A method and device for shale sequence stratigraphic division based on sediment noise simulation

Abstract

The present invention discloses a method and device for shale sequence stratigraphic division based on sediment noise simulation, including: (1) selecting various logging curves, performing preprocessing and then spectral analysis to identify astronomical orbital periods; (2) performing Gaussian band-pass filtering on the thickness corresponding to the astronomical orbital periods in the logging curves, selecting the filtered curves to establish a time-depth conversion model, and obtaining the time-domain sequences of various logging curves; (3) calculating the time-domain sequences of various logging curves through a sediment noise model to obtain several different relative sea / lake level change curves; (4) performing principal component analysis on the median curve of several different relative sea / lake level change curves to extract the maximum principal component PC1 curve; (5) identifying the maximum and minimum values at the 1200 kyr or 2400 kyr periods of the PC1 curve, and identifying the maximum sea / lake flooding surface and sequence stratigraphic interface. The present invention can solve the problem of constructing an isochronous sequence stratigraphic framework for mudstone shale in the prior art.

Description

Technical Field

[0001] The present invention belongs to the technical field of shale formation information processing, and in particular relates to a method and device for dividing mud shale sequence stratigraphy based on sediment noise simulation. Background Art

[0002] The division and construction of traditional sequence stratigraphy mainly rely on the identification of "sedimentary sequences" defined by unconformity surfaces and corresponding conformity surfaces. For mud shale sequences with weak intuitive lithological changes and continuous deposition, it is difficult to identify the sequence boundaries within the mud shale. Usually, the mud shale section is treated as a marine (lake) transgressive system tract within a third-order sequence, making it difficult to conduct fine horizon calibration and analysis within it.

[0003] With the progress of unconventional shale oil and gas exploration and development, several methods for dividing mud shale sequence stratigraphy have emerged in recent years: genetic sequence stratigraphy division method, gamma energy spectrum cycle analysis method, chemical sequence stratigraphy method, etc.

[0004] For example, the Chinese patent document with the publication number CN107817260A discloses a method for identifying high-frequency sequences of mud shale, and the steps are as follows: obtaining core element geochemical information; selecting parameters that can effectively reflect lithological and paleoenvironmental changes to make data curves; performing sliding average processing on the data curves at different scales; analyzing the relationship between the data curves after sliding average processing and each sequence, and combining the cyclicity of the data curves to identify sequences at different scales.

[0005] The Chinese patent document with the publication number CN107515957A discloses a method for dividing mud shale sequence stratigraphy, including: extracting core samples and obtaining element data and fabric characteristics; analyzing the environmental evolution characteristics such as paleoclimate, paleo water depth, paleosalinity, and paleo redox according to the characteristics of the elements; determining the characteristics of parasequence set boundaries according to the sequence cycle standard of sequence stratigraphy; establishing the correspondence between the natural gamma curve and the parasequence set and sequence boundaries, establishing a stratigraphic framework profile; performing sequence stratigraphy division in the whole area, drawing a planar distribution map of sequence stratigraphy, and analyzing the distribution law of sequence stratigraphy.

[0006] However, current methods mostly identify boundaries and sequence units based on the change patterns (mutations, inflection points, upward and downward trend changes, etc.) of data sequences. The influence of subjective human division factors on the correct identification of sequence boundaries is relatively large, and it is difficult to achieve the division of "isochronous" stratigraphic cells. Summary of the Invention

[0007] The present invention provides a method and device for dividing mud shale sequence stratigraphy based on sediment noise simulation, which solves the problem of constructing an isochronous sequence stratigraphic framework for mudstone shale in the prior art.

[0008] A method for dividing shale sequence stratigraphy based on sediment noise simulation, comprising the following steps:

[0009] (1) Select various logging curve data for preprocessing to make the data equally spaced and remove background white noise; perform spectral analysis on the preprocessed data to identify astronomical orbital periods;

[0010] (2) Perform Gaussian band-pass filtering on the thickness corresponding to the astronomical orbital period in the logging curve, select the filtered curve to establish a time-depth conversion model, and obtain the time-domain sequences of various logging curves based on this time-depth conversion model;

[0011] (3) Calculate the time-domain sequences of various logging curves through a sediment noise model to obtain several different relative sea / lake level change curves;

[0012] (4) Perform principal component analysis on the median curve of several different relative sea / lake level change curves, and extract the maximum principal component PC1 curve;

[0013] (5) Filter the PC1 curve on a scale of 1200 kyr or 2400 kyr, and identify the maximum sea / lake flooding surface and sequence stratigraphic interface at the maximum and minimum values of the filtered curve.

[0014] Preferably, in step (1), the selection of various logging curve data includes: natural gamma curve, resistivity curve, density curve, spontaneous potential curve, and porosity curve.

[0015] In step (2), the identified astronomical orbital period is the long eccentricity period, and the relationship between this period and the corresponding thickness should satisfy the estimation of the formation sedimentation background and sedimentation rate. The central frequency participating in the filtering calculation is the reciprocal of this thickness.

[0016] Preferably, in step (2), the reciprocal of the thickness in the natural gamma data is used as the central frequency of the filtering. When establishing the time-depth conversion model, a time-depth conversion model for all logging curves is uniformly established according to the filtered curve obtained from the natural gamma.

[0017] Due to the slight difference in the formation thickness corresponding to the long eccentricity period preserved in different logging sequences, usually, the thickness period in the natural gamma data should be referred to as the central frequency of the filtering, and a time-depth conversion model for all logging sequences is uniformly established for the filtered curve of the natural gamma logging sequence.

[0018] The specific process of step (3) is:

[0019] (3-1) After importing the time-domain sequences of various logging curves into the Matlab program Acycle software package, a window size of 300 - 500 thousand years, or a window of 400 thousand years, can be selected for comparison with the long eccentricity period. A smaller running window will result in higher resolution, but at the same time, the low-frequency period variance and the total variance will decrease, increasing the uncertainty in the estimation of the non-astronomical signal ratio.

[0020] (3-2) Determine the relative change in water depth by detecting the noise in the time-domain sequence;

[0021] Among them, the noise refers to the non-orbital period signal in the time-domain sequence, which is obtained during the process of detecting the astronomical orbital periods in the target time-domain sequence; the target orbital parameters during the detection process are the long eccentricity period, short eccentricity period, slope period, and precession period respectively. The long eccentricity period corresponds to 405 thousand years, the short eccentricity period corresponds to 125 or 95 thousand years, the slope period corresponds to 40 thousand years or shorter, and the precession period corresponds to 23.6, 22.3, 19.1 thousand years or shorter;

[0022] (3-3) Assume that the noise distribution conforms to a normal distribution. Based on Monte Carlo simulation, compare the astronomical cycle signals and noise saved in the time-domain sequence with the astronomical orbital periods, analyze the contribution of the signals related to water depth saved in the time-domain sequence to the astronomical periods, use the variance of the noise as an index of the relative sea-level change, calculate the ratio of the non-orbital signal variance to the total variance in a sliding time window, and perform more than 1500 iterations of calculation to obtain the sediment noise model of each logging curve, and use the median line of the model as the relative sea / lake-level change curve.

[0023] In step (3-2), the noise described includes dating error, sampling error, differential compaction, unstable sedimentation, and the influence of later diagenesis.

[0024] In step (5), the central frequency used for filtering on the 1200 thousand-year scale is 8.33×10 -4 (the reciprocal of 1200 thousand years), corresponding to the ice ages in the geological history period (Cenozoic, Late Ordovician, Late Permian, etc.), and the central frequency used for filtering on the 2400 thousand-year scale is 4.16×10 -4 (the reciprocal of 2400 thousand years), corresponding to the greenhouse periods in the geological history period (Cretaceous, Early Triassic, Late Cambrian, etc.).

[0025] The present invention also provides a device for shale sequence stratigraphic division based on sediment noise simulation, including a memory and one or more processors. Executable code is stored in the memory, and when the one or more processors execute the executable code, it is used to implement the above-mentioned shale sequence stratigraphic division method.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] 1. Based on the Milankovitch theory, the present invention converts deep-domain data into time-domain data, providing a basic unit with time significance for the isochronous division of sequence stratigraphy.

[0028] 2. Based on sediment noise simulation, by calculating the variance between the noise and the total energy, the present invention plots the relative sea (lake) level change curve; since the noises preserved in different logging data are different, the maximum principal component in each sediment noise curve can be identified according to the principal component analysis, and the water depth change it reflects is closer to the real water depth change. According to the extreme value of the change of the filtering curve of the maximum principal component curve at a specific frequency, the sequence interface and flooding surface inside the shale can be visually identified, thus realizing the division of the isochronous stratigraphic framework. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0030] Figure 1 It is a schematic flow chart of a method for dividing the sequence stratigraphy of shale based on sediment noise simulation provided by the present invention;

[0031] Figure 2 It is a spectrum analysis diagram of the natural gamma curve after preprocessing of a certain well in the embodiment of the present invention;

[0032] Figure 3 It is a sediment noise model of various time-domain logging curves of a certain well in the embodiment of the present invention;

[0033] Figure 4 It is the principal component analysis result of the median curve of various sediment noise models of a certain well in the embodiment of the present invention;

[0034] Figure 5 It is the PC1 curve, PC1 filtering curve and sequence stratigraphy division result of a certain well in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0035] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0036] As Figure 1 shown, a method for dividing the shale sequence stratigraphy based on sediment noise simulation includes the following steps:

[0037] S101, Preprocess various selected logging curves (natural gamma, resistivity, density, spontaneous potential, porosity) to make the data equally spaced and remove background white noise, import them into the open-source software Matlab for spectral analysis, and identify the astronomical orbital periods.

[0038] Taking the natural gamma logging curve as an example, the selected natural gamma logging curve is processed such as equally spacing and removing background white noise, imported into the open-source software for spectral analysis, and the corresponding astronomical orbital period is identified (as Figure 2 shown, the test interval of the frequency is 0 - 1.2m -1 . The numbers in the figure are the reciprocals of the corresponding abscissas (frequencies), that is, the thickness periods. According to the ratio relationship of each period of the astronomical orbital parameters and the ratio relationship of each thickness, it can be inferred that the 21.27m thickness period corresponds to the long eccentricity period (405 thousand years), and the estimated average sedimentation rate meets the judgment of the regional geological background (sedimentary environment).

[0039] S102, Perform Gaussian band-pass filtering on the depths corresponding to the astronomical orbital periods identified, select a suitable filtering curve according to the number of sine waves in the filtering curve and the preliminary judgment of the formation sedimentation background and sedimentation rate, establish a time-depth conversion model, and obtain the time-domain sequences of various logging curves (natural gamma, resistivity, density, spontaneous potential, porosity) based on the time-depth conversion model.

[0040] Perform Gaussian band-pass filtering on the depth corresponding to the astronomical orbital period identified, that is, 21.27m, and the center frequency of the filtering is selected as 1 / 21.27, that is, 0.047m -1 . Figure 2 The number of sine waves in the filtering curve in

[0041] is 15, the single-well sedimentation thickness is 298m, and its sedimentation rate is consistent with the preliminary judgment in step S101. Therefore, a time-depth conversion model can be established according to this filtering curve, that is, one depth period corresponds to 405 thousand years. Based on this time-depth conversion model, the time-domain sequences of each logging curve are obtained.

[0042] Taking the time-domain natural gamma sequence as an example, calculate the relative sea-level change curve by the sediment noise model for the time-domain natural gamma sequence. The specific process is as follows:

[0043] ① After importing the data, the window size is selected as 300 - 500 ka, which can be compared with the long eccentricity period.

[0044] ② By detecting the signal without astronomical orbital control in the sequence, i.e., noise, to discriminate the relative change of water depth. The detection of noise mainly depends on the detection of astronomical orbital periods in the target time-domain data. Among them, the target layer section is the Early Carboniferous, so the selected target orbital parameters are: long eccentricity (405 ka), short eccentricity (125 and 95 ka), slope (38 ka and 32 ka), precession (21.4, 20.4 and 18).

[0045] ③ Based on Monte Carlo simulation, compare the signals and noise preserved in the time-domain data with the astronomical cycle signals, analyze the contribution of the signals related to water depth preserved in the time-domain sequence to the astronomical cycle. The variance of the noise can be used as an indicator of the relative sea-level change. Calculate the ratio of the non-orbital signal variance to the total variance in the sliding time window and perform 2000 iterative calculations to obtain the sediment noise model of each logging curve (as Figure 3 shown), where the median line can indicate the relative water depth change.

[0046] S104, perform principal component analysis on the relative water depth change curves of various loggings and extract the maximum principal component PC1.

[0047] Perform principal component analysis on the median lines of the sediment noise models of the above logging curves (natural gamma, resistivity, density, spontaneous potential, porosity). Five principal components can be extracted from the five median lines, and the maximum principal component PC1 accounts for 62.57% (as Figure 4 shown), which is selected as the curve that can best indicate the relative water depth change of this well (as Figure 5 shown in a).

[0048] S105, identify the maximum and minimum values in the 1200 or 2400 kyr cycle of the PC1 curve, and divide the internal sequence boundaries and flooding surfaces of the shale.

[0049] Since the selected target layer section is mainly developed during the ice age, therefore, perform filtering processing on PC1 with a central frequency of 8.33×10 -4 to obtain the filtered curve (as Figure 5 shown in b). According to the positions of the maximum and minimum values of the filtered curve, divide the sequence boundaries and the maximum flooding surface, so as to construct the internal sequence stratigraphic framework of the mud shale (as Figure 5 shown in c).

[0050] The above-described embodiments have elaborated in detail the technical solutions and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the present invention. Any modification, supplement, and equivalent replacement made within the scope of the principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for dividing shale sequence stratigraphy based on sediment noise simulation, characterized by It includes the following steps: (1) Select various logging curve data for preprocessing to make the data equally spaced and remove background white noise; Perform spectral analysis on the preprocessed data to identify astronomical orbital periods; (2) Perform Gaussian band-pass filtering on the thickness corresponding to the astronomical orbital period in the logging curve, select the filtered curve to establish a time-depth conversion model, and obtain the time-domain sequences of various logging curves based on this time-depth conversion model; (3) Calculate the time-domain sequences of various logging curves through a sediment noise model to obtain several different relative sea / lake level change curves; the specific process is as follows: (3-1) After importing the time-domain sequences of various logging curves into the Matlab program Acycle software package, select a window size of 300 - 500 thousand years; (3-2) Determine the relative change in water depth by detecting the noise in the time-domain sequence; Among them, the noise refers to the non-orbital period signal in the time-domain sequence, which is obtained during the process of detecting the astronomical orbital period in the target time-domain sequence; the target orbital parameters during the detection process are the long eccentricity period, short eccentricity period, slope period, and precession period respectively. The long eccentricity period corresponds to 405 thousand years, the short eccentricity period corresponds to 125 or 95 thousand years, the slope period corresponds to 40 thousand years, and the precession period corresponds to 23.6, 22.3, or 19.1 thousand years; (3-3) Assume that the noise distribution conforms to a normal distribution. Based on Monte Carlo simulation, compare the astronomical period signal and noise saved in the time-domain sequence with the astronomical orbital period, analyze the contribution of the signal related to water depth saved in the time-domain sequence to the astronomical period, use the variance of the noise as an index of relative sea level change, calculate the ratio of the variance of the non-orbital period signal to the total variance in a sliding time window, and perform iterative calculations more than 1500 times to obtain the sediment noise model of each logging curve, and use the median line of the model as the relative sea / lake level change curve; (4) Perform principal component analysis on the median curves of several different relative sea / lake level change curves to extract the maximum principal component PC1 curve; (5) Filter the PC1 curve on a scale of 1200 thousand years or 2400 thousand years, and identify the maximum sea / lake flooding surface and sequence stratigraphic interface at the maximum and minimum values of the filtered curve.

2. The method for dividing shale sequence stratigraphy based on sediment noise simulation according to claim 1, wherein In step (1), the selection of various logging curve data includes: natural gamma curve, resistivity curve, density curve, spontaneous potential curve, and porosity curve.

3. The method for dividing the shale sequence stratigraphy based on deposition noise simulation according to claim 1, wherein In step (2), the identified astronomical orbital period is the long eccentricity period, and the relationship between the thickness corresponding to this period should satisfy the estimation of the formation sedimentation background and sedimentation rate. The central frequency participating in the filtering calculation is the reciprocal of this thickness.

4. The method for shale sequence stratigraphic division based on sediment noise simulation according to claim 1, wherein In step (2), use the reciprocal of the thickness in the natural gamma data as the central frequency of the filtering. When establishing the time-depth conversion model, based on the filtered curve obtained from the natural gamma, establish the time-depth conversion model for all logging curves uniformly.

5. The method for dividing the shale sequence stratigraphy based on sediment noise simulation according to claim 1, wherein, In step (3-2), the noise mentioned includes dating error, sampling error, differential compaction, unstable sedimentation, and the influence of late diagenesis.

6. The method for dividing shale sequence stratigraphy based on sediment noise simulation according to claim 1, wherein In step (5), the central frequency used for the 1200 kyr scale filtering is 8.33×10 -4 , corresponding to the ice age in the geological history period, and the central frequency used for the 2400 kyr scale filtering is 4.16×10 -4 , corresponding to the greenhouse period in the geological history period.

7. A device for dividing the sequence stratigraphy of mud shale based on sediment noise simulation, characterized in that, Comprising a memory and one or more processors, wherein executable code is stored in the memory, and when the one or more processors execute the executable code, it is used to implement the shale sequence stratigraphic division method described in any one of claims 1-6.

Images