Method for improving prediction precision of deep clastic rock thin reservoir and application thereof

Abstract

The invention provides a method for improving prediction precision of a deep clastic rock thin reservoir and application of the method, and relates to the technical field of oil exploration and development. The method comprises the following steps of: aiming at a deep clastic rock reservoir section, constructing guided filtering processing on three-dimensional seismic data; then carrying out spectral bluing frequency expansion processing to obtain a processed three-dimensional seismic data volume; carrying out correction and logging rock physical parameter intersection analysis on the logging curve; constructing an isochronous stratigraphic framework model, and using the 90-degree phase rotation three-dimensional seismic data volume to carry out waveform indication inversion; line drawing is carried out on the top and bottom interfaces of the predicted reservoir, interlayer attributes are extracted, and the boundary of the reservoir is depicted finely; and performing time-depth conversion by using the top-bottom interface of the predicted reservoir after line drawing, subtracting the top depth from the bottom depth, and generating a reservoir plane equal-thickness map by using a difference value, thereby realizing plane distribution and feature fine prediction of the complex clastic rock thin reservoir. The matching degree of the prediction result and the actual drilling well is high, and technical support is provided for efficient exploration and development of deep clastic rock oil and gas reservoirs.

Description

Technical Field

[0001] The invention belongs to the technical field of petroleum exploration and development, and in particular relates to a method for improving the prediction accuracy of deep clastic rock thin reservoirs and an application thereof. Background Art

[0002] Clastic reservoirs are the most important type of oil and gas reservoirs in oil and gas exploration and development. A large number of oil and gas reservoirs controlled by such reservoirs have been discovered in many basins. With the continuous deepening of exploration and development, higher requirements are placed on the prediction accuracy of such reservoirs. The accuracy of the prediction results is directly related to the economic benefits.

[0003] The concept of reservoir prediction has been proposed for nearly half a century. During this period, many prediction methods have emerged, such as seismic multi-attribute, frequency division technology, geostatistical inversion, prestack P- and S-wave joint inversion, and fluid detection. However, due to the different geological targets and exploration and development stages, the actual problems to be solved are also different, which makes the ideas and technical routes of reservoir prediction also have certain differences. At present, the deep clastic reservoirs targeted by exploration and development are often thin (less than 1 / 4 of the wavelength of seismic waves), strong heterogeneity, fast lateral changes, rapid thinning and pinching out, and vertical interlayer superposition. At the same time, due to the deep burial depth, low main frequency of seismic data, and poor quality of basic data, it is difficult to make a detailed prediction of such reservoirs.

[0004] Chinese invention patent 201811080757.0 discloses a method and device for predicting the thickness of thin reservoirs, which includes: calibrating the acquired seismic data and well logging data to determine the location of the target layer; performing three-dimensional seismic horizon interpretation to determine the first time when the seismic wave reaches the top of the target layer and the second time at the bottom; determining the forward model of the change of the thickness of the thin reservoir according to the well logging data, and determining the change of the seismic wave amplitude with the change of the thickness of the thin reservoir; determining the forward model of the change of the thickness of the overlying strata of the target layer according to the well logging data, and determining the law of the change of the seismic wave amplitude with the thickness of the overlying strata; determining the stratum thickness of the overlying stratum according to the seismic inversion and well logging data; setting a time window to extract the amplitude attribute of the target layer; correcting the amplitude attribute of the target layer according to the law of the change of the amplitude with the thickness of the overlying strata and the stratum thickness; predicting the thickness of the thin reservoir according to the corrected amplitude attribute. This application can improve the accuracy of the prediction of the thickness of thin reservoirs.

[0005] Chinese invention patent 201610379833.2 provides a method and device for predicting the thickness of thin reservoirs, wherein the method includes: convolving the formation reflection coefficient sequence of multiple first thin reservoirs within the work area with the seismic wavelet to obtain the geological basis function space, expanding the seismic data of multiple second thin reservoirs based on the geological basis function space to obtain multiple expansion coefficients, predicting the thickness of the second thin reservoir according to the expansion coefficients to obtain the maximum probability thickness, judging whether the difference between the maximum probability thickness and the actual thickness of the corresponding second thin reservoir is within a preset threshold, if so, the thickness prediction of the second thin reservoir is valid; judging whether the probability of the prediction being valid meets the preset accuracy, if so, predicting the thickness of the third thin reservoir using the geological basis function space. The method of the invention directly estimates the thickness of thin reservoirs from time domain seismic records, the method is easy to operate, has strong applicability, and has a high prediction accuracy for thin sandstone.

[0006] At present, the conventional technical means have low prediction accuracy and strong multi-solution of prediction results, which seriously restricts the efficient and rolling exploration and development of clastic oil and gas reservoirs. In addition, the low main frequency of deep seismic data, poor quality of basic data, and thin and complex clastic reservoirs make it difficult to accurately predict reservoirs. Therefore, it is urgent to provide a method for predicting thin reservoirs that can improve the prediction accuracy of deep clastic thin reservoirs. Summary of the invention

[0007] The present invention aims at the problems in the prior art such as low main frequency of deep seismic data, poor quality of basic data and great difficulty in fine reservoir prediction caused by thin and complex clastic reservoirs, and provides a method for improving the prediction accuracy of deep clastic thin reservoirs and its application. The method of the present invention formulates corresponding technical strategies for various factors that restrict the prediction accuracy, systematically improves the prediction accuracy of deep clastic thin reservoirs, and the prediction results are highly consistent with actual drilling, which provides technical support for the efficient exploration and development of deep clastic oil and gas reservoirs and achieves higher economic benefits.

[0008] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0009] In one aspect, the present invention provides a method for improving the prediction accuracy of deep clastic thin reservoirs, comprising the steps of:

[0010] S1. For the target deep clastic reservoir section, select the guide body calculation method and filter size, and perform structural guide filtering on the 3D seismic data;

[0011] S2, performing spectrum bluening and spectrum spreading processing on the three-dimensional seismic data obtained by the full-guided filtering processing in step S1 to obtain a processed three-dimensional seismic data volume;

[0012] S3, performing outlier correction and environmental correction on the logging curve to obtain the corrected logging rock physical parameters;

[0013] S4, performing crosstalk analysis on the well logging rock physical parameters after correction processing in step S3;

[0014] S5. Based on the fine calibration and interpretation results of the horizon, an isochronous stratigraphic framework model is constructed, and waveform indication inversion based on reservoir sensitive parameters is performed using the three-dimensional seismic data volume obtained by the 90-degree phase rotation step S2;

[0015] S6, based on the waveform indication inversion result of the reservoir sensitive parameter in step S5, line drawing is performed on the top and bottom interfaces of the predicted reservoir, and interlayer attributes are extracted to finely depict the boundary of the reservoir;

[0016] S7. Use the layer velocity to perform time-depth conversion on the top and bottom interfaces of the predicted reservoir after line drawing in step S6, and subtract the top depth from the bottom depth, and use the difference to generate a reservoir plane isopach map to achieve a detailed prediction of the plane distribution and characteristics of complex clastic rock thin reservoirs.

[0017] Preferably, in step S1, the guide body needs to be calculated before the filtering process, and the calculation of the guide body includes three key parameters: algorithm, aperture size and guide mode.

[0018] Further preferably, the algorithm is selected from at least one of a BG algorithm and a FFT algorithm.

[0019] More preferably, the algorithm is a BG algorithm.

[0020] Further preferably, the aperture size is 3 lines×3 channels×3 sampling points.

[0021] Further preferably, the guiding method is selected from at least one of horizontal guiding, center guiding and full guiding.

[0022] More preferably, the guiding method is a full guiding method.

[0023] Preferably, in step S1, the filter size is 1 line×1 channel×1 sampling point.

[0024] Preferably, in step S2, the spectrum blueing and frequency extension processing is specifically as follows: the three-dimensional seismic data obtained in step S1 is converted into a reflection coefficient by using circular deconvolution, that is, the three-dimensional seismic data is resampled to generate a new sparse pulse reflection coefficient, which is derived from the maximum and minimum amplitude values ​​in the three-dimensional seismic data; then the spectrum of the seismic signal is matched with the spectrum of the logging curve to design a spectrum blueing operator, and the designed spectrum blueing operator is used to convolve with the three-dimensional seismic data to generate the three-dimensional seismic data after frequency extension.

[0025] Further preferably, the spectrum blueing and frequency extension processing is specifically as follows: the three-dimensional seismic data obtained in step S1 is converted into a reflection coefficient by using circular deconvolution, that is, the three-dimensional seismic data is resampled to 0.5 ms to generate a new sparse pulse reflection coefficient, which is derived from the maximum and minimum amplitude values ​​in the three-dimensional seismic data; then the spectrum of the seismic signal is matched with the spectrum of the logging curve to design a spectrum blueing operator, and the designed spectrum blueing operator is used to convolve with the three-dimensional seismic data to generate the three-dimensional seismic data after frequency extension.

[0026] Preferably, in step S3, the abnormal value correction includes, but is not limited to, correcting the acoustic time difference (AC) parameter abnormality caused by wellbore expansion or other factors using a multivariate linear formula.

[0027] Further preferably, the outlier correction is an outlier correction of the AC curve, and a multivariate regression linear correction method is applied. The method is implemented as follows:

[0028] (1) Clarify the stratigraphic information of the wellbore expansion section, select multiple adjacent wells with better borehole quality, and calculate the correlation between the AC curves of the same stratigraphic section and adjacent depth sections and other curves that are less affected by the wellbore expansion;

[0029] (2) Select the three curves with good correlation with the AC curve, namely, the spontaneous potential curve (SP), the induction resistivity curve (RILM), and the neutron curve (CNL), to establish the multivariate linear equation AC = f (SP, RILM, CNL);

[0030] (3) The acoustic wave curve of the wellbore expansion section is fitted by the SP curve, RILM curve, CNL curve and the multivariate linear equation AC = f(SP, RILM, CNL) to replace the abnormal AC curve of the wellbore expansion section, and then spliced ​​with other AC curve segments of the well to obtain a new corrected AC curve.

[0031] Preferably, in step S3, the environmental correction includes but is not limited to spontaneous potential (SP) curve baseline correction.

[0032] Further preferably, the SP curve baseline correction is specifically: the SP curve baseline correction is 0 value.

[0033] Preferably, in step S4, the intersection analysis is specifically to perform an intersection analysis on the logging rock physical parameters of the clastic rock and the surrounding rock (mudstone) in the target layer.

[0034] Preferably, in step S5, the waveform indication inversion based on reservoir sensitive parameters refers to the inversion of the SP curve waveform indication combined with well-seismic data on the basis of fine stratum calibration and tracking to obtain the SP curve inversion volume result. This method is jointly driven by the model and data; specifically,

[0035] (1) Select the upper and lower adjacent marker layers of the target layer and establish a stratigraphic framework model based on the actual stratigraphic contact relationship;

[0036] (2) performing a 90-degree phase rotation on the three-dimensional seismic data volume obtained in step S2 to obtain a new three-dimensional seismic data volume;

[0037] (3) The waveform indication inversion method is used to invert the SP curve based on the idea of ​​combining well and seismic data. There are two important parameters that need to be analyzed and selected in the inversion operation process, namely the effective sample number and the optimal cutoff frequency.

[0038] Further preferably, the effective sample number described in (3) above in step S5 is set with reference to the statistical results of known wells, and statistical analysis is performed through the sample number and seismic correlation. The correlation gradually increases with the increase of the sample number, and the sample number when the correlation is maximum is the optimal sample parameter.

[0039] Further preferably, the optimal cutoff frequency parameter described in the above (3) in step S5 is related to the effective sample number parameter, and the optimal cutoff frequency parameter needs to be determined after the effective sample number is determined; if the inversion certainty is more preferred, the parameter should not be set too high; on the contrary, if the inversion resolution is more preferred and random results can be accepted, a higher cutoff frequency can be set.

[0040] Further preferably, for the optimal cut-off frequency, after the relevant curve enters the horizontal section, it indicates that everything except the frequency component is random, and the frequency before entering the horizontal section is the maximum effective frequency.

[0041] Preferably, step S6 is specifically as follows: on the basis of the SP inversion volume result obtained in step S5, a profile display is performed through the threshold value determined in step S4, and in combination with the well, the top and bottom interfaces of the predicted clastic reservoir S layer are manually line-drawn in the entire area, the top line-drawn layer is named T, and the bottom line-drawn layer is named B, and then the interlayer properties between the T layer and the B layer are extracted to obtain the planar distribution of the clastic reservoir and the planar characteristics of the SP parameters related to the physical properties and fluids.

[0042] Preferably, step S7 is specifically as follows: based on the velocity spectrum of the study area, the layer velocity of the S sublayer is extracted, and the top and bottom T layers and B layers of the layer are time-to-depth converted, the depth of the T layer is subtracted from the depth of the B layer to obtain the depth difference between the two layers, and the depth difference is used to make contour lines, thereby obtaining a plane isopach map of the S sublayer, and then the plane isopach map is combined with the plane isopach map to make a detailed description of the planar distribution characteristics of the complex clastic thin reservoir from the aspects of reservoir thickness, physical properties and fluid.

[0043] On the other hand, the present invention provides application of the above method in clastic reservoir prediction and / or carbonate layered reservoir prediction.

[0044] Preferably, the application is: using the reservoir plane isopach map established by the method to achieve detailed prediction of the plane distribution and characteristics of complex clastic thin reservoirs.

[0045] Finally, the present invention provides a reservoir plane isopach map established by the above method.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] 1. The present invention has formulated corresponding technical strategies for various factors that restrict the accuracy of reservoir prediction, systematically improving the prediction accuracy of deep clastic thin reservoirs, and the prediction results are highly consistent with actual drilling; the method of the present invention provides technical support for the efficient exploration and development of deep clastic oil and gas reservoirs, and has achieved higher economic benefits.

[0048] 2. The method described in the present invention can effectively improve the accuracy of prediction of various types of thin reservoirs, thereby reducing drilling risks and costs, and has a relatively broad application prospect.

[0049] 3. The method of the present invention, by selecting appropriate algorithms and guidance methods, makes the dip and azimuth information of the formation more clearly described; by spectrum blueing and frequency spreading processing, the resolution ability of seismic data for deep clastic reservoirs and the resolution limit of the formation are improved; through correction, not only the subsequent logging rock physical parameter analysis results are made more reliable, but also the calibration of the layer is made more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is a method and application flow chart for improving the prediction accuracy of deep clastic thin reservoirs as described in Example 1 of the present invention.

[0051] Figure 2 It is the guide volume profile extracted from three-dimensional seismic data using the FFT algorithm in Example 1 of the present invention.

[0052] Figure 3 It is the guide volume profile extracted from three-dimensional seismic data using the BG algorithm in Example 1 of the present invention.

[0053] Figure 4 It is a schematic diagram of three different guiding modes in the guiding body calculation process in Example 1 of the present invention.

[0054] Figure 5 It is a schematic diagram of the profile effect and main frequency of the original three-dimensional seismic data in Example 1 of the present invention.

[0055] Figure 6It is a schematic diagram of the profile effect and main frequency of three-dimensional seismic data after spectrum blueing and frequency enhancement processing in Example 1 of the present invention.

[0056] Figure 7 It is a schematic diagram comparing the acoustic wave curve (AC) before and after the wellbore expansion correction in Example 1 of the present invention.

[0057] Figure 8a It is a schematic diagram of the comparison of synthetic record calibration before the acoustic wave curve (AC) correction in Example 1 of the present invention.

[0058] Figure 8b It is a schematic diagram of the calibration comparison of the synthetic record after the acoustic wave curve (AC) correction in Example 1 of the present invention.

[0059] Fig. 9 It is a schematic diagram of the analysis and selection of the effective sample number for SP parameter waveform indication inversion in Example 1 of the present invention.

[0060] Fig.10 It is a schematic diagram of the analysis and selection of the best cutoff frequency for SP parameter waveform indication inversion in Example 1 of the present invention.

[0061] Fig.11 It is a schematic line drawing of the top and bottom interfaces of the S sublayer in Example 1 of the present invention based on the well and SP parameter inversion prediction results.

[0062] Fig.12 It is a plane distribution prediction map obtained by extracting interlayer attributes based on SP parameter inversion and line drawing of the S sublayer in Example 1 of the present invention.

[0063] Fig.13 It is a reservoir plane isopach map obtained based on SP parameter inversion and line drawing of the S sublayer in Example 1 of the present invention. DETAILED DESCRIPTION

[0064] The following non-limiting examples can make those of ordinary skill in the art understand the present invention more comprehensively, but do not limit the present invention in any way. The following content is merely an exemplary description of the scope of the present invention, and those skilled in the art can make various changes and modifications to the present invention according to the disclosed content, and it should also belong to the scope of the present invention. When the embodiment gives a numerical range, it should be understood that, unless otherwise specified in the present invention, the two endpoints of each numerical range and any numerical value between the two endpoints can be selected. Unless otherwise defined, all technical and scientific terms used in the present invention have the same meanings as those of ordinary skill in the art to which the present invention belongs. The present invention will be further described below in the form of specific embodiments.

[0065] The method for improving the prediction accuracy of deep clastic rock thin reservoirs according to the present invention and its application in Example 1 are shown in the flowchart. Figure 1 .

[0066] Example 1

[0067] Step S1: for the target deep clastic reservoir section, select the guide body calculation method and filter size, and perform structural full guide filter processing on the three-dimensional seismic data.

[0068] Specifically, in this embodiment, the structural guidance filtering technology is applied to filter the existing three-dimensional seismic data to improve the signal-to-noise ratio of the seismic data. This step is implemented by Opendtect software. Before filtering, the guidance body needs to be calculated, which mainly includes three key parameters: algorithm, aperture size, and guidance method. A good guidance body requires good inversion of the dip and azimuth information of the formation. Through the calculation and comparison of different parameters, it is found that the algorithm and guidance method have a greater impact on the calculation results, such as Figure 2 , Figure 3 As shown in the figure, the guide body calculated by the BG algorithm depicts the dip and azimuth information of the formation more clearly than that calculated by the FFT algorithm.

[0069] In addition, if Figure 4 As shown, this embodiment selects three different guidance methods: horizontal guidance, center guidance and full guidance. Among them, the full guidance method picks up the most complete and accurate inclination and azimuth information, so in the subsequent steps, the BG algorithm, full guidance method, 3 lines × 3 channels × 3 sampling points aperture extraction guide body are selected for constrained filtering processing.

[0070] At the same time, because the research object is a reservoir, the filtering process should highlight the denoising effect, so the size of the filter cannot be set too large. This embodiment uses a filter with a size of 1 line × 1 channel × 1 sampling point to perform filtering processing on three-dimensional seismic data.

[0071] Step S2: Perform spectrum blueing and frequency extension processing on the three-dimensional seismic data obtained by the full-guided filtering processing in step S1 to improve the resolution of the seismic data for deep clastic reservoirs, laying a foundation for improving the prediction accuracy of thin reservoirs by combining well and seismic data.

[0072] In this step, the spectrum blueing and frequency extension processing is specifically as follows: the three-dimensional seismic data obtained in step S1 is converted into a reflection coefficient by using circular deconvolution, that is, the three-dimensional seismic data is resampled to 0.5ms to generate a new sparse pulse reflection coefficient, which is derived from the maximum and minimum amplitude values ​​in the three-dimensional seismic data; then the spectrum of the seismic signal is matched with the spectrum of the logging curve to design a spectrum blueing operator, and the designed spectrum blueing operator is used to convolve with the three-dimensional seismic data to generate the three-dimensional seismic data after frequency extension.

[0073] Specifically, the 3D seismic data after the full-guided filtering obtained in step S1 is subjected to spectrum blueing and frequency extension processing, in order to improve the resolution of the seismic data for deep clastic reservoirs. Figure 5 , Figure 6 As shown in the figure, through spectrum blueing and frequency extension processing, the main frequency of three-dimensional seismic data is increased from 33Hz to 50Hz, the vertical resolution of reservoir is improved, and the seismic response characteristics are kept good. Taking the average formation velocity of 3300m / s as an example, the vertical resolution limit of a reflection wave to the formation is increased from 25m to 16m. The method of the present invention lays a foundation for improving the prediction accuracy of thin reservoirs by combining well and seismic.

[0074] On the basis of using the BG algorithm full-guided filtering to improve the signal-to-noise ratio of three-dimensional seismic data, the spectrum blueing technology is used to perform frequency extension processing on the seismic data to obtain three-dimensional seismic data with higher signal-to-noise ratio and resolution, and the three-dimensional seismic data is phase-rotated by 90 degrees to participate in the characteristic parameter waveform indication inversion of well-seismic combination (including but not limited to the natural potential parameters used in the embodiment of the present invention), and the threshold value obtained by the intersection analysis of logging rock physical parameters is used to adjust the profile display of the inversion result. On this basis, the top and bottom interfaces are manually drawn and the interlayer attributes are extracted to predict the planar distribution range of the clastic reservoir, and the layer velocity is used to perform time-to-depth conversion on the line-drawn top and bottom interfaces to calculate the top and bottom interface depth difference, and the plane isopach map of the clastic reservoir is obtained by contour mapping.

[0075] Step S3: perform outlier correction and environmental correction on the logging curves of the wells drilled in the study area to ensure the accuracy and reliability of the logging rock physical parameters.

[0076] The purpose of the outlier correction is to ensure the accuracy of the stratum calibration and tracking involved in the establishment of the grid model. The multivariate linear formula needs to be fitted based on multiple parameters in the same stratum segment and adjacent depth segment that have a high correlation with the AC parameters and are less affected by the measurement environment; the logging rock physical parameters involved in the intersection analysis need to be corrected for outliers (including but not limited to the multivariate linear correction of the acoustic time difference (AC) in the embodiment of the present invention) and environmental correction (including but not limited to the baseline correction of the spontaneous potential (SP) curve in the embodiment of the present invention).

[0077] The stratigraphic framework model used in the inversion should be established based on the actual stratigraphic contact relationship; the number of effective samples and the optimal cutoff frequency selected for waveform indication inversion should be determined based on the statistical analysis of known wells, and the optimal cutoff frequency parameter should not differ too much from the maximum effective frequency obtained by analysis, so as to ensure that the inversion result has high resolution and high credibility; complex clastic reservoirs have rapid lateral changes and strong heterogeneity, resulting in large lateral velocity changes. Time-to-depth conversion should be performed through depth calculation of the layer velocity obtained by velocity spectrum calculation to ensure the accuracy of the obtained reservoir plane isopach map.

[0078] Specifically, outlier correction and environmental correction are performed on the logging curves of existing drilled wells. Since the wells selected in this embodiment were drilled at similar times and the curves were measured using the same equipment, the main corrections are made to the acoustic time difference (AC) anomaly caused by wellbore expansion and the natural potential (SP) baseline offset caused by the measurement environment.

[0079] The baseline correction of the SP curve is relatively simple. It is only necessary to correct the baseline of the SP curve of all wells to 0.

[0080] The abnormal correction of AC curve uses the multivariate regression linear correction method. The implementation process of this method is as follows:

[0081] (1) Clarify the stratigraphic information of the wellbore expansion section, select multiple adjacent wells with better borehole quality, and calculate the correlation between the AC curves of the same stratigraphic section and adjacent depth sections and other curves that are less affected by the wellbore expansion;

[0082] (2) Select the three curves with good correlation with the AC curve, namely, the spontaneous potential curve (SP), the induction resistivity curve (RILM), and the neutron curve (CNL), to establish the multivariate linear equation AC = f (SP, RILM, CNL);

[0083] (3) The acoustic wave curve of the wellbore expansion section is obtained by fitting the SP curve, RILM curve, CNL curve and the multivariate linear equation AC = f(SP, RILM, CNL) to replace the abnormal AC curve of the wellbore expansion section and splice it with other AC curve segments of the well to obtain the following: Figure 7 The new AC curve is shown.

[0084] The processed AC curve not only makes the subsequent logging rock physical parameter analysis results more reliable, but also makes the layer calibration more accurate. Figure 8a As shown in Figure 1, the synthetic record of the original synthetic record of the SHB5-1X well at the T33 layer has a poor correlation with the seismic traces near the well. The conventional approach is to perform local tension and compression, and then calibrate the T33 layer at the trough above the strong peak. After AC curve correction (such as Figure 8b As shown in the figure, the synthetic record at the T33 layer has a high correlation with the seismic trace near the well, and there is no need to stretch or compress the synthetic record. The T33 layer is accurately calibrated at the zero phase on the strong wave peak.

[0085] Step S4: performing cross-analysis on the logging rock physical parameters corrected in step S3 to clarify the reservoir sensitive parameters and the threshold values ​​for distinguishing the clastic reservoir from the surrounding rock.

[0086] Specifically, the logging rock physical parameters of the target layer clastic rock and the surrounding rock (mudstone) are cross-analyzed. Due to the overlapping phenomenon of wave impedance, the distinction of lithology is poor. In this embodiment, the natural potential (SP) curve is selected for the subsequent well-seismic combined inversion, and the threshold value for distinguishing clastic rock from mudstone is determined to be -10mv. Clastic rock is less than -10mv, and mudstone is between -10mv-0mv.

[0087] Step S5: Based on the fine calibration and interpretation results of the horizons, an isochronous stratigraphic framework model is constructed, and waveform indication inversion based on reservoir sensitive parameters is performed using the three-dimensional seismic data volume obtained by the 90-degree phase rotation step S2.

[0088] Specifically, based on the fine calibration and tracking of the horizon, the well-seismic combined SP curve waveform indication inversion is carried out to obtain the SP curve inversion volume result. This method is jointly driven by the model and data, and makes full use of the existing data to predict the complex clastic thin reservoir. The prediction accuracy and credibility are higher than those of traditional methods. The specific implementation process is as follows:

[0089] (1) Select the upper and lower adjacent marker layers of the target layer and establish a stratigraphic framework model based on the actual stratigraphic contact relationship;

[0090] (2) Performing a 90-degree phase rotation on the three-dimensional seismic data volume obtained in step S2 to obtain a new three-dimensional seismic data volume. The purpose of this is to give the seismic reflection phase axis a certain geological significance. The change of a complete seismic reflection phase axis represents the response of a single clastic reservoir or the superposition response of multiple reservoirs, rather than the lithology or physical interface information before phase rotation. This makes the correspondence between the seismic reflection phase axis and the spontaneous potential (SP) curve better;

[0091] (3) The waveform indication inversion method is used to invert the SP curve based on the idea of ​​combining well and seismic data. This method not only takes advantage of the high resolution of the well, but also takes advantage of the high lateral resolution of seismic data. The basic idea is to refer to the two factors of waveform similarity and spatial distance when screening statistical samples, and to sort the samples according to the distribution distance on the basis of ensuring the consistency of the sample structure characteristics, so that the inversion results can reflect the constraints of the sedimentary facies in space and be more in line with the sedimentary laws and characteristics in the plane. There are two important parameters that need to be analyzed and selected in the inversion operation process. One is the number of effective samples, which mainly characterizes the degree of influence of the spatial variation of seismic waveforms on the reservoir. The setting of this parameter is mainly based on the statistical results of known wells. Through statistical analysis of "number of samples" and "seismic correlation", the correlation gradually increases with the increase of sample number. After reaching a certain level, the correlation no longer increases with the increase of sample number, indicating that more samples do not help improve the prediction accuracy. The number of samples when the correlation is the largest is the optimal sample parameter. Fig. 9As shown in the figure, the effective sample number is 5 through analysis; the second is the optimal cutoff frequency, which is related to the "effective sample number" parameter. The "optimal cutoff frequency" parameter must be determined after the effective sample number is determined. If you prefer the certainty of the inversion, this parameter should not be set too high. On the contrary, if you prefer the inversion resolution and can accept random results, you can set a higher cutoff frequency. Fig.10 As shown, basically, after the correlation curve enters the horizontal section, it shows that everything except its frequency components is random, and the frequency before entering the horizontal section is the maximum effective frequency. Since step S1 improves the resolution of seismic data, this embodiment only needs to set the maximum effective frequency to the optimal cutoff frequency, which can not only ensure higher recognition accuracy for complex clastic thin reservoirs, but also ensure that the inversion prediction results have a high degree of credibility. Step S6, for key oil-bearing strata, based on the waveform indication inversion results of the reservoir sensitive parameters in step S5, the top and bottom interfaces of the predicted reservoir are line-drawn, and the interlayer attributes are extracted to finely depict the boundaries of the reservoir.

[0092] Specifically, based on the SP inversion volume result obtained in step S5, the section display is performed according to the threshold value determined in step S4, and Fig.11 The top and bottom interfaces of the predicted clastic reservoir S layer are manually drawn in the whole area. The top line drawing layer (dashed line) is named T, and the bottom line drawing layer (solid line) is named B. Then, the interlayer attributes between the T layer and the B layer are extracted to obtain the following: Fig.12 The figure shows the planar distribution of clastic reservoirs and the planar characteristics of SP parameters related to physical properties and fluids.

[0093] Step S7, using the layer velocity to perform time-depth conversion on the top and bottom interfaces of the predicted reservoir after line drawing in step S6, and subtracting the top depth from the bottom depth, and using the difference to generate a reservoir plane isopach map.

[0094] Specifically, due to the complexity of the thin clastic reservoir, there is no strict positive correlation between the reservoir thickness and physical properties. Therefore, the interval velocity of the S layer is extracted based on the velocity spectrum of the study area, and the top and bottom T layers and B layers of the layer are converted into time depths. The depth of the T layer is subtracted from the depth of the B layer to obtain the depth difference between the two layers. The depth difference is used to perform the following operations: Fig.13 The contour line diagram shown in the figure is obtained, and the plane thickness map of the S layer is obtained. Fig.12 The planar distribution characteristics of complex clastic thin reservoirs can be described in detail in terms of reservoir thickness, physical properties and fluids.

[0095] Finally, it should be noted that the above content is only used to illustrate the technical solution of the present invention, rather than to limit the scope of protection of the present invention. Simple modifications or equivalent substitutions of the technical solution of the present invention by ordinary technicians in this field do not deviate from the essence and scope of the technical solution of the present invention.

Claims

1. A method to improve the prediction accuracy of deep clastic thin reservoirs, It is characterized in that Includes steps: S1. For the target deep clastic reservoir section, select the guide body calculation method and filter size, and perform structural guide filtering on the 3D seismic data; S2, performing spectrum bluening and spectrum spreading processing on the three-dimensional seismic data obtained by the full-guided filtering processing in step S1 to obtain a processed three-dimensional seismic data volume; S3, performing outlier correction and environmental correction on the logging curve to obtain the corrected logging rock physical parameters; S4, performing crosstalk analysis on the well logging rock physical parameters after correction processing in step S3; S5. Based on the fine calibration and interpretation results of the horizon, an isochronous stratigraphic framework model is constructed, and waveform indication inversion based on reservoir sensitive parameters is performed using the three-dimensional seismic data volume obtained by the 90-degree phase rotation step S2; S6, based on the waveform indication inversion result of the reservoir sensitive parameter in step S5, line drawing is performed on the top and bottom interfaces of the predicted reservoir, and interlayer attributes are extracted to finely depict the boundary of the reservoir; S7. Use the layer velocity to perform time-depth conversion on the top and bottom interfaces of the predicted reservoir after line drawing in step S6, and subtract the top depth from the bottom depth, and use the difference to generate a reservoir plane isopach map to achieve a detailed prediction of the plane distribution and characteristics of complex clastic rock thin reservoirs.

2. The method according to claim 1, It is characterized in that In step S1, the guide body needs to be calculated before the filtering process, and the calculation of the guide body includes three key parameters: algorithm, aperture size and guide mode.

3. The method according to claim 2, It is characterized in that The algorithm is selected from at least one of a BG algorithm and a FFT algorithm.

4. The method according to claim 3, It is characterized in that The algorithm is the BG algorithm.

5. The method according to claim 2, It is characterized in that The aperture size is 3 lines×3 channels×3 sampling points.

6. The method according to claim 2, It is characterized in that The guiding method is selected from at least one of horizontal guiding, center guiding and full guiding.

7. The method according to claim 6, It is characterized in that The guiding method is a full guiding method.

8. The method according to claim 1, It is characterized in that In step S1, the filter size is 1 line×1 channel×1 sampling point.

9. The method according to claim 1, It is characterized in that In step S2, the spectrum blueing and frequency extension processing is specifically as follows: the three-dimensional seismic data obtained in step S1 is converted into a reflection coefficient by using circular deconvolution, that is, the three-dimensional seismic data is resampled to generate a new sparse pulse reflection coefficient, which is derived from the maximum and minimum amplitude values ​​in the three-dimensional seismic data; then the spectrum of the seismic signal is matched with the spectrum of the logging curve to design a spectrum blueing operator, and the designed spectrum blueing operator is used to convolve with the three-dimensional seismic data to generate the three-dimensional seismic data after frequency extension.

10. The method according to claim 1, It is characterized in that In step S3, the abnormal value correction includes, but is not limited to, correcting the abnormal acoustic time difference parameters caused by wellbore expansion or other factors using a multivariate linear formula.

11. The method according to claim 10, It is characterized in that The abnormal value correction is an abnormal correction of the acoustic wave time difference curve, and a multivariate regression linear correction method is applied. The implementation of this method is as follows: (1) Clarify the stratigraphic information of the wellbore expansion section, select multiple adjacent wells with better borehole quality, and calculate the correlation between the AC curves of the same stratigraphic section and adjacent depth sections and other curves that are less affected by the wellbore expansion; (2) Select three curves with good correlation with the acoustic time difference curve, namely, the natural potential curve, the induction resistivity curve, and the neutron curve, to establish a multivariate linear equation AC=f(SP, RILM, CNL); (3) The acoustic wave curve of the wellbore expansion section is fitted by the natural potential curve, the induction resistivity curve, the neutron curve and the multivariate linear equation AC=f(SP, RILM, CNL) to replace the abnormal acoustic wave time difference curve of the wellbore expansion section, and is spliced ​​with other acoustic wave time difference curve sections of the well to obtain a new corrected acoustic wave time difference curve.

12. The method according to claim 1, It is characterized in that In step S3, the environmental correction includes but is not limited to the baseline correction of the natural potential curve; the baseline correction of the natural potential curve is specifically: the baseline of the natural potential curve is corrected to 0 value.

13. The method according to claim 1, It is characterized in that In step S4, the intersection analysis is specifically to perform an intersection analysis on the logging rock physical parameters of the clastic rock and the surrounding rock in the target layer.

14. The method according to claim 1, It is characterized in that In step S5, the waveform indication inversion based on reservoir sensitive parameters refers to the inversion of the natural potential curve waveform indication combined with well-seismic data on the basis of fine layer calibration and tracking, so as to obtain the natural potential curve inversion volume result. This method is jointly driven by the model and data; specifically: (1) Select the upper and lower adjacent marker layers of the target layer and establish a stratigraphic framework model based on the actual stratigraphic contact relationship; (2) performing a 90-degree phase rotation on the three-dimensional seismic data volume obtained in step S2 to obtain a new three-dimensional seismic data volume; (3) The waveform indication inversion method is used to invert the natural potential curve based on the idea of ​​combining well and seismic data. There are two important parameters that need to be analyzed and selected in the inversion operation process, namely the effective sample number and the optimal cutoff frequency.

15. The method according to claim 14, It is characterized in that The effective sample number mentioned in (3) is set by referring to the statistical results of known wells. The statistical analysis is performed through the sample number and seismic correlation. The correlation gradually increases with the increase of the sample number. The sample number with the maximum correlation is the optimal sample parameter. The optimal cutoff frequency described in (3) is related to the effective sample number parameter. The optimal cutoff frequency parameter needs to be determined after the effective sample number is determined. If the inversion certainty is more preferred, this parameter should not be set too high. On the contrary, if the inversion resolution is more preferred and random results can be accepted, a higher cutoff frequency can be set.

16. The method according to claim 15, It is characterized in that The optimal cut-off frequency, after the relevant curve enters the horizontal section, indicates that everything except its frequency components is random, and the frequency before entering the horizontal section is the maximum effective frequency.

17. The method according to claim 1, It is characterized in that Step S6 is specifically as follows: based on the natural potential inversion volume result obtained in step S5, a profile display is performed through the threshold value determined in step S4, and the top and bottom interfaces of the predicted clastic reservoir S layer are manually drawn in the entire area in combination with the well. The top line drawing layer is named T, and the bottom line drawing layer is named B. Then, by extracting the interlayer properties between the T layer and the B layer, the planar distribution of the clastic reservoir and the planar characteristics of the natural potential parameters related to the physical properties and fluids are obtained.

18. The method according to claim 1, It is characterized in that Step S7, specifically, is as follows: based on the velocity spectrum of the study area, the layer velocity of the S sublayer is extracted, and the top and bottom T layers and B layers of the layer are time-to-depth converted, the depth of the T layer is subtracted from the depth of the B layer to obtain the depth difference between the two layers, and the depth difference is used to make contour lines, thereby obtaining a plane isopach map of the S sublayer, and then combined with the plane isopach map to make a detailed description of the plane distribution characteristics of the complex clastic thin reservoir from the aspects of reservoir thickness, physical properties and fluid.

19. Application of the method according to any one of claims 1 to 18 in clastic reservoir prediction and / or carbonate layered reservoir prediction.

20. A reservoir plane isopach map established by the method according to any one of claims 1 to 18.

Images