阅读说明：本技术 一种套管最小内径检测方法 (Method for detecting minimum inner diameter of casing ) 是由 黄华 张�诚 刘晓敏 张文昌 朱玲玲 李宏魁 万龙 谭肖 于 2020-04-26 设计创作，主要内容包括：本发明涉及一种套管最小内径检测方法,属于油气田工程测井技术领域。本发明首先制作测量套管变形井,并测量套管变形井中不同变形程度的套管的最小内径；然后在套管变形井的套管内下入油管,在所述油管内设置电磁测井仪,测量套管变形井中不同井深处的脉冲涡流响应信号；再采用机器学习算法建立脉冲涡流响应信号和套管最小内径之间的数据模型,通过样本训练的方式,得到较为精确的脉冲涡流响应信号与套管最小井径之间的数据模型,更为精确的拟合脉冲涡流响应信号与套管最小井径之间的非线性关系；最后将待测井套管的脉冲涡流响应信号带入上述模型,从而实现在不取油管的条件下,准确得到套管最小内径值。(The invention relates to a method for detecting the minimum inner diameter of a casing, and belongs to the technical field of oil and gas field engineering logging. Firstly, manufacturing a sleeve deformation measuring well, and measuring the minimum inner diameters of sleeves with different deformation degrees in the sleeve deformation well; then, an oil pipe is put into a sleeve of the sleeve deformation well, an electromagnetic logging instrument is arranged in the oil pipe, and pulse eddy current response signals at different well depths in the sleeve deformation well are measured; establishing a data model between the impulse eddy current response signal and the minimum inner diameter of the casing by adopting a machine learning algorithm, obtaining a more accurate data model between the impulse eddy current response signal and the minimum well diameter of the casing in a sample training mode, and more accurately fitting a nonlinear relation between the impulse eddy current response signal and the minimum well diameter of the casing; and finally, substituting the pulse eddy current response signal of the well casing to be tested into the model, thereby accurately obtaining the minimum inner diameter value of the casing under the condition of not taking the oil pipe.)

1. A method for detecting the minimum inner diameter of a casing is characterized by comprising the following steps:

1) the method comprises the steps of carrying out statistical analysis on the deformation type and the deformation degree of a casing of each real logging in a research area, manufacturing a plurality of deformation casings with different deformation types and different deformation degrees according to the deformation type and the deformation degree of the casing of each real logging in the research area to form a casing deformation well, and measuring the minimum inner diameter of the casings with different deformation degrees in the casing deformation well;

2) an oil pipe is arranged in a sleeve of the sleeve deformation well, an electromagnetic logging instrument is arranged in the oil pipe, and pulse eddy current response signals at different well depths in the sleeve deformation well are measured;

3) establishing a nonlinear relation model representing the minimum inner diameter of the casing and the impulse eddy current response signal by using a machine learning algorithm, wherein the nonlinear relation model is trained by samples formed by the minimum inner diameters of the casings with different deformation degrees in the casing deformation well and the corresponding impulse eddy current response signals;

4) and (3) putting an electromagnetic logging instrument into the underground oil pipe to be detected, measuring the pulse eddy current response signals at different depths of the well, and inputting the pulse eddy current response signals at different depths of the underground oil pipe to be detected into the nonlinear relation model so as to obtain the minimum inner diameter of the sleeve corresponding to each depth of the underground oil pipe to be detected.

2. The casing pipe minimum inner diameter detection method according to claim 1, wherein the machine learning algorithm is a regression random forest algorithm, and the nonlinear relation model is obtained by a classification regression tree fitting mode.

3. The casing minimum inside diameter inspection method of claim 1, further comprising baseline drift correction of the measured pulsed eddy current response signal.

4. The method of claim 3, wherein the baseline wander correction is performed using a least squares method.

5. The casing minimum inside diameter inspection method according to any one of claims 1 to 4, wherein the electromagnetic logging tool employs an eddy current electromagnetic logging tool of a three-directional probe.

6. The method of claim 5, further comprising performing feature extraction on each probe signal of the eddy current electromagnetic logging tool at each depth by using a principal component analysis method, selecting a first principal component of each probe as a statistical feature, combining the first principal components of each probe to obtain a combined feature, and using the combined feature as a corresponding pulsed eddy current response signal at the depth.

7. The casing minimum inner diameter detection method according to claim 1 or 2, wherein the casing deformation well is formed by connecting a plurality of deformation casings with different deformation types and different deformation degrees in series and then putting the deformation casings into a real logging well.

Technical Field

The invention relates to a method for detecting the minimum inner diameter of a casing, and belongs to the technical field of oil and gas field engineering logging.

Background

In the modern oil and gas well industry, the downhole tubular column structure of an oil and gas production well mainly comprises a casing and an oil pipe, wherein the casing is used for keeping a well wall stable and separating an oil and gas layer and is cemented and sealed by cement slurry and the well wall; the oil pipe is used as an oil-gas transmission channel and is positioned inside the casing pipe. Due to the influence of factors such as formation stress, perforation, pressure change of a production zone, corrosion of downhole fluid and the like, the casing serving as a protective layer often has defects of different degrees, such as: extrusion, deformation, etc., which directly affect the safety, oil and gas production and production efficiency of the downhole operation.

The minimum inner diameter of the casing is a key parameter in deformation detection of the casing of the oil-gas well, and the numerical value of the minimum inner diameter of the casing directly reflects the deformation degree of the casing and is an important basis for evaluating the integrity of the shaft of the oil-gas well. The method for detecting the minimum inner diameter of the downhole tubular column commonly used at home and abroad comprises ultrasonic downhole television well logging, eagle-eye television well logging, multi-arm borehole diameter well logging and eddy-current electromagnetic well logging.

The ultrasonic underground television logging, the eagle-eye television logging and the multi-arm caliper logging belong to caliper imaging logging methods, and measurement principles of ultrasonic, optics and mechanical arm expansion are respectively adopted, so that measurement of an oil pipe or a casing single-layer pipe column can be realized only. In an oil-gas production well, an oil pipe is arranged in a sleeve and is limited by a well body structure, and when the deformation of the inner diameter of the sleeve is tested, the oil pipe needs to be taken out first to enable an instrument to be in direct contact with the inner wall of the sleeve for measurement. The depth under the oil pipe of a general oil and gas well exceeds 3000m, the construction period of the operation of taking the oil pipe is long, the cost is high, especially under the condition of high pressure and high sulfur content, a permanent packer is adopted in the underground pipe column, when the oil pipe is taken, a series of measures such as killing the well, temporarily blocking the oil and gas layer, milling the packer and the like are needed to be taken for avoiding the leakage of hydrogen sulfide, the process is complex, the construction time is long, the operation risk is large, even the situations that the oil pipe and a downhole tool fall, the production layer is killed, the later period cannot be produced again and the like can be caused, and the application of the well logging technologies such as ultrasonic underground television well logging, eagle eye television well logging, multi-arm well diameter and the like is limited.

The eddy current electromagnetic logging is based on the electromagnetic mutual inductance principle, adopts an indirect measurement mode, excites a group of bipolar current signals by an electromagnetic coil to form a primary magnetic field around the coil, the magnetic field generates an eddy current ring when encountering an annular medium to further form a secondary magnetic field, receives the magnetic field signals at a gap for emitting excitation signals, and can realize the defect condition measurement of an oil pipe, a casing single-layer pipe column structure and an oil pipe and casing double-layer pipe column structure according to the distribution condition and the size of the eddy current electromagnetic field in the oil pipe and the casing. According to the measurement principle, the test data directly acquired by the eddy current electromagnetic logging instrument can only represent the electromagnetic field size of the casing pipe defect and cannot directly obtain the minimum inner diameter of the casing pipe, quantitative calculation of the minimum inner diameter of the casing pipe can be realized only by carrying out feature extraction and processing calculation on the test data, and reliable basis is provided for the formulation of measures such as later-stage well repair, operation, well sealing and the like.

For example, the invention patent with the application number of 201811217762.1 and the invention name of "a method for detecting the deformation degree of the inner diameter of a casing of a downhole string" adopts a mode of establishing a casing deformation standard well and utilizes a least square linear fitting method to obtain a mathematical model between the casing inner diameter size values of different well depths and the casing electromagnetic response characteristics of the corresponding well depths, thereby realizing the quantitative detection of the deformation degree of the inner diameter of the casing under the condition of not taking the oil pipe. However, the method adopts a linear fitting mode, the impulse eddy current response signal and the minimum borehole diameter of the casing are actually in a nonlinear relation, and when a fitting model is established, the fitting relation is established only according to the maximum electromagnetic response value of each casing variable section and the size of the inner diameter of the casing, the electromagnetic response characteristic of the whole casing variable section is not considered, so that the inner diameter of the casing of the downhole tubular column obtained through measurement is not consistent with the inner diameter of the actual casing of the downhole tubular column.

Disclosure of Invention

The invention aims to provide a method for detecting the minimum inner diameter of a casing, which aims to solve the problem of inaccurate measurement result caused by the fact that the minimum inner diameter is detected by adopting a linear fitting relation at present.

The invention provides a method for detecting the minimum inner diameter of a casing for solving the technical problems, which comprises the following steps:

1) the method comprises the steps of carrying out statistical analysis on the deformation type and the deformation degree of a casing of each real logging in a research area, manufacturing a plurality of deformation casings with different deformation types and different deformation degrees according to the deformation type and the deformation degree of the casing of each real logging in the research area to form a casing deformation well, and measuring the minimum inner diameter of the casings with different deformation degrees in the casing deformation well;

2) an oil pipe is arranged in a sleeve of the sleeve deformation well, an electromagnetic logging instrument is arranged in the oil pipe, and pulse eddy current response signals at different well depths in the sleeve deformation well are measured;

3) establishing a nonlinear relation model representing the minimum inner diameter of the casing and the impulse eddy current response signal by using a machine learning algorithm, wherein the nonlinear relation model is trained by samples formed by the minimum inner diameters of the casings with different deformation degrees in the casing deformation well and the corresponding impulse eddy current response signals;

4) and (3) putting an electromagnetic logging instrument into the underground oil pipe to be detected, measuring the pulse eddy current response signals at different depths of the well, and inputting the pulse eddy current response signals at different depths of the underground oil pipe to be detected into the nonlinear relation model so as to obtain the minimum inner diameter of the sleeve corresponding to each depth of the underground oil pipe to be detected.

According to the method, a data model between the pulse eddy current response signal and the minimum inner diameter of the casing is obtained by adopting a machine learning algorithm, and due to the nonlinear relation between the pulse eddy current response signal and the minimum inner diameter of the casing, the accurate data model between the pulse eddy current response signal and the minimum inner diameter of the casing can be obtained by adopting the machine learning algorithm in a sample training mode, the nonlinear relation between the pulse eddy current response signal and the minimum inner diameter of the casing can be fitted more accurately, and the minimum inner diameter value of the casing can be accurately obtained under the condition that an oil pipe is not taken.

Further, in order to describe the relationship between the impulse eddy current response signal and the minimum inner diameter of the casing more accurately, the machine learning algorithm is a regression random forest algorithm, and the nonlinear relationship model is obtained in a classification regression tree fitting mode.

Further, in order to avoid the problem that the electromagnetic logging instrument is interfered and inaccurate in measurement, the method further comprises the step of carrying out baseline drift correction on the measured impulse eddy current response signals.

Further, the baseline drift correction adopts a least square method.

Furthermore, in order to obtain complete spatial three-dimensional electromagnetic information of different well depths of the casing deformation well, the electromagnetic logging instrument adopts an eddy current electromagnetic logging instrument with probes in three directions.

Furthermore, in order to reduce the dimension of the original signal, retain most information of the signal and eliminate the mutual influence among the dimensions of the signal part, the method also comprises the steps of respectively using a principal component analysis method to extract the characteristics of each probe signal of the eddy current electromagnetic logging instrument at each depth, selecting the first principal component of each probe as a statistical characteristic, combining the first principal components of each probe to obtain a combined characteristic, and using the combined characteristic as a corresponding pulse eddy current response signal at the depth.

Furthermore, the casing deformation well is formed by connecting a plurality of deformation casings with different deformation types and different deformation degrees in series and then putting the casings into a real logging well.

Drawings

FIG. 1 is a flow chart of a method of detecting the minimum inside diameter of a casing according to the present invention;

FIG. 2 is a schematic representation of casing minimum inside diameter measurement data for each well depth sample point of a casing deformation well in an embodiment of the present invention;

FIG. 3 is a schematic diagram of an eddy current electromagnetic tool configuration for a three directional probe used in an embodiment of the present invention;

FIG. 4 is a schematic representation of a comparison of a minimum inside diameter measurement to be logged to a corresponding multi-arm caliper measurement in an embodiment of the present invention;

wherein 1 is a longitudinal probe A, 2 is a first transverse probe B, 3 is a transverse second transverse probe C, 4 is an auxiliary structure, and 5 is a centralizer.

Detailed Description

The following further describes embodiments of the present invention with reference to the drawings.

Firstly, manufacturing a sleeve deformation measuring well, and measuring the minimum inner diameters of sleeves with different deformation degrees in the sleeve deformation well; then, an oil pipe is put into a sleeve of the sleeve deformation well, an electromagnetic logging instrument is arranged in the oil pipe, and pulse eddy current response signals at different well depths in the sleeve deformation well are measured; and then, a data model between the impulse eddy current response signal and the minimum inner diameter of the casing is obtained by adopting a machine learning algorithm, and due to the nonlinear relation between the impulse eddy current response signal and the minimum inner diameter of the casing, a more accurate data model between the impulse eddy current response signal and the minimum inner diameter of the casing can be obtained by adopting the machine learning algorithm in a sample training mode, the nonlinear relation between the impulse eddy current response signal and the minimum inner diameter of the casing is more accurately fitted, and the minimum inner diameter value of the casing can be accurately obtained under the condition of not taking an oil pipe. The detection flow of the method is shown in fig. 1, and comprises the following steps:

1. and manufacturing a plurality of deformed sleeves with different deformation types and different deformation degrees to form the sleeve deformed well.

And (5) carrying out statistical analysis on the deformation type and the deformation degree of each actual well log in the research area. Through carrying out statistical analysis on the hole diameter monitoring historical data of each actual well logging in the research area, including the historical hole diameter data interpretation result, the historical well repairing construction record, the geological data of the research area and the like, the casing deformation type and the deformation degree of each actual well logging in the research area are obtained, and the simulated casing deformation well is formed. According to the deformation type and the deformation degree of the casing of each real logging in a research area, a plurality of deformation casings with different deformation types and different deformation degrees are manufactured to form a casing deformation well, the casing deformation well is connected in series to form a simulation casing, and the simulation casing is put into the real logging (namely a real well) in the research area to form a simulation casing deformation well.

As other embodiments, several deformation casings of different deformation types and different deformation degrees can be separately lowered into the research area for logging without series connection.

2. The minimum inside diameter of casing of different deformation degrees in a casing deformation well is measured.

In the embodiment, the multi-arm caliper logging method is used for measuring the inner diameter size values of the casing pipe simulating different well depths of the casing pipe deformation well. The multi-arm well diameter logging method is characterized in that the well diameter measuring equipment used by the multi-arm well diameter logging method is a plurality of groups of mechanical arms which are unfolded outwards from the center of the instrument, and the multi-arm curves measured by the mechanical arms are utilized to obtain the inner diameter size values of the casings at different well depths.

In this embodiment, the MFC-24 twenty-four arm caliper is used to measure the inner diameter of the casing at different depths of the casing deformation well, and R is min_0≤i≤24r_i(h) And further solving the minimum inner diameter of the corresponding downhole tubular column casing. Wherein h represents the well depth and is m; r represents the minimum value of the inner diameter of the casing pipe when the well depth is h, and the unit is mm; r is_iRepresents the value of the inside diameter of the casing measured in mm for this multiple arm. The minimum inside diameter measurement data of the casing for each sample point is shown in fig. 2; as can be seen from fig. 2, the minimum casing inner diameter data of the casing deformation well is not uniformly distributed because the minimum casing inner diameter of the casing deformation well is obviously changed only when the casing has defects such as deformation and bending.

As other embodiments, the inside diameter of the casing of the deformed well of the simulated casing may be measured by other borehole diameter imaging logging methods, for example: an ultrasonic downhole television logging method or an eagle eye television logging method.

3. And (3) putting an oil pipe into a simulation casing pipe of the simulation casing pipe deformation well, putting an eddy current electromagnetic logging instrument into the oil pipe, measuring the electromagnetic response characteristics of the casing pipe of the simulation casing pipe deformation well at different well depths, and performing pretreatment and characteristic extraction.

3.1 the invention puts the oil pipe in the middle of the simulation casing of the casing deformation well, in order to obtain the complete space three-dimensional electromagnetic information of different well depths of the casing deformation well, put MTD-J eddy current electromagnetic logging instrument in the oil pipe, measure the impulse eddy current response signal of different well depths of the casing deformation well, the eddy current electromagnetic logging instrument used has A, B, C three direction detection probes, the spatial distribution of the three direction probes used is as shown in figure 3, the space axes of the first transverse probe B2 and the second transverse probe C3 are horizontal and vertical, the space axis of the longitudinal probe A1 is horizontal and vertical, the eddy current electromagnetic logging instrument also includes the auxiliary structure 4 and the centralizer 5.

3.2, obtaining the preprocessed pulse eddy current signal by adopting a baseline drift removal preprocessing method based on a least square method for the signal obtained in the step 3.1. The input is the original acquired pulsed eddy current signal, the sample data of this embodiment includes 24-dimensional data including X depth data, 23-dimensional data included in the pulsed eddy current signal data, and the data dimensions from different detection probes A, B, C are 11-dimensional, 6-dimensional and 6-dimensional, respectively, where U ═ U₁…U₂₃]And (3) adopting a least square fitting method (a, b) ═ argmin for the baseline drift of the eddy current signals at different depths of the 23-dimensional pulse eddy current signals_a,b(‖aX+b-U‖²) Calculating to obtain a vector and a vector b, and removing baseline drift;

3.3 obtaining a 3.2 preprocessed pulsed eddy current signal, the pulsed eddy current signal data comprising 23-dimensional data, the data dimensions from different inspection probes A, B, C being 11-dimensional, 6-dimensional, respectively. Where the data is represented as a matrix:

wherein U is_a、U_b、U_cRespectively representing A, B, C the signal components measured by the probe; respectively extracting first principal component components by using principal component analysis methods for signal components of different probes and combining the first principal component components as combined characteristic quantity D ═ D₁,D₂,D₃]Wherein D is₁、D₂、D₃Respectively representing A, B, C probe eddy current signals U_a、U_b、U_cFirst principal component components obtained by principal component analysis, respectively.

The invention utilizes different detection capabilities of each probe to respectively use a principal component analysis method to extract the characteristics of each probe signal, selects the first principal component of each probe capable of fully representing the minimum inner diameter information of the casing as statistical characteristics, and combines the statistical characteristics to obtain combined characteristics. The first principal component of each probe is selected to obtain a combined characteristic, so that the dimension of an original signal is reduced, most information of the signal is kept, and mutual influence among the dimensions of signal parts is eliminated.

4. And constructing a quantitative analysis data model between the impulse eddy current response signal and the minimum inner diameter of the casing.

The embodiment is based on a regression random forest method to construct a quantitative analysis data model between the impulse eddy current response signal and the minimum inner diameter of the casing. The training set data is 3201 groups of combined characteristic quantity obtained by extracting the pulse eddy current response signals and the minimum inner diameter value of the casing under the corresponding well depth, and the training data set E { (U)₁,R₁),(U₂,R₂),...,(U₃₂₀₁,R₃₂₀₁) In which U is_iRepresenting combined characteristic quantities, R, at a certain depth of a casing deformation well_iRepresenting the minimum inner diameter value of the casing corresponding to the well depth, and constructing a quantitative analysis data model between the impulse eddy current response signal and the minimum inner diameter of the casing through training based on a random forest regression method: combining feature quantity U of sample at certain depth_iAnd a minimum inner diameter R corresponding to the sample_iInputting a regression random forest algorithm, wherein the algorithm fits the nonlinear relation between the combined characteristic quantity of the input sample and the minimum inner diameter by establishing a classification regression tree; after the input 3201 group of samples are trained, a mathematical model of the nonlinear relation between the impulse eddy current response signal U and the minimum inner diameter R of the casing based on the regression random forest algorithm is established.

The invention adopts random forest regression to carry out quantitative analysis and detection to obtain a data model between the impulse eddy current response signal and the minimum inner diameter of the casing. Due to the nonlinear relation between the impulse eddy current response signal and the minimum inner diameter of the casing, a random forest method is adopted, a classifier comprising a plurality of decision trees can be trained, a more accurate data model between the impulse eddy current response signal and the minimum inner diameter of the casing can be obtained, and the nonlinear relation between the impulse eddy current response signal and the minimum inner diameter of the casing can be fitted more accurately.

As another implementation, the random forest algorithm in this embodiment may also adopt other machine learning algorithms, for example, algorithms such as a neural network.

5. And (3) putting an electromagnetic logging instrument into the underground oil pipe to be detected, measuring the pulse eddy current response signals at different depths of the well, and inputting the pulse eddy current response signals at different depths of the underground oil pipe to be detected into the nonlinear relation model so as to obtain the minimum inner diameter of the sleeve corresponding to each depth of the underground oil pipe to be detected.

In this embodiment, before fishing an oil pipe in a gas field D405-3 well for well sealing, the sleeve pulse eddy current responses of the sleeve deformation well at different well depths are measured in the oil pipe, 1067 sets of field experimental data are obtained in total, and the calculation result of the minimum inner diameter of the sleeve at the corresponding well depth of the sleeve deformation well is obtained by using the quantitative analysis data model obtained in step 4.

After the calculation result of the minimum inner diameter of the casing at the corresponding well depth of the casing deformation well is obtained by the method, after the oil pipe is salvaged in the well, the minimum inner diameter of the casing at the corresponding well depth after the oil pipe is salvaged in the well is directly measured by a multi-arm well diameter logging method, and a comparison graph of the calculation result of the minimum well diameter of the well and the result of measured data of the multi-arm well diameter is obtained, as shown in fig. 4, it can be seen that: the error between the minimum hole diameter calculation result obtained by the method and the multi-arm hole diameter data is basically less than 10 mm.

In the embodiment of the invention, the average prediction error of 067 sample points is 1.69mm, 1043 sample points with the error smaller than 10mm account for 97.75% of the total sample points; compared with the application number of 201811217762.1, the invention is named as a method for detecting the deformation degree of the inner diameter of a casing of a downhole tubular column, and the method can enable the calculation error of the minimum value of the inner diameter of the casing of the same depth of the same well to be less than 5 percent, so the method can more accurately fit the nonlinear relation between the impulse eddy current response signal and the minimum well diameter of the casing, realize the average prediction error of the minimum inner diameter of a sample point to be less than 2 percent, and even if the data distribution of the minimum inner diameter of the casing is not uniform, the method still has good generalization capability, and in all the sample points, the sample point with the error of less than 10mm accounts for 97.75 percent of the total sample point; the method can accurately and quantitatively calculate the minimum well diameter value of the well casing, and has good application prospect.

It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the invention and is not intended to limit the invention, which has been described in detail with reference to the foregoing examples, but it will be apparent to those skilled in the art that various changes in the form and details of the invention may be made and equivalents may be substituted for elements thereof. All modifications, equivalents and the like which come within the spirit and principle of the invention are intended to be included within the scope of the invention.