Search

CN-122020433-A - scCO2Geological storage multi-mode security assessment system and method

CN122020433ACN 122020433 ACN122020433 ACN 122020433ACN-122020433-A

Abstract

The invention discloses a sco 2 geological storage multi-mode safety evaluation system and a method, which relate to the technical field of carbon dioxide geological storage safety monitoring and risk evaluation, and the sco 2 geological storage multi-mode safety evaluation system mainly comprises a numerical simulation construction module, a multi-source monitoring data acquisition and analysis module, a simulation space feature coding module, a monitoring data time feature coding module, a cross-mode attention feature fusion module, a physical constraint and mechanical degradation modeling module and a safety state classification and early warning module. By implementing the system and the method for evaluating the multi-mode security of the geological storage of the scCO 2 , the geological storage security state of the scCO 2 can be accurately identified and early warned in advance.

Inventors

  • ZHANG GUOHUA
  • LIN HUAYING
  • XIONG FENG
  • Hua Dongjie
  • Tu Fubin

Assignees

  • 中国地质大学（武汉）

Dates

Publication Date
20260512
Application Date
20260414

Claims (8)

  1. 1. A SCCO 2 geological storage multi-mode safety evaluation system is characterized by comprising a numerical simulation construction module, a multi-source monitoring data acquisition and analysis module, a simulation space feature coding module, a monitoring data time feature coding module, a cross-modal attention feature fusion module, a physical constraint and mechanical degradation modeling module and a safety state classification and early warning module, wherein the numerical simulation construction module is used for simulating a fluid flow-geomechanical-crack evolution coupling process by using a numerical simulation method to construct a multi-channel simulation image and simulation detection data, the multi-source monitoring data acquisition and analysis module is used for acquiring on-site monitoring data and organizing the on-site monitoring data into a multi-element time sequence to capture dynamic response signals in the CO 2 sealing process, the simulation space feature coding module is used for extracting space features of the multi-channel simulation image to obtain a potential space feature representation, the monitoring data time feature coding module is used for processing the multi-element time sequence, capturing time feature representation by using self-attention mechanism long-time dependency, mutation abnormality and potential leakage precursor information, the cross-modal attention feature fusion module is used for capturing time feature representation by using the potential space feature representation and the time feature representation, the cross-stress corrosion state analysis module is used for capturing dynamic response signals in the special stress state and the dynamic constraint and the safety constraint and mechanical degradation modeling module, the multi-modal attention feature fusion module is used for obtaining a comprehensive analysis of the safety state constraint and mechanical degradation analysis module is used for acquiring the safety constraint and dynamic response coefficient in the integrated state in the safety analysis, the CO 2 mass conservation constraint and total loss function guide model prediction accords with basic physical laws, and the safety state classification and early warning module is used for carrying out safety state recognition and risk early warning on a CO 2 storage site according to the combined characteristic representation of the fused multi-mode information.
  2. 2. The scCO 2 geological storage multimodal safety assessment system of claim 1, wherein the simulated detection data is used to correct a multichannel simulated image and complement a multivariate time sequence.
  3. 3. The scCO 2 geological sequestration multimodal safety assessment system of claim 1, wherein the in situ monitoring data includes monitoring well CO 2 concentration, surface station CO 2 concentration, microseismic waveform derived attributes.
  4. 4. The scCO 2 geological storage multi-modal security assessment system as claimed in claim 1, wherein the analog spatial signature encoding module includes a multi-channel convolutional neural network.
  5. 5. The scCO 2 geological storage multimodal security assessment system of claim 1, wherein the monitoring data temporal feature coding module includes a plurality of fransformer encoder layers and a self-attention mechanism.
  6. 6. The scCO 2 geological storage multi-modal security assessment system according to claim 1, wherein the cross-modal attention feature fusion module includes a multi-modal feature input unit, a modal projection and alignment unit, a cross-modal attention calculation unit, and a fusion feature output unit, the multi-modal feature input unit is configured to receive a temporal feature representation of a plurality of modal data, the modal projection and alignment unit is configured to project the temporal feature representation to a unified feature space using a linear mapping, the cross-modal attention calculation unit is configured to obtain a comprehensive feature representation of the fusion multi-modal information by performing cross-modal attention weight calculation using a target modality as a query and the rest of auxiliary modalities as keys and values, and the fusion feature output unit is configured to output the comprehensive feature representation of the fusion multi-modal information.
  7. 7. The scCO 2 geological sequestration multimodal safety assessment system according to claim 1, wherein the total loss function is specifically: Wherein, the As a total loss function; Is a cross entropy loss function; Is a loss of pressure-stress consistency; Is the permeability evolution loss; is a conservation of mass constraint; Loss terms of the mechanical degradation model induced for CO 2 erosion; 、 、 、 Weights of loss clauses of pressure-stress consistency loss, permeability evolution loss, mass conservation constraint, and CO 2 corrosion-induced mechanical degradation model; is the fluid pressure; Is the maximum principal stress; Is critical pore pressure; Is tensile strength; Is the L2 norm; means ensuring that the permeability gradient is a non-negative function; represents the partial conductance of the concentration of CO 2 with respect to time; Represents the net flow rate of CO 2 mass or volume fraction per unit volume caused by fluid migration, N is the number of samples, C is the number of safety classes; is in a real safe state; predicting the probability that the sample i belongs to the category j for the model; to take into account the tensile strength at time t of chemical attack; tensile strength at time t-1 to account for chemical attack; is the chemical corrosion coefficient; the concentration of CO 2 at time t.
  8. 8. A method for evaluating the multi-modal security of a scCO 2 geological sequestration, characterized in that the system for evaluating the multi-modal security of a scCO 2 geological sequestration according to any one of claims 1 to 7 is used for monitoring the security of the geological sequestration of carbon dioxide and evaluating the risk.

Description

System and method for evaluating multi-mode security of SCCO 2 geological storage Technical Field The invention relates to the technical field of carbon dioxide geological sequestration safety monitoring and risk assessment, in particular to a system and a method for SCCO 2 geological sequestration multi-mode safety assessment. Background ScCO 2 geological sequestration is one of the important technological paths for achieving large-scale carbon abatement. In geological storage of carbon dioxide, the carbon source is often from a production industry such as a factory (e.g., a power plant), captures and converts the carbon dioxide into a liquid state, and then injects into the depths of a closed geological structure, such as a brine layer, depleted oil and gas fields, or non-productive coal seams, to limit its release into the atmosphere. While this method is considered the most viable sequestration method for geological storage of carbon dioxide, it involves the risk that injection of high-pressure liquid carbon dioxide into the subsurface, the reservoir-cap system will undergo a variety of coupled physical and chemical processes such as pore pressure changes, effective stress redistribution, fracture initiation and propagation, permeability evolution, and scCO 2 -coal-rock-moisture interactions, which can disrupt the mechanical balance of the reservoir and are extremely prone to induce cap damage or CO 2 leak risks. Therefore, an assessment of the security of the geological sequestration of scCO 2 is highly necessary. In the prior art, the evaluation method for the geological storage safety of scCO 2 mainly comprises a threshold criterion method based on numerical simulation, an experience or statistical analysis method based on monitoring data and a data driving method based on machine learning. The above-described methods have the following general disadvantages: (1) Single information source dependence is difficult to reflect underground physical evolution and monitoring response at the same time; (2) The safety threshold is fixedly set, and mechanical parameter degradation caused by the corrosion of the scCO 2 is not considered; (3) The pure data driving method lacks physical constraint and has the risk that the predicted result is inconsistent with the basic physical rule. How to accurately identify and early warn the geological storage safety state of the scCO 2 is a problem to be solved urgently. Disclosure of Invention The invention aims to provide a system and a method for evaluating the security of the geological storage multimode of scCO 2, which can accurately identify and early warn the security state of the geological storage of scCO 2 in advance. The invention provides a sco 2 geological storage multi-mode safety evaluation system which comprises a numerical simulation construction module, a multi-source monitoring data acquisition and analysis module, a simulation space feature coding module, a monitoring data time feature coding module, a cross-modal attention feature fusion module, a physical constraint and mechanical degradation modeling module and a safety state classification and early warning module, wherein the numerical simulation construction module is used for simulating a fluid flow-geomechanical-crack evolution coupling process by using a numerical simulation method to construct a multi-channel simulation image and simulation detection data so as to reflect strong coupling relation of underground pore pressure, stress, permeability and CO 2 distribution, the multi-source monitoring data acquisition and analysis module is used for acquiring field monitoring data and organizing the field monitoring data into a multi-element time sequence, capturing dynamic response signals in the CO 2 storage process provides real-time monitoring data support for safety evaluation, the simulation space feature coding module is used for carrying out space feature extraction on the multi-channel simulation image so as to obtain potential space feature representation, the multi-channel simulation image is used for representing a pressure zone, a high stress zone, a guide flow zone and CO 2 crack evolution coupling process, the data time feature module is used for processing the multi-channel simulation image and the multi-channel simulation image so as to reflect strong coupling relation of underground pore pressure, the stress, the multi-source monitoring data acquisition and the multi-modal attention feature coding module is used for capturing the dynamic response signals in the safety evaluation in the process, the safety evaluation process, the dynamic response signals are obtained by using the special stress and the special stress concentration time feature, the critical stress and the critical stress crack flow monitoring system, and the critical stress crack flow monitoring system is used for representing the safety state observation state, the system comprises a physical constraint and mechan