Method for improving thin reservoir prediction precision under complex structure

Abstract

The invention relates to a method for improving thin reservoir prediction precision under a complex structure, and belongs to the technical field of oilfield development. The method comprises the following specific steps: finding out and determining a target layer influencing a reservoir; determining main frequency, frequency bandwidth and time window parameters; performing wavelet decomposition on the original seismic data, and removing redundant seismic response interference; carrying out standardization processing on the logging curve; carrying out sensitive curve analysis, and determining a wave impedance curve; constructing a wave impedance curve by adopting a wavelet transform-based reconstruction technology; seismic wavelets are extracted again, and then seismic synthetic records are made with the wave impedance curve; correcting the initial model in combination with earthquakes and wells so as to obtain an inversion result, and further identifying reservoir distribution conditions in the longitudinal direction and the transverse direction; and comparing a prediction result with a known drilling well, and comprehensively analyzing a reservoir prediction coincidence rate and a prediction precision improvement degree. According to the method, the resolution ratio of the inversion section is improved, and meanwhile, the reservoir boundary is depicted finely.

Description

Technical Field

[0001] The invention relates to a method for improving the prediction accuracy of thin reservoirs under complex structures, and belongs to the technical field of oil field development. Background Art

[0002] Predicting thin layers is a major difficulty in oil and gas seismic exploration. From the current development status of thin reservoir prediction technology, research methods are generally based on traditional wave impedance inversion, geostatistical inversion, multi-attribute fusion and other methods to carry out reservoir prediction, but conventional reservoir prediction does not consider the change law of wavelets. Moreover, the sensitivity of impedance to reservoirs is low, and the precision of prediction is greatly limited, which still cannot provide a reliable reference for subsequent research. The seismic data collected are usually affected by various factors such as seismic wave absorption and attenuation. Among them, thin reservoir wavelets can represent local seismic signal characteristics, but they must be different from other wavelets. Under complex structures, the amplitude of seismic response of thin reservoirs cannot effectively represent the real changes in reservoir development, which is mainly attributed to comprehensive problems such as burial depth, geological conditions, lateral change laws, and seismic data quality. Especially under the condition of limited resolution of seismic data, it is more difficult to improve the prediction accuracy of thin reservoirs. However, in the face of the growing demand for exploration and development, the search for benefit sweet spots and the research on how to improve the prediction accuracy of thin reservoirs under complex structures have a very long-term significance. Summary of the invention

[0003] In order to solve the above problems existing in the prior art, the present invention discloses a method for improving the prediction accuracy of thin reservoirs under complex structures, based on the reconstruction of seismic data using wavelet decomposition technology, combined with the optimization reconstruction of well logging sensitivity curves, to establish a method for improving the prediction accuracy of thin reservoirs based on wavelet decomposition technology and sensitivity curve optimization technology, and then combine the reservoir signal reconstructed by wavelet with the new advantageous reservoir curve to construct a thin reservoir prediction method suitable for complex structures. This reservoir prediction method can not only break the limitations of the selection method of thin reservoir prediction, but also optimize and enrich the details of the target layer, and make the sand body response of the thin reservoir clearer. At the same time, it takes advantage of the lateral continuity of the reservoir to further improve the vertical identification ability of the effective reservoir.

[0004] The technical solution adopted by the present invention is: a method for improving the prediction accuracy of thin reservoirs under complex structures, and the specific steps are as follows:

[0005] Step 1: Analyze the reflection characteristics of the original seismic data and the degree of its influence on the reservoir response characteristics, and qualitatively identify the thin reservoir development zone by combining the well-seismic response characteristics of the forward model; find and determine the target layer that affects the reservoir;

[0006] Step 2: Combined well-seismic fine calibration, analyze the seismic response characteristics of the target layer, and clarify the main frequency, frequency band width and time window parameters;

[0007] Step 3: Based on step 2, wavelet decomposition and reconstruction technology is used to weaken the influence of strong reflections above and below the reservoir on the reflection waves of the target reservoir, making the thin sand body more obvious;

[0008] Step 4: Standardize the logging curves and unify the same type of logging data to the same scale level;

[0009] Step 5: Comprehensively compare the logging curves related to the reservoir, conduct sensitivity curve analysis, and determine the curve that is most suitable for the target layer and can distinguish lithology;

[0010] Step 6: Using wavelet transform reconstruction technology, the two logging curves with the largest correlation coefficient are fitted into a relationship, and then transformed and reconstructed to construct an optimized wave impedance curve that is beneficial to reservoir prediction;

[0011] Step 7: Re-extract the seismic wavelet and then make a synthetic seismic record with the wave impedance curve;

[0012] Step 8: After analyzing the waveform characteristics of the seismic traces near the well, perform frequency decomposition in the wavelet domain, extract the common characteristics of this set of curves to form an isochronous grid model, and then modify the initial model in combination with seismic and well data under the constraints of the Bayesian framework to obtain the inversion results, thereby identifying the vertical and horizontal reservoir distribution;

[0013] Step 9: Compare the prediction results with known wells and comprehensively analyze the reservoir prediction compliance rate and the degree of improvement in prediction accuracy.

[0014] Furthermore, in step 1, the target layer having an impact on the reservoir refers to the event axis that can reflect the real seismic response characteristics of the reservoir.

[0015] Furthermore, in step 2, the calibration is mainly based on the principle of matching waveform, phase and energy characteristics. First, the marker layer that can be used for stratigraphic comparison is identified, and then the target reservoir is calibrated; after arbitrary well line inspection and repeated comparison, the stratigraphic layers with different lithology and electrical properties are linked to the reflection characteristics of the seismic profile to verify whether the calibration is accurate.

[0016] Furthermore, in step 3, the wavelet decomposition and reconstruction technology is specifically: based on the multi-wavelet seismic trace model, the seismic trace matching is decomposed into multiple wavelets with different main frequencies and shapes, and then selectively reconstructed to identify favorable reservoirs or fluid types. This process requires multiple experiments and continuous adjustment of parameters to achieve the best presentation effect.

[0017] Furthermore, in step 4, the logging curve is standardized to eliminate the influence of factors such as different instruments, scales, time spans, and operating methods on the authenticity and reliability of the logging curve; a variety of methods such as overlapping curve method, structural depth trend method, histogram method, trend surface method, etc. can be used for standardization.

[0018] In step 5, it is known that the natural gamma value reflects the amount of mud content and can be used to classify lithology, while the acoustic time difference, neutron porosity and density curves mainly reflect the quality of physical properties. After the intersection and analysis of different curves, the best sensitive curve can be determined.

[0019] In step 6, wavelet transform reconstruction technology is an effective method for analyzing non-stationary random signals, which can decompose seismic data into multi-layer, multi-frequency band signals, and finally obtain a result after multi-scale analysis; through wavelet transform reconstruction technology, the two logging curves with the largest correlation coefficient are fitted into a relationship, and then converted and reconstructed, an optimized curve that is beneficial to reservoir prediction can be constructed, that is, a wave impedance curve, and the optimization result is finally confirmed through sensitive curve intersection analysis. The wave impedance curve can be used to constrain the inversion model during the inversion process.

[0020] In step 7, since the accuracy of the layer calibration directly affects the subsequent prediction effect, it is necessary to ensure that the correlation between the actual seismic profile and the generated seismic synthetic record is maximized, and it is essential to readjust the seismic synthetic record.

[0021] In step 8, seismic waveform indication inversion is used, which is an improved high-precision inversion based on geostatistics and is performed on the basis of the wave impedance model. The input curve of the inversion can only be the wave impedance curve, that is, the curve after re-optimization and integration. The parameters in the inversion process must also be tested repeatedly. The size of the parameters required for the earthquake work area under different geological conditions depends on the actual situation, and finally the desired prediction effect is obtained.

[0022] Step 9 is the final verification of the use of the entire method. By comparing the predicted data with the known reservoir thickness of multiple wells that have been drilled, the size of the reservoir prediction compliance rate can be obtained, thereby verifying whether the reservoir prediction accuracy has been effectively improved through the application of this method, and whether the degree of improvement can meet actual production needs.

[0023] The present invention discloses a method for improving the prediction accuracy of thin reservoirs under complex structures, and its beneficial effect is to fully consider the seismic response characteristics of thin reservoirs and the characterization of logging curves, accurately start the two key links before inversion, and provide favorable support for subsequent inversion work. Combining effective seismic data bodies and favorable reservoir curves, a benign reservoir prediction environment is built. This method enriches the details of lithological description, not only improves the resolution of inversion profiles, but also finely depicts the reservoir boundaries, which has important guiding significance for thin reservoir prediction methods, thereby improving the prediction accuracy of thin reservoirs, accurately and truly reflecting the actual situation of the reservoir, and providing important references for subsequent fan characterization, closure evaluation, favorable zone division and other work. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0025] Figure 1 The figure shows the comparison of seismic sections of the target layer before and after wavelet decomposition technology;

[0026] Figure 2 Shown is the AC contour map after quadratic trend correction;

[0027] Figure 3 The figure shows the intersection of the optimized wave impedance curve and the GR curve;

[0028] Figure 4 Shown is a comparison of the inversion profiles using conventional wave impedance and the new method.

[0029] Specific implementation methods

[0030] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present invention and its application or use. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] In order to further understand the content of the present invention, the following is a detailed description of the process of improving the prediction accuracy of the bottom conglomerate thin reservoir in the study area by taking the X work area as an example in combination with the specific implementation method, so as to further illustrate the present invention. Unless otherwise specified, it is a conventional method. Example 1

[0032] like Figures 1 to 4 As shown in FIG. 1 , a method for improving the prediction accuracy of thin reservoirs under complex structures is shown in FIG. 1 . The specific steps are as follows:

[0033] Step 1: Prepare relevant seismic data, well information and logging data of the work area, and then combine the seismic profile and formation model for comprehensive analysis. It can be seen that the bottom conglomerate reservoir in the study area is relatively stable, but the burial depth is large, the geological conditions are complex, and the seismic response characteristics are obvious. The development zone can be qualitatively identified, and the target layer can be determined, but the vertical reservoir resolution is poor, and the accuracy of conventional prediction methods is insufficient, which needs to be further improved.

[0034] Step 2: After fine calibration of well-seismic joint analysis, it is determined that the target layer segment shows a trough reflection with a relatively strong amplitude on the seismic scale and is located below the trough. However, it is difficult to effectively distinguish the reservoir from the seismic response characteristics. The reservoir segment is subjected to spectrum analysis to determine the main frequency and bandwidth, and the time window parameters are recorded to prepare for subsequent work.

[0035] Step 3: In case the internal information of the reservoir is not obvious, perform wavelet decomposition on the original seismic data according to the main frequency and bandwidth information recorded in the previous step, decompose it into wavelets of different frequencies within the effective frequency band, and remove unnecessary seismic response interference. Combined with the wells that have been drilled, select the frequency band that can characterize the reservoir, and then re-screen the remaining seismic wavelets to select favorable seismic track sets, and form seismic data that is favorable to the reservoir within the appropriate frequency band. The entire process must fully consider the well profile and plane spectrum distribution characteristics, and multiple tests must be conducted to select the best parameter combination to identify favorable reservoirs, such as Figure 1 .

[0036] Step 4: Due to the characteristics of the target layer section in this work area, such as deep burial depth and wide distribution of wells, combined with the above actual situation, this embodiment uses the trend surface analysis method. This method believes that the logging value changes with the spatial position and forms a certain change trend. After multiple tests, the best fitting number is obtained, thereby obtaining the trend value that conforms to the trend, and generating the corrected AC contour map, that is, Figure 2 As shown, where:

[0037]

[0038] In the formula, m is the best fitting number, (x, y) is the well location coordinate, and for the variable n, an integer can be substituted. After summing the function A(i), the trend value Z is finally obtained.

[0039] Step 5: Conduct intersection analysis on conventional logging curves such as natural gamma, density, compensated neutron, and sonic time difference, conduct a comprehensive analysis and study on the sensitivity, physical properties, or lithology of the reservoir, and determine that the curves that best characterize the reservoir and describe the lithology details of the reservoir are sonic curves and natural gamma curves.

[0040] Step 6: According to step 5, the acoustic wave curve and the natural gamma curve have the strongest correlation. The acoustic wave curve and the natural gamma curve are used to convert and reconstruct, and a wave impedance curve reflecting the reservoir characteristics is re-optimized. This curve can better judge the changes in the lithology of the formation, such as Figure 3 As shown in the cross-plot, it can be seen intuitively that the optimized curve can distinguish the bottom conglomerate from the mudstone.

[0041] Step 7: Combined with the reservoir wave impedance curve, re-extract the wavelet to complete the fine calibration of the seismic synthetic record, laying the foundation for subsequent inversion work.

[0042] Step 8: Use the re-optimized impedance curve to establish an impedance model, and perform reservoir inversion on this basis. During the inversion process, adjust the target sampling rate, optimal cutoff frequency, filter and other parameters to achieve the best effect. The selection of parameters depends on the actual work area conditions.

[0043] Step 9: Combined with known wells, the new method has a higher consistency rate in the predicted values ​​presented in the plane compared with the conventional wave impedance inversion. Figure 4 As shown in the figure, the blue box corresponding to the new method is discontinuous, and the vertical resolution at the same position is significantly improved, which can fully reflect the vertical and horizontal distribution characteristics of the reservoir. The accuracy of lithology prediction in reservoir inversion has indeed been effectively improved.

[0044] The above method for improving the prediction accuracy of thin reservoirs under complex structures is proposed for the first time. It uses the seismic data body reconstructed by wavelet decomposition, makes full use of the lithology and physical property characterization characteristics of the curve sensitive to the reservoir, performs optimization and depth optimization on the original sensitive curve, adjusts the appropriate parameters during the reservoir inversion process, and finally completes the process of improving the prediction accuracy of thin reservoirs under complex structures. This method enriches the details of lithology description, not only improves the resolution of the inversion profile, but also finely depicts the reservoir boundary, which has important guiding significance for the thin reservoir prediction method. For the effective characterization of the seismic response characteristics of real thin reservoirs under complex geological backgrounds, the focus is on how to effectively improve the vertical and lateral resolution of thin reservoirs. At the same time, the research results of this method can provide favorable technical support for the identification of thin reservoirs in the future, which will help improve the benefits of exploration and development. At present, this technical method has been applied in actual production blocks and has achieved good results, which has guiding significance for future deep exploration research goals.

[0045] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for improving the prediction accuracy of thin reservoirs under complex structures, characterized in that: The following steps are involved: Step 1: Analyze the reflection characteristics of the original seismic data and its influence on the reservoir response characteristics, and combine the well-seismic response characteristics of the forward model to qualitatively identify the thin reservoir development zone, find out and determine the target layer that affects the reservoir; Step 2: Fine calibration of well-seismic joint analysis to analyze the seismic response characteristics of the target layer and to determine the main frequency, frequency band width and time window parameters; Step 3: Based on step 2, use the wavelet decomposition and reconstruction technology to reduce the impact of strong reflections above and below the reservoir on the reflection waves of the target reservoir; Step 4: Standardize the logging curves and unify the same type of logging data to the same scale level; Step 5: Comprehensively compare the logging curves related to the reservoir, conduct sensitivity curve analysis, and determine the curve that is most suitable for the target layer and can distinguish lithology; Step 6: Using wavelet transform reconstruction technology, fit the two logging curves with the largest correlation coefficient into a relationship, and then perform transformation and reconstruction to construct a wave impedance curve; Step 7: Re-extract the seismic wavelet and make a synthetic seismic record with the wave impedance curve; Step 8: After analyzing the waveform characteristics of the seismic traces near the well, perform frequency decomposition in the wavelet domain and extract the common characteristics of this set of curves to form an isochronous grid model. Then, under the constraints of the Bayesian framework, the initial model is modified in combination with seismic and well data to obtain the inversion results, thereby identifying the vertical and horizontal reservoir distribution.

2. A method for improving the prediction accuracy of thin reservoirs under complex structures according to claim 1, characterized in that: In step 1, the target layer that has an impact on the reservoir refers to the event axis that can reflect the true seismic response characteristics of the reservoir.

3. A method for improving the prediction accuracy of thin reservoirs under complex structures according to claim 1, characterized in that: In step 2, the calibration is mainly based on the principle of matching waveform, phase and energy characteristics. First, the marker layer that can be used for stratigraphic comparison is identified, and then the target reservoir is calibrated. After checking any well line and repeated comparisons, the stratigraphic layers with different lithology and electrical properties are linked to the reflection characteristics of the seismic profile to verify whether the calibration is accurate.

4. A method for improving the prediction accuracy of thin reservoirs under complex structures according to claim 1, characterized in that: In step 3, the wavelet decomposition and reconstruction technology is specifically as follows: based on the multi-wavelet seismic trace model, the seismic trace matching is decomposed into multiple wavelets with different main frequencies and shapes, and then selectively reconstructed to identify favorable reservoirs or fluid types.

5. The method for improving the prediction accuracy of thin reservoirs under complex structures according to claim 1, characterized in that: In step 4, the logging curve is standardized to eliminate the influence of different instruments, scales, time spans, and operating methods on the authenticity and reliability of the logging curve.

6. A method for improving the prediction accuracy of thin reservoirs under complex structures according to claim 1, characterized in that: In step 4, the standardization processing methods include overlapping curve method, depth trend construction method, histogram method, and trend surface method.

7. A method for improving the prediction accuracy of thin reservoirs under complex structures according to claim 1, characterized in that: In step 5, the logging curves include natural gamma curves, density curves, compensated neutron curves, and acoustic time difference curves.

8. The method for improving the prediction accuracy of thin reservoirs under complex structures according to claim 1, characterized in that: The method also includes step 9: comparing the prediction results with known wells, and comprehensively analyzing the reservoir prediction compliance rate and the degree of improvement in prediction accuracy.

Images