Drill jamming identification method and device for well drilling, electronic equipment, storage medium and program product

Abstract

The embodiment of the invention provides a drilling jamming identification method and device for well drilling, electronic equipment, a storage medium and a program product, and relates to the technical field of well drilling engineering. The method comprises the following steps: acquiring drilling data, and acquiring a hook load actual value and a torque actual value at the current moment in the drilling data; and according to the drilling data, the underground drilling state is determined. And determining a preset hook load prediction model and a torque prediction model corresponding to the drilling state, and respectively calling the hook load prediction model and the torque prediction model to perform drilling prediction on the drilling data to obtain a hook load prediction value and a torque prediction value at the current moment. According to the hook load actual value, the torque actual value, the hook load predicted value and the torque predicted value, an underground drill jamming index is determined; wherein the drill jamming index represents whether drill jamming occurs underground or not. The method achieves the effect of improving the drilling jamming recognition accuracy of well drilling.

Description

Technical Field

[0001] This application relates to the field of drilling engineering technology, and in particular to a method, device, electronic equipment, storage medium and program product for identifying stuck drill bits in drilling. Background Technology

[0002] Currently, stuck pipe is one of the most common downhole drilling accidents. It not only severely impacts drilling speed and well construction quality, but also, if not handled promptly or properly, can lead to other serious accidents such as drill string breakage, well collapse, and blowouts. Based on the cause, stuck pipe can be categorized into differential pressure stuck pipe, reduced diameter stuck pipe, well collapse stuck pipe, keyway stuck pipe, sand stuck pipe, and object stuck pipe. If one type of stuck pipe is not addressed promptly, it may trigger another type, creating a compound stuck pipe accident that significantly increases the difficulty of handling. Therefore, timely and accurate identification of stuck pipe signs and appropriate handling in the early stages of stuck pipe are of great significance for improving drilling safety and efficiency.

[0003] In existing technologies, most stuck drill bit identification and prediction techniques employ physical model prediction and single-parameter or multi-parameter threshold setting methods. However, due to the complexity of downhole conditions, physical models struggle to accurately describe and predict the stress on downhole tools. Furthermore, when input data is incomplete or inaccurate, the calculation results of the physical model can deviate significantly from the actual situation. Single-parameter and multi-parameter threshold setting methods generally suffer from high false alarm and false negative rates, resulting in low accuracy in existing stuck drill bit identification methods. Summary of the Invention

[0004] This application provides a method, apparatus, electronic device, storage medium, and program product for identifying stuck drill pipe in drilling, in order to improve the accuracy of stuck drill pipe identification in drilling.

[0005] In a first aspect, embodiments of this application provide a method for identifying stuck drill pipe in wells, including:

[0006] Acquire drilling data, and acquire the actual values ​​of the hook load and torque at the current moment from the drilling data; wherein, the drilling data is real-time data generated by the drilling operation;

[0007] Based on the drilling data, determine the downhole drilling status;

[0008] Determine the preset hook load prediction model and torque prediction model corresponding to the drilling state, and call the hook load prediction model and torque prediction model respectively to perform drilling prediction on the drilling data to obtain the hook load prediction value and torque prediction value at the current moment.

[0009] The stuck pipe index is determined based on the actual value of the hook load, the actual value of the torque, the predicted value of the hook load, and the predicted value of the torque; wherein, the stuck pipe index indicates whether stuck pipe will occur downhole.

[0010] In one possible implementation, determining the sticking index downhole based on the actual hook load value, the actual torque value, the predicted hook load value, and the predicted torque value includes:

[0011] Based on the actual values ​​of the hook load and torque, determine the actual changing trends of the hook load and torque; and based on the predicted values ​​of the hook load and torque, determine the predicted changing trends of the hook load and torque.

[0012] The stuck pipe index in the well is determined based on the actual and predicted trends.

[0013] In one possible implementation, determining the actual changing trends of the hook load and torque based on the actual values ​​of the hook load and torque, and determining the predicted changing trends of the hook load and torque based on the predicted values ​​of the hook load and torque, includes:

[0014] Obtain historical time series of hook load and torque;

[0015] Based on the preset adjacent sliding window average difference algorithm, the actual change trend of the hook load and the torque is determined by the actual value of the hook load, the actual value of the torque, the historical hook load time series, and the torque time series.

[0016] Based on the adjacent sliding window average difference algorithm, the predicted change trends of the hook load and the torque are determined by the hook load prediction value, the torque prediction value, the historical hook load time series, and the torque time series.

[0017] In one possible implementation, determining the downhole stuck index based on the actual trend and the predicted trend includes:

[0018] The actual trend of change is compared with the predicted trend of change. If it is determined that the actual trend of change and the predicted trend of change are inconsistent, the change in hook load between the actual value of hook load and the predicted value of hook load, and the change in torque between the actual value of torque and the predicted value of torque are calculated.

[0019] The stuck pipe index in the well is determined based on the change in hook load and the change in torque.

[0020] In one possible implementation, the drilling condition includes multiple types; determining the hook load prediction model and torque prediction model corresponding to the drilling condition includes:

[0021] For each type of drilling condition, determine the hook load prediction model and / or torque prediction model corresponding to the drilling condition;

[0022] The drilling state includes any one or more of the following: drilling, tripping, running down, reaming, and reverse reaming.

[0023] In one possible implementation, the method further includes:

[0024] Obtain datasets for each drilling state;

[0025] For each dataset under each drilling condition, the value of the target parameter is determined based on the dataset.

[0026] A training set for predicting the load of a large hook is generated based on the values ​​of the target parameters and the values ​​of the preset parameters in the dataset used to generate the large hook load prediction model.

[0027] A torque prediction training set is generated based on the values ​​of the target parameters and the values ​​of the preset parameters used to generate the torque prediction model in the dataset.

[0028] The hook load prediction model is trained based on the hook load prediction training set until it converges; and the torque prediction model is trained based on the torque prediction training set until it converges.

[0029] Secondly, embodiments of this application provide a stuck drill bit identification device, comprising:

[0030] The acquisition module is used to acquire drilling data, and to acquire the actual values ​​of the hook load and torque at the current moment from the drilling data; wherein, the drilling data is real-time data generated by the drilling operation;

[0031] The first determining module is used to determine the downhole drilling status based on the drilling data;

[0032] The second determining module is used to determine the preset hook load prediction model and torque prediction model corresponding to the drilling state;

[0033] The prediction module is used to call the hook load prediction model and torque prediction model respectively to perform drilling prediction on the drilling data, and obtain the hook load prediction value and torque prediction value at the current moment.

[0034] The third determining module is used to determine the stuck pipe index in the well based on the actual value of the hook load, the actual value of the torque, the predicted value of the hook load, and the predicted value of the torque; wherein, the stuck pipe index indicates whether stuck pipe will occur in the well.

[0035] In one possible implementation, the third determining module includes:

[0036] The first determining unit is used to determine the actual changing trend of the hook load and torque based on the actual value of the hook load and the actual value of the torque; and to determine the predicted changing trend of the hook load and the torque based on the predicted value of the hook load and the predicted value of the torque.

[0037] The second determining unit is used to determine the stuck pipe index downhole based on the actual change trend and the predicted change trend.

[0038] In one possible implementation, the first determining unit is specifically used for:

[0039] Obtain historical time series of hook load and torque;

[0040] Based on the preset adjacent sliding window average difference algorithm, the actual change trend of the hook load and the torque is determined by the actual value of the hook load, the actual value of the torque, the historical hook load time series, and the torque time series.

[0041] Based on the adjacent sliding window average difference algorithm, the predicted change trends of the hook load and the torque are determined by the hook load prediction value, the torque prediction value, the historical hook load time series, and the torque time series.

[0042] In one possible implementation, the second determining unit is specifically used for:

[0043] The actual trend of change is compared with the predicted trend of change. If it is determined that the actual trend of change and the predicted trend of change are inconsistent, the change in hook load between the actual value of hook load and the predicted value of hook load, and the change in torque between the actual value of torque and the predicted value of torque are calculated.

[0044] The stuck pipe index in the well is determined based on the change in hook load and the change in torque.

[0045] In one possible implementation, the drilling state includes multiple types; the second determining module is specifically used for:

[0046] For each type of drilling condition, determine the hook load prediction model and / or torque prediction model corresponding to the drilling condition;

[0047] The drilling state includes any one or more of the following: drilling, tripping, running down, reaming, and reverse reaming.

[0048] In one possible implementation, the device is further specifically used for:

[0049] Obtain datasets for each drilling state;

[0050] For each dataset under each drilling condition, the value of the target parameter is determined based on the dataset.

[0051] A training set for predicting the load of a large hook is generated based on the values ​​of the target parameters and the values ​​of the preset parameters in the dataset used to generate the large hook load prediction model.

[0052] A torque prediction training set is generated based on the values ​​of the target parameters and the values ​​of the preset parameters used to generate the torque prediction model in the dataset.

[0053] The hook load prediction model is trained based on the hook load prediction training set until it converges; and the torque prediction model is trained based on the torque prediction training set until it converges.

[0054] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0055] The memory stores computer-executed instructions;

[0056] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0057] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0058] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0059] The drilling stuck pipe identification method, device, electronic equipment, storage medium, and program product provided in this application acquire drilling data, and obtain the actual values ​​of hook load and torque at the current moment from the drilling data; wherein, the drilling data is real-time data generated by drilling operations. Based on the drilling data, the downhole drilling state is determined. A preset hook load prediction model and torque prediction model corresponding to the drilling state are determined, and the hook load prediction model and torque prediction model are respectively called to perform drilling prediction on the drilling data, obtaining the predicted values ​​of hook load and torque at the current moment. Based on the actual values ​​of hook load, actual torque, predicted hook load, and predicted torque, the downhole stuck pipe index is determined; wherein, the stuck pipe index characterizes whether stuck pipe will occur downhole. In this scheme, drilling prediction is performed on the drilling data using the hook load prediction model and torque prediction model to obtain the predicted values ​​of hook load and torque at the current moment, and the stuck pipe index is comprehensively determined based on the predicted hook load and torque values ​​and the obtained actual values ​​of hook load and torque. Therefore, compared with existing technologies, this application solves the problem that traditional single-parameter and multi-parameter pre-set threshold methods are difficult to universally apply to stuck pipe identification under different working conditions, well types, and stuck pipe types, resulting in high false alarm and false alarm rates. It also solves the problem that current physical models and purely data-driven models require detailed drilling dynamic and static parameters as input or require the construction of complete, high-quality stuck pipe sample data for model training. By constructing independent intelligent prediction models for hook load and torque under different drilling conditions, drilling engineering knowledge and machine learning algorithms are fully integrated to automatically identify stuck pipe, avoiding the subjectivity and lag of manual monitoring, and eliminating the cumbersome data collection and input and the construction of a large amount of stuck pipe sample data. This is of great significance for improving the accuracy of stuck pipe identification and the degree of automation and intelligence, effectively improving the accuracy and timeliness of stuck pipe prediction, thereby improving the accuracy of stuck pipe identification in drilling. Attached Figure Description

[0060] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0061] Figure 1 A flowchart illustrating a stuck pipe identification method for drilling provided in this application embodiment. Figure 1 ;

[0062] Figure 2 A flowchart illustrating another drilling stuck pipe identification method provided in this application embodiment. Figure 2 ;

[0063] Figure 3 A flowchart illustrating a stuck pipe identification method for drilling provided in this application. Figure 3 ;

[0064] Figure 4 This is a schematic diagram of a stuck drill identification device provided in an embodiment of this application;

[0065] Figure 5 A schematic diagram of another stuck drill identification device provided in this application embodiment;

[0066] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0067] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0068] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0069] Currently, stuck pipe is one of the most common downhole drilling accidents. It not only severely impacts drilling speed and well construction quality, but also, if not handled promptly or properly, can lead to other serious accidents such as drill string breakage, well collapse, and blowouts. Based on the cause, stuck pipe can be categorized into differential pressure stuck pipe, reduced diameter stuck pipe, well collapse stuck pipe, keyway stuck pipe, sand stuck pipe, and object stuck pipe. If one type of stuck pipe is not addressed promptly, it may trigger another type, creating a compound stuck pipe accident that significantly increases the difficulty of handling. Therefore, timely and accurate identification of stuck pipe signs and appropriate handling in the early stages of stuck pipe are of great significance for improving drilling safety and efficiency.

[0070] In one example, current stuck pipe identification and prediction technologies mostly employ physical model prediction and single-parameter or multi-parameter threshold setting methods. However, due to the complexity of downhole conditions, physical models are difficult to accurately describe and predict the stress on downhole tools. Moreover, when the input data is incomplete or inaccurate, the calculation results of the physical model will deviate significantly from the actual situation. Single-parameter and multi-parameter threshold setting methods generally suffer from high false alarm and false negative rates, resulting in low accuracy of existing stuck pipe identification methods.

[0071] Based on the above scenarios, it can be seen that the existing technology has a technical problem of low accuracy in identifying stuck drill bits.

[0072] The stuck pipe identification method for drilling provided in this application effectively improves the accuracy and timeliness of stuck pipe prediction, thereby improving the accuracy of stuck pipe identification in drilling.

[0073] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0074] Figure 1 This application provides a flowchart of a method for identifying stuck pipe in drilling. Figure 1 ,like Figure 1 The method includes:

[0075] S101 acquires drilling data, and acquires the actual values ​​of the hook load and torque at the current moment from the drilling data; where the drilling data is real-time data generated by the drilling operation.

[0076] For example, the executing entity of this embodiment can be an electronic device, a terminal device, a stuck drill identification device or device, or other device or device capable of executing this embodiment, and there is no limitation thereto. In this embodiment, the executing entity is described as an electronic device.

[0077] First, drilling data is acquired in real time during drilling operations. This data includes well depth, bit depth, rotary table speed, riser pressure, outlet flow rate, hook height, etc., without limitation. The actual hook load and torque values ​​at the current moment are then extracted from the drilling data. Specifically, drilling data (i.e., comprehensive logging data) can be retrieved from real-time data servers and historical data files, and this comprehensive logging data is cleaned. The cleaning algorithm includes processing duplicate data, missing data, outlier data, and data exceeding a threshold range. The threshold settings are based on the physical meaning of the parameters, such as well depth > bit depth, bit depth > 0, hook load < the limit value specified by the drilling rig performance, etc., without limitation. If the bit position is missing in the drilling data, linear interpolation can be used to fill the gaps, resulting in filled drilling data.

[0078] S102. Determine the downhole drilling status based on drilling data.

[0079] For example, based on drilling data, the drilling status can be automatically identified. The identified drilling status includes: drilling, tripping, running in, reaming, and reverse reaming. The drilling status identification adopts a method based on preset rules: if the drill bit is at the bottom of the well, the pump is on, and the drill bit is rotating, it is determined to be drilling; if the drill bit is moving upward and the drill string is not rotating, it is determined to be tripping; if the drill bit is moving downward and the drill string is not rotating, it is determined to be running in; if the drill bit is moving upward and the drill string is rotating, it is determined to be reverse reaming; if the drill bit is moving downward and the drill string is rotating, it is determined to be reaming.

[0080] S103. Determine the preset hook load prediction model and torque prediction model corresponding to the drilling state, and call the hook load prediction model and torque prediction model respectively to perform drilling prediction on the drilling data to obtain the hook load prediction value and torque prediction value at the current moment.

[0081] For example, the hook load prediction model is divided into four independent models: a trip hook load prediction model, a run hook load prediction model, a reaming hook load prediction model, and a reverse reaming hook load prediction model. The torque prediction model is divided into three independent models: a drilling torque prediction model, a reaming hook load prediction model, and a reverse reaming hook load prediction model. The obtained drilling states include drilling, tripping, running, reaming, and reverse reaming. Each type of drilling state corresponds to a hook load prediction model, and each type of drilling state corresponds to a torque prediction model. It is necessary to determine the preset hook load prediction model and / or torque prediction model corresponding to each type of state. Then, the hook load prediction model is called to perform drilling prediction on the drilling data to obtain the hook load prediction value at the current moment, and the torque prediction model is called to perform drilling prediction on the drilling data to obtain the torque prediction value at the current moment.

[0082] For example, the drilling state corresponds to a drilling torque prediction model with a torque prediction model; the tripping state corresponds to a tripping hook load prediction model with a hook load prediction model and a torque prediction model with a hook load prediction model; the running-in state corresponds to a running-in hook load prediction model with a hook load prediction model; the reaming state corresponds to a reaming hook load prediction model with a hook load prediction model and a torque prediction model with a reaming start model; and the reverse reaming state corresponds to a reverse reaming hook load prediction model with a hook load prediction model and a torque prediction model with a reverse reaming hook load prediction model. Then, each model is called to perform drilling prediction on the drilling data to obtain the hook load prediction value and torque prediction value at the current moment.

[0083] S104. Determine the sticking index downhole based on the actual value of hook load, the actual value of torque, the predicted value of hook load, and the predicted value of torque; whereby the sticking index indicates whether sticking will occur downhole.

[0084] For example, based on the actual values ​​of the hook load and torque, the actual trends of both the hook load and torque are determined, and based on the predicted values ​​of the hook load and torque, the predicted trends of both the hook load and torque are determined. Based on the actual and predicted trends, a sticking index is determined. The sticking index indicates whether sticking will occur downhole; for example, if the sticking index is greater than a preset sticking threshold, it indicates that sticking will occur downhole.

[0085] The stuck pipe identification method for drilling provided in this application embodiment acquires drilling data, and obtains the actual values ​​of hook load and torque at the current moment from the drilling data; wherein, the drilling data is real-time data generated by drilling operations. Based on the drilling data, the downhole drilling state is determined. A preset hook load prediction model and torque prediction model corresponding to the drilling state are determined, and the hook load prediction model and torque prediction model are respectively called to perform drilling prediction on the drilling data to obtain the predicted values ​​of hook load and torque at the current moment. Based on the actual values ​​of hook load, actual torque, predicted hook load, and predicted torque, the downhole stuck pipe index is determined; wherein, the stuck pipe index characterizes whether stuck pipe will occur downhole. In this scheme, drilling prediction is performed on the drilling data using the hook load prediction model and torque prediction model to obtain the predicted values ​​of hook load and torque at the current moment, and the stuck pipe index is comprehensively determined based on the predicted hook load and torque values ​​and the obtained actual values ​​of hook load and torque. Therefore, compared with existing technologies, this application solves the problem that traditional single-parameter and multi-parameter pre-set threshold methods are difficult to universally apply to stuck pipe identification under different working conditions, well types, and stuck pipe types, resulting in high false alarm and false alarm rates. It also solves the problem that current physical models and purely data-driven models require detailed drilling dynamic and static parameters as input or require the construction of complete, high-quality stuck pipe sample data for model training. By constructing independent intelligent prediction models for hook load and torque under different drilling conditions, drilling engineering knowledge and machine learning algorithms are fully integrated to automatically identify stuck pipe, avoiding the subjectivity and lag of manual monitoring, and eliminating the cumbersome data collection and input and the construction of a large amount of stuck pipe sample data. This is of great significance for improving the accuracy of stuck pipe identification and the degree of automation and intelligence, effectively improving the accuracy and timeliness of stuck pipe prediction, thereby improving the accuracy of stuck pipe identification in drilling.

[0086] Figure 2 A flowchart illustrating a stuck pipe identification method for drilling provided in this application. Figure 2 ,like Figure 2 As shown, in this embodiment... Figure 1 Based on the examples, a method for identifying stuck pipe in drilling is described in detail. This method includes:

[0087] S201. Obtain the dataset for each drilling state.

[0088] For example, real-time drilling data is acquired, namely, the real-time drilling dataset, which includes hook load data and torque data. Based on the real-time drilling dataset, the drilling status is calculated, and the real-time drilling dataset is divided into five categories according to the drilling status: drilling dataset, tripping dataset, running-in dataset, reaming dataset, and reverse reaming dataset.

[0089] S202. For each drilling condition dataset, determine the value of the target parameter based on the dataset.

[0090] For example, to form training sets for the hook load prediction model and torque prediction model, feature engineering is performed. Feature engineering involves calculating the values ​​of target parameters based on the data in each dataset. Specifically, target parameters include tripping speed, reaming speed, and single-joint / unjoint identification results. Feature engineering includes calculating tripping speed and reaming speed; calculating single-joint / unjoint identification results; and marking the number of single joints currently added / removed based on these results. For instance, for the hook load prediction model, the drill bit movement speed is calculated using the collected drill bit position and time, and the number of single joints added / removed is calculated using the calculated drilling state; for the torque prediction model, the drill bit movement speed is calculated using multi-dimensional data such as drill bit depth and time.

[0091] S203. Generate a hook load prediction training set based on the values ​​of the target parameters and the values ​​of the preset parameters in the dataset used to generate the hook load prediction model.

[0092] For example, for the hook load prediction model, based on the values ​​of the target parameters, the physical meaning of the measured parameters in the dataset, and the correlation analysis between the parameters, the inlet density, rotary table speed, riser pressure, outlet flow rate, drill bit movement speed, and single-joint increase / decrease number are preferred as inputs to the hook load prediction model, and a hook load prediction training set is established under the conditions of pulling up, pulling down, reaming, and reverse reaming.

[0093] S204. Generate a torque prediction training set based on the values ​​of the target parameters and the values ​​of the preset parameters used to generate the torque prediction model in the dataset.

[0094] For example, for the torque prediction model, based on the values ​​of the target parameters, the physical meaning of the measured parameters in the dataset, and the correlation analysis between the parameters, the drilling pressure, rotary table speed, riser pressure, and drill bit speed are preferred as inputs to the torque prediction model, and a torque prediction training set is established under drilling, reaming, and reverse reaming states.

[0095] S205. Train the hook load prediction model based on the hook load prediction training set until the hook load prediction model converges; and train the torque prediction model based on the torque prediction training set until the torque prediction model converges.

[0096] For example, a support vector machine (SVM) prediction algorithm, specifically a SVM regression algorithm, is preferred. Based on the established training sets for hook load prediction and torque prediction, hook load prediction models and torque prediction models are trained respectively. The model parameters are then tuned and optimized according to the prediction accuracy until convergence, resulting in the final hook load prediction model and torque prediction model. The hook load prediction model consists of four independent models: a trip hook load prediction model, a down-drilling hook load prediction model, a reaming hook load prediction model, and a reverse reaming hook load prediction model. The torque prediction model consists of three independent models: a drilling torque prediction model, a reaming hook load prediction model, and a reverse reaming hook load prediction model.

[0097] S206. Obtain drilling data, and obtain the actual value of the hook load and the actual value of the torque at the current moment from the drilling data; wherein, the drilling data is the real-time data generated by the drilling operation.

[0098] For example, this step can be referred to Figure 1 Step 101 in the text will not be repeated here.

[0099] S207. Determine the downhole drilling status based on drilling data.

[0100] For example, this step can be referred to Figure 1 Step 102 in the text will not be repeated here.

[0101] S208. Drilling conditions include multiple types; for each type of drilling condition, determine the hook load prediction model and / or torque prediction model corresponding to the drilling condition; wherein, drilling conditions include any one or more of the following: drilling, tripping, running down, reaming, and reverse reaming.

[0102] For example, the hook load prediction model is divided into four independent models: a trip hook load prediction model, a run hook load prediction model, a reaming hook load prediction model, and a reverse reaming hook load prediction model. The torque prediction model is divided into three independent models: a drilling torque prediction model, a reaming hook load prediction model, and a reverse reaming hook load prediction model. The obtained drilling states include drilling, tripping, run-in, reaming, and reverse reaming. Each type of drilling state corresponds to one hook load prediction model, and each type of drilling state corresponds to one torque prediction model. It is necessary to determine the preset hook load prediction model and / or torque prediction model corresponding to each type of state.

[0103] For example, the drilling state corresponds to the drilling torque prediction model with a torque prediction model; the tripping state corresponds to the tripping hook load prediction model with a hook load prediction model and the tripping hook load prediction model with a torque prediction model; the running-in state corresponds to the running-in hook load prediction model with a hook load prediction model; the reaming state corresponds to the reaming hook load prediction model with a hook load prediction model and the reaming hook load prediction model with a torque prediction model; and the reverse reaming state corresponds to the reverse reaming hook load prediction model with a hook load prediction model and the reverse reaming hook load prediction model with a torque prediction model.

[0104] S209. Call the hook load prediction model and torque prediction model respectively to perform drilling prediction on the drilling data, and obtain the hook load prediction value and torque prediction value at the current moment.

[0105] For example, the electronic device can call the hook load prediction model to perform drilling prediction on the drilling data to obtain the hook load prediction value at the current moment, and call the torque prediction model to perform drilling prediction on the drilling data to obtain the torque prediction value at the current moment.

[0106] S210. Based on the actual values ​​of the hook load and torque, determine the actual trends of the hook load and torque; and based on the predicted values ​​of the hook load and torque, determine the predicted trends of the hook load and torque.

[0107] In one example, step S210 includes: obtaining historical hook load time series and torque time series; determining the actual changing trends of hook load and torque based on the actual value of hook load, the actual value of torque, the historical hook load time series, and the torque time series using a preset adjacent sliding window average difference algorithm; and determining the predicted changing trends of hook load and torque based on the adjacent sliding window average difference algorithm using the predicted value of hook load, the predicted value of torque, the historical hook load time series, and the torque time series.

[0108] For example, historical hook load time series and torque time series are obtained. Then, based on a preset adjacent sliding window average difference algorithm, the actual changing trends of hook load and torque are determined using the actual hook load value, actual torque value, historical hook load time series, and torque time series. Furthermore, based on the adjacent sliding window average difference algorithm, the predicted changing trends of hook load and torque are determined using the predicted hook load value, predicted torque value, historical hook load time series, and torque time series. Specifically, the changing trends include: sudden increase, gradual increase, no change, gradual decrease, and sudden decrease.

[0109] S211. Determine the stuck pipe index downhole based on the actual and predicted trends.

[0110] In one example, step S211 includes: comparing the actual trend of change with the predicted trend of change; if it is determined that the actual trend of change and the predicted trend of change are inconsistent, calculating the change in hook load between the actual value of hook load and the predicted value of hook load, and the change in torque between the actual value of torque and the predicted value of torque; and determining the stuck pipe index downhole based on the change in hook load and the change in torque.

[0111] For example, the electronic device can overlay and compare the actual trend and the predicted trend. If it is determined that the actual trend and the predicted trend are inconsistent, the device calculates the change in hook load between the actual value and the predicted value, and the change in torque between the actual value and the predicted value, based on the deviation of the hook load and torque change trends. Based on the change in hook load, the change in torque, and the preset sticking probability calculation formula, the sticking index in the well is calculated.

[0112] For example, by applying the moving window averaging algorithm, the predicted and actual trends of hook load and torque for the two time periods of 14:27 and 14:26 on July 9, 2024 can be calculated. Table 1 below shows the actual, predicted, actual, and predicted hook load values, as well as the actual and predicted torque values ​​for each time period:

[0113] Table 1

[0114]

[0115] Calculate the Possibility of Pipe Stuck using the following formula:

[0116] Possibility of Pipe Stuck

[0117] =Min(2,Abs(VariationOfHKLD) measurment

[0118] -VariationOfHKLD predict ) / TH hkld )*W hkld

[0119] +Min(2,Abs(VariationOfTORQ measurment

[0120] -VariationOfTROQ predict ) / TH torq )*W torq

[0121] Among them, THhkld TH represents the threshold value for the hook load. torq This represents the torque threshold; Abs is the absolute value within the parentheses; VariationOfHKLD measurment The variation of the average value of adjacent sliding windows in the measured sequence of hook loads; VariationOfHKLD predict W represents the change in the average value of adjacent sliding windows in the large hook load prediction sequence. hkld Weights for the hook load factor used to calculate the likelihood of stuck drill bit; VariationOfTORQ measurment VariationOfTROQ is the change in the average value of adjacent sliding windows in the measured torque sequence. predict W represents the change in the average value of adjacent sliding windows in the torque prediction sequence. torq The torque factor weights are used to calculate the likelihood of a stuck drill.

[0122] Take TH hkld =1102.62 * 0.02 = 22.054 kN;

[0123] TH torq =14.54 * 0.1 = 1.45 kN;

[0124] The stuck drill index for hook load can be calculated:

[0125] Abs(VariationOfHKLD measurment -VariationOfHKLD predict ) / TH hkld =(15.25-0.12)

[0126] / 22.054=0.686;

[0127] Calculate the likelihood of a stuck drill bit based on torque:

[0128] Abs(VariationOfTORQ measurment -VariationOfTROQ predict ) / TH torq =(2.27-0.47)

[0129] / 1.45=1.24;

[0130] Take W hkld =0.5, W torq =0.5, calculate the comprehensive jamming index:

[0131] The Possibility of Pipe Stuck is calculated as Min(2, 0.686)*0.5 + Min(2, 1.24)*0.5 = 0.34 + 0.62 = 0.96, indicating a high risk of pipe sticking at the current moment. Therefore, by utilizing machine learning algorithms and time-series data processing techniques to fully extract useful information from real-time surface measurements, and comprehensively analyzing key parameters of pipe sticking, automatic early identification of pipe sticking signs can be achieved. This avoids tedious data collection and input work, eliminates the need to construct labeled pipe sticking accident datasets, reduces manpower and material resources, and is more convenient.

[0132] The stuck pipe identification method for drilling provided in this application embodiment acquires datasets for various drilling states. For each dataset under each drilling state, the values ​​of target parameters are determined based on the dataset. A hook load prediction training set is generated based on the values ​​of the target parameters and the values ​​of preset parameters used to generate a hook load prediction model in the dataset. A torque prediction training set is generated based on the values ​​of the target parameters and the values ​​of preset parameters used to generate a torque prediction model in the dataset. The hook load prediction model is trained using the hook load prediction training set until it converges; and the torque prediction model is trained using the torque prediction training set until it converges. Drilling data is acquired, and the actual values ​​of the hook load and torque at the current moment are acquired from the drilling data; wherein, the drilling data is real-time data generated by drilling operations. The downhole drilling state is determined based on the drilling data. Drilling states include multiple types; for each type of drilling state, a hook load prediction model and / or torque prediction model corresponding to the drilling state is determined; wherein, the drilling state includes any one or more of the following: drilling, tripping, running downhole, reaming, and reverse reaming. The drilling data is predicted using both the hook load prediction model and the torque prediction model to obtain the predicted values ​​of the hook load and torque at the current moment. Based on the actual values ​​of the hook load and torque, the actual trends of their changes are determined; and based on the predicted values ​​of the hook load and torque, the predicted trends of their changes are determined. Based on the actual and predicted trends, the downhole stuck pipe index is determined. Therefore, compared to existing technologies, this application solves the problem that traditional single-parameter and multi-parameter pre-set threshold methods are difficult to universally apply to stuck pipe identification under different working conditions, well types, and stuck pipe types, leading to high false alarm and false negative rates; it also solves the problem that current physical models and purely data-driven models require detailed drilling dynamic and static parameters as input or require the construction of complete, high-quality stuck pipe sample data for model training. By constructing independent intelligent prediction models for hook load and torque under different drilling conditions, drilling engineering knowledge and machine learning algorithms are fully integrated to automatically identify stuck pipe. This avoids the subjectivity and lag of manual monitoring, and also saves the cumbersome data collection and input and the construction of a large amount of stuck pipe sample data. It is of great significance to improve the accuracy of stuck pipe identification and the degree of automation and intelligence, and effectively improves the accuracy and timeliness of stuck pipe prediction, thereby improving the accuracy of stuck pipe identification in drilling.

[0133] In one example Figure 3 A flowchart illustrating a stuck pipe identification method for drilling provided in this application. Figure 3 ,like Figure 3As shown, the process includes: real-time data acquisition; real-time data cleaning; obtaining cleaned real-time data; drilling status identification; dividing the dataset according to the drilling status; performing feature engineering on the dataset; obtaining training datasets for training prediction models: hook load training datasets and torque training datasets divided by drilling status; obtaining hook load prediction models and torque prediction models based on the support vector machine prediction algorithm and the training datasets. The hook load prediction models include tripping hook load prediction models, running hook load prediction models, reaming hook load prediction models, and reverse reaming hook load prediction models. The torque prediction models include drilling torque prediction models, reaming hook load prediction models, and reverse reaming hook load prediction models; obtaining prediction results: hook load prediction values ​​and torque prediction values; performing trend analysis of predicted value changes to obtain predicted trends; performing trend analysis of measured hook load and torque changes to obtain actual trends; comparing the actual trends of measured values ​​with the predicted trends of predicted values ​​to calculate the stuck pipe index, i.e., the stuck pipe index.

[0134] Figure 4 A schematic diagram of a stuck drill bit identification device provided in this application is shown below. Figure 4 As shown, the stuck drill identification device 30 provided in this embodiment includes:

[0135] The acquisition module 31 is used to acquire drilling data, and to acquire the actual value of the hook load and the actual value of the torque at the current moment in the drilling data; wherein, the drilling data is real-time data generated by the drilling operation.

[0136] The first determining module 32 is used to determine the downhole drilling status based on drilling data.

[0137] The second determining module 33 is used to determine the preset hook load prediction model and torque prediction model corresponding to the drilling state.

[0138] The prediction module 34 is used to call the hook load prediction model and the torque prediction model respectively to perform drilling prediction on the drilling data and obtain the hook load prediction value and torque prediction value at the current moment.

[0139] The third determining module 35 is used to determine the stuck pipe index in the well based on the actual value of the hook load, the actual value of the torque, the predicted value of the hook load, and the predicted value of the torque; wherein, the stuck pipe index indicates whether stuck pipe will occur in the well.

[0140] Figure 5 This is a schematic diagram of another stuck drill identification device provided in an embodiment of this application. Figure 4 Based on the illustrated embodiments, as Figure 5 As shown, the third determining module 35 includes:

[0141] The first determining unit 351 is used to determine the actual changing trend of the hook load and torque based on the actual value of the hook load and the actual value of the torque; and to determine the predicted changing trend of the hook load and torque based on the predicted value of the hook load and the predicted value of the torque.

[0142] The second determining unit 352 is used to determine the stuck pipe index downhole based on the actual and predicted changing trends.

[0143] In one possible implementation, the first determining unit 351 is specifically used for:

[0144] Obtain historical time series of hook load and torque.

[0145] Based on the preset average difference algorithm of adjacent sliding windows, the actual changing trends of hook load and torque are determined by the actual value of hook load, actual value of torque, historical hook load time series, and torque time series.

[0146] Based on the average difference algorithm of adjacent sliding windows, the predicted change trends of hook load and torque are determined by using the predicted values ​​of hook load, torque, historical hook load time series, and torque time series.

[0147] In one possible implementation, the second determining unit 352 is specifically used for:

[0148] Compare the actual trend with the predicted trend. If the actual trend and the predicted trend are inconsistent, calculate the change in hook load between the actual value and the predicted value, as well as the change in torque between the actual value and the predicted value.

[0149] The stuck pipe index in the well is determined based on the changes in hook load and torque.

[0150] In one possible implementation, the drilling status includes multiple types; the second determining module 33 is specifically used for:

[0151] For each type of drilling condition, determine the hook load prediction model and / or torque prediction model corresponding to the drilling condition.

[0152] The drilling status includes any one or more of the following: drilling, tripping, running down, reaming, and reverse reaming.

[0153] In one possible implementation, the device is also specifically used for:

[0154] Obtain datasets for each drilling state.

[0155] For each dataset under each drilling condition, the values ​​of the target parameters are determined based on the dataset.

[0156] A training set for predicting the load of a large hook is generated based on the values ​​of the target parameters and the values ​​of the preset parameters in the dataset used to generate the large hook load prediction model.

[0157] A torque prediction training set is generated based on the values ​​of the target parameters and the values ​​of the preset parameters in the dataset used to generate the torque prediction model.

[0158] The hook load prediction model is trained using the hook load prediction training set until it converges; and the torque prediction model is trained using the torque prediction training set until it converges.

[0159] The stuck drill identification device provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0160] Figure 6 A schematic diagram of the structure of the electronic device provided in this application. Figure 6 As shown, the electronic device 50 provided in this embodiment includes at least one processor 501 and a memory 502. Optionally, the device 50 further includes a communication component 503. The processor 501, memory 502, and communication component 503 are connected via a bus 504.

[0161] In a specific implementation, at least one processor 501 executes computer execution instructions stored in memory 502, causing at least one processor 501 to perform the above-described method.

[0162] The specific implementation process of processor 501 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0163] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0164] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0165] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0166] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0167] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0168] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0169] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0170] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0171] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0172] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0173] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0174] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0175] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for identifying stuck drill pipe in well drilling, characterized in that, include: Acquire drilling data, and acquire the actual values ​​of the hook load and torque at the current moment from the drilling data; wherein, the drilling data is real-time data generated by the drilling operation; Based on the drilling data, determine the downhole drilling status; Determine the preset hook load prediction model and torque prediction model corresponding to the drilling state, and call the hook load prediction model and torque prediction model respectively to perform drilling prediction on the drilling data to obtain the hook load prediction value and torque prediction value at the current moment. The stuck pipe index is determined based on the actual value of the hook load, the actual value of the torque, the predicted value of the hook load, and the predicted value of the torque; wherein, the stuck pipe index indicates whether stuck pipe will occur downhole.

2. The method according to claim 1, characterized in that, The determination of the stuck pipe index in the well based on the actual value of the hook load, the actual value of the torque, the predicted value of the hook load, and the predicted value of the torque includes: Based on the actual values ​​of the hook load and torque, determine the actual changing trends of the hook load and torque; and based on the predicted values ​​of the hook load and torque, determine the predicted changing trends of the hook load and torque. The stuck pipe index in the well is determined based on the actual and predicted trends.

3. The method according to claim 2, characterized in that, The actual changing trends of the hook load and torque are determined based on the actual values ​​of the hook load and torque. And based on the predicted values ​​of the hook load and torque, determine the predicted trends of the changes in both the hook load and the torque, including: Obtain historical time series of hook load and torque; Based on the preset adjacent sliding window average difference algorithm, the actual change trend of the hook load and the torque is determined by the actual value of the hook load, the actual value of the torque, the historical hook load time series, and the torque time series. Based on the adjacent sliding window average difference algorithm, the predicted change trends of the hook load and the torque are determined by the hook load prediction value, the torque prediction value, the historical hook load time series, and the torque time series.

4. The method according to claim 3, characterized in that, Determining the sticking index downhole based on the actual trend and the predicted trend includes: The actual trend of change is compared with the predicted trend of change. If it is determined that the actual trend of change and the predicted trend of change are inconsistent, the change in hook load between the actual value of hook load and the predicted value of hook load, and the change in torque between the actual value of torque and the predicted value of torque are calculated. The stuck pipe index in the well is determined based on the change in hook load and the change in torque.

5. The method according to claim 1, characterized in that, The drilling conditions include multiple types; determining the hook load prediction model and torque prediction model corresponding to the drilling conditions includes: For each type of drilling condition, determine the hook load prediction model and / or torque prediction model corresponding to the drilling condition; The drilling state includes any one or more of the following: drilling, tripping, running down, reaming, and reverse reaming.

6. The method according to any one of claims 1-5, characterized in that, The method further includes: Obtain datasets for each drilling state; For each dataset under each drilling condition, the value of the target parameter is determined based on the dataset. A training set for predicting the load of a large hook is generated based on the values ​​of the target parameters and the values ​​of the preset parameters in the dataset used to generate the large hook load prediction model. A torque prediction training set is generated based on the values ​​of the target parameters and the values ​​of the preset parameters used to generate the torque prediction model in the dataset. The hook load prediction model is trained based on the hook load prediction training set until it converges; and the torque prediction model is trained based on the torque prediction training set until it converges.

7. A stuck drill bit identification device for drilling, characterized in that, include: The acquisition module is used to acquire drilling data, and to acquire the actual values ​​of the hook load and torque at the current moment from the drilling data; wherein, the drilling data is real-time data generated by the drilling operation; The first determining module is used to determine the downhole drilling status based on the drilling data; The second determining module is used to determine the preset hook load prediction model and torque prediction model corresponding to the drilling state; The prediction module is used to call the hook load prediction model and torque prediction model respectively to perform drilling prediction on the drilling data, and obtain the hook load prediction value and torque prediction value at the current moment. The third determining module is used to determine the stuck pipe index in the well based on the actual value of the hook load, the actual value of the torque, the predicted value of the hook load, and the predicted value of the torque; wherein, the stuck pipe index indicates whether stuck pipe will occur in the well.

8. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-6.

10. A computer program product comprising a computer program that, when executed by a processor, implements the method of any one of claims 1-6.

Images