Method and device for determining a safe threshold of the fracture pressure of a rock of a carbon dioxide reservoir

Abstract

The application provides a method and device for determining a carbon dioxide reservoir rock fracture pressure safety threshold, and relates to the technical field of carbon dioxide storage, and comprises the following steps: obtaining geological parameters of a target reservoir and injection monitoring data and an injection stage when carbon dioxide is injected into the target reservoir; performing feature extraction on the injection monitoring data to obtain injection feature parameters; inputting the geological parameters and the injection feature parameters into a pre-trained reservoir mechanical state recognition model to obtain a mechanical state of the target reservoir; inputting the injection feature parameters into a pre-constructed rock fracture threshold prediction model to obtain a first rock fracture pressure safety threshold; if the first rock fracture pressure safety threshold is not greater than a preset fracture threshold, then the first rock fracture pressure safety threshold is corrected according to the mechanical state, the geological parameters and the injection stage to obtain a target rock fracture pressure safety threshold. The application can realize dynamic and accurate determination of the carbon dioxide reservoir rock fracture pressure safety threshold.

Description

Technical Field

[0001] This application relates to the field of carbon dioxide sequestration technology, and more specifically, to a method and apparatus for determining the safety threshold of rock fracture pressure in carbon dioxide reservoirs. Background Technology

[0002] With the large-scale application of carbon dioxide geological storage technology in greenhouse gas emission reduction, the safe management of the reservoir injection process has become a core critical factor for project implementation. During carbon dioxide injection operations, the mechanical state of the reservoir rock dynamically evolves with operating conditions such as injection pressure and flow rate. Rock fracturing pressure is a key indicator for defining safe injection boundaries and preventing reservoir fracturing and carbon dioxide leakage. Therefore, developing accurate methods for identifying reservoir mechanical states and a rock fracturing pressure threshold correction mechanism is crucial for ensuring the long-term stable and safe operation of carbon dioxide geological storage projects.

[0003] Currently, existing technologies all use empirical formulas combined with fixed coefficient corrections to calculate the fracture pressure threshold. They do not incorporate dynamic monitoring data during the injection process, and therefore cannot reflect the real-time impact of injection pressure, flow rate, and other operating conditions on the reservoir stress field. Consequently, they fail to reflect the reservoir's real-time pressure-bearing capacity, and the calculated fracture pressure threshold is unreliable. Furthermore, the formula parameters are uniform empirical values, which cannot be adapted to the different properties of different reservoirs, such as lithology, porosity, and fracture development. This results in a systematic deviation between the calculated results and the actual fracture pressure of the reservoir, leading to low accuracy. Summary of the Invention

[0004] The purpose of this application is to provide a method and apparatus for determining the safety threshold of rock fracture pressure in carbon dioxide reservoirs, thereby solving the above-mentioned problems existing in the prior art and enabling dynamic and accurate determination of the safety threshold of rock fracture pressure in carbon dioxide reservoirs.

[0005] Firstly, a method for determining the safety threshold of rock fracture pressure in a carbon dioxide reservoir is provided, the method including: Acquire geological parameters of the target reservoir, as well as injection monitoring data and injection stages during the injection of carbon dioxide into the target reservoir; Based on the geological parameters, feature extraction is performed on the injection monitoring data to obtain injection feature parameters; The geological parameters and the injection feature parameters are input into a pre-trained reservoir mechanical state recognition model to obtain the mechanical state of the target reservoir; The injected feature parameters are input into a pre-built rock fracture threshold prediction model to obtain the first rock fracture pressure safety threshold. If the first rock fracture pressure safety threshold is not greater than the preset fracture threshold, then the first rock fracture pressure safety threshold is corrected according to the mechanical state, the geological parameters and the injection stage to obtain the target rock fracture pressure safety threshold of the target reservoir.

[0006] In an optional implementation, the geological parameters include: geological type, lithological characteristic parameters, reservoir physical property parameters, and geological structure; The injection monitoring data includes: injection operating parameters, geophysical monitoring data, and rock mechanical response data; The injection characteristic parameters include: microseismic characteristics, injection dynamic characteristics, acoustic characteristics, and reservoir mechanical characteristics.

[0007] In an optional implementation, feature extraction is performed on the injection monitoring data based on the geological parameters to obtain injection feature parameters, including: Based on the geological parameters, spatiotemporal clustering calculations and magnitude-frequency fitting processing are performed on the geophysical monitoring data to obtain microseismic characteristics; Based on the geological parameters, a nonlinear model is constructed and a dynamic gradient is solved for the injection working condition parameters to obtain the injection dynamic characteristics; Based on the geological parameters, the acoustic signal data of the geophysical monitoring data is subjected to noise reduction, separation, and characteristic parameter calculation to obtain acoustic characteristics; Based on the geological parameters, the rock mechanical response data and the geophysical monitoring data are processed by mechanical parameter inversion, fracture feature quantification and response relationship fitting to obtain the reservoir mechanical characteristics.

[0008] In an optional implementation, the reservoir mechanical state identification model includes: The input layer is used to input the geological parameters and the injection feature parameters; A dual-parameter collaborative input verification layer is used to verify the injected feature parameters and the target reservoir geological parameters to obtain the first injected feature parameters and the first geological parameters that pass the verification, as well as the second injected feature parameters and the corresponding abnormal states that fail the verification, and the second geological parameters and the corresponding abnormal states. A cross-domain geological feature fusion layer is used to perform feature fusion on the first geological parameter, the first injected feature parameter, the second injected feature parameter, and the second geological parameter based on the abnormal state corresponding to the second injected feature parameter and the second geological parameter, so as to obtain a fused feature vector and geological and feature association weights; A multi-scale dynamic inference layer is used to perform bi-branch collaborative inference and anomaly feature-oriented enhancement processing on the fused feature vector and geological and feature correlation weights to obtain the initial reservoir mechanical state. The geological constraint dynamic calibration layer is used to verify and dynamically calibrate the initial reservoir mechanical state, the first geological parameter, the second injection characteristic parameter, and the second geological parameter using a geological constraint adaptive calibration method, so as to obtain the mechanical state of the target reservoir.

[0009] In an optional implementation, the rock fracture threshold prediction model may include: An injection feature input verification layer is used to preprocess the first injected feature parameters to obtain the third injected feature parameters; An injection feature quantization and parsing layer is used to perform feature parsing, weight allocation, and coupling calculation on the third injected feature parameters to obtain a feature coupling vector. A geological constraint embedded calculation layer is used to embed the first geological parameter into the feature coupling vector and then calculate the initial value and threshold calculation error range of the first rock fracture pressure safety threshold according to the configured carbon dioxide-rock coupled elastoplastic constitutive equation. The threshold precision calibration output layer is used to calibrate the initial value of the first rock fracture pressure safety threshold, the threshold calculation error range, and the third injection characteristic parameter to obtain the first rock fracture pressure safety threshold.

[0010] In an optional implementation, the first rock fracture pressure safety threshold is modified based on the mechanical state, the geological parameters, and the injection stage to obtain the target rock fracture pressure safety threshold for the target reservoir, including: Based on the mechanical state, the geological parameters, the injection characteristic parameters, and the injection stage, the first rock fracture pressure safety threshold is dynamically corrected in a zoned manner to obtain the second rock fracture pressure safety threshold. Based on the injection stage and the geological parameters, the second rock fracture pressure safety threshold is corrected to obtain the target rock fracture pressure safety threshold.

[0011] In an optional implementation, the method further includes: If the first rock fracture pressure safety threshold is greater than the preset fracture threshold, the second rock fracture pressure safety threshold is corrected according to the injection stage and geological parameters to obtain the target rock fracture pressure safety threshold.

[0012] Secondly, an apparatus for determining the safety threshold of rock fracture pressure in a carbon dioxide reservoir is provided, the apparatus comprising: The acquisition unit is used to acquire geological parameters of the target reservoir, as well as injection monitoring data and injection stages during the injection of carbon dioxide into the target reservoir. The extraction unit is used to extract features from the injection monitoring data based on the geological parameters to obtain injection feature parameters; The identification unit is used to input the geological parameters and the injection feature parameters into a pre-trained reservoir mechanical state identification model to obtain the mechanical state of the target reservoir. The prediction unit is used to input the injected feature parameters into a pre-built rock fracture threshold prediction model to obtain the first rock fracture pressure safety threshold. The correction unit is used to correct the first rock fracture pressure safety threshold according to the mechanical state, the geological parameters and the injection stage if the first rock fracture pressure safety threshold is not greater than the preset fracture threshold, so as to obtain the target rock fracture pressure safety threshold of the target reservoir.

[0013] Thirdly, an electronic device is provided, which includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; When a processor executes a program stored in memory, it implements any of the steps described in the first aspect above.

[0014] Fourthly, a computer-readable storage medium is provided, wherein a computer program is stored therein, and when executed by a processor, the computer program implements the steps of any of the methods described in the first aspect above.

[0015] This application integrates geological parameters and multi-source injection monitoring data, utilizes a reservoir mechanical state identification model to dynamically determine the current mechanical state of the reservoir, and combines this with a rock fracture threshold prediction model to generate a preliminary safety threshold. Then, it adaptively corrects the threshold based on the injection stage, mechanical state, and geological conditions, achieving real-time, accurate, and phased adjustment of the fracture pressure safety threshold. This significantly improves the safety margin and operational efficiency of the CO2 injection process, effectively avoiding engineering risks caused by threshold lag or inaccuracy. This application achieves dynamic and accurate determination of the rock fracture pressure safety threshold in carbon dioxide reservoirs, significantly improving the consistency between the threshold and the actual mechanical state of the reservoir; and ensuring the safety and efficiency of the injection process. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1An architecture diagram of a system for determining the safety threshold of rock fracture pressure in a carbon dioxide reservoir, provided in an embodiment of this application; Figure 2 A flowchart illustrating a method for determining the safety threshold of rock fracture pressure in a carbon dioxide reservoir, provided as an embodiment of this application; Figure 3 A schematic diagram of a device for determining the safety threshold of rock fracture pressure in a carbon dioxide reservoir, provided in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. Unless otherwise defined, the technical or scientific terms used in this application should have the ordinary meaning understood by those skilled in the art. The words "first," "second," and similar terms used in this application do not indicate any order, quantity, or importance, but are only used to distinguish different components. The words "comprising" or "including," etc., mean that the element or object preceding the word covers the element or object listed after the word and its equivalents, but do not exclude other elements or objects. The words "connected," "coupled," or "connected," etc., are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up," "down," "left," "right," etc., are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0019] The method for determining the safety threshold of carbon dioxide reservoir rock fracture pressure provided in this application embodiment can be applied to... Figure 1 In the system architecture shown, such as Figure 1As shown, the system may include a server and a data acquisition device. The server can be a physical server, a server cluster consisting of multiple physical servers, or a distributed system. It can also be a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms. The data acquisition device and the server can be connected directly or indirectly through wired or wireless communication methods; this application does not impose any limitations on this connection.

[0020] The preferred embodiments of this application are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit this application. Furthermore, the embodiments and features in the embodiments of this application can be combined with each other without conflict.

[0021] Figure 2 This is a flowchart illustrating a method for determining the safety threshold of rock fracture pressure in a carbon dioxide reservoir, as provided in an embodiment of this application. Figure 2 As shown, the method may include: Step S210: Obtain the geological parameters of the target reservoir, as well as the injection monitoring data and injection stages when injecting carbon dioxide into the target reservoir.

[0022] The geological parameters may include: geological type, lithological characteristics, reservoir physical properties, and geological structure; lithological characteristics may include: lithological type (limestone, sandstone, interbedded sandstone and mudstone, and basalt, etc.) and rock mechanical parameters (uniaxial / triaxial compressive strength, internal friction angle, cohesion, and elastic modulus); reservoir physical properties may include: porosity, permeability, pore type (primary pores / secondary pores), reservoir depth, and effective thickness; geological structure may include: whether faults / folds are present, fault strike / scale, and natural fracture development characteristics (fracture density, connectivity, strike, and opening). The data includes: injection condition parameters, geophysical monitoring data, and rock mechanical response data; injection condition parameters may include: injection pressure, injection flow rate, injection rate, real-time reservoir temperature and pressure, and cumulative injection volume; geophysical monitoring data may include: microseismic event data (source location, magnitude, occurrence time, and source mechanism) and acoustic signal data (original acoustic amplitude, propagation time, and wave train characteristics); rock mechanical response data may include: rock strain or displacement in the near-wellbore zone of the reservoir, real-time changes in rock permeability, and pore pressure around the wellbore.

[0023] In practice, geological parameters are determined through exploration data of the target reservoir (such as core analysis, logging curves, and seismic interpretation) and regional geological surveys. Among these, rock mechanics parameters can be obtained through laboratory rock sample testing or logging inversion. Injection condition parameters are collected in real time by wellhead and downhole sensors (such as pressure gauges, flow meters, and thermometers) of the injection well in the target reservoir. Geophysical monitoring data rely on continuous recording by microseismic monitoring arrays and acoustic logging / drilling acoustic systems deployed at the injection well site or surface of the target reservoir. Rock mechanics response data are dynamically obtained through fiber optic sensing (DAS / DTS), downhole strain gauges, repeated fracturing tests, or numerical inversion based on monitoring data.

[0024] Step S220: Based on geological parameters, feature extraction is performed on the injection monitoring data to obtain injection feature parameters.

[0025] The injection characteristic parameters may include: microseismic characteristics, injection dynamic characteristics, acoustic characteristics, and reservoir mechanical characteristics.

[0026] In practice A dual-mode quantization extraction method was used to perform spatiotemporal clustering calculations and magnitude-frequency fitting on the microseismic event-related data from the injected monitoring data, yielding microseismic characteristics. Specifically, based on the geological type of geological parameters, the reservoir's microseismic response sensitivity was predicted. If the geological type was low-microseismic-response, the microseismic characteristics only included b-value evolution characteristics, excluding spatiotemporal clustering morphological characteristics. If the geological type was not low-microseismic-response, the extracted microseismic characteristics included both b-value evolution characteristics and spatiotemporal clustering morphological characteristics. When valid source information (source location, magnitude, occurrence time, focal mechanism) was acquired, then... If a microseismic event is detected, it is considered to be present; if no microseismic event is detected, it is considered to be absent. When a microseismic event is present, the real-time b-value, the difference between the b-value and the historical mean, and the rate of change of the b-value are calculated using the magnitude-frequency linear fitting method, and combined to form the b-value evolution feature. The spatial clustering degree and time series frequency of the event are calculated using a spatiotemporal clustering algorithm, and combined to form the microseismic spatiotemporal clustering morphological feature. When there is no microseismic event, the background baseline value of the block's b-value is assigned as the b-value evolution feature, and the absence of microseismic state attribute label is simultaneously marked. The spatiotemporal clustering degree baseline value of the block is assigned as the microseismic spatiotemporal clustering morphological feature, and the absence of microseismic state attribute label is simultaneously marked. Nonlinear fitting and gradient calculation methods were used to construct nonlinear models and solve dynamic gradients for injection condition parameters in injection monitoring data to obtain injection dynamic characteristics. Specifically, reservoir temperature and pressure response characteristics were determined based on geological parameters and reservoir physical properties. For reservoirs with low temperature and pressure sensitivity, the injection dynamic characteristics included the pressure-flow nonlinear exponent, the rate of change of injection pressure, and the rate of change of flow rate. For reservoirs with low temperature and pressure sensitivity, the injection dynamic characteristics included the pressure-flow nonlinear exponent, reservoir temperature and pressure gradient characteristics, the rate of change of injection pressure, and the rate of change of flow rate. In particular, nonlinear fitting and gradient calculation methods were used to construct nonlinear models for injection condition parameters in injection monitoring data. The process involves constructing and dynamically solving gradient data to obtain the pressure-flow nonlinear index, injection pressure change rate, flow rate change rate, and reservoir temperature-pressure gradient characteristics. Specifically, this includes: constructing an injection pressure-flow fitting model using a nonlinear fitting algorithm, calculating the correlation coefficient and curvature of the fitted curves, and combining them to form the pressure-flow nonlinear index; dynamically calculating the real-time injection pressure and flow rate time series data using a first-order gradient solution method to obtain the real-time parameter change rates, which are used as the injection pressure change rate and flow rate change rate, respectively; and dynamically calculating the real-time reservoir temperature and pressure data using a first-order gradient solution method to obtain the real-time reservoir temperature-pressure gradient values, which are used as the reservoir temperature-pressure gradient characteristics. Wavelet transform and feature quantization methods are used to denoise and separate acoustic signals from geophysical monitoring data injected into the monitoring data, and to calculate characteristic parameters to obtain acoustic features. Specifically, the homogeneity of the acoustic response of lithology is determined based on geological parameters. If the lithology has a highly homogeneous acoustic response, the acoustic features can include acoustic velocity, real-time rate of change of acoustic velocity, and acoustic attenuation coefficient. If the lithology does not have a highly homogeneous acoustic response, the acoustic features can include acoustic velocity, real-time rate of change of acoustic velocity, acoustic attenuation coefficient, and real-time rate of change of acoustic attenuation coefficient. Specifically, the original acoustic signal is first processed using a wavelet transform algorithm to separate the effective acoustic signal from noise, resulting in a denoised effective signal. Based on the denoised effective signal, the real-time value of the acoustic velocity is calculated according to the propagation distance / propagation time, and the real-time rate of change of acoustic velocity is obtained through time series analysis. Based on the denoised effective signal, the real-time value of the acoustic attenuation coefficient is calculated according to the received amplitude / original amplitude, and the real-time rate of change of acoustic attenuation coefficient is obtained through time series analysis. The method employs parameter inversion and linear quantization to perform mechanical parameter inversion, fracture characteristic quantification, and response relationship fitting on geological parameters (lithological characteristic parameters, geological structures) and injection monitoring data (rock mechanical response data, geophysical monitoring data) to obtain reservoir mechanical characteristics. Based on the geological structure derived from the geological parameters, the integrity of the rock mass is determined. If the rock mass is of extremely intact type, the reservoir mechanical characteristics can include rock mass strength characteristics and injection response sensitivity characteristics; if it is not of extremely intact type, the reservoir mechanical characteristics can include rock mass strength characteristics, fracture development characteristics, and injection response sensitivity characteristics. Specifically, based on acoustic characteristics... Real-time wave velocity values ​​are obtained by inverting uniaxial compressive strength and internal friction angle using a rock mechanics wave velocity inversion model. These uniaxial compressive strength and internal friction angle are used as rock mass strength characteristics. Fracture development characteristics are obtained by combining the basic parameters of natural fractures in geological structures (density, connectivity) with microseismic spatiotemporal clustering morphology characteristics (or their benchmark values) and calculating a comprehensive fracture development index using a linear quantization method. This index serves as the fracture development characteristic. Based on real-time injection pressure and real-time rock strain values ​​from injection monitoring data, a pressure-strain response relationship is obtained through linear fitting. The ratio of strain change to pressure change is calculated and used as the injection response sensitivity characteristic.

[0027] Step S230: Input the geological parameters and injection characteristic parameters into the pre-trained reservoir mechanical state recognition model to obtain the mechanical state of the target reservoir.

[0028] The mechanical state can include: elastically stable state, pore pressure-dominated plastic deformation state, stable microcrack propagation state, or critical state of unstable crack propagation.

[0029] In one embodiment of this application, the reservoir mechanical state identification model may include: The input layer is used to input geological parameters and injection feature parameters; A dual-parameter collaborative input verification layer is used to perform joint compliance verification of injected feature parameters and target reservoir geological parameters using a dual-parameter linked compliance verification method. This results in the first injected feature parameter and the first geological parameter that pass verification, as well as the second injected feature parameter and its corresponding anomaly status, and the second geological parameter and its corresponding anomaly status, which fail verification. Anomaly status can include anomaly type and numerical deviation range. Specifically, for the injected feature parameters, based on preset feature dimension standards and reasonable numerical ranges (set based on rock mechanics theory and multi-reservoir historical data), dimension consistency and numerical rationality verification are performed. The first injected feature parameter that passes verification is selected, and the anomaly type (data error / reservoir mutation) and numerical deviation range are marked for the second injected feature parameter that fails verification. Next, for the geological parameters, based on the target reservoir exploration specifications, parameter integrity verification is performed. The first geological parameter that passes verification is selected, and the anomaly type (parameter missing / numerical anomaly) and numerical deviation range are marked for the second geological parameter that fails verification. Finally, based on the first geological parameter, the reasonable range of the corresponding injected feature parameter is predicted, and the accuracy of the second injected feature parameter marking is verified. The geological feature cross-domain fusion layer employs a geological-feature cross-domain attention fusion algorithm to quantize, encode, process anomalies, and perform deep fusion on the first geological parameter, first injected feature parameter, second injected feature parameter, and second geological parameter, resulting in a fused feature vector and geological and feature association weights. Specifically, the first geological parameter is first converted into a geological feature vector with the same dimensions as the first injected feature parameter using a normalized quantization encoding method. Then, for the second geological parameter, only its effective dimensional information is extracted (completely invalid missing parameters are removed) to assist in adjusting the encoding weights of the geological feature vector. Subsequently, based on the anomaly type of the second injected feature parameter, the first injected feature parameter is subjected to weighted avoidance processing (the fusion weight of the associated features corresponding to the second injected feature parameter of the "data error" category is reduced to a preset low value, while the associated features corresponding to the second injected feature parameter of the "reservoir mutation" category retain the original weights and are labeled with reservoir mutation tags). Finally, a cross-domain attention mechanism is introduced to calculate the association weights between the first geological parameter and the processed injected feature parameter. Based on the association weights, the geological feature vector and the processed injected feature parameter are deeply fused dimension by dimension to generate a fused vector while retaining the original core feature information. The multi-scale dynamic inference layer employs a multi-scale hierarchical inference algorithm to perform dual-branch collaborative inference and anomaly feature-oriented enhancement processing on the fused feature vector and geological and feature association weights, obtaining the initial reservoir mechanical state, initial identification confidence, and fusion weights of the dual-branch inference results. Specifically, the multi-scale dynamic inference layer can include: a coarse-grained inference branch and a fine-grained inference branch. The coarse-grained inference branch uses a lightweight random forest algorithm to perform fast classification inference on the fused feature vector based on geological and feature association weights to obtain the initial mechanical state. The fine-grained inference branch uses an improved SVM algorithm to first determine the specific dimensions and anomaly degree of reservoir mutation features based on the second injected feature parameters and the second geological parameters, then performs targeted weight enhancement on the mutation feature dimensions in the fused vector, focusing on fine-grained classification inference for mechanical states with small feature differences to obtain the initial reservoir mechanical state. Finally, the inference results of the two branches are dynamically fused based on the association weight table to calculate the initial identification confidence. The geologically constrained dynamic calibration layer is used to perform geological rule verification, anomaly correlation verification, and dynamic calibration processing on the initial reservoir mechanical state, first geological parameter, second injection characteristic parameter, and second geological parameter using a geologically constrained adaptive calibration method to obtain the mechanical state of the target reservoir. Specifically, firstly, based on the first geological parameter, a preset geological type-mechanical state judgment rule (set based on rock mechanics test data) is invoked to compare the initial identification confidence level with the confidence threshold of the corresponding geological type. Then, based on the second injection characteristic parameter and the second geological parameter, the correlation between the preliminary mechanical state and reservoir abrupt change characteristics and geological anomalies is verified (e.g., if the anomalous injection characteristic is a sudden drop in microseismic b-value, it is necessary to verify whether the preliminary state matches microfracture propagation / critical state). If the confidence level meets the standard and the state is consistent with the anomalous characteristics and geological anomalies, the result is directly confirmed. If it does not meet the standard, the fusion weight of the dual-branch inference result is adjusted, and the fine-grained inference branch is recalculated until the result meets the standard.

[0030] In one embodiment of this application, the reservoir mechanical state identification model adopts an improved random forest-SVM fusion model. Through a multi-task joint training mechanism, a large-scale multi-scenario training sample set is used to complete the training of the pre-trained basic model to determine the core parameters of the model. Different injection feature parameters of different reservoirs are input into the reservoir mechanical state identification model. A transfer learning mechanism is adopted to fine-tune the model feature weight parameters based on the pre-trained basic model. The model identification threshold (judgment boundary of four types of mechanical states) is set, and the judgment threshold of "critical state of unstable fracture propagation" is calibrated by combining rock mechanics test data of different reservoirs to ensure the accuracy of critical state identification.

[0031] Step S240: Input the injected feature parameters into the pre-built rock fracture threshold prediction model to obtain the first rock fracture pressure safety threshold.

[0032] The rock fracture threshold prediction model may include: The injection feature input verification layer is used to perform threshold prediction-specific verification processing on the first injection feature parameters using the injection feature-specific verification method to obtain the third injection feature parameters. Specifically, based on the feature requirements of rock fracture pressure calculation, the layer verifies the numerical integrity and dimensional uniformity of each first injection feature parameter (converted to engineering standard dimensions), marks null values ​​for differentially removed feature items, and performs standardization and normalization processing on feature values ​​(mapping to the 0-1 interval) to eliminate dimensional and magnitude differences, thereby obtaining the third injection feature parameters, a feature value normalization comparison table (marking the mapping relationship between the original value and the standardized value of each feature), and null value feature annotation items (only marked, not involved in subsequent calculations). The injection feature quantization analysis layer is used to perform feature analysis, weight allocation, and coupling calculation on the third injected feature parameters and feature value normalization lookup table using a multi-feature coupling quantization analysis algorithm to obtain feature coupling vectors. Specifically, based on rock mechanics theory (the correlation between rock mass fracture and various features), quantization weights are first assigned to each third injected feature parameter (e.g., rock mass strength feature weight 0.25, fracture development feature weight 0.2, b-value evolution feature weight 0.15). Then, feature parameters corresponding to null value feature annotations are removed, and coupling calculations are performed on the remaining effective feature parameters. Through weighted summation and feature interaction analysis, core feature components directly related to rock fracture pressure are extracted, and the influence of each feature on fracture pressure is quantified to obtain a feature quantization weight table. The geological constraint embedded calculation layer is used to perform geological constraint analysis and preliminary threshold calculation on the feature coupling vector, feature quantization weight table, and first geological parameters using a geological constraint quantization embedded algorithm. This yields the initial value of the first rock fracture pressure safety threshold and the threshold calculation error range. Specifically, the first geological parameters (lithology and physical properties) are first quantized and encoded into threshold calculation constraint coefficients (e.g., constraint coefficient of 1.05 for limestone reservoirs, 1.12 for tight reservoirs, and 0.98 for high-porosity reservoirs). Then, the constraint coefficients are embedded into the feature coupling vector, and combined with the feature quantization weight table, the initial value of the first rock fracture pressure safety threshold is calculated using the configured carbon dioxide-rock coupled elastoplastic constitutive equation. The threshold precision calibration output layer is used to perform precise calibration processing on the initial value of the first rock fracture pressure safety threshold, the threshold calculation error range, and the third injection characteristic parameters using an error adaptive calibration algorithm to obtain the first rock fracture pressure safety threshold. Specifically, the threshold calculation error range is first corrected based on the sensitive features (sonic wave velocity, rock mass strength) in the third injection characteristic parameters. Then, combined with the geological constraint coefficient reference table, the preliminary first rock fracture pressure safety threshold is adaptively calibrated to eliminate the calculation deviation in the geological constraint embedding process and ensure that the calibrated threshold is within a reasonable engineering range (based on multi-reservoir historical calculation data).

[0033] In another embodiment of this application, after obtaining the first rock fracture pressure safety threshold, the method may further include: A mechanical rationality verification + original model iterative correction algorithm is adopted to verify and iteratively correct the first rock fracture pressure safety threshold, the final mechanical state of the target reservoir, and the verified injection characteristic parameters. Specifically, based on the mechanical state of the target reservoir, a mechanical rationality judgment standard for the corresponding working condition is constructed (e.g., under the elastic steady state, the first threshold should be within the preset reasonable range of [15-25MPa]; under the critical state of unstable fracture propagation, the first threshold should be lower than 60% of the rock mass compressive strength). The first rock fracture pressure safety threshold is then compared with this standard to determine its mechanical rationality. If the verification passes, the verified first rock fracture pressure safety threshold is directly output. If the verification fails, the verified injection characteristic parameters are re-inputted based on the rock fracture threshold prediction model for iterative calculation to correct the initial threshold deviation until the verification passes, and the verified first rock fracture pressure safety threshold is obtained.

[0034] Step S250: If the first rock fracture pressure safety threshold is not greater than the preset fracture threshold, then the first rock fracture pressure safety threshold is corrected according to the mechanical state, geological parameters, injection characteristic parameters and injection stage to obtain the target rock fracture pressure safety threshold of the target reservoir.

[0035] The fracture threshold is pre-set based on regional reservoir safety exploitation standards.

[0036] In practice (1) When the first rock fracture pressure safety threshold is not greater than the preset fracture threshold, the sub-region fracture risk weight fusion and mechanical state constraint correction algorithm is adopted. According to the mechanical state, geological parameters, injection characteristic parameters and injection stage, the first rock fracture pressure safety threshold is dynamically corrected in a zone-differentiated manner to obtain the second rock fracture pressure safety threshold. Specifically, the rock mass strength characteristics, fracture development characteristics and injection response sensitivity characteristics in the injection characteristic parameters are fused and calculated through the weighted summation normalization algorithm to generate the sub-region fracture risk weight factor (which is dynamically updated with real-time injection characteristic data. The higher the risk, the smaller the factor, with a value range of 0.7-0.95). Then, according to the geological structure and lithological characteristics in the geological parameters, different sub-regions of the target reservoir are divided (such as fracture development sub-region, tight sub-region and high porosity sub-region). Combined with the physical constraints of the final mechanical state of the target reservoir (fracture unstable propagation critical state strengthening correction, elastic stable state moderate weakening correction), a corresponding correction magnitude is assigned to each sub-region. The mechanically verified threshold is multiplied by the fracture risk weight factor of the corresponding sub-region to complete the zone-differentiated correction and obtain the second rock fracture pressure safety threshold. Based on the injection stage and geological parameters, the safety threshold for the second rock fracture pressure is corrected to obtain the safety threshold for the target rock fracture pressure. Specifically, the target safety redundancy coefficient corresponding to the injection stage is matched from the configured table of different injection stages and different safety redundancy coefficients. Then, combined with the reservoir physical parameters in the geological parameters, the target safety redundancy coefficient is fine-tuned (the coefficient for tight reservoirs is increased by 5%, the coefficient for high-porosity reservoirs is decreased by 5%, and the coefficient for neutral reservoirs remains unchanged). Finally, the safety threshold for the second rock fracture pressure is multiplied by the fine-tuned target safety redundancy coefficient to obtain the safety threshold for the target rock fracture pressure. Among them, the safety redundancy coefficient for the pressurization stage is 1.15, the safety redundancy coefficient for the stabilization stage is 1.05, and the safety redundancy coefficient for the depressurization stage is 0.95. (2) When the first rock fracture pressure safety threshold is greater than the preset fracture threshold, the first rock fracture pressure safety threshold is corrected according to the injection stage and geological parameters to obtain the target rock fracture pressure safety threshold. Specifically, the target safety redundancy coefficient corresponding to the injection stage is matched from the configuration table of different injection stages and different safety redundancy coefficients. Then, the target safety redundancy coefficient is fine-tuned in combination with the reservoir physical parameters in the geological parameters (the coefficient of tight reservoir is increased by 5%, the coefficient of high porosity reservoir is decreased by 5%, and the coefficient of neutral reservoir remains unchanged). Finally, the second rock fracture pressure safety threshold is multiplied by the fine-tuned target safety redundancy coefficient to obtain the target rock fracture pressure safety threshold. Among them, the safety redundancy coefficient of the pressurization stage is 1.15, the safety redundancy coefficient of the steady injection stage is 1.05, and the safety redundancy coefficient of the depressurization stage is 0.95.

[0037] Corresponding to the above method, this application also provides a device for determining the safety threshold of rock fracture pressure in carbon dioxide reservoirs, such as... Figure 3 As shown, the device includes: The acquisition unit 310 is used to acquire geological parameters of the target reservoir, as well as injection monitoring data and injection stages when injecting carbon dioxide into the target reservoir. Extraction unit 320 is used to extract features from injection monitoring data based on geological parameters to obtain injection feature parameters; The identification unit 330 is used to input geological parameters and injection characteristic parameters into a pre-trained reservoir mechanical state identification model to obtain the mechanical state of the target reservoir. Prediction unit 340 is used to input the injected feature parameters into a pre-built rock fracture threshold prediction model to obtain the first rock fracture pressure safety threshold. The correction unit 350 is used to correct the first rock fracture pressure safety threshold according to the mechanical state, geological parameters, injection characteristic parameters and injection stage if the first rock fracture pressure safety threshold is not greater than the preset fracture threshold, so as to obtain the target rock fracture pressure safety threshold of the target reservoir.

[0038] The functions of each unit in the device for determining the safety threshold of carbon dioxide reservoir rock fracture pressure provided in the above embodiments of this application can be implemented through the above method steps. Therefore, the specific working process and beneficial effects of each unit in the device for determining the safety threshold of carbon dioxide reservoir rock fracture pressure provided in the embodiments of this application will not be repeated here.

[0039] This application also provides an electronic device, such as... Figure 4 As shown, it includes a processor 410, a communication interface 420, a memory 430, and a communication bus 440, wherein the processor 410, the communication interface 420, and the memory 430 communicate with each other through the communication bus 440.

[0040] Memory 430 is used to store computer programs; When the processor 410 executes the program stored in the memory 430, it performs the following steps: Acquire geological parameters of the target reservoir, as well as injection monitoring data and injection stages during the injection of carbon dioxide into the target reservoir; Based on geological parameters, feature extraction is performed on the injection monitoring data to obtain injection feature parameters; Geological parameters and injection characteristic parameters are input into a pre-trained reservoir mechanical state recognition model to obtain the mechanical state of the target reservoir. The injected feature parameters are input into a pre-built rock fracture threshold prediction model to obtain the first rock fracture pressure safety threshold. If the first rock fracture pressure safety threshold is not greater than the preset fracture threshold, then the first rock fracture pressure safety threshold is corrected according to the mechanical state, geological parameters, injection characteristic parameters and injection stage to obtain the target rock fracture pressure safety threshold of the target reservoir.

[0041] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0042] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0043] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0044] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0045] The implementation methods and beneficial effects of the various components of the electronic device in the above embodiments for solving the problem can be found in [reference needed]. Figure 2 The steps in the illustrated embodiments are used to implement the electronic device. Therefore, the specific working process and beneficial effects of the electronic device provided in this application will not be repeated here.

[0046] In another embodiment provided in this application, a computer-readable storage medium is also provided, which stores instructions that, when executed on a computer, cause the computer to perform the method for determining the safety threshold of carbon dioxide reservoir rock fracture pressure as described in any of the above embodiments.

[0047] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute the method for determining the safety threshold of rock fracture pressure in carbon dioxide reservoirs as described in any of the above embodiments.

[0048] Those skilled in the art will understand that the embodiments in this application can be provided as methods, systems, or computer program products. Therefore, the embodiments in this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the embodiments in this application can take the form of a computer program product implemented 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.

[0049] This application describes embodiments of methods, apparatus (systems), and computer program products according to embodiments of this application with reference to flowchart illustrations and / or block diagrams. 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. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0050] 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.

[0051] 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.

[0052] Although preferred embodiments have been described in this application, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of this application.

[0053] Obviously, those skilled in the art can make various modifications and variations to the embodiments of this application without departing from the spirit and scope of the embodiments of this application. Therefore, if these modifications and variations to the embodiments of this application fall within the scope of this application and its equivalents, then these modifications and variations are also intended to be included in the embodiments of this application.

Claims

1. A method for determining the safety threshold of rock fracture pressure in a carbon dioxide reservoir, characterized in that, The method includes: Acquire geological parameters of the target reservoir, as well as injection monitoring data and injection stages during the injection of carbon dioxide into the target reservoir; Based on the geological parameters, feature extraction is performed on the injection monitoring data to obtain injection feature parameters; The geological parameters and the injection feature parameters are input into a pre-trained reservoir mechanical state recognition model to obtain the mechanical state of the target reservoir; The injected feature parameters are input into a pre-built rock fracture threshold prediction model to obtain the first rock fracture pressure safety threshold. If the first rock fracture pressure safety threshold is not greater than the preset fracture threshold, then the first rock fracture pressure safety threshold is corrected according to the mechanical state, the geological parameters and the injection stage to obtain the target rock fracture pressure safety threshold of the target reservoir.

2. The method as described in claim 1, characterized in that, The geological parameters include: geological type, lithological characteristic parameters, reservoir physical property parameters, and geological structure; The injection monitoring data includes: injection operating parameters, geophysical monitoring data, and rock mechanical response data; The injection characteristic parameters include: microseismic characteristics, injection dynamic characteristics, acoustic characteristics, and reservoir mechanical characteristics.

3. The method as described in claim 2, characterized in that, Based on the geological parameters, feature extraction is performed on the injection monitoring data to obtain injection feature parameters, including: Based on the geological parameters, spatiotemporal clustering calculations and magnitude-frequency fitting processing are performed on the geophysical monitoring data to obtain microseismic characteristics; Based on the geological parameters, a nonlinear model is constructed and a dynamic gradient is solved for the injection working condition parameters to obtain the injection dynamic characteristics; Based on the geological parameters, the acoustic signal data of the geophysical monitoring data is subjected to noise reduction, separation, and characteristic parameter calculation to obtain acoustic characteristics; Based on the geological parameters, the rock mechanical response data and the geophysical monitoring data are processed by mechanical parameter inversion, fracture feature quantification and response relationship fitting to obtain the reservoir mechanical characteristics.

4. The method as described in claim 1, characterized in that, The reservoir mechanical state identification model includes: The input layer is used to input the geological parameters and the injection feature parameters; A dual-parameter collaborative input verification layer is used to verify the injected feature parameters and the target reservoir geological parameters to obtain the first injected feature parameters and the first geological parameters that pass the verification, as well as the second injected feature parameters and the corresponding abnormal states that fail the verification, and the second geological parameters and the corresponding abnormal states. A cross-domain geological feature fusion layer is used to perform feature fusion on the first geological parameter, the first injected feature parameter, the second injected feature parameter, and the second geological parameter based on the abnormal state corresponding to the second injected feature parameter and the second geological parameter, so as to obtain a fused feature vector and geological and feature association weights; A multi-scale dynamic inference layer is used to perform bi-branch collaborative inference and anomaly feature-oriented enhancement processing on the fused feature vector and geological and feature correlation weights to obtain the initial reservoir mechanical state. The geological constraint dynamic calibration layer is used to verify and dynamically calibrate the initial reservoir mechanical state, the first geological parameter, the second injection characteristic parameter, and the second geological parameter using a geological constraint adaptive calibration method, so as to obtain the mechanical state of the target reservoir.

5. The method as described in claim 4, characterized in that, The rock fracture threshold prediction model includes: An injection feature input verification layer is used to preprocess the first injected feature parameters to obtain the third injected feature parameters; An injection feature quantization and parsing layer is used to perform feature parsing, weight allocation, and coupling calculation on the third injected feature parameters to obtain a feature coupling vector. The geological constraint embedded calculation layer is used to embed the first geological parameter into the feature coupling vector and then calculate the initial value and threshold calculation error range of the first rock fracture pressure safety threshold according to the configured carbon dioxide-rock coupled elastoplastic constitutive equation. The threshold precision calibration output layer is used to calibrate the initial value of the first rock fracture pressure safety threshold, the threshold calculation error range, and the third injection characteristic parameter to obtain the first rock fracture pressure safety threshold.

6. The method as described in claim 1, characterized in that, Based on the mechanical state, the geological parameters, and the injection stage, the first rock fracture pressure safety threshold is modified to obtain the target rock fracture pressure safety threshold of the target reservoir, including: Based on the mechanical state, the geological parameters, the injection characteristic parameters, and the injection stage, the first rock fracture pressure safety threshold is dynamically corrected in a zoned manner to obtain the second rock fracture pressure safety threshold. Based on the injection stage and the geological parameters, the second rock fracture pressure safety threshold is corrected to obtain the target rock fracture pressure safety threshold.

7. The method as described in claim 6, characterized in that, The method further includes: If the first rock fracture pressure safety threshold is greater than the preset fracture threshold, the second rock fracture pressure safety threshold is corrected according to the injection stage and geological parameters to obtain the target rock fracture pressure safety threshold.

8. A device for determining the safety threshold of rock fracture pressure in a carbon dioxide reservoir, characterized in that, The device includes: The acquisition unit is used to acquire geological parameters of the target reservoir, as well as injection monitoring data and injection stages during the injection of carbon dioxide into the target reservoir. The extraction unit is used to extract features from the injection monitoring data based on the geological parameters to obtain injection feature parameters; The identification unit is used to input the geological parameters and the injection feature parameters into a pre-trained reservoir mechanical state identification model to obtain the mechanical state of the target reservoir. The prediction unit is used to input the injected feature parameters into a pre-built rock fracture threshold prediction model to obtain the first rock fracture pressure safety threshold. The correction unit is used to correct the first rock fracture pressure safety threshold according to the mechanical state, the geological parameters and the injection stage if the first rock fracture pressure safety threshold is not greater than the preset fracture threshold, so as to obtain the target rock fracture pressure safety threshold of the target reservoir.

9. An electronic device, characterized in that, The electronic device includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the method of any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1-7.

Images