Well leakage risk early warning method and device

Abstract

This invention discloses a method and device for early warning of well leakage risk. The method includes: real-time acquisition of logging data during drilling operations when drill string micro-elements are vented, and obtaining analytical logging data after time-depth conversion; obtaining a well leakage judgment index value of a first preset window length from the analytical logging data; calculating a well leakage judgment index base value based on the well leakage judgment index value of the first preset window length; calculating the deviation between each well leakage judgment index value and the well leakage judgment index base value for each acquisition point of a second preset window length, and obtaining a first well leakage judgment result based on the current well depth, the deviations corresponding to all acquisition points, a first well leakage judgment threshold, and a second well leakage judgment threshold; if the first well leakage judgment result indicates that there is a suspected well leakage in the current wellbore, determining whether a well leakage warning needs to be issued based on the hook height corresponding to the current well depth and the maximum value of the historical hook height. This invention can promptly detect wellbore leakage-prone points.

Description

Technical Field

[0001] This invention relates to the field of oil drilling technology, and in particular to a method and device for early warning of well leakage risk. Background Technology

[0002] During drilling, when the bottomhole pressure exceeds the formation leakage pressure, drilling fluid can leak into the formation, creating complex downhole conditions. The hazards of well leakage are multifaceted. First, it disrupts the integrity of the formation, impairing fluid storage and seepage conditions in the target formation; it reduces the drilling fluid's proppant-carrying capacity, forcing drilling to stop; it also easily leads to more serious and complex accidents such as well collapse, blowouts, and stuck pipe; and it significantly increases drilling fluid costs and prolongs the drilling cycle.

[0003] As oil and gas exploration and development deepens, oil and gas reservoirs become increasingly deep, drilling geology becomes increasingly complex, drilling fluid safety density windows become narrow, and well leakage becomes increasingly frequent. The coexistence of severe well leakage and blowout has become one of the most significant and complex problems hindering safe and efficient drilling.

[0004] Currently, there are two main methods for monitoring well leakage. The first is the surface parameter monitoring method for drilling fluid circulation systems. This method mainly relies on two key parameters—tank volume and outlet flow rate—to determine whether well leakage has occurred. In most field operations, ultrasonic level gauges are typically used to monitor the liquid level in the circulation tank in real time and continuously. Changes in the liquid level are used to estimate the increase or decrease in tank volume, thereby determining whether well leakage exists. However, this method has significant limitations. When there are large fluctuations in the liquid level in the circulation tank, such as mechanical disturbances caused by the agitator or the presence of air bubbles on the liquid surface, the measurement accuracy of the ultrasonic level gauge will be affected. The measurement signal will be interfered with, leading to data deviations and potentially false alarms. Such false alarms may mislead personnel into making incorrect decisions, failing to detect actual well leakage in a timely manner, posing certain safety hazards.

[0005] Another monitoring method is downhole monitoring while drilling (MWD). Compared to surface parameter monitoring, MWD can detect leakage more quickly and earlier. It uses specialized measuring instruments installed downhole to acquire real-time parameters such as pressure and flow rate, thus promptly detecting signs of well leakage. However, this method is not without its flaws. Firstly, downhole MWD instruments are expensive, requiring significant investment in equipment purchase, installation, and maintenance. This undoubtedly increases the overall cost of drilling operations, potentially creating financial strain for projects with limited budgets. Secondly, the harsh downhole environment—high temperature, high pressure, and strong vibrations—constantly threatens the normal operation of MWD instruments, posing a risk of failure. If the instrument malfunctions, timely and accurate monitoring of well leakage becomes impossible, affecting the assessment and handling of complex downhole situations.

[0006] While these two existing well leakage monitoring methods can detect well leakage problems to some extent, they both have significant shortcomings. Monitoring results are often only available at the alarm stage, indicating a lag in leakage identification. By the time leakage is detected, the downhole situation may already be quite serious, increasing the difficulty and cost of subsequent handling. Neither method can accurately pinpoint the location of the leak. The inability to quickly locate the leak after it occurs hinders the development of targeted well leakage control measures and the timely implementation of effective treatment plans, posing a serious threat to drilling safety. Therefore, a more advanced and accurate well leakage monitoring technology is urgently needed to solve these problems. Summary of the Invention

[0007] In a first aspect, embodiments of the present invention provide a well leakage risk early warning method, which can promptly detect leakage-prone points and achieve well leakage risk early warning through logging data during drill string micro-element venting, with good application effect. The method includes:

[0008] Real-time acquisition of logging data during drilling operations when drill string micro-elements are vented, and time-depth conversion of the logging data to obtain logging data for analysis;

[0009] Using the latest well depth in the analytical logging data as the starting point, a sliding window analysis is performed on the shallow well section to obtain the well leakage judgment index value of the first preset window length from the analytical logging data.

[0010] Calculate the base value of the well leakage judgment index based on the well leakage judgment index value of the first preset window length;

[0011] For each collection point of the second preset window length, calculate the deviation between each well leakage judgment index value and the well leakage judgment index base value. If the current well depth is not greater than the first depth, obtain the first well leakage judgment result based on the relationship between the deviations corresponding to all collection points and the first well leakage judgment threshold. Otherwise, obtain the first well leakage judgment result based on the relationship between the deviations corresponding to all collection points and the second well leakage judgment threshold.

[0012] If the first well leakage assessment result indicates that there is a suspected well leakage in the current wellbore, a well leakage warning is issued based on the maximum value of the hook height corresponding to the current well depth and the historical hook height.

[0013] Secondly, embodiments of the present invention also provide a well leakage risk early warning device, which can promptly detect leakage-prone points and achieve well leakage risk early warning through logging data during drill string micro-element venting, with good application effect. The device includes:

[0014] The logging data acquisition module is used to acquire logging data in real time when the drill string micro-element is vented during drilling operations, and to perform time-depth conversion on the logging data to obtain logging data for analysis.

[0015] The well leakage judgment index value calculation module is used to perform sliding window analysis from the latest well depth in the analysis logging data to the shallow well section, and obtain the well leakage judgment index value of the first preset window length from the analysis logging data.

[0016] The well leakage judgment index base value calculation module is used to calculate the well leakage judgment index base value based on the well leakage judgment index value of the first preset window length;

[0017] The well leakage detection module is used to calculate the deviation between each well leakage detection index value and the well leakage detection index base value for each collection point of the second preset window length. If the current well depth is not greater than the first depth, the first well leakage detection result is obtained based on the relationship between the deviations corresponding to all collection points and the first well leakage detection threshold. Otherwise, the first well leakage detection result is obtained based on the relationship between the deviations corresponding to all collection points and the second well leakage detection threshold. If the first well leakage detection result indicates that there is a suspected well leakage in the current wellbore, the module determines whether a well leakage warning needs to be issued based on the maximum value of the hook height corresponding to the current well depth and the historical hook height.

[0018] Thirdly, embodiments of the present invention also provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned well leakage risk early warning method.

[0019] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described well leakage risk warning method.

[0020] Fifthly, embodiments of the present invention also provide a computer program product, the computer program product including a computer program, which, when executed by a processor, implements the above-mentioned well leakage risk early warning method.

[0021] In this embodiment of the invention, real-time logging data is collected during the venting of drill string micro-elements. Venting of drill string micro-elements is often one of the important early warning signs of well leakage. By collecting and analyzing data at this specific moment, monitoring can be initiated as soon as signs of well leakage appear. Compared to the traditional method of identifying well leakage based on changes in overall parameters after a certain period of time has passed, this method can detect potential well leakage risks much earlier, thus gaining more time for subsequent countermeasures. The collected logging data undergoes time-depth conversion, transforming the time-dimensional data into depth sequence data based on well depth. This better reflects the actual situation in drilling operations where well depth is the key reference, allowing subsequent data mining and parameter analysis to be conducted on a more scientific and reasonable data foundation. This improves the accuracy of data analysis and provides more reliable data support for well leakage assessment. Using sliding window analysis to determine well leakage assessment index values ​​and calculate baseline values ​​allows for the dynamic construction of well leakage assessment benchmarks that align with current formation conditions and drilling operations, based on real-time drilling progress and actual data from different well sections. Compared to using fixed thresholds or empirical parameters, this approach better adapts to various complex drilling scenarios, reducing the probability of false alarms and missed alarms. By selecting different well leakage assessment thresholds based on well depth and comparing the hook height with historical maximum hook heights, this hierarchical, multi-dimensional assessment model comprehensively considers the differences in formation characteristics at different well depths and the actual operating status of the drilling tools. This makes the well leakage early warning assessment process more rigorous and reliable, avoiding inaccuracies caused by single indicators or fixed standards, and further improving the accuracy of well leakage early warning. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0023] Figure 1 This is a flowchart of the well leakage risk early warning method in an embodiment of the present invention;

[0024] Figure 2 This is a flowchart illustrating well leakage risk warning based on specific data in an embodiment of the present invention;

[0025] Figure 3 This is a schematic diagram of the well leakage risk early warning device in an embodiment of the present invention;

[0026] Figure 4 This is another structural schematic diagram of the well leakage risk early warning device in an embodiment of the present invention;

[0027] Figure 5 This is a schematic diagram of a computer device in an embodiment of the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0029] The principle of the solution proposed in this invention is based on intelligent early warning of well leakage risk through drill string micro-element venting. By collecting and analyzing the logging data during drilling operations when drill string micro-element venting occurs in real time, it is possible to more accurately identify downhole leakage points, quickly capture subtle changes in well leakage risk, and achieve early warning of well leakage risk.

[0030] Micro-element venting is a term used in drilling engineering to describe a subtle anomaly during drilling. It primarily refers to the momentary weightlessness or sudden movement of the drill bit at a specific, minute depth (micro-element), where the drill bit suddenly moves downwards a short distance without obstruction, similar to venting but on a much smaller scale. Micro-element venting is often an early signal of formation anomalies, potentially indicating the presence of: loosely permeable formations; micro-fractures or caves (especially in carbonate formations); or abnormal formation pressure zones. These conditions can increase the risk of lost circulation and blowouts. Therefore, micro-element venting phenomena need to be carefully recorded and analyzed during logging and drilling monitoring to provide a basis for subsequent drilling parameter adjustments (such as reducing drilling pressure and controlling flow rate).

[0031] Figure 1 This is a flowchart of a well leakage risk early warning method in an embodiment of the present invention. The method includes:

[0032] Step 101: Real-time acquisition of logging data during drilling operations when drill string micro-elements are vented, and time-depth conversion of the logging data to obtain logging data for analysis.

[0033] Step 102: Using the latest well depth in the analysis logging data as the starting point, perform sliding window analysis towards the shallow well section, and obtain the well leakage judgment index value of the first preset window length from the analysis logging data;

[0034] Step 103: Calculate the well leakage judgment index base value based on the well leakage judgment index value of the first preset window length;

[0035] Step 104: For each collection point of the second preset window length, calculate the deviation between each well leakage judgment index value and the well leakage judgment index base value. If the current well depth is not greater than the first depth, obtain the first well leakage judgment result based on the relationship between the deviations corresponding to all collection points and the first well leakage judgment threshold; otherwise, obtain the first well leakage judgment result based on the relationship between the deviations corresponding to all collection points and the second well leakage judgment threshold.

[0036] Step 105: If the first well leakage judgment result is that there is a suspected well leakage in the current wellbore, determine whether a well leakage warning needs to be issued based on the maximum value of the hook height corresponding to the current well depth and the historical hook height.

[0037] In this embodiment of the invention, real-time logging data is collected during the venting of drill string micro-elements. Venting of drill string micro-elements is often one of the important early signals of well leakage. By collecting and analyzing data at this specific moment, monitoring can be carried out when well leakage first appears. Compared with the traditional method of identifying well leakage based on changes in overall parameters after a certain period of time has passed, this method can detect potential well leakage risks earlier, thus gaining more time for subsequent countermeasures. The collected logging data undergoes time-depth conversion, transforming the time-dimensional data into depth sequence data based on well depth. This better reflects the actual situation in drilling operations where well depth is the key reference, allowing subsequent data mining and parameter analysis to be conducted on a more scientific and reasonable data foundation. This improves the accuracy of data analysis and provides more reliable data support for well leakage judgment. Using sliding window analysis to determine well leakage assessment index values ​​and calculate baseline values ​​allows for the dynamic construction of well leakage assessment benchmarks that align with current formation conditions and drilling operations, based on real-time drilling progress and actual data from different well sections. Compared to using fixed thresholds or empirical parameters, this approach better adapts to various complex drilling scenarios, reducing the probability of false alarms and missed alarms. By selecting different well leakage assessment thresholds based on well depth and comparing the hook height with historical maximum hook heights, this hierarchical, multi-dimensional assessment model comprehensively considers the differences in formation characteristics at different well depths and the actual operating status of the drilling tools. This makes the well leakage early warning assessment process more rigorous and reliable, avoiding inaccuracies caused by single indicators or fixed standards, and further improving the accuracy of well leakage early warning.

[0038] Each step is described in detail below.

[0039] In step 101, logging data is collected in real time when the drill string micro-element is vented during drilling operations, and the logging data is converted from time to depth to obtain logging data for analysis.

[0040] In practice, a comprehensive logging instrument can be used to collect logging data from the target well area in real time during drilling operations and input it into a database.

[0041] In one embodiment, logging data is acquired in real time during drilling operations when drill string micro-elements are vented, and the logging data is converted to depth over time to obtain logging data for analysis, including:

[0042] When the real-time logging data meets the drilling conditions, the logging data is standardized.

[0043] The standardized logging data is converted from time series to depth series to obtain logging data for analysis.

[0044] In the above embodiments, non-real-time logging data is filtered out and not monitored by identifying drilling conditions. Only logging data that meets the drilling conditions is subsequently monitored. The well leakage risk warning for micro-element venting is based on the fact that when drilling encounters formations with a risk of well leakage but very thin thickness, compared to the upper well section under similar drilling conditions, the micro-element venting section (i.e., the leakage-prone point) will simultaneously exhibit data characteristics such as fast drilling time, low drilling pressure, and low MSE (mechanical specific energy). Therefore, intelligent well leakage risk warning needs to be implemented during the drilling process.

[0045] To address the issues of missing and noisy logging data caused by equipment failure, acquisition errors, and unit inconsistencies, data standardization can be implemented. Missing logging data can be filled using methods such as weighted interpolation, mean interpolation, linear interpolation, and nearest neighbor interpolation. Machine learning interpolation models are also introduced: algorithms such as random forest and LSTM are used, combined with historical similar well data characteristics, to predictively fill in missing values ​​(more suitable for non-linear data than traditional interpolation). Abnormal logging data is corrected using techniques such as fixed threshold judgment and K-Sigma methods to identify and process outliers and abnormal fluctuations, ensuring the accuracy and continuity of the logging data. Specifically, for the systematic errors of different logging instrument models, error compensation models can be established using historical calibration data (such as correction formulas for ultrasonic level gauges affected by temperature) to eliminate noise caused by equipment differences. Wavelet transform or Kalman filtering is used to separate high-frequency noise (such as agitator vibration interference) from effective signals (such as actual liquid level changes), improving the signal-to-noise ratio.

[0046] For example, the logging data is converted from time series to depth series based on a depth interval of 0.1 meters (e.g., instantaneous drilling time is collected and calculated every 0.1m, in min / m). The converted logging data is then properly stored in a database for further analysis and research.

[0047] In practice, the standardized logging data can be converted from time series to depth series according to the preset depth interval. The depth interval can be adaptively adjusted based on formation changes. When encountering homogeneous formations (such as thick sandstone) during drilling operations, the first depth interval (e.g., 0.1 m - 0.5 m) is used, and the second depth interval (e.g., 0.05 m) is used in heterogeneous or thin interbedded formations (such as shale and limestone interbedded), thereby improving the identification accuracy of thin and easily leaked layers.

[0048] When collecting logging data in real time, the sampling frequency can be dynamically adjusted according to the drill bit type (such as PDC drill bit, roller cone drill bit): increase the sampling frequency (e.g., 2 times per second) when the PDC drill bit is drilling fast, and decrease the frequency when the roller cone drill bit is slow, so as to balance the amount of data and the calculation efficiency.

[0049] In one embodiment, the logging data used for analysis includes one or any combination of time, well depth, drill bit position, instantaneous rotation speed, rotary table rotation speed, drilling pressure, and hook height. Examples of logging data are shown in Table 1.

[0050] Table 1

[0051]

[0052] Steps 102-105 describe the specific process of monitoring and analyzing wellbore leakage based on drill string micro-elements. This process incorporates expert research experience into the intelligent early warning system, continuously and automatically monitoring real-time data from logging and drilling operations. Through intelligent algorithms and data analysis, it identifies downhole leakage risks and issues timely warning signals, replacing manual monitoring and analysis, thus empowering and improving wellbore leakage monitoring. This, in turn, helps optimize drilling measures, reduces wellbore leakage risks, lowers the incidence of wellbore leakage, and ensures drilling safety. Steps 102-105 can be considered as steps in the intelligent early warning model.

[0053] In step 102, a sliding window analysis is performed on the shallow well section using the latest well depth in the analysis logging data as the starting point, and the well leakage judgment index value of the first preset window length is obtained from the analysis logging data.

[0054] In one embodiment, the well leakage detection index values ​​include instantaneous drilling rate, drilling pressure, and rotary table speed.

[0055] Using the latest well depth D in the logging data as the starting point, a sliding window analysis is performed on the shallow well section. Taking the first preset window length of 5 as an example, the current instantaneous drilling rate is recorded as ROP1, the current drilling pressure as WOB1, and the current rotary table speed as RPM1; and the instantaneous drilling rate of the previous acquisition point, i.e., the second acquisition point, is recorded as ROP2, drilling pressure as WOB2, and rotary table speed as RPM2; from the second acquisition point, the previous acquisition point is traced back to the history, i.e., the third acquisition point, and the instantaneous drilling rate is recorded as ROP3, drilling pressure as WOB3, and rotary table speed as RPM3; and so on, continuously tracing back 5 more instantaneous drilling rates, drilling pressures, and rotary table speeds.

[0056] In step 103, the well leakage judgment index base value is calculated based on the well leakage judgment index value of the first preset window length;

[0057] The arithmetic mean of five instantaneous drilling speeds, drilling pressures, and rotary table speeds are calculated as base values. The arithmetic mean of the instantaneous drilling speed is denoted as ROP0, the arithmetic mean of the drilling pressure is denoted as WOB0, and the arithmetic mean of the rotary table speed is denoted as RPM0, thus forming the instantaneous drilling speed base value, drilling pressure base value, and rotary table speed base value.

[0058] In step 104, the first well leakage judgment result is obtained based on the current well depth, the well leakage judgment index value of the second preset window length, and the well leakage judgment index base value.

[0059] In one embodiment, a first well leakage judgment result is obtained based on the current well depth, the well leakage judgment index value of the second preset window length, and the well leakage judgment index base value, including:

[0060] If the current well depth is not greater than the first depth, for each well leakage judgment index value at each collection point in the second preset window length, calculate the deviation between the well leakage judgment index value and the well leakage judgment index base value. If the deviations corresponding to the preset number of collection points are all less than the first well leakage judgment threshold, it is determined that the well leakage judgment index value can determine that there is a suspicious well leakage.

[0061] If the current well depth is greater than the first depth, for each well leakage judgment index value, at each collection point within the second preset window length, calculate the deviation between the well leakage judgment index value and the well leakage judgment index base value. If the deviations corresponding to the preset number of collection points are all less than the second well leakage judgment threshold, it is determined that the well leakage judgment index value can be used to judge the existence of a suspicious well leakage.

[0062] If all well leakage judgment index values ​​can determine that there is a suspected well leakage, the first well leakage judgment result is determined to be that there is a suspected well leakage in the current wellbore.

[0063] In one embodiment, if the well leakage judgment index value is the instantaneous drilling rate, the deviation is a first ratio of the instantaneous drilling rate to the instantaneous drilling rate baseline value;

[0064] If the well leakage judgment index is the drilling pressure, the deviation is the second ratio of the drilling pressure to the drilling pressure baseline value;

[0065] If the well leakage judgment index value is the rotary table speed, the deviation is the absolute value of the difference between the rotary table speed and the base value of the rotary table speed.

[0066] Since the characteristics of drilling speed, drilling pressure and rotary table speed at leakage points will change differently as well depth increases, the first well depth (which can be 2500m) is used as the standard for dividing shallow and deep well sections, and different leakage point judgment conditions are assigned to each section.

[0067] Taking a second preset window length of 3 (i.e., including 3 collection points) and a preset quantity of 2 as an example, different logical judgments are made for different well depths.

[0068] When the well depth is less than or equal to the first well depth, the following analysis and judgment are performed:

[0069] a. The first ratio of the instantaneous drilling rate ROP1 / instantaneous drilling rate base value ROP0 at the current (first sampling point) is less than the first well leakage judgment threshold corresponding to the instantaneous drilling rate (taking 0.50 as an example), or the first ratio of the instantaneous drilling rate ROP2 / instantaneous drilling rate base value ROP0 at the second sampling point is less than 0.50, or the first ratio of the instantaneous drilling rate ROP3 / instantaneous drilling rate base value ROP0 at the third sampling point is less than 0.50. If two of the above three conditions are met, the judgment condition of increased drilling rate is met.

[0070] b. If the second ratio of the current drilling pressure WOB1 / drilling pressure baseline WOB0 is less than the first well leakage judgment threshold corresponding to the drilling pressure (taking 0.70 as an example), or the second ratio of the drilling pressure WOB2 / drilling pressure baseline WOB0 at the second sampling point is less than 0.70, or the second ratio of the drilling pressure WOB3 / drilling pressure baseline WOB0 at the third sampling point is less than 0.70, then the judgment condition of decreasing drilling pressure is met.

[0071] c. The absolute value of the current rotary table speed RPM1 - the base value of the rotary table speed RPM0 is less than (the first well leakage judgment threshold corresponding to the rotary table speed, which is 5 here), i.e., ABS(RPM1-RPM0) < 5; or the absolute value of the rotary table speed RPM2 - the base value of the rotary table speed RPM0 at the second collection point is less than 5, i.e., ABS(RPM2-RPM0) < 5; or the absolute value of the rotary table speed RPM3 - the base value of the rotary table speed RPM0 at the third collection point is less than 5, i.e., ABS(RPM3-RPM0) < 5. If two of the above three conditions are met, the judgment condition of stable rotary table speed is met.

[0072] When the well depth is greater than the first well depth, the following analysis and judgment are performed:

[0073] a. If the first ratio of the current drilling rate ROP1 to the instantaneous drilling rate base value ROP0 is <0.70, or the first ratio of the second drilling rate ROP2 to the instantaneous drilling rate base value ROP0 is < the second well leakage judgment threshold corresponding to the instantaneous drilling rate (taken as 0.70 here), or the first ratio of the third drilling rate ROP3 to the instantaneous drilling rate base value ROP0 is <0.70, then the judgment condition of faster drilling time is met.

[0074] b. If the second ratio of the current drilling pressure WOB1 / drilling pressure baseline WOB0 is less than the second well leakage judgment threshold corresponding to the drilling pressure (taken as 0.80 here), or the second ratio of the second drilling pressure WOB2 / drilling pressure baseline WOB0 is less than 0.80, or the second ratio of the third drilling pressure WOB3 / drilling pressure baseline WOB0 is less than 0.80, then the drilling pressure decreases.

[0075] c. The absolute value of the current rotary table speed RPM1 minus the base value RPM0 is less than the second well leakage judgment threshold (taken as 5 here), i.e., ABS(RPM1-RPM0) < 5; or the absolute value of the second rotary table speed RPM2 minus the base value RPM0 is less than 5, i.e., ABS(RPM2-RPM0) < 5; or the absolute value of the third rotary table speed RPM3 minus the base value RPM0 is less than 5, i.e., ABS(RPM3-RPM0) < 5. If two of the above three conditions are met, the judgment condition of stable rotary table speed is met.

[0076] In addition to instantaneous drilling rate, drilling pressure, and rotary table speed, the following indicators can also be considered for well leakage assessment:

[0077] Mechanical specific energy (MSE): MSE is calculated directly (MSE = rotational energy + drilling energy / volumetric removal). When well leakage occurs, MSE decreases significantly, so it is used as one of the core indicators, complementing drilling rate and drilling pressure.

[0078] Drill bit torque (T): Lost formations are usually accompanied by abnormal torque fluctuations (such as a sudden drop in torque in fractured formations). The ratio of drill bit torque to the base value of drill bit torque is increased as an auxiliary judgment condition.

[0079] Hook load change rate: When well leakage occurs, the friction between the drill string and the well wall changes, causing an abnormal trend in the hook load. The hook load change rate per unit depth (ΔF / ΔD) is calculated and incorporated into the judgment system.

[0080] In this embodiment of the invention, the first preset window length and the second preset window length can be adaptively adjusted based on the formation complexity. In one embodiment, the method further includes:

[0081] The formation entropy value of the current well section is calculated in real time. When the formation entropy value is less than the entropy threshold, the length of the first preset window and the length of the second preset window are increased according to a preset ratio.

[0082] Formation entropy reflects the heterogeneity of the formation. When the entropy value is high (complex formations), a short window (such as 3-5 sampling points) is used, and when the entropy value is low (homogeneous formations), a long window (such as 8-10 sampling points) is used to avoid missing complex features by a single window.

[0083] In step 105, if the first well leakage judgment result is that there is a suspected well leakage in the current wellbore, it is determined whether a well leakage warning needs to be issued based on the maximum value of the hook height corresponding to the current well depth and the historical hook height.

[0084] In one embodiment, determining whether a well leakage warning needs to be issued based on the current well depth corresponding to the hook height and the maximum historical hook height includes:

[0085] Calculate the first difference between the hook height corresponding to the current well depth and the maximum value of the historical hook height;

[0086] When the first difference is less than the difference threshold, it is determined that a well leakage warning needs to be issued.

[0087] Record the hook height corresponding to the current well depth as HH, and the historical maximum hook height as HHmax. If HH - HHmax < the difference threshold (e.g., -0.3m), it indicates that this leakage point characteristic is not caused by the connection of the drill string, and a well leakage warning can be issued. At this time, under the condition of stable rotary table speed, the drilling speed increases, the drilling pressure decreases, and the mechanical specific energy (MSE) decreases, indicating that the drill bit has encountered a leakage point, and there is a greater possibility of fractures or highly permeable porous formations, thus the risk of well leakage is greater.

[0088] Then, the time when a well leakage warning needs to be issued is determined as the warning time, and a warning message is generated. The warning message includes the warning time and the warning content. For example, the warning message can be expressed as: 2024 / 05 / 16 16:52:08, well depth 5659.95m, suspected encounter with a highly permeable and easily leaking formation, pay attention to prevent well leakage.

[0089] The warning information can then be stored in a well leakage warning database. Backend developers can create a query interface, which can be called by the frontend and integrated and displayed on the page. This will ensure that engineering and technical personnel can quickly obtain key warning information and formulate drilling risk management strategies accordingly to ensure drilling operation safety.

[0090] The following is a specific embodiment to illustrate the effectiveness of the method proposed in this invention, using the LTX2 well as an example. Figure 2 This is a flowchart illustrating well leakage risk warning based on specific data in an embodiment of the present invention.

[0091] (1) First, see Figure 2 The system collects logging data in real time when the drill string micro-element is vented during drilling operations. When the real-time logging data meets the drilling conditions, the logging data is standardized. The standardized logging data is then converted from a time series to a depth series to obtain logging data for analysis.

[0092] Subsequently, the instantaneous drilling rate, drilling pressure, and rotary table speed were analyzed and calculated to form the baseline values ​​for instantaneous drilling rate, drilling pressure, and rotary table speed. When the LTX2 well was in drilling operation, the corresponding well depth was 7356.55m. The current instantaneous drilling rate ROP1 was 15m / hr, the current drilling pressure WOB1 was 5.26kN, and the current rotary table speed RPM1 was 27.56RPM. With a depth of 0.1m as the sampling interval, the corresponding well depth of the second sampling point was traced back from the current well depth to the previous sampling point, which was 7356.45m. The instantaneous drilling rate was 3.16m / hr, the drilling pressure WOB2 was 6.73kN, and the rotary table speed RPM2 was 27.4RPM. Using a depth interval of 0.1 meters, the well depth corresponding to the previous historical sampling point (the third sampling point) was traced back from the second sampling point to a depth of 7356.35 m. The instantaneous drilling rate was 3.47 m / hr, the drilling pressure (WOB3) was 16.29 kN, and the rotary table speed (RPM3) was 27.66 RPM. This process was repeated, tracing back five consecutive historical sampling points. The calculated baseline values ​​for these five data points were: instantaneous drilling rate (ROP0) of 5.73 m / hr, drilling pressure (WOB0) of 16.08 kN, and rotary table speed (RPM0) of 27.57 RPM. These values ​​were then imported into the instantaneous drilling rate baseline database, drilling pressure baseline database, and rotary table speed baseline database.

[0093] (2) Analysis of well leakage early warning of drilling tool micro-element venting

[0094] The current well depth is 7356.55m > 2500m. The following analysis and judgment are made:

[0095] ① The current instantaneous drilling rate ROP1 / instantaneous drilling rate baseline ROP0 = 15 / 5.73 = 2.62 > 0.70, or the instantaneous drilling rate ROP2 / instantaneous drilling rate baseline ROP0 = 3.16 / 5.73 = 0.55 < 0.70, or the instantaneous drilling rate ROP3 / instantaneous drilling rate baseline ROP0 = 3.47 / 5.73 = 0.61 < 0.70, and two of the above three threshold conditions are met;

[0096] ② The current drill pressure WOB1 / drill pressure baseline WOB0 = 5.26 / 16.08 = 0.33 < 0.80, or the drill pressure WOB2 / drill pressure baseline WOB0 at the second sampling point = 6.73 / 16.08 = 0.42 < 0.80, or the drill pressure WOB3 / drill pressure baseline WOB0 = 16.29 / 16.08 = 1.01 > 0.80. Two of the above three threshold conditions are met.

[0097] ③ The absolute value of the current turntable speed RPM1 - the base value of turntable speed RPM0 is <5, i.e., ABS(RPM1-RPM0) = ABS(27.56-27.57) = 0.01 < 5; or the absolute value of the turntable speed RPM2 - the base value of turntable speed RPM0 at the second sampling point is <5, i.e., ABS(RPM2-RPM0) = ABS(27.4-27.57) = 0.17 < 5; or the absolute value of the turntable speed RPM3 - the base value of turntable speed RPM0 at the third sampling point is <5, i.e., ABS(RPM3-RPM0) = ABS(27.66-27.57) = 0.09 < 5. All three threshold conditions are met.

[0098] ④ In summary, the LTX2 well exhibited characteristics of increased drilling speed, decreased drilling pressure, and reduced mechanical energy (MSE) within the range of 7355.95m to 7356.55m, while maintaining a stable rotary table speed. Furthermore, the recorded hook height HH corresponding to the current well depth was 13.81m, and the historical maximum hook height HHmax was 16m. This satisfies the condition that HH-HHmax=13.81-16.11=-2.3<-0.3m, which is considered normal drilling condition. This aligns with the judgment logic for well leakage warning identification based on drill string micro-element venting, indicating a high probability that the drill bit encountered a leakage-prone point and a significant well leakage risk. Consequently, a well leakage warning was issued.

[0099] (3) Send overflow warning information

[0100] The warning interface is shown in Table 2 below.

[0101] Table 2

[0102]

[0103] The drilling log for April 17, 2023 shows: "From 08:00 on April 16, 2023 to 05:00 on April 17, 2023, composite drilling reached a depth of 7359.27m. Well leakage occurred; the standpipe pressure decreased by 18.23 MPa from 19.61 MPa; the outlet flow rate decreased by 65.41% from 67.03%; other parameters remained unchanged; there was no loss of return at the wellhead. From 05:00 to 06:00, the drill string was lifted from the bottom of the well, and the flow rate was reduced to circulate and measure the leakage velocity (flow rate 10 L / s, leakage velocity 5.1 m / s)." 3 / h, 1.23m of drilling fluid lost, 4.7m 3From 08:00 onwards, directional composite drilling (drilling through the choke channel) was carried out. A comparison of the early warning results and the drilling log shows that well LTX2 successfully issued a leak warning 51 minutes and 2.61 meters in advance.

[0104] Through the above steps, well leakage early warning can be achieved. In order to further improve the accuracy of well leakage early warning, this embodiment of the invention proposes a method that integrates statistical models with machine learning models and spatiotemporal correlation analysis. Through the logical progression of data layer, analysis layer, verification layer and correlation layer, accurate early warning of well leakage risk can be achieved. The required data are logging data and historical logging data and well leakage records of adjacent wells in the current wellbore. Among them, adjacent wells are adjacent wells in the same block. The selection criteria are that the distance from the current well is ≤5km, the geological strata are the same, and the deviation of the completed well depth is ≤500m to ensure similarity of geological conditions.

[0105] In one embodiment, the method further includes:

[0106] When the first difference is less than the difference threshold, the deviation between the well leakage judgment index value and the well leakage judgment index base value is input into the random forest model to obtain the well leakage probability of the current well. The random forest model is trained based on the historical logging data and well leakage records of the adjacent wells of the current well.

[0107] If the probability of well leakage is greater than the probability threshold, a well leakage warning needs to be issued.

[0108] In this embodiment of the invention, the training set for training the random forest model consists of historical logging data and well leakage records of neighboring wells within the same block (corresponding to the associated data in step 1). The input feature is the deviation between the well leakage judgment index value and the well leakage judgment index base value in the historical logging data of neighboring wells in the current wellbore. The label data is the well leakage records of neighboring wells in the current wellbore (1 indicates well leakage has occurred, 0 indicates normal). By training the random forest model and optimizing parameters (such as the number of decision trees and maximum depth), the random forest model can identify well leakage characteristic deviation patterns (such as a combination of sudden increase in drilling speed + sudden decrease in drilling pressure + sudden decrease in MSE). During prediction, the real-time deviation of the current wellbore is input into the random forest model, and the well leakage probability P of the current wellbore is output. The well leakage probability P ranges from 0 to 1; a larger P indicates a higher risk of well leakage. If P ≥ 0.7 (the probability threshold is determined by verification through historical data), it indicates that there is a suspected well leakage. Moreover, the first difference value has been determined to be less than the difference threshold. At this time, the two methods make the same judgment, so it is determined that a well leakage warning needs to be issued. This realizes the elimination of misjudgment by a single model through consistency verification and the retention of highly reliable initial screening results.

[0109] In one embodiment, the method further includes:

[0110] After determining that a well leakage warning needs to be issued, the deviations between the well leakage judgment index values ​​and the well leakage judgment index base values ​​of all the nearest neighbor wells in the current wellbore within the current well depth range are extracted from the well leakage feature map of the block, forming a query deviation vector;

[0111] The similarity between the deviation vector to be analyzed, formed by the deviations of all well leakage judgment index values ​​of the current wellbore at the current well depth from the base value of the well leakage judgment index, and the query deviation vector;

[0112] The credibility of the warning is determined based on the similarity.

[0113] Warning information is generated based on the warning time, warning type, current well depth, warning credibility, and warning description.

[0114] In this embodiment of the invention, when constructing the block well leakage feature map, it can be based on the historical logging data of adjacent wells and constructed according to depth intervals (e.g., every 100 meters is an interval). Each depth interval stores the deviation vector formed by the average deviation of all well leakage judgment index values ​​from the well leakage judgment index base value; when calculating the similarity S, the cosine similarity calculation method can be used.

[0115] The following rules can be used to determine the credibility of an alert based on the similarity:

[0116] If S≥0.6, the confidence level of the early warning is determined to be 0.9; (the thresholds of 0.6 and 0.9 are calibrated using historical data from the block), indicating that the suspicious characteristics of the current well are highly similar to the well leakage characteristics of neighboring wells at the same depth, and the confidence level of the early warning is high;

[0117] If 0.4 ≤ S < 0.6, the confidence level of the early warning is determined to be 0.7;

[0118] If S < 0.4, the confidence level of the early warning is determined to be 0.5.

[0119] In this embodiment of the invention, warnings can be classified into different levels:

[0120] High-risk warning: The confidence level of the warning is 0.9;

[0121] Medium-risk warning: The confidence level of the warning is 0.7;

[0122] Low-risk warning: The credibility of the warning is 0.5.

[0123] The warning level can be included in the warning information.

[0124] This invention also proposes a well leakage risk early warning device, the principle of which is similar to the well leakage risk early warning method, and will not be described in detail here.

[0125] Figure 3This is a schematic diagram of the well leakage risk early warning device in an embodiment of the present invention. The device includes:

[0126] The logging data acquisition module 301 is used to acquire logging data in real time when the drill string micro-element is vented during drilling operations, and to perform time-depth conversion on the logging data to obtain logging data for analysis.

[0127] The well leakage judgment index value calculation module 302 is used to perform sliding window analysis from the latest well depth in the analysis logging data to the shallow well section, and obtain the well leakage judgment index value of the first preset window length from the analysis logging data.

[0128] The well leakage judgment index base value calculation module 303 is used to calculate the well leakage judgment index base value based on the well leakage judgment index value of the first preset window length.

[0129] The well leakage judgment module 304 is used to calculate the deviation between each well leakage judgment index value and the well leakage judgment index base value for each collection point of the second preset window length. If the current well depth is not greater than the first depth, the first well leakage judgment result is obtained according to the relationship between the deviations corresponding to all collection points and the first well leakage judgment threshold. Otherwise, the first well leakage judgment result is obtained according to the relationship between the deviations corresponding to all collection points and the second well leakage judgment threshold. If the first well leakage judgment result indicates that there is a suspected well leakage in the current wellbore, the module determines whether a well leakage warning needs to be issued based on the maximum value of the hook height corresponding to the current well depth and the historical hook height.

[0130] In one embodiment, the well logging data acquisition module is used for:

[0131] When the real-time logging data meets the drilling conditions, the logging data is standardized.

[0132] The standardized logging data is converted from time series to depth series to obtain logging data for analysis.

[0133] In one embodiment, the logging data used for analysis includes one or any combination of time, well depth, drill bit position, instantaneous rotation speed, rotary table rotation speed, drilling pressure, and hook height;

[0134] The indicators for judging well leakage include instantaneous drilling rate, drilling pressure, and rotary table speed.

[0135] In one embodiment, the well leakage detection module is used for:

[0136] If the current well depth is not greater than the first depth, for each well leakage judgment index value at each collection point in the second preset window length, calculate the deviation between the well leakage judgment index value and the well leakage judgment index base value. If the deviations corresponding to the preset number of collection points are all less than the first well leakage judgment threshold, it is determined that the well leakage judgment index value can determine that there is a suspicious well leakage.

[0137] If the current well depth is greater than the first depth, for each well leakage judgment index value, at each collection point within the second preset window length, calculate the deviation between the well leakage judgment index value and the well leakage judgment index base value. If the deviations corresponding to the preset number of collection points are all less than the second well leakage judgment threshold, it is determined that the well leakage judgment index value can be used to judge the existence of a suspicious well leakage.

[0138] If all well leakage judgment index values ​​can determine that there is a suspected well leakage, the first well leakage judgment result is determined to be that there is a suspected well leakage in the current wellbore.

[0139] In one embodiment, if the well leakage judgment index value is the instantaneous drilling rate, the deviation is a first ratio of the instantaneous drilling rate to the instantaneous drilling rate baseline value;

[0140] If the well leakage judgment index is the drilling pressure, the deviation is the second ratio of the drilling pressure to the drilling pressure baseline value;

[0141] If the well leakage judgment index value is the rotary table speed, the deviation is the absolute value of the difference between the rotary table speed and the base value of the rotary table speed.

[0142] In one embodiment, the well leakage detection module is used for:

[0143] Calculate the first difference between the hook height corresponding to the current well depth and the maximum value of the historical hook height;

[0144] When the first difference is less than the difference threshold, it is determined that a well leakage warning needs to be issued.

[0145] Figure 4 This is another structural schematic diagram of the well leakage risk early warning device in an embodiment of the present invention. In one embodiment, the device further includes a well leakage probability acquisition module 401, used for:

[0146] When the first difference is less than the difference threshold, the deviation between the well leakage judgment index value and the well leakage judgment index base value is input into the random forest model to obtain the well leakage probability of the current well. The random forest model is trained based on the historical logging data and well leakage records of the adjacent wells of the current well.

[0147] The well leakage detection module is also used to: determine whether a well leakage warning needs to be issued if the probability of well leakage is greater than the probability threshold.

[0148] In one embodiment, the device further includes a warning information generation module 402, used for:

[0149] After determining that a well leakage warning needs to be issued, the deviations between the well leakage judgment index values ​​and the well leakage judgment index base values ​​of all the nearest neighbor wells in the current wellbore within the current well depth range are extracted from the well leakage feature map of the block, forming a query deviation vector;

[0150] The similarity between the deviation vector to be analyzed, formed by the deviations of all well leakage judgment index values ​​of the current wellbore at the current well depth from the base value of the well leakage judgment index, and the query deviation vector;

[0151] The credibility of the warning is determined based on the similarity.

[0152] Warning information is generated based on the warning time, warning type, current well depth, warning credibility, and warning description.

[0153] The beneficial effects achieved by the method and apparatus proposed in the embodiments of the present invention are as follows:

[0154] The effectiveness of early warning has been significantly enhanced. Traditional well leakage monitoring mainly relies on tracking surface parameters of the drilling fluid circulation system and post-drilling logging interpretation. These methods often suffer from delayed monitoring results and the inability to accurately identify leakage points in real time, leading to the inability to formulate and implement well leakage control measures in advance, thus posing a significant threat to drilling safety. However, the intelligent well leakage risk early warning method based on drill string micro-element venting, by collecting drilling data in real time and analyzing it using artificial intelligence algorithms, can more accurately identify potential leakage points downhole, quickly capture subtle changes in well leakage risk, and achieve early warning of well leakage risk, improving the timeliness and accuracy of early warnings.

[0155] The automation and intelligence levels of risk monitoring have been significantly improved. This invention achieves continuous and automated monitoring of drilling operations by collecting drilling data in real time using an integrated logging instrument and combining it with an intelligent early warning model for data analysis. This intelligent monitoring method not only reduces reliance on manual monitoring and alleviates the workload of operators, but also provides real-time guidance for operators to take optimization measures and respond promptly to well leakage risks, thereby significantly improving drilling operation safety and efficiency.

[0156] This invention also provides a computer device. Figure 5 This is a schematic diagram of a computer device in an embodiment of the present invention. The computer device 500 includes a memory 510, a processor 520, and a computer program 530 stored in the memory 510 and executable on the processor 520. When the processor 520 executes the computer program 530, it implements the above-mentioned well leakage risk warning method.

[0157] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described well leakage risk warning method.

[0158] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-described well leakage risk early warning method.

[0159] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0160] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0161] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0162] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0163] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for early warning of well leakage risk, characterized in that, include: Real-time acquisition of logging data during drilling operations when drill string micro-elements are vented, and time-depth conversion of the logging data to obtain logging data for analysis; Using the latest well depth in the analytical logging data as the starting point, a sliding window analysis is performed on the shallow well section to obtain the well leakage judgment index value of the first preset window length from the analytical logging data. Calculate the base value of the well leakage judgment index based on the well leakage judgment index value of the first preset window length; For each collection point of the second preset window length, calculate the deviation between each well leakage judgment index value and the well leakage judgment index base value. If the current well depth is not greater than the first depth, obtain the first well leakage judgment result based on the relationship between the deviations corresponding to all collection points and the first well leakage judgment threshold. Otherwise, obtain the first well leakage judgment result based on the relationship between the deviations corresponding to all collection points and the second well leakage judgment threshold. If the first well leakage assessment result indicates that there is a suspected well leakage in the current wellbore, a well leakage warning is issued based on the maximum value of the hook height corresponding to the current well depth and the historical hook height.

2. The method according to claim 1, characterized in that, Real-time acquisition of logging data during drill string venting in drilling operations, followed by time-depth conversion of the logging data to obtain analytical logging data, including: When the real-time logging data meets the drilling conditions, the logging data is standardized. The standardized logging data is converted from time series to depth series to obtain logging data for analysis.

3. The method according to claim 1, characterized in that, The logging data used for analysis includes one or any combination of time, well depth, drill bit position, instantaneous rotation speed, rotary table rotation speed, drilling pressure, and hook height; The indicators for judging well leakage include instantaneous drilling rate, drilling pressure, and rotary table speed.

4. The method according to claim 3, characterized in that, For each acquisition point within the second preset window length, calculate the deviation between each well leakage judgment index value and the well leakage judgment index base value. If the current well depth is not greater than the first depth, obtain the first well leakage judgment result based on the relationship between the deviations corresponding to all acquisition points and the first well leakage judgment threshold; otherwise, obtain the first well leakage judgment result based on the relationship between the deviations corresponding to all acquisition points and the second well leakage judgment threshold, including: If the current well depth is not greater than the first depth, for each well leakage judgment index value at each collection point in the second preset window length, calculate the deviation between the well leakage judgment index value and the well leakage judgment index base value. If the deviations corresponding to the preset number of collection points are all less than the first well leakage judgment threshold, it is determined that the well leakage judgment index value can determine that there is a suspicious well leakage. If the current well depth is greater than the first depth, for each well leakage judgment index value, at each collection point within the second preset window length, calculate the deviation between the well leakage judgment index value and the well leakage judgment index base value. If the deviations corresponding to the preset number of collection points are all less than the second well leakage judgment threshold, it is determined that the well leakage judgment index value can be used to judge the existence of a suspicious well leakage. If all well leakage judgment index values ​​can determine that there is a suspected well leakage, the first well leakage judgment result is determined to be that there is a suspected well leakage in the current wellbore.

5. The method according to claim 4, characterized in that, If the well leakage judgment index value is the instantaneous drilling rate, the deviation is the first ratio of the instantaneous drilling rate to the instantaneous drilling rate base value; If the well leakage judgment index is the drilling pressure, the deviation is the second ratio of the drilling pressure to the drilling pressure baseline value; If the well leakage judgment index value is the rotary table speed, the deviation is the absolute value of the difference between the rotary table speed and the base value of the rotary table speed.

6. The method according to claim 1, characterized in that, Based on the current well depth corresponding to the hook height and the historical maximum hook height, determine whether a well leakage warning needs to be issued, including: Calculate the first difference between the hook height corresponding to the current well depth and the maximum value of the historical hook height; When the first difference is less than the difference threshold, it is determined that a well leakage warning needs to be issued.

7. The method according to claim 6, characterized in that, The method further includes: When the first difference is less than the difference threshold, the deviation between the well leakage judgment index value and the well leakage judgment index base value is input into the random forest model to obtain the well leakage probability of the current well. The random forest model is trained based on the historical logging data and well leakage records of the adjacent wells of the current well. If the probability of well leakage is greater than the probability threshold, a well leakage warning needs to be issued.

8. The method according to claim 1, characterized in that, The method further includes: After determining that a well leakage warning needs to be issued, the deviations between the well leakage judgment index values ​​and the well leakage judgment index base values ​​of all the nearest neighbor wells in the current wellbore within the current well depth range are extracted from the well leakage feature map of the block, forming a query deviation vector; The similarity between the deviation vector to be analyzed, formed by the deviations of all well leakage judgment index values ​​of the current wellbore at the current well depth from the base value of the well leakage judgment index, and the query deviation vector; The credibility of the warning is determined based on the similarity. Warning information is generated based on the warning time, warning type, current well depth, warning credibility, and warning description.

9. A well leakage risk early warning device, characterized in that, include: The logging data acquisition module is used to acquire logging data in real time when the drill string micro-element is vented during drilling operations, and to perform time-depth conversion on the logging data to obtain logging data for analysis. The well leakage judgment index value calculation module is used to perform sliding window analysis from the latest well depth in the analysis logging data to the shallow well section, and obtain the well leakage judgment index value of the first preset window length from the analysis logging data. The well leakage judgment index base value calculation module is used to calculate the well leakage judgment index base value based on the well leakage judgment index value of the first preset window length; The well leakage detection module is used to calculate the deviation between each well leakage detection index value and the well leakage detection index base value for each collection point of the second preset window length. If the current well depth is not greater than the first depth, the first well leakage detection result is obtained based on the relationship between the deviations corresponding to all collection points and the first well leakage detection threshold. Otherwise, the first well leakage detection result is obtained based on the relationship between the deviations corresponding to all collection points and the second well leakage detection threshold. If the first well leakage detection result indicates that there is a suspected well leakage in the current wellbore, the module determines whether a well leakage warning needs to be issued based on the maximum value of the hook height corresponding to the current well depth and the historical hook height.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method of any one of claims 1 to 8.

12. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 8.

Images