Water drilling hole wall stability early warning method based on mud parameters and drilling tool posture

Abstract

The invention relates to the technical field of drilling engineering and geological safety monitoring, in particular to a water drilling hole wall stability early warning method based on slurry parameters and drilling tool postures, which comprises the following steps of: 1, sensing and acquiring multi-source heterogeneous parameters in real time; 2, constructing a hole wall stability evaluation index system; step 3, constructing a virtual drilling digital twinborn model and performing stress simulation; according to the method, synchronous acquisition and fusion analysis are carried out by integrating the mud performance parameters, the drilling tool attitude parameters and the stratum response parameters, a multi-source heterogeneous hole wall stability evaluation index system is constructed, multi-source information fusion algorithms such as a D-S evidence theory and Kalman filtering are adopted, the limitation of dependence on a single parameter in the prior art is overcome, and the method is suitable for large-scale popularization and application. According to the method, the hole wall state can be comprehensively judged from multiple dimensions of mechanical balance, slurry wall protection efficiency and drilling tool disturbance, the comprehensiveness of risk identification and the accuracy of early warning signals are remarkably improved, and misjudgment or missed judgment caused by information islands is avoided.

Description

Technical Field

[0001] This invention relates to the field of drilling engineering and geological safety monitoring technology, specifically to a method for early warning of borehole wall stability based on mud parameters and drill bit posture. Background Technology

[0002] Water drilling is a drilling process that uses water or mud as a circulating medium and borehole wall proppant, and is widely used in water conservancy projects, mineral exploration, foundation construction, and other fields. The stability of the borehole wall in water drilling directly affects the safety, efficiency, and quality of drilling operations. If the borehole wall becomes unstable, it can lead to accidents such as collapse, borehole diameter reduction, and stuck drill bits, resulting in project delays, increased costs, and even casualties. Therefore, water drilling borehole stability early warning systems, through real-time monitoring and analysis of key parameters, can identify instability risks in advance and take control measures, playing a crucial role in ensuring the smooth progress of drilling operations.

[0003] For example, application number CN202010333681.9, authorized announcement date 20210604, describes a device and method for rapidly determining the wall-protecting performance of drilling mud. It includes an outer cylinder, a first partition at the lower part of the outer cylinder, an inner cylinder fixed to the top of the first partition, and an annular cavity formed between the outer and inner cylinders. A water injection device is connected to the top of the annular cavity. A drilling rig is positioned at the center of the upper part of the inner cylinder, and a drill string is mounted on the output shaft of the drilling rig. The drilling rig is connected to a drilling mud circulation system for providing wall-protecting mud. A sediment box is located at the bottom directly below the drill string. To address the destructive effect of pore water pressure on the borehole wall after drilling and the effectiveness of freshly prepared drilling mud in wall protection, a precise indicator parameter is needed. This parameter should guide the optimization of wall protection for drilling holes in different regions and geological conditions. However, this method only focuses on the static evaluation of mud performance and does not involve the real-time correlation analysis of drill string dynamic behavior and formation response, making it difficult to cope with the risk of sudden instability in complex formations.

[0004] For example, a drilling process accident monitoring method, with application number CN200910272585.1 and publication date 20100421, includes well leakage and overflow accidents. The difference lies in that the alarm judgment for well leakage and overflow is achieved by monitoring abnormal changes in the total volume of mud in the circulating mud tank using a working condition identification model and an alarm model. When the change in the total mud volume exceeds the alarm threshold for well leakage or overflow, a corresponding alarm is triggered. This drilling process accident monitoring method can accurately and reliably determine the occurrence of well leakage, overflow, and other alarm accidents by monitoring drilling engineering parameters in real time. However, this method is limited to mud volume parameters and does not comprehensively consider multi-source information such as drill string attitude and borehole wall displacement, resulting in insufficient accuracy and comprehensiveness in early warning.

[0005] Most of the aforementioned and existing borehole stability monitoring technologies rely on mud properties or single mechanical parameters, lacking synchronous acquisition and fusion analysis of multi-source heterogeneous data such as mud, drilling tools, and formations. This makes it difficult to comprehensively reflect the borehole stability state. Furthermore, relying on threshold alarm mechanisms without introducing dynamic prediction models, it is impossible to simulate borehole stress distribution and plastic zone development in advance, resulting in delayed early warning response.

[0006] In view of this, it is urgent to design a water-based borehole wall stability early warning method based on mud parameters and drill bit attitude to solve the above problems. Summary of the Invention

[0007] The purpose of this invention is to provide a method for early warning of borehole wall stability in water drilling based on mud parameters and drill bit attitude, so as to overcome the above-mentioned shortcomings in the prior art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: A water-based borehole wall stability early warning method based on mud parameters and drill string attitude includes the following steps: Step 1. Real-time sensing and acquisition of multi-source heterogeneous parameters: The mud performance parameters, drill bit attitude parameters and formation response parameters are acquired in real time through a sensor network, and the parameters are sent to the central processing unit at a preset sampling frequency through the deployed data transmission modules. The mud performance parameters include density (ρ, measured in the range of 1.15-1.40 g / cm³). 3 ), viscosity (μ, Marshall funnel viscosity 18-50s), filtration loss (FL, API standard ≤15mL / 30min) and pH value (8-10). The drill bit attitude parameters include inclination angle (α, measurement accuracy ±0.1°), azimuth angle (β, measurement accuracy ±0.5°), rotational speed (n, range 0-200rpm), and drilling pressure (Fd, range 0-20t). The formation response parameters include borehole wall displacement (δ, monitoring threshold ≥5mm) and formation lithology data (such as rock compressive strength UCS, internal friction angle φ). Among them, the mud performance parameters were obtained by online density meter (accuracy ±0.01g / cm³). 3 Data was collected using a rotational viscometer, an API filtration meter, and a pH sensor. Drill bit attitude parameters are acquired using a MEMS inertial measurement unit and a strain gauge drill pressure sensor; Formation response parameters are acquired using distributed fiber optic sensors (spatial resolution ≤1m) and logging-while-drilling tools (including resistivity logging and sonic logging modules).

[0009] The distributed fiber optic sensors are deployed along the borehole axis at intervals of no more than 1 meter to monitor the radial displacement (δ) of the borehole wall and the temperature field distribution (temperature measurement accuracy ±0.5°C), wherein the displacement anomaly threshold is set to δ≥5mm; The logging-while-drilling tool includes resistivity logging and sonic logging modules, which are used to identify formation interfaces and lithological changes in real time. When the sonic transit time change rate exceeds 15%, a lithological change warning is triggered. The data transmission module adopts a hybrid network architecture consisting of industrial Ethernet and wireless transmission modules to ensure the reliability of data transmission in the complex environment of the drilling site. The data transmission sampling frequency f≥1Hz and the packet loss rate is controlled to ≤0.1%.

[0010] Step 2. Construction of the pore wall stability assessment index system: Based on mechanical theory and historical data, a pore wall stability assessment index system is constructed, and the indexes are quantified through preset thresholds and weights; The borehole wall stability assessment index system includes mechanical balance index, mud performance index, and drill string disturbance index; The mechanical equilibrium indices are calculated based on elastoplastic mechanics theory to determine formation collapse pressure and fracturing pressure, and Biot's consolidation theory is introduced to consider the time effect of pore water pressure dissipation on pore wall stability. The formulas for calculating formation collapse pressure and fracturing pressure are shown below: Collapse pressure: ; Rupture pressure: ; Where, σ h and σ H These are the minimum and maximum horizontal principal stresses, respectively. K is the rock mass structure coefficient, with a value ranging from 0.8 to 1.2; α is the Biot coefficient, with a value ranging from 0.6 to 0.9; P p Pore ​​pressure; Φ is the internal friction angle; UCS is the uniaxial compressive strength; The mud performance indicators are set with a viscosity mutation threshold Δμ≥10s and a filtration loss critical value FL. crit ≥15mL / 30min; The drill string disturbance index establishes a correlation model between the drill string eccentricity e and the allowable vibration acceleration a_max: Sand layer: e≤0.02·D (D is the borehole diameter) bedrock layer: a max ≤0.3g (g is the acceleration due to gravity); The weights of the indicators are assigned using the analytic hierarchy process (AHP) or the random forest algorithm. The weights of the mechanical balance indicator are ω1=0.4, the mud performance indicator is ω2=0.35, and the drill string disturbance indicator is ω3=0.25. The weights are dynamically adjusted based on real-time data (with an adjustment range not exceeding ±0.05).

[0011] Step 3. Construction of Virtual Borehole Digital Twin Model and Stress Simulation: Construct a virtual borehole digital twin model, and predict the stress distribution and plastic zone development of the borehole wall through stress simulation and seepage-stress coupling analysis, and output the stability safety factor; The virtual borehole digital twin model is constructed based on the finite element method, including geometric modeling, material constitutive relation definition, and boundary condition setting; In the definition of material constitutive relations, different combinations of constitutive models are used for different lithological strata, including but not limited to the Drucker-Prager model, the Mohr-Coulomb model, and the strain softening model, among which: Clay layer: Strain softening model adopted, yield criterion is Where τ is shear stress, c is cohesion, and σ n Normal stress; Sand layer: The Mohr-Coulomb model was used, and the yield function was... Where σ1 and σ3 are the maximum and minimum principal stresses, ; Fractured bedrock: The Drucker-Prager model was used, with the yield function being... Where I1 is the first stress invariant, J2 is the second deviatoric stress invariant, and α and k are material parameters; The stress simulation performs seepage-stress coupling analysis to calculate the distribution of the plastic zone around the hole and the stress concentration factor K. t Output stability safety factor F s ,in: ; When F s When the value is less than 1.0, it is considered to be in the plastic zone; The digital twin model supports multi-resolution modeling, uses a simplified model for rapid evaluation when computing resources are limited, and the model parameters are calibrated through real-time data, with the calibration error controlled within ≤10%.

[0012] Step 4. Multi-source information fusion and dynamic risk assessment of stability coefficient: The multi-source information fusion algorithm is used to integrate real-time monitoring data and digital twin simulation results, calculate the comprehensive stability coefficient, and classify the risk level according to the stability coefficient to realize dynamic risk assessment of the hole wall stability; The multi-source information fusion algorithm uses DS evidence theory or Kalman filtering to integrate real-time monitoring data and digital twin simulation results; The comprehensive stability coefficient is obtained through weighted calculation and risk levels are divided according to preset thresholds, including safe, low risk, medium risk, and high risk, where: The weighted formula for the overall stability coefficient S is shown below: ; Among them, F s For safety factor, Δμ is viscosity deviation, μ0 is initial viscosity, and a is actual vibration acceleration. max FL is the allowable vibration acceleration, and FL is the actual filtration loss. crit This is the critical filtration loss rate; Risk levels are determined based on the S-value: S≥0.8: Safe; 0.6≤S<0.8: Low risk; 0.4≤S<0.6: Medium risk; S<0.4: High risk; The risk level classification also takes into account the drilling depth factor, and different threshold standards are used in different depth ranges; The risk level also uses time series analysis methods (such as the ARIMA model) to predict the changing trend of the stability coefficient, identify the trend of stability deterioration in advance, and introduce confidence assessment (confidence level ≥90%) to improve the reliability of prediction.

[0013] Step 5. Generation and execution of adaptive intelligent control strategy driven by early warning: An adaptive intelligent control strategy is generated based on the risk level, and the control strategy is executed in real time by the automatic control system; The adaptive intelligent control strategy includes: When the risk level is low, adjust the mud performance parameters by injecting a viscosity improver (such as polyacrylamide, with an addition ratio of 0.1%-0.3%) or a filtration loss reducer (such as sulfonated asphalt, with an addition ratio of 0.5%-1%). At medium risk levels, collaboratively optimize drill string parameters and limit drilling speed (v≤1.5m / h) or rotation speed (n≤100rpm). In high-risk situations, implement emergency wall protection procedures by injecting composite grout containing fiber-reinforced materials or quick-setting cement grout (water-cement ratio 0.6-0.8), or lowering temporary casing (casing diameter D). c ≥D+100mm, where D is the borehole diameter), and the setting time t of the composite grout. set Adjustments can be made based on the risk level (t) set =10-30min); For karst strata, the control strategy also includes adjusting the plugging performance of the drilling mud in real time during drilling and adding plugging materials while drilling. Set a priority mechanism for the control strategy. When multiple parameters are abnormal at the same time, control measures will be executed according to the preset priority order (priority order: mud parameters > drill string parameters > emergency measures). The control strategy also takes into account changes in formation lithology and dynamically adjusts mud formulation and drilling tool operating parameters.

[0014] The automatic control system uses a PLC controller to achieve real-time execution of the control strategy, with a response delay not exceeding a preset time threshold (t). response ≤5s); The automatic control system establishes a human-machine collaborative decision-making mechanism, providing decision suggestions and intervention interfaces for field engineers while automatically adjusting the system. The automatic control system also includes a visualization of the borehole wall status based on augmented reality technology, which overlays early warning information and control suggestions onto the actual drilling scenario and supports remote access via mobile devices (data transmission latency ≤100ms).

[0015] Step 6. Closed-loop feedback and continuous optimization and iteration of model parameters: The control effect is evaluated through a closed-loop feedback mechanism, and the model parameters and knowledge base are continuously optimized based on the evaluation results. At the same time, machine learning methods are used to iteratively update the early warning model.

[0016] The closed-loop feedback mechanism evaluates the control effect by comparing the change in the stability coefficient before and after control, wherein the formula for calculating the change in the stability coefficient ΔS is as follows:

[0017] The parameters of the soil and rock mass in the digital twin model are updated based on the reinforcement learning algorithm, and the constitutive relation with an error of more than 15% is corrected. Establish a drilling data storage system based on blockchain technology to ensure the traceability and immutability of monitoring data and control records, and to achieve data sharing and auditing (data block generation time interval ≤ 1min).

[0018] The knowledge base stores historical drilling cases and control schemes, and uses machine learning methods to iteratively update the early warning model to improve prediction accuracy and adaptability. Establish a cloud-based remote monitoring center to support centralized monitoring of multiple drilling projects and remote expert consultations, and integrate an artificial intelligence-assisted decision-making module; The model is retrained monthly, and the machine learning model parameters are updated using a sliding time window (30-day window) to maintain a prediction accuracy of ≥90%.

[0019] In the above technical solution, the water-based borehole wall stability early warning method based on mud parameters and drill string attitude provided by the present invention has the following beneficial effects: (1) This invention integrates mud performance parameters, drill bit attitude parameters and formation response parameters for synchronous acquisition and fusion analysis, and constructs a multi-source heterogeneous borehole wall stability evaluation index system. Furthermore, it adopts multi-source information fusion algorithms such as DS evidence theory and Kalman filtering to overcome the limitations of existing technologies that rely on a single parameter. It can comprehensively evaluate the borehole wall status from multiple dimensions such as mechanical balance, mud wall protection effectiveness and drill bit disturbance, significantly improving the comprehensiveness of risk identification and the accuracy of early warning signals, and avoiding misjudgment or omission caused by information silos.

[0020] (2) The present invention constructs a virtual borehole digital twin model, and performs seepage-stress coupling analysis and stress simulation through the finite element method. It can dynamically simulate and predict the stress distribution and plastic zone development around the borehole wall, and output the stability safety factor. This technology breaks through the static and lag limitations of traditional threshold alarms, realizes the advanced judgment of the borehole wall instability trend, and predicts the change trend of stability coefficient by combining time series analysis, so that the system can issue an early warning before the physical instability phenomenon occurs, which buys valuable time for taking control measures and greatly enhances the safety pre-control capability of the construction process.

[0021] (3) This invention establishes an adaptive intelligent control strategy driven by early warning, which can automatically generate and execute corresponding control measures based on the results of dynamic risk assessment, such as adjusting mud performance, optimizing drill string parameters or starting emergency wall protection program. The PLC controller ensures rapid response and introduces a human-machine collaboration mechanism. In addition, the system has closed-loop feedback and continuous optimization functions. Through machine learning methods such as reinforcement learning, the digital twin model parameters and early warning model are continuously updated based on real-time feedback of control effects, so that the entire system has self-learning and adaptive capabilities, becomes more and more accurate with use, and significantly improves the long-term reliability and adaptability of the method under different geological conditions.

[0022] (4) This invention constructs an integrated monitoring and execution system. It ensures the credibility and traceability of data through blockchain-based data storage. It supports centralized management of multiple projects and remote expert consultation by utilizing cloud platforms and remote monitoring centers. In particular, it uses augmented reality technology to overlay early warning information and control suggestions onto the actual drilling scene for visualization and supports mobile access. This greatly improves the human-computer interaction experience and provides intuitive and efficient decision support tools for on-site engineers and remote experts. It connects the entire chain from data perception to intelligent decision-making to on-site execution, and improves the overall project management efficiency and collaborative handling capabilities of drilling projects. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0024] Figure 1 This is a flowchart illustrating the method for early warning of borehole wall stability based on mud parameters and drill bit attitude, as provided in this invention. Detailed Implementation

[0025] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0026] like Figure 1 As shown, the water-based borehole wall stability early warning method based on mud parameters and drill string attitude includes the following steps: Step 1. Real-time sensing and acquisition of multi-source heterogeneous parameters: The mud performance parameters, drill bit attitude parameters and formation response parameters are acquired in real time through a sensor network, and the parameters are sent to the central processing unit at a preset sampling frequency through the deployed data transmission modules. The mud performance parameters include density (ρ, measured in the range of 1.15-1.40 g / cm³). 3 ), viscosity (μ, Marshall funnel viscosity 18-50s), filtration loss (FL, API standard ≤15mL / 30min) and pH value (8-10). The drill bit attitude parameters include inclination angle (α, measurement accuracy ±0.1°), azimuth angle (β, measurement accuracy ±0.5°), rotational speed (n, range 0-200rpm), and drilling pressure (Fd, range 0-20t). The formation response parameters include borehole wall displacement (δ, monitoring threshold ≥5mm) and formation lithology data (such as rock compressive strength UCS, internal friction angle φ). Among them, the mud performance parameters were obtained by online density meter (accuracy ±0.01g / cm³). 3 Data was collected using a rotational viscometer, an API filtration meter, and a pH sensor. Drill bit attitude parameters are acquired using a MEMS inertial measurement unit and a strain gauge drill pressure sensor; Formation response parameters are acquired using distributed fiber optic sensors (spatial resolution ≤1m) and logging-while-drilling tools (including resistivity logging and sonic logging modules).

[0027] The distributed fiber optic sensors are deployed along the borehole axis at intervals of no more than 1 meter to monitor the radial displacement (δ) of the borehole wall and the temperature field distribution (temperature measurement accuracy ±0.5°C), wherein the displacement anomaly threshold is set to δ≥5mm; The logging-while-drilling tool includes resistivity logging and sonic logging modules, which are used to identify formation interfaces and lithological changes in real time. When the sonic transit time change rate exceeds 15%, a lithological change warning is triggered. The data transmission module adopts a hybrid network architecture consisting of industrial Ethernet and wireless transmission modules to ensure the reliability of data transmission in the complex environment of the drilling site. The data transmission sampling frequency f≥1Hz and the packet loss rate is controlled to ≤0.1%.

[0028] Step 2. Construction of the pore wall stability assessment index system: Based on mechanical theory and historical data, a pore wall stability assessment index system is constructed, and the indexes are quantified through preset thresholds and weights; The borehole wall stability assessment index system includes mechanical balance index, mud performance index, and drill string disturbance index; The mechanical equilibrium indices are calculated based on elastoplastic mechanics theory to determine formation collapse pressure and fracturing pressure, and Biot's consolidation theory is introduced to consider the time effect of pore water pressure dissipation on pore wall stability. The formulas for calculating formation collapse pressure and fracturing pressure are shown below: Collapse pressure: ; Rupture pressure: ; Where, σ h and σ H These are the minimum and maximum horizontal principal stresses, respectively. K is the rock mass structure coefficient, with a value ranging from 0.8 to 1.2; α is the Biot coefficient, with a value ranging from 0.6 to 0.9; P p Pore ​​pressure; Φ is the internal friction angle; UCS is the uniaxial compressive strength; The mud performance indicators are set with a viscosity mutation threshold Δμ≥10s and a filtration loss critical value FL. crit ≥15mL / 30min; The drill string disturbance index establishes a correlation model between the drill string eccentricity e and the allowable vibration acceleration a_max: Sand layer: e≤0.02·D (D is the borehole diameter) bedrock layer: a max ≤0.3g (g is the acceleration due to gravity); The weights of the indicators are assigned using the analytic hierarchy process (AHP) or the random forest algorithm. The weights of the mechanical balance indicator are ω1=0.4, the mud performance indicator is ω2=0.35, and the drill string disturbance indicator is ω3=0.25. The weights are dynamically adjusted based on real-time data (with an adjustment range not exceeding ±0.05).

[0029] Step 3. Construction of Virtual Borehole Digital Twin Model and Stress Simulation: Construct a virtual borehole digital twin model, and predict the stress distribution and plastic zone development of the borehole wall through stress simulation and seepage-stress coupling analysis, and output the stability safety factor; The virtual borehole digital twin model is constructed based on the finite element method, including geometric modeling, material constitutive relation definition, and boundary condition setting; In the definition of material constitutive relations, different combinations of constitutive models are used for different lithological strata, including but not limited to the Drucker-Prager model, the Mohr-Coulomb model, and the strain softening model, among which: Clay layer: Strain softening model adopted, yield criterion is Where τ is shear stress, c is cohesion, and σ n Normal stress; Sand layer: The Mohr-Coulomb model was used, and the yield function was... Where σ1 and σ3 are the maximum and minimum principal stresses, ; Fractured bedrock: The Drucker-Prager model was used, with the yield function being... Where I1 is the first stress invariant, J2 is the second deviatoric stress invariant, and α and k are material parameters; The stress simulation performs seepage-stress coupling analysis to calculate the distribution of the plastic zone around the hole and the stress concentration factor K. t Output stability safety factor F s ,in: ; When F s When the value is less than 1.0, it is considered to be in the plastic zone; The digital twin model supports multi-resolution modeling, uses a simplified model for rapid evaluation when computing resources are limited, and the model parameters are calibrated through real-time data, with the calibration error controlled within ≤10%.

[0030] Step 4. Multi-source information fusion and dynamic risk assessment of stability coefficient: The multi-source information fusion algorithm is used to integrate real-time monitoring data and digital twin simulation results, calculate the comprehensive stability coefficient, and classify the risk level according to the stability coefficient to realize dynamic risk assessment of the hole wall stability; The multi-source information fusion algorithm uses DS evidence theory or Kalman filtering to integrate real-time monitoring data and digital twin simulation results; The comprehensive stability coefficient is obtained through weighted calculation and risk levels are divided according to preset thresholds, including safe, low risk, medium risk, and high risk, where: The weighted formula for the overall stability coefficient S is shown below: ; Among them, F s For safety factor, Δμ is viscosity deviation, μ0 is initial viscosity, and a is actual vibration acceleration. max FL is the allowable vibration acceleration, and FL is the actual filtration loss. crit This is the critical filtration loss rate; Risk levels are determined based on the S-value: S≥0.8: Safe; 0.6≤S<0.8: Low risk; 0.4≤S<0.6: Medium risk; S<0.4: High risk; The risk level classification also takes into account the drilling depth factor, and different threshold standards are used in different depth ranges; The risk level also uses time series analysis methods (such as the ARIMA model) to predict the changing trend of the stability coefficient, identify the trend of stability deterioration in advance, and introduce confidence assessment (confidence level ≥90%) to improve the reliability of prediction.

[0031] Step 5. Generation and execution of adaptive intelligent control strategy driven by early warning: An adaptive intelligent control strategy is generated based on the risk level, and the control strategy is executed in real time by the automatic control system; The adaptive intelligent control strategy includes: When the risk level is low, adjust the mud performance parameters by injecting a viscosity improver (such as polyacrylamide, with an addition ratio of 0.1%-0.3%) or a filtration loss reducer (such as sulfonated asphalt, with an addition ratio of 0.5%-1%). At medium risk levels, collaboratively optimize drill string parameters and limit drilling speed (v≤1.5m / h) or rotation speed (n≤100rpm). In high-risk situations, implement emergency wall protection procedures by injecting composite grout containing fiber-reinforced materials or quick-setting cement grout (water-cement ratio 0.6-0.8), or lowering temporary casing (casing diameter D). c ≥D+100mm, where D is the borehole diameter), and the setting time t of the composite grout. set Adjustments can be made based on the risk level (t) set =10-30min); For karst strata, the control strategy also includes adjusting the plugging performance of the drilling mud in real time during drilling and adding plugging materials while drilling. Set a priority mechanism for the control strategy. When multiple parameters are abnormal at the same time, control measures will be executed according to the preset priority order (priority order: mud parameters > drill string parameters > emergency measures). The control strategy also takes into account changes in formation lithology and dynamically adjusts mud formulation and drilling tool operating parameters.

[0032] The automatic control system uses a PLC controller to achieve real-time execution of the control strategy, with a response delay not exceeding a preset time threshold (t). response ≤5s); The automatic control system establishes a human-machine collaborative decision-making mechanism, providing decision suggestions and intervention interfaces for field engineers while automatically adjusting the system. The automatic control system also includes a visualization of the borehole wall status based on augmented reality technology, which overlays early warning information and control suggestions onto the actual drilling scenario and supports remote access via mobile devices (data transmission latency ≤100ms).

[0033] Step 6. Closed-loop feedback and continuous optimization and iteration of model parameters: The control effect is evaluated through a closed-loop feedback mechanism, and the model parameters and knowledge base are continuously optimized based on the evaluation results. At the same time, machine learning methods are used to iteratively update the early warning model.

[0034] The closed-loop feedback mechanism evaluates the control effect by comparing the change in the stability coefficient before and after control, wherein the formula for calculating the change in the stability coefficient ΔS is as follows:

[0035] The parameters of the soil and rock mass in the digital twin model are updated based on the reinforcement learning algorithm, and the constitutive relation with an error of more than 15% is corrected. Establish a drilling data storage system based on blockchain technology to ensure the traceability and immutability of monitoring data and control records, and to achieve data sharing and auditing (data block generation time interval ≤ 1min).

[0036] The knowledge base stores historical drilling cases and control schemes, and uses machine learning methods to iteratively update the early warning model to improve prediction accuracy and adaptability. Establish a cloud-based remote monitoring center to support centralized monitoring of multiple drilling projects and remote expert consultations, and integrate an artificial intelligence-assisted decision-making module; The model is retrained monthly, and the machine learning model parameters are updated using a sliding time window (30-day window) to maintain a prediction accuracy of ≥90%.

[0037] The water-drilled borehole wall stability early warning method based on mud parameters and drill bit attitude provided by this invention has beneficial effects.

[0038] In another embodiment of the present invention, sensor configuration and data transmission reliability are enhanced for complex geological conditions (such as high-pressure aquifers or fractured zones).

[0039] Sensor network enhancement: In the deployment of distributed fiber optic sensors, a dual-ring redundant structure is adopted, with a spacing of 0.5 meters, to improve the accuracy of borehole wall displacement monitoring (δ resolution improved to ±1mm). At the same time, a microwave remote sensing module is integrated to assist in monitoring surface subsidence around the borehole, complementing the fiber optic data.

[0040] Formation response parameter expansion: A gamma logging module has been added to the logging-while-drilling tool to identify mudstone content in real time, and lithology is rapidly determined by combining the sonic transit time rate (threshold adjusted to 10%). A pore pressure monitoring unit has been introduced to directly measure Pp, improving the accuracy of mechanical equilibrium index calculations.

[0041] Data transmission architecture upgrade: Adopting the 5G-RedCap industrial IoT protocol, supporting multi-hop self-organizing networks, ensuring data transmission through relay nodes in signal dead zones, increasing the sampling frequency to 5Hz, and achieving a packet loss rate of ≤0.05%. Data compression uses lossless algorithms to reduce bandwidth consumption.

[0042] Real-time calibration mechanism: Design online self-calibration modules for sensors, such as automatic zero-point calibration of densitometers every 10 minutes and periodic calibration of viscometers using reference fluids to ensure data accuracy.

[0043] In another embodiment of the present invention, the construction and adaptive control strategy of the virtual borehole digital twin model are refined and optimized for special strata (such as expansive clay or high-stress rock strata).

[0044] Digital twin model enhancement: Multiphysics coupling, including thermo-fluid-structure interaction (TFS), is introduced into the finite element model to simulate the effect of mud temperature changes on borehole wall stress. Material constitutive relations are enhanced with the Cam-Clay model for soft clay, with the yield function being... Where p' is the effective stress, p0' is the initial stress, λ is the compressibility index, and ε is the initial stress. v The strain is represented by volumetric strain. Model parameters are dynamically updated using real-time well logging data, and a particle filtering algorithm is employed to reduce uncertainties.

[0045] Improved stress simulation accuracy: Damage mechanics theory is introduced into the calculation of the plastic zone, defining a damage variable D. A high-risk warning is triggered when D ≥ 0.8. The stability safety factor Fs is adjusted for time dependence. ; Where k is the formation creep coefficient, which is fitted using historical data.

[0046] Adaptive control strategy: For high-stress rock formations, active vibration control is introduced at medium-risk levels, suppressing resonance by adjusting the phase difference between drill bit speed and drilling pressure. In the emergency wall protection procedure, the composite slurry formula is dynamically optimized according to lithology: sodium silicate-based quick-setting agent (tset=5-10min) is added to sand layers, and bentonite is added to clay layers to enhance suspension.

[0047] Human-computer interaction interface optimization: Augmented reality visualization integrates BIM models, overlays geological profiles and real-time stress cloud maps, and supports gesture-based interactive control. The mobile application integrates voice alarms and automatic generation of control reports.

[0048] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A water-based borehole wall stability early warning method based on mud parameters and drill string attitude, characterized in that, Includes the following steps: Step 1. Real-time sensing and acquisition of multi-source heterogeneous parameters: The mud performance parameters, drill bit attitude parameters and formation response parameters are acquired in real time through a sensor network, and the parameters are sent to the central processing unit at a preset sampling frequency through the deployed data transmission modules. Step 2. Construction of the pore wall stability assessment index system: Based on mechanical theory and historical data, a pore wall stability assessment index system is constructed, and the indexes are quantified through preset thresholds and weights; Step 3. Construction of Virtual Borehole Digital Twin Model and Stress Simulation: Construct a virtual borehole digital twin model, and predict the stress distribution and plastic zone development of the borehole wall through stress simulation and seepage-stress coupling analysis, and output the stability safety factor; Step 4. Multi-source information fusion and dynamic risk assessment of stability coefficient: The multi-source information fusion algorithm is used to integrate real-time monitoring data and digital twin simulation results, calculate the comprehensive stability coefficient, and classify the risk level according to the stability coefficient to realize dynamic risk assessment of the hole wall stability; Step 5. Generation and execution of adaptive intelligent control strategy driven by early warning: An adaptive intelligent control strategy is generated based on the risk level, and the control strategy is executed in real time by the automatic control system; Step 6. Closed-loop feedback and continuous optimization and iteration of model parameters: The control effect is evaluated through a closed-loop feedback mechanism, and the model parameters and knowledge base are continuously optimized based on the evaluation results. At the same time, machine learning methods are used to iteratively update the early warning model.

2. The water-based borehole wall stability early warning method based on mud parameters and drill string attitude according to claim 1, characterized in that, In step one: The mud performance parameters include density, viscosity, filtration loss, and pH value; The drill bit attitude parameters include inclination angle, azimuth angle, rotational speed, and drilling pressure; The formation response parameters include borehole wall displacement and formation lithology data; Among them, the mud performance parameters were collected by online density meter, viscosity meter, filtration loss meter and pH sensor; Drill bit attitude parameters are acquired via an inertial measurement unit and a drill pressure sensor; Formation response parameters were acquired using distributed fiber optic sensors and logging-while-drilling tools.

3. The water-based borehole wall stability early warning method based on mud parameters and drill string attitude according to claim 2, characterized in that, The distributed fiber optic sensors are deployed along the borehole axis at intervals of no more than 1 meter to monitor the radial displacement and temperature field distribution of the borehole wall. The logging-while-drilling tool includes resistivity logging and sonic logging modules for real-time identification of formation interfaces and lithological changes; The data transmission module adopts a hybrid network architecture consisting of industrial Ethernet and wireless transmission modules to ensure the reliability of data transmission in the complex environment of the drilling site.

4. The water-based borehole wall stability early warning method based on mud parameters and drill string attitude according to claim 1, characterized in that, In step two: The borehole wall stability assessment index system includes mechanical balance index, mud performance index, and drill string disturbance index; The mechanical equilibrium index is calculated based on the elastoplastic mechanics theory to determine the formation collapse pressure and fracture pressure, and the Biot consolidation theory is introduced to consider the time effect of pore water pressure dissipation on pore wall stability. The mud performance indicators are set with viscosity mutation threshold and filtration loss critical value; The drill bit disturbance index establishes a correlation model between drill bit eccentricity and vibration acceleration; The indicators are weighted using the analytic hierarchy process or machine learning algorithms, and the weights are dynamically adjusted based on real-time data.

5. The water-based borehole wall stability early warning method based on mud parameters and drill string attitude according to claim 1, characterized in that, In step three: The virtual borehole digital twin model is constructed based on the finite element method, including geometric modeling, material constitutive relation definition, and boundary condition setting; In the definition of material constitutive relations, different combinations of constitutive models are used for different lithological strata, including but not limited to the Drucker-Prager model, the Mohr-Coulomb model and the strain softening model. The stress simulation performs seepage-stress coupling analysis, calculates the distribution of the plastic zone around the hole and the stress concentration factor, and outputs the stability safety factor. The digital twin model supports multi-resolution modeling, uses a simplified model for rapid evaluation when computing resources are limited, and the model parameters are calibrated through real-time data.

6. The water-based borehole wall stability early warning method based on mud parameters and drill string attitude according to claim 1, characterized in that, In step four: The multi-source information fusion algorithm uses DS evidence theory or Kalman filtering to integrate real-time monitoring data and digital twin simulation results; The comprehensive stability coefficient is obtained through weighted calculation and risk levels are divided according to preset thresholds, including safe, low risk, medium risk and high risk; The risk level classification also takes into account the drilling depth factor, and different threshold standards are used in different depth ranges; The risk level also uses time series analysis to predict the trend of stability coefficient changes, identify the trend of stability deterioration in advance, and introduces confidence assessment to improve the reliability of prediction.

7. The water-based borehole wall stability early warning method based on mud parameters and drill string attitude according to claim 1, characterized in that, In step five: The adaptive intelligent control strategy includes: Adjust mud performance parameters and inject viscosifiers or filtration reducers when the risk level is low; At medium-risk levels, collaboratively optimize drill string parameters and limit drilling speed or rotation speed; In high-risk situations, an emergency wall protection procedure is implemented, which involves injecting a composite grout containing fiber-reinforced materials or a quick-setting cement grout, or lowering a temporary casing. The setting time of the composite grout can be adjusted according to the risk level. For karst strata, the control strategy also includes adjusting the plugging performance of the drilling mud in real time during drilling and adding plugging materials while drilling. Set a priority mechanism for the control strategy. When multiple parameters are abnormal at the same time, control measures will be executed according to the preset priority order. The control strategy also takes into account changes in formation lithology and dynamically adjusts mud formulation and drilling tool operating parameters.

8. The water-based borehole wall stability early warning method based on mud parameters and drill string attitude according to claim 1, characterized in that, In step five: The automatic control system uses a PLC controller to realize the real-time execution of the control strategy, and the response delay does not exceed a preset time threshold. The automatic control system establishes a human-machine collaborative decision-making mechanism, providing decision suggestions and intervention interfaces for field engineers while automatically adjusting the system. The automatic control system also includes a visualization of the borehole wall status based on augmented reality technology, which overlays early warning information and control suggestions onto the actual drilling scenario and supports remote access via mobile devices.

9. The water-based borehole wall stability early warning method based on mud parameters and drill string attitude according to claim 1, characterized in that, In step six: The closed-loop feedback mechanism evaluates the control effect by comparing the changes in stability coefficients before and after control, and updates the soil and rock parameters in the digital twin model based on reinforcement learning algorithms. Establish a drilling data storage system based on blockchain technology to ensure the traceability and immutability of monitoring data and control records, and to achieve data sharing and auditing.

10. The water-based borehole wall stability early warning method based on mud parameters and drill string attitude according to claim 1, characterized in that, In step six: The knowledge base stores historical drilling cases and control schemes, and uses machine learning methods to iteratively update the early warning model to improve prediction accuracy and adaptability. Establish a cloud-based remote monitoring center to support centralized monitoring of multiple drilling projects and remote expert consultations, and integrate an artificial intelligence-assisted decision-making module; The model is retrained monthly, and the machine learning model parameters are updated using a sliding time window.

Images