An abnormality elimination method of horizon based on dominant frequency band
By employing a layer anomaly removal method based on advantageous frequency bands, the problem of skipping points in automatic layer tracking was solved, achieving high-precision layer tracking and improving the efficiency and accuracy of seismic data interpretation.
Patent Information
- Application Number
- CN202111551352.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-12-17
- Publication Date
- 2025-12-09
- Estimated Expiration
- 2041-12-17
AI Technical Summary
Existing technologies suffer from skipping points in automatic layer tracking in areas with sparse seed point density and complex target layer wave group characteristics, resulting in low interpretation accuracy and making it difficult to meet the interpretation needs of massive seismic data.
By employing a method for eliminating stratigraphic anomalies based on advantageous frequency bands, including data filtering, encrypted stratigraphy, stratigraphic snap correction, outlier removal, and fine automatic tracking, and combining instantaneous phase cosine value calculation to identify and delete jump point areas, the accuracy of seed points is improved, ultimately achieving high-precision stratigraphic tracking.
It significantly improves the accuracy of automatic horizon tracking, enhancing the precision and accuracy of horizon tracking and meeting the need for efficient interpretation of massive amounts of seismic data.
Smart Images
Figure CN116338780B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application belongs to the technical field of seismic data processing, and relates to a horizon automatic tracking method based on a seismic data processing algorithm, in particular to a horizon abnormality elimination method based on a dominant frequency band. BACKGROUND
[0002] With the development of wide-azimuth, wide-band, high-density and multi-component seismic exploration technology, the obtained seismic data area is larger and larger, the data volume is more and more, and the carried geological information is richer. Geophysical interpretation is both an opportunity and a challenge. The traditional seismic horizon interpretation is mainly artificial, and the efficiency is low, which is difficult to meet the interpretation demand of such mass data. Therefore, geophysicists propose more new technologies and methods to improve the accuracy and efficiency of geophysical interpretation. Automatic identification and reasonable processing of data from seismic information is a key link for the progress of seismic interpretation technology.
[0003] GeoEast software independently developed by East Geophysical Company of PetroChina integrates and is compatible with geophysical acquisition, processing and interpretation modules and technologies. The most typical technology for improving seismic interpretation efficiency is horizon automatic tracking (GeoAutoTrackHorizon). The basic principle is to realize horizon automatic matching tracking based on seed point stratum event (or seismic reflection interface) feature recognition (as shown in Figure 1a The accuracy of the horizon tracked by the method is often related to the seed point density and event continuity: ① the denser the seed layer, the higher the tracking horizon accuracy; ② the better the event continuity, the higher the tracking horizon accuracy. However, in the seed layer density sparse and target layer wave group feature complex blocks, the automatic tracking horizon still has the problem of jumping points, and the low interpretation accuracy defect cannot be overcome.
[0004] Two horizon correction methods are commonly used for horizon jumping point problems: ① Horizon Smoothing, which smoothes the position relationship between seed points to narrow the displacement gap between adjacent seed points (as shown in Figure 1b Although the displacement of the wrong seed point is corrected, the wrong seed point cannot be fundamentally identified and solved; ② Horizon Snap, which is a method of returning seed points to the set wave group feature position according to the set event features (wave peak, wave trough, zero point, etc.) of the horizon (as shown in Figure 1c The accuracy of this method is related to the target layer wave group feature: the stronger the amplitude, the better the homing effect; on the contrary, the target layer wave group amplitude is weak, and the correction effect is poor. SUMMARY
[0005] In order to solve the above problems in the prior art, the application aims to provide a layer position abnormality removing method based on dominant frequency bands, which establishes high-density seed points, increases layer position skip point identification and removing functions, improves seed point accuracy, and automatically tracks again to achieve the purpose of accurate layer position tracking.
[0006] In order to achieve the above-mentioned purpose, the technical scheme adopted by the application is as follows:
[0007] A layer position abnormality removing method based on dominant frequency bands, which comprises the following steps performed in sequence:
[0008] S1. Data filtering
[0009] The main frequency of the data body is reduced to obtain a dominant data body;
[0010] S2. Encryption of layer position
[0011] The seed points of the target layer are encrypted based on the dominant data body to obtain an encrypted layer position;
[0012] S3. Layer position snap correction
[0013] The snap correction is performed on the encrypted layer position, so that most of the seed points in the encrypted layer position coincide with the reflection events of the target layer, and only local skip points exist;
[0014] S4. Layer position abnormal value removing
[0015] The instantaneous phase cosine value of the seed points in the corrected encrypted layer position is calculated to determine a skip point area, and the layer position in the skip point area is deleted to obtain a layer position alpha;
[0016] S5. Fine automatic tracking
[0017] The layer position alpha is taken as a seed point to perform automatic tracking of the layer position in the dominant frequency band data to obtain a high-precision layer position beta;
[0018] S6. Layer position fine adjustment
[0019] The high-precision layer position beta is substituted into the data body in S1 to perform layer position fine adjustment to obtain a final layer position gamma.
[0020] As a limitation, in step S1, the main frequency is reduced by filtering, and the main frequency of the data of the target layer after processing is 15-24 Hz.
[0021] As another limitation, in step S2, the target layer is a geological interface or an oil and gas or mineral top interface determined by drilling information and logging information.
[0022] The encryption is performed by an automatic tracking method or an interpolation method.
[0023] As a third limitation, in step S4, the formula for calculating the instantaneous phase cosine value is:
[0024] cos(θ) = ΔX / (ΔX 2 + ΔZ 2 ) 1 / 2 Equation (I)
[0025] In equation (I), θ is the mutation angle of the horizon data, ΔX is the interval between adjacent horizon data in the main line direction, and ΔZ is the difference between the time values or depth values of adjacent horizon data in the vertical direction.
[0026] Or cos(θ) = ΔY / (ΔY 2 + ΔZ 2 ) 1 / 2 Equation (II)
[0027] In equation (II), θ is the mutation angle of the horizon data, ΔY is the interval between adjacent horizon data in the survey line direction, and ΔZ is the difference between the time values or depth values of adjacent horizon data in the vertical direction.
[0028] The jump point area is an area in which the cosine value of the jump point anomaly value is less than 0.53 in the work area.
[0029] After deleting the horizon in the jump point area, seed points need to be added to fill the blank blocks.
[0030] As a fourth limitation, in step S5, the seed point is consistent with the reflection axis with an accuracy of 100%, and the sum of the seed point density and the horizon anomaly point density in S4 is 1*1.
[0031] As a fifth limitation, in step S6, the horizon fine tuning is performed by setting a small time window for horizon homing.
[0032] Compared with the prior art, the beneficial effects obtained by the present application are:
[0033] The horizon anomaly elimination method based on the advantage frequency band provided by the present application constructs a horizon data anomaly value identification method, quickly and conveniently predicts the jump point area by calculating the instantaneous phase cosine attribute of adjacent horizon data based on the advantage frequency band, greatly improves the horizon automatic tracking accuracy, and further improves the accuracy of horizon tracking.
[0034] The present application is suitable for the horizon anomaly elimination process based on the advantage frequency band. BRIEF DESCRIPTION OF DRAWINGS
[0035] The present application will be described in further detail below with reference to the accompanying drawings and specific embodiments.
[0036] Figure 1 is a schematic diagram of the principle of the horizon tracking and correction method in the background of the present application, Figure 1a Figure 2 is a schematic diagram of the principle of the horizon tracking and correction method in the background of the present application, Figure 1b Figure 3 is a schematic diagram of the principle of the horizon snap, Figure 1c Figure 4 is a schematic diagram of the principle of the horizon smoothing,
[0037] Figure 5 is a comparison chart of the data filtering results in the embodiment of the present application, Figure 2a Figure 6 is a comparison chart of the horizon characteristics of the profile in each process in the embodiment of the present application, Figure 2b Figure 7 is a comparison chart of the horizon characteristics of the profile in each process in the embodiment of the present application, Figure 2c Figure 8 is a comparison chart of the horizon characteristics of the profile in each process in the embodiment of the present application,
[0038] Figure 3 Figure 9 is a schematic diagram of the principle of removing abnormal data by the instantaneous phase cosine value in the embodiment of the present application,
[0039] Figure 4 Figure 10 is a plan view of the instantaneous phase cosine attribute in the research area in the embodiment of the present application,
[0040] Figure 5 Figure 11 is a horizon plan view after removing abnormal data in the embodiment of the present application,
[0041] Figure 12 is a comparison chart of the horizon characteristics of the profile in each process in the embodiment of the present application, Figure 6a Figure 13 is a profile feature of the horizon tracked manually, Figure 6b Figure 14 is a profile feature of the horizon after horizon densification and snapping, Figure 6c Figure 15 is a profile feature of the horizon after removing the jump point abnormal value, Figure 6d Figure 16 is a profile feature of the horizon after fine tracking, Figure 6e Figure 17 is a profile feature of the horizon after fine tracking. DETAILED DESCRIPTION
[0042] The preferred embodiments of the present application will be described below with reference to the accompanying drawings, and it should be understood that the preferred embodiments described herein are only used to illustrate and understand the present application, and are not used to limit the present application.
[0043] The embodiment is an abnormal horizon removal method based on the dominant frequency band in the Honghaoershut Sag of the Eren Basin in Xilinhot City, Inner Mongolia. The target layer of the research area is buried at a depth of 600-3500 meters. The seismic data is affected by lateral heterogeneous lithology, and the amplitude intensity is low and the continuity is poor.
[0044] The embodiment is an abnormal horizon removal method based on the dominant frequency band in the Honghaoershut Sag of the Eren Basin in Xilinhot City, Inner Mongolia. The target layer of the research area is buried at a depth of 600-3500 meters. The seismic data is affected by lateral heterogeneous lithology, and the amplitude intensity is low and the continuity is poor.
[0045] Based on the actual situation of the study area, the embodiment is implemented in GeoEast software, which is a seismic data processing and interpretation collaborative integrated software system independently developed by China Petroleum East Geophysical Company.
[0046] The embodiment includes the following steps performed in sequence:
[0047] S1. Data filtering
[0048] Conventional filtering techniques are used to filter the data volume to improve the signal-to-noise ratio of the data, make the seismic reflection axis continuous and discontinuous features more obvious, and highlight the waveform differences between adjacent traces. The results are shown in FIG. 2, wherein, Figure 2a is the original profile with a frequency of 0-24Hz, Figure 2b is the frequency profile of 0-12Hz, Figure 2c is the frequency profile of 0-18Hz;
[0049] As can be seen from FIG. 2, the low-frequency data wave group is more continuous, but Figure 2b the profile and Figure 2a the phase axis form of the original profile changes, and the local actual horizon changes from a wave peak to a wave trough, while Figure 2c the overall profile is consistent with the original profile in form, so the advantage frequency band selection should be based on the premise of ensuring the consistency of the wave group form, and the data volume with the more obvious characteristics of the target layer wave group should be selected, that is, Figure 2c as the advantage data volume for subsequent tests, and the main frequency of the original data volume before filtering is 14-24Hz, and the main frequency of the advantage data volume after filtering is 14-18Hz;
[0050] S2. Encryption horizon
[0051] On the basis of the advantage data volume, the seed points of the target layer are encrypted using the automatic tracking method, but the encrypted horizon has problems such as jump points and inconsistency with the reflection axis. Through horizon snapping, the problem is attributed to the horizon jump point problem. Most of the horizon points in the snapped horizon are attributed to the reflection axis of the target layer, and the encrypted horizon and the jump points in the encrypted layer are obtained, as shown in Figure 6b ;
[0052] S3. Horizon outlier (jump point) rejection
[0053] On the basis of the advantage frequency band, the instantaneous phase cosine value of the adjacent point data is calculated. The jump point and the adjacent horizon have a sudden displacement, and the sudden displacement value is usually greater than one reflection axis phase height (20 milliseconds in time domain and 30 meters in depth domain). The sudden displacement is geometrically represented as a sudden change in dip angle. The instantaneous phase cosine value function of the adjacent (trace or line direction) data represents the phase change between points, as shown in Figure 3As shown in the calculation of the instantaneous phase cosine plane attribute of horizon, a certain direction can be selected to calculate the adjacent horizon points, such as the X direction (main line direction) or the Y direction (measuring line direction);
[0054] The formula for calculating the instantaneous phase cosine value is:
[0055] cos(θ) = ΔX / (ΔX 2 + ΔZ 2 ) 1 / 2 Formula (i)
[0056] In formula (i), θ is the mutation angle of horizon data, ΔX is the interval between adjacent horizon data in the main line direction, and ΔZ is the difference between the time value or depth value of adjacent horizon data in the vertical direction;
[0057] In this embodiment, θ is 0-72.5°, ΔX is 25m, and ΔZ is 0-79.5m, so that cos(θ) is 0.3-0.99;
[0058] As shown in the instantaneous phase cosine plane attribute in the main line direction, the cosine value of the jump point anomaly in the work area will suddenly change to 0.53 or less, such as Figure 4 As shown, so the attribute value less than 0.53 is defined as the jump point area;
[0059] The mutation point will cause the mutation of the cosine value, which is represented as a random point in the instantaneous phase cosine attribute plane. By defining the phase cosine value, the horizon in the jump point area is deleted, as shown in Figure 5 and Figure 6c The instantaneous phase cosine value near the fault will also be represented as a random point. During the data removal process, one horizon point on both sides of the fault will be deleted, but this does not affect the final tracking result of the horizon. If there is a large range of blank blocks after deletion, seed points need to be added to fill the blank blocks to obtain the horizon α;
[0060] S4. Fine automatic tracking
[0061] The horizon α is used as a seed point to automatically track the horizon in the dominant frequency band data to obtain a high-precision horizon β, as shown in Figure 6d As shown, the seed point and the event after automatic tracking of the horizon are consistent with 100% accuracy, and the sum of the seed point density and the S3 horizon anomaly point density is 1*1;
[0062] S5. Horizon fine tuning
[0063] The high-precision horizon β is always an intermediate horizon in the dominant frequency band, and needs to be snapped back (Horizon Snap) based on the original data. The Snap snapping window parameter is set to -30-30ms, and the horizon is fine-tuned to obtain the final horizon, as shown in Figure 6e .
[0064] As shown in Figure 6, through example comparison, the layers are encrypted ( Figure 6a ), classified as ( Figure 6b Outlier removal Figure 6c ), fine automatic tracking ( Figure 6d ) and fine-tuning ( Figure 6e The resulting final strata achieve the goal of accurate and automatic strata tracking. Furthermore, the strata obtained by this invention have a 100% accuracy in matching the seismic phase axis and a density of 1*1, which can be applied to different geophysical interpretation scenarios.
[0065] In this embodiment, interpolation can also be used in S2 to encrypt the seed points of the target layer;
[0066] The formula for calculating the instantaneous phase cosine value in S3 can also be:
[0067] cos(θ)=ΔY / (ΔY 2 +ΔZ 2 ) 1 / 2 ――――――――Form (ii)
[0068] In equation (ii), θ is the dip angle of the abrupt change in the stratigraphic data, ΔY is the spacing between adjacent stratigraphic data along the survey line direction, and ΔZ is the difference in time or depth values between adjacent stratigraphic data in the vertical direction.
[0069] This invention can also be applied to other layer tracking scenarios, such as: ① layer conversion of data volumes from different years within the same block; ② layer conversion of time-depth domain data volumes within the same block; ③ layer conversion of P-wave and S-wave data volumes within the same block; ④ quality checks of interpreted layers, etc.
Claims
1. A method for removing abnormal horizon based on dominant frequency band, characterized in that, The method comprises the following steps in sequence: S1. Data filtering The main frequency of the data body is reduced to obtain a dominant data body; S2. Encryption horizon The seed points of the target layer are encrypted based on the dominant data body to obtain an encrypted horizon; S3. Horizon snap correction The encrypted horizon is subjected to snap correction, so that most of the seed points in the encrypted horizon coincide with the event of the target layer, and only local points jump; S4. Horizon outlier removal The seed points in the corrected encrypted horizon are subjected to instantaneous phase cosine value calculation to determine a jump point area, and the horizon in the jump point area is deleted to obtain a horizon a; In step S4, the calculation formula of the instantaneous phase cosine value is: - Formula (I) In formula (i), θ is the mutation angle of the horizon data, ΔX is the distance between adjacent horizon data in the main line direction, and ΔZ is the difference between the time values or depth values of adjacent horizon data in the vertical direction; or Formula (ii) In formula (ii), θ is the mutation angle of the horizon data, ΔY is the distance between adjacent horizon data in the survey line direction, and ΔZ is the difference between the time values or depth values of adjacent horizon data in the vertical direction; The jump point area is an area in which the cosine value of the jump point anomaly point is less than 0.53 in the work area; After the horizon in the jump point area is deleted, the seed points need to be supplemented to fill the blank area; S5. Fine automatic tracking The horizon a is taken as a seed point to perform automatic tracking of the horizon in the dominant frequency band data to obtain a high-precision horizon β; S6. Horizon fine tuning The high-precision horizon β is substituted into the data body in S1 to perform horizon fine tuning to obtain a final horizon γ.
2. The method according to claim 1, wherein: In step S1, the main frequency is reduced by filtering, and the main frequency of the data of the target layer after processing is 15-24 Hz.
3. The method according to claim 1, wherein: In step S2, the target layer is a geological interface or an oil and gas or mineral top interface determined through drilling information and logging information. The encryption is performed by automatic tracking or interpolation.
4. The method according to any one of claims 1-3, wherein: In step S5, the coincidence accuracy of the seed point with the event is 100%, and the sum of the seed point density and the horizon anomaly point density in S4 is 1*1.
5. The method according to any one of claims 1-3, wherein: In step S6, the horizon fine tuning is performed by setting a time window for horizon homing.
6. The method according to claim 4, characterized in that: In step S6, the horizon fine tuning is performed by setting a time window for horizon homing.