less than the number of array receivers, i.e., L < N. Therefore, only the least squares solution of A can be calculated, obtained from the following formula: (6) in Let G be the complex conjugate transpose, ε be the damping factor, and I be the identity matrix; introducing the damping factor can improve the stability of the algorithm. The wavefield separation using the above linear prediction method is divided into two steps: the first step is to separate the combined CSG data, and the second step is to separate the COG data. For each depth point, the data is arranged into a CSG data combination according to the receiver array; the CSG data combination contains an up-going wave and a down-going wave. The up-going wave contains a direct wave from the source to the receiver and a reflected wave reflected back from the lower reflector. The down-going wave is a wave reflected back from the upper reflector. The time offset of the two up-going waves is the same, which is equal to the Stoneley wave slowness "+s". The time offset of the down-going wave is negative. The propagation of the wave in the array is regarded as having an apparent slowness "-s". Therefore, in equation (4), let L = 2, that is, s1 = +s, s2 = -s, and perform the analysis of the two modes. After separating the uplink and downlink waves, data from intermediate receivers at each depth are collected and combined to form uplink COG data and downlink COG data. The uplink COG data includes a direct wave with a slowness of 0 and an uplink reflected wave with a slowness of "+2s". The downlink COG data includes a residual direct wave with a slowness of 0 and a downlink reflected wave with a slowness of "-2s". The two sets of data are separated to obtain the direct wave and downlink reflected wave required for calculation. Step five specifically involves: The data from the first receiver obtained in step one is combined into a COG data combination, and the Stoneley wave attenuation is calculated using linear fitting, based on the following principle: Assume that the amplitude spectrum of the waveform received by a receiver at depth z satisfies Equation 8 In the formula, S(w) and R(w) represent the spectrum of the sound source and the receiver, respectively; Green's function is used to control the geometric spread of sound waves from the source to the receiver; T is the propagation time; Q is the quality factor. The spectral ratio at different depths z1 and z2 satisfies Formula 9. In the formula, f is the frequency and V is the wave speed. It is assumed that the geometric diffusion G depends only on the depth and is independent of the frequency; therefore, it can be... If treated as a frequency-independent constant, then at the same frequency, the natural logarithm of the spectral ratio can be used. A linear fit is performed on the distance (z2 — z1) between the two receivers, and the slope of the fitted line is... That is, it is defined as the attenuation factor; For acoustic logging instruments, since they have multiple receivers, there is a linear relationship between any two receivers, satisfying Equation 10. in The attenuation coefficient within the span of the receiver array can be numerically determined by the least squares linear fitting method, which represents the Stoneley wave attenuation. The attenuation of permeable fractures, based on theoretical calculations, varies with frequency and shows an increasing trend towards lower frequencies. Based on the frequency-dependent characteristic of Stoneley wave attenuation, the average attenuation in the low-frequency band and the average attenuation in the high-frequency band were calculated separately. When the average attenuation in the low-frequency band is significantly higher than that in the high-frequency band, it indicates that the attenuation is likely caused by infiltrative cracks.

2. The method for evaluating crack permeability based on the high and low frequency reflection coefficients and attenuation of Stoneley waves according to claim 1, wherein step one specifically comprises: The noise frequency differs significantly from the Stoneley wave frequency; a bandpass filter can be used to remove the noise. The response function of the windowed bandpass filter is: (1) 。 3. The method for evaluating crack permeability based on the high and low frequency reflection coefficients and attenuation of Stoneley waves according to claim 1, wherein step four specifically comprises: For each depth point, the ratio of the downlink reflected wave spectrum to the direct wave spectrum is calculated to obtain the reflection coefficient as a function of frequency; the average reflection coefficients in the ranges of 0-1kHz and 1-2kHz are calculated to obtain the low-frequency reflection coefficient and the high-frequency reflection coefficient. Theories for calculating Stoneley wave reflection In the formula, R is the reflection coefficient, z is the fracture zone thickness, the fracture zone may contain one or more fractures perpendicular or inclined to the well axis, and kz and k0 are the Stoneley wave numbers of the fracture zone and the formation, respectively. The reflection coefficient of the permeable crack, calculated using the above theory, shows a characteristic of increasing towards lower frequencies. Based on the frequency-dependent characteristic of the Stoneley wave reflection coefficient, the average reflection coefficients at low and high frequencies were calculated separately. When the low-frequency reflection coefficient is significantly higher than the high-frequency reflection coefficient, it indicates that the reflection is caused by the permeable crack.

4. The method for evaluating crack permeability based on the high and low frequency reflection coefficients and attenuation of Stoneley waves according to claim 1, wherein step six specifically comprises: The difference between the normalized reflection coefficients and the difference in attenuation are multiplied to define the crack seepage factor, defined as follows: ( 11) A high fracture permeability factor indicates strong formation fracture permeability, while a low value indicates weak permeability.

Images