Full-frequency-domain hydrocarbon detection method and system

Abstract

The invention discloses a full-frequency-domain hydrocarbon detection method and system. The method comprises the following steps: acquiring full-frequency-domain hydrocarbon detection data through seismic data and underground data; training a hydrocarbon detection model based on the hydrocarbon detection data of the full frequency domain, and detecting a thin oil layer through hydrocarbon detection to obtain a detection result; calculating evaluation parameters of the hydrocarbon detection model, evaluating and analyzing the detection result by adopting the evaluation parameters, and judging the detection precision of the hydrocarbon detection model; and optimizing and training the hydrocarbon detection model to improve the detection precision of the hydrocarbon detection model. By detecting the thin oil layer, the development efficiency and economic benefits of an oil field are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of oilfield development, and in particular to a full-frequency-domain hydrocarbon detection method and system. Background Art

[0002] Accurately identifying thin oil layers is a crucial step in oilfield development. However, due to the inherent characteristics of thin oil layers, traditional hydrocarbon detection techniques often face difficulties in identification. This results in insufficient identification accuracy, which in turn impacts oilfield development efficiency. Therefore, optimizing hydrocarbon detection methods to improve the accuracy of thin oil layer identification, thereby further enhancing oilfield development efficiency and economic benefits, has become a major challenge facing oilfield development. Summary of the Invention

[0003] The present invention aims to provide a full-frequency hydrocarbon detection method and system for detecting thin oil layers, thereby improving the development efficiency and economic benefits of oil fields.

[0004] To achieve the above objectives, an embodiment of the present invention discloses a full-frequency-domain hydrocarbon detection method, comprising:

[0005] Obtain full-frequency domain hydrocarbon detection data through seismic data and downhole data;

[0006] Training a hydrocarbon detection model based on the hydrocarbon detection data in the full frequency domain, and detecting thin oil layers through the hydrocarbon detection to obtain detection results;

[0007] Calculating evaluation parameters of the hydrocarbon detection model, evaluating and analyzing the detection results using the evaluation parameters, and determining the detection accuracy of the hydrocarbon detection model;

[0008] The hydrocarbon detection model is optimized and trained to improve the detection accuracy of the hydrocarbon detection model.

[0009] Furthermore, full-frequency domain hydrocarbon detection data is obtained through seismic data and downhole data, including:

[0010] The seismic data and downhole data are processed by using spectral decomposition technology, and the processed seismic data and downhole data are used to perform full-frequency domain hydrocarbon detection to obtain the full-frequency domain hydrocarbon detection data.

[0011] Furthermore, training a hydrocarbon detection model based on the full-frequency-domain hydrocarbon detection data includes:

[0012] Based on the hydrocarbon detection data in the full frequency domain, extracting feature recognition parameters using a neural network, and training the hydrocarbon detection model using the feature recognition parameters;

[0013] The characteristic recognition parameters include: frequency attenuation gradient, low-frequency energy, frequency corresponding to a specified energy ratio, and energy ratio of a specified frequency segment.

[0014] Furthermore, the evaluation parameters include accuracy, recall, precision, F1 score and AUC-ROC;

[0015] Among them, F1 score is the harmonic mean of precision and recall, and AUC-ROC is the area under the receiver operating characteristic curve.

[0016] Furthermore, the evaluation parameters of the hydrocarbon detection model are calculated, and the detection results are evaluated and analyzed using the evaluation parameters to determine the detection accuracy of the hydrocarbon detection model, including:

[0017] Using AI technology, the detection results of the hydrocarbon detection model are displayed in the form of graphics or images;

[0018] Analyzing the decision-making process of the hydrocarbon detection model and understanding the detection process of the hydrocarbon detection model;

[0019] Based on the detection results, evaluation parameters of the hydrocarbon detection model are calculated, and the detection performance and accuracy of the hydrocarbon detection model are judged by the evaluation parameters, so as to further optimize the hydrocarbon detection model.

[0020] Furthermore, the hydrocarbon detection model is optimized and trained to improve the recognition accuracy of the hydrocarbon detection model, including:

[0021] Based on the comparison and analysis of drilling test and oilfield production data and extracted near-well seismic attribute characteristics, the extracted feature recognition parameters are optimized; based on the optimized feature recognition parameters, the hydrocarbon detection model is optimized and trained to obtain the optimized hydrocarbon detection model;

[0022] The back propagation and optimization algorithms are used to further optimize the selected feature recognition parameters, and the hydrocarbon detection model is trained again based on the optimized feature recognition parameters to obtain the final hydrocarbon detection model, so as to improve the detection accuracy of the hydrocarbon detection model for thin oil layers.

[0023] Based on the same inventive concept, an embodiment of the present invention further discloses a full-frequency-domain hydrocarbon detection system, comprising:

[0024] An acquisition unit, configured to acquire hydrocarbon detection data in the full frequency domain through seismic data and downhole data;

[0025] a detection unit, configured to train a hydrocarbon detection model based on the hydrocarbon detection data in the full frequency domain, and to detect the thin oil layer through the hydrocarbon detection to obtain a detection result;

[0026] an evaluation unit, configured to calculate evaluation parameters of the hydrocarbon detection model, evaluate and analyze the detection results using the evaluation parameters, and determine the detection accuracy of the hydrocarbon detection model;

[0027] The optimization unit is used to optimize and train the hydrocarbon detection model to improve the detection accuracy of the hydrocarbon detection model.

[0028] Based on the same inventive concept, an embodiment of the present invention further provides an electronic device, comprising: a memory and a processor; the processor is configured to read and execute a computer program stored in the memory to implement the aforementioned full-frequency domain hydrocarbon detection method.

[0029] Based on the same inventive concept, an embodiment of the present invention further provides a computer storage medium, wherein the computer storage medium stores computer executable instructions, and when the computer executable instructions are executed, the aforementioned full-frequency domain hydrocarbon detection method is implemented.

[0030] The technical effects and advantages of the present invention are as follows: The present invention fully utilizes the full frequency spectrum of seismic data to capture more details and features of underground geological structures. By analyzing a wider frequency range, it can provide higher data resolution, which is more accurate in revealing complex geological structures and identifying fluid properties. It can also provide a more comprehensive assessment of physical properties such as formation porosity, fracture density, and fluid saturation.

[0031] The present invention combines full-frequency hydrocarbon detection with artificial intelligence methods to provide more comprehensive and accurate data analysis capabilities than traditional geostatistical inversion, and performs particularly well in handling complex geological environments and refined oil and gas exploration.

[0032] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0034] Figure 1 This is a flow chart of a full-frequency-domain hydrocarbon detection method according to an embodiment of the present invention;

[0035] Figure 2This is an example diagram of the hydrocarbon detection principle according to an embodiment of the present invention;

[0036] Figure 3 This is an example diagram of the frequency attenuation gradient according to an embodiment of the present invention;

[0037] Figure 4 This is an example diagram of low-frequency energy according to an embodiment of the present invention;

[0038] Figure 5 This is an example diagram of frequencies corresponding to specified energy ratios according to an embodiment of the present invention;

[0039] Figure 6 This is an example diagram of energy ratios corresponding to specified frequencies according to an embodiment of the present invention;

[0040] Figure 7 5 examples of evaluation parameters according to an embodiment of the present invention;

[0041] Figure 8 This is a schematic structural diagram of a full-frequency-domain hydrocarbon detection system according to an embodiment of the present invention;

[0042] Figure 9 The figure is a schematic structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0044] To address the deficiencies of the prior art, the present invention discloses a full-frequency hydrocarbon detection method. Figure 1 As shown, the following steps are included:

[0045] Step S1: Acquire hydrocarbon detection data in the full frequency domain through seismic data and downhole data; specifically including:

[0046] The seismic data and downhole data are processed by using spectral decomposition technology, and the processed seismic data and downhole data are used to perform full-frequency domain hydrocarbon detection to obtain the full-frequency domain hydrocarbon detection data.

[0047] like Figure 2 As shown, the principles of the spectrum decomposition method and hydrocarbon detection technology are:

[0048] The principle of reflection wave seismic exploration indicates that seismic wave signals gradually attenuate as they propagate through underground media. Many factors influence seismic signal attenuation, primarily reflection mechanisms at adjacent lithofacies interfaces, faults, and fractures, spherical diffusion within the same-phase medium, and changes in physical properties within the same-phase medium (such as the presence of oil, gas, and water). This study focuses on the last of these attenuation factors, namely, the attenuation of seismic signals caused by changes in physical properties within the same-phase medium.

[0049] Spectral decomposition technology provides a means for analyzing seismic wave attenuation in the frequency domain. Spectral decomposition technology can decompose seismic signals into different frequency components, thus obtaining seismic signals in the full frequency domain. Generally speaking, in the high-frequency band, under the same geological background conditions, the presence of oil and gas increases the energy attenuation of the seismic signal. Compared with the frequency domain characteristics without attenuation, the entire frequency band after attenuation will shrink toward the low-frequency band. Energy attenuation is often indicated by physical parameters (i.e., characteristic identification parameters) such as the energy attenuation gradient with frequency, low-frequency energy, the frequency corresponding to a specified energy ratio, and the energy ratio of a specified frequency band. Different physical parameters reflect the possibility of the presence of oil and gas from different perspectives.

[0050] Step S2: training a hydrocarbon detection model based on the hydrocarbon detection data in the full frequency domain, and detecting thin oil layers through the hydrocarbon detection to obtain detection results.

[0051] The method of training a hydrocarbon detection model based on the hydrocarbon detection data in the full frequency domain includes:

[0052] Based on the hydrocarbon detection data in the full frequency domain, extracting feature recognition parameters using a neural network, and training the hydrocarbon detection model using the feature recognition parameters;

[0053] The characteristic recognition parameters include: frequency attenuation gradient, low-frequency energy, frequency corresponding to a specified energy ratio, and energy ratio of a specified frequency segment.

[0054] like Figure 3 As shown in the figure, the frequency attenuation gradient is one of the attenuation properties, which shows how the energy of seismic waves in the high-frequency band changes with frequency. The frequency attenuation gradient can indicate how quickly seismic waves attenuate during propagation. In addition to the diffusion effect of seismic waves propagating in a single-phase medium and the reflection mechanism of seismic waves at the reflection interface of a multi-phase medium, the attenuation gradient value (ATN_GRT) increases if there are attenuation factors such as oil and gas. Figure 3 Where ATN_GRT represents the attenuation gradient, ΔE represents the change in energy (amplitude), and Δf represents the change in frequency.

[0055] like Figure 4As shown in Figure 2, low-frequency energy is another important attenuation property, indicating its intensity. Due to the presence of oil and gas, the attenuation of high-frequency energy in seismic waves is greater than that of low-frequency energy. After attenuation, the entire frequency band shrinks toward the low-frequency band, causing the low-frequency energy to increase.

[0056] like Figure 5 As shown, the third geophysical parameter indicating the attenuation properties of oil and gas is the frequency corresponding to the specified energy ratio. The total energy of the effective frequency band is 1. If the energy of the specified low frequency band is 85%, the frequency corresponding to when the energy reaches 85% can be searched within the effective frequency band starting from the starting frequency of the effective frequency band. This is called F1; if the energy of the specified low frequency band is 65%, the searched frequency is called F2. If there are attenuation factors such as oil and gas, the frequency corresponding to the specified energy ratio will become smaller. According to the same principle, the frequency corresponding to when the energy reaches the maximum value can be searched within the entire effective frequency band, that is, the starting attenuation frequency. The smaller the starting attenuation frequency, the greater the possibility of the existence of oil and gas. Figure 5 and Figure 6 In the equation, E represents energy, and F and f both represent frequency.

[0057] like Figure 6 As shown in Figure 2, another geophysical parameter indicating the oil and gas attenuation property is the energy ratio corresponding to a specified frequency band ( Figure 5 If a frequency F is specified within the effective frequency band, the energy from the effective band's starting frequency to that frequency F can be calculated, known as the low-frequency energy (E low ); the percentage of the total energy of the entire frequency band (E full ) is known as the energy ratio corresponding to the specified frequency. If attenuation factors such as oil and gas are present, the energy in the high-frequency band will decrease, increasing the ratio of the low-frequency energy to the total energy at the specified frequency. In other words, if the energy ratio corresponding to the specified frequency increases, the likelihood of oil and gas being present increases. The likelihood of oil and gas being present can be determined by changes in the energy ratio corresponding to the specified frequency.

[0058] Step S3: Calculating the evaluation parameters of the hydrocarbon detection model, using the evaluation parameters to evaluate and analyze the detection results, and determining the detection accuracy of the hydrocarbon detection model; specifically including:

[0059] By using AI technology, the detection results of the hydrocarbon detection model are displayed in the form of graphics or images, so that the spatial distribution of hydrocarbons can be quickly understood;

[0060] Analyze the decision-making process of the hydrocarbon detection model, understand the detection process of the hydrocarbon detection model, and understand the advantages and disadvantages of the model;

[0061] Based on the detection results, the evaluation parameters of the hydrocarbon detection model are calculated, the detection performance and accuracy of the hydrocarbon detection model are judged by the evaluation parameters, key influencing factors are analyzed, and corresponding countermeasures are proposed to further optimize the hydrocarbon detection model.

[0062] Among them, such as Figure 7 As shown, the evaluation parameters include accuracy, recall, precision, F1 score and AUC-ROC;

[0063] Precision refers to the ratio of correctly detected samples to the total number of samples. In a binary classification problem, precision is the number of true positives (TP) and true negatives (TN) divided by the total number of samples. Precision is a fundamental metric for measuring the overall detection capability of a hydrocarbon detection model.

[0064] Recall, also known as sensitivity or true positive rate, is the ratio of the number of positive samples correctly identified by a hydrocarbon detection model to the total number of actual positive samples. In a binary classification problem, recall is equal to the number of true positives (TP) divided by the total number of actual positive samples (i.e., TP + FN). Recall is an important metric for measuring a hydrocarbon detection model's ability to detect positive samples.

[0065] Precision refers to the ratio of the number of positive samples correctly identified by a hydrocarbon detection model to the total number of samples identified as positive by the model. In a binary classification problem, precision is equal to the number of true positives (TP) divided by the number of samples predicted as positive by the model (i.e., TP + FP). Precision is a key metric for measuring the accuracy of hydrocarbon detection models in detecting positive samples.

[0066] The F1 score is the harmonic mean of precision and recall, which considers both precision and recall. The F1 score ranges from 0 to 1, with higher values ​​indicating better hydrocarbon detection model performance. The F1 score is an important metric for measuring the overall performance of hydrocarbon detection models.

[0067] AUC-ROC refers to the area under the receiver operating characteristic curve (AUC). The ROC curve is plotted with the false positive rate (FPR) on the horizontal axis and the true positive rate (TPR) on the vertical axis. AUC values ​​range from 0 to 1, with higher values ​​indicating better model performance. AUC-ROC is a comprehensive indicator that measures the performance of hydrocarbon detection models at different thresholds.

[0068] Step S4: Optimizing and training the hydrocarbon detection model to improve the detection accuracy of the hydrocarbon detection model. Specifically including:

[0069] Based on actual production data and wellside seismic measurement results, the extracted feature recognition parameters are optimized; based on the optimized feature recognition parameters, the hydrocarbon detection model is optimized and trained to obtain the optimized hydrocarbon detection model; wherein, the hydrocarbon detection model is a deep learning model of a convolutional neural network, a recurrent neural network or other applicable neural networks.

[0070] The back propagation and optimization algorithms are used to further optimize the selected feature recognition parameters, and the hydrocarbon detection model is trained again based on the optimized feature recognition parameters to obtain the final hydrocarbon detection model, so as to improve the detection accuracy of the hydrocarbon detection model for thin oil layers.

[0071] After drilling and completion of a new well, testing is performed before formal production to assess individual well production and determine an optimal production schedule. Actual production data is then determined through field statistics. Wellside seismic measurements are essentially seismic attributes extracted along the wellbore. Attributes that closely match test data and production data are selected as the basis for further hydrocarbon detection. For example, if a seismic attribute extracted along the wellbore corresponds to a high-yield interval in both testing and production, it can be considered sensitive to oil and gas activity.

[0072] In the embodiment, the frequency attenuation gradient and low-frequency energy attributes are more sensitive to the oil and gas content of the formation, have a certain correlation with the initial production of the production wells, better match the initial production of the production wells in the work area, and are ultimately selected as the prediction basis for hydrocarbon detection.

[0073] Based on the same inventive concept, the embodiment of the present invention also provides a full-frequency hydrocarbon detection system, such as Figure 8 As shown, including:

[0074] An acquisition unit, configured to acquire hydrocarbon detection data in the full frequency domain through seismic data and downhole data;

[0075] a detection unit, configured to train a hydrocarbon detection model based on the hydrocarbon detection data in the full frequency domain, and to detect the thin oil layer through the hydrocarbon detection to obtain a detection result;

[0076] an evaluation unit, configured to calculate evaluation parameters of the hydrocarbon detection model, evaluate and analyze the detection results using the evaluation parameters, and determine the detection accuracy of the hydrocarbon detection model;

[0077] The optimization unit is used to optimize and train the hydrocarbon detection model to improve the detection accuracy of the hydrocarbon detection model.

[0078] Regarding the system in the above embodiment, the specific manner in which each unit module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0079] Based on the same inventive concept, an embodiment of the present invention further provides an electronic device, the structure of which is as follows: Figure 9 As shown, it includes: a memory and a processor, wherein the processor is used to read and execute the computer program stored in the memory to implement the above-mentioned full-frequency domain hydrocarbon detection method.

[0080] Based on the same inventive concept, an embodiment of the present invention further provides a computer storage medium, wherein the computer storage medium stores computer executable instructions, and when the computer executable instructions are executed, the aforementioned full-frequency domain hydrocarbon detection method is implemented.

[0081] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A full-frequency hydrocarbon detection method, characterized in that: include: Obtain full-frequency domain hydrocarbon detection data through seismic data and downhole data; Training a hydrocarbon detection model based on the hydrocarbon detection data in the full frequency domain, and detecting thin oil layers through the hydrocarbon detection to obtain detection results; Calculating evaluation parameters of the hydrocarbon detection model, evaluating and analyzing the detection results using the evaluation parameters, and determining the detection accuracy of the hydrocarbon detection model; The hydrocarbon detection model is optimized and trained to improve the detection accuracy of the hydrocarbon detection model.

2. A full frequency domain hydrocarbon detection method according to claim 1, characterized in that: Obtain full-frequency hydrocarbon detection data from seismic and downhole data, including: The seismic data and downhole data are processed by using spectral decomposition technology, and the processed seismic data and downhole data are used to perform full-frequency domain hydrocarbon detection to obtain the full-frequency domain hydrocarbon detection data.

3. A full frequency domain hydrocarbon detection method according to claim 1 or 2, characterized in that: Training a hydrocarbon detection model based on the full-frequency-domain hydrocarbon detection data includes: Based on the hydrocarbon detection data in the full frequency domain, extracting feature recognition parameters using a neural network, and training the hydrocarbon detection model using the feature recognition parameters; The characteristic recognition parameters include: frequency attenuation gradient, low-frequency energy, frequency corresponding to a specified energy ratio, and energy ratio of a specified frequency segment.

4. The full-frequency-domain hydrocarbon detection method according to claim 1, characterized in that: The evaluation parameters include accuracy, recall, precision, F1 score and AUC-ROC; Among them, F1 score is the harmonic mean of precision and recall, and AUC-ROC is the area under the receiver operating characteristic curve.

5. The full-frequency-domain hydrocarbon detection method according to claim 1, characterized in that: Calculating evaluation parameters of the hydrocarbon detection model, evaluating and analyzing the detection results using the evaluation parameters, and determining the detection accuracy of the hydrocarbon detection model, including: Using AI technology, the detection results of the hydrocarbon detection model are displayed in the form of graphics or images; Analyzing the decision-making process of the hydrocarbon detection model and understanding the detection process of the hydrocarbon detection model; Based on the detection results, evaluation parameters of the hydrocarbon detection model are calculated, and the detection performance and accuracy of the hydrocarbon detection model are judged by the evaluation parameters, so as to further optimize the hydrocarbon detection model.

6. The full-frequency-domain hydrocarbon detection method according to claim 3, characterized in that: Optimizing and training the hydrocarbon detection model to improve the recognition accuracy of the hydrocarbon detection model includes: Based on the comparison and analysis of drilling test and oilfield production data and extracted near-well seismic attribute characteristics, the extracted feature recognition parameters are optimized; based on the optimized feature recognition parameters, the hydrocarbon detection model is optimized and trained to obtain the optimized hydrocarbon detection model; The back propagation and optimization algorithms are used to further optimize the selected feature recognition parameters, and the hydrocarbon detection model is trained again based on the optimized feature recognition parameters to obtain the final hydrocarbon detection model, so as to improve the detection accuracy of the hydrocarbon detection model for thin oil layers.

7. A full-frequency hydrocarbon detection system, characterized in that: include: An acquisition unit, configured to acquire hydrocarbon detection data in the full frequency domain through seismic data and downhole data; a detection unit, configured to train a hydrocarbon detection model based on the hydrocarbon detection data in the full frequency domain, and to detect the thin oil layer through the hydrocarbon detection to obtain a detection result; an evaluation unit, configured to calculate evaluation parameters of the hydrocarbon detection model, evaluate and analyze the detection results using the evaluation parameters, and determine the detection accuracy of the hydrocarbon detection model; The optimization unit is used to optimize and train the hydrocarbon detection model to improve the detection accuracy of the hydrocarbon detection model.

8. An electronic device, characterized in that: include: Memory, processor; The processor is used to read and execute the computer program stored in the memory to implement the full-frequency domain hydrocarbon detection method described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which, when executed, implement the full-frequency-domain hydrocarbon detection method according to any one of claims 1 to 6.

Images