Real-time monitoring and early warning system for stability of deepwater underwater wellhead

Abstract

The invention provides a deepwater underwater wellhead stability real-time monitoring and early warning system, and belongs to the technical field of early warning systems. The deepwater underwater wellhead stability real-time monitoring and early warning system comprises a monitoring scheme design module, a data acquisition and storage module, a state observation module, a state early warning module, a signal wireless transmission module and a platform remote monitoring module. The deepwater underwater wellhead stability real-time monitoring and early warning system comprises the following steps that the monitoring scheme design module determines layout of sensors and a host; the data acquisition and storage module acquires signals and sends the signals to the state observation module; the state observation module fuses the mechanism model and the data driving model to form a state prediction model; the state early warning module forms a three-level early warning mechanism and transmits early warning data; the signal wireless transmission module encodes, modulates and transmits the early warning data; the platform remote monitoring module receives and processes the signals to achieve stability early warning, and the problems that in the prior art, a deepwater underwater wellhead stability real-time monitoring and early warning system lacks, and safe and efficient development of deepwater oil and gas resources is difficult to guarantee are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of early warning systems, and in particular relates to a real-time monitoring and early warning system for deepwater underwater wellhead stability. Background Art

[0002] Marine oil and gas resources, as a core area of ​​global energy strategic planning, play a crucial role in ensuring energy security and promoting economic development. With its enormous development potential, my country's deepwater South China Sea has gradually become the main battlefield for future exploration in the region and a replacement area for new reserves. The development of its oil and gas resources is of great significance to the implementation of the national energy strategy. As the core equipment of drilling operations, the subsea wellhead system shoulders the important responsibility of supporting the blowout preventer stack, riser stack, surface conductor pipe, and casing. It is the first line of defense for ensuring the safety of drilling operations, and its stable operation is directly related to the safety and efficiency of the entire drilling project.

[0003] However, deepwater subsea wellhead systems operate in extremely harsh environments. Operating in deepwater environments for extended periods, they are constantly subjected to complex loads such as waves, currents, and offshore platform motion. These combined loads pose a significant challenge to the structural stability of these systems. Under these extreme conditions, subsea wellhead systems are highly susceptible to instability. Once destabilized, these systems can trigger serious safety incidents such as blowouts and oil and gas leaks, resulting in not only significant economic losses but also incalculable damage to the marine ecosystem. Currently, monitoring and early warning technologies for deepwater subsea wellhead stability are still incomplete, and a mature, systematic deepwater subsea wellhead stability monitoring and early warning system has yet to be established. This hinders effective real-time monitoring of the operational status of deepwater subsea wellhead systems and the difficulty in providing early warning of potential risks. During drilling operations, operators lack timely visibility into the mechanical changes and structural status of the subsea wellhead system. This makes it difficult to implement effective countermeasures in the event of potential system failures or stability issues, severely hindering the safe and efficient exploration and development of deepwater oil and gas resources. Summary of the Invention

[0004] In view of this, the present invention provides a real-time monitoring and early warning system for deepwater underwater wellhead stability, which solves the problem in the existing technology that there is a lack of a real-time monitoring and early warning system for deepwater underwater wellhead stability, which makes it impossible to monitor its operating status in real time and warn of risks, making it difficult to ensure the safe and efficient development of deepwater oil and gas resources.

[0005] The present invention is achieved in that: The present invention provides a real-time monitoring and early warning system for deepwater underwater wellhead stability, which includes a monitoring scheme design module, a data acquisition and storage module, a status observation module, a status early warning module, a signal wireless transmission module, and a platform remote monitoring module. The execution steps are as follows: The monitoring scheme design module establishes a mechanism model and a simulation model based on the environmental and structural characteristics of the underwater wellhead system to locate key mechanical sensitive areas and determine the layout of sensors and hosts; the data acquisition and storage module collects signals through strain and tilt sensors, transmits them to the underwater monitoring host via optical fiber for integration and storage, and sends them to the status observation module; the status observation module integrates the mechanism model and the data-driven model to establish a deep learning model to form a status prediction model; the status warning module judges stability based on dynamic response data, forms a three-level warning mechanism and transmits warning data; the signal wireless transmission module encodes, modulates and transmits the warning data; the platform remote monitoring module receives and processes signals to realize stability warning.

[0006] The monitoring scheme design module is responsible for determining the sensitive areas of key mechanical characteristics of the underwater wellhead system and establishing a monitoring scheme based on this; the data acquisition and storage module collects the bending moment and inclination data of the monitoring points in real time and integrates and stores them in the database; the state observation module integrates the underwater wellhead system mechanism model and the data-driven deep learning algorithm to establish a deepwater underwater wellhead system state observation model to predict the global dynamic response of the underwater wellhead system through local monitoring data; the state warning module introduces a variety of evaluation indicators, analyzes the current operating status of the underwater wellhead system based on real-time global response data, and issues warning information; the signal wireless transmission module converts the warning data into acoustic warning signals based on long-distance wireless communication technology, and transmits them to the platform remote monitoring module; the platform remote monitoring module derives the underwater wellhead stability risk status by analyzing the warning signals, and issues warnings according to different risk states.

[0007] The core foundation is the structural dynamic response mechanism model of the underwater wellhead system (i.e., the underwater wellhead system mechanism model). This simulation model is built using the finite element method and relies on the mechanism model for simulation analysis of computational conditions. The mechanism model provides the foundation for the simulation model's mechanical response prediction capabilities. The simulation model, through detailed calculations, assists in locating key mechanically sensitive areas, further validating and refining the application of the mechanism model.

[0008] The structural dynamics mechanism model (i.e., the structural dynamics response mechanism model of the underwater wellhead system) and the data-driven model are the two main components of the deep learning model for underwater wellhead system stability. The data-driven model provides data regularity support, while the mechanism model provides a mechanical theoretical framework. The two are integrated and embedded in the LSTM deep learning algorithm to form the deep learning stability model, providing both theoretical support and the ability to analyze real-world data.

[0009] The deep learning model for underwater wellhead system stability is a direct source for the underwater wellhead system state prediction model. This deep learning model processes sensor data, reconstructs local mechanical properties into global mechanical properties, and obtains global real-time dynamic response characteristics. This foundation forms the state prediction model, which reflects the data processing and analysis of the deep learning model and can be directly used to predict and evaluate system states.

[0010] On the basis of the above technical solution, the real-time monitoring and early warning system for deepwater underwater wellhead stability of the present invention can also be improved as follows: Among them, the underwater wellhead system environment and structural characteristic parameters in the monitoring scheme design module database identify water depth, flow profile, soil parameters, underwater wellhead system configuration information, structural coupling relationship, and lock the monitoring leading variables and auxiliary variables; support users to copy and modify database configuration information to realize fast parameter input.

[0011] The benefits of this improved approach include: the monitoring scheme design module locks monitoring variables based on database environment and structural characteristic parameters, and supports users to copy and modify database configuration information, significantly reducing the complexity of parameter input. By directly accessing key parameters such as water depth and flow profile already in the database, errors caused by manual re-entry are reduced, the efficiency of monitoring scheme design is accelerated, the scheme design is more closely aligned with actual working conditions, and the monitoring scheme is more targeted and reliable.

[0012] Furthermore, the monitoring scheme design module pre-calculates the mechanical model based on the finite element method, obtains the coordinates of the key mechanical characteristic areas and the sensor layout type, location and quantity, and provides two preset schemes: The inclination sensor uses MEMS inclinometers, which are arranged in two groups along the axial direction of the pipe string, with the spacing being the minimum unit length of the simulation model of the key mechanical sensitive area; The strain sensor uses two groups of fiber Bragg grating strain gauges, which are symmetrically arranged along the radial direction of the pipe string, with the spacing being the diameter of the pipe string at the corresponding position.

[0013] The benefits of this improved solution include: A pre-calculated mechanical model based on the finite element method determines key mechanical characteristic regions and sensor placement parameters, and two preset sensor placement schemes are provided, ensuring the scientific and rationality of sensor installation locations. The precise placement of the MEMS inclinometer and fiber Bragg grating strain gauges efficiently senses lateral bending deformation at the underwater wellhead, avoiding unpredictable sensor placement, improving the accuracy of mechanical signal acquisition, and providing high-quality raw data support for subsequent data processing and condition assessment.

[0014] Furthermore, the data acquisition and storage module collects strain data and inclination data through strain and inclination sensors, and transmits them to the wellhead monitoring host via a wired optical fiber composite cable. The monitoring host has a built-in data integration and data storage sub-module, which converts the optical fiber signal into a digital signal, integrates the data type and time to form a monitoring data stream, and the data storage sub-module stores the monitoring data stream in real time and establishes a historical monitoring database for use in subsequent state prediction model correction.

[0015] The benefits of this improved solution include: the data acquisition and storage module collects data from strain-tilt sensors, transmits it to the host computer via a fiber-optic composite cable, and converts and integrates it into a data stream, enabling real-time and stable transmission of monitoring data. The historical database established by the built-in storage unit not only provides a rich historical sample for model calibration but also ensures data traceability. The data integration process standardizes the data format, reduces data redundancy, improves data processing efficiency, and provides a reliable data foundation for real-time analysis in the state observation module.

[0016] Furthermore, the state observation module embeds the mechanism model into the LSTM deep learning algorithm, extracts the inclination and strain data of key characteristic areas, combines historical monitoring data to train the mechanism model, reconstructs local monitoring data into global dynamic predictions, and introduces real-time verification, uncertainty management and self-correction mechanisms.

[0017] The beneficial effects of this improved approach are as follows: the state observation module embeds the structural dynamic response mechanism model into the LSTM deep learning algorithm, combines historical data training with the model, and introduces real-time verification and self-correction mechanisms, achieving precise reconstruction of local monitoring data into global dynamic predictions. This approach balances the theoretical rigor of the mechanism model with the practical adaptability of the data-driven model, improving the accuracy and timeliness of global dynamic response predictions, enabling timely capture of potential structural changes in the underwater wellhead system, and providing a scientific basis for stability assessment.

[0018] Furthermore, the status warning module introduces short-term risk index and long-term instability probability as evaluation indicators. The short-term risk index is used to capture the risk of sudden instability, and the long-term instability probability quantifies the comprehensive risk of environmental load and structural degradation. The two jointly support stability judgment.

[0019] The benefits of this improved solution include: the status warning module incorporates dual evaluation indicators: a short-term risk index and a long-term instability probability. The short-term indicator quickly captures the risk of sudden instability, while the long-term indicator quantifies the combined impact of environmental and structural degradation, enabling a comprehensive assessment of the stability of the subsea wellhead system. These two indicators complement each other, avoiding the limitations of single-indicator assessments, improving the comprehensiveness and accuracy of stability assessments, and providing a targeted basis for risk prevention and control at different time scales.

[0020] Furthermore, the status warning module designs a three-level warning mechanism based on evaluation indicators, namely: green warning status represents stable operation, yellow warning status represents the need to investigate the cause of instability, and red warning status represents the need to immediately implement emergency plans. Each level of warning corresponds to specific judgment conditions.

[0021] The beneficial effects of adopting this improved solution include: a three-level early warning mechanism designed based on evaluation indicators clearly distinguishes between stable operation, investigation required, and emergency response status. Each level of warning corresponds to a specific response measure, achieving a precise classification of risk levels. This mechanism enables operators to quickly identify the urgency of risks, avoids ambiguity in early warning information, ensures timely investigation during yellow warnings, and efficient emergency response during red warnings, reducing the probability of accidents and the extent of losses.

[0022] Furthermore, the signal wireless transmission module includes a coding submodule, a modulation submodule and a transmission submodule. The coding submodule receives the monitoring data stream from the status warning module, and uses LDPC and RS cascade to perform signal encoding to generate a warning digital signal; the modulation submodule generates multi-carriers through OFDM modulation technology, and each subcarrier is orthogonal to each other, dividing the high-speed data stream into multiple parallel low-speed data streams, which are modulated onto the subcarriers of each channel for transmission; the transmission submodule is designed based on the 4×4 MIMO solution, and outputs the acoustic wave signal in parallel to the platform remote monitoring module.

[0023] The benefits of this improved solution are as follows: The signal wireless transmission module utilizes LDPC and RS concatenated coding, OFDM modulation, and a 4×4 MIMO transmission scheme, significantly improving the reliability, anti-interference capabilities, and transmission efficiency of early warning data transmission. Concatenated coding enhances data error correction capabilities, OFDM technology enables stable high-speed data transmission, and the MIMO scheme improves signal transmission stability. This ensures accurate and timely transmission of early warning signals to the platform in complex underwater environments, guaranteeing real-time remote monitoring.

[0024] Furthermore, the platform's remote monitoring module responds according to the warning status: a green logo is displayed in the green warning status and a daily stability report is generated; an audible and visual alarm is triggered in the yellow warning status, and a diagnostic report is generated and synchronized to the safety engineer; an audible and visual alarm is triggered in the red warning status, and a strong decibel buzzer alarm signal and voice broadcast are issued, emergency operations are automatically executed and information is pushed to the command center.

[0025] The platform remote monitoring module executes different early warning responses based on the real-time operating status of the underwater wellhead system decoded from the early warning signal. The early warning response content is:

[0026] If the remote monitoring module of the underwater wellhead system platform determines that it is in a green warning state, the warning response is: the platform remote monitoring host operation page displays a green icon; routine data recording is performed, and a system stability report is generated daily; If the remote monitoring module of the underwater wellhead system platform determines that the state is yellow, the warning response is as follows: the platform remote monitoring host operation page displays a yellow mark, immediately triggers the sound and light alarm, and issues a warning signal; automatically generates a diagnostic report that lists the main factors causing the risk and the location of the occurrence. The alarm information and diagnostic report are synchronously sent to the responsible safety engineer for verification; If the underwater wellhead system determines that the platform remote monitoring module is in a red warning state, the warning response is: a red mark is displayed on the platform remote monitoring host operation page, and an audible and visual alarm is immediately triggered, emitting a strong decibel buzzer alarm signal and voice broadcast; emergency operations are automatically executed, including starting the blowout preventer shutdown procedure, system emergency release and other user-preset emergency plans; and accident plans and risk status are pushed to the command center in real time.

[0027] Furthermore, it also includes hardware subsystems and software subsystems: the hardware subsystem includes the platform remote monitoring host and the underwater monitoring host; the software subsystem contains underlying code and user pages, supporting physical modeling, scheme design, model update, data analysis, report export and early warning query operations.

[0028] The hardware subsystem includes a platform remote monitoring host consisting of a display, memory and processor, and an underwater monitoring host consisting of sensors, processors, memory, fiber optic composite cables, modems, and signal transmitters; the software subsystem includes the underlying function implementation code and user pages, supporting users to establish physical models, design monitoring plans, update deep learning models, extract and analyze key data, export stability reports, and query early warning information through the display of the platform remote monitoring host.

[0029] Compared with the existing technology, the beneficial effects of the real-time monitoring and early warning system for deepwater underwater wellhead stability provided by the present invention are: From a cost-effectiveness perspective, the system adopts a single monitoring point design concept, leveraging a state prediction model to extrapolate local monitoring data to a global dynamic response, eliminating the need for redundant monitoring points along the entire length of the underwater wellhead. This precise deployment significantly reduces the investment in hardware equipment such as sensors and host computers, lowering system deployment and maintenance costs. Compared to traditional global monitoring solutions, it offers significant economic advantages and is particularly suitable for large-scale deepwater oil and gas development scenarios. In terms of monitoring accuracy, the system constructs a state prediction model through the deep integration of mechanism models and deep learning algorithms. Combined with model prediction and correction technology, it achieves real-time and accurate global assessment of underwater wellhead stability. The mechanical theoretical framework provided by the mechanism model ensures the rigor of the analysis, while the deep learning algorithm's ability to mine multi-source data improves the adaptability of the prediction. The synergistic effect of the two effectively eliminates the limitations of local monitoring data, making the reconstruction of global mechanical properties more consistent with actual working conditions and providing a reliable basis for risk assessment. In terms of response efficiency, the system achieves a deep integration of hardware monitoring and software prediction. The underwater monitoring host performs in-situ processing and early warning analysis on the collected data, and only transmits key information wirelessly to the platform's remote monitoring host. The simplified transmission link combined with the RLS equalizer to compensate for signal delay significantly reduces the overall delay of data transmission and processing, ensuring that early warning information can be quickly delivered to the decision-making terminal. Furthermore, the system's three-level early warning mechanism and the differentiated response of the platform's remote monitoring enable refined risk management. Routine recording of green alerts, timely investigation of yellow alerts, and emergency initiation of red alerts enable operators to take targeted measures based on risk levels, effectively avoiding delays in handling caused by ambiguous alerts and significantly reducing the probability of accidents and the extent of losses. The clear division between hardware and software subsystems enhances the system's stability and ease of use. Diverse software functions support flexible model updates and data analysis, meeting monitoring needs in diverse scenarios and providing full-process, highly efficient technical support for safe deepwater subsea wellhead operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of the deepwater underwater wellhead real-time monitoring and early warning system of the present invention; Figure 2 This is a schematic diagram of the layout of the deepwater underwater wellhead real-time monitoring and early warning system of the present invention; Figure 3 This is a schematic diagram of the hardware structure of the deepwater underwater wellhead real-time monitoring and early warning system of the present invention; Figure 4 It is a calculation flow chart of the deepwater underwater wellhead system state observation module of the present invention. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0032] like Figure 1The figure shows a flow chart of a real-time monitoring and early warning system for deepwater underwater wellhead stability provided by the present invention, which includes a monitoring scheme design module, a data acquisition and storage module, a status observation module, a status early warning module, a signal wireless transmission module, and a platform remote monitoring module, and includes the following steps: Step 1: The monitoring scheme design module locks the monitoring dominant variables and auxiliary variables according to the environmental and structural characteristics of the underwater wellhead system, establishes a deepwater underwater wellhead system mechanism model based on the structural dynamics method, and establishes an underwater wellhead system simulation model based on the finite element method to locate the key mechanical sensitive areas of the underwater wellhead system, and finally determines the layout of sensors, underwater monitoring hosts, and platform remote monitoring hosts to complete the design of the deepwater underwater wellhead system monitoring scheme.

[0033] Step 2: The data acquisition and storage module includes data acquisition, data integration and data storage functional submodules. According to the monitoring plan, the strain-tilt sensor senses the signal to complete data acquisition, and transmits the data to the underwater monitoring host via wired optical fiber transmission. The underwater monitoring host then integrates the monitoring data into a data stream form and stores it in the database through the built-in storage unit. Finally, it sends it to the status observation module in real time for data processing. Step 3: The state observation module integrates the structural dynamics mechanism model and the data-driven model to establish a deep learning model for the stability of the underwater wellhead system, processes the data transmitted by the underwater monitoring sensor in real time, reconstructs the response characteristics of the deepwater underwater wellhead system from local mechanical characteristics to global mechanical characteristics in real time, obtains the global real-time dynamic response characteristics of the deepwater underwater wellhead system, and forms a deepwater underwater wellhead system state prediction model.

[0034] Step 4: The status warning module introduces multiple stability evaluation indicators to judge the short-term and long-term stability of the underwater wellhead system based on the global dynamic response data of the underwater wellhead system, and forms a three-level risk warning response mechanism for underwater wellhead stability according to the evaluation indicator threshold setting; further forms a warning data stream and transmits it to the platform remote monitoring host for warning.

[0035] Step 5: The signal wireless transmission module includes a coding unit, a modulation unit and a transmitting unit, which are integrated inside the wellhead monitoring host; the coding unit is responsible for encoding and converting the warning data transmitted by the warning module to obtain a digital signal, thereby improving the reliability of data transmission; the modulation unit is responsible for converting the encoded digital signal into an acoustic signal, and improving the signal anti-interference ability and signal transmission efficiency based on multi-carrier modulation technology; the transmitting unit is based on MIMO multi-array transmission technology, and further improves the stability of the acoustic signal transmission process through signal gain.

[0036] Step 6: The platform remote monitoring module is integrated into the platform remote monitoring host and includes three sub-modules: signal reception, signal processing, and early warning response. Among them: the signal reception sub-module compensates for the signal transmission delay through the RLS equalizer, and the built-in LDPC decoder decodes the analog signal, and the decoded digital signal is transmitted to the signal processing sub-module; the signal processing sub-module confirms the risk level represented by the early warning data and passes it to the early warning response sub-module, ultimately realizing the stability early warning of the deepwater underwater wellhead system.

[0037] The design process of step 1 is as follows: Identify the structural characteristics of each component of the system, establish a structural dynamic response mechanism model of the underwater wellhead system, and conduct constraint analysis, load analysis, structural coupling analysis, and mechanical property analysis on the entire underwater wellhead system to enable accurate prediction of the dynamic response of deepwater underwater wellhead systems; Read the environmental data and structural data from the deepwater underwater wellhead monitoring system database, and generate a calculation condition table based on the input water depth, flow profile, wave parameters, soil parameters, underwater wellhead system component type, configuration information, and structural coupling relationship; The calculation condition table is input into the structural dynamic response mechanism model of the deepwater underwater wellhead system. The time and space history information of the underwater wellhead system displacement, bending moment, and stress are extracted based on the monitoring variables to identify the key mechanical sensitive areas of the underwater wellhead system. The sensor installation position is calculated according to the key mechanically sensitive areas of the underwater wellhead system, and a group of strain-tilt sensors are jointly deployed based on the structural characteristics to realize the perception of the lateral bending deformation of the underwater wellhead system. At the same time, the layout of the underwater monitoring host and the platform remote monitoring host is determined according to the layout of the strain-tilt sensors, and finally the design of the deepwater underwater wellhead system monitoring plan is completed.

[0038] The calculation process of step 3 includes: The underwater wellhead system structural dynamic response mechanism model is embedded in the LSTM deep learning algorithm, making it suitable for deepwater underwater wellhead stability prediction. At the same time, it is further integrated with the data-driven model to realize the coupled analysis of underwater wellhead structural deformation and multi-source monitoring data, forming a deep learning model for deepwater underwater wellhead system stability. The inclination and strain data of the key characteristic areas of the system are extracted and input into the deepwater underwater wellhead system stability deep learning model for training together with historical monitoring data. The local monitoring data is reconstructed into a global dynamic prediction, and the global dynamic response of the deepwater underwater wellhead system is finally obtained. The key data of the global dynamic response of the deepwater underwater wellhead system are extracted, the dynamic data of the displacement and bending moment of the entire length of the underwater wellhead are output, and the inclination angle and strain value of the monitoring point in the next time step are synchronously predicted. Real-time verification, uncertainty management and self-correction mechanisms are introduced to ensure that the results of the global reconstruction of the model can truly reflect the stability of the underwater wellhead system.

[0039] Among them, see Figure 2 , the layout diagram of the deepwater underwater wellhead real-time monitoring and early warning system of the present invention is introduced.

[0040] The deepwater underwater wellhead stability monitoring and early warning system includes a hardware subsystem and a software subsystem, among which: the monitoring scheme design module and the status observation module are integrated in the software subsystem, supporting users to input system configuration, design monitoring schemes, update deep learning models, extract and analyze key data, export stability reports, and query early warning information through the display of the platform remote monitoring host; the data acquisition and storage module, the status early warning module, the signal wireless transmission module, and the platform remote monitoring module are integrated in the hardware subsystem, including a platform remote monitoring host consisting of a display, memory and processor, and a wellhead monitoring service host consisting of sensors, processors, memory, optical fiber composite cables, modems, and signal transmitters.

[0041] In the above technical solution, the underwater wellhead system environment and structural characteristic parameters in the monitoring solution design module database are used to identify water depth, flow profile, soil parameters, underwater wellhead system configuration information, and structural coupling relationships, and lock the monitoring leading variables and auxiliary variables; users are supported to copy and modify database configuration information to realize rapid parameter input.

[0042] The monitoring scheme design module pre-calculates the mechanical model based on the finite element method to obtain the coordinates of the key mechanical characteristic areas and the sensor layout type, location, and quantity. The monitoring scheme design module establishes the underwater wellhead system mechanism model based on structural dynamics. The modeling method is as follows: ; Where, is the lateral deformation of the underwater wellhead string, It is the damping of underwater wellhead string structure; For time; is the vertical coordinate of the underwater wellhead string; is the elastic modulus of the underwater wellhead string material; is the environmental load acting on the pipe element per unit length; is the cross-sectional area of ​​the underwater wellhead string; is the mass per unit length of the underwater wellhead string; is the axial tension of the underwater wellhead string; is the moment of inertia of the underwater wellhead string section;

[0043] The finite element method is used to pre-calculate the structural dynamics mechanism model of the underwater wellhead system. The coordinates of the key mechanical characteristic areas are accurately calculated according to the configuration of the underwater wellhead system, and the sensor layout type, location, and quantity are further determined. Two preset schemes are provided: The inclination sensor uses MEMS inclinometers, which are arranged in two groups along the axial direction of the pipe string, with the spacing being the minimum unit length of the simulation model of the key mechanical sensitive area; The strain sensor uses two groups of fiber Bragg grating strain gauges, which are symmetrically arranged along the radial direction of the pipe string, with the spacing being the diameter of the pipe string at the corresponding position.

[0044] See also Figure 3 This section describes the hardware structure of the deepwater underwater wellhead real-time monitoring and early warning system. The wellhead monitoring hardware group includes a wellhead monitoring host and monitoring sensors. The platform remote hardware group is connected to the wellhead monitoring hardware group via wireless communication.

[0045] The hardware subsystem consists of a platform remote monitoring host and a wellhead monitoring host, covering data acquisition and storage modules, status warning modules, signal transmission modules, and remote monitoring modules.

[0046] The underwater wellhead system data acquisition and storage module collects data based on the strain and inclination sensors on the monitoring system. The strain sensor collects strain data at the underwater wellhead monitoring point, and the inclination sensor collects inclination data at the monitoring point, and transmits the data to the wellhead monitoring host via a wired optical fiber composite cable. The wellhead monitoring host has a built-in data integration and data storage submodule, which converts the optical fiber signal sent by the sensor into a digital signal, and integrates the data type and time on this basis to form a monitoring data stream. The data storage submodule stores the monitoring data stream in real time and establishes a historical monitoring database for use in subsequent prediction model correction.

[0047] The data acquisition and storage module collects strain data and inclination data through strain and inclination sensors, and transmits them to the wellhead monitoring host via a wired optical fiber composite cable. The monitoring host has a built-in data integration and data storage sub-module, which converts the optical fiber signal into a digital signal, integrates the data type and time to form a monitoring data stream. The data storage sub-module stores the monitoring data stream in real time and establishes a historical monitoring database for use in subsequent status prediction model correction.

[0048] The status warning module introduces short-term risk index and long-term instability probability as evaluation indicators. The short-term risk index is used to capture the risk of sudden instability, and the long-term instability probability quantifies the comprehensive risk of environmental loads and structural degradation. The calculation method is: The short-term risk index is used to capture sudden instability risks in real time and provide real-time warnings for potential short-term risks. It is calculated using the following model: ; Where: It is a short-term risk index, which is obtained by extracting the strain, stress and bending moment of the global response of the structure and the output range is [0, 1]; 、 、 are weight coefficients respectively; is the structural strain, dimensionless; is the critical strain value, which is the failure threshold of the material or structure and is dimensionless; Predict the equivalent stress at time t; The maximum equivalent stress allowed for the structure; Predict the bending moment at time t; is the maximum bending moment limit of the material; The long-term instability probability is the combined risk of quantified environmental loads and structural degradation. It is related to environmental loads and structural strength and is calculated as follows: ; ; ; Where, is the probability of long-term instability; is the standard normal distribution function; is the environmental load margin factor; is the structural strength margin coefficient; is the uncertainty factor; D is the cumulative damage; The wave height that the system can be designed to withstand; is the monitored significant wave height; is the corrosion loss factor; the two together support the stability judgment.

[0049] The status warning module has a three-level warning mechanism based on evaluation indicators: green warning status indicates stable operation, yellow warning status indicates that the cause of instability needs to be investigated, and red warning status indicates that emergency plans need to be implemented immediately. Each level of warning corresponds to specific judgment conditions; Green warning means that the deepwater underwater wellhead is operating stably. The judgment conditions are: ; A yellow warning indicates that the underwater wellhead system is in an unstable state and requires further warning to investigate the cause. The judgment conditions are: ; A red alert indicates that the underwater wellhead system is in a serious emergency state and requires immediate implementation of the risk emergency plan. The criteria for this are: .

[0050] The signal wireless transmission module includes a coding submodule, a modulation submodule, and a transmission submodule. The coding submodule receives the monitoring data stream from the status warning module and uses LDPC and RS cascade to encode the signal to generate a warning digital signal; the modulation submodule generates multi-carriers through OFDM modulation technology, and each subcarrier is orthogonal to each other, dividing the high-speed data stream into multiple parallel low-speed data streams, which are modulated onto the subcarriers of each channel for transmission; the transmission submodule is designed based on a 4×4 MIMO solution and outputs the acoustic wave signal in parallel to the platform remote monitoring module.

[0051] The platform remote monitoring module responds according to the warning status: If the underwater wellhead system remote monitoring module determines that the warning status is green, the warning response is: the platform remote monitoring host operation page displays a green icon; routine data recording is performed, and a system stability report is generated daily; If the underwater wellhead system remote monitoring module determines that the state is yellow, the warning response is as follows: the platform remote monitoring host operation page displays a yellow mark, immediately triggers the sound and light alarm, and issues a warning signal; automatically generates a diagnostic report that lists the main factors causing the risk and the location of the occurrence. The alarm information and diagnostic report are synchronously sent to the responsible safety engineer for verification; If the remote monitoring module of the underwater wellhead system determines that it is in a red warning state, the warning response is: a red mark is displayed on the platform remote monitoring host operation page, and an audible and visual alarm is immediately triggered, emitting a strong decibel buzzer alarm signal and voice broadcast; emergency operations are automatically executed, including starting the blowout preventer shutdown procedure, system emergency release and other user-preset emergency plans; and accident plans and risk status are pushed to the command center in real time.

[0052] See also Figure 4 , the calculation flow chart of the deepwater underwater wellhead system status observation module of the invention is introduced.

[0053] The state observation module integrates the structural dynamics mechanism model and the data-driven model to establish a deep learning model for underwater wellhead stability. It performs real-time data processing based on the data transmitted by underwater monitoring sensors, reconstructs the response characteristics of the deepwater underwater wellhead system from local mechanical characteristics to global mechanical characteristics in real time, obtains the global real-time dynamic response characteristics of the deepwater underwater wellhead system, and forms a deepwater underwater wellhead system state prediction model. The calculation process includes: The underwater wellhead system structural dynamic response mechanism model is embedded in the LSTM deep learning algorithm to make it suitable for deepwater underwater wellhead stability prediction. At the same time, the data-driven model is further integrated to realize the coupled analysis of wellhead structure deformation and multi-source monitoring data, forming a deep learning model for deepwater underwater wellhead system stability. The real-time curvature of the discrete calculation underwater wellhead system is calculated. As input features to modify the gate enhancement formula, the calculation method is: ; ; ; ; Where, For the Gate of Forgetfulness; is the input gate; is the cell state to be updated; is the cell state at time t; is the Sigmoid activation function; W is the set weight matrix, and the subscripts correspond to different gates; is the input variable at time t; is the hidden variable at time t-1; is the real-time curvature, which is calculated by the finite element discretization of the underwater wellhead system. The calculation formula is: ; b is a bias term, which is used to adjust the threshold of gate activation, and the subscripts correspond to different gates; is the deformation influence factor, which is obtained by training historical data and is used to quantify the direct correction effect of curvature on the cell state; The deepwater subsea wellhead system stability deep learning model uses mean square error (MSE) and coefficient of determination (R²) as metrics to evaluate the performance of the trained LSTM model. The mean absolute percentage error (MAPE) and R² are used to analyze the model prediction results. The calculation methods of MAPE, MSE, and R² are as follows: ; ; ; Where, is the true value, Take the average of the true values, is the predicted value, is the number of predicted points.

[0054] The inclination and strain data of the key characteristic areas of the system are extracted and input into the deepwater underwater wellhead system stability deep learning model for training together with historical monitoring data. The local monitoring data is reconstructed into a global dynamic prediction, and the global dynamic response of the deepwater underwater wellhead system is finally obtained. The key data of the global dynamic response of the deepwater underwater wellhead system are extracted, the dynamic data of the displacement and bending moment of the entire length of the underwater wellhead are output, and the inclination angle and strain value of the monitoring point in the next time step are synchronously predicted. Real-time verification, uncertainty management and self-correction mechanisms are introduced to ensure that the results of the global reconstruction of the model can truly reflect the stability of the underwater wellhead system.

[0055] Specifically, the principle of the present invention is as follows: The real-time monitoring and early warning system for deepwater underwater wellhead stability is centered around the collaborative work of multiple modules, building a complete technical system from data acquisition to early warning response. Its technical principle is based on the deep integration of structural mechanics theory and intelligent data analysis, achieving precise monitoring and early warning of underwater wellhead system stability. During the monitoring plan design phase, the system relies on the environmental characteristics and structural parameters of the underwater wellhead system to establish a mechanism model through structural dynamics methods, and simultaneously uses the finite element method to construct a simulation model. The mechanism model has the ability to analyze system constraints, loads, structural coupling, and mechanical characteristics, and can accurately predict dynamic responses. The simulation model, by inputting operating parameters such as water depth and flow profile, locates key mechanically sensitive areas, providing a scientific basis for sensor and host layout, and ensuring the accuracy and necessity of monitoring point settings. The data acquisition and storage module serves as the system's "sensory nerves," capturing structural mechanical signals through strain-tilt sensors deployed in key sensitive areas and transmitting the data via a wired fiber-optic composite cable to the underwater monitoring host. The host formats and stores the data, forming a standardized data stream. This not only ensures the integrity of the original data but also provides a high-quality data foundation for subsequent analysis. The established historical database also provides sample support for model correction. The state observation module is the system's "intelligent brain." It embeds the structural dynamics mechanism model into the LSTM deep learning algorithm and integrates the data-driven model to build a stability deep learning model. This model is trained on local monitoring data and historical data to reconstruct mechanical properties from local to global. Combined with real-time verification and self-correction mechanisms, it generates a predictive model that truly reflects the system state and accurately outputs global dynamic response characteristics. The status warning module uses the dynamic response data output by the prediction model, introducing short-term risk index and long-term instability probability as evaluation indicators. By setting thresholds, it forms a three-level warning mechanism, achieving a comprehensive assessment of the system's short-term sudden risk and long-term degradation risk. The signal wireless transmission module uses LDPC and RS concatenated coding, OFDM modulation, and MIMO transmission technology to ensure efficient and stable transmission of warning data in complex underwater environments. The platform remote monitoring module completes the final implementation and disposal of warning information through signal processing and differentiated response.

[0056] The following provides Example 1 of a specific application scenario of the present invention: This embodiment is applied to a deepwater oil and gas field development project in the South China Sea. The water depth in this area is about 1,500 meters. It is affected by monsoons and ocean currents all year round. The ocean current speed can reach 1.5 meters per second, the wave period is 5-12 seconds, and the seabed soil is mainly clay silt with certain rheological properties. The project uses an underwater wellhead system for oil and gas production. The system consists of surface conduits, casing heads, tubing heads, blowout preventer groups and other components. The total height is about 18 meters and the maximum outer diameter is 1.2 meters. It undertakes the important functions of oil and gas transportation, pressure control and equipment support. Due to the complex geological conditions in this area, the underwater wellhead system is subjected to multiple effects such as wave loads, ocean current forces, soil reaction forces and internal pressures for a long time. There is a risk of structural deformation, fatigue damage and even instability. A precise and efficient stability monitoring and early warning system is urgently needed to ensure its safe operation.

[0057] The real-time monitoring and early warning system for deepwater underwater wellhead stability in this embodiment comprises six core modules: a monitoring scheme design module, a data acquisition and storage module, a status observation module, a status early warning module, a signal wireless transmission module, and a platform remote monitoring module. The hardware and software of each module work together to form a complete monitoring and early warning system. The system's hardware subsystem includes a monitoring host and sensor array deployed underwater, as well as a remote monitoring host located on the offshore platform. The software subsystem, integrated within the host, supports the entire process, including monitoring scheme design, data processing, model training, and early warning response.

[0058] The monitoring scheme design module operates based on the environmental characteristics of the project area and the structural parameters of the underwater wellhead system. First, the module retrieves environmental data for the deepwater area from a database, including data on a water depth of 1,500 meters, flow profile distribution, wave parameters, and soil mechanical properties. It also obtains the structural characteristics, configuration information, and coupling relationships of each component of the underwater wellhead system. Using structural dynamics methods, a system mechanism model is established, and constraint, load, structural coupling, and mechanical property analyses are performed on the underwater wellhead system, enabling prediction of the system's dynamic response.

[0059] The module then generates a calculation table based on the input environmental and structural parameters. After inputting this into the mechanism model, it extracts the temporal and spatial history of the system's displacement, bending moment, and stress, pinpointing the key mechanically sensitive areas at the connection between the casing head and the surface conduit and at the bottom of the BOP stack. Based on the characteristics of these sensitive areas, the module calculates and determines the sensor installation locations and deploys a strain-tilt sensor system: two MEMS inclinometers are arranged axially along the tubing, spaced by the minimum unit length of the simulation model for the key mechanically sensitive areas; and two fiber Bragg grating strain gauges are symmetrically arranged radially along the tubing, spaced by the tubing diameter at the corresponding locations. The underwater monitoring host is also installed on a fixed base 5 meters from the wellhead, and the platform remote monitoring host is deployed at the control center of the offshore platform, completing the overall monitoring solution design. The module allows users to access existing configuration information from the database through the platform remote monitoring host's interactive interface, copying or modifying it, allowing for rapid adaptation of the monitoring solution to the specific project.

[0060] The data acquisition and storage module operates according to the layout designed for the monitoring plan. Strain sensors and inclination sensors installed in key mechanically sensitive areas of the underwater wellhead system detect changes in pipe string strain and inclination in real time, capturing lateral bending deformation signals. The raw signals collected by the sensors are transmitted to the underwater monitoring host via a dedicated wired fiber-optic composite cable. This cable is resistant to high pressure and corrosion, ensuring stable data transmission at water depths of 1,500 meters.

[0061] After the underwater monitoring host receives sensor signals, the built-in data integration submodule unifies the format of different sensor data types, adds timestamp information, and integrates them into a standardized monitoring data stream. The data storage submodule uses a built-in storage unit to store the data stream in real time in a local database. It also establishes a historical monitoring database to record all monitoring data during system operation, providing data support for subsequent model calibration and trend analysis. This integrated data stream is sent to the status observation module in real time to ensure timely data processing.

[0062] The state observation module is the core of the system's precise monitoring capabilities. It integrates a structural dynamics model with a data-driven model to construct a deep learning model. The module embeds the underwater wellhead system model into an LSTM deep learning algorithm and integrates the data-driven model to achieve coupled analysis of structural deformation and multi-source monitoring data. By extracting inclination and strain data from key characteristic areas and combining it with data samples from a historical monitoring database, the deep learning model is trained to enable the reconstruction of local monitoring data into global mechanical properties.

[0063] During system operation, the module receives monitoring data streams from the data acquisition and storage module in real time and feeds them into the trained deep learning model. This model calculates and extends local mechanical properties to a global scale, outputting dynamic displacement and bending moment data along the entire length of the underwater wellhead system and simultaneously predicting the inclination angle and strain values ​​at the monitoring point for the next time step. The module incorporates a real-time verification mechanism, comparing the predicted results with the theoretical values ​​calculated by the mechanism model. Model parameters are adjusted through uncertainty management and self-correction mechanisms to ensure that the global reconstruction results truly reflect the stability state of the underwater wellhead system, forming a reliable state prediction model.

[0064] The State Warning Module conducts stability assessments based on the global dynamic response data output by the State Observation Module. It uses a short-term risk index and a long-term instability probability as evaluation indicators. The short-term risk index is calculated by weighting the strain, stress, and bending moment extracted from the global structural response to capture the risk of sudden instability. The long-term instability probability comprehensively considers the cumulative effects of environmental loads and structural material degradation to quantify the system's long-term stability.

[0065] Based on project safety standards, the module sets three levels of warning thresholds: a green warning indicates stable system operation, with the corresponding short-term risk index and long-term instability probability within a safe range; a yellow warning indicates potential system instability, with indicators exceeding the safe range but not reaching an emergency level; and a red warning indicates a serious risk of instability, with indicators reaching or exceeding critical values. Once the module determines the system status, it generates a corresponding warning data stream, which is transmitted in real time to the signal wireless transmission module, providing a basis for subsequent warning responses.

[0066] The wireless signal transmission module, integrated into the underwater monitoring host, is responsible for reliably transmitting warning data to the platform's remote monitoring module. The encoding submodule receives the warning data stream from the status warning module and uses LDPC and RS cascade coding to encode the signal, converting the raw data into a digital signal with error correction capabilities, thereby improving data transmission reliability.

[0067] The modulation submodule uses OFDM modulation technology to convert the encoded high-speed digital signal into a multi-carrier acoustic signal. The subcarriers are orthogonal to each other, splitting the high-speed data stream into multiple parallel low-speed data streams and modulating them onto the subcarriers, enhancing the signal's anti-interference capability and transmission efficiency. The transmission submodule, based on a 4×4 MIMO design, transmits the signal through a multi-array acoustic transmitter. Signal gain improves transmission stability, ensuring accurate signal transmission to the platform's receiver even in complex underwater environments.

[0068] The platform remote monitoring module, deployed in the control center of the offshore platform, is responsible for receiving and processing warning signals, ultimately implementing stability warning functions. The signal reception submodule acquires the transmitted signal through an underwater acoustic receiver, employs an RLS equalizer to compensate for transmission delays, and uses a built-in LDPC decoder to decode the analog signal, converting it back into a digital signal for transmission to the signal processing submodule.

[0069] The signal processing submodule analyzes the digital signal, confirms the risk level represented by the early warning data, and passes it to the early warning response submodule to perform corresponding operations: when it is judged as a green early warning, the monitoring host interface displays a green mark, performs routine data recording and generates a system stability report every day; when it is judged as a yellow early warning, the interface displays a yellow mark and triggers an audible and visual alarm, automatically generates a diagnostic report listing the leading risk factors and locations, and simultaneously pushes it to the safety engineer for verification; when it is judged as a red early warning, the interface displays a red mark, triggers a strong decibel buzzer alarm and voice broadcast, and automatically activates the preset emergency plan, including the blowout preventer shutdown procedure, system emergency release, etc., and pushes the accident plan and risk status to the command center in real time to ensure rapid response and disposal.

[0070] After the system was installed and deployed, it underwent comprehensive testing. The results showed that all modules operated normally, data collection was accurate, transmission was stable, and model prediction accuracy met design requirements. During a three-month trial run, the system successfully monitored the mechanical state of the underwater wellhead system in real time. When strong currents caused increased local strain, the system promptly issued a yellow warning. After verification and confirmation by engineers, appropriate reinforcement measures were implemented, preventing further risk.

[0071] Compared with traditional monitoring methods, this system utilizes a single monitoring point design and global reconstruction technology, eliminating the need for global deployment and significantly reducing hardware costs. The integration of mechanism models and deep learning prediction methods improves the accuracy of stability assessments. Optimized signal transmission links and delay compensation ensure real-time early warning information. The system's three-level early warning mechanism enables precise, graded responses to risks, providing reliable assurance for the safe operation of deepwater subsea wellhead systems and effectively enhancing the safety and economic efficiency of deepwater oil and gas development.

[0072] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed by the present invention, which should be covered by the scope of protection of the present invention.

Claims

1. A deepwater underwater wellhead stability real-time monitoring and early warning system, characterized in that: It includes monitoring scheme design module, data acquisition and storage module, status observation module, status warning module, signal wireless transmission module and platform remote monitoring module. The execution steps are as follows: The monitoring scheme design module establishes a mechanism model and simulation model based on the environmental and structural characteristics of the underwater wellhead system to locate key mechanically sensitive areas and determine the layout of sensors and hosts. The data acquisition and storage module collects signals through strain and tilt sensors, transmits them via optical fiber to the underwater monitoring host for integration and storage, and then sends them to the status observation module. The state observation module integrates the mechanism model and the data-driven model to establish a deep learning model to form a state prediction model; the state warning module judges stability based on dynamic response data, forms a three-level warning mechanism and transmits warning data; the signal wireless transmission module encodes, modulates and transmits the warning data; the platform remote monitoring module receives and processes signals to realize stability warning.

2. A deepwater underwater wellhead stability real-time monitoring and early warning system according to claim 1, characterized in that: The monitoring scheme design module database contains the underwater wellhead system environment and structural characteristic parameters, identifies water depth, flow profile, soil parameters, underwater wellhead system configuration information, and structural coupling relationships, and locks in the monitoring leading and auxiliary variables; it supports users to copy and modify database configuration information to achieve rapid parameter input.

3. A deepwater underwater wellhead stability real-time monitoring and early warning system according to claim 2, characterized in that: The monitoring scheme design module pre-calculates the mechanical model based on the finite element method, obtains the coordinates of the key mechanical characteristic areas and the sensor layout type, location, and quantity, and provides two preset schemes: The inclination sensor uses MEMS inclinometers, which are arranged in two groups along the axial direction of the pipe string, with the spacing being the minimum unit length of the simulation model of the key mechanical sensitive area; The strain sensor uses two groups of fiber Bragg grating strain gauges, which are symmetrically arranged along the radial direction of the pipe string, with the spacing being the diameter of the pipe string at the corresponding position.

4. A deepwater underwater wellhead stability real-time monitoring and early warning system according to claim 3, characterized in that: The data acquisition and storage module collects strain data and inclination data through strain and inclination sensors, and transmits them to the wellhead monitoring host via a wired optical fiber composite cable. The monitoring host has a built-in data integration and data storage sub-module, which converts the optical fiber signal into a digital signal, integrates the data type and time to form a monitoring data stream. The data storage sub-module stores the monitoring data stream in real time and establishes a historical monitoring database for use in subsequent status prediction model correction.

5. A deepwater underwater wellhead stability real-time monitoring and early warning system according to claim 4, characterized in that: The state observation module embeds the mechanism model into the LSTM deep learning algorithm, extracts the inclination and strain data of key characteristic areas, combines historical monitoring data to train the mechanism model, reconstructs local monitoring data into global dynamic predictions, and introduces real-time verification, uncertainty management and self-correction mechanisms.

6. A deepwater underwater wellhead stability real-time monitoring and early warning system according to claim 5, characterized in that: The status warning module introduces short-term risk index and long-term instability probability as evaluation indicators. The short-term risk index is used to capture the risk of sudden instability, and the long-term instability probability quantifies the comprehensive risk of environmental loads and structural degradation. The two together support stability judgment.

7. A deepwater underwater wellhead stability real-time monitoring and early warning system according to claim 6, characterized in that: The status warning module designs a three-level warning mechanism based on evaluation indicators, namely: green warning status represents stable operation, yellow warning status represents the need to investigate the cause of instability, and red warning status represents the need to immediately implement emergency plans. Each level of warning corresponds to specific judgment conditions.

8. A deepwater underwater wellhead stability real-time monitoring and early warning system according to claim 7, characterized in that: The signal wireless transmission module includes a coding submodule, a modulation submodule and a transmission submodule. The coding submodule receives the monitoring data stream from the status warning module and uses LDPC and RS cascade to encode the signal to generate a warning digital signal; the modulation submodule generates multi-carriers through OFDM modulation technology, and each subcarrier is orthogonal to each other, dividing the high-speed data stream into multiple parallel low-speed data streams, which are modulated onto the subcarriers of each channel for transmission; the transmission submodule is designed based on a 4x4 MIMO solution and outputs the acoustic wave signal in parallel to the platform remote monitoring module.

9. A deepwater underwater wellhead stability real-time monitoring and early warning system according to claim 8, characterized in that: The platform's remote monitoring module responds according to the warning status: a green logo is displayed and a daily stability report is generated in the green warning status; an audible and visual alarm is triggered in the yellow warning status, and a diagnostic report is generated and synchronized to the safety engineer; a red warning status triggers an audible and visual alarm, emits a strong decibel buzzer alarm signal and voice broadcast, automatically executes emergency operations and pushes information to the command center.

10. A deepwater underwater wellhead stability real-time monitoring and early warning system according to claim 9, characterized in that: It also includes hardware subsystems and software subsystems: the hardware subsystem includes the platform remote monitoring host and underwater monitoring host; the software subsystem contains underlying code and user pages, supporting physical modeling, scheme design, model update, data analysis, report export and early warning query operations.

Images