A seismic forward and reservoir favorable area selection method

Abstract

The present application belongs to the field of seismic exploration, and particularly relates to a method for selecting a favorable reservoir area by seismic forward modeling. The present application selects a typical well crossing a target layer, intercepts a seismic profile above and below the well, reads amplitude values and amplitude ratios of wave crests, resamples a well logging curve to establish a forward modeling, performs forward modeling on the model by using different primary frequency wavelets, selects a primary frequency wavelet with the minimum amplitude ratio error, performs forward modeling by transforming the target layer velocity, and summarizes waveform amplitude values and amplitude ratio changes as a favorable waveform feature. A region with the same waveform as the feature is a favorable reservoir area. The present application establishes a quantitative comparison relationship between a forward modeling and an actual seismic profile by using amplitude ratios, and thus can quantitatively analyze seismic forward modeling results, optimizes a primary frequency establishment process of the forward modeling, indirectly enhances the accuracy of well-crossing seismic forward modeling analysis, and improves the reliability of favorable reservoir area prediction.

Description

Technical Field

[0001] This invention belongs to the field of seismic exploration, specifically relating to a method for seismic forward modeling and selection of favorable reservoir areas. Background Technology

[0002] In seismic exploration, seismic forward modeling serves as a crucial link between earthquakes and geology. Forward modeling analysis is used to interpret and validate models and identify geological bodies. A sound interpretation of the geological forward model is essential throughout the entire reservoir prediction process. Seismic forward modeling meticulously analyzes the changes in reflection waveforms caused by different lithological combinations, organically combining geology and seismology.

[0003] In actual earthquake forward modeling, researchers often describe the model forward modeling results by observing the similarity between the earthquake waveforms obtained by forward modeling with different values ​​of specific parameters and the waveforms of actual seismic traces. For example, by enhancing or weakening the reflection axis, or shifting the time upward or downward, the optimal value of the parameter is determined, thereby characterizing the earthquake characteristics corresponding to the parameter. This characteristic can then be used to predict or evaluate other earthquake data.

[0004] For example, Chinese invention patent application CN111090117A discloses an effective reservoir prediction method and system under phase-controlled forward modeling constraints. The method matches the forward model with actual seismic data under different interlayer thickness parameters until the frequency and waveform of the obtained forward model are closest to those of the actual seismic data. At this point, the interlayer thickness parameters that achieve the best match are recorded and used in subsequent effective reservoir prediction.

[0005] Although the above method can obtain the interlayer thickness parameters through forward modeling, it only considers the waveform similarity and qualitatively describes the forward modeling results, resulting in low reliability when predicting or evaluating effective reservoirs. Summary of the Invention

[0006] The purpose of this invention is to provide a method for seismic forward modeling and reservoir favorable area selection, which solves the problem that existing methods for seismic forward modeling and reservoir favorable area selection can only perform qualitative analysis by comparing the similarity between the seismic waveform obtained from forward modeling and the waveform of the actual seismic trace, resulting in low reliability of forward modeling and thus affecting the reliability of reservoir favorable area prediction.

[0007] To achieve the above objectives, the present invention provides a seismic forward modeling method, comprising the following steps:

[0008] 1) Obtain seismic, geological and well logging data of the seismic work area to be evaluated, determine the seismic frequency band range of the target layer and the average thickness of a single layer of the target layer, select wells that pass through the target layer in the work area as typical wells, extract the seismic profiles through the wells within the first depth range above and below the target layer, and read the amplitude values ​​of the peaks and / or troughs above and below the target layer and their amplitude ratios from the seismic profiles.

[0009] 2) Determine the corresponding formation velocity based on the sonic transit time of a typical well, resample the logging curves, and select the resampled formation velocity and density curves of the first depth range above and below the target layer to establish a depth domain layered forward model.

[0010] 3) Perform seismic forward modeling on the established depth-domain layered forward modeling model using wavelets of different dominant frequencies within the frequency band. Read the amplitude ratio of the upper and lower peaks and / or troughs of the target layer in each forward modeling result and compare it with the amplitude ratio of the target layer in the well-pass seismic profile. Select the dominant frequency value that has the smallest absolute error between the amplitude ratio and the amplitude ratio of the well-pass seismic profile, and use the wavelet of the dominant frequency as the seismic wavelet used in the model forward modeling.

[0011] After determining the seismic wavelet to be used in the forward modeling, the seismic forward modeling simulation is performed using the wavelet and the established depth-domain layered forward modeling model.

[0012] The beneficial effects are as follows: This seismic forward modeling method establishes a quantitative comparison relationship between the forward model and the actual seismic profile by using the amplitude ratio as a bridge. The dominant frequency value used when the absolute error between the amplitude ratio and the amplitude ratio of the well-pass seismic profile is minimized is selected for seismic forward modeling, making the results obtained from the forward modeling closer to the actual seismic data. Therefore, it can improve the accuracy of subsequent well-pass seismic forward modeling analysis. At the same time, the setting of resampling well logging curve modeling enhances the generalizability of forward modeling for strata of different thicknesses in different blocks, and enhances the applicability of the forward modeling analysis method.

[0013] Furthermore, to optimize the process of establishing the dominant frequency of the forward model, and to balance the selection efficiency and accuracy of the dominant frequency of the forward model, so as to make the forward modeling results more reliable, in step 3), when determining the dominant frequency value of the seismic wavelet used in the forward modeling, the results of seismic forward modeling of the model are first tested with wavelets of different dominant frequencies using a first set step size. After determining the wavelet range in which the forward modeling results are consistent with the waveform shape and peak-valley number of the well-pass seismic profile, a second set step size is used to compare the amplitude ratio with the target layer of the well-pass seismic profile within the wavelet range. The first set step size is larger than the second set step size.

[0014] Since a forward modeling process is required for each dominant frequency value tested during the selection of the dominant frequency, performing a forward modeling for all frequency values ​​within the seismic frequency band would increase the selection time and cost. Therefore, a qualitative coarse screening is performed by first selecting a portion of the dominant frequency values ​​with a larger step size and then comparing the waveforms to select the frequency band range that may contain the optimal dominant frequency value between the step sizes. Then, the dominant frequency values ​​within this range are forward modeled and the amplitude ratios are compared to perform a quantitative fine screening to select the optimal dominant frequency value. This method avoids performing multiple forward models and can obtain more accurate selection results.

[0015] Furthermore, the first depth range is 50-100ms.

[0016] The selected depth range includes at least one seismic reflection wavelength. The magnitude of the seismic reflection wavelength is mainly affected by the dominant seismic frequency and the seismic velocity. The higher the dominant seismic frequency, the smaller the reflection wavelength; the higher the seismic velocity, the larger the reflection wavelength.

[0017] Furthermore, the first set step size is 5 Hz, and the second set step size is 1 Hz.

[0018] The value of the first set step size is related to the seismic frequency band range of the target layer. If the seismic frequency band range of the target layer is large, the first set step size can be set to be larger. The value of the second set step size is related to the requirements for selection time and the accuracy of forward modeling results. If a shorter selection time is required, the second set step size can be set to be larger. If a more accurate forward modeling result is required, the second set step size can be set to be smaller.

[0019] Furthermore, the resampling uses a depth sampling interval of 1 / 2 of the average thickness of a single layer of the target layer.

[0020] The resampling depth interval should be set so as not to miss the velocity and density features of a single layer of the target layer while reducing the number of sampling points. Therefore, the advantage of setting the depth sampling interval to 1 / 2 of the average thickness is that the resampled data can both simplify the model and not miss the features of typical layers.

[0021] The present invention also provides a method for selecting favorable reservoir areas, comprising the following steps:

[0022] 1) Obtain seismic, geological, and well logging data of the seismic work area to be evaluated, determine the seismic frequency band range of the target layer and the average thickness of a single layer of the target layer, select high-yield wells that pass through the target layer in the work area as typical wells, the high-yield wells being drilled wells whose production reaches the set threshold; extract the well-passing seismic profile within the first depth range above and below the target layer, and read the amplitude values ​​and amplitude ratios of the peaks and / or troughs above and below the target layer in the seismic profile;

[0023] 2) Determine the corresponding formation velocity based on the sonic transit time of a typical well, resample the logging curves, and select the resampled velocity and density curves of the first depth range above and below the target layer to establish a depth domain layered forward model.

[0024] 3) Perform seismic forward modeling on the established depth-domain layered forward modeling model using wavelets of different dominant frequencies within the frequency band. Read the amplitude ratio of the upper and lower peaks and / or troughs of the target layer in each forward modeling result and compare it with the amplitude ratio of the target layer in the well-pass seismic profile. Select the dominant frequency value that has the smallest absolute error between the amplitude ratio and the amplitude ratio of the well-pass seismic profile, and use the wavelet of the dominant frequency as the seismic wavelet used in the model forward modeling.

[0025] 4) Using the seismic wavelet selected in step 3), by changing the velocity of the target layer in the depth domain layered forward model, perform seismic forward modeling respectively, record the waveform morphology above and below the target layer, as well as the amplitude values ​​and amplitude ratios of the peaks and / or troughs in each forward modeling result, and summarize the waveform changes as characteristics of favorable waveforms.

[0026] 5) Apply the characteristics of the favorable waveform to the seismic work area to be evaluated, find the area in the work area with the same waveform shape and amplitude ratio as the favorable waveform, and determine the area as the favorable reservoir area.

[0027] The beneficial effects are as follows: This method for selecting favorable reservoir areas establishes a quantitative comparison between the forward model and the actual seismic profile using the amplitude ratio as a bridge. The dominant frequency value used when the absolute error between the amplitude ratio and the well-pass seismic profile is minimized is selected for seismic forward modeling, making the results obtained from the forward modeling closer to the actual seismic data. This improves the accuracy of subsequent well-pass seismic forward modeling analysis. Simultaneously, the resampling logging curve modeling setting reduces sampling points without omitting the velocity and density characteristics of individual layers in the target layer, simplifying the model. This enhances the generalizability of forward modeling for different blocks and formations of varying thicknesses, thus improving the applicability of the forward analysis method. Without resampling, the model would have too many layers, making it overly complex and lacking in abstraction. Furthermore, by obtaining a more accurate seismic forward modeling dominant frequency to more accurately determine the waveforms above and below the target layer, the amplitude ratio of the waveforms above and below the target layer quantitatively characterizes the waveform characteristics of high-yield wells, making the identified favorable reservoir areas more reliable.

[0028] Furthermore, in order to optimize the process of establishing the dominant frequency of the forward model and take into account both the selection efficiency and accuracy of the dominant frequency of the forward model, so as to make the forward modeling results more reliable, in step 3), when determining the dominant frequency value of the seismic wavelet used in the forward modeling, the results of seismic forward modeling of the model with wavelets of different dominant frequencies are first tested with a first set step size. After determining the wavelet range in which the forward modeling results are consistent with the waveform shape and peak-valley number of the well-pass seismic profile, a second set step size is used to compare the amplitude ratio with the target layer of the well-pass seismic profile within the wavelet range. The first set step size is larger than the second set step size.

[0029] Since a forward modeling process is required for each dominant frequency value tested during the selection of the dominant frequency, performing a forward modeling for all frequency values ​​within the seismic frequency band would increase the selection time and cost. Therefore, a qualitative coarse screening is performed by first selecting a portion of the dominant frequency values ​​with a larger step size and then comparing the waveforms to select the frequency band range that may contain the optimal dominant frequency value between the step sizes. Then, the dominant frequency values ​​within this range are forward modeled and the amplitude ratios are compared to perform a quantitative fine screening to select the optimal dominant frequency value. This method avoids performing multiple forward models and can obtain more accurate selection results.

[0030] Furthermore, the first depth range is 50-100ms.

[0031] Furthermore, the first set step size is 5 Hz, and the second set step size is 1 Hz.

[0032] The value of the first set step size is related to the seismic frequency band range of the target layer. If the seismic frequency band range of the target layer is large, the first set step size can be set to be larger. The value of the second set step size is related to the requirements for selection time and the accuracy of forward modeling results. If a shorter selection time is required, the second set step size can be set to be larger. If a more accurate forward modeling result is required, the second set step size can be set to be smaller.

[0033] Furthermore, the resampling uses a depth sampling interval of 1 / 2 of the average thickness of a single layer of the target layer. Attached Figure Description

[0034] Figure 1 This is a flowchart of a seismic forward modeling method according to the present invention;

[0035] Figure 2 This is a flowchart of a method for selecting a favorable reservoir area according to the present invention;

[0036] Figure 3a This is a schematic diagram of the seismic profile taken 80ms above and below the target layer in an embodiment of the reservoir favorable area selection method of the present invention.

[0037] Figure 3b This is a frequency-amplitude diagram of the target layer's 80ms well-pass seismic profile captured in an embodiment of the reservoir favorable area selection method of the present invention.

[0038] Figure 4 This is a schematic diagram of a depth-domain layered forward model established by resampling the velocity and density curves 150m above and below the target layer in an embodiment of the reservoir favorable area selection method of the present invention.

[0039] Figure 5 This is a schematic diagram of the forward modeling results under different dominant frequency Ricker wavelets and the amplitude ratio of the upper and lower peaks of the target layer in an embodiment of the reservoir favorable area selection method of the present invention.

[0040] Figure 6 In an embodiment of the reservoir favorable area selection method of the present invention, different forward modeling results are obtained by changing the target layer velocity in the depth domain layered forward model when the absolute error between the amplitude ratio and the amplitude ratio of the well seismic profile is the smallest.

[0041] Figure 7 In an embodiment of the reservoir favorable area selection method of the present invention, different forward modeling results are obtained by changing the target layer velocity in the depth domain layered forward model when using the 35 Hz dominant frequency of the Ricker wavelet. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0043] Examples of Seismic Forward Modeling Methods

[0044] This embodiment provides a seismic forward modeling method, referring to... Figure 1 The specific steps are as follows:

[0045] 1) Obtain seismic, geological and well logging data of the seismic work area to be evaluated, determine the seismic frequency band range of the target layer and the average thickness of a single layer of the target layer, select wells that pass through the target layer in the work area as typical wells, extract the seismic profiles through the wells within the first depth range above and below the target layer, and read the amplitude values ​​of the peaks and / or troughs above and below the target layer and their amplitude ratios from the seismic profiles.

[0046] This step is the data preparation step. The selected typical well is a drilled well that has passed through the target layer and for which density and sonic logging data have been obtained. To ensure data reliability and provide more accurate data support for seismic forward modeling, the selected typical well should have a high signal-to-noise ratio (SNR) for the seismic data. Subsequent seismic forward modeling is mainly based on the seismic characteristics of this target layer. Typically, the first depth range is 50-100 ms. The selected depth range includes at least one seismic reflection wavelength. The magnitude of the seismic reflection wavelength is mainly affected by the dominant seismic frequency and seismic velocity; the higher the dominant seismic frequency, the smaller the reflection wavelength; the higher the seismic velocity, the larger the reflection wavelength.

[0047] 2) Determine the corresponding formation velocity based on the sonic transit time of a typical well, resample the logging curves, and select the resampled formation velocity and density curves of the first depth range above and below the target layer to establish a depth domain layered forward model.

[0048] Because well logging curves involve numerous sampling points and corresponding layers, resampling would make the established forward model overly complex and lacking in abstraction. Therefore, resampling of the well logging curves is necessary. During resampling, a depth sampling interval of half the average thickness of a single layer in the target formation is used. This resampling well logging curve modeling and the corresponding depth sampling interval reduce the number of sampling points while retaining the velocity and density characteristics of the single layer in the target formation, simplifying the model. This enhances the generalizability of forward modeling for formations of different thicknesses in different blocks and improves the applicability of the forward analysis method.

[0049] 3) Initialize the model using wavelets of different dominant frequencies within the frequency band, i.e., perform seismic forward modeling using the depth-domain layered forward model established in step 2). Read the amplitude ratios of the upper and lower peaks and / or troughs of the target layer in each forward modeling result, and compare them with the amplitude ratios of the target layer in the well-pass seismic profile. Select the dominant frequency value with the smallest absolute error between the amplitude ratio and the amplitude ratio in the well-pass seismic profile, and use the wavelet of this dominant frequency as the seismic wavelet used in the model forward modeling. In this embodiment, the Ricker wavelet is used for seismic forward modeling. The Ricker wavelet has the advantages of concentrated energy and low sidelobes, and can better highlight seismic characteristics.

[0050] To optimize the process of establishing the dominant frequency of the forward model and balance the efficiency and accuracy of selecting the dominant frequency, making the forward modeling results more reliable, when determining the dominant frequency value of the seismic wavelet used in the forward modeling, the results of seismic forward modeling of the model with wavelets of different dominant frequencies are first tested with a first set step size. After determining the wavelet range in which the forward modeling results are consistent with the waveform shape and peak-valley number of the well-pass seismic profile, a second set step size is used to compare the amplitude ratio with the target layer of the well-pass seismic profile within the wavelet range. The first set step size is larger than the second set step size.

[0051] Since a forward modeling process is required for each dominant frequency value tested during the selection of the dominant frequency, performing a forward modeling for all frequency values ​​within the seismic frequency band would increase the selection time and cost. Therefore, a method can be adopted to first select a portion of the dominant frequency values ​​for forward modeling with a larger step size. The waveform of the forward modeling result is then compared with the approximate shape of the waveform in the well-pass seismic profile. This is equivalent to a preliminary qualitative screening to select the frequency band range that may contain the optimal dominant frequency value within the step size. Then, the dominant frequency values ​​within this range are forward modeled with a smaller compensation. The waveform shape above and below the target layer, as well as the amplitude and amplitude ratio of the peaks and / or troughs, corresponding to the forward modeling result are compared with the amplitude and amplitude ratio corresponding to the waveform in the well-pass seismic profile. This quantitative fine screening is then performed to select the optimal dominant frequency value. This method avoids multiple forward modeling iterations and yields more accurate selection results.

[0052] The first set step size is related to the seismic frequency band range of the target layer. If the target layer seismic frequency band range is large, the first set step size can be set larger. The second set step size is related to the requirements for selection time and the accuracy of forward modeling results. If a shorter selection time is required, the second set step size can be set larger; if a more accurate forward modeling result is required, the second set step size can be set smaller. In this embodiment, the known target layer seismic frequency band range is 9-55 Hz, so the corresponding first set step size is set to 5 Hz and the second set step size is set to 1 Hz.

[0053] The above process allows us to select the wavelet dominant frequency that makes the forward modeling results closer to the actual earthquake data. Using the wavelet with this dominant frequency in earthquake forward modeling can improve the reliability of earthquake forward modeling analysis.

[0054] Example of reservoir favorable area selection method

[0055] This embodiment provides a seismic forward modeling method, referring to... Figure 2 The specific steps are as follows:

[0056] 1) Obtain seismic, geological, and well logging data of the seismic work area to be evaluated, determine the seismic frequency band range of the target layer and the average thickness of a single layer of the target layer, select high-yield wells that pass through the target layer in the work area as typical wells, the high-yield wells being drilled wells whose production reaches the set threshold; extract the well-passing seismic profile within the first depth range above and below the target layer, and read the amplitude values ​​and amplitude ratios of the peaks and / or troughs above and below the target layer in the seismic profile.

[0057] 2) Determine the corresponding formation velocity based on the sonic transit time of a typical well, resample the logging curves, and select the resampled velocity and density curves of the first depth range above and below the target layer to establish a depth-domain layered forward model.

[0058] 3) Use wavelets of different dominant frequencies within the frequency band to perform seismic forward modeling on the initial model, i.e. the depth domain layered forward modeling model established in step 2). Read the amplitude ratio of the upper and lower peaks and / or valleys of the target layer in each forward modeling result, compare it with the amplitude ratio of the target layer in the well seismic profile, and select the dominant frequency value that has the smallest absolute error between the amplitude ratio and the amplitude ratio of the well seismic profile. Use the wavelet of this dominant frequency as the seismic wavelet used in the model forward modeling.

[0059] The only difference between the above steps and the seismic forward modeling method embodiment is that in this embodiment, the selected typical wells also need to meet the conditions of high-yield wells. Here, a high-yield well refers to a well with a production rate greater than a certain threshold, such as a natural gas production rate of more than 10,000 cubic meters per day, which is considered a high-yield well. Since the purpose of seismic forward modeling in this embodiment is to find favorable reservoir areas, the seismic features extracted and used as templates for favorable reservoir areas must be seismic features of high-yield wells.

[0060] The remaining steps described above have been explained in detail in the seismic forward modeling method embodiments, and will not be repeated here.

[0061] 4) Using the seismic wavelet selected in step 3), by changing the velocity of the target layer in the depth domain layered forward model, perform seismic forward modeling respectively, record the waveform morphology above and below the target layer, as well as the amplitude values ​​and amplitude ratios of the peaks and / or troughs in each forward modeling result, and summarize the waveform changes as characteristics of favorable waveforms.

[0062] This step mainly uses the waveform morphology above and below the target layer, as well as the amplitude values ​​and amplitude ratios of the peaks and / or troughs, as seismic features to quantitatively describe the characteristics of high-yield wells, thereby obtaining a characteristic template for high-yield areas (favorable reservoir areas).

[0063] 5) Apply the characteristics of the favorable waveform to the seismic work area to be evaluated, find the area in the work area with the same waveform shape and amplitude ratio as the favorable waveform, and determine the area as the favorable reservoir area.

[0064] The following describes this embodiment in detail with reference to the accompanying drawings, using the seismic forward modeling of well B in work area A as an example; wherein, the wavelet selected for the seismic forward modeling is the Ricker wavelet.

[0065] 1) Based on the acquired seismic, geological, and well logging data from work area A, the seismic frequency band of the target layer was determined to be 9-55 Hz, and the average thickness of a single layer was 4 m. A typical well, well B, with a high seismic signal-to-noise ratio, penetrating the target layer, and for which density and sonic logging data have been obtained, was selected from work area A. This well is a high-yield well. For example... Figure 3a As shown, the seismic profiles through the well are taken 80ms above and below the target layer in well B. Figure 3b The diagram shows the frequency-amplitude profile of the rectangular area of ​​the well-pass seismic profile, which reflects the frequency band range of the seismic profile. The amplitude values ​​of the main seismic reflection interface T9 and the wave peak below it near the target layer of the seismic profile are read as 31430 and 2722, respectively. The amplitude ratio of the main seismic reflection interface T9 to the wave peak below T9 is calculated to be 31430 / 2722 = 11.5.

[0066] 2) Calculate the formation velocity corresponding to the target layer based on the sonic transit time curve of well B. Resample the logging curves using half the average thickness of the target layer (2m) as the depth sampling interval. Establish the velocity and density curves from the resampled values ​​at depths of 150m above and below the target layer, as shown in the figure. Figure 4 The depth-domain layered forward model shown is as follows, where Figure 4The horizontal axis represents the model's length, an extension made for aesthetic purposes; the vertical axis represents the model's depth, corresponding one-to-one with the depth of the logging curves. Each layer of the model carries velocity and density information for that layer, with all data points (logging curve depths) having consistent values ​​in the horizontal direction. Color codes indicate the velocity of each layer. The arithmetic mean of the sonic transit time values ​​of the resampling points of the target layer was calculated, and the reciprocal was taken and standardized to m / s, determining the average velocity of the target layer to be 5200 m / s.

[0067] 3) Seismic forward modeling is performed on the initial model, i.e., the depth-domain layered forward model established in step 2), using Ricker wavelets of different dominant frequencies within the frequency band. Initially, different dominant frequencies of Ricker wavelets are tested with a step size of 5 Hz. For example... Figure 5 As shown, the forward modeling results of the Ricker wavelet with a dominant frequency of 30-35 Hz are basically consistent with the waveform morphology and peak-valley number of the well-pass seismic profile. Then, the dominant frequency step size is adjusted to 1 Hz, and the amplitude ratio of the upper and lower peaks of the target layer in each forward modeling result is read and compared with the amplitude ratio of 11.5 of the target layer in the seismic profile. It is found that the peak-amplitude ratio of 9.8 of the Ricker wavelet with a dominant frequency of 32 Hz has the smallest absolute error with the peak-amplitude ratio of 11.5 of the actual well-pass seismic profile. Therefore, the dominant frequency value of the Ricker wavelet for the subsequent model is selected as 32 Hz.

[0068] 4) Using the 32Hz dominant frequency Ricker wavelet as the seismic wavelet for subsequent models, the velocity of the target layer in the depth-domain layered forward model was transformed (from 4000m / s to 6000m / s, specifically 4000m / s, 4500m / s, 5200m / s, and 6000m / s), and seismic forward modeling was performed accordingly. The forward modeling results are as follows: Figure 6 As shown.

[0069] Record the amplitude values ​​and ratios of the peaks above and below the target layer in each forward modeling result, and summarize the following waveform variation patterns: As the velocity of the target layer increases (from 4000m / s to 6000m / s in the forward modeling model, with the increase process being 4000m / s, 4500m / s, 5200m / s, and 6000m / s), the amplitude of the T9 peak first weakens and then strengthens (from 0.0263 to 0.02616 and then to 0.0287 in the forward modeling model), and the amplitude of the peak below T9 also gradually strengthens (from 0.00152 to 0.00536 in the forward modeling model). The calculated ratio of the amplitudes of the peaks below T9 decreases from 17.6 to 5.4. It can be seen that for the high-yield well type corresponding to well B, if the formation velocity decreases only due to the gas content of the target layer, then the waveform with a strong T9 peak, a weaker lower T9 peak, and a large amplitude ratio between T9 and the lower peak is determined to be the favorable waveform. This is similar to the pattern verified by other high-yield wells.

[0070] If the wavelet frequency is not precisely determined and forward modeling is performed directly, such as using a 35 Hz Ricker wavelet, the forward modeling result will be as follows: Figure 7 As shown, the observed variation pattern is as follows: as the velocity of the target layer increases (from 4000 m / s to 6000 m / s in the forward model, with the increase occurring at 4000 m / s, 4500 m / s, 5200 m / s, and 6000 m / s), the amplitude of the T9 peak first weakens and then strengthens (from 0.01885 to 0.01787 and then to 0.01825 in the forward model), while the amplitude of the peak below T9 first weakens and then strengthens (from 0.00457 to 0.00377 and then to 0.00512 in the forward model). The calculated favorable waveform characteristics are: the ratio of the amplitude of T9 to the peak below it first increases and then decreases (from 4.12 to 4.74 and then to 3.56 in the forward model). Therefore, the conclusion of selecting favorable reservoir areas using the 35Hz Ricker wavelet forward model is that the waveform with a medium-strong T9 peak and a relatively strong lower T9 peak is favorable. This is different from the 35Hz Ricker wavelet forward model and contradicts the pattern verified by other high-yield wells. For example, the characteristics of other high-yield wells in area A are also the same: a medium-strong T9 peak and a relatively weak lower T9 peak. In contrast, areas with a medium-strong T9 peak and a relatively strong lower T9 peak contain many unsuccessful wells, and the reservoirs are mostly underdeveloped.

[0071] Comparative studies have shown that the method for selecting favorable reservoir areas, based on amplitude ratio, quantitatively selects the dominant frequency of the forward model, which can effectively enhance the reliability of forward modeling and improve the accuracy of selecting effective reservoirs.

[0072] 5) Apply the summarized waveform change patterns to the A seismic work area to find regions with the same waveform shape and amplitude ratio as the favorable waveform, namely, regions with strong T9 peaks, weaker lower T9 peaks, and larger amplitude ratios between T9 and the lower peaks, and determine these regions as favorable target areas.

[0073] This invention establishes a quantitative comparison between the forward model and the actual seismic profile by using the amplitude ratio of the upper and lower peaks and / or troughs of the target layer as a comparison parameter. This enables quantitative analysis of the seismic forward modeling results, thereby enhancing their reliability. Simultaneously, it optimizes the process of establishing the dominant frequency in the forward modeling, balancing accuracy and efficiency. By selecting the dominant frequency with the smallest error in the forward modeling results for forward simulation, the accuracy of well-through seismic forward modeling analysis is indirectly enhanced. Based on the above, by using the dominant frequency with the smallest error for forward simulation, combined with the quantitative comparison parameters, waveform characteristics of high-yield areas that better reflect reality can be obtained, thus improving the reliability of reservoir favorable area prediction.

[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A seismic forward modeling method, characterized in that, Includes the following steps: 1) Obtain seismic, geological and well logging data of the seismic work area to be evaluated, determine the seismic frequency band range of the target layer and the average thickness of a single layer of the target layer, select wells that pass through the target layer in the work area as typical wells, extract the seismic profiles through the wells within the first depth range above and below the target layer, and read the amplitude values ​​of the peaks and / or troughs above and below the target layer and their amplitude ratios from the seismic profiles. 2) Determine the corresponding formation velocity based on the sonic transit time of a typical well, resample the logging curves, and select the resampled formation velocity and density curves of the first depth range above and below the target layer to establish a depth domain layered forward model. 3) Perform seismic forward modeling on the established depth-domain layered forward modeling model using wavelets of different dominant frequencies within the frequency band. Read the amplitude ratio of the upper and lower peaks or the amplitude ratio of the upper and lower valleys of the target layer in each forward modeling result, and compare them with the amplitude ratio of the target layer in the well-pass seismic profile. Select the dominant frequency value that has the smallest absolute error between the amplitude ratio and the amplitude ratio of the well-pass seismic profile, and use the wavelet of the dominant frequency as the seismic wavelet used in the model forward modeling, so that the results obtained from the forward modeling are closer to the actual seismic data. After determining the seismic wavelet to be used in the forward modeling, the seismic forward modeling simulation is performed using the wavelet and the established depth-domain layered forward modeling model.

2. The seismic forward modeling method according to claim 1, characterized in that, In step 3), when determining the dominant frequency value of the seismic wavelet used in the forward modeling, the results of the seismic forward modeling of the model are first tested with wavelets of different dominant frequencies using a first set step size. After determining the wavelet range in which the forward modeling results are consistent with the waveform shape and peak-valley number of the well-pass seismic profile, a second set step size is used to compare the amplitude ratio with the target layer of the well-pass seismic profile within the wavelet range. The first set step size is larger than the second set step size.

3. The seismic forward modeling method according to claim 1, characterized in that, The first depth range is 50-100ms.

4. The seismic forward modeling method according to claim 2, characterized in that, The first set step size is 5 Hz, and the second set step size is 1 Hz.

5. The seismic forward modeling method according to claim 1, characterized in that, The resampling interval is half the average thickness of a single layer of the target layer.

6. A method for selecting favorable reservoir areas, characterized in that, Includes the following steps: 1) Obtain seismic, geological, and well logging data of the seismic work area to be evaluated, determine the seismic frequency band range of the target layer and the average thickness of a single layer of the target layer, select high-yield wells that pass through the target layer in the work area as typical wells, the high-yield wells being drilled wells whose production reaches the set threshold; extract the well-passing seismic profile within the first depth range above and below the target layer, and read the amplitude values ​​and amplitude ratios of the peaks and / or troughs above and below the target layer in the seismic profile; 2) Determine the corresponding formation velocity based on the sonic transit time of a typical well, resample the logging curves, and select the resampled velocity and density curves of the first depth range above and below the target layer to establish a depth domain layered forward model. 3) Perform seismic forward modeling on the established depth-domain layered forward modeling model using wavelets of different dominant frequencies within the frequency band. Read the amplitude ratio of the upper and lower peaks or the amplitude ratio of the upper and lower valleys of the target layer in each forward modeling result, and compare them with the amplitude ratio of the target layer in the well-pass seismic profile. Select the dominant frequency value that has the smallest absolute error between the amplitude ratio and the amplitude ratio of the well-pass seismic profile, and use the wavelet of the dominant frequency as the seismic wavelet used in the model forward modeling, so that the results obtained from the forward modeling are closer to the actual seismic data. 4) Using the seismic wavelet selected in step 3), by changing the velocity of the target layer in the depth domain layered forward model, perform seismic forward modeling respectively, record the waveform morphology above and below the target layer and the amplitude values ​​and amplitude ratios of the peaks and / or troughs in each forward modeling result, and summarize the waveform changes as characteristics of favorable waveforms. 5) Apply the characteristics of the favorable waveform to the seismic work area to be evaluated, find the area in the work area with the same waveform shape and amplitude ratio as the favorable waveform, and determine the area as the favorable reservoir area.

7. The method for selecting favorable reservoir areas according to claim 6, characterized in that, In step 3), when determining the dominant frequency value of the seismic wavelet used in the forward modeling, the results of the seismic forward modeling of the model are first tested with wavelets of different dominant frequencies using a first set step size. After determining the wavelet range in which the forward modeling results are consistent with the waveform shape and peak-valley number of the well-pass seismic profile, a second set step size is used to compare the amplitude ratio with the target layer of the well-pass seismic profile within the wavelet range. The first set step size is larger than the second set step size.

8. The method for selecting favorable reservoir areas according to claim 7, characterized in that, The first depth range is 50-100ms.

9. The method for selecting favorable reservoir areas according to claim 8, characterized in that, The first set step size is 5 Hz, and the second set step size is 1 Hz.

10. The method for selecting favorable reservoir areas according to claim 9, characterized in that, The resampling interval is half the average thickness of a single layer of the target layer.

Images