A method for intelligently judging the best time for drilling short start

Abstract

The present invention relates to the technical field of short-start drilling, and discloses a method for intelligently judging the best time for short-start drilling, comprising: S01: obtaining logging related parameters related to drill cuttings accumulation, and performing correlation and normalization processing on each logging related parameter; S02: establishing a normal value prediction model for related logging parameters, and using the model to obtain normal values ​​of related logging parameters; S03: calculating fusion deviation values ​​of logging parameters, comparing the predicted normal values ​​of some logging parameters with the actual parameter measurement values, and calculating the fusion deviation value Δ STF ; S04: Determine whether a short start is required, based on the fusion deviation value Δ STF , determine whether a short start is needed; S05: evaluate the effect of the short start, compare some logging parameters after the short start with the normal values ​​of the prediction model, evaluate the effect of the short start by integrating the degree of deviation, and delete the abnormal values ​​of parameter measurement after the short start. The present invention can accurately determine the timing of the short start, greatly improve the efficiency of the short start, greatly reduce the drilling cost, and has strong stability and fast response speed.

Description

Technical Field

[0001] The invention relates to the technical field of short-start drilling, and specifically discloses a method for intelligently judging the best time for short-start drilling. Background Art

[0002] An important goal of drilling is to increase the drilling speed, shorten the drilling and completion time, and minimize the drilling cost to improve economic benefits. Drilling cuttings are inevitably generated during the drilling process. Excessive drill cuttings will accumulate and cause drill bit jams, making drilling difficult, terminating, and even causing safety accidents such as well kicks and blowouts, increasing unnecessary drilling and completion time, and causing huge economic losses. In order to solve the drill bit jam caused by drill cuttings accumulation and avoid more serious accidents caused by drill bit jams, an effective method is urgently needed to solve this problem.

[0003] According to the literature, short-lift is an effective method to remove the accumulation of cuttings, but it is expensive, and each short-lift is more expensive than the previous one. The number of short-lifts needs to be strictly controlled, which requires finding the most suitable time for short-lifts to reduce drilling costs. When the accumulation of cuttings in the well is enough to cause drill bit sticking and make drilling difficult, and the wellbore pressure is less than the formation pressure, it is the best time for short-lift. However, finding the best time for short-lift is not the best way to improve drilling efficiency and reduce costs. The accumulation of cuttings is related to the poor wellbore cleaning effect caused by the cuttings generated during the drilling process and the low return rate of cuttings. The return of cuttings is directly related to the length of cuttings migration, and the length of cuttings migration is related to the drilling fluid flow rate, annular velocity, and fluid turbulence state. Control parameters to keep the wellbore in a dynamic cleaning state and achieve the best wellbore cleaning effect until the cuttings that cannot be returned by the annulus accumulate to the extent of causing drill bit sticking or the wellbore pressure is about to be less than the formation pressure, and then short-lift is performed. This is the best way to reduce the number of short-lifts. While the methods of wellbore cleaning and judging the wellbore cleaning status are already very rich and complete, there is little research on short-lift to remove rock cuttings. Basically, it is based on experience to judge whether a short-lift is needed, which makes it impossible to accurately judge the timing of the short-lift. Summary of the invention

[0004] The present invention aims to provide a method for intelligently judging the best time for a short-start drilling operation, aiming to solve the problem that the time for a short-start drilling operation cannot be accurately judged.

[0005] In order to achieve the above object, the basic scheme of the present invention is as follows: a method for intelligently judging the best time for short-start drilling, comprising the following steps:

[0006] S01: Obtain logging parameters related to drilling cuttings accumulation

[0007] Perform correlation and normalization processing on various logging related parameters to obtain relevant actual measurement parameter values;

[0008] S02: Establish a normal value prediction model for relevant logging parameters

[0009] Use the model to obtain normal values ​​of relevant logging parameters;

[0010] S03: Calculate the fusion deviation value of logging parameters

[0011] Compare the predicted normal values ​​of some logging parameters with the measured values ​​of relevant actual parameters to calculate the fusion deviation value Δ STF ;

[0012] S04: Determine whether a short start is required

[0013] According to the fusion deviation value Δ STF , determine whether a short start is needed;

[0014] S05: Evaluate the effect of short start

[0015] Some logging parameters after short-term logging are compared with the normal values ​​of the prediction model, the short-term logging effect is evaluated by integrating the degree of deviation, and the abnormal values ​​of parameter measurements are deleted after short-term logging.

[0016] Furthermore, in S01, the logging related parameters are hook load, hook height, standpipe pressure, measured well depth, vertical well depth, drill bit measured depth, and drill bit vertical depth.

[0017] Furthermore, in S02, the established model is a generalized regression neural network model based on the forest optimization algorithm, the model input is the measured well depth, vertical well depth, drill bit measured depth, drill bit vertical depth, and the model output is the big hook load, big hook height, and riser pressure.

[0018] Further, in S03, after obtaining the normal value and the actual measurement value of the logging parameter, the fusion deviation value Δ is calculated by the following formula: STF :

[0019]

[0020]

[0021]

[0022]

[0023] D m is the weight of the hook load, D h is the weight of the big hook height, D p is the weight of casing pressure, M f is the predicted normal value of the hook load, M a is the actual measured value of the hook load, H f H is the predicted normal value of the hook height,a is the actual measured value of the hook height, P f is the predicted normal value of casing pressure, P a is the actual measured value of casing pressure.

[0024] Furthermore, in S04, the fusion deviation value is related to the degree of drill cuttings accumulation and the wellbore pressure, and can directly reflect the drilling status and determine whether a short start is required. The relationship is as follows:

[0025]

[0026] No need to short-start, safe drilling;

[0027]

[0028] If the annular flow rate needs to be adjusted, drilling can continue;

[0029]

[0030] A short start is required.

[0031] Furthermore, in S05, the relationship between the degree of fusion deviation and the short-term effect is as follows:

[0032]

[0033] The short start effect is excellent;

[0034]

[0035] Short start has better effect;

[0036]

[0037] The effect of short start is poor.

[0038] Principles and beneficial effects of the present invention: (1) In this solution, the short start timing is determined by relevant parameters rather than by experience, and compared with the existing method, the short start timing is more accurate.

[0039] (2) In this scheme, a generalized regression neural network based on the forest optimization algorithm is used to predict the normal values ​​of some parameters under abnormal conditions, and after each short-term, the abnormal values ​​are deleted to ensure the accuracy of the predicted normal values;

[0040] (3) In this scheme, the weighted sum of the fusion deviation values ​​of the three parameters of hook load, hook height, and casing pressure can be calculated to accurately reflect the bottom hole drilling status, and the fusion deviation value is derived to determine whether a short start is needed and to control the annular flow; finally, the effect of the short start is determined by the fusion deviation value derivative. Compared with the existing technology, this scheme greatly reduces the number of short starts, greatly improves the efficiency of short starts, greatly reduces the drilling cost, has strong stability, and fast response speed.

[0041] Of course, the implementation of the application plan does not necessarily need to achieve all the technical effects mentioned above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 is a flow chart in an embodiment of the present invention;

[0043] Figure 2 It is a schematic diagram of a short start in an embodiment of the present invention;

[0044] Figure 3 is the measured value of short-start logging in the embodiment of the present invention;

[0045] Figure 4 It is the short start logging prediction value in the embodiment of the present invention. DETAILED DESCRIPTION

[0046] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0047] In the description of the present invention, it is necessary to understand that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0048] The reference numerals in the drawings of the specification include: drill pipe 1, wellbore 2, annulus 3, drill bit 4.

[0049] Example:

[0050] Basically as attached Figure 1 As shown: A method for intelligently judging the best time for drilling short start, comprising the following steps:

[0051] S01: Obtain logging related parameters related to drilling cuttings accumulation. In this embodiment, as shown in the attached Figure 2As shown, it is an actual working condition. The drill pipe 1 and the drill bit 4 form an annulus 3 with the wellbore 2. Cuttings will be generated during the drilling process. Even if the annulus flow is changed to clean the wellbore, larger cuttings will be deposited. In this embodiment, a comprehensive logging system is used to collect the measured well depth, vertical well depth, drill bit measured depth, drill bit vertical depth, big hook load, big hook height, and riser pressure parameters related to the accumulation of drill cuttings, and the related logging parameters are correlated and normalized to obtain the related actual measurement parameters.

[0052] S02: Establish a generalized regression neural network model based on the forest optimization algorithm to predict the normal values ​​of relevant logging parameters. Specifically, the training data is 50,000 groups, and all the abnormal data measured in it are replaced with normal values. The smooth parameters of the generalized regression neural network are optimized by the forest optimization algorithm, which greatly improves the prediction accuracy of the output parameters and makes the predicted values ​​of the output parameters all normal values. In this embodiment, the model input is the measured well depth, vertical well depth, drill bit measured depth, and drill bit vertical depth. The model output is the big hook load, big hook height, and standpipe pressure. Figure 3 , Figure 4 As shown. (A / B / C are respectively: standpipe pressure, hook load, and hook height)

[0053] S03: Compare the predicted normal values ​​of relevant logging parameters with the actual measured parameters to calculate the fusion deviation value Δ STF , the specific calculation method is as follows:

[0054]

[0055]

[0056]

[0057]

[0058] D m is the weight of the hook load, D h is the weight of the big hook height, D p is the weight of casing pressure, M f is the predicted normal value of the hook load, M a is the actual measured value of the hook load, H f H is the predicted normal value of the hook height, a is the actual measured value of the hook height, P f is the predicted normal value of casing pressure, P a is the actual measured value of casing pressure.

[0059] S04: According to the fusion deviation value Δ STF , to determine whether a short start is needed, to determine whether a short start is needed, the relationship is as follows:

[0060]

[0061] No need to short-start, safe drilling;

[0062]

[0063] If the annular flow rate needs to be adjusted, drilling can continue;

[0064]

[0065] A short start is required.

[0066] S05: Compare the relevant logging parameters after short-start with the normal values, and evaluate the short-start effect by integrating the degree of deviation. The relationship between the degree of integration deviation and the short-start effect is as follows:

[0067]

[0068] The short start effect is excellent;

[0069]

[0070] Short start has better effect;

[0071]

[0072] The effect of short start is poor.

[0073] In summary, in this embodiment, the short start timing is determined by relevant parameters rather than by experience, and compared with the existing method, the short start timing is more accurate.

[0074] In this embodiment, a generalized regression neural network based on a forest optimization algorithm is used to predict the normal values ​​of some parameters under abnormal conditions, and after each short-term, the abnormal values ​​are deleted to ensure the accuracy of the predicted normal values;

[0075] In this embodiment, the weighted sum of the fusion deviation values ​​of the three parameters of hook load, hook height, and casing pressure is calculated to accurately reflect the bottom hole drilling status, and the fusion deviation value is derived to determine whether a short start is needed and to control the annular flow rate; finally, the effect of the short start is determined by the fusion deviation value derivative. Compared with the prior art, this solution greatly reduces the number of short starts, greatly improves the efficiency of short starts, greatly reduces the drilling cost, has strong stability, and fast response speed.

[0076] The above is only an embodiment of the present invention, and the common knowledge such as the known specific structure and / or characteristics in the scheme is not described in detail here. It should be pointed out that for those skilled in the art, several deformations and improvements can be made without departing from the structure of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A method for intelligently judging the best time for drilling short start, characterized by: The following steps are involved: S01: Obtain logging parameters related to drilling cuttings accumulation Perform correlation and normalization processing on various logging related parameters to obtain the measured values ​​of related actual parameters; Specifically: Use the integrated logging system to collect the measured well depth, vertical well depth, drill bit measured depth, drill bit vertical depth, hook load, hook height, and riser pressure parameters related to drill cuttings accumulation, and perform correlation and normalization processing on each logging related parameter to obtain the relevant actual measurement parameters; S02: Establish a normal value prediction model for relevant logging parameters The model is used to obtain the normal values ​​of relevant logging parameters; the established model is a generalized regression neural network model based on the forest optimization algorithm. The model input is the measured well depth, vertical well depth, drill bit measured depth, drill bit vertical depth, and the model output is the big hook load, big hook height, and standpipe pressure. S03: Calculate the fusion deviation value of logging parameters Compare the predicted normal values ​​of some logging parameters with the measured values ​​of relevant actual parameters to calculate the fusion deviation value Δ STF ; The fusion deviation value Δ is calculated by the following formula STF : D m is the weight of the hook load, D h is the weight of the big hook height, D p is the weight of casing pressure, M f is the predicted normal value of the hook load, M a is the actual measured value of the hook load, H f H is the predicted normal value of the hook height, a is the actual measured value of the hook height, P f is the predicted normal value of casing pressure, P a is the actual measured value of casing pressure; S04: Determine whether a short start is required According to the fusion deviation value Δ STF , determine whether a short start is needed; The fusion deviation value is related to the degree of cuttings accumulation and wellbore pressure. It can directly reflect the drilling status and determine whether a short start is needed. The relationship is as follows: No need to short-start, safe drilling; If the annular flow rate needs to be adjusted, drilling can continue; A short start is required; S05: Evaluate the effect of short start Compare some logging parameters after short-term logging with the normal values ​​of the prediction model, evaluate the short-term logging effect by integrating the degree of deviation, and delete the abnormal values ​​of parameter measurements after short-term logging; The relationship between the degree of fusion deviation and the short-term effect is as follows: The short start effect is excellent; Short start has better effect; The effect of short start is poor.

Images