Shale gas well casing screwing-on state real-time detection method

Abstract

The invention belongs to the technical field of electric digital data processing, and particularly relates to a shale gas well casing screwing-on state real-time detection method. According to the method, a single-feature or multi-feature analysis method is established on the basis of torque change rules of a thread slight interference section ab and a shoulder abutting section bc in the screwing-on process of the shale gas well special thread casing and relative position relations between three torque change key points a, b and c and four threshold values Tmax, Tmin, Tstepmin and Tstepmax; in the casing running construction process, whether the change characteristics of torque key points and lines in an actual make-up torque curve are normal or not is analyzed in real time, if the make-up state is abnormal, it is indicated that shale gas well casing make-up construction does not meet the shaft integrity requirement, make-up needs to be stopped, make-up is conducted again by adopting rectification measures, and the process is completed. Therefore, gas tightness and integrity of a shale gas well drilling shaft are guaranteed.

Description

Technical Field

[0001] This invention belongs to the field of electronic digital data processing technology, specifically relating to a method for real-time detection of the casing fastening status in shale gas wells. Background Technology

[0002] In recent years, with the development of shale gas, the number of shale gas wells has been increasing. Shale gas wells require fracturing before production, and the fracturing process necessitates wellbore pressure buildup, placing higher demands on wellbore integrity. After production begins, wellbore airtightness is also crucial for ensuring safe operation. To improve wellbore airtightness, special threaded casing is widely used in shale gas well drilling. However, the special thread type itself cannot completely guarantee airtightness; it also requires appropriate threading torque. Therefore, the casing threading process is critical to the sealing performance of the special threaded joints and the integrity of the wellbore.

[0003] The American Petroleum Institute (API) and Chinese national standards have set strict torque requirements for threading of casing and tubing of different specifications and steel grades. Excessive threading torque increases the likelihood of thread damage and sticking; insufficient torque leads to slippage and leakage. Therefore, improving the accuracy of threading is crucial for ensuring optimal connection performance of the casing's internal and external threads and guaranteeing the airtightness of shale gas wellbores.

[0004] Currently, most drilling teams are equipped with torque meters to collect torque curves in real time for manual monitoring. However, the torque during casing connection and threading is affected by various factors, and the threading curve can exhibit various abnormal states. Manual torque curve detection is inefficient and easily affected by subjective human factors, often resulting in missed or misjudged abnormal threading states, leading to damage and leakage of shale gas well casing and reducing wellbore integrity. If the gas tightness of a shale gas well fails to meet standards, it will affect subsequent fracturing and production safety, increasing the cost and safety risks of oil and gas extraction. Therefore, real-time monitoring of casing threading torque status is crucial for improving the gas tightness and integrity of shale gas wells.

[0005] Chinese patent document CN120781551A discloses a method for predicting the threading torque of a direct-connection threaded joint in casing. This method uses an empirical formula formed by fitting a structural mechanics model with experimental data to directly calculate the required threading torque value of the direct-connection threaded joint in casing by measuring the structural parameters of the steel pipe foundation. However, this method is only applicable to the construction design stage before casing installation and cannot perform real-time analysis of the actual threading status during construction, nor can it guarantee the airtightness of the casing in shale gas wells.

[0006] Chinese patent document CN120929894A discloses a method for constructing and evaluating a quality assessment model for tubing snap-in. This method utilizes a convolutional neural network to identify the shape of the torque curve to diagnose abnormal tubing snap-in torque curves. However, convolutional neural networks inherently have some errors and require a large number of abnormal snap-in torque curve samples to ensure a certain level of accuracy. Because this method cannot provide sufficient actual abnormal snap-in torque data samples, it uses a data augmentation model to artificially increase the sample size. However, the augmented samples themselves contain errors, and the accumulation and propagation of these errors lead to a high probability of misjudgment. Furthermore, this method primarily identifies abnormal torque curves based on their shape. Some torque curves may appear normal in shape, but if the final snap-in torque exceeds the permissible torque, it is still considered an abnormal torque curve. Such abnormal torque curves cannot be correctly identified based on shape alone, thus reducing wellbore integrity, affecting subsequent fracturing production, increasing production safety risks, and raising oil and gas extraction costs. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a real-time detection method for the casing threading status in shale gas wells. Based on the torque variation patterns of the slight interference segment (ab) and the shoulder-to-top segment (bc) during the special casing threading process in shale gas wells, and the relative positional relationships between the three key torque variation points (a, b, c) and the four thresholds (T_max, T_min, T_step_min, T_step_max), a single-feature or multi-feature analysis method is established. During casing installation, the method analyzes in real-time whether the changes in the key torque points and lines in the actual threading torque curve are normal. If the threading status is abnormal, it indicates that the casing threading construction in the shale gas well does not meet the wellbore integrity requirements, necessitating termination of threading and re-threading with corrective measures to ensure the airtightness and integrity of the shale gas well drilling wellbore.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: A method for real-time detection of casing fastening status in shale gas wells includes the following steps: S1. Obtain the time series dataset ,in, For time-series indexes, For the first The time truncation of each time-series index For the first The upper torque of each time-series index; S2, from time series datasets Identify feature point a in the dataset, and determine the time series dataset based on feature point a. If the corresponding upper clamping torque curve is determined to be in a normal upper clamping state, step S3 is executed; if it is determined to be in an abnormal upper clamping state, the current detection and upper clamping operation is terminated. S3. Determining Time Series Datasets Based on Vector Product The shoulder of the platform is opposite to vertex b, and the upper torque T at point b is determined. b If the threshold condition is met, proceed to step S4; otherwise, end the current detection and deduction operation. S4. Based on the shoulder-to-vertex b determined in step S3, construct the torque dataset for the mild interference segment. And adopt stability The stability of the upper clamping torque in the slightly interfered section is judged. If it is judged to be an abnormal upper clamping state, the current test and upper clamping operation are ended; if it is judged to be a normal upper clamping state, step S5 is executed. S5, from time series datasets Identify point c and determine the upper clamping torque T at point c. c Determine whether the threshold condition is met to identify the deduction status category. Based on the determined deduction status category, decide whether to execute step S6 or end the current detection and deduction operation. S6. Based on the shoulder apex b determined in step S3 and point c identified in step S5, construct a time-series dataset of the buckling torque on the shoulder apex segment. And adopt stability The stability of the upper winding torque of the shoulder section of the upper winding torque curve after step S5 is judged to achieve real-time detection of the upper winding status of the bushing.

[0009] Preferably, in step S2, the time-series dataset is used to... Feature point a is identified in the data, where feature point a is the starting point of the upper torque curve and corresponds to the time series dataset. In the (t1, T1) step, determine whether the feature point a starts from the origin. If T1≠0 or t1≠0, the torque curve is determined to be in an abnormal torque curve state. Based on this single feature, T1≠0 or t1≠0, the abnormal torque curve state type is identified as an incomplete torque curve, and the detection and torque curve operation ends. If the feature point a starts from the origin, i.e., T1=0 and t1=0, it is determined to be in a normal torque curve state, and step S3 is executed.

[0010] Preferably, step S3 includes: S31. Starting from feature point a, traverse the time series dataset. For each satisfied and The index is used to select three consecutive data points (t) in ascending order of the index. i-1 T i-1 ), (t) i T i ), (t) i+1 T i+1 ), and construct vectors and : (1) (2) S32. Calculate the cross product result of vectors and , if | | > C 临界 , then preliminarily determine that the point corresponding to index i is a candidate shoulder pair vertex. The formula is as follows: (3) Among the candidate shoulder pair vertices obtained by traversing in ascending order of the index, select the first point that meets the index condition as the shoulder pair vertex b. The coordinates of point b are (t b , T b ); S33. Determine whether the make-up torque T b of point b meets the threshold condition. If it meets the threshold condition, execute step S4. If it does not meet the threshold condition, end this detection and make-up operation.

[0011] Further preferably, C 临界 described in step S32 is 0.05 - 0.1. Different values result in different recognition accuracies, and the critical value is set according to actual needs.

[0012] Preferably, the threshold condition described in step S33 is T_step_min ≤ T b ≤ T_step_max.

[0013] Preferably, in step S33, if the threshold condition is not met, it is determined as an abnormal make-up state, and the type of abnormal make-up state is identified based on a single feature, and this detection and make-up operation is ended, including: If T b [[ID=5i]]< T_step_min, the type of abnormal make-up state is identified as the shoulder torque being too small by this single feature, and this detection and make-up operation is ended; If T b > T_step_max, the type of abnormal make-up state is identified as the shoulder torque being too large by this single feature, and this detection and make-up operation is ended.

[0014] Preferably, in step S3, if there is no solution for b after traversing the time series data set , the make-up torque curve corresponding to the time series data set is determined as an abnormal make-up state, and the type of abnormal make-up state is identified as no torque step by this single feature, and this detection and make-up operation is ended. <00*0143> Preferably, the smoothness described in step S4 is calculated according to the following formula: (4) where n b is the timing index of the shoulder to the vertex b, and is the recognition accuracy control variable; Further preferably, for step S4, if < 0, the make-up torque on the light interference section is not stable, and it is determined as an abnormal make-up state. From this single feature, the abnormal make-up state type is thread sticking; if ≥0, it is determined as a normal make-up state, and step S5 is executed.

[0016] Further preferably, the is 0.8 - 1.

[0017] Preferably, step S5 includes: S51. Identify point c from the timing data set : (5) where is the make-up torque at point c; S52. Judge whether the make-up torque T at point c c meets the threshold condition: If it meets the threshold condition, that is, T_min ≤ T c ≤ T_max, it is determined as a normal make-up state, and step S6 is executed; If T c > T_max, it is determined as an abnormal make-up state. From this single feature, the abnormal make-up state type is excessive torque in the thread section, and step S6 is executed; If T c < T_min, it is determined as an abnormal make-up state. From this single feature, the abnormal make-up state type is that the sealing surface and the torque shoulder are not in contact, and this detection and make-up operation is ended.

[0018] Preferably, the smoothness described in step S6 is calculated according to the following formula: (6) where n c is the timing index of point c; Preferably, the smoothness described in step S6 is used to judge the smoothness of the make-up torque on the shoulder-to-shoulder section of the make-up torque curve detected in step S5. Specifically: For the make-up torque curve determined as a normal make-up state in step S52, if For the make-up torque curve determined as a normal make-up state in step S52, if If CV2 ≥ 0, it is still considered a normal buckle state, the detection ends, and the buckle operation is completed; if CV2 < 0, the abnormal buckle state type is identified by this single feature as the torque not rising linearly after the inflection point, the detection and buckle operation end. In step S52, because T c The upper torque curve determined by T_max to be in an abnormal upper buckling state, if CV2 < 0, then the abnormal upper buckling state type is identified as thread step surface yielding based on this multi-feature identification; if If the value is ≥0, the abnormal threading status is still classified as excessive torque in the thread segment, and the test ends.

[0019] In the method of this invention, T_step_min <T_step_max<T_min<T_max。

[0020] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention proposes a real-time detection method for casing connection status in shale gas wells. This method is applicable to real-time monitoring of the airtightness and integrity of casing connections during the casing running process in shale gas well drilling. This method addresses the influence of factors such as thread structure, thread grease, connection parameters, connection tools, and sensors on the torque curve during the connection process of special-thread casing in shale gas wells. It mainly analyzes the change characteristics of the torque curve in the slight interference section (ab section) and the shoulder-to-top section (bc section) during the casing connection process in shale gas wells, and realizes real-time detection of connection status. This method can detect more types of connection anomalies and has higher accuracy in detecting abnormal connection status. The construction team can take targeted measures to reconnect according to the type of abnormal connection status until the connection status is normal, end the current casing connection construction, and continue to repeat the connection status detection of the next casing. This ensures the integrity and airtightness of the shale gas wellbore from the construction stage and improves the well construction quality of shale gas wells.

[0021] (2) This invention breaks through the limitations of existing technologies that rely solely on curve shape recognition. It detects the threading state of the slightly interfering section based on the fluctuation, fluctuation amplitude, and torque magnitude of the ab segment of the torque curve; and detects the threading state of the shoulder-top section based on the fluctuation, fluctuation amplitude, maximum torque, and minimum torque of the bc segment of the torque curve. It can identify threading states with abnormal curve shapes and accurately capture hidden anomalies such as normal curve shapes but final threading torque exceeding the permissible range. This solves the problems of high misjudgment probability and missed detection of hidden anomalies in existing methods, and reduces the hidden dangers of thread damage, sticking, slippage, and leakage caused by threading anomalies. The method can realize real-time monitoring and dynamic early warning of the sleeve threading process, breaking through the limitations of traditional methods that rely solely on the final torque value for post-event judgment, and significantly improving the controllability and reliability of threading quality.

[0022] (3) The method of the present invention establishes the relative positional relationship between the torque key point and the multi-level threshold, thereby realizing the accurate identification and classification of the abnormal types of the upper buckle (such as thread interference abnormality, shoulder misalignment abnormality, over-torque / under-torque abnormality, etc.), providing clear rectification direction for on-site construction personnel, and avoiding sleeve damage or construction delay caused by blindly repeating the upper buckle.

[0023] (4) This invention can be directly applied to the casing construction process of shale gas well drilling. It can collect the torque curve of the casing in real time, analyze the key points of torque and the curve change characteristics. Once an abnormal casing state is detected, it can issue an early warning and prompt to terminate the casing. After taking corrective measures, the casing can be re-cased to avoid problems such as casing damage and wellbore leakage caused by the continuous abnormal casing state, thereby reducing rework costs and casing wear. Compared with the traditional method of manually monitoring the torque curve, this method does not require subjective judgment by humans, which greatly improves the detection efficiency, avoids missed judgment and misjudgment caused by human subjective factors, and reduces the risk of subsequent fracturing production obstruction and production safety accidents caused by abnormal casing, thereby reducing the additional costs and unsafe factors of oil and gas extraction. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of an incomplete torque curve of the special casing for shale gas wells in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the curve showing the small torque of the special casing step in the shale gas well in Embodiment 1 of the present invention; Figure 3 This is a schematic diagram of the curve showing excessive torque on the special casing step in a shale gas well according to Embodiment 1 of the present invention; Figure 4 This is a schematic diagram of the torque-free curve of the special casing for shale gas wells in Embodiment 1 of the present invention; Figure 5 This is a schematic diagram of the threaded connection of the special sleeve pipe in the shale gas well in Embodiment 1 of the present invention; Figure 6 This is a schematic diagram of the excessive torque curve of the special threaded section of the shale gas well in Embodiment 1 of the present invention; Figure 7 This is a schematic diagram showing the non-contact curve between the sealing surface of the special casing and the torque shoulder in Embodiment 1 of the present invention. Figure 8 This is a schematic diagram of the torque curve after the inflection point of the special casing in the shale gas well in Embodiment 1 of the present invention, showing that the torque does not increase linearly. Figure 9 This is a schematic diagram of the yield curve of the special threaded stepped surface of the shale gas well in Embodiment 1 of the present invention; Figure 10 This is a schematic diagram of the buckling torque curve on the sleeve in Embodiment 2 of the present invention; Figure 11 This is a schematic diagram of the sleeve clamping torque curve after identifying the shoulder to vertex b in Embodiment 2 of the present invention. Detailed Implementation

[0025] This invention provides a real-time detection method for the casing threading status of shale gas wells. It is based on the torque variation patterns of the slight interference segment (ab) and the shoulder-to-top segment (bc) during the special casing threading process in shale gas wells, as well as the relative positional relationships between three key points (a, b, c) and four thresholds (T_max, T_min, T_step_min, T_step_max). A single-feature or multi-feature analysis method is established to analyze the changes in the torque key points and lines in the actual threading torque curve in real time during casing installation to determine if they are normal. If the threading status is abnormal, it indicates that the casing threading construction of the shale gas well does not meet the wellbore integrity requirements, and corrective measures need to be taken to re-thread the casing to ensure the airtightness and integrity of the shale gas well drilling well.

[0026] In the method of this invention, T_step_min, T_step_max, T_min, and T_max can be obtained from the instruction manual of the sleeve used.

[0027] Example 1 A method for real-time detection of casing fastening status in shale gas wells includes the following steps: S1. Obtain the time series dataset ,in, For time-series indexes, For the first The time truncation of each time-series index For the first The upper torque of each time-series index; S2, from time series datasets Identify feature point a in the dataset, and determine the time series dataset based on feature point a. If the corresponding upper clamping torque curve is determined to be in a normal upper clamping state, step S3 is executed; if it is determined to be in an abnormal upper clamping state, the current detection and upper clamping operation is terminated. Specifically, as follows: from the time series dataset Feature point a is identified in the data, where feature point a is the starting point of the upper torque curve and corresponds to the time series dataset. In the context of (t1, T1), determine whether feature point a starts from the origin. If T1≠0 or t1≠0, then the torque curve is determined to be in an abnormal torque curve state. The abnormal torque curve state type is identified by a single feature, i.e., T1≠0 or t1≠0, as an incomplete torque graph. Figure 1 As shown, the detection and deduction operation ends; if feature point a starts from the origin, i.e., T1=0 and t1=0, it is determined to be a normal deduction state, and step S3 is executed.

[0028] S3. Determining Time Series Datasets Based on Vector Product The shoulder of the platform is opposite to vertex b, and the upper torque T at point b is determined. b If the threshold condition is met, proceed to step S4; otherwise, end the current detection and deduction operation. Specifically, it includes: S31. Starting from feature point a, traverse the time series dataset. For each satisfied and The index is used to select three consecutive data points (t) in ascending order of the index. i-1 T i-1 ), (t) i T i ), (t) i+1 T i+1 ), and construct vectors and : (1) (2) S32, Calculate vectors and Cross product result If | |>C 临界 If the point corresponding to index i is initially determined to be a candidate shoulder pair vertex; (3) Among the candidate shoulder pairs traversed in ascending index order, select the first one that satisfies the index condition. Point is taken as the shoulder opposite vertex b, and the coordinates of point b are (t b T b ); S33. Determine the upper torque T at point b. b If the threshold condition is met, proceed to step S4; otherwise, end the current detection and deduction operation.

[0029] Step S32, C 临界 The value ranges from 0.05 to 0.1, and different values ​​result in different recognition accuracies. The threshold value should be set according to actual needs.

[0030] The threshold condition mentioned in step S33 is T_step_min≤T b ≤ T_step_max.

[0031] T_step_min is the minimum step torque, and T_step_max is the maximum step torque.

[0032] Specifically, in step S33, if the threshold condition is not met, it is determined as an abnormal make-up state, and the type of abnormal make-up state is identified based on a single feature, and this detection and make-up operation are ended, including: If T b < T_step_min, the type of abnormal make-up state is identified as too small step torque based on this single feature, as Figure 2 shown, and this detection and make-up operation are ended; If T b > T_step_max, the type of abnormal make-up state is identified as too large step torque based on this single feature, as Figure 3 shown, and this detection and make-up operation are ended; If there is no solution for b after traversing the time series data set , the make-up torque curve corresponding to the time series data set is determined as an abnormal make-up torque state, and the type of abnormal make-up state is identified as no torque step based on this single feature, and this detection and make-up operation are ended, as Figure 4 shown.

[0033] S4. Construct a make-up torque data set for the slightly interference section based on the shoulder pair vertex b determined in step S3 , and use the smoothness to determine the smoothness of the make-up torque in the slightly interference section. If it is determined as an abnormal make-up state, this detection and make-up operation are ended; if it is determined as a normal make-up state, step S5 is executed; Specifically, the ab section is the slightly interference section of the thread and the sealing surface, and the curve should be slow, smooth and continuous. Construct a make-up torque data set for the slightly interference section . In the actual wellsite data, the torque curve of the slightly interference section of the normal curve, that is, the ab section, is not necessarily a strict straight line. Therefore, the smoothness CV1 is used to evaluate the smoothness of the make-up torque in the slightly interference section. If CV1 < 0, the make-up torque in the slightly interference section is not smooth, which is an abnormal make-up state: The smoothness mentioned above is calculated according to the following formula: (4) where n b is the time series index of the shoulder pair vertex b, is the recognition accuracy control variable; If < 0, the make-up smoothness and smoothness in the slightly interference section are poor, which is an abnormal make-up state, and the type of abnormal make-up state is identified as thread sticking based on a single feature, as Figure 5 shown; If ≥ 0, it is determined as a normal make-up state, and step S5 is executed.

[0034] The is 0.8 - 1.

[0035] S5. Identify point c from the time - series data set and determine whether the make - up torque T c at point c meets the threshold condition to determine the make - up status category. According to the determined make - up status category, judge whether to execute step S6 or end the current inspection and make - up operation; Specifically, step S5 includes: S51. Identify point c from the time - series data set : (5) where is the make - up torque at point c; S52. Judge whether the make - up torque T c at point c meets the threshold condition: If it meets the threshold condition, that is, T_min ≤ T c ≤ T_min, it is determined to be a normal make - up status, and step S6 is executed; If T c > T_max, it is determined to be an abnormal make - up status. The abnormal make - up status type is identified as too high torque in the thread section by this single feature, as Figure 6 shown, and step S6 is executed; If T c [[ID=4*]]< T_min, it is determined to be an abnormal make - up status. The abnormal make - up status type is identified as the sealing surface and torque shoulder not contacting by this single feature, as Figure 7 shown, and the current inspection and make - up operation are ended.

[0036] S6. Based on the shoulder - to - shoulder vertex b determined in step S3 and point c identified in step S5, construct a time - series data set of the make - up torque of the shoulder - to - shoulder section and use the smoothness [[ID=*0]] to determine the smoothness of the make - up torque of the shoulder - to - shoulder section of the make - up torque curve after being judged in step S5, so as to realize the real - time detection of the casing make - up status.

[0037] Specifically, the bc section is the shoulder - to - shoulder section. After the shoulders of the male and female connectors contact, the torque of the shoulder section rises rapidly to reach the maximum torque. The make - up torque of the shoulder - to - shoulder section should rise rapidly and linearly relative to the make - up torque of the slightly interference section. Construct a time - series data set of the make - up torque of the shoulder - to - shoulder section and calculate the rising smoothness CV2 of the make - up torque of the shoulder - to - shoulder section. If CV2 < 0, it means that the make - up torque of the shoulder - to - shoulder section does not rise linearly, which is an abnormal make - up status.

[0038] The smoothness Calculate using the following formula: (6) Where, n c The time-series index for point c; The use of stability The stability of the upper buckling torque at the top section of the shoulder of the upper buckling torque curve detected in step S5 is judged, specifically as follows: For the upper clamping torque curve determined to be in a normal upper clamping state in step S52, if If CV2 ≥ 0, it is still considered a normal buckling state, the detection ends, and the buckling operation is completed; if CV2 < 0, the abnormal buckling state type is identified by this single feature as the torque not increasing linearly after the inflection point, such as... Figure 8 As shown; In step S52, because T c The upper torque curve that determines T_max as an abnormal upper buckling state, if CV2 < 0, then the abnormal upper buckling state type is identified as thread step surface yielding based on this multi-feature identification. Figure 9 As shown; if ≥0, the abnormal threading status type is still excessive torque of the thread segment.

[0039] In the aforementioned real-time detection method for casing connection status in shale gas wells, for cases where an abnormal connection status is determined, the detection and connection operation are terminated based on the connection torque curve. Subsequently, corresponding measures are taken to reconnect the casing based on the determined abnormal connection status type, thereby ensuring the airtightness of the shale gas well casing connection.

[0040] Example 2 During the casing running in a shale gas well, the following data was obtained for one of the casing lines: casing running depth 877.3m, casing steel grade BG140V, outer diameter 139.7mm, length 11.25m, wall thickness 12.34mm, thread type BGT2, and special thread type. Its torque values ​​are: T_max = 27.570kN.m, T_min = 22.550kN.m, T_step_max = 18.452kN.m, and T_step_min = 2.260kN.m.

[0041] S1. Obtain the time series dataset ; The real-time upper clamping torque data collected by the torque meter is shown in Table 1 below, which is a time series dataset composed of upper clamping torque T and upper clamping time t. With the time series dataset The corresponding upper torque curve is as follows Figure 10 As shown: Table 1 Upper clamp torque data .

[0042] S2, from time series datasets Identify feature point a in the dataset, and determine the time series dataset based on feature point a. If the corresponding upper clamping torque curve is determined to be in a normal upper clamping state, step S3 is executed; if it is determined to be in an abnormal upper clamping state, the current detection and upper clamping operation is terminated. Point a is the starting point of the torque curve, corresponding to the time series dataset. In the case of (t1, T1), as shown in Table 1, T1=t1=0, so the upper clamping torque curve is in the normal upper clamping state, and step S3 is executed.

[0043] S3. Determining Time Series Datasets Based on Vector Product The shoulder of the platform is opposite to vertex b, and the upper torque T at point b is determined. b If the threshold condition is met, proceed to step S4; otherwise, end the current detection and deduction operation. Specifically, it includes: S31. Starting from feature point a(t1, T1), traverse the time series dataset. For each satisfied and The index is used to select three consecutive data points (t) in ascending order of the index. i-1 T i-1 ), (t) i T i ), (t) i+1 T i+1 ), and construct vectors and : (1) (2) S32, Calculate vectors and cross product result If | |>C 临界 If the point corresponding to index i is initially determined to be a candidate shoulder pair vertex, the formula is as follows; (3) Among the candidate shoulder pairs traversed in ascending index order, select the first one that satisfies the index condition. Point is taken as the shoulder opposite vertex b, and the coordinates of point b are (t b T b ).

[0044] Calculate sequentially starting from i=2 | The values ​​are shown in the table below: Table 2 | |Value Calculation Results 。

[0045] C 临界 = 0.1。

[0046] From the calculation results, when i = 28, |C| > 0.1, and it is the first point that satisfies the index condition (26), so this point is the opposite vertex of the shoulder, and at this time t b = 6.7098, T b = 4.4174。

[0047] S33. Judge whether the make-up torque T b at point b meets the threshold condition. If it meets the threshold condition, execute step S4. If it does not meet the threshold condition, end this detection and make-up operation; T_step_min < T b < T_step_max, it is determined as the normal make-up state. As Figure 11 shown, execute step S4; S4. According to the opposite vertex b of the shoulder determined in step S3, construct the make-up torque data set for the slightly interfering section , and use the smoothness to judge the smoothness of the make-up torque in the slightly interfering section. If it is determined as an abnormal make-up state, end this detection and make-up operation; if it is determined as a normal make-up state, execute step S5; (4) From the data of the opposite vertex of the shoulder, it can be seen that n b = 28. Intercept the data of the slightly interfering section from the data in Table 1 and calculate its smoothness CV1. After calculation, CV1 = 0.1490. In this embodiment, the value is 0.8.

[0048] Since CV1 > 0 indicates that the make-up torque in this section is stable, the make-up torque curve is determined as the normal make-up state, and execute step S5.

[0049] S5. Identify point c from the time series data set , and judge whether the make-up torque T c at point c meets the threshold condition to determine the make-up state. According to the determined make-up state category, judge whether to execute step S6 or end this detection and make-up operation; S51. Identify point c from the time series data composed of the make-up torque T and the make-up time t.

[0050] (5) S52. Since T c>T_max, the upper torque curve is determined to be an abnormal upper torque state, the type is excessive torque in the thread section, execute step S6.

[0051] S6. Based on the shoulder apex b determined in step S3 and point c identified in step S5, construct a time-series dataset of the buckling torque on the top segment of the shoulder. And adopt stability The stability of the upper winding torque of the shoulder section of the upper winding torque curve after step S5 is judged to achieve real-time detection of the upper winding status of the bushing.

[0052] A time-series dataset of torque deduction on the top section of the shoulder is constructed by extracting data from the torque step point and beyond in Table 1. ,in, =52, calculate the stability of the buckling torque CV2 on the top section of the shoulder; (6) The calculated value is CV2 = -0.0201.

[0053] Since CV2 < 0 and Tc > T_max has been obtained in step S52, the abnormal type of the torque curve is identified as thread step surface yielding by multi-feature identification, and the detection ends.

[0054] Based on the above test results, the possible causes of the abnormal torque curve are: (1) incorrect torque parameter setting; (2) use of incorrect threading grease; (3) malfunction of the threading equipment.

[0055] The solution to this abnormal curve is as follows: Discard the male and female threaded pipes, replace them with new sleeves and couplings, set the parameters correctly, apply the thread grease correctly, and operate correctly until the upper thread torque curve test result is normal, and then complete the installation of the sleeve.

Claims

1. A method for real-time detection of casing fastening status in shale gas wells, characterized in that, Includes the following steps: S1. Obtain the time series dataset ,in, For time-series indexes, For the first The time truncation of each time-series index For the first The upper torque of each time-series index; S2, from time series datasets Identify feature point a in the dataset, and determine the time series dataset based on feature point a. If the corresponding upper clamping torque curve is determined to be in a normal upper clamping state, step S3 is executed; if it is determined to be in an abnormal upper clamping state, the current detection and upper clamping operation is terminated. S3. Determining Time Series Datasets Based on Vector Product The shoulder of the platform is opposite to vertex b, and the upper torque T at point b is determined. b If the threshold condition is met, proceed to step S4; otherwise, end the current detection and deduction operation. S4. Based on the shoulder-to-vertex b determined in step S3, construct the torque dataset for the mild interference segment. And adopt stability The stability of the upper clamping torque in the slightly interfered section is judged. If it is judged to be an abnormal upper clamping state, the current test and upper clamping operation are ended; if it is judged to be a normal upper clamping state, step S5 is executed. S5, from time series datasets Identify point c and determine the upper clamping torque T at point c. c Determine whether the threshold condition is met to identify the deduction status category. Based on the determined deduction status category, decide whether to execute step S6 or end the current detection and deduction operation. S6. Based on the shoulder apex b determined in step S3 and point c identified in step S5, construct a time-series dataset of the buckling torque on the shoulder apex segment. And adopt stability The stability of the upper winding torque of the shoulder section of the upper winding torque curve after step S5 is judged to achieve real-time detection of the upper winding status of the bushing.

2. The method for real-time detection of casing fastening status in shale gas wells according to claim 1, characterized in that, Step S2 involves extracting data from the time-series dataset. Feature point a is identified in the data, where feature point a is the starting point of the upper torque curve and corresponds to the time series dataset. In the (t1, T1) step, determine whether the feature point a starts from the origin. If T1≠0 or t1≠0, the torque curve is determined to be in an abnormal torque curve state. Based on this single feature, T1≠0 or t1≠0, the abnormal torque curve state type is identified as an incomplete torque curve, and the detection and torque curve operation ends. If the feature point a starts from the origin, i.e., T1=0 and t1=0, it is determined to be in a normal torque curve state, and step S3 is executed.

3. The method for real-time detection of casing fastening status in shale gas wells according to claim 1, characterized in that, Step S3 includes: S31. Starting from feature point a, traverse the time series dataset. For each satisfied and The index is used to select three consecutive data points (t) in ascending order of the index. i-1 T i-1 ), (t) i T i ), (t) i+1 T i+1 ), and construct vectors and : （1） （2） S32, Calculate vectors and cross product result If | |>C 临界 Then, the point corresponding to index i is initially determined to be a candidate shoulder pair vertex, as shown in the following formula: （3） Among the candidate shoulder pairs obtained by traversing in ascending index order, select the first one that satisfies the index condition. Point is taken as the shoulder opposite vertex b, and the coordinates of point b are (t b T b ); S33. Determine the upper torque T at point b. b If the threshold condition is met, proceed to step S4; otherwise, end the current detection and deduction operation. The threshold condition is T_step_min≤T b ≤T_step_max.

4. The method for real-time detection of casing fastening status in shale gas wells according to claim 3, characterized in that, Step S32, C 临界 The threshold value is 0.05-0.1; in step S33, if the threshold condition is not met, it is determined to be an abnormal deduction state, and the abnormal deduction state type is identified based on a single feature, ending the current detection and deduction operation, including: If T b < T_step_min, the abnormal make-up state type is identified as the step torque being too small by this single feature, and this detection and make-up operation are ended; If T b >T_step_max, based on this single feature, the abnormal buckling state type is identified as excessive step torque, and the current detection and buckling operation ends.

5. The method for real-time detection of casing fastening status in shale gas wells according to claim 3, characterized in that, In step S3, if the time series dataset is traversed... If there is no solution for b, then the time series dataset will be... The corresponding upper clamping torque curve is determined to be an abnormal upper clamping state. Based on this single feature, the abnormal upper clamping state type is identified as no torque step, and the detection and upper clamping operation are terminated.

6. The method for real-time detection of casing fastening status in shale gas wells according to claim 1, characterized in that, The stability described in step S4 Calculate using the following formula: （4） Where, n b This is the temporal index of the shoulder to vertex b. To identify the accuracy control variables.

7. The method for real-time detection of casing fastening status in shale gas wells according to claim 6, characterized in that, In step S4, if If the torque is less than 0, the torque on the threaded section is unstable due to slight interference, which is determined to be an abnormal threading state. Therefore, this single feature identifies the abnormal threading state as thread sticking. If the value is ≥0, it is determined to be a normal buckle state, and step S5 is executed; It is 0.8-1.

8. The method for real-time detection of casing fastening status in shale gas wells according to claim 1, characterized in that, Step S5 includes: S51, From time series datasets Identify point c in the middle: （5） in, The torque at point c is the torque applied during the upward movement. S52. Determine the torque T at point c. c Does the threshold condition meet? If the threshold condition is met, i.e., T_min ≤ T c If the value is less than or equal to T_max, then it is determined to be a normal deduction state, and step S6 is executed; If T c If T_max is found, it is determined to be an abnormal threading state. The abnormal threading state type is identified by this single feature as excessive torque in the thread segment, and step S6 is executed. If T c < is less than T_min, it is determined as an abnormal make-up state. The type of the abnormal make-up state is identified as the sealing surface and the torque shoulder not contacting by this single feature, and this inspection and make-up operation are ended.

9. The method for real-time detection of casing fastening status in shale gas wells according to claim 8, characterized in that, The stability described in step S6 Calculate using the following formula: （6） Where, n c This is the time sequence index for point c.

10. The method for real-time detection of casing fastening status in shale gas wells according to claim 9, characterized in that, Step S6 describes the use of stability. The stability of the upper buckling torque at the top section of the shoulder of the upper buckling torque curve detected in step S5 is judged, specifically as follows: For the upper clamping torque curve determined to be in a normal upper clamping state in step S52, if If CV2 ≥ 0, it is still considered a normal buckle state, the detection ends, and the buckle operation is completed; if CV2 < 0, the abnormal buckle state type is identified by this single feature as the torque not rising linearly after the inflection point, the detection and buckle operation end. In step S52, because T c The upper torque curve determined by T_max to be in an abnormal upper buckling state, if CV2 < 0, then the abnormal upper buckling state type is identified as thread step surface yielding based on this multi-feature identification; if If the value is ≥0, the abnormal threading status is still classified as excessive torque in the thread segment, and the test ends.

Images