Method for rapidly detecting unconformity surface structure by using logging information

Abstract

The invention relates to the technical field of oil exploration, and provides a method for rapidly detecting an unconformity surface structure by using logging information, which comprises the following steps of: 1, determining a stratum contact relationship by combining a tectonic motion history of a to-be-researched area and using reflection response characteristics of seismic data, and identifying an unconformity surface of the to-be-researched area on a seismic section; step 2, identifying and determining the exact depth position of the unconformity surface on a well by using a sound wave time difference method; step 3, after the depth position of the unconformity surface is determined, performing principal component analysis on the logging curve within a certain range of surrounding rocks above and below the unconformity surface; and step 4, classifying the principal component curves by using a K-means clustering method, and further determining an unconformity surface weathered clay layer, semi-weathered rocks and unweathered rocks. According to the invention, the unconformity surface structure can be well and rapidly detected.

Description

Technical Field

[0001] The present invention relates to the technical field of petroleum exploration, in particular to a method for rapidly detecting unconformity surface structure by utilizing well logging data. Background Art

[0002] Unconformities often reflect regional crustal movements, sea (or lake) level fluctuations, or local tectonic events over a specific period. They serve as an important basis for studying geological development and determining the periods of crustal movement. The structural classification of unconformities is quite extensive, with 16 types of spatial structural configurations of unconformities established. Recent research has also demonstrated that the structure of unconformities and their control over oil and gas accumulation hold important geological significance for oil and gas exploration.

[0003] After uplift, erosion, and weathering, unconformity rocks undergo changes in their chemical composition and rock characteristics. This is primarily manifested in two ways. First, physical fragmentation occurs, altering physical characteristics such as loosening of the rock structure and increased porosity. Second, chemical decomposition occurs, leading to the enrichment of Al and Fe ions and an increase in the content of clay minerals like kaolinite.

[0004] Currently, some researchers use outcrop data to classify unconformities. However, outcrop data are often missing or do not correlate well with the underlying strata, making the use of outcrop data to study unconformities in the underlying strata somewhat limited and unintelligent. Other researchers use the logging response characteristics of unconformities to determine unconformity structures. However, this method cannot effectively identify wells with low weathering levels and unclear unconformity logging response characteristics. Furthermore, there is no clear standard for unconformity logging response characteristics, making this method highly subjective.

[0005] Currently, the most common method is to use rock analysis and testing data to calculate the rock weathering index to determine the degree of rock weathering and, therefore, the structure of the unconformity. However, this method requires analytical data, has a long experimental cycle, cannot quickly identify the unconformity structure, and requires coring data for analysis and testing, which is expensive. In the implementation of oil and gas exploration, the proportion of coring wells is also relatively small, and there may be situations where no physical data is available for analysis.

[0006] Other scholars use the principal component analysis method to reduce the dimension of the logging curve to analyze the unconformity surface structure. However, the principal component curve is a dimensionless parameter and has no clear geological meaning. Therefore, some scholars combine the analytical test data with the logging data, calculate the rock weathering index by analyzing the test data to determine the unconformity surface structure of the known well, and then use the principal component analysis curve of the well logging data of the known well combined with the unconformity surface structure determined by the rock weathering index to establish an identification template, and determine the unconformity surface structure of other wells based on the identification template. Among them, the invention patent with publication number CN103744109A, "A method for identifying the weathering crust structure of clastic rocks in areas without well coverage", uses this method. However, this method still requires the use of analytical test data for prior analysis, and only after the identification template is established can it provide a basis for the identification of unconformities in subsequent wells. There are still problems such as long analytical test cycles, few samples, and high costs. Summary of the Invention

[0007] The present invention provides a method for quickly detecting unconformity surface structure using well logging data, which can preferably quickly detect unconformity surface structure.

[0008] According to the present invention, a method for quickly detecting unconformity surface structure using well logging data includes the following steps:

[0009] Step 1: Combined with the tectonic movement history of the area to be studied, the reflection response characteristics of the seismic data are used to determine the stratigraphic contact relationship and identify the unconformity surface of the area to be studied on the seismic section;

[0010] Step 2: Use the acoustic transit time method to identify the exact depth of the unconformity surface in the well;

[0011] Step 3: After determining the depth of the unconformity surface, perform principal component analysis on the well logging curves within a certain range of the surrounding rocks above and below the unconformity surface;

[0012] Step 4: Use the K-means clustering method to classify the principal component curves, and then determine the weathered clay layer, semi-weathered rock and unweathered rock on the unconformity surface.

[0013] Preferably, in step 2, after the unconformity surface undergoes uplift and erosion, the compaction trends of the mudstones in the upper and lower strata are inconsistent, and the exact depth position of the unconformity surface can be determined using the acoustic time difference method.

[0014] Preferably, in step three, the well logging curve is a comprehensive response of the physical and chemical characteristics of the rock. The dimensionality reduction characteristics of the principal component analysis method are used to extract the main information in the well logging curve and then conduct unconformity surface structure research and analysis.

[0015] Preferably, in step three, natural gamma ray GR, acoustic time difference DT, density curve RHOB and deep lateral resistivity RLA5 are selected as logging curves for extracting parameters reflecting the degree of rock weathering.

[0016] As a preference, in step 3, let p-dimensional random vector Vector X k The mathematical expectation is E(X k )=μ k , the variance is Var(X k )=σ k ; The steps for performing principal component analysis on the n-dimensional observation sample matrix of random vector X are as follows:

[0017] 1) Standardization of original data; let the jth observation data of the i-th sampling point be x ij , i=1,2,…,n;j=1,2,…,p, after standardization, we get the normalized matrix Z:

[0018]

[0019] 2) Calculate the correlation coefficient matrix R of the normalized matrix:

[0020]

[0021] 3) Calculate the non-negative eigenvalues and eigenvectors of the correlation coefficient matrix, eigenvalue λ i The corresponding eigenvector C of is:

[0022]

[0023] And meet the following requirements:

[0024]

[0025] C (i) 、C (j) Determine the eigenvectors corresponding to different eigenvalues;

[0026] 4) Calculate the cumulative contribution rate, determine the principal components, and further analyze; the cumulative contribution rate of the first m principal components is When a is close to 1, the first m eigenvectors are selected to generate the principal component Y:

[0027] Y=CZ

[0028] C=(C1 C2 … C m ).

[0029] Preferably, in step 4, the K-means clustering method is specifically as follows:

[0030] (1) Initialize the class center;

[0031] Assume that the sample is data W=(w1 w2 w3...w n ), randomly select K samples from the sample as the center of K classes, denoted as U = (u1 u2 u3...u k ),u k Represents the center of each class; assign n samples in data W to K classes, sample w i The k-th allocation is represented by γ ik ∈{0,1}, that is, if it is assigned to this class, it is 1, and if it is not assigned to this class, it is 0.

[0032] Use the least squares method to determine the sample w i with u k The optimal distance, the criterion function WCSS is written as:

[0033]

[0034] (2) First, fix u k , calculate the sample w i Assign to the u closest to it k γ ik :

[0035]

[0036] (3) Refix γ ik , calculate u k The optimal solution, using the least squares principle, the gradient direction of the criterion function is to differentiate the criterion function, and get u k The analytical expression is:

[0037]

[0038] (4) Repeat steps (2) and (3) until the criterion function converges, that is, the classification of the sample data W is determined.

[0039] The beneficial effects of the present invention are as follows:

[0040] The present invention can use logging data to quickly detect the unconformity surface structure without the need for analysis and testing, so as to further study the control effect of the unconformity surface on oil and gas, thereby accelerating the oil and gas exploration research process.

[0041] This invention uses mathematical transformation methods to improve the current situation where accurate unconformity structure analysis requires analytical and testing data. By integrating the advantages of principal component analysis (PCA) dimensionality reduction and K-means cluster analysis, it not only saves analytical and testing time and accelerates research progress, but also eliminates the need for physical data such as cores, reducing the costs of coring and analytical testing. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a flow chart of a method for rapidly detecting unconformity surface structure using well logging data in an embodiment;

[0043] Figure 2 This is the reflection characteristic map of the unconformity surface in the seismic profile of Panyu 4 North Sag in Xijiang Sag in the embodiment;

[0044] Figure 3 This is the unconformity surface position map determined by the acoustic transit time method for well PY1 in the embodiment;

[0045] Figure 4 This is the principal component analysis cross-plot of Well PY1 in the embodiment;

[0046] Figure 5 This is a cluster analysis diagram of the principal component curve of Well PY1 in the embodiment;

[0047] Figure 6 This is a comprehensive diagram of the unconformity structure analysis of the PY1 well in the example. DETAILED DESCRIPTION

[0048] In order to further understand the content of the present invention, the present invention is described in detail with reference to the accompanying drawings and embodiments. It should be understood that the embodiments are merely for explaining the present invention and are not intended to limit the present invention.

[0049] Example

[0050] like Figure 1 As shown, this embodiment provides a method for quickly detecting unconformity surface structure using well logging data, which includes the following steps:

[0051] Step 1: Combine the tectonic history of the study area with the reflection response characteristics of the seismic data to determine the stratigraphic contact relationships and identify the unconformity surface in the study area on the seismic profile. Compared with outcrop data, seismic reflection data can more intuitively reflect the subsurface stratigraphic contact relationships in the study area and can quickly identify unconformities.

[0052] like Figure 2 The northern part of Panyu 4 Subsag was affected by multiple phases of tectonic movements. The tectonic movements were strong, causing the basin to be uplifted and eroded as a whole, accompanied by magma diapirism. A regional unconformity surface of T80 was developed between the Wenchang Formation and the Enping Formation. The central strata of the northern subsag were significantly uplifted, with strong transformation and obvious truncation characteristics of the interface.

[0053] Step 2: Use the acoustic transit time method to identify and determine the exact depth of the unconformity surface at the wellbore. After the unconformity surface undergoes uplift and erosion, the compaction trends of the mudstones in the upper and lower strata are inconsistent. The acoustic transit time method can be used to determine the exact depth of the unconformity surface.

[0054] like Figure 3The mudstone compaction curve from Well PY1 shows the depth of the T80 interface. The compaction characteristics of the mudstones above and below the interface are clearly two-stage, indicating that this interface is a large-scale unconformity. Above the T80 interface, the Enping Formation to the Neogene exhibits normal compaction. Below the T80 interface, due to factors such as undercompaction of the mudstones, the compaction trend is inconsistent with that above the interface. Based on this principle, the unconformity can be determined to be located at 3272 m.

[0055] Step 3: After determining the depth of the unconformity, principal component analysis is performed on the well log curves within a certain range of the surrounding rock above and below the unconformity. Well log curves represent a comprehensive response to the physical and chemical characteristics of the rock. By utilizing the dimensionality reduction properties of principal component analysis, key information from the well log curves can be extracted for studying and analyzing the unconformity structure. However, principal component curves lack a clear physical meaning and are still unable to effectively identify the unconformity structure.

[0056] Principal component analysis, also known as principal component analysis, is a multivariate statistical method. The ultimate goal of principal component analysis is to use mathematical transformations to reduce dimensionality and simplify data, so as to achieve the goal of using fewer variables to explain most of the variables in the original data.

[0057] Let p-dimensional random vector The mathematical expectation of vector Xk is E(X k )=μ k , the variance is Var(X k )=σ k ; The steps for performing principal component analysis on the n-dimensional observation sample matrix of random vector X are as follows:

[0058] 1) Standardization of original data; let the jth observation data of the i-th sampling point be x ij , i=1,2,…,n;j=1,2,…,p, after standardization, we get the normalized matrix Z:

[0059]

[0060]

[0061] 2) Calculate the correlation coefficient matrix R of the normalized matrix:

[0062]

[0063] 3) Calculate the non-negative eigenvalues and eigenvectors of the correlation coefficient matrix, eigenvalue λ i The corresponding eigenvector C of is:

[0064]

[0065] And meet the following requirements:

[0066]

[0067] C (i) 、C (j) Determine the eigenvectors corresponding to different eigenvalues;

[0068] 4) Calculate the cumulative contribution rate, determine the principal components, and further analyze; the cumulative contribution rate of the first m principal components is When a is close to 1 (a ≥ 90%), select the first m eigenvectors to generate the principal component Y:

[0069] Y=CZ

[0070] C=(C1 C2 … C m ).

[0071] Natural gamma ray (GR), acoustic time difference (DT), density curve (RHOB) and deep lateral resistivity (RLA5) were selected as logging curves to extract parameters reflecting the degree of rock weathering. First, principal component analysis was performed to obtain the eigenvalues and contribution rates of each principal component eigenvector (Table 1). The cumulative contribution rate of the first (PC1), second (PC2) and third (PC3) principal components is 94.19%, which can represent the characteristics of the four logging curves and achieve the purpose of dimensionality reduction. Figure 4 , the principal component crossplot has no clear physical meaning, so it is impossible to determine the weathering degree of rocks at different depths.

[0072] Table 1. Principal component analysis eigenvalues, eigenvectors, and contribution rates of Well PY1

[0073]

[0074]

[0075] Step 4: Use the K-means clustering method to classify the principal component curves, thereby determining the weathered clay layer, semi-weathered rock, and unweathered rock on the unconformity. After dimensionality reduction, the principal component of the well logging curve becomes a dimensionless data, making it impossible to distinguish the unconformity structure. Conventional methods require creating an identification template based on the rock weathering index and principal component curves of wells with known unconformities before identifying unknown wells. After cluster analysis, the identification template is skipped and classification is performed using mathematical transformation methods to directly identify the unconformity structure.

[0076] Cluster analysis is an unsupervised learning method whose core idea is to divide samples in a dataset into several similar groups or "clusters." Common cluster analysis methods include K-means clustering, hierarchical clustering, and spectral clustering. In this example, K-means clustering is used for research. The basic process is as follows:

[0077] (1) Initialize the class center;

[0078] Assume that the sample is data W=(w1 w2 w3...w n ), randomly select K samples from the sample as the center of K classes, denoted as U = (u1 u2 u3...u k ),u k Represents the center of each class; assign n samples in data W to K classes, sample w i The k-th allocation is represented by γ ik ∈{0,1}, that is, if it is assigned to this class, it is 1, and if it is not assigned to this class, it is 0.

[0079] Use the least squares method to determine the sample w i with u k The optimal distance, the criterion function WCSS is written as:

[0080]

[0081] (2) First, fix u k , calculate the sample w i Assign to the u closest to it k γ ik :

[0082]

[0083] (3) Refix γ ik , calculate u k The optimal solution, using the least squares principle, the gradient direction of the criterion function is to differentiate the criterion function, and get u k The analytical expression is:

[0084]

[0085] (4) Repeat steps (2) and (3) until the criterion function converges, that is, the classification of the sample data W is determined.

[0086] The K-means cluster analysis method is used to perform cluster analysis on the principal components, such as Figure 5 , we can quantitatively determine the weathering degree at different depths and then determine the unconformity surface structure. Figure 6 The principal component-cluster analysis method determined that the PY1 well is composed of semi-weathered rocks at a depth of 3272-3274.1m, unweathered rocks below 3274.1m, and a binary structural model of the unconformity surface. This structure is conducive to the lateral and vertical migration of oil and gas.

[0087] This study clarified the unconformity structure of the Panyu 4 North Subsag and confirmed the development of semi-weathered rocks. Based on this understanding, the oil and gas accumulation model of the A48 oilfield was promoted, which guided the drilling of the A481 well and resulted in the oil and gas discovery.

[0088] The above is a schematic description of the present invention and its embodiments, which is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. Therefore, if a person skilled in the art is inspired by this and, without departing from the purpose of the present invention, designs a structure and embodiment similar to this technical solution without inventiveness, they shall fall within the scope of protection of the present invention.

Claims

1. A method for rapidly detecting unconformity structures using well logging data, characterized by: The following steps are involved: Step 1: Combined with the tectonic movement history of the area to be studied, the reflection response characteristics of the seismic data are used to determine the stratigraphic contact relationship and identify the unconformity surface of the area to be studied on the seismic section; Step 2: Use the acoustic transit time method to identify the exact depth of the unconformity surface in the well; Step 3: After determining the depth of the unconformity surface, perform principal component analysis on the well logging curves within a certain range of the surrounding rocks above and below the unconformity surface; Step 4: Use the K-means clustering method to classify the principal component curves, and then determine the weathered clay layer, semi-weathered rock and unweathered rock on the unconformity surface.

2. The method for rapidly detecting unconformity surface structure using well logging data according to claim 1, characterized in that: In step 2, after the unconformity surface undergoes uplift and erosion, the compaction trends of the mudstones in the upper and lower strata are inconsistent. The exact depth position of the unconformity surface can be determined using the acoustic time difference method.

3. The method for rapidly detecting unconformity surface structure using well logging data according to claim 2, characterized in that: In step three, the well logging curve is a comprehensive response of the rock physical and chemical characteristics. The dimensionality reduction characteristics of the principal component analysis method are used to extract the main information in the well logging curve and then conduct unconformity surface structure research and analysis.

4. The method for rapidly detecting unconformity surface structure using well logging data according to claim 3, characterized in that: In step 3, natural gamma ray (GR), acoustic time difference (DT), density curve (RHOB) and deep lateral resistivity (RLA5) are selected as logging curves for extracting parameters reflecting the degree of rock weathering.

5. The method for rapidly detecting unconformity surface structure using well logging data according to claim 4, characterized in that: In step 3, let p-dimensional random vector Vector X k The mathematical expectation is E(X k )=μ k , the variance is Var(X k )=σ k ; The steps for performing principal component analysis on the n-dimensional observation sample matrix of random vector X are as follows: 1) Standardization of original data; let the jth observation data of the i-th sampling point be x ij , i=1,2,…,n;j=1,2,…,p, after standardization, we get the normalized matrix Z: 2) Calculate the correlation coefficient matrix R of the normalized matrix: 3) Calculate the non-negative eigenvalues and eigenvectors of the correlation coefficient matrix, eigenvalue λ i The corresponding eigenvector C of is: And meet the following requirements: C (i) 、C (j) Determine the eigenvectors corresponding to different eigenvalues; 4) Calculate the cumulative contribution rate, determine the principal components, and further analyze; the cumulative contribution rate of the first m principal components is When a is close to 1, the first m eigenvectors are selected to generate the principal component Y: Y=CZ C=(C1 C2…C m )。 6. The method for rapidly detecting unconformity surface structure using well logging data according to claim 5, characterized in that: In step 4, the K-means clustering method is specifically as follows: (1) Initialize the class center; Assume that the sample is data W=(w1 w2 w3...w n ), randomly select K samples from the sample as the center of K classes, denoted as U = (u1 u2 u3...u k ),u k Represents the center of each class; assign n samples in data W to K classes, sample w i The k-th allocation is represented by γ ik ∈{0,1}, that is, if it is assigned to this class, it is 1, and if it is not assigned to this class, it is 0. Use the least squares method to determine the sample w i with u k The optimal distance, the criterion function WCSS is written as: (2) First, fix u k , calculate the sample w i Assign to the u closest to it k γ ik : (3) Refix γ ik , calculate u k The optimal solution, using the least squares principle, the gradient direction of the criterion function is to differentiate the criterion function, and get u k The analytical expression is: (4) Repeat steps (2) and (3) until the criterion function converges, that is, the classification of the sample data W is determined.

Images