Noise logging single-point signal synthesis method

Abstract

The invention discloses a noise logging single-point signal synthesis method, and belongs to the technical field of geophysical logging, and the method specifically comprises the following steps: S1, calculating a gain correction coefficient of single-point measurement data, and carrying out the gain correction of a measurement point frequency spectrum; s2, performing uniform equalization and enhanced equalization on the frequency spectrum of the measuring point after gain correction to obtain the frequency spectrum of each single-point data after equalization; and S3, performing interpolation synthesis on the frequency spectrum data subjected to equalization processing in a depth dimension by using an interpolation algorithm to obtain continuous frequency spectrum data in the depth dimension. According to the method, the single-point measurement spectrum signals which cannot be used for analyzing the string groove leakage or the absorption intensity of the absorption layer are processed, the spectrum data equalization technology and the interpolation technology are used for performing depth dimension synthesis on the single-point measurement spectrum signals, the noise spectrum with the continuous depth dimension is obtained, and the signal-to-noise ratio is increased.

Description

Technical Field

[0001] This invention belongs to the field of geophysical logging technology, specifically relating to a method for synthesizing single-point signals in noise logging. Background Technology

[0002] Noise logging instruments utilize spectrum analysis technology to clearly identify the sources of noise generated by fluid flow through imaging analysis of the downhole noise frequency spectrum and noise intensity. This effectively identifies the location of channeling and the effective injection sites, providing valuable guidance for subsequent engineering operations. However, currently, there is no universally applicable commercial software for processing, interpreting, and evaluating noise spectrum logging data, both domestically and internationally. Data interpretation methods remain at the level of qualitative analysis, failing to fully realize the application potential of noise spectrum logging and hindering the widespread application of noise spectrum logging instruments. Especially under current conditions, many oilfield production wells are mostly horizontal wells with screen completions. For this type of completion, conventional flow imaging logging technology cannot detect fluid leakage points, measure and calculate the production intensity of the producing layers, or determine the stratigraphic position, thus affecting the final quantitative evaluation. Since noise spectrum logging instruments can detect fluid flow information in the annular space outside the screen and within the formation, in order to better understand the downhole production situation and determine the location of the producing layer under screen completion conditions, it is necessary to use a combination of noise logging and conventional logging techniques to effectively identify the flow state of downhole fluids and diagnose abnormalities.

[0003] Noise signals are a type of acoustic signal. Traditional acoustic signal processing techniques focus on the study of acoustic wave propagation laws. Improving the signal-to-noise ratio of single-point noise logging signals and effectively synthesizing the noise spectrum of single-point signals to obtain noise signals reflecting the surface, and then studying fluid leakage and the production intensity of the producing layers throughout the entire downhole depth range, represents a new technical challenge in the field of petroleum logging. For single-point frequency synthesis techniques, common signal synthesis techniques include linear interpolation, cubic spline interpolation, and parabolic interpolation. For single-point spectral signal extraction, techniques such as calculating the average value of the spectrum are used. However, common spectral signal synthesis techniques have significant drawbacks. They do not consider the differences in acoustic amplitude detection at different frequencies during single-point depth measurement, thus losing the characteristics of acoustic amplitude jitter. The extracted single-point spectral signals cannot characterize the changes in acoustic amplitude at a single frequency, i.e., they cannot characterize the amplitude jitter characteristics at the same frequency. Therefore, there is an urgent need to find a technique that can both extract single-point spectral signals and accurately extract the spectrum that conforms to the characteristics of sound amplitude variation at a single frequency, so as to synthesize the single-point spectrum into a depth-dimensional spectrum, reflect the noise situation within the measurement range, obtain the specific depth of the fluid leakage point or the generation and absorption layer, provide information for the formulation of subsequent measures, sealing the leakage point, obtaining the generation and absorption intensity and amount of each generation and absorption layer, and analyzing the properties of the fluid produced in the generation layer.

[0004] Regarding well logging data, current noise logging instruments have evolved from traditional instruments that record a few frequency curves to spectral noise logging instruments. These instruments can record complete two-dimensional spectral data, and through data processing, can clearly present a complete spectral image that varies with depth and location, making subsequent interpretation and analysis more convenient and intuitive, and the results more accurate. However, compared to the development of the instruments themselves, the development of spectral data processing, interpretation, and application in noise logging instruments lags behind. Current focus remains on data spectrum processing and image display, with data interpretation and evaluation still at the level of qualitative analysis. Quantitative interpretation and evaluation have not yet been researched domestically or internationally, and there are no mature and effective methods. Furthermore, in terms of data application, the lack of quantitative interpretation methods limits the scope of data application, currently confined to leak and cross-contamination detection, and has not yet been effectively expanded to the effective evaluation of injection and production profiles. Regarding noise spectrum processing, how to correctly and effectively synthesize a continuous depth-dimensional spectrum from a single-point spectrum to highlight fluid leakage points and the production intensity and location of the producing layers is also a subject for which no effective technical reports have been found domestically or internationally. Summary of the Invention

[0005] The purpose of this invention is to provide a method for synthesizing single-point signals in noise logging, which solves the technical problem in the prior art that single-point spectra cannot be correctly and effectively synthesized into continuous spectra with depth-dimensional continuity.

[0006] The technical solution adopted in this invention is a method for synthesizing single-point signals in noise logging, which is implemented according to the following steps:

[0007] S1, calculate the gain correction coefficient of the single-point measurement data and perform gain correction on the measurement point spectrum;

[0008] S2, perform uniform equalization and enhancement equalization on the spectrum of the measurement points after gain correction in S1, to obtain the equalized spectrum of each single-point data.

[0009] S3 uses an interpolation algorithm to interpolate and synthesize the equalized spectral data of S2 in the depth dimension, resulting in continuous spectral data in the depth dimension.

[0010] The invention is further characterized by:

[0011] S1 specifically involves: using the gain of each single-point measurement data within the measurement depth range and the maximum gain of all single-point measurement data, calculating the gain correction coefficient of the single-point measurement data, and using the gain correction coefficient to perform gain correction on the measurement point spectrum.

[0012] S1 specifically involves: performing gain correction on the measurement data at each single point within the measurement depth range, using the maximum gain of all single-point measurement data within the measurement depth range as the base value for gain correction, with the correction coefficient being:

[0013] GainCo= GainBase / GainNow (1)

[0014] In equation (1), GainCo is the gain correction product coefficient of the current stationary point, GainBase is the maximum gain among all stationary points, and GainNow is the gain value of the current stationary point;

[0015] The spectral gain correction for each single-point data is shown in equation (2):

[0016] NVDL ij =GainCo k ×NVDL ′ ij (2)

[0017] In equation (2), NVDL ij GainCo presents the spectrum after gain correction following a single-point measurement at the current depth. k NVDL is the gain correction factor for the current single-point depth measurement. ′ ij Let i ∈ (1, ..., m), where m is the number of repeated samples of the single-point measurement spectrum, j ∈ (1, ..., n), where n is the number of frequency samples of the single-point measurement spectrum, with 1 amplitude sampled at each frequency, and n is also the number of amplitudes in the same sample, k ∈ (1, ..., N), where N is the number of single-point measurements within the measurement depth range.

[0018] In S2, uniform equalization specifically involves calculating the average amplitude across all channels of the entire single-point spectrum data, resulting in the n average amplitude values ​​of the single-point spectrum data, denoted as NVDLAVG. j If j∈(1,…,n), then:

[0019]

[0020] Removing the s1 repeated sampling points at the start time and the s2 repeated sampling points at the end time, which are not included in the calculation of the average value of the single-point spectrum data, we get:

[0021]

[0022] In equation (4), s1 and s2 take different values ​​depending on the specific spectral data of each single point measurement. When the single point measurement starts to repeat the measurement, the amplitude of the repeated measurement samples of s1 increases or decreases, and the color displayed on the spectrum changes. At this time, s1 is the number of sampling points whose color changes when the repeated measurement starts on the spectrum. s2 is the number of invalid sampling points of the repeated measurement when the measurement ends.

[0023] The enhanced equilibrium in S2 specifically refers to:

[0024] Calculate the average signal value for each channel at all sampling points:

[0025]

[0026] In equation (5), j is the channel number and m is the number of repeated sampling points in the selected time.

[0027] Calculate the standard deviation of each sampling point:

[0028]

[0029] In equation (6), n is the number of channels;

[0030] The formula for calculating enhanced equalization is obtained by adding the average amplitude of each channel to the standard deviation of all sampled amplitudes of that channel:

[0031]

[0032] Substituting the average signal value and standard deviation of each channel into the above formula, we obtain the formula for calculating the single-point spectrum data after enhanced equalization:

[0033]

[0034] In equation (8), j is the channel number, m is the number of repeated sampling points at the selected time, n is the number of channels, s1 is the repeated sampling point at the start time that was removed, and s2 is the repeated sampling point at the end time that was removed.

[0035] The interpolation methods in S3 include Lagrange interpolation, linear interpolation, parabolic interpolation, exponential interpolation, or cubic spline interpolation.

[0036] In S3, when the equalized spectral data is interpolated and synthesized in the depth dimension, the starting depth and ending depth of all single-point measurement data are selected during interpolation. The starting depth is the minimum depth of all single-point measurements, and the ending depth is the maximum depth of all single-point measurements. The sampling interval for depth dimension synthesis is set, and the maximum spacing during interpolation is set. When the depth of two adjacent single-point data exceeds the maximum spacing, the data of this segment is not interpolated, and the spectral data of this segment is all set to 0, i.e., a blank segment. When the depth interval of two single-point measurement data exceeds 10 meters, the segment between the depths of these two single-point data is not interpolated.

[0037] The beneficial effects of this invention are:

[0038] This invention processes single-point measurement spectrum signals that cannot be used for analyzing leakage in a channel or the absorption intensity of a fluid-generating layer. It employs spectrum equalization and interpolation techniques to synthesize the single-point measurement spectrum signals in the depth dimension, obtaining a continuous noise spectrum and improving the signal-to-noise ratio. The equalized and interpolated spectrum data reflects the noise situation within the measurement range and can be used to analyze leakage points in a channel or the absorption conditions of a fluid-generating layer. It provides the specific depth of the fluid leakage point or the fluid-generating layer, aiding in the development of subsequent measures, including sealing the leakage point, obtaining the absorption intensity and quantity of each layer of the fluid-generating layer, and analyzing the properties of the fluid produced in the producing layer. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the spectrum of single-point noise logging measurements in existing technologies;

[0040] Figure 2 This is a schematic diagram of the spectrum with stable sound amplitude on the same channel in the existing technology;

[0041] Figure 3 This is a schematic diagram of the spectrum where the sound amplitude is unstable on the same channel in the existing technology;

[0042] Figure 4 This is a schematic diagram of the spectrum of a single point at 3871 meters in a certain well in the existing technology;

[0043] Figure 5 This is a schematic diagram of the synthesized spectrum of a well at a depth of 3811-3910 meters using the single-point signal synthesis method of this invention for noise logging in the prior art.

[0044] Figure 6 This is a flowchart illustrating the single-point signal synthesis method for noise logging according to the present invention.

[0045] Figure 7 This is the data file for the noise measurement section of Well Z in Embodiment 5 of the present invention;

[0046] Figure 8 This is a schematic diagram of the noise spectrum of a single point measurement at 3866 meters in Well Z in Embodiment 5 of the present invention;

[0047] Figure 9 This is a schematic diagram of the noise spectrum of a single point measurement at 3868 meters in Well Z in Embodiment 5 of the present invention;

[0048] Figure 10 This is a schematic diagram of the noise spectrum of a single point measurement at 3899 meters in Well Z in Embodiment 5 of the present invention;

[0049] Figure 11 This is a schematic diagram of the noise spectrum of a single point measurement at 3806 meters in Well Z in Embodiment 5 of the present invention;

[0050] Figure 12This is a schematic diagram of the spectrum before noise gain correction for a single-point measurement of well Z at 3901 meters in Embodiment 5 of the present invention;

[0051] Figure 13 This is a schematic diagram of the noise spectrum equalization and depth dimension interpolation synthesis results of well Z from 3800 meters to 3910 meters in Embodiment 5 of the present invention. Detailed Implementation

[0052] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments.

[0053] In oil well logging using noise logging instruments, to improve the accuracy of noise signal measurement, reduce interference from noise at non-leakage or non-production / absorption points, and eliminate interference from noise generated by friction between the instrument and casing arm during continuous measurement, thus improving the signal-to-noise ratio, noise logging instruments are typically designed in a single-point measurement mode. This means the noise logging instrument remains stationary at a certain depth within the casing, and the instrument detects the noise at that depth, obtaining the noise spectrum at that depth. During logging, a measurement point is designed at regular intervals (e.g., every 1 meter) within a certain depth range downhole. Measurement points are denser at suspected fluid leakage points or production / absorption zones to improve noise signal detection accuracy, while measurement points are reduced at non-fluid leakage points or non-production / absorption zones, resulting in multiple single-point noise spectra. Single-point measurement spectra cannot be used to analyze leakage in a trough or the production of a septic layer because the spectrum of a single-point measurement is a repeated measurement, and each sample on the spectrum graph has consistency. In order to obtain a depth-continuous noise spectrum that reflects the noise situation within the measurement range, and to analyze the leakage point in a trough or the production and absorption situation of a septic layer, to obtain the specific depth of the fluid leakage point or the septic layer, for the next step of measures to be formulated, to seal the leakage point, to obtain the generation and absorption intensity and generation amount of each layer of the septic layer, and to analyze the properties of the fluid produced in the production layer, it is necessary to synthesize the noise spectra of all single-point measurements to obtain a depth-dimensional continuous noise spectrum.

[0054] Example 1

[0055] like Figure 6 As shown, the noise logging single-point signal synthesis method disclosed in this invention is implemented according to the following steps:

[0056] S1: Using the gain of each single-point measurement data within the measurement depth range and the maximum gain of all single-point measurement data, calculate the gain correction coefficient of the single-point measurement data. Use the gain correction coefficient to perform gain correction on the measurement point spectrum. The single-point measurement spectrum signal gain correction part uses the gain of each single-point measurement data and the maximum gain of all measurement point data to calculate the gain correction coefficient of the single-point measurement data. Use the gain correction coefficient to perform gain correction on the measurement point spectrum. After gain correction, the amplitude of each single-point measurement spectrum data is at the same level, which can avoid abnormal data segments in the synthesized spectrum due to different gains, affecting the subsequent analysis and calculation of the generation and absorption characteristics of the slot point and the generation and absorption layer.

[0057] S2 performs uniform equalization and enhancement equalization on the measured point spectrum after gain correction in S1, obtaining the equalized spectrum of each individual point data. The single-point measurement spectrum signal equalization processing uses uniform equalization and enhancement equalization techniques to obtain the equalized spectrum of each individual point data. Uniform equalization is suitable when the amplitude changes of repeated measurements on the same channel are not significant. When the amplitude on the same channel is unstable, enhancement equalization should be used to equalize the spectrum data. Enhancement equalization, also known as enhancement averaging, is the uniform average of uniform equalization plus the variance of the amplitude on that channel. The variance of the amplitude change on a single channel represents the degree of amplitude jitter on that channel; small jitter reflects the change of ground noise over time. During equalization processing, the unstable spectrum data at the beginning and end of the single-point measurement are removed to obtain the spectrum signal reflecting the true noise at the measured point.

[0058] S3 uses an interpolation algorithm to interpolate and synthesize the equalized spectral data from S2 in the depth dimension, resulting in continuous spectral data in the depth dimension. The depth-dimensional interpolation and synthesis part of the single-point measurement spectral signal uses the starting and ending depths of all measurement points and a given sampling interval. It employs Lagrange interpolation, linear interpolation, parabolic interpolation, exponential interpolation, or cubic spline interpolation algorithms to interpolate and synthesize the equalized spectral data from all single-point measurements in the depth dimension, obtaining continuous spectral data in the depth dimension.

[0059] Figure 6 The depth-dimensional continuous spectral data obtained in the final step can be further used to determine the stratigraphic position of the groove point and the layer of the absorbing layer, as well as to calculate the absorbing intensity of the absorbing layer.

[0060] Example 2

[0061] Based on Example 1, because the spectrum of single-point noise logging measurements is repeated, the sampled measurements on the spectrum diagram are consistent and cannot reflect the noise leakage and the generation characteristics of the generating layer. The repeated sampled signals in each time dimension of the single-point noise spectrum signal have good consistency; therefore, the amplitude of the sound on the same channel in each repeated sampled spectrum signal is basically the same, which cannot reflect the depth and intensity of the leakage point in the cross-channel, nor can it reflect the layer position and generation intensity of the generating layer. Figure 4 This is a typical single-point measurement spectrum signal diagram of a certain well. The single-point measurement depth of 3871 meters corresponds to the producing layer, which has a large water production. However, this single-point spectrum signal diagram cannot reflect the production situation of the producing layer. The depth-dimensional continuous spectrum diagram synthesized by the technology of this invention is as follows: Figure 5 As shown, Figure 5 It can accurately reflect the output characteristics of the above-mentioned output layer, from Figure 5 As can be seen from the spectrum diagram, there is a strong production signal of the producing layer at 3871 meters in this well. Compared with the single-point noise spectrum signal at 3871 meters, the synthesized depth-dimensional continuous spectrum diagram can not only reflect the production depth of the producing layer, but also calculate the production intensity of the producing layer from the spectrum signal.

[0062] Depend on Figure 1 It can be seen that single-point measurement spectrum cannot be used for analyzing channel leakage or production conditions of the producing layer, because the spectrum of single-point measurement is a repeated measurement, and each sample on the spectrum graph has consistency. Therefore, the single-point measurement data in S1 of this invention is: the data obtained by the noise logging instrument when it is stationary at a certain depth in the well, hence it is called single-point measurement data.

[0063] In the single-point measurement mode of noise logging, the spectrum signal obtained by the instrument's static measurement reflects the noise characteristics at that depth point. The main sources of noise include fluid cross-flow behind casing, friction perforation arm of produced fluid from the producing layer, fluid sucked into the suction layer, and fluid in formation pores (especially fractured pores). Different types of noise sources produce different noise frequency distributions and different amplitudes. Current research results at home and abroad show that the noise frequency generated by fluid flow in the well (oil casing) is less than 1kHz, the noise frequency range of completion tools (perforation, oil and casing perforation) is 1-3kHz, the noise frequency range of back casing trenches (cement cross-flow, reservoir fractures) is 3-5kHz, and the noise frequency range of reservoir fluid flow is 10-15kHz. The frequency range of the spectrum signals acquired by mainstream noise logging instruments at single points is generally around 1kHz-12.8kHz. Within this frequency range, one amplitude is measured every 0.1kHz, for a total of 128 amplitudes. Measurements are repeated multiple times within the 1kHz-12.8kHz frequency range to acquire multiple amplitudes for subsequent calculation of the single-point amplitude. The purpose of acquiring multiple time samples is to perform statistical averaging of the acquired time samples, reducing measurement errors, improving measurement accuracy, and making the single-point spectrum calculation result closer to the true spectrum at that single point. A typical noise logging single-point measurement spectrum diagram is shown below. Figure 1 As shown in the figure, the frequency range of the spectrum signal collected at this point is 1kHz-12.8kHz. Starting from 1kHz, an amplitude is collected every 0.1kHz, resulting in 128 amplitudes at different frequencies. This process is repeated 101.7 times, resulting in 101.7 amplitudes within the 1kHz-12.8kHz frequency range. The time interval between repeated collections is 0.1 seconds, meaning that the amplitude within the 1kHz-12.8kHz frequency range is measured and collected once every 0.1 seconds. The spectrum measurement at this point takes 10.7 seconds.

[0064] Specifically, S1 involves: performing gain correction on the measurement data at each single point within the measurement depth range to ensure that the amplitude of the sound spectrum at each single point is at the same level. This avoids abnormal data segments in the synthesized spectrum due to differences in gain, which could affect the analysis of the absorption and generation at the slotting point and the absorption layer. The gain correction is performed using the maximum gain of all single-point measurement data within the measurement depth range as the base value, and the correction coefficient is:

[0065] GainCo= GainBase / GainNow (1)

[0066] In equation (1), GainCo is the gain correction product coefficient of the current stationary point, GainBase is the maximum gain among all stationary points, and GainNow is the gain value of the current stationary point;

[0067] Therefore, the spectral gain correction for each single-point data is shown in equation (2):

[0068] NVDL ij =GainCo k ×NVDL ′ ij (2)

[0069] In equation (2), NVDL ij GainCo presents the spectrum after gain correction following a single-point measurement at the current depth. k NVDL is the gain correction factor for the current single-point depth measurement. ′ ij Let i ∈ (1, ..., m), where m is the number of repeated samples of the single-point measurement spectrum, j ∈ (1, ..., n), where n is the number of frequency samples of the single-point measurement spectrum, with 1 amplitude sampled at each frequency, and n is also the number of amplitudes in the same sample, k ∈ (1, ..., N), where N is the number of single-point measurements within the measurement depth range.

[0070] After correction using the above gain correction formula, the smaller the gain of the single-point measurement spectrum, the greater the amplification of the spectrum data after gain correction, and vice versa. If the gain values ​​of all single-point measurement spectra within the measurement depth range are at the same level, or if the gain values ​​of all single-point measurement spectra are the same, then the spectrum data of each single-point measurement spectrum after gain correction will not change much or will not change at all.

[0071] Example 3

[0072] Based on Example 1, the uniform equalization in S2 specifically involves: calculating the average amplitude of all channels in the entire single-point spectrum data to obtain the n average amplitude values ​​of the single-point spectrum data, denoted as NVDLAVG. j If j∈(1,…,n), then:

[0073]

[0074] To eliminate measurement instability during the start and end time periods of single-point measurements, the average sound amplitude (NVDLAVG) of n single-point spectral data is calculated. j When the repeated sampling points at the start time s1 and the repeated sampling points at the end time s2 are removed and not included in the calculation of the average value of the single-point spectrum data, equation (3) is obtained after correction:

[0075]

[0076] In equation (4), s1 and s2 have different values ​​depending on the specific spectral data of each single-point measurement. Usually, the amplitude of the repeated measurement samples s1 at the beginning of the single-point measurement will increase or decrease, and the color displayed on the spectrum will change. At this time, s1 is the number of sampling points whose color changed when the repeated measurement started on the spectrum. s2 is the number of invalid sampling points in the repeated measurement when the measurement ends. The sampling points change color on the spectrum, and the amplitude value increases or decreases. An average value is calculated for the amplitude value at each frequency. For example, 128 frequency sampling points will result in 128 average values.

[0077] The above-mentioned single-point spectrum equalization process is called uniform equalization, or uniform averaging. Uniform equalization is suitable for situations where the amplitude of repeated measurements on the same channel does not change significantly. When the amplitude of repeated measurements on the same channel changes significantly, i.e., the stability of repeated measurements is poor, it is necessary to emphasize the role of larger amplitude values, improve signal strength, reflect the jitter of the amplitude value at that frequency, and further improve the signal-to-noise ratio of the amplitude signal at that frequency. This invention proposes enhanced equalization, i.e., enhanced averaging. Enhanced averaging is the uniform average value plus the variance of all effective amplitude values ​​at that frequency. The variance can reflect the jitter of the amplitude value at that frequency. During noise logging, the amplitude of the sound wave sometimes changes with different measurement media and sound sources. The variance can reflect this change. Adding this jitter value to the normal value can better reflect the change in amplitude. The standard deviation of the entire single-point noise spectrum can better reflect the stability of single-point measurements.

[0078] Specifically, the enhanced equalization in S2 is as follows:

[0079] Calculate the average signal value for each channel at all sampling points:

[0080]

[0081] In equation (5), j is the channel number and m is the number of repeated sampling points in the selected time.

[0082] Calculate the standard deviation of each sampling point:

[0083]

[0084] In equation (6), n is the number of channels, which is also the number of frequency samples;

[0085] The formula for calculating enhanced equalization is obtained by adding the average amplitude of each channel to the standard deviation of all sampled amplitudes of that channel:

[0086]

[0087] Substituting the average signal value and standard deviation of each channel into the above formula, we obtain the formula for calculating the single-point spectrum data after enhanced equalization:

[0088]

[0089] In equation (8), j is the channel number, m is the number of repeated sampling points at the selected time, n is the number of channels, i.e. the number of frequency samples, s1 is the repeated sampling point at the start time that was removed, and s2 is the repeated sampling point at the end time that was removed.

[0090] When performing single-point measurements, the sound amplitude at the same frequency should be stable. When the measured sound amplitude on the same channel is unstable, an enhanced average can be used, which is the center value (arithmetic mean) plus the variance. For example, when there is flowing sound, the sound amplitude on the same channel will vary in size, and the variance can be used to represent the degree of this variation.

[0091] When the amplitude of each channel on the spectrum of a certain measurement point is relatively stable compared to other amplitudes on that channel, and the color change of the amplitude of the same channel on the spectrum image is not significant and relatively uniform, then uniform averaging is selected to equalize the spectrum of that stationary point. If the amplitude of each channel on the spectrum of a certain stationary point changes drastically compared to other amplitudes on that channel, and the color change of the amplitude of the same channel on the spectrum image is large and uneven, then enhanced averaging, i.e., the center value (uniform averaging) plus variance, is selected to equalize the spectrum of that single point. The variance proposed in this invention, representing the amplitude change on a single channel, represents the degree of amplitude jitter on that channel. Small jitter reflects the change of ground noise over time.

[0092] Figure 2 In the spectrum diagram, if the amplitude of the sound on the same channel is stable, then uniform equalization should be used for the spectrum of that single point during equalization processing. Figure 3 If the amplitude of the same channel is unstable in the spectrum diagram, then the single-point spectrum will be enhanced by averaging during equalization.

[0093] Furthermore, the interpolation methods in S3 include Lagrange interpolation, linear interpolation, parabolic interpolation, exponential interpolation, or cubic spline interpolation. These methods interpolate all single-point spectral data after equalization in the depth dimension, resulting in noise spectral data with continuous depth (i.e., uniform depth intervals). The interpolated spectral data is a two-dimensional spectral curve, while the interpolation results of the total noise curve and other one-dimensional curves are one-dimensional total noise curves or other one-dimensional curves. Linear methods are generally recommended. The number of interpolation points is the number of single points needed for interpolation; linear interpolation requires a minimum of 2 points, while parabolic and exponential methods require 3 points to establish the interpolation equation for spectral data interpolation.

[0094] Example 4

[0095] Based on Example 1, in S3, when the equalized spectral data is interpolated and synthesized in the depth dimension, the starting depth and ending depth of all single-point measurement data are selected during interpolation. The starting depth is the minimum depth of all single-point measurements, and the ending depth is the maximum depth of all single-point measurements. The sampling interval for depth dimension synthesis is set. The sampling interval can be set to a commonly used interval value, and it is recommended to match it with other logging curves, such as 0.1 meters. The maximum spacing during interpolation is set. When the depth of two adjacent single-point data exceeds the maximum spacing, the data of this segment is not interpolated, and the spectral data of this segment is all set to 0, i.e., a blank segment. When the depth interval of two single-point measurement data exceeds 10 meters, the segment between the depths of these two single-point data is not interpolated.

[0096] Example 5

[0097] This embodiment takes the signal synthesis of single-point spectrum data from noise logging at well Z in the X oilfield abroad as an example. It performs gain correction, equalization processing, and depth-dimensional interpolation synthesis on the spectrum data of each single point measurement, and provides a detailed description of a specific embodiment of the noise logging single-point signal synthesis method of the present invention.

[0098] The well noise logging depth range is 3800-3910 meters. Within this depth range, multiple water-absorbing layers and cross-channel leakage points were predicted before logging. Within a 110-meter well section, one noise measurement point was designed at 1-meter intervals along the depth, totaling 111 measurement points within the 3800-3910 meter depth range. The total noise curves and two-dimensional spectrum data for these 111 measurement points are stored in 111 separate noise logging files. Some individual point noise measurement data files are shown below. Figure 7 As shown.

[0099] Each noise single-point measurement yields one single-point measurement spectrum data. The 111 noise single-point measurement files for Well Z contain 111 noise spectrum data. The spectrum diagram of the single-point measurement data at 3866 meters is shown below. Figure 8 As shown, the spectrum of single-point measurement data at 3868 meters is as follows. Figure 9 As shown, the spectrum of single-point measurement data at 3899 meters is as follows. Figure 10 As shown, the spectrum of single-point measurement data at 3806 meters is as follows. Figure 11 As shown, the spectrum diagrams of the remaining single-point measurement data are similar.

[0100] From above Figure 8 , Figure 9 , Figure 10 and Figure 11It can be seen that the repeated sampling signals on the same channel in each time dimension of the spectrum diagram of each single point measurement of the noise logging in this well have good consistency. The sound amplitude on the same channel is basically the same. The color reflecting the change of sound amplitude changes very little or not at all on the same channel. It cannot reflect the depth and intensity of the leakage point of the channel, nor can it reflect the layer position and production intensity of the production layer. It cannot reflect the noise leakage and production characteristics of the production layer.

[0101] Gain correction was first performed on the spectral signals of 111 noise logging points in this well. The gain of each of the 111 point measurement data was 1024. According to the single-point measurement spectrum gain correction method of this invention, the gain correction coefficient for each measurement point of the 111 point measurement data was...

[0102] GainCo = GainBase / GainNow

[0103] GainCo - Gain correction product coefficient of the current stagnation point, GainBase - Maximum gain among all stagnation points (the maximum gain in this well is 1024), GainNow - Gain value of the current stagnation point. Since the gain of data from each measurement point in this well is 1024, the gain correction coefficient for the spectrum of each single-point measurement is GainCo = 1024 / 1024. That is, the gain correction coefficient for the spectrum of the 111 single-point measurement data in this well is 1. Therefore, the formula for the spectrum gain correction of each single-point data is:

[0104] NVDL ij =1×NVDL ′ ij

[0105] NVDL ij The spectrum after gain correction for the current depth single-point measurement, NVDL ′ ij The spectrum before gain correction is shown below. Taking the measurement spectrum at 3901 meters as an example, the time repetition sampling number is 652 times, i.e., i = 652. The frequency sampling range is 1kHz-12.8kHz. Starting from 1kHz, one amplitude is collected every 0.1kHz, for a total of 128 amplitudes collected in the frequency range, i.e., j = 128. The spectrum before gain correction is shown below. Figure 12 As shown, because the gain correction factor is 1, the spectrum after gain correction is the same as the spectrum before gain correction. The gain correction for the spectrum data of the remaining 110 measurement points is similar.

[0106] Using the single-point noise spectrum equalization processing technology of this invention, equalization processing is performed on 111 single-point measurement spectrum data after gain correction to obtain equalized noise spectrum. According to the technology of this invention, after analyzing the 111 single-point measurement spectrum data, it was found that the amplitude of 80 point measurement data samples at different times on the same channel was relatively consistent. Therefore, uniform equalization was used for these 80 single-point measurement spectrum data. The amplitude of 31 point measurement data samples at different times on the same channel was unstable, with significant fluctuations in amplitude and different colors on the spectrum. Therefore, enhanced equalization was used for these 31 single-point measurement spectrum data. After equalization processing, a total of 111 equalized single-point spectrum data were obtained.

[0107] Using the depth-dimensional interpolation synthesis technique of this invention, the 111 single-point spectral data points after equalization are interpolated and synthesized in the depth dimension. Linear interpolation synthesis is selected as the interpolation synthesis method. After equalization and depth-dimensional interpolation synthesis, the 111 single-point measurement spectral data points of this well yield continuous noise spectral data in the depth dimension, as follows: Figure 13 As shown.

[0108] from Figure 13 The noise spectrum equalization and depth-dimensional interpolation synthesis results from the 3800-3910 meter depth range of this well show that the equalized and interpolated spectrum data reflects the noise situation across the entire well section. This data can be used to analyze leakage points or the production and absorption conditions of producing layers, determining the specific depth of the leakage point or producing layer, which is crucial for subsequent measures, including sealing the leakage point. Further calculations can also determine the production intensity and quantity of each producing layer, and analyze the properties of the produced fluid. This well is a typical producing well; the logging objective is to locate the water production point and analyze the relative water production. From... Figure 13 The spectrum results after equalization and interpolation show a strong production signal at a depth of 3871 meters, indicating a major production layer with the highest water production. Three production signals are observed at depths of 3816.1 meters, 3865.9 meters, and 3868.1 meters, respectively. The strength of these three signals is lower than that at 3871 meters, and their relative water production is also lower. Figure 13 In the noise spectrum equalization and depth-dimensional interpolation synthesis result diagram, the noise spectrum signal within the depth range between the four output layer depths shows little variation in amplitude and a relatively uniform distribution, which is typical interference noise. From this embodiment, it can be concluded that...

[0109] The depth-dimensional continuous spectrum synthesized by the technology of this invention can not only reflect the leakage point of the channel, but also reflect the depth of the generation and absorption layer and the relative generation and absorption amount. Further analysis and research on the equalization and interpolation synthesized spectrum data can calculate the relative generation and absorption intensity of the generation and absorption layer, i.e., the sound source point. Combined with the total output and absorption amount, the absolute output and absorption amount of the generation and absorption layer can be calculated.

Claims

1. A method for synthesizing single-point signals in noise logging, characterized in that, The specific steps are as follows: S1, calculate the gain correction coefficient of the single-point measurement data and perform gain correction on the measurement point spectrum; S2, perform uniform equalization and enhancement equalization on the spectrum of the measurement points after gain correction in S1, to obtain the equalized spectrum of each single point data. S3 uses an interpolation algorithm to interpolate and synthesize the equalized spectral data from S2 in the depth dimension, resulting in continuous spectral data in the depth dimension.

2. The method for synthesizing single-point noise logging signals according to claim 1, characterized in that, Specifically, S1 involves: using the gain of each single-point measurement data within the measurement depth range and the maximum gain of all single-point measurement data, calculating the gain correction coefficient of the single-point measurement data, and using the gain correction coefficient to perform gain correction on the measurement point spectrum.

3. The method for synthesizing single-point noise logging signals according to claim 2, characterized in that, Specifically, S1 involves: performing gain correction on the measurement data of each single point within the measurement depth range, using the maximum gain of all single-point measurement data within the measurement depth range as the base value for gain correction, with the correction coefficient being: GainCo = GainBase / GainNow (1) In equation (1), GainCo is the gain correction product coefficient of the current stationary point, GainBase is the maximum gain among all stationary points, and GainNow is the gain value of the current stationary point; The spectral gain correction for each single-point data is shown in equation (2): NVDL ij =GainCo k ×NVDL′ ij (2) In equation (2), NVDL ij GainCo presents the spectrum after gain correction following a single-point measurement at the current depth. k NVDL is the gain correction factor for the current single-point depth measurement. ′ ij Let i ∈ (1, ..., m), where m is the number of repeated samples of the single-point measurement spectrum, j ∈ (1, ..., n), where n is the number of frequency samples of the single-point measurement spectrum, with 1 amplitude sampled at each frequency, and n is also the number of amplitudes in the same sample, k ∈ (1, ..., N), where N is the number of single-point measurements within the measurement depth range.

4. The method for synthesizing single-point noise logging signals according to claim 2, characterized in that, The uniform equalization in S2 specifically involves calculating the average amplitude across all channels of the entire single-point spectrum data to obtain the n average amplitude values ​​of the single-point spectrum data, denoted as NVDLAVG. j If j∈(1,...,n), then: Removing the s1 repeated sampling points at the start time and the s2 repeated sampling points at the end time, which are not included in the calculation of the average value of the single-point spectrum data, we get: In equation (4), s1 and s2 take different values ​​depending on the specific spectral data of each single point measurement. When the single point measurement starts to repeat the measurement, the amplitude of the repeated measurement samples of s1 increases or decreases, and the color displayed on the spectrum changes. At this time, s1 is the number of sampling points whose color changes when the repeated measurement starts on the spectrum. s2 is the number of invalid sampling points of the repeated measurement when the measurement ends.

5. The method for synthesizing single-point noise logging signals according to claim 2, characterized in that, The enhanced equalization in S2 specifically refers to: Calculate the average signal value for each channel at all sampling points: In equation (5), j is the channel number and m is the number of repeated sampling points in the selected time. Calculate the standard deviation of each sampling point: In equation (6), n is the number of channels; The formula for calculating enhanced equalization is obtained by adding the average amplitude of each channel to the standard deviation of all sampled amplitudes of that channel: Substituting the average signal value and standard deviation of each channel into the above formula, we obtain the formula for calculating the single-point spectrum data after enhanced equalization: In equation (8), j is the channel number, m is the number of repeated sampling points at the selected time, n is the number of channels, s1 is the repeated sampling point at the start time that was removed, and s2 is the repeated sampling point at the end time that was removed.

6. The method for synthesizing single-point noise logging signals according to claim 2, characterized in that, The interpolation method in S3 is Lagrange interpolation algorithm, linear interpolation algorithm, parabolic interpolation algorithm, exponential interpolation algorithm, or cubic spline function interpolation algorithm.

7. The method for synthesizing single-point noise logging signals according to claim 6, characterized in that, In step S3, when the equalized spectral data is interpolated and synthesized in the depth dimension, the starting depth and ending depth of all single-point measurement data are selected during interpolation. The starting depth is the minimum depth of all single-point measurements, and the ending depth is the maximum depth of all single-point measurements. The sampling interval for depth dimension synthesis is set, and the maximum spacing during interpolation is set. When the depth of two adjacent single-point data exceeds the maximum spacing, no interpolation processing is performed on the data of this segment, and all spectral data of this segment is set to 0, i.e., a blank segment. When the depth interval of two single-point measurement data exceeds 10 meters, no interpolation processing is performed on the segment between the depths of these two single-point data.

Images