CN-122020285-A - Real-time identification system and method for stick-slip working conditions of horizontal well drilling bit

CN122020285ACN 122020285 ACN122020285 ACN 122020285ACN-122020285-A

Abstract

The invention belongs to the technical field of horizontal well drilling optimization, and discloses a system and a method for identifying the stick-slip working condition of a horizontal well drilling bit in real time. The method comprises the steps of obtaining multi-source data such as drilling pressure, rotating speed, torque and mechanical drilling speed based on a logging while drilling technology, constructing an obtaining method of effective torque of a drill bit combined with a drilling test to obtain effective rock breaking torque of the drill bit under different moment and well depth conditions, constructing a logging data normalization model, normalizing the multi-source logging data, and obtaining the change rule of real rock breaking energy utilization rate of the drill bit along with time and well depth. The invention constructs a real-time automatic identification method for the stick-slip working condition of the horizontal well drilling bit, further applies manual intervention in time, avoids the stick-slip working condition from developing into the early damage of the bit, and has important significance for saving non-production time, prolonging the service life of the bit and sliding a drilling ruler and ensuring the drilling safety and production benefit.

Inventors

SONG WEIQIANG
LI SHUPENG
YU HAIYANG
ZHANG WEI
LI YUANHANG
Mu Huidong
LIU TONGSHENG
WANG XU

Assignees

山东科技大学

Dates

Publication Date: 20260512
Application Date: 20260108

Claims (10)

1. The method for identifying the stick-slip working condition of the drilling bit of the horizontal well in real time is characterized by comprising the following steps of: s1, acquiring multisource data of weight on bit, rotation speed, torque and mechanical drilling speed based on a logging while drilling technology, constructing a drill bit effective rock breaking torque acquisition method combined with a while drilling test, and reading a torque value recorded during normal rotary drilling And torque recorded when the drilling tool is lifted off the bottom of the well and kept at the original rotation speed to idle Obtaining effective rock breaking torque at the drill bit under different drilling moments and well depth conditions; S2, constructing a logging data normalization model, taking the effective rock breaking torque of the drill bit and logging data under different moments and well depth conditions into the normalization model, and calculating to obtain the real rock breaking energy utilization rate A law of variation with time and well depth; S3, analyzing the obtained real rock breaking energy utilization rate at corresponding moment and well depth The method comprises the steps of carrying out physical priori knowledge normalization on data, constructing a bp neural network, carrying out parameter optimization, training verification, constructing a horizontal well drill bit stick-slip working condition real-time identification model based on logging data, and determining And (3) judging whether the characteristic parameter threshold values of the amplitude, the frequency and the time domain change rate are the stick-slip working conditions, and carrying out real-time identification and regulation of the stick-slip working conditions of the horizontal well drill bit.
2. The method for identifying the stick-slip condition of the drilling bit of the horizontal well in real time according to claim 1, wherein in the step S1, obtaining the multi-source data of the weight on bit, the rotation speed, the torque and the mechanical drilling rate comprises: s101, reading torque sum of normal rotary drilling , Equal to the rotational torque of the drill string And effective rock breaking torque And (3) summing; S102, lifting the drill string, separating the drill bit from the bottom of the well, reducing the drill bit to 0 but keeping the rotation speed unchanged, and reading the torque recorded by the logging system at the moment ; And in step S101 The values of (2) are the same; S103, lowering the drill bit back to the bottom of the well, adding the weight to a normal value, and effectively breaking the rock torque at the drill bit Obtained by the following calculation; ; In the formula, In order to be effective in terms of the breaking torque, In the form of a sum of the torques, Torque recorded for a logging system; Torque recorded by logging system The well depth is increased and changed, but the daily footage of the horizontal section is limited, short-lift rock clearing is needed, and a plurality of well depth positions are obtained Further, the linear interpolation is adopted to obtain Functional relation with well depth, i.e. in the above Is a function of well depth, is to be obtained Logging data normalization model with physical priori knowledge is carried in to obtain rock breaking effect of drill bit at corresponding moment and well depth 。
3. The method for identifying the stick-slip condition of the horizontal well drilling bit in real time according to claim 1, wherein in step S2, a logging data normalization model is constructed as follows: ; In the formula, For the true rock breaking energy utilization rate of the drill bit, Is weight on bit, unit is ; Is the cross-sectional area of the well bore in units of ; The unit is the rotating speed of the turntable/top drive ; For the output rotation speed of the underground power drilling tool, the power drilling tool producer can provide the relation between the output rotation speed and the displacement, wherein the unit is ; Real-time torque recorded by a logging system during normal drilling is expressed in units of ; Average torque recorded for logging system during normal circulation and rotation of drill bit lifted off bottom of well in units of ; Is the mechanical drilling speed, and the unit is ; For the true rotational speed of the drill bit, The real rock breaking torque of the drill bit is obtained.
4. The method for identifying the stick-slip condition of the horizontal well drilling bit in real time according to claim 1, wherein in step S3, constructing the horizontal well drilling bit stick-slip condition real-time identification model based on logging data comprises: S301, transmitting the obtained real rock breaking energy utilization rate of the drill bit at the corresponding moment and the well depth to a first input layer of a neural network, wherein each input variable corresponds to one input neuron; s302, transmitting the real rock breaking energy utilization rate of the drill bit at the corresponding moment and well depth to the next hidden layer or the output layer of the network through the weighted sum nonlinear transformation of the input layer; S303, comparing the output result with an actual label, calculating average relative error of the output value and a true value by using MAE and R2, dividing a data set according to the ratio of a training set to a test set of 7:3, combining physical priori knowledge, and constructing a real-time recognition model of the stick-slip working condition of the horizontal well drill bit based on logging data according to a neural network modeling step.
5. The method for identifying the stick-slip working conditions of the horizontal well drilling bit in real time according to claim 4 is characterized in that after a horizontal well drilling bit stick-slip working condition real-time identification model based on logging data is constructed, if the output value of the model is 2, the model is the stick-slip working condition, when the deviation between the predicted value and the actual value of the model exceeds a set threshold, the characteristic parameters of amplitude and frequency of the real rock breaking energy utilization rate of the bit along with the change of time and well depth are required to be compared with the threshold, whether the working conditions are stick-slip is judged, and the reliability of the horizontal well drilling bit stick-slip working condition real-time identification model based on the logging data is further verified.
6. The method for identifying the stick-slip condition of the horizontal well drilling bit in real time according to claim 4, wherein in the step S3, the construction of the model for identifying the stick-slip condition of the horizontal well drilling bit based on logging data in real time specifically comprises the following steps: (1) Marking the true value of the well section under the normal working condition as 1, marking the true value of the stick-slip generating well section as 2, setting the number of hidden layer nodes as 4, selecting trainscg algorithm by training algorithm, selecting tansing and purelin excitation functions, and introducing multiple parameters of mechanical drilling speed, torque and drilling weight as input vectors into a BP neural network for training and testing; (2) The method comprises the steps of marking the true value of a well section under a normal working condition as 1, marking the true value of a well section under a stick-slip working condition as 2, setting the number of hidden layer nodes as 4, selecting trainscg algorithm by training algorithm, selecting tansing and purelin excitation functions, and selecting multiple parameters of mechanical drilling speed, torque and drilling weight as input vectors for training and testing; (3) Comparing the step (1) with the step (2), and taking MSEb single parameters as input vectors; (4) Selecting optimal node number by changing hidden layer node number, selecting maximum iteration step number as 1000 by adopting Trainlm/LMs algorithm, and inputting MSE at hidden layer node number of 4-20 to train neural network; (5) Fixing the maximum iteration step number to be 1000, selecting Trianscg as an optimal algorithm, and optimizing the node number of the hidden layer; (6) And under the optimal algorithm and the optimal node number, only changing the maximum iteration step number for testing, and selecting the maximum iteration step number as 100-1500 for training the BP artificial neural network.
7. The method for identifying the stick-slip condition of the horizontal well drilling bit in real time according to claim 1, wherein in the step S3, a real rock breaking energy utilization rate of the bit corresponding to the moment and the well depth and a horizontal well drilling bit stick-slip condition real-time identification model based on logging data are embedded into a logging system, and the stick-slip condition is identified in real time and a prompt is given in drilling.
8. A system for identifying the stick-slip condition of a drilling bit of a horizontal well in real time, which is characterized in that the system implements the method for identifying the stick-slip condition of the drilling bit of the horizontal well in real time according to any one of claims 1 to 7, and the system comprises: the method comprises the steps of establishing a method for acquiring the effective torque of the drill bit by combining with the while-drilling test, obtaining the effective torque of the drill bit under the conditions of different moments and well depths, fitting the drill bit breaking torque into a function of the well depths, predicting the breaking torque under the subsequent drilling working conditions in real time, and continuously and iteratively correcting the fitting function of the drill bit breaking torque and the well depths by combining with the subsequent while-drilling test results; the system comprises a real rock breaking energy utilization rate calculation module, a logging data normalization module, a data analysis module and a data analysis module, wherein the real rock breaking energy utilization rate calculation module is used for constructing a logging data normalization model, normalizing multi-source logging data of effective rock breaking torque, top drive rotating speed, displacement and drilling pressure of the drill bit, and obtaining real rock breaking energy utilization rate reflecting the drill bit; According to the method, by combining well Shi Xianyan data, evolution data of the utilization rate of rock breaking energy of the drill bit along with time and well depth under normal drilling and stick-slip working conditions are obtained respectively, a real-time recognition model of the stick-slip working conditions of the drill bit of the horizontal well based on logging data is built according to training and testing neural network modeling steps, and then real-time recognition and regulation of the stick-slip working conditions of the drill bit of the horizontal well are carried out.
9. The system for identifying the stick-slip condition of the drilling bit of the horizontal well according to claim 8, wherein the system is mounted on computer equipment, and the computer equipment comprises at least one processor, a memory and a computer program stored in the memory and running on the at least one processor, and the processor realizes the functions when executing the computer program.
10. The system for identifying the stick-slip condition of the drilling bit of the horizontal well in real time according to claim 8, wherein the system is mounted on a computer readable storage medium, and the computer readable storage medium stores a computer program, and the computer program can realize the functions when being executed by a processor.

Description

Real-time identification system and method for stick-slip working conditions of horizontal well drilling bit Technical Field The invention belongs to the technical field of horizontal well drilling optimization, and particularly relates to a system and a method for identifying the stick-slip working condition of a horizontal well drilling bit in real time. Background Horizontal well drilling can obviously increase the contact area between a well bore and a reservoir to improve the recovery ratio, and is a necessary technology for oil and gas benefit development. Taking drilling of a deep coal-rock gas reservoir horizontal well as an example, a PDC drill bit can quickly shear rock to complete rock breaking after the drill bit is used for immersing coal rock, the actual rotation speed and rock breaking torque of the drill bit are stable, but because the gangue in the coal-rock reservoir is obviously higher in strength than the gangue and easy to fall off, after the drill bit is immersed in the gangue under the action of the bit pressure, the drill bit torque is required to be accumulated for a certain time (the torque increasing speed is slower) to be enough to shear the gangue, the drill bit torque is further quickly released, the rotation speed and rock breaking torque are suddenly high and suddenly low, and the fluctuation frequency is obviously reduced, so that the unfavorable working condition of sticky and slippery drill bit appears. The stick-slip working condition is a main cause of early damage of the PDC drill bit in horizontal well drilling, is easy to cause unexpected tripping and drill bit replacement, and seriously affects the drilling aging and construction cost of the horizontal well. After the drill bit is in a stick-slip working condition, if the drill bit weight and the rotating speed can be timely reduced, the early damage of the drill bit can be completely avoided. At present, drilling constructors mainly analyze fluctuation conditions of mechanical drilling speed and ground torque in drilling, observe the appearance of a drill bit after drilling, develop the analysis after drilling by combining working experience, and judge the working conditions of the upper drilling and the lower drilling so as to guide subsequent operation, and in recent years, scientific researchers try to identify the stick-slip of the drill bit by installing various vibration sensors on the drill bit. The method for identifying the stick-slip by analyzing the mechanical drilling speed and the torque fluctuation in drilling is excessively dependent on subjective experience of engineers, reliability is low, and actual requirements cannot be met, the stick-slip working condition can be accurately identified by a drill bit vibration sensor, but the cost of the drill bit is increased, ground and underground are difficult to communicate in real time in the drilling process, measured vibration data are analyzed and applied after the drilling is started under the most conditions unless real-time coding and underground-ground communication systems and equipment are additionally added, and cost is high. Therefore, a quantitative prediction method capable of accurately judging the stick-slip of the drill bit in real time under the technical conditions of the process is needed in horizontal well drilling. Logging while drilling is almost standard of all drilling construction, a constructor can acquire corresponding data of drilling parameters (drilling pressure, torque, displacement, rotating speed and the like) and mechanical drilling speed in real time through logging while drilling, and adjacent well logging data can be kept in file and a large-scale database is built, well logging while drilling information comprises well bottom drilling working condition information, and the current well logging database needs to be further dug in the aspect of supporting well drilling optimization. Disclosure of Invention In order to overcome the problems in the related art, the disclosed embodiments of the invention provide a system and a method for identifying the stick-slip condition of a horizontal well drilling bit in real time, and in particular relate to a model and a method for identifying the stick-slip condition of the horizontal well drilling bit in real time based on logging data. The technical scheme is that the method for identifying the stick-slip working condition of the drilling bit of the horizontal well in real time comprises the following steps: s1, acquiring multisource data of weight on bit, rotation speed, torque and mechanical drilling speed based on a logging while drilling technology, constructing a drill bit effective rock breaking torque acquisition method combined with a while drilling test, and reading a torque value recorded during normal rotary drilling And torque recorded when the drilling tool is lifted off the bottom of the well and kept at the original rotation speed to idleObtaining effective rock bre