A method for predicting complex drilling situations

Abstract

The present invention relates to the field of geological exploration technology, and in particular to a method for predicting complex drilling situations, including: generating a time domain spectrum based on historical time domain parameters when a wellhead drill string vibrates; performing Fourier transform on the time domain parameters in the time domain spectrum and generating a frequency domain spectrum; obtaining corresponding weight parameters and amplitude parameters when a complex drilling situation occurs according to the frequency domain spectrum, and further establishing a complex situation prediction calculation model; and using the complex situation prediction calculation model to estimate the risk of complex drilling situations. This technical solution has low implementation cost, simple and convenient operation, and strong real-time performance. It can timely discover the development trend of complex drilling situations and make predictions, reduce the risk of drill bit sticking and drill bit mud packing, and is of great significance to reducing drilling costs.

Description

Technical Field

[0001] The invention relates to the technical field of geological exploration, and in particular to a method for predicting complex drilling situations. Background Art

[0002] During the drilling process, drilling complications are likely to occur due to formation factors, technical factors or improper operation. The so-called drilling complications include:

[0003] 1) The drill string is difficult to rotate or is stuck, resulting in drill stuck, which needs to be released before drilling can continue. Drill stuck will cause great harm to the drilling project and require a lot of manpower and material resources to deal with. At the least, it will waste drilling time and increase costs. At the worst, it will cause part of the drill bit to fall to the bottom of the well, making the wellbore scrapped. At present, the main prediction methods for drill stuck include comprehensive logging method, decision tree method, Bayesian method, neural network method, support vector machine method, etc. Most of these methods require learning and training of drill stuck cases that have occurred in the area to effectively predict. If there are not many cases, it will affect the prediction accuracy, especially for new blocks, where there are not many wells drilled, there are no cases or very few cases, and the adaptability is subject to certain limitations; these methods mainly use drilling parameters, and the size and change of drilling parameters are closely related to well depth, wellbore structure, formation lithology, drill bit type, drill tool structure, and drilling technology, all of which may affect the accuracy of the prediction.

[0004] 2) Mud packing on the drill bit reduces the drilling efficiency or makes drilling impossible, increasing the well cost and well construction cycle. Mud packing on the drill bit will also cause great harm to the drilling project. The mechanical drilling speed of the drill bit after mud packing will be significantly reduced, which will affect the drilling efficiency. Forcibly starting drilling with a mud-packed drill bit may also induce well kicks or even blowouts, which requires a lot of manpower and material resources to deal with. At the least, it will waste drilling time and increase costs, and at worst, it will cause serious downhole accidents. At present, the prediction method for drill bit mud packing is mainly to predict through the changes in comprehensive logging parameters. It mainly relies on manual identification and fails to form a quantitative model. It is closely related to the quality and sense of responsibility of on-site operators, and the prediction accuracy is not high. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a method for predicting drilling complex situations. The method has low cost, simple operation, and strong real-time performance. The method can timely discover the development trend of drilling complex situations and make forecasts, thereby reducing the risks of drill sticking and drill bit mud packing, which is of great significance to reducing drilling costs.

[0006] This is achieved specifically through the following technical solutions:

[0007] A method for predicting drilling complexity, characterized by comprising:

[0008] Generate a time domain spectrum based on the historical time domain parameters of the wellhead drill string vibration;

[0009] Perform Fourier transformation on the time domain parameters in the time domain spectrum and generate a frequency domain spectrum;

[0010] Obtain corresponding weight parameters and amplitude parameters when a complex drilling situation occurs according to the frequency domain spectrum, and establish a complex situation prediction calculation model based on the weight parameters and amplitude parameters;

[0011] Monitor the vibration of the drill string at the target wellhead in real time, obtain relevant vibration parameters, and use the complex situation prediction calculation model to estimate the risk of complex drilling situations.

[0012] Preferably, the generating of the time domain spectrum comprises the following steps:

[0013] The vibration signal of the wellhead drill string is collected by using a vibration acceleration sensor;

[0014] Using a computer to remotely and wirelessly receive the vibration signal collected by the vibration acceleration sensor;

[0015] According to the distribution of the vibration direction of the wellhead drill string in space, a three-dimensional coordinate system including the X-axis, Y-axis and Z-axis is established by computer, and the vibration signal is converted into time domain parameters corresponding to the three directions of the X-axis, Y-axis and Z-axis;

[0016] According to the distribution of the time domain parameters on the X-axis, Y-axis and Z-axis, a time domain spectrum including vibration in the X-axis direction, vibration in the Y-axis direction and vibration in the Z-axis direction is generated by computer.

[0017] Preferably, the complex situation includes drill sticking and / or drill bit balling.

[0018] Preferably, the complex situation prediction calculation model includes a drill sticking prediction calculation model and / or a drill bit balling prediction calculation model.

[0019] Preferably, establishing the stuck pipe prediction calculation model comprises the following steps:

[0020] According to the situation of whether there is a stuck drill case, the weight values ​​are obtained in the three directions of the X-axis, Y-axis and Z-axis of the frequency domain spectrum, including the first weight W j1 and the second weight W j2 ; Wherein, j∈m, m represents the vibration direction, and the value is 1, 2 or 3, 1 represents the X-axis direction, 2 represents the Y-axis direction, and 3 represents the Z-axis direction;

[0021] According to the drill bit type and drill bit speed, determine the frequency range f in the corresponding frequency domain spectrum when the drill is stuck j1 ~f j2 , f j1 is the starting frequency, and its value is f j2 : is the end frequency, the value is Hz; n is the drill bit speed; k is the number of drill teeth or blades;

[0022] Get the frequency range f j1 ~f j2 The maximum amplitude value MAXA in j1 And the frequency range f j1 ~f j2 The maximum amplitude value MAXA in the frequency domain j2 ;

[0023] Establish a calculation model for predicting stuck drill:

[0024]

[0025] Among them, F q A is the stuck drill risk index; ji A represents the amplitude value in the j direction when the frequency is i; jmax and A jmin are the maximum amplitude and minimum amplitude in the frequency domain spectrum of the drill string vibration at the target wellhead, respectively;

[0026] Establish the drill stuck alarm threshold value α.

[0027] Preferably, when obtaining the weight value, if there is a drill stuck case, the weight values ​​of the three directions of X-axis, Y-axis and Z-axis are obtained through a neural network algorithm based on the relevant vibration parameters of the drill stuck case; if there is no drill stuck case, the weight values ​​of the three directions of X-axis, Y-axis and Z-axis are obtained according to the contribution degree of the directions of X-axis, Y-axis and Z-axis to the drill stuck. j1 The values ​​in the X-axis, Y-axis and Z-axis directions are 30%, 20% and 10% respectively. j2 The values ​​in the X-axis, Y-axis and Z-axis directions are 20%, 15% and 5% respectively.

[0028] Preferably, estimating the risk of drill bit sticking comprises the following steps:

[0029] The vibration signal of the drill string at the target wellhead is measured in real time using a vibration acceleration sensor;

[0030] Generate a real-time frequency spectrum based on the acquired vibration signal using a computer;

[0031] Get the maximum amplitude A of the target wellhead drill string in the frequency spectrum jmax and minimum amplitude A jmin ;

[0032] Based on the maximum amplitude A jmax and minimum amplitude A jmin , the stuck pipe risk index F is calculated using the stuck pipe prediction calculation model q ;

[0033] The stuck pipe risk index Fq Compare with the alarm threshold value α; when F q >α, it means the risk index of drill sticking exceeds the safety range, and the computer issues a warning. q ≤α, indicating that the stuck pipe risk index is within the safe range, and the computer only displays the specific stuck pipe risk index value.

[0034] Preferably, establishing the drill bit balling prediction calculation model comprises the following steps:

[0035] Based on the frequency domain spectrum, the average amplitude A in the X-axis direction when the drill bit does not have mud balls is obtained. xn , and the average amplitude reduction value ΔA when mud balls appear x , and use the neural network algorithm to obtain the amplitude weight W4 in the X-axis direction;

[0036] Based on historical comprehensive logging data, the average pump pressure p when the drill bit is free of mud balls is obtained. n , Average drilling time ROP n and average torque N n , as well as the pump pressure increase Δp, drilling increase ΔROP and torque reduction ΔN when mud balls appear;

[0037] Based on the historical comprehensive logging data, the neural network algorithm is used to obtain the pump pressure weight W1, drilling time weight W2 and torque weight W3;

[0038] Establish drill bit mud balling prediction calculation model:

[0039]

[0040] Among them, F p is the risk index of mud balling; p is the actual pump pressure; ROP is the actual drilling time; N is the actual torque; A x is the actual amplitude in the X-axis direction.

[0041] Establish mud bag alarm threshold value β.

[0042] Preferably, estimating the risk of drill bit balling includes the following steps:

[0043] Collect comprehensive logging data in real time and obtain the actual amplitude A of the drill string at the target wellhead in the X-axis direction x ; Comprehensive logging data includes the actual pump pressure p, actual drilling ROP and actual torque N corresponding to the drill bit operation;

[0044] Based on the collected comprehensive logging data, the drill bit balling prediction calculation model is used to calculate the balling risk index F. p ;

[0045] The risk index of mud packing F pCompare with the alarm threshold value β; when F p >β, it means that the mud bag risk index exceeds the safety range, and the computer issues a warning; when F p ≤β, indicating that the balling risk index is within a safe range, and the computer only displays the specific balling risk index value.

[0046] Preferably, the real-time collection of comprehensive logging data is performed through a comprehensive logging instrument data interface.

[0047] Beneficial effects brought by this technical solution:

[0048] 1) This technical solution aims at the complex drilling situations of drill bit sticking and / or mud balling. It provides corresponding prediction conditions by establishing a complex situation prediction calculation model. By collecting relevant parameters corresponding to the target drill bit, the real-time prediction of the complex drilling situation can be directly realized through the computer. No special prediction equipment is required. Under the condition of not increasing the equipment cost and the on-site investigation cost, the well crew can take timely response measures according to the prediction results, which reduces the risk of drill bit sticking and mud balling, and is of great significance to increasing the service life of the drill bit and improving the drilling efficiency. Based on this, the drilling cost can be further reduced.

[0049] 2) The entire technical solution starts from the vibration signal of the wellhead drill string, uses historical time domain parameters to establish the corresponding time domain spectrum, and combines Fourier transform to obtain the vibration frequency domain diagram (frequency domain spectrum) of the wellhead drill string. This not only facilitates the acquisition of the parameters required to establish a complex situation prediction calculation model, but also ensures the correlation between the parameters and the target drill bit, the effectiveness and reliability of the parameter use, and the accuracy of the prediction results.

[0050] 3) Vibration signals can be collected remotely using a vibration acceleration sensor in conjunction with a computer, and comprehensive logging data can be collected through the comprehensive logging instrument data interface. The method of obtaining the parameters required for predicting this complex situation is simple and direct, ensuring that this technical solution is easy to implement. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The foregoing and following detailed description of the present invention will become more apparent when read in conjunction with the following drawings, in which:

[0052] Figure 1 A schematic diagram of a preferred implementation process of the present invention;

[0053] Figure 2 This is the vibration spectrum diagram of Well X1 before and after the drill bit got stuck;

[0054] Figure 3 This is the vibration spectrum diagram of Well X1 before and after the jam; DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solution and advantages of the invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments.

[0056] Therefore, the following detailed description of the invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] Example 1

[0058] This embodiment provides a method for predicting drilling complexity, which, as a preferred implementation scheme of the technical solution, includes the following steps:

[0059] A time domain spectrum is generated based on the historical time domain parameters of the wellhead drill string vibration. The historical time domain parameters can be time domain parameters of the same well or time domain parameters of different wells. Specifically, based on the database of the current drilling project, find out whether the target well area (the well area where the target well is located, the so-called target well, that is, the well that currently needs to be predicted for drilling complexity) has previously used the same drilling equipment as the target well in the reference well (the well whose drilling project precedes the target well and has more comprehensive drilling project data). If yes, the time domain parameters of the wellhead drill string vibration during the drilling project of the reference well are used as the historical time domain parameters required by this technical solution; if not, the drilling equipment in the target well works for a period of time, and the time domain parameters of the wellhead drill string vibration during this period are collected as the historical time domain parameters required by this application.

[0060] The time domain parameters in the time domain spectrum are Fourier transformed to generate a frequency domain spectrum. Among them, the Fourier transform can express a function that meets certain conditions as a linear combination of trigonometric functions (sine and / or cosine functions) or their integrals. In different research fields, the Fourier transform has many different variants, such as continuous Fourier transform and discrete Fourier transform.

[0061] The corresponding weight parameters and amplitude parameters when a drilling complex situation occurs are obtained according to the frequency domain spectrum, and a complex situation prediction calculation model is established based on the weight parameters and amplitude parameters. Preferably, the complex situation prediction calculation model includes a drill stuck prediction calculation model and a drill bit mud balling prediction calculation model.

[0062] The vibration of the drill string at the target wellhead is monitored in real time, relevant vibration parameters are obtained, and the risk of drilling complex situations (i.e., drill bit sticking and / or drill bit mud balling) is estimated using a complex situation prediction calculation model.

[0063] This technical solution aims at the complex drilling situations of drill bit sticking and / or mud balling. It provides corresponding prediction conditions by establishing a complex situation prediction calculation model. By collecting relevant parameters corresponding to the target drill bit, the real-time prediction of the complex drilling situation can be directly realized through the computer. No special prediction equipment is required. Under the condition of not increasing the equipment cost and the on-site investigation cost, the well crew can take timely response measures according to the prediction results, thereby reducing the risk of drill bit sticking and mud balling, which is of great significance to increasing the service life of the drill bit and improving the drilling efficiency. Based on this, the drilling cost can be further reduced.

[0064] Example 2

[0065] This embodiment provides a method for predicting complex drilling situations. As a preferred implementation of this technical solution, according to actual cases, the target well is named as X1 well, and complex situation prediction is achieved for the working conditions in the target well. Specifically, the following steps are included:

[0066] S1, generating a time domain spectrum based on the historical time domain parameters of the wellhead drill string vibration, specifically comprising the following steps:

[0067] S1-1, using a vibration acceleration sensor to collect vibration signals of a wellhead drill string, where the wellhead drill string may be a target wellhead drill string corresponding to the X1 well, or a wellhead drill string corresponding to a reference well in the same target well area using the same drilling equipment;

[0068] S1-2, using a computer to remotely and wirelessly receive a vibration signal collected by a vibration acceleration sensor;

[0069] S1-3, according to the distribution of the vibration direction of the wellhead drill string in space, a three-dimensional coordinate system including an X-axis (tangential), a Y-axis (transverse) and a Z-axis (vertical) is established by a computer (the three-dimensional coordinate system may be a virtual three-dimensional coordinate system, that is, the three-dimensional coordinate system here is only used to distinguish the vibration direction of the wellhead drill string, so as to record the corresponding time domain parameters according to the direction), and the vibration signal is converted into time domain parameters corresponding to the three directions of the X-axis, the Y-axis and the Z-axis;

[0070] S1-4, according to the distribution of the time domain parameters on the X-axis, Y-axis and Z-axis, a time domain spectrum including vibration in the X-axis direction, vibration in the Y-axis direction and vibration in the Z-axis direction is generated by computer.

[0071] S2, Fourier transforms the time domain parameters in the time domain spectrum and generates a frequency domain spectrum, which has the same directionality as the time domain spectrum.

[0072] S3, obtaining corresponding weight parameters and amplitude parameters when a complex drilling situation occurs according to the frequency domain spectrum, and establishing a complex situation prediction calculation model based on the weight parameters and amplitude parameters, the complex situation prediction calculation model is a stuck drill prediction calculation model.

[0073] The establishment of a stuck pipe prediction calculation model includes the following steps:

[0074] S3-1, according to whether there is a stuck drill case, obtain weight values ​​in the three directions of the X-axis, Y-axis and Z-axis of the frequency domain spectrum, including the first weight W j1 and the second weight W j2 ; Where j∈m, m represents the vibration direction, and its value is 1, 2 or 3, 1 represents the X-axis direction, 2 represents the Y-axis direction, and 3 represents the Z-axis direction. Where W j1 and W j2 The values ​​in the three directions of X-axis, Y-axis and Z-axis may be different. There are different ways to obtain the values ​​depending on whether there is a drill stuck case. Specifically, if there is a drill stuck case, the weight values ​​of the three directions of X-axis, Y-axis and Z-axis are obtained through the neural network algorithm based on the relevant vibration parameters of the drill stuck case. If there is no drill stuck case, then W j1 The values ​​in the X-axis, Y-axis and Z-axis directions are 30%, 20% and 10% respectively. j2 The values ​​in the X-axis, Y-axis and Z-axis directions are 20%, 15% and 5% respectively. j1 With W j2 The relative size of W is determined according to their contribution to the stuck pipe prediction calculation model. j1 and W j2 The values ​​in the X-axis, Y-axis, and Z-axis directions are determined according to the contribution of the wellhead drill string vibration in the X-axis, Y-axis, and Z-axis directions. j1 The three values ​​corresponding to the X-axis, Y-axis, and Z-axis directions are related to W j2 The sum of the three values ​​corresponding to the X-axis, Y-axis, and Z-axis directions is 100%. In the case of the five stuck drill cases in Well X1, according to the contribution of the X-axis, Y-axis, and Z-axis directions to the stuck drill, the weight values ​​of the three directions of X-axis, Y-axis, and Z-axis are as shown in Table 1 (table of parameter values).

[0075] S3-2, according to the drill bit type and drill bit speed, determine the frequency range f in the corresponding frequency domain spectrum when the drill is stuck j1 ~f j2 Among them, f j1 is the starting frequency, f j2 is the end frequency, both of which are related to the drill type and drill speed, specifically: f j1 The value is Hz (round to integer); f j2 The value is Hz (round integer); n is the drill bit speed; k is the number of drill teeth (for roller bit) or the number of blades (for PDC bit).

[0076] S3-3, obtain the frequency range f j1 ~f j2 The maximum amplitude value MAXA in j1 And the frequency range f j1 ~f j2 The maximum amplitude value MAXA in the frequency domain j2 Among them, MAXA j1 and MAXA j2 The values ​​on the X-axis, Y-axis, and Z-axis may be different, which are related to the well depth, drilling tool structure, drill bit type, and drilling parameters. They can be obtained by measuring the parameters during normal drilling on site (without drilling complications) and analyzing the drill bit stuck cases. When there is no drill bit stuck case, MAXA j1 and MAXA j2 Take a value three times the maximum amplitude value in the corresponding vibration direction during normal drilling. The relevant parameter values ​​of Well X1 are shown in Table 1 (Table of Parameter Values):

[0077] Table 1 Parameter value table

[0078]

[0079] S3-4, establish a prediction model for stuck drill:

[0080]

[0081] Among them, F q A is the stuck drill risk index; ji A represents the amplitude value in the j direction when the frequency is i; jmax and A jmin They are the maximum amplitude and minimum amplitude in the frequency domain spectrum of the drill string vibration at the target wellhead, respectively.

[0082] S3-5, establishing a drill stuck alarm threshold value α. Specifically, the drill stuck alarm threshold value may be set to 60.

[0083] S4, real-time monitoring of the vibration of the target wellhead drill string, obtaining relevant vibration parameters, and using the stuck drill calculation model to estimate the risk of drilling complications. Specifically, the following steps are included:

[0084] S4-1, using a vibration acceleration sensor to measure the vibration signal of the target wellhead drill string in real time;

[0085] S4-2, using a computer to generate a real-time frequency spectrum based on the acquired vibration signal;

[0086] S4-3, obtain the maximum amplitude A of the target wellhead drill string in the frequency spectrum jmax and minimum amplitude A jmin ;

[0087] S4-4, based on the maximum amplitude A jmax and minimum amplitude A jmin , the stuck pipe risk index F is calculated using the stuck pipe prediction calculation model q ;

[0088] S4-5 will be stuck drill risk index F q Compare with the alarm threshold value α; when F q >α, it means the risk index of drill sticking exceeds the safety range, and the computer issues a warning. q ≤α, indicating that the stuck pipe risk index is within the safe range, and the computer only displays the specific stuck pipe risk index value.

[0089] In this example, the drill bit got stuck due to block loss at 17:40 in Well X1. Figure 2 It is the vibration spectrum diagram (frequency spectrum) of the well before and after the drill stuck. It can be seen from the figure that before the drill stuck (before 17:35), the vibration amplitudes in the three directions were small and stable. The drill stuck risk index was calculated using the drill stuck prediction calculation model, and its value was also very small, mostly below 10, and the maximum was less than 20, indicating that the drill stuck risk was very low; from 17:35 to 17:40, it can be seen from the figure that the frequency of the 12-20Hz segment in the Y direction increased sharply, and the amplitudes in the X and Z directions also increased significantly. The drill stuck risk index in this case also rose rapidly to above 63.2. Based on this, the computer immediately issued a drill stuck warning, and the well crew immediately took up and down movable drilling tools. At 18:40, the drill stuck risk index rose to 87.7 at its maximum, and the drill stuck still occurred. Because the upper well wall fell off, it would naturally sink downward under the action of gravity. This kind of drill stuck was difficult to avoid. Fortunately, after the computer issued a drill stuck prediction, the well crew took appropriate measures, the drill stuck was not very serious, and it was quickly handled and successfully unstuck. Figure 3 This is the vibration spectrum diagram (frequency spectrum) of Well X1 before and after it was unstuck. After 18:33, the vibration spectrum of the drill string at the wellhead quickly returned to normal, and the stuck drill risk index dropped below 5.

[0090] Example 3

[0091] This embodiment provides a method for predicting drilling complexity, which, as a preferred implementation scheme of the technical solution, includes the following steps:

[0092] S1, generates a time domain spectrum based on the historical time domain parameters of the wellhead drill string vibration.

[0093] S2, performing Fourier transform on the time domain parameters in the time domain spectrum and generating a frequency domain spectrum.

[0094] S3, obtaining corresponding weight parameters and amplitude parameters when a complex drilling situation occurs according to the frequency domain spectrum, and establishing a complex situation prediction calculation model based on the weight parameters and amplitude parameters, the complex situation prediction calculation model is a drill bit mud balling prediction calculation model.

[0095] The establishment of a drill bit balling prediction calculation model includes the following steps:

[0096] S3-1, based on the frequency domain spectrum, obtain the average amplitude A in the X-axis direction when the drill bit does not have mud balls xn , and the average amplitude reduction value ΔA when mud balls appear x , and use the neural network algorithm to obtain the first weight W4 in the X-axis direction. For example, the value of W4 in a certain actual case is 34%.

[0097] S3-2 Based on historical comprehensive logging data, obtain the average pump pressure p when the drill bit does not have mud balls n , Average drilling time ROP n and average torque N n , and the pump pressure increase value Δp when mud balls appear, the increase value ΔROP during drilling, and the torque reduction value ΔN. Among them, the historical comprehensive logging data can be the comprehensive logging data in the target well (the well where the target drill bit is located), or it can be the comprehensive logging data in other reference wells that use the same drill bit in the same well area as the target well. Δp, ΔROP, ΔA x The values ​​of and ΔN are related to the formation, well depth, drilling tool structure, wellbore structure, drill bit type and drilling parameters, and can be obtained through the drill bit balling case using statistical analysis methods.

[0098] S3-3, based on the historical comprehensive logging data, the neural network algorithm is used to obtain the pump pressure weight W1, drilling time weight W2 and torque weight W3. The corresponding values ​​of W1, W2 and W3 in the actual case are 12%, 23% and 31% respectively. In short, W1, W2, W3 and W4 are 100%.

[0099] S3-4 Establish a drill bit mud ball prediction calculation model:

[0100]

[0101] Among them, F p is the risk index of mud balling; p is the actual pump pressure; ROP is the actual drilling time; N is the actual torque; A x is the actual amplitude in the X-axis direction.

[0102] Drilling time is one of the indicators to measure the drillability of rock formations, that is, the time (minutes) required to drill one meter. Drilling time is exactly the reciprocal of mechanical drilling speed. Oil drilling determines the changes in the formation by measuring drilling time, which is called drilling time logging. Drilling time logging is an important part of drilling geological work.

[0103] Pump pressure refers to the surface pressure at the discharge port of the pump. The pump pressure of the core drilling pump is generated by the various resistances encountered by the flushing fluid flowing through the drilling flushing circulation system. Therefore, the pump pressure is equal to the sum of the pressure losses of the flushing fluid flowing through the circulation system. This part of the pressure loss includes the pressure loss of the flushing fluid flowing through the core tube and the annular space between the cores, the pressure loss when the flushing fluid turns and changes direction when flowing through the drill bit, and the pressure loss of the flushing fluid flowing through the annular space between the core tube and the hole wall. It is usually obtained by experimental methods or empirical data.

[0104] S3-4, establishing a mud ball alarm threshold value β, and the specific mud ball alarm threshold value β can be 60-80.

[0105] S4, real-time monitoring of the vibration of the drill string at the target wellhead, obtaining relevant vibration parameters, and using the drill bit balling prediction calculation model to estimate the risk of drill bit balling, specifically including the following steps:

[0106] S4-1, real-time acquisition of comprehensive logging data, and acquisition of the actual amplitude A of the drill string at the target wellhead in the X-axis direction x ; Comprehensive logging data includes the actual pump pressure p, actual drilling ROP and actual torque N corresponding to the drill bit working. Get the actual amplitude A x The method is as follows: first, based on the vibration model of the target wellhead drill string, a corresponding frequency spectrum is generated, and then based on the frequency spectrum, the actual amplitude A of the vibration signal of the target wellhead drill string in the X-axis direction is obtained. x Preferably, the comprehensive logging data can be collected through a comprehensive logging instrument data interface.

[0107] S4-2, based on the collected comprehensive logging data, the drill bit balling prediction calculation model is used to calculate the balling risk index F p ; The mud bag risk index F p Compare with the alarm threshold value β; when F p >β, it means that the mud bag risk index exceeds the safety range, and the computer issues a warning; when F p ≤β, indicating that the balling risk index is within a safe range, and the computer only displays the specific balling risk index value.

Claims

1. A method for predicting drilling complexity, characterized in that: include: Generating a time domain spectrum based on historical time domain parameters of a wellhead drill string during vibration, specifically comprising the following steps: using a vibration acceleration sensor to collect vibration signals of the wellhead drill string; using a computer to remotely and wirelessly receive the vibration signals collected by the vibration acceleration sensor; using a computer to establish a three-dimensional coordinate system including an X-axis, a Y-axis and a Z-axis according to the distribution of the vibration direction of the wellhead drill string in space, and converting the vibration signal into time domain parameters corresponding to the three directions of the X-axis, the Y-axis and the Z-axis; using a computer to generate a time domain spectrum including vibration in the X-axis direction, the Y-axis direction and the Z-axis direction according to the distribution of the time domain parameters on the X-axis, the Y-axis and the Z-axis; Perform Fourier transformation on the time domain parameters in the time domain spectrum and generate a frequency domain spectrum; According to the frequency domain spectrum, corresponding weight parameters and amplitude parameters are obtained when a complex drilling situation occurs, and a complex situation prediction calculation model is established based on the weight parameters and amplitude parameters; the complex situation includes a stuck drill, and the complex situation prediction calculation model includes a stuck drill prediction calculation model. Establishing the stuck drill prediction calculation model includes the following steps: according to whether there is a stuck drill case, corresponding to the frequency domain spectrum, the weight values ​​in the three directions of the X-axis, Y-axis and Z-axis are obtained, including the first weight and the second weight ;in, , Indicates the vibration direction, with values ​​of 1, 2, or 3, 1 for the X-axis direction, 2 for the Y-axis direction, and 3 for the Z-axis direction; Determine the frequency range in the corresponding frequency domain spectrum when the drill is stuck based on the drill type and drill speed , is the starting frequency, and its value is Hz; is the end frequency, and its value is Hz; where is the drill bit speed; is the number of drill teeth or blades; get the frequency range The maximum amplitude value in And frequency range The maximum amplitude value in the frequency domain ; Establish a calculation model for prediction of stuck drill: ;in, is the stuck drill risk index; Indicated in Direction, frequency is The amplitude value at time ; and are the maximum amplitude and minimum amplitude in the frequency domain spectrum of the drill string vibration at the target wellhead, respectively; a drill stuck alarm threshold value α is established; Monitor the vibration of the target wellhead drill string in real time, obtain relevant vibration parameters, and use the complex situation prediction calculation model to estimate the risk of drilling complex situations, including the risk of drill bit sticking. The specific steps are as follows: use a vibration acceleration sensor to measure the vibration signal of the target wellhead drill string in real time; use a computer to generate a real-time frequency spectrum based on the obtained vibration signal; obtain the maximum amplitude of the target wellhead drill string in the frequency spectrum and minimum amplitude ; Based on maximum amplitude and minimum amplitude , use the stuck pipe prediction calculation model to calculate the stuck pipe risk index ; Set the stuck drill risk index Compare with the alarm threshold value α; when When , it means that the risk index of drill bit sticking exceeds the safety range, and the computer issues a warning; when , indicating that the stuck drill risk index is within the safe range, and the computer only displays the specific stuck drill risk index value.

2. A method for predicting drilling complexity as claimed in claim 1, characterized in that: When obtaining the weight values, if there is a drill stuck case, the weight values ​​of the three directions of X-axis, Y-axis and Z-axis are obtained through a neural network algorithm based on the relevant vibration parameters of the drill stuck case; if there is no drill stuck case, the weight values ​​of the three directions of X-axis, Y-axis and Z-axis are obtained based on the contribution of the directions of X-axis, Y-axis and Z-axis to the drill stuck. The values ​​in the X-axis, Y-axis and Z-axis directions are 30%, 20% and 10% respectively. The values ​​in the X-axis, Y-axis and Z-axis directions are 20%, 15% and 5% respectively.

3. A method for predicting drilling complexity as claimed in claim 1, characterized in that: The complex situation also includes drill bit mud balling, and the corresponding complex situation prediction calculation model also includes a drill bit mud balling prediction calculation model.

4. A method for predicting drilling complexity as claimed in claim 3, characterized in that: Establishing the drill bit balling prediction calculation model includes the following steps: Based on the frequency domain spectrum, the average amplitude in the X-axis direction when the drill bit does not have mud balls is obtained. , and the average amplitude reduction when mud balls appear , and use the neural network algorithm to obtain the amplitude weight in the X-axis direction ; Based on historical comprehensive logging data, obtain the average pump pressure when the drill bit is not mud-balled , Average drilling time and average torque , and the pump pressure increase value when mud balls appear , drilling time increase value and torque reduction ; Based on historical comprehensive logging data, the pump pressure weight is obtained using a neural network algorithm , Drilling Time Weight and torque weight ; Establish drill bit mud balling prediction calculation model: ； in, is the mud bag risk index; is the actual pump pressure; is the actual drilling time; is the actual torque; is the actual amplitude in the X-axis direction; Establish mud bag alarm threshold value β.

5. A method for predicting drilling complexity as claimed in claim 4, characterized in that: Estimating the risk of drilling complications also includes estimating the risk of drill bit balling, which includes the following steps: Collect comprehensive logging data in real time and obtain the actual amplitude of the drill string in the X-axis direction at the target wellhead ; Comprehensive logging data includes the actual pump pressure corresponding to the drill bit operation , Actual drilling time and actual torque ; Based on the collected comprehensive logging data, the drill bit balling prediction calculation model is used to calculate the balling risk index. ; The risk index of mud pack Compare with the alarm threshold value β; when When , it means that the mud bag risk index exceeds the safety range, and the computer issues a warning; when , indicating that the mud balling risk index is within the safe range, and the computer only displays the specific mud balling risk index value.

6. A method for predicting drilling complexity as claimed in claim 5, characterized in that: The real-time collection of comprehensive logging data is carried out through the comprehensive logging instrument data interface.

Images