High resolution inversion method based on sensitive well logs

Abstract

The present application relates to a kind of high-resolution inversion methods based on sensitive well logging curve, which includes: obtaining well curve data;Determine whether the well curve data is abnormal, to obtain standardized curve data;Analysis and screening standardized curve data, obtain sensitive curve data;Curve reconstruction is carried out to sensitive curve data, form pseudo-acoustic curve data;Pseudo-acoustic curve data is carried out seismic inversion.The present application adds sensitive curve which is more sensitive to lithology to original curve, forms new pseudo-acoustic curve, and carries out seismic inversion, to obtain three-dimensional impedance body that can reflect reservoir, and then facilitate to predict reservoir.

Description

Technical Field

[0001] This invention relates to the field of geophysical inversion, and more particularly to a high-resolution inversion method based on sensitive well logging curves. Background Technology

[0002] Seismic inversion is the process of imaging (solving) the spatial structure and physical properties of underground rock strata using surface seismic observation data and constrained by known geological laws and drilling / logging data. During seismic inversion, artificially generated seismic waves are emitted from a series of points on the ground. These waves propagate underground, and when they encounter a wave impedance interface (i.e., a surface where the wave impedances of upper and lower strata are unequal), reflection occurs, changing the direction of propagation and causing the waves to begin propagating upwards. Receivers are placed at a series of receiving points on the ground to receive these upward-propagating seismic waves.

[0003] Currently, the post-stack acoustic impedance inversion method based on the convolution model is commonly used. However, this method only utilizes acoustic logging information and density logging information, and does not use electrical and radiometric logging data. Therefore, it is difficult to distinguish between subsurface reservoirs and non-reservoirs by directly using this acoustic impedance for inversion, thus failing to effectively identify lithology and affecting the understanding of reservoirs. Summary of the Invention

[0004] The purpose of this invention is to provide a high-resolution inversion method based on sensitive logging curves, which aims to solve the problems of unclear lithology identification, inaccurate reservoir thickness prediction, and low resolution of current conventional impedance inversion methods.

[0005] In a first aspect, embodiments of the present invention provide a high-resolution inversion method based on sensitive logging curves, comprising:

[0006] Acquire well curve data;

[0007] Preprocess the well curve data to obtain standardized curve data;

[0008] Analyze and filter the standardized curve data to obtain the sensitivity curve data;

[0009] The sensitive curve data is reconstructed to form pseudo-acoustic curve data;

[0010] Seismic inversion was performed on the pseudo-acoustic curve data.

[0011] Optionally, the curve reconstruction steps include: extracting sensitive curve data, which includes multiple sensitive curves corresponding to low, medium, and high frequency bands; calculating the matching degree of the sensitive curves in each frequency band; comparing the magnitude of the matching degree, and if the matching degree value is large, mapping the corresponding sensitive curve to the acoustic curve data of the original frequency band to form pseudo-acoustic curve data.

[0012] Optionally, the formula for calculating the degree of matching is:

[0013]

[0014] in, This represents the j-th sensitivity curve in the i-th frequency band. Let J and M represent the m-th sensitivity curve in the i-th frequency band, where J and M both belong to the sensitivity curve set A. Let cor represent the correlation between the sensitivity curves, and x+n represent the target layer segment. P i，j This represents the overall matching degree of the sensitivity curve for the i-th frequency band and the j-th curve.

[0015] Optionally, the steps of seismic inversion include: calibrating fine reservoirs; constraining pseudo-acoustic curve data based on fine reservoirs to construct a geological framework model; and performing seismic inversion in conjunction with the geological framework model.

[0016] Optionally, methods for calibrating fine reservoirs include: selecting a strong seismic reflection interface as a standard layer; using standard wavelets for initial calibration of synthetic seismic records; extracting and statistically analyzing wavelets in well-bypass seismic data volumes based on the wave group relationships corresponding to large strata and performing secondary calibration; and monitoring and adjusting the correspondence of internal reflection wave groups through slowness curves.

[0017] Optionally, it also includes: before performing seismic inversion, constraining the dynamic range of well impedance and performing interpolation and extrapolation based on the established geological framework model to determine the constraint conditions; and during seismic inversion, performing sparse constraint pulse inversion on the geological framework model based on the constraint conditions.

[0018] Optional steps include preprocessing the well curve data.

[0019] Optionally, preprocessing steps include: outlier handling, curve resampling, environmental correction, baseline correction, normalization, and standardization.

[0020] Optionally, the well logging data may be preprocessed, including: environmental correction of the acoustic logging data and density logging data acquired in the enlarged well section; and / or baseline correction of the spontaneous potential logging data.

[0021] Optionally, the preprocessed well curve data can be standardized.

[0022] The beneficial effects of this invention are as follows: by acquiring sensitive curve data that is more sensitive to lithology, and then adding the sensitive curve data to the original acoustic curve data to obtain new pseudo-acoustic curve data, and then performing seismic inversion on the newly obtained pseudo-acoustic curve data, the identification of lithology and the accuracy of reservoir thickness are improved, thereby obtaining a three-dimensional impedance volume that can reflect the reservoir, which facilitates reservoir prediction. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 Histogram of acoustic impedance to distinguish lithology in the study area;

[0025] Figure 2 A flowchart of a high-resolution inversion method based on sensitive logging curves provided in Embodiment 1 of the present invention is shown;

[0026] Figures 3A-3B The figure shows a comparison of the curve preprocessing before and after in Embodiment 1 of the present invention;

[0027] Figures 4A-4B The resistivity-neutron logging cross plot and the resistivity-acoustic impedance cross plot are shown in Embodiment 1 of the present invention.

[0028] Figures 5A-5C A schematic diagram illustrating the fusion of sensitive curve data and original acoustic wave curve in Embodiment 1 of the present invention is shown.

[0029] Figure 6 The pseudo-acoustic curve data distinguishing lithology histogram in Embodiment 1 of the present invention is shown;

[0030] Figure 7 The diagram shows the seismic inversion results of pseudo-acoustic curve data in Embodiment 1 of the present invention. Detailed Implementation

[0031] Preferred embodiments of the invention will now be described in more detail. While preferred embodiments of the invention are described below, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0032] Figure 1 This is a histogram of acoustic impedance to differentiate lithology within the study area. Figure 1It can be seen that the acoustic impedance of the target layer is not sensitive to reservoir differentiation and the overlap is serious. Therefore, a high-resolution inversion method based on sensitive logging curves is proposed. By screening out sensitive logging curves, the sensitive curve information reflecting lithology is added to the acoustic logging curve to form new pseudo-acoustic curve data, and seismic inversion is carried out on it to obtain a three-dimensional impedance volume that can reflect the reservoir, and then reservoir prediction is performed.

[0033] One embodiment of the present invention provides a high-resolution inversion method based on sensitive logging curves. Figure 2 The flowchart of a high-resolution inversion method based on sensitive logging curves provided in Embodiment 1 of the present invention is shown below. Figure 2 This high-resolution inversion method based on sensitive logging curves includes:

[0034] S01: Obtain well curve data;

[0035] S02: Determine whether the well curve data is abnormal in order to obtain standardized curve data;

[0036] S03: Analyze and filter the standardized curve data to obtain the sensitive curve data;

[0037] S04: Reconstruct the sensitive curve data to form pseudo-acoustic curve data;

[0038] S05: Perform seismic inversion on pseudo-acoustic curve data.

[0039] Refer to Figures 3 to 6 for a detailed explanation of the high-resolution inversion method based on sensitive logging curves provided in this embodiment.

[0040] S01, Acquire well logging data. In this embodiment, there are at least three wells in the study area, and well logging data needs to be collected separately for each well in the study area beforehand. Well logging data includes acoustic logging data, density logging data, spontaneous potential logging data, resistivity logging data, etc., corresponding to low, medium, and high frequency bands.

[0041] S02, determine whether the well curve data is abnormal in order to obtain standardized curve data.

[0042] Since the acquired well logging data is unprocessed, it may contain anomalies. Therefore, it is necessary to determine whether the well logging data is abnormal to ensure the rationality of each curve. If necessary, after determining whether the well logging data is abnormal, preprocessing is required. Preprocessing steps include: outlier handling, curve resampling, environmental correction, baseline correction, normalization, and standardization. It is worth noting that for enlarged well sections, environmental correction is required for both sonic logging and density logging data to ensure the correctness of subsequent impedance inversion; additionally, baseline correction is required for spontaneous potential logging data.

[0043] In addition, the logging curves of the target layer need to be standardized, including: selecting a standard layer, which should be a layer with a stable distribution, a certain thickness, and obvious curve characteristics; normalizing the standard deviation and mean of each logging curve on the standard layer to obtain the offset; and superimposing the offset onto the well curve data after the above correction to standardize it. It should be noted that the standardization process for normal well curve data can refer to the standardization process for corrected abnormal well curves, and will not be elaborated here. The standardized abnormal well curve data and normal well curve data are then combined to form standardized curve data.

[0044] refer to Figures 3A-3B Taking resistivity logging curves as an example, a comparison was made before and after curve preprocessing. First, the resistivity was logarithmed to remove outliers, and then standardization was performed on the resulting curves. Figure 3A This is an unprocessed resistivity logging curve. Figure 3B The image shows a preprocessed resistivity logging curve. As can be seen from the comparison, the range of values ​​in the preprocessed resistivity curve is more concentrated.

[0045] S03: Analyze and filter standardized curve data to obtain sensitive curve data.

[0046] In this embodiment, the steps of analyzing and screening standardized curve data include: obtaining well logging interpretation results; extracting electrical and radiometric well logging curve data from the standardized curve data; performing histogram or cross-plot analysis on the electrical and radiometric well logging curve data based on the well logging interpretation results; and screening out lithology-sensitive curve data based on the analysis results. It should be noted that the first processing result needs to be obtained beforehand using conventional inversion methods. Furthermore, the well logging interpretation results are geological results obtained beforehand using conventional inversion methods.

[0047] In practical analysis, taking the cross-plot analysis of resistivity curves and neutron logging curves, and the cross-plot analysis of resistivity curves and sonic impedance curves as examples, a comparison shows that resistivity and neutron logging have a greater degree of reservoir differentiation compared to sonic impedance curves. Figures 4A-4B As shown, it can be seen that selecting well curve data such as electrical properties and radioactivity for histogram or cross plot analysis can yield more sensitive curve data that are more sensitive to lithology.

[0048] S04: Reconstruct the sensitive curve data to form pseudo-acoustic curve data.

[0049] Since acoustic curves cannot effectively identify the research object, it is necessary to reconstruct a pseudo-acoustic curve that reflects the geophysical characteristics of the reservoir using certain mathematical methods for seismic inversion. The reconstruction method adopts a wavelet transform-based reconstruction technique, which specifically includes: acquiring sensitive curve data, which includes multiple sensitive curves corresponding to low, medium, and high frequency bands; calculating the matching degree of the sensitive curves in each frequency band; comparing the magnitude of the matching degree, and if the matching degree is large, mapping the corresponding sensitive curve to the acoustic curve data of the original frequency band to form pseudo-acoustic curve data.

[0050] In this embodiment, when comparing the matching degree, the matching degree of multiple sensitive curves needs to be compared to determine the sensitive curve with the larger matching degree value, and then the sensitive curve is mapped to the original frequency band acoustic wave curve data. In other embodiments, when comparing the matching degree, a threshold can also be set, and the matching degree of multiple sensitive curves can be compared with the threshold. When it is greater than the threshold, the sensitive curve can be mapped to the original frequency band acoustic wave curve data.

[0051] In this embodiment, the formula for calculating the matching degree is:

[0052]

[0053] in, This represents the j-th sensitivity curve in the i-th frequency band. Let J and M represent the m-th sensitivity curve in the i-th frequency band, where J and M both belong to the sensitivity curve set A. Let cor represent the correlation between the sensitivity curves, and x+n represent the target layer segment. P i，j This represents the overall matching degree of the sensitivity curve for the i-th frequency band and the j-th curve.

[0054] It should be noted that during curve reconstruction, to facilitate reservoir inversion targeting geological objectives, acoustic logging data is selected for reconstruction as the low-frequency portion of the pseudo-acoustic logging data; sensitive logging data is selected for reconstruction as the mid-to-high frequency portion of the pseudo-acoustic logging data. Specifically, taking resistivity curves as sensitive logging curves as an example, the high-frequency band of the resistivity curve is fused with the low-frequency band of the acoustic logging curve, such as... Figures 5A-5C As shown in the figure, the solid line represents the original curve, and the dashed line represents the filtered curve. Figure 5A For the sound wave curve, Figure 5B The resistivity curve is shown. Figure 5C The newly obtained pseudo-acoustic curve is shown. The new impedance curve obtained by multiplying the pseudo-acoustic curve by the density provides better reservoir differentiation, and the overlapping regions are significantly reduced. Figure 6 As shown.

[0055] S05: Perform seismic inversion on pseudo-acoustic curve data.

[0056] Specifically, high-resolution seismic inversion is carried out using the pseudo-acoustic curves obtained in step S04. The specific steps are as follows: calibrating the fine reservoir; constraining the pseudo-acoustic curve data based on the fine reservoir to construct a geological framework model; and performing seismic inversion in combination with the geological framework model.

[0057] First, it is necessary to finely calibrate the reservoir to facilitate the subsequent construction of a more accurate geological framework model. Fine-calibration of the reservoir specifically includes: selecting a strong seismic reflection interface as a standard layer; initial calibration of the synthetic seismic record using a standard wavelet; extracting wavelets from the statistical well bypass seismic data volume based on the wave group relationships corresponding to the large set of strata, and performing secondary calibration; and monitoring and adjusting the correspondence of internal reflection wave groups using slowness curves. It should be noted that the standard wavelet can be a Ricker wavelet, a bandpass wavelet, etc.; furthermore, when adjusting the correspondence of internal reflection wave groups, wavelets can be re-extracted using well and seismic data for fine-tuning.

[0058] Secondly, a geological framework model is constructed. This involves interpolating the pseudo-acoustic curve data within each stratum based on pre-calibrated fine reservoirs, resulting in a smooth, closed geological framework model.

[0059] Finally, seismic inversion is performed. Before seismic inversion, it is necessary to constrain the dynamic range of well impedance based on the established geological framework model and perform interpolation and extrapolation to determine the constraint conditions. This facilitates sparse-constrained pulse inversion of the geological framework model based on the constraint conditions during seismic inversion. It should be noted that by performing interpolation and extrapolation, the dynamic range of impedance at each sampling point of each seismic trace to be inverted can be determined, thus providing accurate constraint conditions for the inversion.

[0060] After inversion using the above method, a new acoustic impedance inversion result is obtained. This inversion result can better distinguish between reservoirs and non-reservoirs, facilitating accurate reservoir prediction. (Reference) Figure 7 As can be seen from the figure, the inversion results can basically distinguish between reservoirs and non-reservoirs, and the effect is good, which ultimately proves the effectiveness of the method.

[0061] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments.

Claims

1. A high-resolution inversion method based on sensitive well logging curves, characterized in that, include: Acquire well curve data; Determine whether the well curve data is abnormal in order to obtain standardized curve data; The standardized curve data is analyzed and filtered to obtain the sensitive curve data; The sensitive curve data is reconstructed to form pseudo-acoustic curve data; Seismic inversion is performed on the pseudo-acoustic curve data; The process of reconstructing the sensitive curve data includes: Extract the sensitive curve data, which includes multiple sensitive curves corresponding to low, medium, and high frequency bands; Calculate the degree of matching of the sensitivity curves in each frequency band; The matching degree is compared. If the matching degree is large, the corresponding sensitive curve is mapped to the acoustic curve data of the original frequency band to form pseudo-acoustic curve data. The formula for calculating the matching degree is: in, This represents the j-th sensitivity curve in the i-th frequency band. Let J and M represent the m-th sensitivity curve in the i-th frequency band, where both J and M belong to the sensitivity curve set A. Let cor represent the correlation between the sensitivity curves, and x+n represent the (x+n)-th sampling point within the target layer. P i，j This represents the overall matching degree of the sensitivity curve for the i-th frequency band and the j-th curve.

2. The high-resolution inversion method based on sensitive logging curves according to claim 1, characterized in that, Seismic inversion is performed on the pseudo-acoustic curve data, including: Detecting fine reservoirs; Based on the detailed reservoir, the pseudo-acoustic curve data is constrained to construct a geological framework model; Seismic inversion was performed using the aforementioned geological framework model.

3. The high-resolution inversion method based on sensitive logging curves according to claim 2, characterized in that, The method for calibrating fine reservoirs includes: Select the interface with stronger seismic reflection as the standard layer; Initial calibration of synthetic seismic records using standard wavelets; Based on the wave group relationship corresponding to the large set of strata, the wavelets in the well bypass seismic data volume are extracted, statistically analyzed, and then calibrated twice. The correspondence between the internal reflected wave groups is adjusted by monitoring the slowness curve.

4. The high-resolution inversion method based on sensitive logging curves according to claim 2, characterized in that, Before performing seismic inversion using the geological framework model, the following steps are also included: Based on the established geological framework model, the dynamic range of well impedance is constrained and interpolated / extrapolated to determine the constraint conditions. The process of performing seismic inversion using the geological framework model also includes: Based on the constraints, sparse-constrained pulse inversion is performed on the geological framework model.

5. The high-resolution inversion method based on sensitive logging curves according to claim 1, characterized in that, The well curve data is preprocessed.

6. The high-resolution inversion method based on sensitive logging curves according to claim 5, characterized in that, The preprocessing steps include: outlier handling, curve resampling, environmental correction, baseline correction, normalization, and standardization.

7. The high-resolution inversion method based on sensitive logging curves according to claim 5, characterized in that, The well curve data is preprocessed, including: Environmental correction is performed on the acoustic logging and density logging data acquired in the enlarged well section; and / or, Baseline correction was performed on the spontaneous potential logging curve data.

8. The high-resolution inversion method based on sensitive logging curves according to claim 5, characterized in that, The preprocessed well curve data is standardized.

Images