A method, device and equipment for early warning of carbon dioxide storage leakage

Abstract

The application provides a carbon dioxide storage leakage early warning method, device and equipment, relates to the technical field of carbon storage engineering safety monitoring, and comprises the following steps: acquiring the space-time temperature field data and space-time strain field data in the wellbore of a carbon dioxide monitoring well, and the formation parameters of a sealing cap layer through which the carbon dioxide monitoring well passes and the cap layer-fault space topological atlas of the area where the carbon dioxide monitoring well is located; performing carbon dioxide leakage analysis and carbon dioxide leakage inversion respectively to obtain carbon dioxide leakage analysis results of the sealing cap layer and a carbon dioxide leakage amount; dynamically simulating the migration path and spatial distribution diagram of carbon dioxide in the sealing cap layer according to the carbon dioxide leakage analysis results, the carbon dioxide leakage amount and the formation parameters; and generating early warning information based on the carbon dioxide leakage analysis results, the migration path and the spatial distribution diagram. The application significantly improves the accuracy and timeliness of carbon dioxide storage leakage early warning.

Description

Technical Field

[0001] This application relates to the field of safety monitoring technology for carbon sequestration projects, and more specifically, to a method, apparatus, and equipment for early warning of carbon dioxide sequestration leaks. Background Technology

[0002] CO2 geological storage, a key negative emission technology for achieving the "dual carbon" goal, has been demonstrated and commercially deployed on a large scale globally. Its core lies in injecting captured CO2 into deep saline aquifers or depleted oil and gas reservoirs, relying on a dense overlying caprock for long-term sealing. However, if CO2 leakage occurs due to caprock rupture, fault activation, or wellbore failure, it will not only weaken the effectiveness of storage but may also pollute shallow groundwater, harm the ecological environment, and even threaten public safety. Therefore, establishing a highly sensitive, highly accurate, and quantifiable risk leakage early warning method is an urgent need to ensure the safe operation of CO2 storage.

[0003] Currently, mainstream monitoring methods include surface atmospheric CO2 concentration monitoring, groundwater chemical sampling, time-shifted seismic (4Dseismic) monitoring, and well geophysical logging. In recent years, distributed fiber optic sensing (DTS / DSS) technology has been widely used for temperature and strain field observation in sealed wells due to its ability to monitor the entire well section continuously and in real time. Some studies have attempted to combine temperature-strain anomalies with formation parameters to identify leakage signs through threshold criteria or simple inversion models, and to predict CO2 migration trends using numerical simulations.

[0004] The aforementioned existing technologies have three specific and interrelated technical problems: First, they rely solely on a single physical field (such as temperature) or a fixed threshold to determine leakage, failing to fully couple the thermo-mechanical-lithological response characteristics. This makes them susceptible to non-leakage factors such as injection-production interference and diurnal variations in geothermal temperature, resulting in a high false alarm rate. Second, leakage inversion generally ignores prior information about the leakage location and directly models the data for the entire well section, causing distortion in source strength estimation and failing to support accurate repair. Third, early warning information generation is mostly based on static distance or empirical rules, failing to dynamically integrate migration path topology, buffer layer retention capacity, and sensitive target exposure scenarios, making it difficult to provide a graded response strategy with engineering guidance value. Summary of the Invention

[0005] The purpose of this application is to provide a method, apparatus, and equipment for early warning of carbon dioxide storage leakage, which solves the above-mentioned problems existing in the prior art. It can generate early warning information reflecting the spatiotemporal evolution characteristics of leakage risk based on multi-source monitoring and geological data fusion, and significantly improve the accuracy and timeliness of early warning.

[0006] Firstly, a method for early warning of carbon dioxide storage leakage is provided, which may include: Acquire spatiotemporal temperature field data and spatiotemporal strain field data inside the carbon dioxide monitoring well, as well as the formation parameters of the caprock through which the carbon dioxide monitoring well passes and the caprock-fault spatial topology map of the area where the carbon dioxide monitoring well is located; Carbon dioxide leakage analysis was conducted based on spatiotemporal temperature field data, spatiotemporal strain field data, formation parameters, and caprock-fault spatial topology map to obtain the carbon dioxide leakage analysis results of the caprock. Based on the carbon dioxide leakage analysis results, spatiotemporal temperature field data, spatiotemporal strain field data, and formation parameters, carbon dioxide leakage was inverted to obtain the amount of carbon dioxide leakage. Based on the carbon dioxide leakage analysis results, carbon dioxide leakage amount and formation parameters, dynamic simulation was used to obtain the migration path and spatial distribution map of carbon dioxide in the caprock. Early warning information is generated based on the analysis results of carbon dioxide leakage, migration paths, and spatial distribution maps.

[0007] In an optional implementation, carbon dioxide leakage analysis is performed based on the spatiotemporal temperature field data, the spatiotemporal strain field data, the formation parameters, and the caprock-fault spatial topology map to obtain the carbon dioxide leakage analysis results of the caprock, including: The spatiotemporal temperature field data, the spatiotemporal strain field data, and the formation parameters are input into a pre-trained thermo-mechanical-lithological coupled anomaly identification model to obtain anomaly identification results for different regions in the caprock; wherein, the anomaly identification results include: coupling characteristics of different regions; The region corresponding to the coupling feature whose similarity to each leakage feature in the configured carbon dioxide leakage feature library is higher than the configured similarity threshold is taken as the target region. The formation parameters, the anomaly identification results of the target area, and the caprock-fault spatial topology map are input into a pre-trained carbon dioxide leakage cause classification and location model to obtain the leakage cause and location of carbon dioxide leakage in the caprock. Based on the anomaly identification results in different regions, the cause of carbon dioxide leakage in the capping layer, and the location of the leakage, the carbon dioxide leakage analysis results of the capping layer are obtained.

[0008] In an optional implementation, the anomaly identification result further includes: the spatial location of different regions; The formation parameters, the anomaly identification results of the target area, and the caprock-fault spatial topology map are input into a pre-trained carbon dioxide leakage cause classification and location model to obtain the leakage causes and locations of carbon dioxide leakage in the caprock, including: Based on the spatial location of the target area, extract the target geological structure information corresponding to the target area from the cap-fault spatial topology map; Based on the spatial location of the target area, extract the target stratigraphic parameters corresponding to the target area from the stratigraphic parameters; By inputting the target stratigraphic parameters of the target area, the anomaly identification, and the target geological structure information into the carbon dioxide leakage cause classification and location model, the leakage cause and location of carbon dioxide leakage in the caprock are obtained.

[0009] In an optional implementation, carbon dioxide leakage is inverted based on carbon dioxide leakage analysis results, spatiotemporal temperature field data, spatiotemporal strain field data, and formation parameters to obtain the carbon dioxide leakage amount, including: The inversion range is determined based on the leak location identified in the carbon dioxide leak analysis results. Target spatiotemporal temperature data and target spatiotemporal strain data within the inversion range are extracted from the spatiotemporal temperature field data and the spatiotemporal strain field data, respectively. From the stratigraphic parameters, extract the stratigraphic parameters of the inversion range to obtain the stratigraphic parameters to be inverted; Based on the formation parameters to be inverted and the leakage causes in the carbon dioxide leakage analysis results, a coupling mechanism model of CO2 leakage amount and anomaly parameters is constructed. Substituting the target spatiotemporal temperature field data and the target spatiotemporal strain field data into the CO2 leakage amount-anomaly parameter coupling mechanism model, the initial CO2 leakage amount is obtained; The initial CO2 leakage amount is corrected using the formation parameters to be inverted, and the carbon dioxide leakage amount is obtained.

[0010] In an optional implementation, the carbon dioxide leakage cause classification and location model includes: The spatial feature fusion layer is used to spatially fuse the target stratigraphic parameters of the input target area, the anomaly identification combination, and the target geological structure information to obtain a fusion feature matrix. The coupled feature enhancement layer is used to extract target fusion features from the fusion feature matrix through shallow neural networks and geological mechanism constraints; The dual-task inference layer includes a leakage cause classification sublayer and a leakage point three-dimensional localization sublayer; it is used to determine the leakage cause and leakage location of carbon dioxide leakage in the capping layer based on the target fusion characteristics.

[0011] In an optional implementation, early warning information is generated based on carbon dioxide leakage analysis results, migration paths, and spatial distribution maps, including: Multi-dimensional feature extraction and spatiotemporal alignment processing are performed on the carbon dioxide leakage analysis results, the migration path, and the spatial distribution map to obtain a comprehensive risk feature matrix. Based on the aforementioned comprehensive risk characteristic matrix, hierarchical early warning information is provided.

[0012] In an optional implementation, based on the comprehensive risk feature matrix, hierarchical early warning information includes: The comprehensive risk feature matrix is ​​input into a pre-trained risk assessment model for dynamic risk level assessment to obtain a risk level sequence that changes over time. The transport path and spatial distribution map are subjected to three-dimensional visualization processing based on adaptive mesh refinement and kriging interpolation to generate a three-dimensional risk heat map that is dynamically updated over time. The three-dimensional risk heat map is spatially overlaid and buffered with a preset database of spatial distribution of sensitive targets containing multiple sensitive targets to obtain the expected impact time, maximum exposure concentration and comprehensive risk index for each sensitive target. The cause of the leak, the risk level sequence, the expected impact time of each sensitive target, the maximum exposure concentration, and the comprehensive risk index are processed by multi-criteria decision analysis based on fuzzy logic reasoning. Combined with a dynamically adjusted early warning rule base, hierarchical early warning information is generated.

[0013] Secondly, a carbon dioxide sequestration leak early warning device is provided, which may include: The acquisition unit is used to acquire spatiotemporal temperature field data and spatiotemporal strain field data inside the carbon dioxide monitoring well, as well as the formation parameters of the caprock through which the carbon dioxide monitoring well passes and the caprock-fault spatial topology map of the area where the carbon dioxide monitoring well is located. The analysis unit is used to perform carbon dioxide leakage analysis based on spatiotemporal temperature field data, spatiotemporal strain field data, formation parameters and caprock-fault spatial topology map, and obtain the carbon dioxide leakage analysis results of the caprock. The inversion unit is used to perform carbon dioxide leakage inversion based on carbon dioxide leakage analysis results, spatiotemporal temperature field data, spatiotemporal strain field data, and formation parameters to obtain the amount of carbon dioxide leakage. The simulation unit is used to dynamically simulate the migration path and spatial distribution map of carbon dioxide in the caprock based on the carbon dioxide leakage analysis results, carbon dioxide leakage amount and formation parameters. The generation unit is used to generate early warning information based on carbon dioxide leakage analysis results, migration paths, and spatial distribution maps.

[0014] 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.

[0015] 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.

[0016] This application utilizes monitoring wellbore and caprock data to conduct leakage analysis, achieving accurate identification of carbon dioxide leakage in the caprock. Based on the leakage analysis results, combined with wellbore monitoring data and caprock formation parameters, leakage inversion is performed, enabling effective quantification of carbon dioxide leakage. Dynamic simulation, combining leakage analysis results, quantified leakage, and formation parameters, accurately reveals the migration path and spatial distribution of carbon dioxide within the caprock, clearly presenting its migration patterns and distribution characteristics. Early warning information is generated based on leakage analysis results, migration paths, and spatial distribution maps, enabling targeted early warning of carbon dioxide storage leakage, timely feedback on caprock leakage risk status, and effective guidance for the prevention and control of carbon dioxide storage leakage. This improves the safety and management efficiency of the carbon dioxide storage process, ensuring the stable operation of carbon dioxide geological storage. Attached Figure Description 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 1 An architecture diagram of an early warning system for carbon dioxide storage leakage provided in an embodiment of this application; Figure 2 A flowchart illustrating an early warning method for carbon dioxide storage leakage provided in an embodiment of this application; Figure 3 A schematic diagram of the structure of an early warning device for carbon dioxide storage leakage 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 early warning method for carbon dioxide sequestration leakage provided in this application embodiment can be applied to... Figure 1 In the system architecture shown, such as Figure 1 As shown, the system may include a server and a terminal. 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 terminal may be a user equipment (UE) such as a mobile phone, smartphone, laptop, digital radio receiver, personal digital assistant (PDA), tablet computer (PAD), handheld device, in-vehicle device, wearable device, computing device, or other processing device connected to a wireless modem, mobile station (MS), mobile terminal, etc. The terminal and server can be directly or indirectly connected via wired or wireless communication methods, which is not limited herein.

[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 an early warning method for carbon dioxide storage leakage provided in an embodiment of this application. Figure 2 As shown, the method may include: Step S210: Obtain spatiotemporal temperature field data and spatiotemporal strain field data inside the carbon dioxide monitoring well, as well as the formation parameters of the caprock through which the carbon dioxide monitoring well passes and the caprock-fault spatial topology map of the area where the carbon dioxide monitoring well is located.

[0022] Among them, the spatiotemporal temperature field data is a three-dimensional data volume (depth-time-temperature) composed of continuously measured temperature values ​​at different depths (space) and time points within the carbon dioxide monitoring well shaft; the spatiotemporal strain field data refers to a three-dimensional data volume composed of minute deformations (strain) or acoustic vibration signals of rock or cement sheath continuously measured at different depths (space) and time points along the carbon dioxide monitoring well shaft; the spatiotemporal strain field data may include strain amplitude / microseismic signals, strain frequency / spectral characteristics, and spatiotemporal evolution patterns of strain / acoustic waves; the caprock may include: buffer layer and caprock; the formation parameters of the caprock may include: reservoir physical properties, rock mechanical parameters, thermophysical parameters, and formation geometric parameters; The physical properties of the caprock can include: porosity, permeability, pore throat distribution, rock density, and pore volume compressibility (which determines the transport capacity of CO2 in the caprock); the rock mechanical parameters can include: elastic modulus, Poisson's ratio, compressive strength, tensile strength, and cohesion (which reflect the mechanical properties of the caprock and are related to strain field changes and formation fracturing risk); the thermophysical parameters can include: rock thermal conductivity, specific heat capacity, and thermal diffusivity (which reflect the thermal conduction characteristics of the caprock and are related to temperature field changes and CO2 phase transition heat effect); and the formation geometric parameters can include: the thickness, burial depth, lateral distribution range, and intersection section and depth of the caprock with the monitoring wellbore (which clarifies the formation spatial range corresponding to the parameters). The caprock-fault spatial topology map is a three-dimensional spatial geometric distribution map of the caprock and faults in the area where the carbon dioxide monitoring well is located. It is used to characterize the topological relationship between the spatial distribution of the caprock and the spatial structure of the fault. The caprock-fault spatial topology map can include: spatial characteristics of the caprock, spatial characteristics of the fault, and the topological relationship between the caprock and the fault. The spatial characteristics of the caprock can include: the three-dimensional spatial burial depth, thickness variation, lateral distribution boundary, bedding attitude (dip angle, strike) of the caprock (buffer layer + caprock), lithological distribution of different layers, and the spatial crossing position of the monitoring well shaft in the caprock. The spatial characteristics of faults can include: the three-dimensional spatial location, strike, dip, dip angle, displacement, extension depth, fault zone width, fault sealing level (closed / semi-closed / open) of all faults within the monitoring well area, as well as the intersection relationship between the fault and the caprock (such as whether the fault cuts through the caprock, the location and depth of the cut, and whether it intersects with the monitoring well shaft); the topological relationship includes: the spatial contact mode between the fault and the caprock, the spatial distribution of microfractures within the caprock, and the relative position of the monitoring well and the key structures of the fault / caprock. All features are accompanied by a three-dimensional coordinate system (such as geodetic coordinates and well shaft coordinates) to achieve precise spatial positioning.

[0023] In practice, the spatiotemporal temperature field data is obtained by measuring the spatiotemporal temperature field data through a distributed optical fiber temperature monitoring system pre-installed in the carbon dioxide monitoring well. The optical fiber sensing cable is laid along the entire section of the carbon dioxide monitoring well (fitted to the inner wall of the casing or fixed to the well wall). Utilizing the Rayleigh scattering or Raman scattering effect of the optical fiber, continuous, distributed, and real-time monitoring of the temperature at each measuring point in the well is achieved. The system automatically collects and stores the temperature data at each measuring point at different times. After data preprocessing (noise removal and missing value completion), standardized spatiotemporal temperature field data is formed. Spatiotemporal strain field data are obtained through a distributed optical fiber strain monitoring system (DVS), which is deployed in the same well section as the temperature monitoring optical fiber (or uses an integrated temperature-strain optical fiber cable). Utilizing the Brillouin scattering effect of optical fiber, it captures the minute deformations of rock or casing caused by formation stress changes and CO2 migration at various measuring points in the wellbore. The system automatically collects strain data at different times at each measuring point. After data calibration (eliminating the interference of temperature on optical fiber strain monitoring) and preprocessing, standardized spatiotemporal strain field data are formed. Formation parameters of the caprock are obtained through drilling, logging, and core experiments.

[0024] In one embodiment of this application, the method for obtaining a caprock-fault spatial topology map may include: Acquire seismic exploration data (two-dimensional profiles or three-dimensional data volumes) of the area where the carbon dioxide monitoring well is located, well logging data (such as sonic, density and resistivity data), well geological stratification data, core description data and regional geological maps. Noise suppression, amplitude recovery, and offset repositioning are performed on the seismic exploration data to obtain preprocessed seismic exploration data; environmental correction and standardization are performed on the well logging curve data to obtain preprocessed well logging curve data; the drilling geological stratification data and the preprocessed well logging curve data are depth aligned to obtain a geological stratification data table. Stratigraphic tracking and fault interpretation were performed on the preprocessed seismic exploration data to obtain data on the top and bottom interfaces of the cap layer and the spatial distribution of faults. The geological stratification data table is fused with the preprocessed well logging curve data. Based on the fused data, the caprock top and bottom interface data, and the fault spatial distribution data, well-seismic joint calibration is performed to obtain the calibrated caprock-fault interpretation results. Based on the calibrated caprock-fault interpretation results, combined with the fault rock type and caprock lithology data in the core description data and the regional geological map, the sealing of each fault was evaluated, and fault sealing attribute data were obtained. This paper spatially integrates caprock top-bottom interface data, fault spatial distribution data, and fault sealing attribute data, and combines them with regional geological maps to construct a caprock-fault spatial topology map of the area where the carbon dioxide monitoring well is located. Specifically, the caprock top-bottom interface data is constructed into a continuous three-dimensional triangular mesh surface to form a caprock geometric model; the fault spatial distribution data is constructed into a fault surface triangular mesh model with spatial location and geometric shape; and the fault sealing attribute data is assigned as attribute fields to the corresponding fault surface triangular mesh units. Based on the stratigraphic attitude, structural strike, and regional stress background provided in the regional geological map, the relative position and contact state of the caprock geometric model and the fault surface triangular mesh model in three-dimensional space are verified for consistency, and local geometric shapes that obviously do not conform to the regional tectonic rules are corrected. Under a unified three-dimensional coordinate system, the verified and corrected caprock geometric model, the attributed fault surface triangular mesh model, and regional geological constraint information are fused to generate a caprock-fault spatial topology map that includes caprock spatial distribution, fault three-dimensional morphology, fault-caprock spatial configuration relationship, and fault sealing attributes.

[0025] Step S220: Based on the spatiotemporal temperature field data, spatiotemporal strain field data, formation parameters, and caprock-fault spatial topology map, perform carbon dioxide leakage analysis to obtain the carbon dioxide leakage analysis results of the caprock.

[0026] In practice, carbon dioxide leakage analysis is performed based on spatiotemporal temperature field data, spatiotemporal strain field data, formation parameters, and caprock-fault spatial topology maps to obtain the carbon dioxide leakage analysis results of the caprock, including: Spatiotemporal temperature field data, spatiotemporal strain field data, and formation parameters are input into a pre-trained thermo-mechanical-lithological coupled anomaly identification model to obtain anomaly identification results for different regions in the caprock. The anomaly identification results include: the anomaly state of different regions and the coupling characteristics, spatial location, and duration of anomalies in regions with anomalies. The regions corresponding to coupling features whose similarity to leakage features in the configured carbon dioxide leakage feature library is higher than a configured similarity threshold are designated as target regions. The carbon dioxide leakage feature library is a pre-built, dedicated feature library that stores a large number of typical thermal-mechanical-lithological coupling features of the caprock caused by CO2 migration / leakage from the storage layer to the caprock. For any region's coupling feature, the similarity between that coupling feature and each leakage feature in the carbon dioxide leakage feature library is calculated, resulting in a similarity value for each coupling feature. If any similarity value is higher than a preset similarity threshold, the region corresponding to that coupling feature is determined to be an abnormal region with CO2 leakage, i.e., the target region; if the similarity is lower than the threshold, it is determined to be a non-leakage abnormal region. The anomaly identification results and caprock-fault spatial topology map of the target area are input into a pre-trained carbon dioxide leakage cause classification and location model to obtain the leakage cause and location of carbon dioxide leakage in the caprock. Specifically, based on the spatial location of the target area, the geological structure information corresponding to the target area is extracted from the caprock-fault spatial topology map. This geological structure information may include: whether the target area is close to a fault, whether there are micro-fractures in the caprock, and the location of casing installations, etc. Based on the spatial location of the target area, the target stratigraphic parameters corresponding to the target area are extracted from the stratigraphic parameters. The target stratigraphic parameters, anomaly identification results, and target geological structure information of the target area are input into the carbon dioxide leakage cause classification and location model to obtain... The model identifies the causes and locations of carbon dioxide leaks in the caprock; it classifies and locates CO2 leaks based on the coupling characteristics of the target area and the integrated geological structure information, accurately classifying them into preset cause types such as caprock microfracture seepage, fault activation gas conduction, or casing damage leakage; based on the coupling characteristics of the target area, the spatial distribution of stratigraphic parameters, and the three-dimensional coordinate system of the caprock-fault spatial topology map, the model calculates and outputs the precise three-dimensional spatial location of the CO2 leak in the caprock (such as the burial depth of the leak point, plane coordinates, and the caprock segment / fault zone to which it belongs). If it is a multi-point leak / fault activation leak, it will also output the extension direction of the leak channel. Based on the anomaly identification results in different regions, the causes and locations of carbon dioxide leakage in the capping layer, the analysis results of carbon dioxide leakage in the capping layer were obtained.

[0027] In one embodiment of this application, the thermo-mechanical-lithological coupling anomaly identification model may include: The mechanism coupling layer, including the heat conduction control equation, rock mechanics constitutive equation, and lithological parameter coupling correlation formula, is used to substitute the input spatiotemporal temperature field data, spatiotemporal strain field data, and formation parameters into the heat conduction control equation, rock mechanics constitutive equation, and lithological parameter coupling correlation formula for coupling calculations, obtaining the thermal-mechanical-lithological coupled field distribution data of each region of the caprock. The thermal-mechanical-lithological coupled field distribution data of each region of the caprock are coupled values ​​with three-dimensional spatial coordinates and time series, which may include: the linked values ​​of effective stress, temperature gradient, and permeability distribution in each region. The heat conduction control equation is used to describe the spatiotemporal variation law of the temperature field within the caprock, combined with CO2... The phase change heat effect (endothermic / exothermic) correction correlates temperature with formation thermophysical parameters (thermal conductivity, specific heat capacity), achieving coupling between the temperature field and lithological parameters; the rock mechanics constitutive equation (linear elastic / elastoplastic) is used to describe the stress-strain relationship of the caprock / well casing, correlating the strain field with rock mechanics parameters (elastic modulus, Poisson's ratio), achieving coupling between the strain field and lithological parameters; the lithological parameter coupling correlation formula is used to describe the linkage between porosity, permeability and stress / temperature (e.g., increased effective stress leads to decreased porosity and permeability; temperature changes cause rock expansion / contraction, altering the pore throat distribution), achieving bidirectional coupling between temperature, strain, and lithological parameters; The feature extraction layer is used to perform feature dimensionality reduction and key feature extraction on the thermal-mechanical-lithological coupling field distribution data of each region using a pre-trained fully connected neural network (FCNN, 2-3 hidden layers) or an autoencoder, so as to obtain the thermal-mechanical-lithological coupling feature vector of each region. The anomaly detection layer uses a pre-trained logistic regression (LR) or support vector machine (SVM) classifier to compare the thermal-mechanical-lithological coupling feature vectors of each region with a pre-trained baseline library of normal coupling features of the caprock, and calculates the feature deviation value. If the deviation value exceeds the model's preset anomaly detection threshold, the region is determined to be an anomaly region; otherwise, it is a normal region. The result of determining whether a region is normal, the thermal-mechanical-lithological coupling feature vector of the region determined to be an anomaly, the basic spatial location of the anomaly region, and the duration of the anomaly are output as the anomaly identification result.

[0028] In another embodiment of this application, when training a fully connected neural network (FCNN, 2-3 hidden layers) or an autoencoder, coupled field data under normal conditions of the capping layer are used as training samples, while coupled field data under perturbation states such as CO2 transport and pure geostress changes are used as validation samples. When training a logistic regression (LR) or support vector machine (SVM) classifier, a large number of coupled feature vector samples of normal / abnormal states are used to optimize the decision threshold and weight of logistic regression / SVM, so that the model can accurately distinguish between normal and abnormal states of the capping layer, while ensuring the robustness of abnormal determination.

[0029] In another embodiment of this application, the carbon dioxide leakage cause classification and location model includes: The spatial feature fusion layer is used to spatially fuse the target stratigraphic parameters, anomaly identification, and target geological structure information of the input target area to obtain a fused feature matrix. Specifically, the spatial location in the anomaly identification results of the target area is converted into geodetic three-dimensional coordinates consistent with the caprock-fault spatial topology map; the caprock is divided into three-dimensional grids according to a preset precision, and the anomaly data and target stratigraphic parameters of the target area are mapped to the corresponding grids according to coordinates to form a grid-level anomaly-geological fusion dataset; the target geological structure features of the target area and its surroundings are extracted from the caprock-fault spatial topology map and converted into standardized structural feature vectors; the coupled feature vectors, anomaly features, and temporal evolution features of the target area are concatenated with the extracted structural feature vectors according to the grid spatial location to generate a grid-level fused feature matrix. A coupling feature enhancement layer is used to extract target fusion features from the fusion feature matrix through shallow neural networks and geological mechanism constraints. Specifically, based on the geological mechanism of CO2 migration and preset feature association rules, features strongly related to the cause and location of the leak are extracted from the fusion feature matrix to obtain the first fusion feature. The first fusion feature is input into a fully connected neural network with two hidden layers. The fully connected neural network aims to maximize the discriminative power of the leak cause and location-related features. Through backpropagation, the weights are iteratively optimized to perform nonlinear mapping and enhancement on the first fusion feature, extracting the leak-specific fusion feature vector to obtain the second fusion feature. The second fusion feature is normalized to eliminate dimensional differences, and then dimensionality is reduced through principal component analysis to obtain the target fusion feature. The dual-task inference layer includes a leakage cause classification sublayer and a leakage point three-dimensional localization sublayer; it is used to determine the leakage cause and leakage location of carbon dioxide leakage in the capping layer based on the target fusion characteristics. The leakage cause classification sublayer uses a Lightweight Gradient Boosting Decision Tree (LightGBM) classifier to perform multi-class classification of leakage causes in the target area based on pre-trained leakage cause-core feature association patterns and target fusion features, and outputs the probability values ​​of each cause. The cause type with the highest probability value is taken as the final leakage cause. If the probability value of a certain cause is higher than a preset confidence threshold, it is directly determined; if all are lower than the threshold, manual correction is performed in combination with geological structural patterns. A three-dimensional leakage point location sublayer is used to determine the quantitative relationship between the distance of anomalies in the target fusion features and the leakage point based on the CO2 seepage mechanics equation (Darcy's law) and the formation stress propagation formula (anomalies are negatively correlated with the distance of the leakage point, i.e., the closer to the leakage point, the higher the anomaly value). The anomalies and permeability anisotropy parameters in the target fusion features are substituted into the quantitative relationship formula, and combined with the three-dimensional coordinates of the grid, the probability value of each grid as a leakage point is calculated, generating a leakage point probability distribution cloud map. The accuracy of the probability distribution cloud map is optimized according to the leakage cause output by branch 1. The three-dimensional center coordinates of the grid with the highest probability in the probability distribution cloud map are taken as the three-dimensional spatial location of the leakage point. If it is a leakage channel (such as fault activation), the extension direction and distribution range of the leakage channel are also output.

[0030] Step S230: Based on the carbon dioxide leakage analysis results, spatiotemporal temperature field data, spatiotemporal strain field data and formation parameters, perform carbon dioxide leakage inversion to obtain the amount of carbon dioxide leakage.

[0031] In practice, the inversion range is determined with the leak location in the carbon dioxide leak analysis results as the core. The inversion range can be a three-dimensional area of ​​30-50m around the leak point, covering the entire target area. Target spatiotemporal temperature data and target spatiotemporal strain data within the inversion range are extracted from spatiotemporal temperature field data and spatiotemporal strain field data, respectively. The extraction method for target spatiotemporal temperature data and target spatiotemporal strain data includes: subtracting the normal geological background value of the capping layer (the normal benchmark value pre-stored by the thermo-mechanical-lithological coupling anomaly identification model) from the spatiotemporal temperature field data and spatiotemporal strain field data within the inversion calculation range, and extracting the temperature anomaly fluctuation component and strain anomaly fluctuation component caused only by CO2 leakage to obtain the target spatiotemporal temperature field data and target spatiotemporal strain field data. Extract the stratigraphic parameters of the inversion range from the stratigraphic parameters to obtain the stratigraphic parameters to be inverted; Based on the formation parameters to be inverted and the leakage causes in the carbon dioxide leakage analysis results, a CO2 leakage amount-anomaly parameter coupling mechanism model is constructed. This involves matching the corresponding leakage mechanism model from different configured leakage causes and different leakage mechanism models. The formation parameters to be inverted and the leakage mechanism models are then fused to obtain the CO2 leakage amount-anomaly parameter coupling mechanism model. Different leakage mechanism models are pre-set according to different leakage causes; these may include: a thermal effect mechanism model: CO2 expands and absorbs heat / dissolves and releases heat as it enters the caprock from the storage layer. The amount of heat released / absorbed is positively correlated with the leakage amount, and the heat change causes formation temperature... The anomaly is quantified by the CO2 phase change heat flow formula (heat flow = leakage × CO2 phase change enthalpy per unit mass); the seepage effect mechanism model: the migration of CO2 in the caprock microfractures / faults / casing damage sites follows Darcy's law of seepage (leakage = permeability × cross-sectional area × pressure gradient / fluid viscosity), the seepage process causes changes in formation pore pressure, which in turn leads to anomalies in strain; the stress-strain effect mechanism model: the formation strain caused by changes in CO2 pore pressure follows the linear elastic constitutive equation of rock mechanics (strain change = pore pressure change × rock volume compressibility coefficient / elastic modulus), and the pore pressure change is positively correlated with the leakage; Substituting the target spatiotemporal temperature field data and the target spatiotemporal strain field data into the CO2 leakage amount-anomaly parameter coupling mechanism model, the initial CO2 leakage amount is obtained; The initial CO2 leakage amount is corrected using the formation parameters to be inverted, resulting in the carbon dioxide leakage amount; specifically, the correction formula is as follows: Q = Q 初 ×K×λ×E; where Q represents the amount of carbon dioxide leakage; Q 初 The initial CO2 leakage is represented by K, λ, and E, which are the heterogeneity correction coefficients for permeability, thermal conductivity, and elastic modulus, respectively (calculated from actual formation parameters, with coefficients ranging from 0.5 to 2.0, adjusted according to the strength of heterogeneity).

[0032] Step S240: Based on the carbon dioxide leakage analysis results, carbon dioxide leakage amount and formation parameters, dynamically simulate the migration path and spatial distribution map of carbon dioxide in the caprock.

[0033] In practice By using formation parameters, the pre-constructed carbon dioxide transport simulation model is modified to obtain the target carbon dioxide transport simulation model. The carbon dioxide transport simulation model is a numerical model that is suitable for describing the transport of carbon dioxide in the underground environment and is constructed using the principle of multiphase fluid dynamics. The pre-constructed carbon dioxide transport simulation model is subjected to parameter assignment and structural embedding processing using formation parameters to obtain the target carbon dioxide transport simulation model that conforms to the actual conditions of the target area. Based on the carbon dioxide leakage analysis results, the simulation boundary conditions of the target carbon dioxide transport simulation model are determined; the source location of the leakage point is used as the input boundary, the geological and physical properties of the surrounding environment are used as the side boundaries, and the far field is assumed to be a flow-free boundary. Based on the amount of carbon dioxide leakage, determine the initial state of the target carbon dioxide transport simulation model at the start of the simulation; specifically, based on the amount of carbon dioxide leakage, determine the leakage rate, initial pressure, and temperature; use the leakage rate, initial pressure, and temperature as the initial state at the start of the simulation. Dynamic simulation calculations are performed on a target carbon dioxide transport simulation model with defined boundary conditions and initial states to obtain the transport path and spatial distribution map of carbon dioxide in the caprock. By solving the corresponding mass conservation equation, momentum conservation equation, and energy conservation equation using the target carbon dioxide transport simulation model, the spatial distribution of carbon dioxide at different time points is obtained. The simulated spatial distribution of carbon dioxide at different time points is then transformed into visualization graphics to generate a carbon dioxide transport path map and spatial distribution map in the caprock, which are used to intuitively show the diffusion trend of carbon dioxide over time and its distribution characteristics in underground space.

[0034] Step S250: Based on the carbon dioxide leakage analysis results, migration path and spatial distribution map, generate early warning information.

[0035] In practice, multi-dimensional feature extraction and spatiotemporal alignment are performed on the carbon dioxide leakage analysis results, migration paths, and spatial distribution maps to obtain a comprehensive risk feature matrix. Specifically, feature extraction is performed on the carbon dioxide leakage analysis results to obtain leakage features. These features may include: leakage cause risk coefficient, leakage channel expansion rate, and integrity coefficient of the 50m grid around the leakage point. The leakage cause risk coefficient is determined based on the cause of the leakage: fault activation is 1.2, caprock microfractures are 0.8, and casing damage is 0.6. Based on the migration path, migration dynamic features are extracted. These features are obtained by extracting the arrival time of the migration front at the grid, diffusion velocity gradient, flux density, and the angle between the migration direction and the fault / caprock boundary at each time point. Based on the spatial distribution map, spatial distribution features are extracted. These features may include: CO2 concentration values ​​at each grid, concentration gradient (along the migration direction), formation permeability coefficient, and straight-line distance to sensitive targets. The above features are extracted at each time stamp with a time step of 1 hour to form a time-series feature sequence. The comprehensive risk feature matrix is ​​input into a pre-trained risk assessment model for dynamic risk level assessment, resulting in a risk level sequence that changes over time. The risk assessment model employs a shallow CNN+LightGBM fusion model, which calculates the risk value for each time step and each grid in the comprehensive risk feature matrix based on a pre-defined feature-risk correlation. For each time step, the risk values ​​of all grids are weighted and summed according to their respective grid weights (0.8 for grids surrounding the leak point, 0.2 for other areas) to obtain the overall regional risk value for that time step. A level sequence is generated based on pre-defined thresholds (0.3-0.5 points: Level 1 warning; 0.5-0.8 points: Level 2 warning; 0.8-1.0 points: Level 3 warning), and the risk level and core contribution features for each time step are labeled (e.g., "t=8h, Level 2 warning, core contribution: increased migration speed + increased concentration gradient"). For cases of abrupt changes in risk level over three consecutive time steps (e.g., jumping from Level 1 to Level 3), verification is performed in conjunction with the cause of the leak to eliminate misjudgments caused by data noise.

[0036] The transport path and spatial distribution map are subjected to three-dimensional visualization processing based on adaptive mesh refinement and kriging interpolation to generate a three-dimensional risk heat map that is dynamically updated over time; in the three-dimensional risk heat map, different color gradients represent the risk level of carbon dioxide concentration. The three-dimensional risk heat map is spatially overlaid with a pre-set spatial distribution database of sensitive targets containing multiple sensitive targets, and buffer analysis is performed to obtain the expected impact time, maximum exposure concentration, and comprehensive risk index for each sensitive target; among them, sensitive targets can be groundwater monitoring wells, ecological protection zone boundaries, and residential areas; The system employs a multi-criteria decision analysis based on fuzzy logic reasoning to process the leakage causes, risk level sequences, and the expected impact time, maximum exposure concentration, and comprehensive risk index of each sensitive target. Combined with a dynamically adjusted early warning rule base, it generates tiered early warning information. Specifically, a three-dimensional fuzzy reasoning model is pre-constructed to define fuzzy rules for different leakage causes, risk levels, and the impact of different sensitive targets. The leakage causes, risk level sequences, and comprehensive risk indices of sensitive targets are substituted into the three-dimensional fuzzy reasoning model to obtain the priority and core content of early warning actions. The dynamic early warning rule base is invoked, and corresponding rules are matched based on the current reasoning results to adjust the early warning content. Tiered early warning information generation: Early warning content is integrated according to the rule base output to form a standardized information package: Level 1: Blue marker, heat map thumbnail, high-frequency monitoring instructions, no sensitive target impact indication; Level 2: Yellow marker, dynamic heat map (including sensitive target overlay), targeted maintenance work order (marking sensitive target locations), emergency standby suggestions; Level 3: Red marker, real-time heat map link, area lockdown instructions, and emergency response plan for sensitive targets (such as groundwater monitoring well sampling, residential area evacuation indication).

[0037] Corresponding to the above method, embodiments of this application also provide an early warning device for carbon dioxide storage leakage, such as... Figure 3 As shown, the device includes: The acquisition unit 310 is used to acquire spatiotemporal temperature field data and spatiotemporal strain field data inside the wellbore of the carbon dioxide monitoring well, as well as the formation parameters of the caprock through which the carbon dioxide monitoring well passes and the spatial topology map of the caprock-fault in the area where the carbon dioxide monitoring well is located. Analysis unit 320 is used to perform carbon dioxide leakage analysis based on spatiotemporal temperature field data, spatiotemporal strain field data, formation parameters and caprock-fault spatial topology map, and obtain carbon dioxide leakage analysis results of the caprock. Inversion unit 330 is used to perform carbon dioxide leakage inversion based on carbon dioxide leakage analysis results, spatiotemporal temperature field data, spatiotemporal strain field data and formation parameters to obtain the amount of carbon dioxide leakage; Simulation unit 340 is used to dynamically simulate the migration path and spatial distribution map of carbon dioxide in the caprock based on the carbon dioxide leakage analysis results, carbon dioxide leakage amount and formation parameters. The generation unit 350 is used to generate early warning information based on the carbon dioxide leakage analysis results, migration path and spatial distribution map.

[0038] The functions of each functional unit of the carbon dioxide storage leakage early warning device provided in the above embodiments of this application can be implemented through the above methods and steps. Therefore, the specific working process and beneficial effects of each unit in the carbon dioxide storage leakage early warning device 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 spatiotemporal temperature field data and spatiotemporal strain field data inside the carbon dioxide monitoring well, as well as the formation parameters of the caprock through which the carbon dioxide monitoring well passes and the spatial topology map of the caprock-fault in the area where the carbon dioxide monitoring well is located; Carbon dioxide leakage analysis was conducted based on spatiotemporal temperature field data, spatiotemporal strain field data, formation parameters, and caprock-fault spatial topology map to obtain the carbon dioxide leakage analysis results of the caprock. Based on the carbon dioxide leakage analysis results, spatiotemporal temperature field data, spatiotemporal strain field data, and formation parameters, carbon dioxide leakage was inverted to obtain the amount of carbon dioxide leakage. Based on the carbon dioxide leakage analysis results, carbon dioxide leakage amount and formation parameters, dynamic simulation was used to obtain the migration path and spatial distribution map of carbon dioxide in the caprock. Early warning information is generated based on the analysis results of carbon dioxide leakage, migration paths, and spatial distribution maps.

[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 any of the carbon dioxide sequestration leakage early warning methods described in 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 any of the carbon dioxide sequestration leakage early warning methods described in 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 1One 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 early warning of carbon dioxide storage leakage, characterized in that, The method includes: Acquire spatiotemporal temperature field data and spatiotemporal strain field data inside the carbon dioxide monitoring well, as well as the formation parameters of the caprock through which the carbon dioxide monitoring well passes and the caprock-fault spatial topology map of the area where the carbon dioxide monitoring well is located; Carbon dioxide leakage analysis was conducted based on spatiotemporal temperature field data, spatiotemporal strain field data, formation parameters, and caprock-fault spatial topology map to obtain the carbon dioxide leakage analysis results of the caprock. Based on the carbon dioxide leakage analysis results, spatiotemporal temperature field data, spatiotemporal strain field data, and formation parameters, carbon dioxide leakage was inverted to obtain the amount of carbon dioxide leakage. Based on the carbon dioxide leakage analysis results, carbon dioxide leakage amount and formation parameters, dynamic simulation was used to obtain the migration path and spatial distribution map of carbon dioxide in the caprock. Early warning information is generated based on the analysis results of carbon dioxide leakage, migration paths, and spatial distribution maps.

2. The method as described in claim 1, characterized in that, Carbon dioxide leakage analysis was performed based on the spatiotemporal temperature field data, the spatiotemporal strain field data, the formation parameters, and the caprock-fault spatial topology map to obtain the carbon dioxide leakage analysis results of the caprock, including: The spatiotemporal temperature field data, the spatiotemporal strain field data, and the formation parameters are input into a pre-trained thermo-mechanical-lithological coupled anomaly identification model to obtain anomaly identification results for different regions in the caprock; wherein, the anomaly identification results include: coupling characteristics of different regions; The region corresponding to the coupling feature whose similarity to each leakage feature in the configured carbon dioxide leakage feature library is higher than the configured similarity threshold is taken as the target region. The formation parameters, the anomaly identification results of the target area, and the caprock-fault spatial topology map are input into a pre-trained carbon dioxide leakage cause classification and location model to obtain the leakage cause and location of carbon dioxide leakage in the caprock. Based on the anomaly identification results in different regions, the cause of carbon dioxide leakage in the capping layer, and the location of the leakage, the carbon dioxide leakage analysis results of the capping layer are obtained.

3. The method as described in claim 2, characterized in that, The anomaly identification results also include: the spatial location of different regions; The formation parameters, the anomaly identification results of the target area, and the caprock-fault spatial topology map are input into a pre-trained carbon dioxide leakage cause classification and location model to obtain the leakage causes and locations of carbon dioxide leakage in the caprock, including: Based on the spatial location of the target area, extract the target geological structure information corresponding to the target area from the cap-fault spatial topology map; Based on the spatial location of the target area, extract the target stratigraphic parameters corresponding to the target area from the stratigraphic parameters; By inputting the target stratigraphic parameters of the target area, the anomaly identification, and the target geological structure information into the carbon dioxide leakage cause classification and location model, the leakage cause and location of carbon dioxide leakage in the caprock are obtained.

4. The method as described in claim 3, characterized in that, Based on the carbon dioxide leakage analysis results, spatiotemporal temperature field data, spatiotemporal strain field data, and formation parameters, carbon dioxide leakage was inverted to obtain the amount of carbon dioxide leakage, including: The inversion range is determined based on the leak location identified in the carbon dioxide leak analysis results. Target spatiotemporal temperature data and target spatiotemporal strain data within the inversion range are extracted from the spatiotemporal temperature field data and the spatiotemporal strain field data, respectively. From the stratigraphic parameters, extract the stratigraphic parameters of the inversion range to obtain the stratigraphic parameters to be inverted; Based on the formation parameters to be inverted and the leakage causes in the carbon dioxide leakage analysis results, a coupling mechanism model of CO2 leakage amount and abnormal parameters is constructed. Substituting the target spatiotemporal temperature field data and the target spatiotemporal strain field data into the CO2 leakage amount-anomaly parameter coupling mechanism model, the initial CO2 leakage amount is obtained; The initial CO2 leakage amount is corrected using the formation parameters to be inverted, and the carbon dioxide leakage amount is obtained.

5. The method as described in claim 3, characterized in that, The carbon dioxide leakage cause classification and location model includes: The spatial feature fusion layer is used to spatially fuse the target stratigraphic parameters of the input target area, the anomaly identification combination, and the target geological structure information to obtain a fusion feature matrix. The coupled feature enhancement layer is used to extract target fusion features from the fusion feature matrix through shallow neural networks and geological mechanism constraints; The dual-task inference layer includes a leakage cause classification sublayer and a leakage point three-dimensional localization sublayer; it is used to determine the leakage cause and leakage location of carbon dioxide leakage in the capping layer based on the target fusion characteristics.

6. The method as described in claim 2, characterized in that, Based on the carbon dioxide leak analysis results, migration paths, and spatial distribution maps, early warning information is generated, including: Multi-dimensional feature extraction and spatiotemporal alignment processing are performed on the carbon dioxide leakage analysis results, the migration path, and the spatial distribution map to obtain a comprehensive risk feature matrix. Based on the aforementioned comprehensive risk characteristic matrix, hierarchical early warning information is provided.

7. The method as described in claim 6, characterized in that, Based on the aforementioned comprehensive risk feature matrix, tiered early warning information includes: The comprehensive risk feature matrix is ​​input into a pre-trained risk assessment model for dynamic risk level assessment to obtain a risk level sequence that changes over time. The migration path and spatial distribution map are subjected to 3D visualization processing based on adaptive mesh refinement and Kriging interpolation to generate a 3D risk heat map that is dynamically updated over time. The three-dimensional risk heat map is spatially overlaid and buffered with a preset database of spatial distribution of sensitive targets containing multiple sensitive targets to obtain the expected impact time, maximum exposure concentration and comprehensive risk index for each sensitive target. The cause of the leak, the risk level sequence, the expected impact time of each sensitive target, the maximum exposure concentration, and the comprehensive risk index are processed by multi-criteria decision analysis based on fuzzy logic reasoning. Combined with a dynamically adjusted early warning rule base, hierarchical early warning information is generated.

8. A pre-warning device for carbon dioxide storage leakage, characterized in that, The device includes: The acquisition unit is used to acquire spatiotemporal temperature field data and spatiotemporal strain field data inside the carbon dioxide monitoring well, as well as the formation parameters of the caprock through which the carbon dioxide monitoring well passes and the caprock-fault spatial topology map of the area where the carbon dioxide monitoring well is located. The analysis unit is used to perform carbon dioxide leakage analysis based on spatiotemporal temperature field data, spatiotemporal strain field data, formation parameters and caprock-fault spatial topology map, and obtain the carbon dioxide leakage analysis results of the caprock. The inversion unit is used to perform carbon dioxide leakage inversion based on carbon dioxide leakage analysis results, spatiotemporal temperature field data, spatiotemporal strain field data, and formation parameters to obtain the amount of carbon dioxide leakage. The simulation unit is used to dynamically simulate the migration path and spatial distribution map of carbon dioxide in the caprock based on the carbon dioxide leakage analysis results, carbon dioxide leakage amount and formation parameters. The generation unit is used to generate early warning information based on carbon dioxide leakage analysis results, migration paths, and spatial distribution maps.

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