Time-lapse seismic reservoir monitoring method based on structure-oriented constraints

Abstract

This invention provides a time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints. The method includes: Step 1, reading seismic data detected at different times and obtaining basic data through seismic data processing; Step 2, calibrating and matching the two periods of data using time-lapse seismic data matching theory to achieve consistency matching of the time-lapse seismic data; Step 3, estimating structural dip information using plane wave decomposition filtering; Step 4, adding structural constraint terms to perform structurally constrained seismic inversion; Step 5, analyzing reservoir differences through the two periods of data and the inversion results to achieve time-lapse seismic reservoir dynamic monitoring. This time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints is simple to operate and improves the computational efficiency and accuracy of time-lapse seismic inversion, providing valuable reference for predicting the distribution of remaining oil.

Description

Technical Field

[0001] This invention relates to the field of seismic dynamic monitoring technology for complex oil and gas reservoirs, and in particular to a time-shifted seismic reservoir dynamic monitoring method based on structurally guided constraints. Background Technology

[0002] With the continuous exploration and development of oil fields, the degree of oil and gas reservoir exploration is constantly increasing. Since the 1980s, many efforts and attempts have been made at home and abroad to improve the oil field recovery rate. However, for some old oil fields with a large degree of exploration and development, the exploration of large structural oil and gas reservoirs is no longer very realistic. The regional water content is constantly increasing, the reservoir parameters are unclear, and the distribution of remaining oil is unknown. Solving these problems from the perspective of development and exploitation technology alone can no longer meet people's needs. Therefore, exploring the fluid change law during the development stage, deepening the understanding of oil and gas reservoirs in the explored areas, and conducting more refined reservoir characterization and description are of great significance for improving the recovery rate in oil and gas exploration and development. This will enable the search for smaller-scale hidden oil and gas reservoirs and the discovery of the distribution law of remaining oil.

[0003] Changes in seismic response characteristics can reflect changes in underground oil reservoirs. During oil and gas field development, the seismic response varies at different stages as the reservoir development progresses. These changes are controlled by the reservoir's static parameters (porosity, etc.) and dynamic parameters (water saturation, etc.). During the development phase, the reservoir's static parameters remain constant, but its dynamic parameters change. Oil production is accompanied by water or gas injection. On the one hand, the extraction of crude oil and the injection of water or gas alter the reservoir's oil saturation, resulting in changes in the seismic response. On the other hand, changes in the reservoir's pressure system during extraction can cause detectable differences in seismic amplitude. Therefore, processing and interpreting seismic data from different periods can effectively reflect the changes associated with oil and gas reservoirs.

[0004] Studies on the amplitude variations of two seismic reflection records began in the 1950s. By the 1970s, research departments of foreign oil companies had begun conducting reservoir comparative studies using two or more seismic observations. In 1982, Acro conducted experimental reservoir seismic testing in the Holt reservoir in Texas, USA, observing differences in seismic responses through multiple time-delayed seismic acquisitions to determine the dynamic changes in the reservoir. However, the high cost at the time severely hampered the application prospects of this technology. With the widespread application of 3D seismic technology, some oilfields now have 3D seismic data acquired at different times, and the analysis of seismic data from different periods has further developed time-shifted seismic technology. Huang et al. proposed seismic data processing constrained by production dynamics information to guide the interpretation and calibration process. Parr et al. conducted time-shifted seismic studies on the Andrew oilfield using production dynamics data, using differences in wave impedance properties to conduct detailed reservoir evaluations and successfully monitor fluid flow direction. Kok et al. conducted theoretical model studies on steam-driven carbonate reservoirs, demonstrating that fluid flow can be studied based on time-shifted seismic profiles.

[0005] Because repeated explorations cannot achieve complete consistency in acquisition systems, near-surface environments, and processing methods, and because changes in oil and gas can lead to variations in seismic propagation velocity and reflectivity, multiple periods of time-lapse seismic data exhibit not only amplitude and waveform disturbances but also discrepancies in seismic wave travel times. To highlight the effective differences truly caused by reservoir development changes, travel time discrepancies in time-lapse seismic data should be eliminated as much as possible, achieving matching processing of seismic data from different periods. High-precision travel time matching is a crucial step in obtaining effective difference information from time-lapse seismic data. After data matching, the reasonable and accurate utilization of amplitude and waveform differences between time-lapse seismic data is another key element in achieving time-lapse seismic exploration, and these differences can be characterized by wave impedance obtained through seismic inversion techniques. Therefore, conducting reasonable time-lapse seismic inversion is particularly important. Time-lapse seismic inversion is an inversion method based on consistent processing of time-lapse seismic data, targeting changes in reservoir dynamic parameters. Seismic inversion technology has always been the most effective bridge between seismic response and reservoir parameters. Applying it to time-lapse seismic exploration can resolve the relationship between seismic response and reservoir property changes, establish the laws governing the changes in reservoir fluid properties over time, conduct dynamic reservoir monitoring, and thus predict remaining oil and improve recovery rates. The essence of time-lapse seismic exploration remains based on the principles of seismic inversion. However, previous studies on time-lapse seismic inversion have been limited by seismic data, resulting in strong uncertainties and poor stability in multi-phase seismic inversions, often leading to ill-conditioned and ill-posed inversion problems.

[0006] Patent application CN110568492B discloses a method for predicting the distribution of remaining oil using time-lapsed seismic data. This method uses an amplitude frequency map as a benchmark to perform consistency correction on other amplitude data, then calculates the relative rate of change of amplitude and the relative rate of change of P-wave impedance based on the amplitude data and P-wave impedance data to predict the distribution of remaining oil. However, many factors can cause relative changes in seismic amplitude, and simple consistency correction using the amplitude frequency map alone cannot eliminate the influence of non-reservoir variations. In contrast, this invention considers the overall time-lapse information of different periods of data during the matching process before inversion, eliminating the influence of non-reservoir information and retaining only the seismic response differences related to changes in reservoir fluid properties.

[0007] Patent application CN110471106B discloses a time-shift seismic inversion method based on filter design. The core of this method is to provide a time-shift seismic inversion approach with constraints based on filter settings. While conventional time-shift seismic inversion techniques utilize regularization to establish constraints and improve the instability of time-shift estimation, this invention, building upon regularization, uses tectonic information as prior information to constrain the inversion process. Real tectonic information helps in studying geological structures and sedimentary environments, reducing inversion ambiguity and providing constraint operators that conform to geological structural characteristics, thus obtaining inversion results that match actual data.

[0008] Patent application CN110187384A discloses a Bayesian time-shifted seismic differential inversion method. This method constructs an inversion objective function by calculating the covariance matrix containing the statistical correlations between multiple elastic parameters and the covariance matrix containing the statistical correlations of the changes in multiple elastic parameters, thereby obtaining the solution expression for the changes in elastic parameters and achieving the optimal inversion result for the changes in elastic parameters. However, this inversion method relies on prior information derived from the statistical correlations of a large amount of well logging data to obtain the covariance matrix, resulting in low utilization of seismic data information and significant plane error. In contrast, the prior constraint information of this invention is based on structural dip information obtained from seismic data using plane wave decomposition. Applying this information to calculations is more adaptable to complex subsurface structures, and the representation of dip information at different angles more realistically reflects the subsurface structure, facilitating more refined reservoir characterization. Furthermore, by constructing constraint terms based on a prediction error filter, invalid information can be filtered out, further reducing the computational load.

[0009] The existing technologies described above are significantly different from this invention. A search reveals no literature in the XY category, indicating the innovativeness of this invention. Since no solution exists in the existing technologies to address the technical problem we seek, we have invented a novel time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints. Summary of the Invention

[0010] The purpose of this invention is to provide a time-shifting seismic reservoir dynamic monitoring method based on structurally guided constraints that can achieve more accurate time-shifting seismic dynamic monitoring.

[0011] The objective of this invention can be achieved through the following technical measures: a time-lapse seismic reservoir dynamic monitoring method based on tectonic guidance constraints, comprising:

[0012] Step 1: Read earthquake data detected at different times and obtain basic data through earthquake data processing;

[0013] Step 2: Use time-lapse seismic data matching theory to calibrate and match the two periods of data to achieve consistent matching of time-lapse seismic data;

[0014] Step 3: Construct tilt angle information using plane wave decomposition filtering estimation;

[0015] Step 4: Add structural constraint terms to perform structurally constrained seismic inversion;

[0016] Step 5: Analyze the differences in reservoirs using data from the two phases and the inversion results to achieve dynamic monitoring of time-shifted seismic reservoirs.

[0017] The objective of this invention can also be achieved through the following technical measures:

[0018] In step 1, input seismic data from different periods within the same work area and analyze the differences in seismic response caused by non-reservoir changes. The two periods of seismic data represent the seismic response of the same area at different times. Due to the influence of different seismic acquisition methods, processing and other non-reservoir change factors, the two periods of seismic data have significant differences in amplitude, frequency and phase.

[0019] Step 2 specifically includes:

[0020] Step 21: Based on the actual data, apply different element reset methods to match the observation system;

[0021] Step 22: Calculate the energy plane ratio of the large time window near the marker layer of the two data periods, and perform overall energy equalization on the two data periods;

[0022] Step 23: Add construction information m and correction coefficient β to the matched filter in the Z-transform domain for time shift correction.

[0023] In step 23, time-shift correction is performed by adding construction information m and correction coefficient β to the matched filter in the Z-transform domain, and the relationship is established as follows:

[0024]

[0025] Where U represents the Z-transform of the plane decomposition filter, F is the polynomial filter, m represents the construction slope; β is the scaling factor, Z1 is the time-shifted Z-transform based on the plane wave decomposition, and Z4 is the Z-transform of the overall displacement of the two data periods; the purpose of time-shift correction is to achieve matching between the two data periods, therefore, a least-squares objective functional is established:

[0026] U(β,m,Z1,Z4)≈0

[0027] The optimization problem described above can be decomposed into linear and nonlinear parts. Given initial values ​​for each part, the inverse problem iterative method is used to solve it until convergence, as detailed below:

[0028] Since the relationship between β and U is linear, the solution is relatively easy, as follows:

[0029]

[0030] The relationship between m and U is non-linear. First, we need to linearize it analytically, as follows:

[0031] U'(β,m0,Z1,Z4)Δk+U(β,m0,Z1,Z4)≈0

[0032] Where Δm is the increment of the local slope, m0 is the initial value of the local slope, and U(β,m,Z1,Z4) is a filter that is differentiated with respect to m. The final correction coefficient can be obtained by cross-iteration of the correction coefficient β and the constructed local slope m. After obtaining the correction coefficient, the two-stage data can be matched and calibrated to obtain the two-stage seismic data after consistency matching.

[0033] In step 3, the construction information is used as prior information to constrain the inversion, and the construction constraint terms are estimated using the plane wave decomposition filter method.

[0034] Step 3 specifically includes:

[0035] Step 31, define a local plane wave differential model of the following form:

[0036]

[0037] Where u(t,x) is the wave field, and σ is the local slope of the continuous space;

[0038] Step 32, transform and Z-transform the above equation to obtain the following form:

[0039]

[0040] Where k = σΔt / Δx, Δx and Δt are the spatial and temporal sampling intervals, and Z x Zt The Z-transform represents the partial derivatives of the wave field along the spatial and temporal directions;

[0041] Step 33, using a filter To approximate time delay operators Where H(Z) t B(1 / Z) is a filter. t ) / B(Z t Let ω be an all-pass digital filter approximating a time-shift operator, where ω is the angular frequency. Combining the above formula, we obtain a nonlinear equation regarding the local slope, which can be expressed in vector form as:

[0042] C(k)d≈0

[0043] Where k is the required construction information, d is the known data, and C is the plane decomposition filter operator; the above equation is analytically linearized, and the linear iterative optimization method is applied to solve it to obtain the construction information of the response, which is used as the constraint term information for subsequent inversion.

[0044] Step 4 includes:

[0045] Step 41: Determine the time-depth relationship and basic elastic parameters using well logging data and seismic-geological interpretation corresponding to the basic data in the work area;

[0046] Step 42: Reconstruct the missing logging curves of the well locations corresponding to the monitoring data using logging or geophysical methods to obtain the changes in their elastic parameters;

[0047] Step 43: Establish two initial models using well logging data and related seismic wavelets. Since there is noise in the actual seismic data and the seismic inversion problem has multiple solutions, in order to obtain a stable and unique approximate solution, the regularization constraint term containing tectonic information in Step 3 is added to the inversion to obtain the inversion result.

[0048] In step 43, the forward model can be written in the following form:

[0049] s = Gx + n

[0050] Where s is the seismic record, G is the forward modeling operator, x is the seismic elastic parameter sequence, and n is the random noise in the seismic data;

[0051] The inversion solution employs regularization constraint terms that incorporate construction information, and introduces an error filter to represent these constraint terms. Based on the solution concept of the inverse problem, the inversion is equivalent to solving the least-squares solution of the following system, as stated below:

[0052]

[0053] Where G is the forward model operator, R(k) is the prediction operator matrix, and R(k) is the regularization operator containing tectonic information; ∈ is a scale parameter; k is the pre-estimated tectonic information, which restricts the seismic inversion to a smooth tectonic transition; m represents the inversion solution, and the result of m is most accurate when the prediction operator error R(k)m approaches 0; its solution can be expressed in the following form:

[0054]

[0055] Structural information can be characterized by dip angle information, as follows:

[0056] θ = arctan -1 k

[0057] The construction information in the prediction error operator R(k)m takes into account the construction characteristics in different directions, and its function with respect to the tilt angle can be expressed as follows:

[0058]

[0059] Where α i These are weighting coefficients used to balance the structural dip angle constraints in different directions, θ i It is the construction tilt angle in the i-th direction; by constructing constraint terms based on the prediction error filter, invalid information can be filtered out, further reducing the amount of computation.

[0060] In step 5, the changes in elastic parameters in the two periods characterize the changes in the reservoir at different times. The seismic inversion results of the two periods are normalized and matched to obtain the difference information and obtain the information on the changes in reservoir elastic parameters over time, so as to realize the dynamic seismic monitoring of the reservoir.

[0061] The objective of this invention can also be achieved through the following technical measures: a time-shifting seismic reservoir dynamic monitoring system based on structural guidance constraints, wherein the time-shifting seismic reservoir dynamic monitoring system based on structural guidance constraints employs a time-shifting seismic reservoir dynamic monitoring method based on structural guidance constraints for reservoir dynamic monitoring.

[0062] The objective of this invention can also be achieved through the following technical measures: a computer-readable storage medium comprising a stored computer program; wherein, when the computer program is running, it controls the device containing the computer-readable storage medium to execute a time-shifted seismic reservoir dynamic monitoring method based on structurally guided constraints.

[0063] The time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints in this invention belongs to the field of oil and gas reservoir dynamic monitoring and involves the problem of time-lapse seismic inversion algorithms. By performing consistency processing on seismic data from different periods, and conducting seismic inversion based on structural constraints, the method obtains the attribute differences of the data, thereby achieving dynamic monitoring of the reservoir and providing a reliable basis and guidance for remaining oil prediction. Research in this field provides important theoretical and technical support for applications such as dynamic development of oil and gas reservoirs and prediction of remaining oil distribution in heavy oil thermal recovery, and its application prospects are very broad.

[0064] This invention fully considers the changes in oil and gas reservoirs at different development stages and eliminates other influencing factors besides reservoir variations in data from different periods. In time-lapse seismic inversion, this invention utilizes real structural information as a constraint term based on plane wave decomposition, providing constraint operators that conform to geological structural characteristics, thereby obtaining inversion results that match actual data. This enables dynamic monitoring of oil reservoirs, is simple to operate, and improves the computational efficiency and accuracy of time-lapse seismic inversion, providing valuable reference for predicting the distribution of remaining oil. Attached Figure Description

[0065] Figure 1 This is a flowchart of a time-shifted seismic reservoir dynamic monitoring method based on structurally guided constraints in a specific embodiment of the present invention;

[0066] Figure 2 This is a schematic diagram of two periods of earthquake data in the same region in a specific embodiment of the present invention;

[0067] Figure 3 This is a schematic diagram of the seismic wave impedance inversion results based on structural guidance constraints in a specific embodiment 1 of the present invention;

[0068] Figure 4 This is a schematic diagram illustrating the differences in time-shifted seismic reservoir monitoring in a specific embodiment 1 of the present invention;

[0069] Figure 5 This is a schematic diagram of the seismic wave impedance inversion results based on structural guidance constraints in a specific embodiment 2 of the present invention;

[0070] Figure 6 This is a schematic diagram illustrating the differences in time-shifted seismic reservoir monitoring in a specific embodiment 2 of the present invention. Detailed Implementation

[0071] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0072] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments of the present invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, and / or combinations thereof.

[0073] like Figure 1 As shown, Figure 1 This is a flowchart of the time-shifted seismic reservoir dynamic monitoring method based on structural guidance constraints of the present invention.

[0074] Step 1: Read earthquake data detected at different times and obtain basic data through earthquake data processing;

[0075] Step 2: Use time-lapse seismic data matching theory to calibrate and match the two periods of data to achieve consistent matching of time-lapse seismic data;

[0076] Step 3: Construct tilt angle information using plane wave decomposition filtering estimation;

[0077] Step 4: Add structural constraint terms to perform structurally constrained seismic inversion;

[0078] Step 5: Analyze the differences in reservoirs using data from the two phases and the inversion results to achieve dynamic monitoring of time-shifted seismic reservoirs.

[0079] In step 1, input seismic data from different periods within the same work area and analyze the differences in seismic response caused by non-reservoir changes. It should be noted that the two periods of seismic data represent the seismic response of the same area at different times. Due to different seismic acquisition methods, processing, and other non-reservoir change factors, the two periods of seismic data have significant differences in amplitude, frequency, phase, etc.

[0080] In step 2, the purpose of matching calibration is to eliminate the influence of acquisition and observation methods, processing and other non-reservoir change factors, and retain only the seismic response differences related to changes in reservoir fluid properties, which can be described as follows.

[0081] First, based on the actual data, different element reset methods are applied to match the observation system.

[0082] Second, calculate the energy plane ratio of the large time window near the marker layer of the two data periods, and perform overall energy balancing on the two data periods.

[0083] Third, time-shift correction is performed by incorporating construction information m and correction coefficient β into the matched filter in the Z-transform domain, establishing the following relationship:

[0084]

[0085] Where U represents the Z-transform of the plane decomposition filter, F is the polynomial filter, and m represents the construction slope. β is the scaling factor, Z1 is the time-shifted Z-transform based on the plane wave decomposition, and Z4 is the Z-transform of the overall displacement of the two data periods. The purpose of time-shift correction is to achieve matching between the two data periods; therefore, a least-squares objective functional is established:

[0086] U(β,m,Z1,Z4)≈0

[0087] The optimization problem described above can be decomposed into linear and nonlinear parts. Given initial values ​​for each part, the inverse problem iterative method is used to solve it until convergence, as detailed below:

[0088] Since the relationship between β and U is linear, the solution is relatively easy, as follows:

[0089]

[0090] The relationship between m and U is non-linear. First, we need to linearize it analytically, as follows:

[0091] U'(β,m0,Z1,Z4)Δk+U(β,m0,Z1,Z4)≈0

[0092] Here, Δm is the increment of the local slope, m0 is the initial value of the local slope, and U(β,m,Z1,Z4) is a filter obtained by differentiating U(β,m,Z1,Z4) with respect to m. The final correction coefficients can be obtained by iteratively solving for the correction coefficients β and the tectonic local slope m. This method can simultaneously obtain the correction coefficients and the tectonic dip angle, and the tectonic dip angle information ensures that the matching correction is performed along the tectonic direction, improving the accuracy of the correction. After obtaining the correction coefficients, the two periods of data can be matched and calibrated to obtain consistent seismic data from both periods.

[0093] In step 3, the seismic inversion problem exhibits multiple solutions, often requiring the addition of prior information as constraints to obtain a stable approximate solution. Real structural information is helpful in studying geological structures and sedimentary environments, facilitating better reservoir description and tracking of underground reservoir distribution. Therefore, structural information can be used as prior information to constrain the inversion. The structural constraint term is estimated using the plane wave decomposition filter method, as described below.

[0094] First, define a local plane wave differential model of the following form:

[0095]

[0096] Where u(t,x) is the wave field and σ is the local slope of the continuous space.

[0097] Second, by transforming and performing a Z-transform on the above equation, we obtain the following form:

[0098]

[0099] Where k = σΔt / Δx, Δx and Δt are the spatial and temporal sampling intervals, and Z x Z t The Z-transform represents the partial derivatives of the wave field along the spatial and temporal directions.

[0100] Third, using filters To approximate time delay operators Where H(Z) t B(1 / Z) is a filter. t ) / B(Z t Let ω be an all-pass digital filter approximating a time-shift operator, where ω is the angular frequency. Combining the above formula, we obtain a nonlinear equation regarding the local slope, which can be expressed in vector form as:

[0101] C(k)d≈0

[0102] Where k is the required construction information, d is the known data, and C is the plane decomposition filter operator. The above equation is analytically linearized, and a linear iterative optimization method is applied to solve it. The resulting construction information is used as the constraint term information for subsequent inversion.

[0103] In step 4, the forward model can be written in the following form:

[0104] s = Gx + n

[0105] Where s is the seismic record, G is the forward modeling operator, x is the seismic elastic parameter sequence, and n is the random noise in the seismic data.

[0106] Due to the presence of noise in seismic data and the multiple solutions inherent in the seismic inversion problem, a regularization constraint term incorporating tectonic information is used to solve the inversion problem in order to obtain a stable and unique approximate solution. An error filter is introduced to represent this constraint term. Based on the solution principle of the inverse problem, the inversion is equivalent to solving the least-squares solution of the following system, as stated below:

[0107]

[0108] Where G is the forward model operator, R(k) is the prediction operator matrix, and R(k) is the regularization operator containing tectonic information; ∈ is a scale parameter; k is the pre-estimated tectonic information, which restricts the seismic inversion to a smooth tectonic profile; m represents the inversion solution, and the result of m is most accurate when the prediction operator error R(k)m approaches 0. The solution can be expressed in the following form:

[0109]

[0110] Furthermore, structural information can be characterized using dip angle information, expressed as follows:

[0111] θ = arctan -1 k

[0112] The construction information in the prediction error operator R(k)m takes into account the construction characteristics in different directions, and its function with respect to the tilt angle can be expressed as follows:

[0113]

[0114] Where α i These are weighting coefficients used to balance the structural dip angle constraints in different directions, θ i It is the structural dip angle in the i-th direction.

[0115] By constructing constraint terms based on prediction error filters, invalid information can be filtered out, further reducing computational load. Incorporating dip angle information makes it more adaptable to complex subsurface structures; the representation of dip angle information at different angles can more realistically reflect the subsurface structure, facilitating more refined reservoir characterization.

[0116] In step 5, the changes in elastic parameters in the two periods characterize the changes in the reservoir at different times. The seismic inversion results of the two periods are normalized and matched to obtain the difference information and obtain the change information of reservoir elastic parameters over time, so as to realize the dynamic seismic monitoring of the reservoir.

[0117] The following are several specific embodiments of the application of the present invention.

[0118] Example 1

[0119] In a specific embodiment 1 of the present invention, the time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints includes the following steps:

[0120] Step 1: Read earthquake data monitored at different times and obtain basic data through earthquake data processing;

[0121] The two sets of input seismic data represent the seismic response of the same region at different times. Due to different seismic acquisition methods, processing, and other non-reservoir variation factors, the two sets of seismic data show significant differences in amplitude, frequency, and phase, such as... Figure 2 As shown.

[0122] Step 2: Based on the time-lapse seismic data matching theory, calibrate and match the two periods of data to achieve consistent matching of time-lapse seismic data;

[0123] The time-shift corrected seismic data is obtained based on the time-shift correction matching theory. The specific processing steps include: matching the observation systems using different element reset methods according to the actual work area data; skipping this step and directly performing energy equalization if the observation systems in the work area are consistent (in this embodiment, the observation systems are consistent); selecting an appropriate time window and calculating the energy plane ratio of the large time window near the marker layer of the two data periods to perform overall energy equalization of the two data periods; adding tectonic information m and correction coefficient β to the Z-transform domain matched filtering for time-shift correction. The purpose of time-shift correction is to achieve matching between the two data periods; therefore, a least-squares objective functional is established.

[0124] U(β,m,Z1,Z4)≈0

[0125] Where U represents the Z-transform of the plane decomposition filter, m represents the construction slope, β is the scaling factor, Z1 is the time-shifted Z-transform based on the plane wave decomposition formula, and Z4 is the Z-transform of the overall displacement of the two data periods.

[0126] The optimization problem described above can be decomposed into linear and nonlinear parts. Given initial values ​​for each part, the inverse problem iterative method is used to solve it until convergence, as detailed below:

[0127] Since the relationship between β and U is linear, the solution is relatively easy, as follows:

[0128]

[0129] The relationship between m and U is non-linear. First, we need to linearize it analytically, as follows:

[0130] U'(β,m0,Z1,Z4)Δk+U(β,m0,Z1,Z4)≈0

[0131] Here, Δm is the increment of the local slope, m0 is the initial value of the local slope, and U(β,m,Z1,Z4) is a filter obtained by differentiating U(β,m,Z1,Z4) with respect to m. The final correction coefficients can be obtained by cross-iteration of the correction coefficients β and the constructed local slope m. After obtaining the correction coefficients, the two sets of data can be matched and calibrated to obtain consistent seismic data for both sets.

[0132] Step 3: Construct tilt angle information using plane wave decomposition filtering estimation;

[0133] Seismic inversion problems often suffer from multiple solutions, requiring the addition of prior information as constraints to obtain stable approximate solutions. Accurate structural information is crucial for studying geological structures and sedimentary environments, facilitating better reservoir description and tracking of underground reservoir distribution. Step 3 uses structural dip information as prior information and constructs the constraint term based on plane wave decomposition theory. The specific processing steps are as follows:

[0134] (1) Define a local plane wave differential model of the following form:

[0135]

[0136] Where u(t,x) is the wave field and σ is the local slope of the continuous space.

[0137] (2) By transforming and Z-transforming the above equation, we obtain the following form:

[0138]

[0139] Where k = σΔt / Δx, Δx and Δt are the spatial and temporal sampling intervals, and Z x Z t The Z-transform represents the partial derivatives of the wave field along the spatial and temporal directions.

[0140] (3) Using filters To approximate time delay operators Where H(Z) t B(1 / Z) is a filter. t ) / B(Z t Let ω be an all-pass digital filter approximating a time-shift operator, where ω is the angular frequency. Combining the above formula, we obtain a nonlinear equation regarding the local slope, which can be expressed in vector form as:

[0141] C(k)d≈0

[0142] Where k is the required construction information, d is the known data, and C is the plane decomposition filter operator. The above equation is analytically linearized, and a linear iterative optimization method is applied to solve it. The resulting construction information is used as the constraint term information for subsequent inversion.

[0143] Step 4: Add structural constraint terms to perform structurally constrained seismic inversion;

[0144] The specific processing steps include: determining the time-depth relationship and basic elastic parameters using well logging data corresponding to the basic data within the work area and seismic-geological interpretation; reconstructing the missing well logging curves corresponding to the well locations in the monitoring data using well logging or geophysical methods to obtain the changes in their elastic parameters; establishing two-stage sub-initial models using well logging data and related seismic wavelets; due to the presence of noise in actual seismic data and the multiple solutions to the seismic inversion problem, in order to obtain a stable and unique approximate solution, the regularization constraint term containing tectonic information from step 3 is added to the inversion to obtain the inversion result, such as... Figure 3 As shown.

[0145] Here, an error filter is introduced to represent the constraint terms. Based on the solution idea of ​​the inverse problem, the inversion is equivalent to solving the least squares solution of the following system, as stated below:

[0146]

[0147] Where s is the seismic record, G is the forward model operator, R(k) is the prediction operator matrix, which contains the regularization operator of tectonic information; ∈ is a scale parameter; k is the pre-estimated tectonic information, which restricts the seismic inversion to change smoothly along the tectonic lines, and m represents the inversion solution. When the prediction operator error R(k)m approaches 0, the result of m is the most accurate.

[0148] Step 5: Analyze the differences in reservoirs using data from the two phases and the inversion results to achieve dynamic monitoring of time-shifted seismic reservoirs.

[0149] The changes in elastic parameters over two periods characterize the reservoir changes at different times. Step 5 normalizes and matches the seismic inversion results from the two periods to obtain the difference information and thus obtain the changes in reservoir elastic parameters over time, achieving dynamic seismic monitoring of the reservoir. Figure 4 As shown.

[0150] Example 2:

[0151] In a specific embodiment 2 of the present invention, a method for time-lapse seismic reservoir dynamic monitoring based on structurally guided constraints is provided, the method including:

[0152] Step 1: Read earthquake data monitored at different times and obtain basic data through earthquake data processing;

[0153] Step 2: Based on the time-lapse seismic data matching theory, calibrate and match the two periods of data to achieve consistent matching of time-lapse seismic data;

[0154] Step 3: Construct tilt angle information using plane wave decomposition filtering estimation;

[0155] Step 4: Add structural constraint terms to perform structurally constrained seismic inversion;

[0156] Step 5: Analyze the differences in reservoirs using data from the two phases and the inversion results to achieve dynamic monitoring of time-shifted seismic reservoirs.

[0157] The two sets of seismic data input in Step 1 represent the seismic response of the same region at different times. Due to different seismic acquisition methods, processing, and other non-reservoir variation factors, the two sets of seismic data show significant differences in amplitude, frequency, and phase, such as... Figure 2 As shown.

[0158] Step 2 involves acquiring time-shift corrected seismic data based on time-shift correction matching theory. The specific processing includes: matching the observation systems using different element reset methods according to the actual work area data; skipping this step and directly performing energy equalization if the observation systems in the work area are consistent (in this embodiment, the observation systems are consistent); selecting an appropriate time window and calculating the energy plane ratio of the large time window near the marker layer of the two data periods to perform overall energy equalization on the two data periods; adding tectonic information m and correction coefficient β to the Z-transform domain matched filtering for time-shift correction. The purpose of time-shift correction is to achieve matching between the two data periods; therefore, a least-squares objective functional is established:

[0159] U(β,m,Z1,Z4)≈0

[0160] Where U represents the Z-transform of the plane decomposition filter, m represents the construction slope, β is the scaling factor, Z1 is the time-shifted Z-transform based on the plane wave decomposition formula, and Z4 is the Z-transform of the overall displacement of the two data periods.

[0161] The optimization problem described above can be decomposed into linear and nonlinear parts. Given initial values ​​for each part, the inverse problem iterative method is used to solve it until convergence, as detailed below:

[0162] Since the relationship between β and U is linear, the solution is relatively easy, as follows:

[0163]

[0164] The relationship between m and U is non-linear. First, we need to linearize it analytically, as follows:

[0165] U'(β,m0,Z1,Z4)Δk+U(β,m0,Z1,Z4)≈0

[0166] Here, Δm is the increment of the local slope, m0 is the initial value of the local slope, and U(β,m,Z1,Z4) is a filter obtained by differentiating U(β,m,Z1,Z4) with respect to m. The final correction coefficients can be obtained by cross-iteration of the correction coefficients β and the constructed local slope m. After obtaining the correction coefficients, the two sets of data can be matched and calibrated to obtain consistent seismic data for both sets.

[0167] Seismic inversion problems often suffer from multiple solutions, requiring the addition of prior information as constraints to obtain stable approximate solutions. Accurate structural information is crucial for studying geological structures and sedimentary environments, facilitating better reservoir description and tracking of underground reservoir distribution. Step 3 uses structural dip information as prior information and constructs the constraint term based on plane wave decomposition theory. The specific processing steps are as follows:

[0168] (1) Define a local plane wave differential model of the following form:

[0169]

[0170] Where u(t,x) is the wave field and σ is the local slope of the continuous space.

[0171] (2) By transforming and Z-transforming the above equation, we obtain the following form:

[0172]

[0173] Where k = σΔt / Δx, Δx and Δt are the spatial and temporal sampling intervals, and Zx and Zt represent the Z-transforms of the wave field's partial derivatives along the spatial and temporal directions.

[0174] (3) Using filters Where H(Z) t B(1 / Z) is a filter. t ) / B(Z t () is an all-pass digital filter that approximates the time-shift operator to approximate the time-delay operator. Where ω is the angular frequency, and combining the above formula, we obtain the nonlinear equation about the local slope, which can be expressed in vector form as:

[0175] C(k)d≈0

[0176] Where k is the required construction information, d is the known data, and C is the plane decomposition filter operator. The above equation is analytically linearized, and a linear iterative optimization method is applied to solve it. The resulting construction information is used as the constraint term information for subsequent inversion.

[0177] Step 4 includes the following specific processing steps: using well logging data corresponding to the basic data within the work area and seismic-geological interpretation to determine the time-depth relationship and basic elastic parameters; reconstructing the missing well logging curves corresponding to the well locations in the monitoring data using well logging or geophysical methods to obtain the changes in their elastic parameters; establishing two-stage sub-initial models using well logging data and related seismic wavelets; due to the presence of noise in actual seismic data and the multiple solutions to the seismic inversion problem, in order to obtain a stable and unique approximate solution, the regularization constraint term containing structural information from Step 3 is added to the inversion to obtain the inversion result. Compared with Example 1, the structural information is obtained using a five-point center filter, and the result is as follows: Figure 5 As shown.

[0178] Here, an error filter is introduced to represent the constraint terms. Based on the solution idea of ​​the inverse problem, the inversion is equivalent to solving the least squares solution of the following system, as stated below:

[0179]

[0180] Where s is the seismic record, G is the forward model operator, R(k) is the prediction operator matrix, which contains the regularization operator of tectonic information; ∈ is a scale parameter; k is the pre-estimated tectonic information, which restricts the seismic inversion to change smoothly along the tectonic lines, and m represents the inversion solution. When the prediction operator error R(k)m approaches 0, the result of m is the most accurate.

[0181] In one embodiment, the changes in elastic parameters over two periods characterize the reservoir changes at different times. Step 5 normalizes and matches the seismic inversion results of the two periods to obtain the difference information and thus obtain the changes in reservoir elastic parameters over time, achieving dynamic seismic monitoring of the reservoir. Figure 6 As shown.

[0182] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0183] Except for the technical features described in the specification, all other technologies are known to those skilled in the art.

Claims

1. A time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints, characterized in that, This time-lapse seismic reservoir dynamic monitoring method based on tectonic-guided constraints includes: Step 1: Read earthquake data detected at different times and obtain basic data through earthquake data processing; Step 2: Use time-lapse seismic data matching theory to calibrate and match the two periods of data to achieve consistent matching of time-lapse seismic data; Step 3: Construct tilt angle information using plane wave decomposition filtering estimation; Step 4: Add structural constraint terms to perform structurally constrained seismic inversion; Step 5: Analyze the differences in reservoirs using data from the two phases and the inversion results to achieve dynamic monitoring of time-shifted seismic reservoirs.

2. The time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints according to claim 1, characterized in that, In step 1, input seismic data from different periods within the same work area and analyze the differences in seismic response caused by non-reservoir changes. The two periods of seismic data represent the seismic response of the same area at different times. Due to the influence of different seismic acquisition methods, processing and other non-reservoir change factors, the two periods of seismic data have significant differences in amplitude, frequency and phase.

3. The time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints according to claim 1, characterized in that, Step 2 specifically includes: Step 21: Based on the actual data, apply different element reset methods to match the observation system; Step 22: Calculate the energy plane ratio of the large time window near the marker layer of the two data periods, and perform overall energy equalization on the two data periods; Step 23: Add construction information m and correction coefficient β to the matched filter in the Z-transform domain for time shift correction.

4. The time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints according to claim 3, characterized in that, In step 23, time-shift correction is performed by adding construction information m and correction coefficient β to the matched filter in the Z-transform domain, and the relationship is established as follows: Where U represents the Z-transform of the plane decomposition filter, F is the polynomial filter, m represents the construction slope; β is the scaling factor, Z1 is the time-shifted Z-transform based on the plane wave decomposition, and Z4 is the Z-transform of the overall displacement of the two data periods; the purpose of time-shift correction is to achieve matching between the two data periods, therefore, a least-squares objective functional is established: U(β,m,Z1,Z4)≈0 The optimization problem described above can be decomposed into linear and nonlinear parts. Given initial values ​​for each part, the inverse problem iterative method is used to solve it until convergence, as detailed below: Since the relationship between β and U is linear, the solution is relatively easy, as follows: The relationship between m and U is non-linear. First, we need to linearize it analytically, as follows: U'(β,m0,Z1,Z4)Δk+U(β,m0,Z1,Z4)≈0 Where Δm is the increment of the local slope, m0 is the initial value of the local slope, and U(β,m,Z1,Z4) is a filter that is differentiated with respect to m. The final correction coefficient can be obtained by cross-iteration of the correction coefficient β and the constructed local slope m. After obtaining the correction coefficient, the two-stage data can be matched and calibrated to obtain the two-stage seismic data after consistency matching.

5. The time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints according to claim 1, characterized in that, In step 3, the construction information is used as prior information to constrain the inversion, and the construction constraint terms are estimated using the plane wave decomposition filter method.

6. The time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints according to claim 5, characterized in that, Step 3 specifically includes: Step 31, define a local plane wave differential model of the following form: Where u(t,x) is the wave field, and σ is the local slope of the continuous space; Step 32, transform and Z-transform the above equation to obtain the following form: Where k = σΔt / Δx, Δx and Δt are the spatial and temporal sampling intervals, and Z x Z t The Z-transform represents the partial derivatives of the wave field along the spatial and temporal directions; Step 33, using a filter To approximate time delay operators Where H(Z) t B(1 / Z) is a filter. t ) / B(Z t Let ω be an all-pass digital filter approximating a time-shift operator, where ω is the angular frequency. Combining the above formula, we obtain a nonlinear equation regarding the local slope, which can be expressed in vector form as: C(k)d≈0 Where k is the required construction information, d is the known data, and C is the plane decomposition filter operator; the above equation is analytically linearized, and the linear iterative optimization method is applied to solve it to obtain the construction information of the response, which is used as the constraint term information for subsequent inversion.

7. The time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints according to claim 6, characterized in that, Step 4 includes: Step 41: Determine the time-depth relationship and basic elastic parameters using well logging data and seismic-geological interpretation corresponding to the basic data in the work area; Step 42: Reconstruct the missing logging curves of the well locations corresponding to the monitoring data using logging or geophysical methods to obtain the changes in their elastic parameters; Step 43: Establish two initial models using well logging data and related seismic wavelets. Since there is noise in the actual seismic data and the seismic inversion problem has multiple solutions, in order to obtain a stable and unique approximate solution, the regularization constraint term containing tectonic information in Step 3 is added to the inversion to obtain the inversion result.

8. The time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints according to claim 7, characterized in that, In step 43, the forward model can be written in the following form: s = Gx + n Where s is the seismic record, G is the forward modeling operator, x is the seismic elastic parameter sequence, and n is the random noise in the seismic data; The inversion solution employs regularization constraint terms that incorporate construction information, and introduces an error filter to represent these constraint terms. Based on the solution concept of the inverse problem, the inversion is equivalent to solving the least-squares solution of the following system, as stated below: Where G is the forward model operator, R(k) is the prediction operator matrix, and R(k) is the regularization operator containing tectonic information; ∈ is a scale parameter; k is the pre-estimated tectonic information, which restricts the seismic inversion to a smooth tectonic transition; m represents the inversion solution, and the result of m is most accurate when the prediction operator error R(k)m approaches 0; its solution can be expressed in the following form: Structural information can be characterized by dip angle information, as follows: θ=arctan -1 k The construction information in the prediction error operator R(k)m takes into account the construction characteristics in different directions, and its function with respect to the tilt angle can be expressed as follows: Where α i These are weighting coefficients used to balance the structural dip angle constraints in different directions, θ i It is the construction tilt angle in the i-th direction; by constructing constraint terms based on the prediction error filter, invalid information can be filtered out, further reducing the amount of computation.

9. The time-lapse seismic reservoir dynamic monitoring method based on structurally guided constraints according to claim 1, characterized in that, In step 5, the changes in elastic parameters in the two periods characterize the changes in the reservoir at different times. The seismic inversion results of the two periods are normalized and matched to obtain the difference information and obtain the information on the changes in reservoir elastic parameters over time, so as to realize the dynamic seismic monitoring of the reservoir.

10. A time-lapse seismic reservoir dynamic monitoring system based on structurally guided constraints, characterized in that, The time-lapse seismic reservoir dynamic monitoring system based on structural guidance constraints uses the time-lapse seismic reservoir dynamic monitoring method based on structural guidance constraints as described in any one of claims 1-9 to perform reservoir dynamic monitoring.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program; wherein, when the computer program is executed, it controls the device on which the computer-readable storage medium is located to perform a time-shifting seismic reservoir dynamic monitoring method based on structurally guided constraints as described in any one of claims 1-9.

Images