Carbon dioxide geological storage leakage monitoring and early warning system

Abstract

The application discloses a carbon dioxide geological storage leakage monitoring and early warning system, comprising a sensor arrangement module, a data acquisition and processing module, a prediction and early warning module and a linkage control module. The sensor arrangement module is arranged with pressure sensors, temperature sensors and microseismic monitoring equipment in injection wells, monitoring wells and the ground surface, and is combined with distributed optical fiber temperature measurement to realize real-time acquisition of reservoir and wellbore pressure and temperature. The data acquisition and processing module performs dynamic fitting and trend analysis on multi-source monitoring data to calculate a pressure evolution function. The prediction and early warning module predicts future pressure states through threshold determination and a prediction model, and automatically sends an early warning signal when approaching a critical value. The linkage control module is combined with an injection control unit, and can automatically or assistively adjust an injection rate according to the early warning result to reduce risks. The system can significantly improve the safety and reliability of carbon dioxide geological storage.

Description

Technical Field

[0001] This invention relates to the field of environmental monitoring technology, and more specifically, to a carbon dioxide geological storage leakage monitoring and early warning system. Background Technology

[0002] With increasing global pressure to reduce emissions, carbon dioxide capture and storage (CCS) technology is considered an important pathway to achieving carbon neutrality. The basic idea is to capture carbon dioxide from industrial emission sources and transport it via pipeline to deep geological structures (such as depleted oil and gas reservoirs, deep saline aquifers, or unused coal seams) for long-term storage. To ensure the safety and stability of the storage process, continuous monitoring of pressure changes in the storage area is essential. If the reservoir pressure rises too rapidly or exceeds the pressure limit of the caprock, it may lead to caprock failure, carbon dioxide leakage, or even induce microseismic events, posing significant risks to engineering safety and the environment.

[0003] Existing methods for monitoring reservoir pressure mainly fall into the following categories: First, direct measurement using downhole pressure gauges, where pressure sensors are installed in injection or monitoring wells to acquire point pressure data; second, distributed fiber optic sensing technology, which indirectly reflects pressure changes by deploying fiber optic cables to measure temperature and stress changes around the wellbore; third, seismic exploration and microseismic monitoring, which uses artificial or natural microseismic data to invert the reservoir stress field and pressure evolution; and fourth, numerical simulation methods, which combine geological parameters and injection data to predict the reservoir pressure field.

[0004] However, these methods have certain limitations in practical applications. While downhole pressure gauges can provide relatively accurate point pressure data, their spatial coverage is limited and cannot reflect the three-dimensional pressure distribution of the entire reservoir; fiber optic sensing is sensitive to temperature and stress changes, but requires extensive deployment and has complex signal interpretation; seismic surveys and microseismic monitoring are costly and are usually used for phased observations, making continuous monitoring difficult; numerical simulations rely on geological model parameters, and the results often deviate significantly from reality, lacking a real-time update mechanism.

[0005] The following problems are common in existing monitoring systems: First, data acquisition is isolated and lacks multi-source fusion, making it difficult to analyze data collected by different sensors in a unified manner, which limits the accuracy of monitoring results; Second, the early warning mechanism is lagging behind, and in most cases it is only discovered after pressure anomalies or even leaks have occurred, lacking foresight; Third, the control link is missing, and most existing systems can only achieve "monitoring" but cannot link the early warning results with the injection process in real time, making it difficult to take timely control measures to reduce risks.

[0006] To address the aforementioned technological gap, this invention proposes a carbon dioxide geological storage leakage monitoring and early warning system. Summary of the Invention

[0007] The purpose of this invention is to solve the technical problems mentioned in the background section and to provide a carbon dioxide geological storage leakage monitoring and early warning system, comprising:

[0008] Sensor deployment module: Used to deploy pressure sensors at the bottom of the injection well in the sealing zone to obtain the current pressure. Temperature sensors are installed inside the wellbore to obtain the current temperature. Microseismic sensors were deployed around the caprock to monitor the propagation of formation fractures, and seepage sensors were deployed around the sealing zone to obtain the seepage coefficient.

[0009] Data acquisition and processing module: used to collect and process the pressure at the current moment. Current temperature Injection rate and permeability coefficient Perform data transmission and filtering processing;

[0010] Prediction and early warning module: used for prediction based on the above , , , Constructing prediction functions :

[0011]

[0012] And calculate the rate of pressure increase. Combine it with the threshold parameter , , Comparisons are made to generate tiered early warning signals;

[0013] Visualization and Control Module: Used to display the three-dimensional pressure field distribution. and according to the Regulating the injection rate ,

[0014] in: Current pressure; : Current temperature; Injection rate; : Permeability coefficient; Predicting stress; Time step; : Pressure growth rate; , , Early warning classification threshold; : Three-dimensional pressure field distribution.

[0015] As a preferred technical solution of the present invention, the prediction function A Long Short-Term Memory (LSTM) network is used for construction and training to improve the accuracy of prediction results and the ability to model long-term dependencies of time series. The optimization objective function is:

[0016]

[0017] in, This is the loss function; the smaller the value, the smaller the error between the predicted result and the actual measurement. The first one obtained according to the prediction function One predicted pressure value; The first data collected by the sensor One measured pressure value; Given the sample size, by minimizing this loss function, the parameters of the prediction function are continuously optimized, making the predicted pressure closer to the measured value and improving the accuracy of the early warning.

[0018] As a preferred technical solution of the present invention, the data acquisition and processing module measures the pressure at the current moment. During signal preprocessing, the Kalman filter method is used to remove noise and outliers, improving data reliability. Its update formula is:

[0019]

[0020] in, For a moment Estimated pressure; This represents the measured pressure collected by the sensor at that moment; The sampling time interval; The formula, which uses a weighted update combining historical estimates and current measured values, effectively suppresses the influence of sensor noise.

[0021] As a preferred technical solution of the present invention, the prediction and early warning module utilizes the predicted pressure Pressure at the moment The difference between them is used to calculate the rate of pressure increase. Its formula is:

[0022]

[0023] in, The rate of pressure increase per unit time, when A high value indicates a rapid increase in pressure in the storage area within a short period of time, posing an abnormal risk. Therefore, this parameter is the core basis for triggering the early warning mechanism.

[0024] As a preferred technical solution of the present invention, the prediction and early warning module compares the pressure growth rate. With different threshold parameters , , The relationship between warning levels is used to classify warning levels, specifically including:

[0025] when At this time, the system is in a normal state and no action is required.

[0026] when When this happens, the system enters a state of alert, prompting operators to strengthen monitoring;

[0027] when When this happens, the system enters a warning state, requiring injection rate adjustment or other intervention.

[0028] when When the system enters an emergency state, injection should be stopped immediately to prevent leakage risks. , , These correspond to different warning thresholds.

[0029] As a preferred technical solution of the present invention, the threshold parameter , , The calculation formula is obtained based on the formation physical parameters and storage conditions:

[0030]

[0031] in, For the first Level threshold; This is an empirical coefficient; Injection rate; The fluid compressibility coefficient; Porosity; The effective reservoir volume is represented by this formula, which reflects the relationship between the threshold and formation permeability, reservoir capacity, and injection conditions, thereby enabling adaptation to different geological conditions.

[0032] As a preferred embodiment of the present invention, the microseismic sensor is used to monitor crack propagation caused by abnormal pressure, and its triggering condition is expressed by an energy criterion as follows:

[0033]

[0034] in, For micro-vibration energy; Energy factor of the earthquake source; This refers to the distance between the sensor and the earthquake source. The energy attenuation coefficient is the measured microseismic energy. When the energy exceeds the background noise level and continues to increase, it indicates that pressure may cause formation fractures to propagate. This module can be used in conjunction with the aforementioned... Cross-verification of criteria improves the reliability of early warnings. is the base of the natural logarithm, and is a mathematical constant.

[0035] As a preferred technical solution of the present invention, the seepage sensor is based on the injection rate. Seepage rate at monitoring point Calculate the breakout risk coefficient:

[0036]

[0037] in, To overcome the risk factor; The actual seepage rate at the monitoring point; For the injection rate, when When the value exceeds a preset threshold (e.g., 0.05), it indicates that the external seepage rate has approached 5% of the injection rate, which may pose a risk of gas leakage. The system will automatically trigger a risk warning.

[0038] As a preferred technical solution of the present invention, the visualization and control module is based on the pressure. With permeability tensor The formula for calculating the three-dimensional pressure field distribution is as follows:

[0039]

[0040] in, Spatial coordinates At any time Pressure distribution; This represents the initial pressure field; This is the permeability tensor. The calculation results are presented through three-dimensional visualization, allowing operators to intuitively view the pressure evolution within the storage area.

[0041] As a preferred technical solution of the present invention, the visualization and control module adjusts the pressure increase rate accordingly. For injection rate Dynamic adjustments are made, and the control formula is as follows:

[0042]

[0043] in, The adjusted injection rate; This represents the original injection rate; This is the control coefficient; The rate of increase in pressure; This is the emergency threshold. When... Approaching or exceeding When this happens, the module will automatically reduce the injection rate to prevent reservoir overpressure and leakage accidents.

[0044] Beneficial Effects: The carbon dioxide geological storage leakage monitoring and early warning system of this invention can achieve real-time monitoring of underground pressure status during the carbon dioxide storage process. By deploying pressure sensors, temperature sensors, and microseismic monitoring equipment in injection wells, monitoring wells, and on the surface, combined with distributed fiber optic temperature measurement technology, the system can continuously collect key parameters, avoiding the problems of traditional monitoring relying on single-point instruments and limited coverage. This allows pressure changes inside the reservoir and around the wellbore to be captured in a timely manner.

[0045] In the data processing stage, this invention incorporates trend analysis and prediction algorithms to dynamically fit the collected pressure, temperature, and injection rate data. This not only reflects the current pressure distribution but also predicts pressure changes over a future period. When the pressure approaches a critical value or abnormal fluctuations occur, the system issues an early warning signal, allowing managers sufficient time to respond and avoiding risks caused by monitoring delays.

[0046] Furthermore, this invention incorporates a linkage mechanism with the injection control unit. When the system detects a risk, it can automatically or assistedly adjust the injection rate, or even switch to a backup injection well, to reduce the pressure level in the storage area. This closed-loop control method integrates monitoring, early warning, and regulation, significantly improving the safety and stability of carbon dioxide storage. Attached Figure Description

[0047] Figure 1 This is a system block diagram of a carbon dioxide geological storage leakage monitoring and early warning system proposed in this invention;

[0048] Figure 2 This is a flowchart of the operation of a carbon dioxide geological storage leakage monitoring and early warning system proposed in this invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0050] See Figure 1 and Figure 2 The implementation of the present invention will be described in detail below with reference to specific embodiments.

[0051] Example 1: This invention provides a carbon dioxide geological storage leakage monitoring and early warning system. This system combines sensor deployment, real-time data acquisition, signal processing, artificial intelligence prediction, threshold determination, micro-seismic and seepage monitoring, as well as three-dimensional visualization and dynamic control to achieve intelligent monitoring and safety early warning of the pressure field during the geological storage process.

[0052] During implementation, pressure sensors are first installed at the bottom of the injection well to collect real-time pressure data within the reservoir. Temperature sensors are installed inside the wellbore to record the current temperature at different depths. Microseismic sensors are deployed around the caprock to capture minute seismic signals that may be triggered by pressure changes; seepage sensors are deployed around the perimeter of the storage area to acquire seepage-related parameters and calculate the seepage coefficient. Meanwhile, the system records the injection rate at the injection site. These data constitute the basic input information for the system's operation.

[0053] The signals acquired by the sensor undergo preprocessing by the data acquisition and processing module. This module not only handles wireless or wired data transmission but also filters the signals and removes outliers. For example, when a pressure signal... When noise occurs, a Kalman filter is used for correction, and its update formula is as follows:

[0054]

[0055] in, To estimate the pressure, For the actual pressure measured by the sensor, The sampling time interval, This is the Kalman gain. This process yields a pressure signal that more closely approximates the true value, contributing to the accuracy of subsequent prediction models.

[0056] Based on this, the prediction and early warning module forecasts future changes in reservoir pressure. The prediction model uses a Long Short-Term Memory (LSTM) network to capture short-term and long-term dependencies in the time series. The basic form of the prediction function is:

[0057]

[0058] in, To predict the pressure at future moments, The prediction step size is used. To optimize the performance of this prediction function, a mean squared error loss function is introduced as the objective:

[0059]

[0060] in, To predict stress, To measure the actual pressure, The sample size is denoted as . Through continuous iterative training, the model gradually approximates the actual evolution trend of formation pressure.

[0061] After obtaining the predicted pressure value, the system further calculates the pressure increase rate. Its definition is

[0062]

[0063] This parameter represents the rate of pressure increase per unit time. If... A persistently high level of sensitivity indicates that the injection process may lead to reservoir overpressure, which is a dangerous precursor. To facilitate risk assessment, the system has set three sets of threshold parameters. , , The specific warning logic is as follows: when When the system determines the state to be normal; when At this time, it is in a state of attention; when The state is in a warning state; when If this occurs, it is considered an emergency, and immediate measures must be taken.

[0064] The selection of the threshold parameter depends not only on experience but also on calculations based on the formation's physical properties. The improved threshold formula is as follows:

[0065]

[0066] in, For the first Level threshold, This is an empirical coefficient. For injection rate, The fluid compressibility coefficient, Porosity The effective volume of the reservoir is defined as the reservoir volume. By introducing reservoir volume and porosity, the threshold is directly related to the reservoir's pressure-bearing capacity, ensuring the scientific validity of the early warning mechanism. In addition to pressure changes themselves, the system also uses microseismic and seepage monitoring modules for auxiliary verification.

[0067] When reservoir pressure rises too rapidly, microseismic sensors can detect the propagation energy of tiny fractures, based on the following criteria:

[0068]

[0069] in, For micro-vibration energy, The source energy factor, The distance between the sensor and the earthquake source. The energy decay coefficient, The base of the natural logarithm is a mathematical constant, when the measured... When the pressure continues to increase, it indicates that the caprock may crack due to overpressure. Comparing abnormal changes in the pressure growth rate can provide cross-validation. Simultaneously, seepage sensors monitor for potential gas breakthrough risks; the criterion formula is as follows:

[0070]

[0071] in, To overcome the risk factor, The measured seepage rate at the peripheral monitoring points. This represents the injection rate.

[0072] When this coefficient exceeds a set threshold (e.g., 0.05), it indicates that the sealed gas may be breaching the sealing zone, necessitating a risk warning. At the data fusion level, the system also incorporates a geological model to invert and visualize the three-dimensional pressure field. The calculation formula is:

[0073]

[0074] in, The pressure distribution in spatial coordinates. For the initial pressure field, This is the permeability tensor. Through a 3D visualization platform, researchers and operators can intuitively observe the pressure distribution evolution in and around the injection zone, providing direct evidence for judging pressure anomalies. In the final control stage of the system, the early warning signal can directly affect the injection process control, dynamically adjusting the injection rate. Its control formula is:

[0075]

[0076] in, The adjusted injection rate, This is the control coefficient. When Approaching or exceeding the emergency threshold At this time, the system automatically reduces the injection rate to avoid further aggravating reservoir pressure.

[0077] Through the above steps, the system of this invention forms a complete closed loop: basic physical parameters are collected by sensors, data is filtered and fused, pressure prediction is achieved using prediction functions and machine learning models, risk classification is performed by determining the pressure growth rate and threshold, cross-validation is conducted using microseismic and seepage monitoring, and finally, real-time monitoring, risk warning, and safety control of geological storage pressure are achieved by combining three-dimensional visualization and dynamic control mechanisms. This system can not only promptly detect potential risks but also automatically control and prevent reservoir overpressure, thereby ensuring the long-term safety and stability of the carbon dioxide storage process.

[0078] Example 2: Taking a carbon dioxide geological sequestration project in a decommissioned oil and gas field in western China as an example, the reservoir depth is approximately 1800m, the reservoir thickness is 60m, and the porosity is... 0.18, effective reservoir volume The ultimate bearing capacity of the cap layer is approximately 35 MPa.

[0079] During implementation, a pressure sensor is first installed at the bottom of the injection well to collect reservoir pressure data in real time. Distributed temperature sensors are deployed in the wellbore to acquire temperature data in real time. Twelve microseismic monitoring points were deployed around the caprock to monitor crack propagation caused by abnormal pressure, and six seepage monitoring points were deployed around the perimeter of the storage area to observe the seepage coefficient of the sealed gas. Meanwhile, the injection pump station records the carbon dioxide injection rate. .

[0080] For example, at a certain moment, pressure is detected. MPa, temperature Injection rate Permeability coefficient The data first enters the data acquisition and processing module, where it undergoes transmission and filtering. Taking pressure signals as an example, when the sensor signal is disturbed by wellbore disturbances, the system uses Kalman filtering for correction.

[0081]

[0082] If set , Then, for the observed pressure at a certain moment The pressure was estimated a moment ago.

[0083]

[0084]

[0085] This result is used as a smoothed stress input into the subsequent prediction model.

[0086] In the prediction and early warning module, the system constructs a prediction function based on an LSTM network:

[0087]

[0088] After model training and validation, the predicted stress for the next hour was obtained. Rate:

[0089]

[0090] The system also calculates thresholds based on reservoir parameters.

[0091] Take the compression factor Porosity Injection rate Effective volume empirical coefficient Then the third threshold The calculation is as follows:

[0092]

[0093] To further verify this, the system reads signals from the microseismic sensor. The distance from a certain monitoring point to the earthquake source... Actual energy Substitute into the formula: If we assume Then the calculation yields: The values ​​are close to the measured values, indicating that microseismic activity has increased, which is consistent with the trend of increasing pressure.

[0094] Meanwhile, the measured seepage rate at the peripheral seepage monitoring points was The risk factor for the breakthrough is:

[0095]

[0096] This indicates a risk of gas breakthrough.

[0097] The visualization module calculates and displays the three-dimensional pressure field:

[0098]

[0099] The 3D model shows that the pressure at the center of the injection zone has increased significantly, and a local high-pressure anomaly has formed near the caprock.

[0100] Finally, the system automatically performs injection rate adjustment. The original injection rate is... The control formula is: If a control coefficient is set Substituting the data, we get: The system automatically reduced the injection rate to approximately 8.0. This is to mitigate the risk of excessively rapid increase in reservoir pressure.

[0101] As can be seen from this embodiment, the system of the present invention can not only realize multi-point real-time monitoring of parameters such as pressure, temperature, and seepage, but also correct data and predict future trends through filtering and prediction models. When approaching a dangerous state, it can also perform cross-verification by combining microseismic and seepage signals. Finally, it can avoid reservoir overpressure through dynamic control, thus realizing intelligent and full-cycle safety assurance for the geological sealing process.

[0102] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A carbon dioxide geological storage leakage monitoring and early warning system, characterized in that, include: Sensor deployment module: Used to deploy pressure sensors at the bottom of the injection well in the sealing zone to obtain the current pressure. Temperature sensors are installed inside the wellbore to obtain the current temperature. Microseismic sensors were deployed around the caprock to monitor the propagation of formation fractures, and seepage sensors were deployed around the sealing zone to obtain the seepage coefficient. Data acquisition and processing module: used to collect and process the pressure at the current moment. Current temperature Injection rate and permeability coefficient Data transmission and filtering are performed; the data acquisition and processing module performs data acquisition and processing on the pressure at the current moment. During signal preprocessing, the Kalman filter method is used to remove noise and outliers. Its update formula is: ； in, For a moment Estimated pressure; This represents the measured pressure collected by the sensor at that moment; The sampling time interval; For the Kalman gain, this formula is updated by weighting the historical estimates and the current measured values; Prediction and early warning module: used for prediction based on the above , , , Constructing prediction functions : And calculate the rate of pressure increase. Combine it with the threshold parameter , , Comparisons are made to generate tiered early warning signals; Visualization and Control Module: Used to display the three-dimensional pressure field distribution. and according to the Regulating the injection rate , in: Current pressure; : Current temperature; Injection rate; : Permeability coefficient; Predicting stress; Time step; : Pressure growth rate; , , Early warning classification threshold; : Three-dimensional pressure field distribution.

2. The carbon dioxide geological storage leakage monitoring and early warning system according to claim 1, characterized in that, The prediction function The network is built and trained using a Long Short-Term Memory (LSTM) network, and its optimization objective function is: ； in, This is the loss function; the smaller the value, the smaller the error between the predicted result and the actual measurement. The first one obtained according to the prediction function One predicted pressure value; The first data collected by the sensor One measured pressure value; Given the sample size, by minimizing this loss function, the parameters of the prediction function are continuously optimized, making the predicted pressure closer to the measured value and improving the accuracy of the early warning.

3. The carbon dioxide geological storage leakage monitoring and early warning system according to claim 1, characterized in that, The prediction and early warning module utilizes the predicted pressure Pressure at the moment The difference between them is used to calculate the rate of pressure increase. Its formula is: ； in, The rate of pressure increase per unit time, when A higher level indicates a rapid increase in pressure in the storage area within a short period of time, posing an abnormal risk.

4. The carbon dioxide geological storage leakage monitoring and early warning system according to claim 3, characterized in that, The prediction and early warning module compares the pressure growth rate. With different threshold parameters , , The relationship between warning levels is used to classify warning levels, specifically including: when At this time, the system is in a normal state and no action is required. when When this happens, the system enters a state of alert, prompting operators to strengthen monitoring; when When this happens, the system enters a warning state, requiring injection rate adjustment or other intervention. when When the system enters an emergency state, injection should be stopped immediately to prevent leakage risks. , , These correspond to different warning thresholds.

5. A carbon dioxide geological storage leakage monitoring and early warning system according to claim 4, characterized in that, The threshold parameter , , The calculation formula is obtained based on the formation physical parameters and storage conditions: ； in, For the first Level threshold; This is an empirical coefficient; Injection rate; The fluid compressibility coefficient; Porosity; This represents the effective volume of the reservoir.

6. The carbon dioxide geological storage leakage monitoring and early warning system according to claim 1, characterized in that, The microseismic sensor is used to monitor crack propagation caused by abnormal pressure, and its triggering condition is expressed by an energy criterion as follows: ； in, For micro-vibration energy; Energy factor of the earthquake source; This refers to the distance between the sensor and the earthquake source. The energy attenuation coefficient is the measured microseismic energy. When the energy exceeds the background noise level and continues to increase, it indicates that pressure is causing the formation fractures to expand. is the base of the natural logarithm, and is a mathematical constant.

7. The carbon dioxide geological storage leakage monitoring and early warning system according to claim 1, characterized in that, The seepage sensor is based on the injection rate. Seepage rate at monitoring point Calculate the breakout risk coefficient: ； in, To overcome the risk factor; The actual seepage rate at the monitoring point; For the injection rate, when When the value exceeds the preset threshold, it indicates a risk of gas leakage from the sealed container, and the system will automatically trigger a risk warning.

8. A carbon dioxide geological storage leakage monitoring and early warning system according to claim 1, characterized in that, The visualization and control module is based on the pressure. With permeability tensor The formula for calculating the three-dimensional pressure field distribution is as follows: ； in, Spatial coordinates At any time Pressure distribution; This represents the initial pressure field; Let be the permeability tensor.

9. A carbon dioxide geological storage leakage monitoring and early warning system according to claim 1, characterized in that, The visualization and control module is based on the pressure growth rate. For injection rate Dynamic adjustments are made, and the control formula is as follows: ； in, The adjusted injection rate; This represents the original injection rate; This is the control coefficient; The rate of increase in pressure; As the emergency threshold, when Approaching or exceeding When this happens, the module will automatically reduce the injection rate.

Images