Carbonate overpressure formation pressure prediction method and device based on cluster analysis

Abstract

The application provides a carbonate overpressure formation pressure prediction method and device based on cluster analysis, which comprises the following steps: extracting a characteristic item extraction value according to a cause characteristic extraction based on multi-source characteristic parameters; determining a corresponding target pressure prediction model and a cause weight distribution according to the characteristic item extraction value by dynamic weighted cluster analysis based on a cause determination rule; and generating a target formation pressure prediction value by pressure prediction and reverse deduction verification according to the multi-source characteristic parameters, the characteristic item extraction value and the cause weight through the target pressure prediction model, so as to realize rapid automatic identification of overpressure causes and self-adaptive matching of the model, significantly improve the data processing efficiency, accurately depict the coupling effect of multi-mechanism overpressure by adopting a cause weight distribution, multi-model collaborative prediction and reverse deduction iteration correction mechanism, and greatly reduce the prediction deviation, thereby maintaining stable and reliable prediction performance in a carbonate formation with strong heterogeneity and various overpressure causes.

Description

Technical Field

[0001] This invention relates to the field of carbonate rock overpressure formation pressure prediction technology, and in particular to a method and apparatus for predicting carbonate rock overpressure formation pressure based on cluster analysis. Background Technology

[0002] Carbonate reservoir overpressure prediction is a crucial aspect of oil and gas exploration and development, and its accuracy directly impacts deep well drilling cost control and safety boundaries. As oil and gas exploration progresses into deeper and ultra-deeper formations, formation pressure prediction models are shifting from traditional "post-hoc interpretation" to "intelligent early warning." However, carbonate reservoirs are highly heterogeneous, with complex overpressure formation mechanisms, and hidden overpressure bodies are difficult to identify. Inaccurate predictions can easily trigger drilling accidents such as well kicks and lost circulation.

[0003] Currently, overpressure prediction in carbonate rocks largely relies on manual interpretation or single genetic models. Manual interpretation typically takes over four hours, making it difficult to meet the real-time early warning requirements of drilling operations. Conventional prediction models often only simulate a single overpressure genetic type, failing to effectively integrate the synergistic effects of multiple overpressure mechanisms, leading to significant prediction biases. For example, patent document CN 110297280A proposes a method for predicting the spatial distribution characteristics of overpressure in carbonate rocks, achieving three-dimensional overpressure prediction by constructing an overpressure calculation model and combining it with geophysical methods. This method has certain advantages in spatial distribution characterization, but it still mainly relies on statistical modeling and combinations of geological parameters, making it difficult to address the prediction needs of complex genetically coupled strata.

[0004] In summary, existing methods for predicting overpressure in carbonate rocks suffer from prominent problems such as poor timeliness, weak coupling ability of genetic mechanisms, and low prediction accuracy. They are unable to effectively address the drilling risk early warning needs of complex overpressure formations, and there is an urgent need to develop new methods that can integrate multiple overpressure causes and achieve rapid and accurate prediction.

[0005] This section is intended to provide background or context for the embodiments of this application set forth in the claims. The description herein is not an admission that it is prior art simply because it is included in this section. Summary of the Invention

[0006] One objective of this invention is to provide a method for predicting overpressure in carbonate rock formations based on cluster analysis. This method achieves rapid and automatic identification of overpressure genesis and adaptive model matching through dynamic weighted clustering. While significantly improving data processing efficiency, it employs a generative weight allocation, multi-model collaborative prediction, and back-inference iterative correction mechanism to accurately characterize the coupling effects of multiple overpressure mechanisms, thereby greatly reducing prediction bias. This maintains stable and reliable prediction performance in carbonate rock formations with strong heterogeneity and diverse overpressure genesis. Another objective of this invention is to provide a device for predicting overpressure in carbonate rock formations based on cluster analysis. A further objective of this invention is to provide a computer-readable medium. A final objective of this invention is to provide a computer device.

[0007] To achieve the above objectives, this invention discloses a method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis, comprising: Monitoring well logging data in overpressured areas of carbonate rocks to obtain multi-source characteristic parameters of carbonate rock formations; Based on the multi-source feature parameters, causal feature extraction is performed to obtain the feature term extraction values; Based on the cause determination rules, dynamic weighted clustering analysis is performed according to the extracted feature values ​​to determine the corresponding target pressure prediction model and cause weight distribution. The target formation pressure prediction value is generated by using the target pressure prediction model to perform pressure prediction and back-inference verification based on multi-source characteristic parameters, feature term extraction values ​​and causal weights.

[0008] Preferably, before extracting causal features based on multi-source feature parameters to obtain feature term extraction values, the method further includes: The multi-source feature parameters are preprocessed to obtain the preprocessed multi-source feature parameters.

[0009] Preferably, the multi-source characteristic parameters include porosity, sonic transit time, rock composition, well logging fracture parameters, and electrical imaging. The extracted characteristic values ​​include undercompaction contribution, thorium-uranium ratio, pyrolysis hydrocarbon ratio, and fracture anisotropy. Based on multi-source feature parameters, causal feature extraction is performed to obtain feature term extraction values, including: The undercompaction contribution is generated based on porosity, sonic transit time, and logging fracture parameters. Based on the rock composition, the thorium-uranium ratio and pyrolytic hydrocarbon ratio values ​​are generated; Based on the electrical imaging, crack anisotropy is generated.

[0010] Preferably, based on the causal determination rules, dynamic weighted clustering analysis is performed according to the extracted feature values ​​to determine the corresponding target pressure prediction model and causal weight distribution, including: Based on the cause determination rules, dynamic weighted clustering analysis is performed according to the extracted feature values ​​to determine the set of overpressure cause types and the cause weight distribution. Based on the set of overpressure cause types, the corresponding target pressure prediction model is determined.

[0011] Preferably, based on the cause determination rules, dynamic weighted cluster analysis is performed according to the extracted feature values ​​to determine the set of overpressure cause types and the cause weight distribution, including: If the contribution of undercompaction is greater than the preset first threshold, the set of overcompression cause types is determined to include undercompaction causes. If the thorium-uranium ratio is greater than the preset second threshold and the pyrolysis hydrocarbon ratio is greater than the preset third threshold, the set of overpressure origin types is determined to include fluid expansion-related origins. If the crack anisotropy is greater than the preset fourth threshold, the set of overpressure origin types is determined to include tectonic pressurization origins. Weights are assigned to the overpressure cause types in the overpressure cause type set to generate a cause weight distribution.

[0012] Preferably, the target formation pressure prediction value is generated by using a target pressure prediction model, based on multi-source feature parameters, feature term extraction values, and genetic weights, to perform pressure prediction and back-analysis verification, including: The initial formation pressure prediction value is generated by using the target pressure prediction model based on multi-source characteristic parameters, feature term extraction values ​​and genetic weights. The inverse extrapolation of the initial formation pressure prediction values ​​yields the inverted genetic type. Determine whether the inversion causal type is the same as the set of overpressure causal types; If so, the initial formation pressure prediction value shall be determined as the target formation pressure prediction value; If not, update the causal weights and re-execute the step of determining the corresponding target pressure prediction model based on the set of overpressure causal types.

[0013] This invention also discloses a device for predicting overpressure formation pressure in carbonate rocks based on cluster analysis, comprising: The feature parameter acquisition unit is used to monitor well logging data in the overpressure zone of carbonate rocks and acquire multi-source feature parameters of carbonate rock layers. The feature extraction unit is used to extract causal features based on multi-source feature parameters to obtain feature term extraction values; The weighted clustering analysis unit is used to perform dynamic weighted clustering analysis based on the feature extraction values ​​according to the cause determination rules, and to determine the corresponding target pressure prediction model and cause weight distribution. The formation pressure prediction unit is used to predict and back-deductively verify the target formation pressure by using the target pressure prediction model based on multi-source characteristic parameters, feature term extraction values ​​and causal weights, and generate the target formation pressure prediction value.

[0014] The present invention also discloses a computer-readable medium having a computer program stored thereon, which, when executed by a processor, implements the method described above.

[0015] The present invention also discloses a computer device, including a memory and a processor, wherein the memory is used to store information including program instructions, and the processor is used to control the execution of the program instructions, wherein the processor executes the program to implement the method described above.

[0016] The present invention also discloses a computer program product, including a computer program / instruction, which, when executed by a processor, implements the method described above.

[0017] This invention monitors well logging data in overpressured areas of carbonate rocks to obtain multi-source characteristic parameters of the carbonate strata; it extracts genetic features based on these multi-source characteristic parameters to obtain feature extraction values; based on genetic determination rules, it performs dynamic weighted clustering analysis on the feature extraction values ​​to determine the corresponding target pressure prediction model and genetic weight distribution; through the target pressure prediction model, it performs pressure prediction and back-draft verification based on the multi-source characteristic parameters, feature extraction values, and genetic weights to generate target formation pressure prediction values. Dynamic weighted clustering enables rapid automatic identification of overpressure causes and adaptive model matching, significantly improving data processing efficiency. Furthermore, by employing genetic weight allocation, multi-model collaborative prediction, and back-draft iterative correction mechanisms, it accurately characterizes the coupling effect of multiple overpressure mechanisms to significantly reduce prediction bias, thereby maintaining stable and reliable prediction performance in carbonate strata with strong heterogeneity and diverse overpressure causes. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart illustrating a method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis, provided as an embodiment of the present invention; Figure 2 A flowchart illustrating another method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis, provided as an embodiment of the present invention; Figure 3A schematic diagram of a carbonate rock overpressure formation pressure prediction device based on cluster analysis provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] To facilitate understanding of the technical solution provided in this application, the relevant content of the technical solution will be explained below. This invention belongs to the sub-field of formation pressure prediction under oil and gas exploration and development technology. Its core innovation lies in integrating geophysical, geomechanical, and machine learning methods to solve the problem of complex overpressure prediction in carbonate rocks. It has strong interdisciplinary nature and directly serves the safety and efficient development of deep well drilling.

[0022] The following example uses a cluster analysis-based carbonate rock overpressure formation pressure prediction device as the execution subject to illustrate the implementation process of the cluster analysis-based carbonate rock overpressure formation pressure prediction method provided in this embodiment of the invention. It is understood that the execution subject of the cluster analysis-based carbonate rock overpressure formation pressure prediction method provided in this embodiment of the invention includes, but is not limited to, a cluster analysis-based carbonate rock overpressure formation pressure prediction device.

[0023] Figure 1 A flowchart illustrating a method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis, as provided in this embodiment of the invention, is shown below. Figure 1 As shown, the method includes: Step 101: Monitor the logging data of the overpressure area of ​​the carbonate rock to obtain multi-source characteristic parameters of the carbonate rock layer.

[0024] In this embodiment of the invention, well logging data of carbonate rock formations within the target area are collected. The well logging data includes, but is not limited to, porosity, density, sonic transit time, rock composition, temperature gradient, gas composition, well logging fracture parameters, and measured pressure.

[0025] Vertical stress is estimated by integrating the density obtained from density logging with the density of the overlying strata; geochemical indices (pyrolytic hydrocarbon ratio) are obtained from rock pyrolysis analysis in rock composition experiments; fracture porosity is calculated based on fracture parameters such as aperture and density from logging; acoustic time difference constant is calculated by subtracting theoretical normal acoustic time difference from measured acoustic time difference; structural curvature is obtained through seismic interpretation or geological modeling; hydrostatic pressure is calculated based on depth and fluid density; and the current stress field is obtained by comprehensively inverting data from wellbore collapse, drilling-induced fractures (from imaging logging), hydraulic fracturing data, and seismic anisotropy. The current stress field includes, but is not limited to, effective stress, and the direction and magnitude of maximum / minimum horizontal stress.

[0026] Multi-source characteristic parameters of carbonate rock formations include, but are not limited to, porosity sonic transit time, sonic transit time constant, vertical stress, geochemical indices, rock composition, temperature gradient, gas composition, well logging fracture parameters, fracture porosity, measured pressure, hydrostatic pressure, and current stress field.

[0027] This step aims to provide basic data support for subsequent causal analysis, ensuring data consistency and comparability.

[0028] Step 102: Extract causal features based on multi-source feature parameters to obtain feature extraction values.

[0029] In this embodiment of the invention, feature terms closely related to the cause of overpressure are extracted from multi-source feature parameters. The extracted feature terms include, but are not limited to, undercompaction contribution, thorium-uranium ratio, pyrolysis hydrocarbon ratio, and fracture anisotropy.

[0030] This invention, by quantifying the causal sensitive features of the raw data, can effectively characterize the developmental degree of different overpressure mechanisms, laying the foundation for subsequent causal identification.

[0031] Step 103: Based on the cause determination rules, perform dynamic weighted cluster analysis according to the extracted feature values ​​to determine the corresponding target pressure prediction model and cause weight distribution.

[0032] In this embodiment of the invention, based on the preset cause determination rules, the dynamic weighted clustering algorithm identifies and classifies the cause type of overpressure according to the extracted feature values, thereby obtaining the type of target pressure prediction model (such as acoustic-effective stress model, geochemical-pressure regression model, geostress field inversion constraint model) and its corresponding cause weight distribution.

[0033] This step enables adaptive identification of complex, multi-genetic overpressured strata, and can dynamically adjust the model combination according to different geological conditions, significantly improving the predictive model's adaptability to complex genesis.

[0034] Step 104: Using the target pressure prediction model, pressure prediction and reverse inference verification are performed based on multi-source characteristic parameters, feature term extraction values ​​and causal weights to generate the target formation pressure prediction value.

[0035] In this embodiment of the invention, a quantitative prediction of formation pressure is performed based on a target pressure prediction model and causal weights, combined with multi-source feature parameters and feature term extraction values. Simultaneously, the prediction results are verified through a reverse inference mechanism: the predicted values ​​are re-inputted into the target pressure prediction model for causal inference, and compared with the causal identification results. If inconsistencies are found, the weights are adjusted and iterative optimization is performed until the causal determination is consistent, at which point the final predicted value of the target formation pressure is output.

[0036] This step, by introducing a feedback iteration mechanism, effectively improves the reliability and accuracy of the prediction results, and is particularly suitable for complex carbonate rock overpressure formations with multiple coupled genesis.

[0037] In the technical solution provided by this invention, well logging data of carbonate rock overpressure areas are monitored to obtain multi-source characteristic parameters of the carbonate rock formation; genetic features are extracted based on the multi-source characteristic parameters to obtain feature extraction values; based on the genetic determination rules, dynamic weighted clustering analysis is performed on the feature extraction values ​​to determine the corresponding target pressure prediction model and genetic weight distribution; through the target pressure prediction model, pressure prediction and back-inference verification are performed based on the multi-source characteristic parameters, feature extraction values, and genetic weights to generate target formation pressure prediction values. Dynamic weighted clustering enables rapid automatic identification of overpressure causes and adaptive model matching, significantly improving data processing efficiency. At the same time, the genetic weight allocation, multi-model collaborative prediction, and back-inference iterative correction mechanism accurately characterize the multi-mechanism overpressure coupling effect to greatly reduce prediction bias, thereby maintaining stable and reliable prediction performance in carbonate rock formations with strong heterogeneity and diverse overpressure causes.

[0038] Figure 2 A flowchart illustrating another method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis, as provided in this embodiment of the invention, is shown below. Figure 2 As shown, the method includes: Step 201: Monitor the logging data of the carbonate rock overpressure area and obtain multi-source characteristic parameters of the carbonate rock layer.

[0039] In this embodiment of the invention, each step is performed by a carbonate rock overpressure formation pressure prediction device based on cluster analysis.

[0040] In this embodiment of the invention, for carbonate strata within the target exploration area, basic data for subsequent overpressure genesis analysis and pressure prediction are collected through well logging, seismic analysis, and core testing, forming a dataset of multi-source characteristic parameters. Specifically, in the target carbonate rock overpressure development zone, one or more wells (including vertical and deviated wells) are selected as data acquisition points. By performing well logging operations on these wells, multi-source characteristic parameters of the carbonate strata are obtained.

[0041] In this embodiment of the invention, the multi-source characteristic parameters include, but are not limited to, porosity acoustic transit time, acoustic transit time constant, vertical stress, geochemical index, rock composition, temperature gradient, gas composition, well logging fracture parameters, fracture porosity, measured pressure, hydrostatic pressure, and current stress field.

[0042] Porosity is calculated using density logging or sonic logging data, combined with rock skeleton parameters, and is used to assess the reservoir performance and compaction state of a formation.

[0043] Sonic transit time can be directly measured by a sonic logging tool, with units of microseconds per foot (μs / ft), reflecting the degree of rock compaction and porosity development.

[0044] The acoustic transit time constant is an intermediate parameter used to quantify the degree to which the actual compaction state deviates from the normal trend. Specifically, it is obtained by establishing a normal compaction trend line on the acoustic transit time curve, and then calculating the difference between the measured acoustic transit time at a certain depth point and the normal trend value at that depth. This difference reflects the degree of undercompaction of the rock caused by abnormally high pressure.

[0045] Vertical stress is calculated by depth integration of density logging curves, i.e., overlying strata pressure, and is the basic parameter for calculating effective stress.

[0046] Rock composition refers to the mineral composition (such as calcite, dolomite, and argillaceous content) in carbonate rocks, which can be obtained through elemental capture energy spectroscopy logging or core X-ray diffraction analysis and is used to correct rock physics models.

[0047] Geochemical indicators include pyrolytic hydrocarbon ratios (such as S1 / S2) and thorium-uranium ratios (Th / U). The pyrolytic hydrocarbon ratio is obtained through rock pyrolysis analysis of rock cuttings or cores; the thorium-uranium ratio is measured by a natural gamma ray spectroscopy logging tool and is used to indicate the sedimentary environment and organic matter abundance, thereby determining whether a fluid expansion mechanism exists.

[0048] Temperature gradients are obtained through bottom hole temperature measurements or geothermal monitoring data and are used to evaluate thermal maturity and fluid expansion potential.

[0049] Gas composition can be obtained through drilling fluid logging or cable formation testing to detect the content of gases such as methane and carbon dioxide, which can help determine the fluid properties and pressure source.

[0050] Well logging fracture parameters are obtained by image processing based on electrical imaging logging to acquire fracture orientation, density, width, and aperture.

[0051] Fracture porosity is calculated based on the interpretation of electrical imaging logging by statistically analyzing the contribution of fracture opening per unit volume. It is an important characterization parameter of structural pressurization.

[0052] The measured pressure is the formation fluid pressure value directly measured by a cable formation tester, and the unit is megapascals (MPa) or pounds per square inch (psi).

[0053] Hydrostatic pressure is a theoretical pressure value calculated based on the density and depth of formation water. It is used to compare with the measured pressure to calculate the total overpressure (measured pressure minus hydrostatic pressure).

[0054] The current stress field includes the effective stress, the magnitude and direction of the maximum / minimum horizontal principal stresses. It is usually obtained through hydraulic fracturing tests, wellbore collapse analysis (using six-arm dip logging or acoustic scanning logging), or dipole shear wave logging inversion.

[0055] This invention constructs a dataset of multi-source characteristic parameters covering the dimensions of "rock physics, geochemistry, and geomechanics." This dataset not only includes conventional pressure calculation elements but also introduces characteristic parameters that can distinguish different overpressure mechanisms (such as acoustic time difference constant, thorium-uranium ratio, and fracture anisotropy). This provides a solid data foundation for accurately identifying the three types of overpressure formation—undercompaction, fluid expansion, and tectonic pressurization—in subsequent steps, solving the technical problem that traditional methods cannot handle multi-causal coupling due to the single input parameter.

[0056] Step 202: Preprocess the multi-source feature parameters to obtain the preprocessed multi-source feature parameters.

[0057] To eliminate systematic errors caused by factors such as instruments, environment, and depth during data acquisition, unify the dimensions and scales of different parameters, and remove abnormal interference signals, preprocessing is performed on multi-source characteristic parameters. Preprocessing includes well logging curve depth alignment, seismic attribute extraction, and geochemical index standardization.

[0058] Well logging curve depth alignment: This involves matching and correcting the depth of well logging curves (such as sonic transit time curves and density curves) from different logging instruments and measurements. Specifically, marker layers with obvious geological characteristics are identified on each curve. Using a standard curve as a reference, other curves are stretched or compressed as a whole to eliminate depth errors caused by cable extension / retraction, instrument jamming, etc. For deviated well data, vertical depth correction is also required, converting the measured depth into vertical depth. This preprocessing operation of well logging curve depth alignment ensures that various parameters collected at the same depth point truly correspond to the same geological stratum, avoiding logical errors in subsequent feature extraction and pressure calculation due to depth misalignment.

[0059] Seismic attribute extraction: This involves extracting attribute parameters related to carbonate reservoir characteristics and pressure response from 3D seismic data volumes. Specifically, along the target formation bedding planes, various seismic attributes are extracted, including but not limited to: layer velocity, amplitude attributes, and tectonic curvature. Layer velocity, obtained through seismic velocity spectrum analysis, reflects the compaction trend of the formation; amplitude attributes include root-mean-square amplitude, used to identify fracture development zones or changes in fluid properties; and tectonic curvature indicates tectonic stress concentration zones. For well-side seismic traces, the extracted seismic attributes are calibrated and correlated with well logging data, prioritizing attributes sensitive to overpressure. The preprocessing operations for seismic attribute extraction extend one-dimensional well logging information to three-dimensional space, compensating for insufficient inter-well information and providing constraints for predicting regional overpressure distribution.

[0060] Geochemical index standardization: This involves normalizing geochemical analysis data, removing outliers, and performing consistency correction to eliminate systematic biases caused by differences in measuring instruments, experimental conditions, or sample types. Specifically, normalization involves mapping ratio parameters such as pyrolytic hydrocarbon ratios (e.g., S1 / S2) and thorium-uranium ratio (Th / U) to a uniform numerical range (e.g., between 0 and 1) using extreme value normalization or mean-variance standardization methods. Outlier removal uses statistical methods (e.g., the 3σ principle) to identify and remove abnormal data points caused by measurement errors or sample contamination, ensuring that the statistical distribution of the input data conforms to geological laws. Consistency correction involves performing consistency correction on identical geochemical indices at different depths within the same well or between different wells through trend surface analysis or cross-plot analysis to eliminate systematic errors.

[0061] This invention eliminates depth misalignment in well logging data through depth alignment, ensuring accurate and reliable correspondence between parameters. Standardization of geochemical indicators enables joint analysis of parameters with different dimensions and magnitudes within a unified mathematical model, providing a prerequisite for subsequent dynamic weighted clustering analysis. Seismic attribute extraction integrates 3D seismic information with well logging information, enhancing the formation pressure prediction model's ability to identify anomalies in inter-well regions. Outlier removal and quality control effectively filter invalid data, improving the stability and reliability of subsequent causal feature extraction and pressure prediction.

[0062] Step 203: Generate the undercompaction contribution based on porosity, sonic transit time, and logging fracture parameters.

[0063] In this embodiment of the invention, the acoustic time difference constant, porosity-acoustic correlation factor, and fracture porosity are extracted based on porosity, acoustic transit time, and logging fracture parameters.

[0064] The acoustic transit time constant is defined as the difference between the measured acoustic transit time and the theoretical acoustic transit time under normal compaction trend at that depth. The normal compaction trend line is established by regression analysis on the acoustic transit time curve for normally compacted mudstone sections or pure carbonate rock baselines, reflecting the variation of acoustic transit time with depth under normal compaction conditions. The magnitude of the acoustic transit time constant directly reflects the degree to which the formation deviates from the normal compaction state; the larger the difference, the stronger the undercompaction.

[0065] The porosity-acoustic correlation factor is defined as the ratio of measured porosity to theoretical porosity under normal compaction trends. The theoretical porosity can be obtained using an empirical formula by converting the acoustic transit time corresponding to the normal compaction trend line. This correlation factor is used to correct the consistency between porosity and acoustic response, eliminating the interference of lithological variations on undercompaction identification.

[0066] Based on this, a calculation function for the overpressure component caused by undercompaction is constructed. The independent variables of this function include the acoustic time difference constant, fracture porosity, and the porosity-acoustic correlation factor. The function form is based on a rock physics model and reflects the quantitative relationship between porosity retention and abnormal pressure accumulation under undercompaction. This function is used to calculate the overpressure value contributed solely by the undercompaction mechanism.

[0067] The difference between the measured pressure and the hydrostatic pressure is defined as the total overpressure value. The measured pressure is obtained directly using a cable formation tester, while the hydrostatic pressure is calculated based on the formation water density and depth according to the hydrostatic pressure gradient.

[0068] Specifically, through The undercompaction contribution is calculated based on porosity, sonic transit time, and well logging fracture parameters. Here, W represents the undercompaction contribution. The overpressure component caused by under-compaction; The total overpressure is the measured pressure minus the hydrostatic pressure. It is a constant for acoustic time difference, that is, the difference between the measured acoustic time difference at the current depth and the normal trend value at that depth; This refers to the fracture porosity calculated based on fracture parameters such as aperture and density from well logging. The porosity-acoustic correlation factor is the ratio of measured porosity to theoretical porosity under normal compaction conditions. The sum of contributions from all overpressure mechanisms, namely: undercompaction pressure, fluid expansion pressure, and structural pressurization pressure; This is a calculation function relating the overpressure component caused by undercompaction to the acoustic time difference constant, crack porosity, and porosity-acoustic correlation factor.

[0069] In this embodiment of the invention, the contribution of undercompaction is between 0 and 1. The closer the value is to 1, the more the overpressure phenomenon at the current depth point is mainly dominated by the undercompaction mechanism; the closer the value is to 0, the smaller the contribution of undercompaction, and the overpressure may originate from other causes.

[0070] This invention achieves a quantitative characterization of the intensity of undercompaction in carbonate formations. Traditional overpressure prediction methods often directly equate acoustic temporal variation with undercompaction, making it difficult to distinguish acoustic response changes caused by other factors. This invention introduces fracture porosity for correction and combines a porosity-acoustic correlation factor to eliminate lithological influences, effectively removing interference from factors such as fracture development on the acoustic response. This ensures that the calculated contribution of undercompaction truly reflects the actual weight of the undercompaction mechanism in the total overpressure. This quantitative index provides accurate input for subsequent steps of dynamically allocating overpressure causal weights and selectively choosing prediction models, solving the technical challenge of quantitatively separating overpressure caused by multiple factors.

[0071] Step 204: Based on the rock composition, generate the thorium-uranium ratio and the pyrolytic hydrocarbon ratio.

[0072] In this embodiment of the invention, natural gamma ray spectroscopy logging data and rock pyrolysis analysis data are extracted based on rock composition.

[0073] In this embodiment of the invention, the thorium-uranium ratio is calculated based on the measurement principle of natural gamma ray spectroscopy logging. Specifically, the natural gamma ray spectroscopy logging instrument records gamma rays released by the decay of radioactive elements in the formation. Through energy spectrum analysis technology, the total gamma ray count is divided according to energy windows, and the contents of thorium, uranium, and potassium in the formation are quantitatively calculated separately, i.e., the natural gamma ray spectroscopy logging data. The thorium content curve and uranium content curve are directly read from the natural gamma ray spectroscopy logging data. Then, the ratio of the thorium content value and the uranium content value at the same depth point is calculated to obtain the thorium-uranium ratio at that depth point.

[0074] In practical processing, data from relatively pure carbonate rock strata are selected for ratio calculation to avoid interference from excessively high clay content on the indicative significance of the thorium-uranium ratio. The thorium-uranium ratio is an important indicator for judging the sedimentary environment and the degree of organic matter development. A high thorium-uranium ratio usually indicates an oxidizing environment or input of terrigenous materials, while a low thorium-uranium ratio may indicate a reducing environment or enrichment of organic matter, the latter being closely related to the fluid expansion overpressure mechanism.

[0075] In this embodiment of the invention, the pyrolysis hydrocarbon ratio is calculated based on rock pyrolysis analysis data. Specifically, a rock pyrolysis analyzer is used to program-heat rock cuttings or core samples, recording the content of hydrocarbons released in different temperature ranges. The S1 peak represents the content of free hydrocarbons already present in the sample, released at approximately 300°C; the S2 peak represents the hydrocarbon content generated by kerogen pyrolysis, released within the temperature range of 300°C to 600°C. Based on the rock pyrolysis analysis experimental data, the S1 and S2 values ​​for each sample are directly read. Then, the S1 value is divided by the S2 value to obtain the pyrolysis hydrocarbon ratio of the sample. For continuous depth profiles, interpolation methods are used to convert discrete sample analysis results into continuous logging curves. The pyrolysis hydrocarbon ratio reflects the generation and discharge state of hydrocarbons. A high pyrolysis hydrocarbon ratio usually indicates strong fluid expansion, suggesting that there may be overpressure in the formation due to hydrocarbon generation pressurization.

[0076] This invention enables the quantitative characterization of fluid expansion overpressure potential in carbonate formations. The thorium-uranium ratio and the pyrolytic hydrocarbon ratio are core geochemical indicators for identifying fluid expansion mechanisms. The thorium-uranium ratio reflects the sedimentary environment background of organic matter development, while the pyrolytic hydrocarbon ratio directly quantifies the intensity of hydrocarbon generation. Traditional overpressure prediction methods typically ignore geochemical information, making it difficult to identify overpressures not caused by undercompaction. This invention, however, systematically processes energy dispersive spectroscopy logging and rock pyrolysis data, transforming geochemical characteristics into quantifiable input parameters. This provides a crucial basis for subsequent steps in identifying fluid expansion-type overpressures and dynamically allocating causal weights, effectively solving the technical problem of the difficulty in quantitatively characterizing geochemical mechanisms in multi-genetic overpressures.

[0077] Step 205: Generate crack anisotropy based on the electrical imaging diagram.

[0078] In this embodiment of the invention, based on high-resolution formation images obtained from electrical imaging logging, fracture anisotropy parameters, used to characterize the directionality and non-uniformity of fracture development, are calculated through image processing and quantitative interpretation of fractures.

[0079] In this embodiment of the invention, step 205 specifically includes: Step 2051: Obtain an electrophysiological image.

[0080] Specifically, electrical imaging logging instruments transmit electric current into the formation and record changes in current intensity through an array of button electrodes on multiple plates attached to the wellbore, generating a high-resolution resistivity image covering 360 degrees of the wellbore. This image can visually display geological features on the wellbore, where fractures typically appear as dark or light sinusoidal curves with a significant difference in resistivity from the background.

[0081] Step 2052: Perform image enhancement processing on the electro-imaging image.

[0082] Specifically, image processing methods such as filtering and contrast stretching are used to highlight the boundary features of cracks, suppress noise and artifacts (such as scratches and mud peeling), and improve the signal-to-noise ratio for crack identification.

[0083] Step 2053: Perform automatic crack picking or interactive interpretation to generate statistical results.

[0084] Specifically, in the enhanced image, image edge detection algorithms or manual interaction are used to identify and track the trajectory of each crack. For each identified crack, three key geometric parameters are calculated based on its sinusoidal curve shape in the image: the crack's depth, its dip angle, and its dip direction. The dip angle is calculated from the amplitude of the sinusoidal curve, and the dip direction is determined from the orientation of the lowest point of the sinusoidal curve.

[0085] Based on the interpretation of individual fractures, a statistical analysis is performed on all fractures throughout the well section. Fractures are grouped according to their dip azimuth, typically in intervals of 10 or 15 degrees. The fracture development density (i.e., the number of fractures per unit length of well section) or fracture porosity contribution value is calculated within each azimuth interval; this is the statistical result. The fracture porosity contribution value is calculated based on the fracture opening width and extension length, reflecting the degree of fracture opening in space.

[0086] Step 2054: Based on the statistical results, generate crack anisotropy parameters.

[0087] In this embodiment of the invention, the crack anisotropy parameter is defined as follows: the main direction of crack development is determined, that is, the azimuth interval with the largest contribution value of crack density or crack porosity; the ratio between the crack development intensity in the main direction and the crack development intensity perpendicular to the main direction is calculated.

[0088] In the specific calculation, the cracks are projected onto the coordinate system according to their dip direction. The vector mean of the crack direction is calculated by the vector synthesis method, and the variance or eigenvalue of the crack development orientation distribution is solved.

[0089] The crack anisotropy parameter is expressed as the ratio of the maximum eigenvalue to the minimum eigenvalue. The larger the ratio, the stronger the directionality of crack development and the more significant the anisotropy.

[0090] This invention transforms the spatial distribution characteristics of cracks into quantifiable anisotropic parameters through detailed interpretation of electro-imaging images and directional statistical analysis. This provides a key basis for identifying tectonic pressurization-type overpressure and dynamically allocating causal weights in subsequent steps, effectively solving the technical problem of the difficulty in quantitatively evaluating tectonic effects in the cause of overpressure.

[0091] Step 206: Based on the cause determination rules, perform dynamic weighted cluster analysis according to the extracted feature values ​​to determine the set of overpressure cause types and the cause weight distribution.

[0092] In this embodiment of the invention, step 206 specifically includes: Step 2061: If the contribution of undercompaction is greater than the preset first threshold, determine that the set of overcompression cause types includes undercompaction causes.

[0093] In this embodiment of the invention, the undercompaction contribution at the current depth point is compared with a first threshold. If the undercompaction contribution is greater than the first threshold, it is determined that the undercompaction effect at the current depth point constitutes a significant overpressure contribution mechanism, and the undercompaction cause is included in the overpressure cause type set. If the undercompaction contribution is less than or equal to the first threshold, it is determined that the undercompaction effect is not significant, and it is not included in the overpressure cause type set.

[0094] It is worth noting that the first threshold corresponds to the critical value of the contribution of undercompaction, and is used to identify whether the effect of undercompaction is significant. The specific value of the first threshold can be obtained by statistical calibration of sample data of known overcompression types in the study area, or it can be set empirically according to regional geological laws. This embodiment of the invention does not limit this. As an optional option, the first threshold is 60%.

[0095] Step 2062: If the thorium-uranium ratio is greater than the preset second threshold and the pyrolysis hydrocarbon ratio is greater than the preset third threshold, determine that the set of overpressure cause types includes fluid expansion causes; In this embodiment of the invention, the thorium-uranium ratio and the pyrolytic hydrocarbon ratio at the current depth point are compared with a preset second threshold and a third threshold, respectively. If the thorium-uranium ratio is greater than the second threshold and the pyrolytic hydrocarbon ratio is greater than the third threshold, then the fluid expansion effect is determined to constitute a significant overpressure contribution mechanism at the current depth point, and the fluid expansion-type cause is included in the overpressure cause type set. If the two conditions cannot be met simultaneously, i.e., the thorium-uranium ratio is less than or equal to the second threshold, or the pyrolytic hydrocarbon ratio is less than or equal to the third threshold, then the fluid expansion effect is determined to be insignificant, and it is not included in the overpressure cause type set.

[0096] It is worth noting that the second threshold corresponds to the critical value of the thorium-uranium ratio, and the third threshold corresponds to the critical value of the pyrolytic hydrocarbon ratio. The two are used together to identify whether the fluid expansion effect is significant. The specific values ​​of the second and third thresholds can be obtained through statistical calibration of sample data of known overpressure origin types within the study area, or they can be set empirically based on regional geological patterns. This embodiment of the invention does not limit these values. As an optional approach, the second threshold is 2.5, and the third threshold is 0.7.

[0097] Step 2063: If the crack anisotropy is greater than the preset fourth threshold, determine that the set of overpressure origin types includes tectonic pressurization origin.

[0098] In this embodiment of the invention, the crack anisotropy value at the current depth point is compared with a preset fourth threshold. If the crack anisotropy is greater than the fourth threshold, it is determined that tectonic pressurization constitutes a significant overpressure contribution mechanism at the current depth point, and the tectonic pressurization-type cause is included in the overpressure cause type set. If the crack anisotropy is less than or equal to the fourth threshold, it is determined that the tectonic pressurization effect is not significant, and it is not included in the overpressure cause type set.

[0099] It is worth noting that the fourth threshold corresponds to the critical value of fracture anisotropy and is used to identify whether the tectonic pressurization effect is significant. The specific value of the fourth threshold can be obtained through statistical calibration of sample data of known overpressure types in the study area, or it can be set empirically based on regional geological patterns. This embodiment of the invention does not limit this. As an optional option, the fourth threshold is 0.3.

[0100] After steps 2061 to 2063, a set of overpressure cause types is obtained. This set of overpressure cause types may exist in four ways: an empty set (containing no overpressure cause types), a set containing only a single overpressure cause type, a set containing two overpressure cause types, or a set containing three overpressure cause types. For any case other than an empty set, subsequent weight allocation processing is performed.

[0101] It is worth noting that if the set of overpressure cause types is empty, it means that no overpressure cause type was detected, and the process ends.

[0102] Step 2064: Assign weights to the overpressure cause types in the overpressure cause type set to generate cause weight distribution.

[0103] In this embodiment of the invention, the principle of weight allocation is that the sum of the weights of each overpressure cause type equals 1. For cases containing only a single overpressure cause type, the weight of that overpressure cause type is directly assigned a value of 1. For cases containing two or three overpressure cause types, weight allocation is performed based on the relative magnitude of the extracted values ​​of each feature item.

[0104] In this embodiment of the invention, for undercompaction-related genesis, the weight is positively correlated with the contribution of undercompaction; for fluid expansion-related genesis, the weight is positively correlated with the combined values ​​of the thorium-uranium ratio and the pyrolytic hydrocarbon ratio; and for tectonic pressurization-related genesis, the weight is positively correlated with the anisotropy of the fractures. Specifically, based on a dynamic weighted clustering algorithm, the weights of the overpressure-related genesis types are iteratively adjusted according to the extracted values ​​of each feature item corresponding to the overpressure-related genesis type to generate a generative weight distribution.

[0105] As an alternative, dynamic weighted clustering algorithms include, but are not limited to, dynamic hierarchical star clustering algorithm, K-means, and ISODATA algorithm.

[0106] For example, if the set of overpressure origin types includes undercompaction and fluid expansion origins, then based on a dynamic weighted clustering algorithm, the weights of undercompaction and fluid expansion origins are allocated proportionally according to the contribution of undercompaction, the thorium-uranium ratio, and the pyrolytic hydrocarbon ratio. Similarly, if the set of overpressure origin types includes undercompaction, fluid expansion, and tectonic pressurization origins, then the weights are allocated according to the relative proportions of undercompaction contribution, thorium-uranium ratio, pyrolytic hydrocarbon ratio, and fracture anisotropy.

[0107] This invention enables automatic identification and quantitative weight allocation of the genetic types of complex overpressure formations in carbonate rocks. Traditional overpressure prediction methods typically assume that overpressure is dominated by a single mechanism or use fixed empirical weights for the superposition of multiple mechanisms, making it difficult to adapt to the complex situation of diverse genetic types and varying contribution ratios in actual formations. This invention, based on multi-source feature extraction values, dynamically identifies the actual genetic types of overpressure at the current depth point through threshold judgment rules, covering all possible cases of single-gene, dual-gene, and trigeneous coupling. Simultaneously, weights are allocated according to the relative magnitudes of feature values, ensuring that the contribution ratio of each genetic type matches the measured geological characteristics. This dynamic weighted clustering analysis method provides accurate input for subsequent steps of selecting appropriate prediction models for different genetic types and performing weighted fusion calculations, effectively solving the technical challenge of quantitative decomposition and weight allocation of multi-genetic coupled overpressure.

[0108] Step 207: Determine the corresponding target pressure prediction model based on the set of overpressure cause types.

[0109] In this embodiment of the invention, the target pressure prediction model includes at least one of the acoustic-effective stress pressure prediction model, the geochemical-pressure regression pressure prediction model, and the geostress field inversion constrained pressure prediction model.

[0110] The acoustic-effective stress model for pressure prediction is based on the physical mechanism of undercompaction leading to rock porosity retention and anomalous acoustic transit time. It calculates pore pressure through the relationship between acoustic transit time and effective stress. The acoustic-effective stress pressure prediction model is as follows:

[0111] in, For under-compacted void pressure; It is vertical stress; Effective stress; For sound wave time difference; This represents the normal trend of acoustic time difference; The compressibility coefficient of carbonate rocks; This is the compaction index.

[0112] The geochemical-pressure regression model for pressure prediction is based on geochemical mechanisms that lead to increased pressure due to hydrocarbon generation or fluid thermal expansion. It calculates pore pressure using statistical regression relationships between geochemical indicators such as the pyrolysis hydrocarbon ratio and thorium-uranium ratio and pressure. The geochemical-pressure regression pressure prediction model is as follows:

[0113] in, It is the pressure of fluid expansion; This represents the ratio of pyrolytic hydrocarbons. Thorium-uranium ratio; and These are the pyrolysis hydrocarbon ratio coefficient and the pyrolysis hydrocarbon ratio coefficient, respectively. and All are greater than 0; This is the base pressure offset.

[0114] The geomechanical model for predicting constrained pressure based on the inversion of the geostress field is based on the geomechanical mechanism of stress concentration and increased pressure in formations caused by tectonic compression. It calculates pore pressure by relating geostress field parameters (such as minimum horizontal stress) to the pressure in the overlying strata. The geostress field inversion model for predicting constrained pressure is as follows:

[0115] in, To construct pressure; It is vertical stress; Effective stress; This represents the minimum horizontal stress.

[0116] In this embodiment of the invention, a target pressure prediction model for subsequent pressure calculation is determined according to a preset model-genetic correspondence. Specifically, the model-genetic correspondence includes: an acoustic-effective stress pressure prediction model corresponding to undercompaction-related genesis; a geochemical-pressure regression pressure prediction model corresponding to fluid expansion-related genesis; and a geostress field inversion constrained pressure prediction model corresponding to structural pressurization-related genesis.

[0117] Specifically, for cases where the set of overpressure origin types contains only a single overpressure origin type, if the set of overpressure origin types contains only undercompactment origin, then the target pressure prediction model is determined to be the acoustic wave-effective stress pressure prediction model; if the set of overpressure origin types contains only fluid expansion origin, then the target pressure prediction model is determined to be the geochemical-pressure regression pressure prediction model; if the set of overpressure origin types contains only tectonic pressurization origin, then the target pressure prediction model is determined to be the geostress field inversion constrained pressure prediction model.

[0118] Specifically, for cases where the set of overpressure genesis types simultaneously contains two types of overpressure genesis, if the set of overpressure genesis types simultaneously contains both undercompaction and fluid expansion genesis, then the target pressure prediction model is determined to be a combination of the acoustic wave-effective stress pressure prediction model and the geochemical-pressure regression pressure prediction model; if the set of overpressure genesis types simultaneously contains both undercompaction and tectonic pressurization genesis, then the target pressure prediction model is determined to be a combination of the acoustic wave-effective stress pressure prediction model and the geostress field inversion constrained pressure prediction model; if the set of overpressure genesis types simultaneously contains both fluid expansion and tectonic pressurization genesis, then the target pressure prediction model is determined to be a combination of the geochemical-pressure regression pressure prediction model and the geostress field inversion constrained pressure prediction model.

[0119] Specifically, for cases where the set of overpressure origin types contains three overpressure origin types simultaneously, if the set of overpressure origin types simultaneously contains undercompaction, fluid expansion, and tectonic pressurization, then the target pressure prediction model is determined to be a combination of the acoustic wave-effective stress pressure prediction model, the geochemical-pressure regression pressure prediction model, and the geostress field inversion constrained pressure prediction model.

[0120] By applying the selection rules described above, a set of target pressure prediction models is obtained that perfectly matches the set of overpressure causal types at the current depth point. Each pressure prediction model in the target pressure prediction model set will be used to calculate the pressure component contributed by the corresponding causal type, providing model support for pressure fusion calculation based on causal weights in subsequent steps.

[0121] This invention dynamically selects the corresponding model combination based on the set of genetic types, ensuring that the model used in the calculation at each depth point is consistent with its actual overpressure genetic mechanism, thus avoiding systematic errors caused by model misuse. This model-gene adaptive matching mechanism lays a correct foundation for subsequent pressure fusion calculations based on genetic weights, significantly improving the prediction reliability of complex multi-genetic overpressure formations.

[0122] Step 208: Using the target pressure prediction model, generate the initial formation pressure prediction value based on multi-source characteristic parameters, feature term extraction values, and causal weights.

[0123] In this embodiment of the invention, multi-source feature parameters and feature term extraction values ​​are input into the determined target pressure prediction model for calculation to obtain the corresponding pressure prediction components (undercompacted void pressure, fluid expansion pressure, tectonic pressure); according to the genetic weight, the pressure prediction components are weighted and calculated to generate the initial formation pressure prediction value.

[0124] This invention employs corresponding target pressure prediction models for different genetic types. The input parameters of each target pressure prediction model are multi-source feature parameters and extracted feature terms, possessing clear physical meaning or statistical basis. Based on this, weighted fusion of each pressure component is performed using genetic weights, ensuring that the final prediction result accurately reflects the actual contribution ratio of different overpressure mechanisms to the total overpressure. This genetically driven, model-matching, and weighted fusion calculation method significantly improves the prediction accuracy of complex, multi-genetic overpressure formations.

[0125] Step 209: Perform reverse deduction on the initial formation pressure prediction value to obtain the inverted genetic type.

[0126] In this embodiment of the invention, the core idea of ​​reverse inference is to use the reverse process of the pressure prediction model to infer the theoretical values ​​of the causal characteristic indicators from the pressure values, and then determine the causal type based on these inverse theoretical values.

[0127] The initial formation pressure prediction value is used as input data and re-inputted into the pressure prediction model system. The overpressure genesis type corresponding to the pressure value is deduced through reverse calculation, thus forming the inverted genesis type.

[0128] Specifically, for the acoustic wave-effective stress pressure prediction model, the initial formation pressure prediction value is substituted into the inverse function of the acoustic wave-effective stress pressure prediction model to calculate the theoretical anomaly degree of acoustic wave transit time, i.e., the contribution of undercompaction. This process essentially verifies how much undercompaction is needed to achieve the current initial formation pressure prediction value.

[0129] For the geochemical-pressure regression pressure prediction model, the initial formation pressure prediction values ​​are substituted into the inverse function of the model to calculate the theoretical intensity of hydrocarbon generation indicators, namely: the inverted pyrolysis hydrocarbon ratio and the inverted thorium-uranium ratio. The inverted pyrolysis hydrocarbon ratio and the inverted thorium-uranium ratio are then compared with the measured pyrolysis hydrocarbon ratio and the measured thorium-uranium ratio, respectively, to assess the dependence of the current initial formation pressure prediction values ​​on hydrocarbon generation intensity.

[0130] For the geostress field inversion constrained pressure prediction model, the initial formation pressure prediction value is substituted into the inverse function of the geostress field inversion constrained pressure prediction model to calculate the theoretical geostress state, that is, to invert the fracture anisotropy. The inverted fracture anisotropy is then compared with the measured fracture anisotropy to invert the degree of contribution of tectonic activity to the current pressure value.

[0131] Based on the genesis determination rules in step 206, the inversion genesis type is obtained according to the inversion undercompaction contribution, inversion pyrolysis hydrocarbon ratio, inversion thorium-uranium ratio, and inversion fracture anisotropy. For example, if the inversion undercompaction contribution is greater than a preset first threshold, the inversion genesis type is determined to include undercompaction genesis.

[0132] This invention can quantitatively assess the matching degree between the initial predicted pressure value and the actual geological characteristics. If the inverted causal type is inconsistent with the initially determined causal type, it indicates that there is a deviation in the initial causal weight allocation. This mechanism is equivalent to performing a self-consistency check on the pressure prediction model, effectively avoiding prediction inaccuracies caused by improper initial weight allocation, and providing a clear physical basis for subsequent weight iteration correction, thereby ensuring that the final prediction results still have high reliability in complex geological environments with multiple coupled causal factors.

[0133] Step 210: Determine whether the inversion cause type is the same as the overpressure cause type set. If yes, proceed to step 211; otherwise, proceed to step 212.

[0134] In this embodiment of the invention, the overpressure origin type set is a collection containing one or more overpressure origin types (undercompaction type, fluid expansion type, tectonic pressurization type). The inversion origin type is also a collection of inversion origin types identified based on the inverse inference results.

[0135] Specifically, compare whether the causal types contained in the inverted causal type set are completely consistent with the causal types contained in the overpressure causal type set. If they are completely consistent, proceed to step 211; if they are not completely consistent, for example, the number of causal types in the set is different, or the number is the same but the types are different, continue to proceed to step 212.

[0136] Step 211: Determine the initial formation pressure prediction value as the target formation pressure prediction value, and the process ends.

[0137] In this embodiment of the invention, if the inversion genetic type is the same as the set of overpressure genetic types, it indicates that the pressure value obtained by the current forward prediction is completely consistent with the geological genetic characteristics verified by the reverse deduction. That is, the predicted pressure value is consistent with the overpressure mechanism reflected by the measured geological data. The initial formation pressure prediction value is output as the final target formation pressure prediction value, and the process ends.

[0138] Step 212: Update the causal weights and re-execute step 207.

[0139] In this embodiment of the invention, if the inverted causal type is different from the set of overpressure causal types, it indicates that the initially assigned causal weights failed to accurately reflect the contribution ratio of the actual overpressure mechanism, leading to a contradiction between positive prediction and reverse verification. This step eliminates this contradiction by iteratively correcting the causal weights.

[0140] The specific principle for updating the causal weights is to reduce the overpressure causal weight corresponding to the term with the largest error in the back-reasoning. The term with the largest error refers to the causal type with the largest deviation between the theoretical eigenvalues ​​calculated and the measured eigenvalues ​​during the back-reasoning process.

[0141] Specifically, based on the inverted undercompaction contribution, inverted pyrolytic hydrocarbon ratio, inverted thorium-uranium ratio, and inverted fracture anisotropy compared with the measured undercompaction contribution, thorium-uranium ratio, pyrolytic hydrocarbon ratio, and fracture anisotropy, the relative error of each characteristic term is calculated: Undercompaction contribution error: ,in, W represents the contribution of undercompaction as determined by inversion; W represents the measured contribution of undercompaction. This contributes to the error in undercompaction.

[0142] Error in pyrolysis hydrocarbon ratio: ,in, To retrieve the ratio of pyrolytic hydrocarbons; The measured ratio of pyrolytic hydrocarbons; This represents the error in the ratio of pyrolytic hydrocarbons.

[0143] Thorium-uranium ratio error: ,in, To retrieve the ratio of pyrolytic hydrocarbons; The measured ratio of pyrolytic hydrocarbons; This represents the error in the ratio of pyrolytic hydrocarbons.

[0144] Crack anisotropy error: ,in, L represents the contribution of undercompaction in the inversion calculation; L represents the measured contribution of undercompaction. This contributes to the error in undercompaction.

[0145] Errors related to undercompaction contribution, pyrolysis hydrocarbon ratio, thorium-uranium ratio, and fracture anisotropy are compared. The maximum value is selected as the largest error term, thus determining the corresponding cause type of the largest error term. The weight of the selected cause type of the largest error term is then attenuated. Let w be the weight of the cause type of the largest error term. i The updated weight w i ′=w i ×λ, where λ is the attenuation factor, typically ranging from 0.5 to 0.9 (e.g., 0.8). The specific value of the attenuation factor can be set based on actual work area calibration or experience, with the aim of gradually reducing the contribution ratio of this cause.

[0146] Furthermore, after attenuating the weight of the term with the largest error, the weights of all causes in the set of overpressure cause types are normalized to ensure that the sum of all weights is 1.

[0147] Based on the updated causal weights, step 207 is executed again. Based on the updated causal weight distribution, the target pressure prediction model combination corresponding to the current set of overpressure causal types is re-determined. Subsequently, forward prediction, backward deduction, and consistency judgment are performed again in the order of steps 208 to 210 until the inverted causal type matches the set of overpressure causal types, or the preset maximum number of iterations (e.g., 10 times) is reached. If inconsistency still exists after reaching the maximum number of iterations, the initial formation pressure prediction value of the last iteration is output as the target value, and a prompt message is issued for manual review.

[0148] This invention, through the aforementioned iterative correction mechanism, can automatically identify and adjust the weights of the causes of error when contradictions arise between forward prediction and reverse verification, gradually approximating the actual geological conditions. This dynamic adjustment process not only eliminates the subjectivity of the initial weight allocation but also ensures that the final prediction results are highly consistent with the measured geophysical, geochemical, and geomechanical characteristics, significantly improving the prediction accuracy and reliability in complex carbonate strata with multi-mechanism overpressure coupling. Furthermore, this iterative process is entirely executed automatically by the algorithm without human intervention, effectively guaranteeing prediction efficiency.

[0149] It is worth noting that the acquisition, storage, use, and processing of data in the technical solution of this application all comply with relevant laws and regulations. The user information in the embodiments of this application was obtained through legal and compliant means, and the acquisition, storage, use, and processing of user information have been authorized and agreed upon by the client.

[0150] It is worth noting that the information collected in this application is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data all comply with the relevant laws, regulations and standards of the relevant countries and regions, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding operation portals are provided for users to choose to authorize or refuse.

[0151] It is worth noting that the technical solution provided in this application provides users with a corresponding operation entry point, allowing users to choose to agree to or reject the automated decision-making result; if the user chooses to reject, the process will proceed to the expert decision-making process.

[0152] The technical solution of the carbonate rock overpressure formation pressure prediction method based on cluster analysis provided in this invention involves monitoring well logging data of the carbonate rock overpressure area to obtain multi-source characteristic parameters of the carbonate rock layer; extracting genetic features based on the multi-source characteristic parameters to obtain feature extraction values; performing dynamic weighted cluster analysis based on the feature extraction values ​​according to the genetic determination rules to determine the corresponding target pressure prediction model and genetic weight distribution; using the target pressure prediction model, performing pressure prediction and back-inference verification based on the multi-source characteristic parameters, feature extraction values, and genetic weights to generate target formation pressure prediction values; achieving rapid automatic identification and adaptive model matching of overpressure causes through dynamic weighted clustering, significantly improving data processing efficiency; and employing genetic weight allocation, multi-model collaborative prediction, and back-inference iterative correction mechanisms to accurately characterize the multi-mechanism overpressure coupling effect to significantly reduce prediction bias, thereby maintaining stable and reliable prediction performance in carbonate rock formations with strong heterogeneity and diverse overpressure causes.

[0153] Figure 3 This is a schematic diagram of a device for predicting overpressure formation pressure in carbonate rocks based on cluster analysis, provided in an embodiment of the present invention. This device is used to execute the aforementioned method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis. Figure 3 As shown, the device includes: a feature parameter acquisition unit 11, a feature extraction unit 12, a weighted clustering analysis unit 13, and a formation pressure prediction unit 14.

[0154] The feature parameter acquisition unit 11 is used to monitor well logging data in the overpressure area of ​​carbonate rocks and acquire multi-source feature parameters of carbonate rock layers.

[0155] The feature extraction unit 12 is used to extract causal features based on multi-source feature parameters to obtain feature term extraction values.

[0156] The weighted clustering analysis unit 13 is used to perform dynamic weighted clustering analysis based on the causal determination rules and the extracted feature values ​​to determine the corresponding target pressure prediction model and causal weight distribution.

[0157] Formation pressure prediction unit 14 is used to predict and back-deductively verify the target formation pressure by using the target pressure prediction model based on multi-source characteristic parameters, feature term extraction values ​​and causal weights, and generate the target formation pressure prediction value.

[0158] In this embodiment of the invention, the device further includes a preprocessing unit 15.

[0159] The preprocessing unit 15 is used to preprocess the multi-source feature parameters to obtain the preprocessed multi-source feature parameters.

[0160] In this embodiment of the invention, the multi-source feature parameters include porosity, acoustic transit time, rock composition, well logging fracture parameters, and electrical imaging. The extracted feature values ​​include undercompaction contribution, thorium-uranium ratio, pyrolytic hydrocarbon ratio, and fracture anisotropy. The feature extraction unit 12 is specifically used to generate undercompaction contribution based on porosity, acoustic transit time, and well logging fracture parameters; generate thorium-uranium ratio and pyrolytic hydrocarbon ratio based on rock composition; and generate fracture anisotropy based on electrical imaging.

[0161] In this embodiment of the invention, the weighted clustering analysis unit 13 is specifically used to perform dynamic weighted clustering analysis based on the feature extraction values ​​according to the cause determination rules, to determine the set of overpressure cause types and the cause weight distribution; and to determine the corresponding target pressure prediction model based on the set of overpressure cause types.

[0162] In this embodiment of the invention, the weighted clustering analysis unit 13 is specifically used to determine that the set of overpressure genesis types includes undercompaction genesis if the contribution of undercompaction is greater than a preset first threshold; to determine that the set of overpressure genesis types includes fluid expansion genesis if the thorium-uranium ratio is greater than a preset second threshold and the pyrolysis hydrocarbon ratio is greater than a preset third threshold; and to determine that the set of overpressure genesis types includes tectonic pressurization genesis if the fracture anisotropy is greater than a preset fourth threshold; and to assign weights to the overpressure genesis types in the set of overpressure genesis types to generate a geneisy weight distribution.

[0163] In this embodiment of the invention, the formation pressure prediction unit 14 is specifically used to generate an initial formation pressure prediction value based on multi-source feature parameters, feature term extraction values, and causal weights using a target pressure prediction model; perform reverse deduction on the initial formation pressure prediction value to obtain the inverted causal type; determine whether the inverted causal type is the same as the overpressure causal type set; if so, determine the initial formation pressure prediction value as the target formation pressure prediction value; if not, update the causal weights and trigger the weighted clustering analysis unit 13 to re-execute the step of determining the corresponding target pressure prediction model based on the overpressure causal type set.

[0164] In this embodiment of the invention, well logging data of overpressured carbonate rock areas are monitored to obtain multi-source characteristic parameters of the carbonate rock formation. Genetic features are extracted based on these multi-source characteristic parameters to obtain feature extraction values. Based on genetic determination rules, dynamic weighted clustering analysis is performed on the feature extraction values ​​to determine the corresponding target pressure prediction model and genetic weight distribution. Through the target pressure prediction model, pressure prediction and back-draft verification are performed based on the multi-source characteristic parameters, feature extraction values, and genetic weights to generate target formation pressure prediction values. Dynamic weighted clustering enables rapid automatic identification of overpressure causes and adaptive model matching, significantly improving data processing efficiency. Furthermore, the use of genetic weight allocation, multi-model collaborative prediction, and back-draft iterative correction mechanisms accurately characterizes the coupling effect of multiple overpressure mechanisms to significantly reduce prediction bias, thereby maintaining stable and reliable prediction performance in carbonate rock formations with strong heterogeneity and diverse overpressure causes.

[0165] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer device, specifically, a computer device can be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0166] This invention provides a computer device, including a memory and a processor. The memory stores information including program instructions, and the processor controls the execution of the program instructions. When the program instructions are loaded and executed by the processor, they implement the steps of the above-described embodiment of the carbonate rock overpressure formation pressure prediction method based on cluster analysis. For a detailed description, please refer to the above-described embodiment of the carbonate rock overpressure formation pressure prediction method based on cluster analysis.

[0167] The following is for reference. Figure 4 It shows a schematic diagram of the structure of a computer device 600 suitable for implementing the embodiments of this application.

[0168] like Figure 4 As shown, the computer device 600 includes a central processing unit (CPU) 601, which can perform various appropriate tasks and processes based on programs stored in read-only memory (ROM) 602 or programs loaded from storage section 608 into random access memory (RAM) 603. The RAM 603 also stores various programs and data required for the operation of the computer device 600. The CPU 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0169] The following components are connected to I / O interface 605: an input section 606 including a keyboard, mouse, etc.; an output section 607 including a cathode ray tube (CRT), liquid crystal feedback (LCD), etc., and speakers, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to I / O interface 605 as needed. A removable medium 611, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 610 as needed so that computer programs read from it can be installed in storage section 608 as needed.

[0170] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program tangibly embodied on a machine-readable medium, the computer program including program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 609, and / or installed from removable medium 611.

[0171] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0172] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0173] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0174] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0175] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0176] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0177] The acquisition, storage, use, and processing of data in this application all comply with the relevant provisions of national laws and regulations.

[0178] It should be noted that in the embodiments of this application, certain software, components, models and other existing solutions in the industry may be mentioned. These should be regarded as exemplary and are only intended to illustrate the feasibility of implementing the technical solution of this application. However, it does not mean that the applicant has used or necessarily used the solution.

[0179] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0180] This application can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0181] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0182] The above description is merely an embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of this application should be included within the scope of the claims of this application.

Claims

1. A method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis, characterized in that, The method includes: Monitoring well logging data in overpressured areas of carbonate rocks to obtain multi-source characteristic parameters of carbonate rock formations; Based on the multi-source feature parameters, causal feature extraction is performed to obtain feature term extraction values; Based on the cause determination rules, dynamic weighted clustering analysis is performed according to the extracted feature values ​​to determine the corresponding target pressure prediction model and cause weight distribution. The target formation pressure prediction value is generated by using the target pressure prediction model to perform pressure prediction and reverse inference verification based on the multi-source feature parameters, feature term extraction values ​​and causal weights.

2. The method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis according to claim 1, characterized in that, Before performing causal feature extraction based on the multi-source feature parameters to obtain feature term extraction values, the method further includes: The multi-source feature parameters are preprocessed to obtain preprocessed multi-source feature parameters.

3. The method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis according to claim 1, characterized in that, The multi-source characteristic parameters include porosity, sonic transit time, rock composition, well logging fracture parameters, and electrical imaging. The extracted characteristic values ​​include undercompaction contribution, thorium-uranium ratio, pyrolysis hydrocarbon ratio, and fracture anisotropy. The step of extracting causal features based on the multi-source feature parameters to obtain feature term extraction values ​​includes: Based on the porosity, sonic transit time, and logging fracture parameters, the undercompaction contribution is generated. Based on the rock composition, the thorium-uranium ratio and pyrolytic hydrocarbon ratio values ​​are generated; Based on the described electrical imaging pattern, crack anisotropy is generated.

4. The method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis according to claim 3, characterized in that, The step of determining the corresponding target pressure prediction model and causal weight distribution based on the causal determination rules and the extracted feature values ​​through dynamic weighted clustering analysis includes: Based on the cause determination rules, dynamic weighted clustering analysis is performed according to the extracted feature values ​​to determine the set of overpressure cause types and the cause weight distribution. Based on the set of overpressure cause types, the corresponding target pressure prediction model is determined.

5. The method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis according to claim 4, characterized in that, The step of determining the set of overpressure cause types and the cause weight distribution based on the extracted feature values ​​according to the cause determination rules includes: If the contribution of undercompaction is greater than a preset first threshold, the set of overcompression cause types is determined to include undercompaction causes. If the thorium-uranium ratio is greater than a preset second threshold and the pyrolysis hydrocarbon ratio is greater than a preset third threshold, the set of overpressure origin types is determined to include fluid expansion-related origins. If the crack anisotropy is greater than a preset fourth threshold, the set of overpressure cause types is determined to include structural pressurization causes. Weights are assigned to the overpressure cause types in the set of overpressure cause types to generate a cause weight distribution.

6. The method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis according to claim 4, characterized in that, Using the target pressure prediction model, pressure prediction and back-analysis are performed based on the multi-source feature parameters, feature term extraction values, and genetic weights to generate target formation pressure prediction values, including: Using the target pressure prediction model, an initial formation pressure prediction value is generated based on the multi-source feature parameters, feature term extraction values, and causal weights. The inverse extrapolation of the initial formation pressure prediction value yields the inverted genetic type; Determine whether the inversion cause type is the same as the set of overpressure cause types; If so, the initial formation pressure prediction value shall be determined as the target formation pressure prediction value; If not, update the causal weights and re-execute the step of determining the corresponding target pressure prediction model based on the set of overpressure causal types.

7. A device for predicting overpressure formation pressure in carbonate rocks based on cluster analysis, characterized in that, The device includes: The feature parameter acquisition unit is used to monitor well logging data in the overpressure zone of carbonate rocks and acquire multi-source feature parameters of carbonate rock layers. The feature extraction unit is used to extract causal features based on the multi-source feature parameters to obtain feature term extraction values. The weighted clustering analysis unit is used to perform dynamic weighted clustering analysis based on the causal determination rules and the extracted values ​​of the feature terms to determine the corresponding target pressure prediction model and causal weight distribution. The formation pressure prediction unit is used to predict and back-deductively verify the target formation pressure by using the target pressure prediction model based on the multi-source feature parameters, feature term extraction values ​​and causal weights, and generate the target formation pressure prediction value.

8. A computer-readable medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the method for predicting overpressure formations in carbonate rocks based on cluster analysis as described in any one of claims 1 to 6.

9. A computer device comprising a memory and a processor, the memory for storing information including program instructions, and the processor for controlling the execution of the program instructions, characterized in that, When the program instructions are loaded and executed by the processor, they implement the method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the method for predicting overpressure formation pressure in carbonate rocks based on cluster analysis as described in any one of claims 1 to 6.

Images