Drilling inefficient well section identification method based on joint constraint of reference drilling speed and energy utilization efficiency

Abstract

This invention relates to a method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency, belonging to the field of oil and gas field drilling engineering and drilling and logging data processing. The method includes: acquiring and preprocessing data to obtain preprocessed sample data; calculating the mechanical specific energy, confining pressure resistance, and energy utilization efficiency coefficient corresponding to each sampling point at each depth; constructing a high-efficiency sample subset; training a reference drilling rate model using drilling data and formation parameters from the high-efficiency sample subset as input; inputting the preprocessed sample data into the trained reference drilling rate model to obtain the reference drilling rate; calculating the drilling rate efficiency coefficient based on the actual mechanical drilling rate and the reference drilling rate; jointly identifying inefficient points based on the drilling rate efficiency coefficient and the energy utilization efficiency coefficient; and merging the inefficient points to obtain the inefficient well section identification result. This invention can reduce misjudgments caused by judging inefficient well sections solely based on the actual mechanical drilling rate or mechanical specific energy, and improve the accuracy of inefficient well section identification.

Description

Technical Field

[0001] This invention relates to a method for identifying inefficient well sections based on a combined constraint of reference drilling rate and energy utilization efficiency, belonging to the field of oil and gas field drilling engineering and drilling and logging data processing technology. Background Technology

[0002] In oil and gas drilling, the rate of penetration (RLP) is a crucial indicator of drilling efficiency. Influenced by various factors such as formation lithology, rock strength, bottom hole pressure, drill bit condition, drilling parameter combinations, and bottom hole cleanliness, the actual RLP varies significantly across different sections of the same well. Field operations typically assess drilling efficiency by observing actual RLP curves, mechanical energy curves, or changes in engineering parameters. However, these methods rely heavily on empirical interpretation and are easily affected by variations in formation drillability.

[0003] For example, in high-strength formations, the actual mechanical drilling rate is naturally lower. Judging solely based on the low mechanical drilling rate might misclassify normally difficult-to-drill sections as inefficient engineering sections. Conversely, in soft formations or sections with abrupt changes in local parameters, although the actual mechanical drilling rate may not be low, if the mechanical energy consumption is significantly higher, inefficiencies such as drill bit wear, unreasonable parameter combinations, insufficient rock clearing, or decreased energy transfer efficiency may still exist. Therefore, a single drilling rate index or a single mechanical energy index is insufficient to accurately distinguish between "low drilling rate caused by difficult formations" and "inefficient drilling caused by insufficient engineering efficiency."

[0004] Existing technologies also employ methods to evaluate drilling efficiency using mechanical specific energy, drilling rate trends, or statistical thresholds. However, most methods fail to construct a reference drilling rate representing a normal, efficient drilling state, making it difficult to quantify the degree of loss in drilling efficiency relative to the expected rate. Furthermore, conventional mechanical specific energy methods do not adequately consider the influence of rock strength and confining pressure conditions, resulting in a lack of comparability in energy consumption levels between different formations. Therefore, there is an urgent need for a method to identify inefficient drilling sections that can utilize real-time drilling data, formation mechanical parameters, and machine learning models to comprehensively consider both the degree of reference drilling rate achievement and the intensity-normalized energy utilization status. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides a method for identifying inefficient well sections based on a joint constraint of reference drilling rate and energy utilization efficiency. This method solves the problem that existing technologies, which rely solely on actual mechanical drilling rate or mechanical specific energy to determine inefficient well sections, are easily affected by formation differences and lack sufficient accuracy in identifying inefficiencies.

[0006] The technical solution of the present invention is as follows:

[0007] A method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency includes the following steps:

[0008] S1: Obtain the original drilling data, formation parameter data, and drill bit parameter data of the target well section, and preprocess the original drilling data, formation parameter data, and drill bit parameter data to obtain preprocessed sample data;

[0009] S2, based on the preprocessed sample data, calculate the mechanical specific energy, confining pressure compressive strength and energy utilization efficiency coefficient corresponding to each depth sampling point;

[0010] S3, based on the quantile of actual mechanical drilling speed and the quantile of energy utilization efficiency coefficient, constructs an efficient sample subset;

[0011] S4 uses drilling data and formation parameters from a high-efficiency sample subset as input features and actual mechanical drilling rate as output to train a reference drilling rate model.

[0012] S5, input the preprocessed sample data in the target well section into the trained reference drilling rate model to obtain the reference drilling rate corresponding to each depth sampling point;

[0013] S6, calculate the drilling efficiency coefficient based on the actual mechanical drilling speed and the reference drilling speed;

[0014] S7, based on the drilling rate efficiency coefficient and energy utilization efficiency coefficient, a dual-index joint judgment is made to obtain the inefficient points;

[0015] S8. The inefficient points are merged based on a continuous depth window to obtain the inefficient well section identification result.

[0016] Preferably, in step S1, the original drilling data includes at least one of the following: well depth, actual mechanical drilling speed, drilling pressure, rotational speed, torque, displacement, stand pressure, drilling fluid density, well inclination, azimuth, and cumulative footage of the drill bit.

[0017] The formation parameter data includes at least one of the following: natural gamma, sonic transit time, lithology, uniaxial compressive strength, bottom hole pressure, and pore pressure.

[0018] The drill bit parameter data includes at least one of the following: drill bit diameter, drill bit model, drill bit type, and well section in which the drill bit is used;

[0019] Preferably, the preprocessing includes at least one of depth alignment, outlier removal, missing value handling, unit unification, and smoothing.

[0020] Depth alignment unifies raw drilling data, formation parameter data, and drill bit parameter data from different sampling frequencies and data sources into the same depth grid. The depth grid refers to a continuous depth sequence formed according to a preset depth interval, with the well depth as a unified reference. For example, unified sampling points are established according to a depth interval of 0.5m or 1m, and data from different sampling frequencies and sources are mapped to the same depth sampling point through interpolation, nearest neighbor matching, or interval assignment.

[0021] Outlier removal includes removing data points corresponding to actual mechanical drilling speed less than or equal to zero, drilling pressure less than or equal to zero, rotation speed less than or equal to zero, torque less than or equal to zero, and abnormal drilling states. Abnormal drilling states include at least one of the following: connecting a single joint, stopping drilling, circulating, tripping in and out of the drill string, reaming, and reverse reaming.

[0022] Missing value handling includes: for continuous parameters such as well depth, drilling pressure, rotation speed, torque, displacement, standpipe pressure, actual mechanical drilling speed, natural gamma, sonic transit time, bottom hole pressure, and pore pressure, linear interpolation, nearest effective value filling, or sliding window midpoint filling are used; for discrete parameters such as lithology, drill bit model, and drill bit type, the corresponding category of adjacent depth points within the same well section or the well section record of the drill bit is used for filling; for data segments with continuous missing lengths exceeding the preset depth length or that cannot be reliably filled, the corresponding sample points are removed.

[0023] Preferably, in step S2, the formula for calculating the mechanical specific energy is as follows:

[0024]

[0025] In the formula, For the first Mechanical specific energy of each depth sampling point; For drilling pressure; This represents the rock-breaking area of ​​the drill bit, and ; The diameter of the drill bit; Rotational speed; Torque; The actual mechanical drilling rate used to calculate the mechanical specific energy; The unit conversion factor is determined based on the units used for drilling pressure, torque, rotational speed, mechanical drilling speed, drill bit area, and mechanical specific energy. It is used to ensure that the calculated mechanical specific energy has a unified dimension. When torque is expressed in N·m, rotational speed in r / min, mechanical drilling speed in m / h, drill bit rock-breaking area in m², and mechanical specific energy in Pa, the unit conversion factor can be taken as... When the mechanical specific energy is expressed in MPa or the input parameters are in engineering units such as kN or kN·m, the above conversion can be further multiplied by the corresponding unit conversion factor.

[0026] To avoid an abnormal increase in mechanical specific energy when the actual mechanical drilling speed approaches zero, The value can be:

[0027]

[0028] In the formula, This refers to the actual mechanical drilling speed. The preset minimum mechanical drilling speed;

[0029] The formula for calculating confining pressure compressive strength is as follows:

[0030]

[0031] in, ;

[0032] In the formula, For the first The confining pressure resistance at each depth sampling point; Uniaxial compressive strength; This refers to the bottom hole pressure. Pore ​​pressure; The effective confining pressure difference at the bottom of the well; The confining pressure correction factor is used to characterize the sensitivity of rock compressive strength to increase with the increase of effective confining pressure at the bottom of the well. It can be obtained by fitting the results of triaxial compression test of cores in the target block, or it can be determined based on the known rock mechanical parameters of the target block or adjacent blocks.

[0033] The formula for calculating the energy efficiency coefficient is as follows:

[0034]

[0035] In the formula, or For the first The energy utilization efficiency coefficient of each depth sampling point indicates that the higher the energy consumption per unit strength under the corresponding rock strength and confining pressure conditions, the worse the energy utilization efficiency.

[0036] Preferably, the confining pressure correction factor is determined based on the Mohr-Coulomb strength criterion. The expression is:

[0037]

[0038] In the formula, It is the internal friction angle of the rock.

[0039] When data on the internal friction angle of rock or triaxial compression test are unavailable, the confining pressure correction factor... The confining pressure correction factor can be obtained from experience or by correction through adjacent well sections. For the same lithology or the same stratigraphic group, the same confining pressure correction factor can be used; for different lithologies or different stratigraphic groups, different confining pressure correction factors can be set respectively.

[0040] Preferably, the implementation process of step S3 is as follows:

[0041] S31. The preprocessed sample data are grouped according to at least one of the following: lithology, stratigraphic position, drill bit type, drill bit size, and well section in which the drill bit is used, to obtain several sample groups g.

[0042] S32, within each sample group g, sort the actual mechanical drilling rate data in ascending order of value, and take the nth... The quantile of the actual mechanical drilling rate is used as the high quantile threshold of the actual mechanical drilling rate, denoted as . ;

[0043] S33, within each sample group g, the energy utilization efficiency coefficients are sorted in ascending order of value, and the nth value is selected. The quantile energy efficiency coefficient is used as the quantile threshold of the energy efficiency coefficient, denoted as . ;

[0044] S34. Samples that simultaneously meet the following conditions will be considered as an efficient sample subset:

[0045] and .

[0046] Preferably, in step S31, the grouping criteria are used to ensure that the data within the same sample group have similar formation drillability, drill bit conditions, and drilling conditions; primary grouping is prioritized according to lithology or formation position, followed by secondary grouping according to drill bit type, drill bit size, or well section where the drill bit is used; when the number of samples within the same group is insufficient, adjacent well sections with similar lithology or the same drill bit conditions can be merged into the same sample group;

[0047] In step S32, Take a value of 0.70 to 0.85, in step S33, Take a value of 0.50 to 0.70.

[0048] The efficient sample subset of the present invention is not obtained directly from the original drilling data, but is determined by the actual mechanical drilling rate quantile and the energy utilization efficiency coefficient quantile after depth alignment, outlier removal, missing value processing and energy utilization efficiency coefficient calculation.

[0049] Preferably, in step S4, the reference drilling speed model includes one of the following: random forest regression model, gradient boosting tree model, support vector regression model, neural network model, multivariate regression model, and quantile regression model; all of the above models are existing machine learning models. The present invention is based on training the reference drilling speed model with an efficient sample subset and limiting the model input features and output targets, so that the model output can represent the reference drilling speed under similar formation, drill bit, and working conditions.

[0050] The input features of the reference drilling rate model include at least one of the following: drilling pressure, rotational speed, torque, displacement, stand pressure, well depth, natural gamma, sonic transit time, lithology, uniaxial compressive strength, bottom hole pressure, pore pressure, drilling fluid density, and cumulative footage of the drill bit.

[0051] The output of the reference drilling speed model is the reference drilling speed. .

[0052] In this invention, the input features of the reference drilling speed model do not include actual mechanical drilling speed, mechanical specific energy, and energy utilization efficiency coefficient, in order to avoid information leakage caused by the target variable or derived indicators containing the target variable entering the model.

[0053] In step S6, the formula for calculating the drilling speed efficiency coefficient is:

[0054]

[0055] In the formula, For the first Drilling efficiency coefficient at each depth sampling point; This refers to the actual mechanical drilling speed; For reference drilling speed. The smaller the value, the more significant the loss in actual mechanical drilling speed compared to the reference drilling speed.

[0056] Preferably, in step S7, an inefficient point is determined when the following conditions are met:

[0057] and ;

[0058] In the formula, To preset the drilling efficiency threshold, a fixed empirical threshold can be used, such as 0.65, or the statistical quantile threshold of the drilling efficiency coefficient of the target well section can be used, such as the 25th quantile of the drilling efficiency coefficient of the target well section.

[0059] To preset the energy utilization efficiency threshold, a high quantile threshold of the energy utilization efficiency coefficient of the target well section can be taken, such as the 80th quantile. The high quantile threshold refers to the value that is located in a higher quantile position after sorting the energy utilization efficiency coefficient of the target well section from smallest to largest. For example, the 80th quantile means that the energy utilization efficiency coefficient of about 80% of the sample points is not greater than this value, and about 20% of the sample points are higher than this value.

[0060] Preferably, in step S8, the merging process is as follows:

[0061] Set the length of the continuous depth window and inefficient point ratio threshold ;

[0062] When continuous depth window The ratio of the number of inefficient points within the window to the total number of sampling points within the window is greater than or equal to... When the continuous depth window corresponds to an inefficient well section, the well section is identified as an inefficient well section. (When the ratio of the number of inefficient points in the continuous depth window to the total number of sampling points in the window is less than the preset inefficient point ratio threshold, it is considered that the distribution of inefficient points in the continuous depth window is insufficient to constitute a continuous inefficient well section, and the well section corresponding to the window is not identified as an inefficient well section.)

[0063] When the interval between adjacent inefficient well sections is less than the preset merging interval Adjacent inefficient well sections are merged, among which... The sampling interval can be set according to the target well section and the on-site identification accuracy requirements, for example, 5m, 10m or 20m; the inefficient point ratio threshold can be set to 0.50, 0.60 or 0.70. The preset merging interval between adjacent inefficient well sections can be set to 3m, 5m or 10m.

[0064] The results of inefficient well section identification include at least one of the following: inefficient well section number, starting depth, ending depth, well section length, average actual mechanical drilling rate, average reference drilling rate, average drilling rate efficiency coefficient, average energy utilization efficiency coefficient, percentage of inefficient points, and inefficiency level.

[0065] Furthermore, the inefficiency level is determined based on the fact that the well section already meets the inefficiency section identification criteria. Inefficiency levels are divided into mild inefficiency, moderate inefficiency, and severe inefficiency. If the proportion of inefficiency points in a continuous depth window or a merged well section is lower than a preset inefficiency point proportion threshold, it will not be output as an inefficiency section, or it will be marked as a suspected inefficiency section and will not participate in the mild, moderate, and severe inefficiency classification. For identified inefficiency sections, the inefficiency level can be divided according to the average drilling efficiency coefficient, and the degree of inefficiency can be further explained by combining the average energy utilization efficiency coefficient and the proportion of inefficiency points; specifically, when the average drilling efficiency coefficient is less than the first-level threshold, it is determined to be severely inefficient; when the average drilling efficiency coefficient is greater than or equal to the first-level threshold and less than the second-level threshold, it is determined to be moderately inefficient; when the average drilling efficiency coefficient is greater than or equal to the second-level threshold and the well section still meets the inefficiency section identification criteria, it is determined to be mildly inefficient. The first and second level thresholds can be determined based on the distribution of the drilling efficiency coefficient of the target well section, field experience, or statistical quantiles.

[0066] The average energy utilization efficiency coefficient is used to assist in judging the inefficiency level. When the average energy utilization efficiency coefficient is higher than the 80th percentile of the energy utilization efficiency coefficient of the target well section, it indicates that there is a significant energy underutilization in the inefficient well section.

[0067] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-described method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency.

[0068] A computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the above-described method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency.

[0069] For any details not covered in this invention, please refer to the prior art.

[0070] The beneficial effects of this invention are as follows:

[0071] 1. This invention establishes a reference drilling rate model based on a high-efficiency sample subset, which can obtain the reference drilling rate of the target well section under similar formation and operating conditions. The drilling rate efficiency coefficient is constructed by the ratio of the actual mechanical drilling rate to the reference drilling rate, thereby quantitatively evaluating the degree of loss of the actual drilling rate relative to the high-efficiency drilling state.

[0072] 2. This invention constructs an energy utilization efficiency coefficient based on mechanical specific energy and confining pressure compressive strength, thereby achieving a normalized characterization of the mechanical energy consumption state under different formation strength conditions.

[0073] 3. By combining the drilling speed efficiency coefficient and the energy utilization efficiency coefficient, this invention can effectively distinguish between low drilling speed caused by difficult formation drilling and inefficient drilling caused by factors such as mismatch of engineering parameters, decreased rock breaking efficiency of drill bit, or insufficient cleaning of the bottom of well.

[0074] 4. This invention merges inefficient points through a continuous depth window, which can reduce the interference of isolated anomalies on the identification results and output continuous inefficient well sections, providing a basis for drilling parameter optimization and on-site speed and efficiency improvement.

[0075] In summary, this invention can reduce misjudgments caused by judging inefficient well sections solely based on actual mechanical drilling rate or mechanical specific energy, and effectively improve the accuracy of inefficient well section identification. Attached Figure Description

[0076] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute an undue limitation of this application.

[0077] Figure 1 This is a flowchart of the method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency according to the present invention.

[0078] Figure 2This is a flowchart illustrating the efficient sample subset construction and reference drilling speed model training process in an embodiment of the present invention.

[0079] Figure 3 This is a comparison chart of the actual mechanical drilling rate and the reference drilling rate in an embodiment of the present invention;

[0080] Figure 4 The graphs show the variation of drilling efficiency coefficient and energy utilization efficiency coefficient with well depth in an embodiment of the present invention, where (a) is the energy utilization efficiency coefficient and (b) is the drilling efficiency coefficient.

[0081] Figure 5 This is a diagram showing the results of identifying inefficient well sections according to an embodiment of the present invention. Detailed Implementation

[0082] The present invention is illustrated below with examples to further illustrate the invention in detail, but the implementation of the invention is not limited thereto. For those skilled in the art, the present invention can be modified and varied in various ways. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0083] Example 1

[0084] A method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency, such as... Figures 1 to 5 As shown, it includes the following steps:

[0085] S1: Obtain the original drilling data, formation parameter data, and drill bit parameter data of the target well section, and preprocess the original drilling data, formation parameter data, and drill bit parameter data to obtain preprocessed sample data;

[0086] S2, based on the preprocessed sample data, calculate the mechanical specific energy, confining pressure compressive strength and energy utilization efficiency coefficient corresponding to each depth sampling point;

[0087] S3, based on the quantile of actual mechanical drilling speed and the quantile of energy utilization efficiency coefficient, constructs an efficient sample subset;

[0088] S4 uses drilling data and formation parameters from a high-efficiency sample subset as input features and actual mechanical drilling rate as output to train a reference drilling rate model.

[0089] S5, input the preprocessed sample data in the target well section into the trained reference drilling rate model to obtain the reference drilling rate corresponding to each depth sampling point;

[0090] S6, calculate the drilling efficiency coefficient based on the actual mechanical drilling speed and the reference drilling speed;

[0091] S7, based on the drilling rate efficiency coefficient and energy utilization efficiency coefficient, a dual-index joint judgment is made to obtain the inefficient points;

[0092] S8. The inefficient points are merged based on a continuous depth window to obtain the inefficient well section identification result.

[0093] Example 2

[0094] A method for identifying inefficient well sections based on the joint constraints of reference drilling rate and energy utilization efficiency, as described in Example 1, except that in step S1, the original drilling data includes at least one of well depth, actual mechanical drilling rate, drilling pressure, rotation speed, torque, displacement, stand pressure, drilling fluid density, well inclination, azimuth, and cumulative footage of the drill bit.

[0095] The formation parameter data includes at least one of the following: natural gamma, sonic transit time, lithology, uniaxial compressive strength, bottom hole pressure, and pore pressure.

[0096] The drill bit parameter data includes at least one of the following: drill bit diameter, drill bit model, drill bit type, and well section in which the drill bit is used.

[0097] Example 3

[0098] A method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency, as described in Example 2, except that the preprocessing includes at least one of depth alignment, outlier removal, missing value processing, unit unification, and smoothing processing;

[0099] Depth alignment unifies raw drilling data, formation parameter data, and drill bit parameter data from different sampling frequencies and data sources into the same depth grid. The depth grid refers to a continuous depth sequence formed at 0.5m depth intervals with well depth as a unified reference. Data from different sampling frequencies and sources are mapped to the same depth sampling point through interpolation, nearest neighbor matching, or interval assignment.

[0100] Outlier removal includes removing data points corresponding to actual mechanical drilling speed less than or equal to zero, drilling pressure less than or equal to zero, rotation speed less than or equal to zero, torque less than or equal to zero, and abnormal drilling states. Abnormal drilling states include at least one of the following: connecting a single joint, stopping drilling, circulating, tripping in and out of the drill string, reaming, and reverse reaming.

[0101] Missing value handling includes: for continuous parameters such as well depth, drilling pressure, rotation speed, torque, displacement, standpipe pressure, actual mechanical drilling speed, natural gamma, sonic transit time, bottom hole pressure, and pore pressure, linear interpolation, nearest effective value filling, or sliding window midpoint filling are used; for discrete parameters such as lithology, drill bit model, and drill bit type, the corresponding category of adjacent depth points within the same well section or the well section record of the drill bit is used for filling; for data segments with continuous missing lengths exceeding the preset depth length or that cannot be reliably filled, the corresponding sample points are removed.

[0102] Example 4

[0103] A method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency, as described in Example 3, differs in that the mechanical specific energy is calculated using the following formula in step S2:

[0104]

[0105] In the formula, For the first Mechanical specific energy of each depth sampling point; For drilling pressure; This represents the rock-breaking area of ​​the drill bit, and ; The diameter of the drill bit; Rotational speed; Torque; The actual mechanical drilling rate used to calculate the mechanical specific energy; The unit conversion factor is determined based on the units used for drilling pressure, torque, rotational speed, mechanical drilling speed, drill bit area, and mechanical specific energy. It is used to ensure that the calculated mechanical specific energy has a unified dimension. When torque is expressed in N·m, rotational speed in r / min, mechanical drilling speed in m / h, drill bit rock-breaking area in m², and mechanical specific energy in Pa, the unit conversion factor can be taken as... When the mechanical specific energy is expressed in MPa or the input parameters are in engineering units such as kN or kN·m, the above conversion can be further multiplied by the corresponding unit conversion factor.

[0106] To avoid an abnormal increase in mechanical specific energy when the actual mechanical drilling speed approaches zero, The value can be:

[0107]

[0108] In the formula, This refers to the actual mechanical drilling speed. The preset minimum mechanical drilling speed;

[0109] The formula for calculating confining pressure compressive strength is as follows:

[0110]

[0111] in, ;

[0112] In the formula, For the first The confining pressure resistance at each depth sampling point; Uniaxial compressive strength; This refers to the bottom hole pressure. Pore ​​pressure; The effective confining pressure difference at the bottom of the well; The confining pressure correction factor characterizes the sensitivity of rock compressive strength to the increase of effective bottom-hole confining pressure. In this embodiment, the confining pressure correction factor is determined based on the Mohr-Coulomb strength criterion. The expression is:

[0113]

[0114] In the formula, It is the internal friction angle of the rock.

[0115] The formula for calculating the energy efficiency coefficient is as follows:

[0116]

[0117] In the formula, or For the first The energy utilization efficiency coefficient of each depth sampling point indicates that the higher the energy consumption per unit strength under the corresponding rock strength and confining pressure conditions, the worse the energy utilization efficiency.

[0118] Example 5

[0119] A method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency, as described in Example 4, differs in that the implementation process of step S3 is as follows:

[0120] S31. The preprocessed sample data are grouped according to at least one of the following: lithology, stratigraphic position, drill bit type, drill bit size, and well section in which the drill bit is used, to obtain several sample groups g.

[0121] The grouping criteria are used to ensure that the data within the same sample group have similar formation drillability, drill bit conditions, and drilling conditions; primary grouping is prioritized according to lithology or formation position, followed by secondary grouping according to drill bit type, drill bit size, or well section where the drill bit is used; when the number of samples in the same group is insufficient, adjacent well sections with similar lithology or the same drill bit conditions can be merged into the same sample group.

[0122] S32, within each sample group g, sort the actual mechanical drilling rate data in ascending order of value, and take the nth... The quantile of the actual mechanical drilling rate is used as the high quantile threshold of the actual mechanical drilling rate, denoted as . ;

[0123] S33, within each sample group g, the energy utilization efficiency coefficients are sorted in ascending order of value, and the nth value is selected. The quantile energy efficiency coefficient is used as the quantile threshold of the energy efficiency coefficient, denoted as . ;

[0124] In this embodiment, Take 0.70, Take 0.60.

[0125] S34. Samples that simultaneously meet the following conditions will be considered as an efficient sample subset:

[0126] and .

[0127] Example 6

[0128] A method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency, as described in Example 5, except that in step S4, the reference drilling rate model is a random forest regression model.

[0129] Using drilling pressure, rotational speed, torque, well depth, natural gamma, sonic transit time, uniaxial compressive strength, bottom hole pressure, and pore pressure from a high-efficiency sample subset as input features, and actual mechanical drilling rate as output, a random forest regression model is trained as a reference drilling rate model. Figure 2 As shown;

[0130] Input all preprocessed samples from the target well section into the reference drilling rate model to obtain the reference drilling rate corresponding to each depth sampling point. ;

[0131] In this embodiment, the input features of the reference drilling speed model do not include actual mechanical drilling speed, mechanical specific energy, and energy utilization efficiency coefficient, in order to avoid information leakage caused by the target variable or derived indicators containing the target variable entering the model.

[0132] Example 7

[0133] A method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency, as described in Example 6, differs in that the calculation formula for the drilling rate efficiency coefficient in step S6 is as follows:

[0134]

[0135] In the formula, For the first Drilling efficiency coefficient at each depth sampling point; This refers to the actual mechanical drilling speed; For reference drilling speed. The smaller the value, the more significant the loss in actual mechanical drilling speed compared to the reference drilling speed.

[0136] Example 8

[0137] A method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency, as described in Example 7, differs in that, in step S7, a section is determined to be inefficient when the following conditions are met:

[0138] and ;

[0139] In the formula, To preset the drilling speed efficiency threshold, this embodiment uses a fixed empirical threshold of 0.65;

[0140] To preset the energy utilization efficiency threshold, the 80th quantile is taken. The high quantile threshold refers to the value that is located in a higher quantile position after sorting the energy utilization efficiency coefficient of the target well section from smallest to largest.

[0141] Example 9

[0142] A method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency, as described in Example 8, differs in that the merging process in step S8 is as follows:

[0143] Set the length of the continuous depth window and inefficient point ratio threshold ;

[0144] When continuous depth window The ratio of the number of inefficient points within the window to the total number of sampling points within the window is greater than or equal to... When the continuous depth window corresponds to an inefficient well section, the well section is identified as an inefficient well section. (When the ratio of the number of inefficient points in the continuous depth window to the total number of sampling points in the window is less than the preset inefficient point ratio threshold, it is considered that the distribution of inefficient points in the continuous depth window is insufficient to constitute a continuous inefficient well section, and the well section corresponding to the window is not identified as an inefficient well section.)

[0145] In this embodiment, the continuous depth window length can be set to 10m, the inefficient point ratio threshold can be set to 0.60, and the preset merging interval can be set to 5m.

[0146] The results of inefficient well section identification include at least one of the following: inefficient well section number, starting depth, ending depth, well section length, average actual mechanical drilling rate, average reference drilling rate, average drilling rate efficiency coefficient, average energy utilization efficiency coefficient, percentage of inefficient points, and inefficiency level.Figure 5 As shown, Figure 5 The oblique or shaded areas in the table represent inefficient well sections obtained after joint identification using two indicators and merging through continuous depth windows. The starting and ending depths of these inefficient well sections correspond to the inefficient well section identification results in Table 1, including 1788m to 1798m, 2304m to 2317m, 2325m to 2360m, 2844m to 2861m, 2880m to 2911m, 2922m to 2932m, 2962m to 3124m, 3203m to 3222m, 3435m to 3447m, and 3591m to 3723m. Figure 5 As can be seen, the identification result of this embodiment is not an isolated anomaly, but an inefficient well section formed within a continuous depth range.

[0147] The inefficiency level is further determined on the basis that the inefficiency well section has already met the identification conditions; for continuous depth windows where the proportion of inefficiency points is lower than the preset inefficiency point proportion threshold, they are not output as inefficiency well sections, or are only considered as suspected inefficiency windows, and are not included in the classification of mild inefficiency, moderate inefficiency and severe inefficiency.

[0148] For identified inefficient well sections, the classification is primarily based on the average drilling efficiency coefficient (RAC) of the section, supplemented by the average energy utilization efficiency coefficient (EEC) and the percentage of inefficient points to further illustrate the degree of inefficiency. In this embodiment, the first-level threshold is 0.40, and the second-level threshold is 0.65. When the RAC is less than 0.40, it is classified as severely inefficient; when the RAC is greater than or equal to 0.40 and less than 0.65, it is classified as moderately inefficient; and when the RAC is greater than or equal to 0.65 and the section still meets the criteria for identifying inefficient well sections, it is classified as slightly inefficient. The average energy utilization efficiency coefficient characterizes the energy consumption state per unit rock strength condition of the inefficient section, and the percentage of inefficient points characterizes the persistence of inefficiency within that section.

[0149] Example 10

[0150] A method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency, using the target section of well A as an example, includes the following steps:

[0151] S1. Obtain comprehensive drilling data for the target section of well A, and perform depth alignment on data such as well depth, actual mechanical drilling speed, drilling pressure, rotation speed, torque, drill bit diameter, bottom hole pressure, pore pressure, and uniaxial compressive strength; remove data points corresponding to actual mechanical drilling speed less than or equal to zero, drilling pressure less than or equal to zero, rotation speed less than or equal to zero, and abnormal drilling conditions to form preprocessed sample data.

[0152] S2. Based on the preprocessed sample data, calculate the mechanical specific energy (MSE), confining pressure compressive strength (CCS), and energy utilization efficiency coefficient (E). Specifically, the mechanical specific energy is used to characterize the mechanical energy input during rock breaking, the confining pressure compressive strength is used to characterize the formation strength after considering the influence of bottom hole pressure and pore pressure, and the energy utilization efficiency coefficient (E) after strength normalization is obtained by the ratio of MSE to CCS.

[0153] S3. Preprocessed samples are grouped according to lithology, drill bit type, or well section window. Within each group, samples with actual mechanical drilling speed not lower than the 70th quantile of that group and energy utilization efficiency coefficient not higher than the 60th quantile of that group are selected as high-efficiency samples. The resulting subset of high-efficiency samples is used to train the reference drilling speed model.

[0154] S4. A random forest regression model is used to establish a reference drilling rate model. The model input features include drilling pressure, rotational speed, torque, well depth, uniaxial compressive strength, bottom hole pressure, and pore pressure; the model output is the reference drilling rate. After the model training is completed, the input features corresponding to the reference drilling rate model from all preprocessed samples of the target well section of well A are input into the model to obtain the reference drilling rate corresponding to each depth sampling point.

[0155] S5: Calculate the drilling efficiency coefficient And set the inefficiency point discrimination condition as follows: A depth sampling point that meets the above dual-indicator joint discrimination criteria is marked as an inefficient point if its energy utilization efficiency coefficient E is less than 0.65 and is higher than the 80th quantile of the target well section.

[0156] S6: Set the continuous depth window length to be no less than 10m, and the proportion of inefficient points within the window to be no less than 60%. When the interval between adjacent inefficient well sections is less than the preset merging interval, the adjacent inefficient well sections are merged to obtain the inefficient well section identification results of the target well section of well A, as shown in Table 1.

[0157] Table 1. Identification results of inefficient sections in target well section A.

[0158]

[0159] As shown in Table 1, a total of 10 inefficient well sections were identified in the target well section of Well A. These inefficient well sections all exhibited an average drilling rate efficiency coefficient lower than that of normal high-efficiency drilling, while their average energy utilization efficiency coefficient was relatively high. This indicates that the actual drilling process experienced a significant drilling rate loss compared to the reference drilling rate, accompanied by high intensity-normalized mechanical energy consumption.

[0160] The seventh inefficient well section, from 2962m to 3124m, is 162m long, with an average actual mechanical drilling rate of 8.23m / h, an average reference drilling rate of 27.74m / h, and an average drilling efficiency coefficient of 0.296. The inefficient point ratio is 93.79%, indicating that this section is a continuous inefficient section rather than an isolated anomaly. The tenth inefficient well section, from 3591m to 3723m, is 132m long, with an average energy utilization efficiency coefficient of 13.68 and an inefficient point ratio of 98.86%. This indicates that the mechanical energy consumption per unit rock strength condition is significantly higher in this section, potentially due to decreased drill bit rock breaking efficiency, unreasonable drilling parameter combinations, or insufficient bottom hole cleaning.

[0161] Example 11

[0162] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any one of the methods for identifying inefficient well sections based on the joint constraints of reference drilling rate and energy utilization efficiency as described in Examples 1-9.

[0163] Example 12

[0164] A computer-readable storage medium for storing a computer program, which, when executed by a processor, implements any of the methods for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency as described in Examples 1-9.

[0165] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency, characterized in that, Includes the following steps: S1: Obtain the original drilling data, formation parameter data, and drill bit parameter data of the target well section, and preprocess the original drilling data, formation parameter data, and drill bit parameter data to obtain preprocessed sample data; S2, based on the preprocessed sample data, calculate the mechanical specific energy, confining pressure compressive strength and energy utilization efficiency coefficient corresponding to each depth sampling point; S3, based on the quantile of actual mechanical drilling speed and the quantile of energy utilization efficiency coefficient, constructs an efficient sample subset; S4 uses drilling data and formation parameters from a high-efficiency sample subset as input features and actual mechanical drilling rate as output to train a reference drilling rate model. S5, input the preprocessed sample data in the target well section into the trained reference drilling rate model to obtain the reference drilling rate corresponding to each depth sampling point; S6, calculate the drilling efficiency coefficient based on the actual mechanical drilling speed and the reference drilling speed; S7, based on the drilling rate efficiency coefficient and energy utilization efficiency coefficient, a dual-index joint judgment is made to obtain the inefficient points; S8. The inefficient points are merged based on a continuous depth window to obtain the inefficient well section identification result.

2. The method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency as described in claim 1, characterized in that, In step S1, the original drilling data includes at least one of the following: well depth, actual mechanical drilling speed, drilling pressure, rotational speed, torque, displacement, stand pressure, drilling fluid density, well inclination, azimuth, and cumulative footage of the drill bit. The formation parameter data includes at least one of the following: natural gamma, sonic transit time, lithology, uniaxial compressive strength, bottom hole pressure, and pore pressure. The drill bit parameter data includes at least one of the following: drill bit diameter, drill bit model, drill bit type, and well section in which the drill bit is used; Preprocessing includes at least one of depth alignment, outlier removal, missing value handling, unit unification, and smoothing. Depth alignment unifies raw drilling data, formation parameter data, and drill bit parameter data from different sampling frequencies and data sources into the same depth grid. A depth grid refers to a continuous depth sequence formed according to a preset depth interval with well depth as a unified reference. Outlier removal includes removing data points corresponding to actual mechanical drilling speed less than or equal to zero, drilling pressure less than or equal to zero, rotation speed less than or equal to zero, torque less than or equal to zero, and abnormal drilling states. Abnormal drilling states include at least one of the following: connecting a single joint, stopping drilling, circulating, tripping in and out of the drill string, reaming, and reverse reaming. Missing value handling includes: for continuous parameters, linear interpolation, nearest-neighbor filling, or sliding window midpoint filling methods are used; For discrete parameters, the corresponding category of adjacent depth points within the same well section or the well section record used by the drill bit is used to fill the data; for data segments with continuous missing lengths exceeding the preset depth length or that cannot be reliably filled, the corresponding sample points are removed.

3. The method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency according to claim 2, characterized in that, In step S2, the formula for calculating mechanical specific energy is as follows: In the formula, For the first Mechanical specific energy of each depth sampling point; For drilling pressure; This represents the rock-breaking area of ​​the drill bit, and ; The diameter of the drill bit; Rotational speed; Torque; The actual mechanical drilling rate used to calculate the mechanical specific energy; Unit conversion factor; The value can be: In the formula, This refers to the actual mechanical drilling speed. The preset minimum mechanical drilling speed; The formula for calculating confining pressure compressive strength is as follows: in, ; In the formula, For the first The confining pressure resistance at each depth sampling point; Uniaxial compressive strength; This refers to the bottom hole pressure. Pore ​​pressure; The effective confining pressure difference at the bottom of the well; This is the confining pressure correction factor, used to characterize the sensitivity of rock compressive strength to increase with increasing effective confining pressure at the bottom of the well; The formula for calculating the energy efficiency coefficient is as follows: In the formula, For the first Energy utilization efficiency coefficient of each depth sampling point.

4. The method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency according to claim 3, characterized in that, Confining pressure correction factor The expression is: In the formula, It represents the internal friction angle of the rock.

5. The method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency according to claim 1, characterized in that, The implementation process of step S3 is as follows: S31. The preprocessed sample data are grouped according to at least one of the following: lithology, stratigraphic position, drill bit type, drill bit size, and well section in which the drill bit is used, to obtain several sample groups g. S32, within each sample group g, sort the actual mechanical drilling rate data in ascending order of value, and take the nth... The quantile of the actual mechanical drilling rate is used as the high quantile threshold of the actual mechanical drilling rate, denoted as . ; S33, within each sample group g, the energy utilization efficiency coefficients are sorted in ascending order of value, and the nth value is selected. The quantile energy efficiency coefficient is used as the quantile threshold of the energy efficiency coefficient, denoted as . ; S34. Samples that simultaneously meet the following conditions will be considered as an efficient sample subset: and .

6. The method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency according to claim 5, characterized in that, In step S31, primary grouping is performed according to lithology or stratigraphic position, followed by secondary grouping according to drill bit type, drill bit size, or well section where the drill bit is used. When the number of samples in the same group is insufficient, adjacent well sections with similar lithology or the same drill bit conditions are merged into the same sample group. In step S32, Take a value of 0.70 to 0.85, in step S33, Take a value of 0.50 to 0.

70.

7. The method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency according to claim 1, characterized in that, In step S4, the reference drilling speed model includes one of the following: random forest regression model, gradient boosting tree model, support vector regression model, neural network model, multiple regression model, and quantile regression model. In step S6, the formula for calculating the drilling speed efficiency coefficient is: In the formula, For the first Drilling efficiency coefficient at each depth sampling point; This refers to the actual mechanical drilling speed; For reference drilling speed.

8. The method for identifying inefficient well sections based on joint constraints of reference drilling rate and energy utilization efficiency according to claim 1, characterized in that, In step S7, a point is identified as inefficient when the following conditions are met: and ; In the formula, The preset drilling speed efficiency threshold; The preset energy utilization efficiency threshold; In step S8, the merging process is as follows: Set the length of the continuous depth window and inefficient point ratio threshold ; When continuous depth window The ratio of the number of inefficient points within the window to the total number of sampling points within the window is greater than or equal to... When this happens, the well section corresponding to the continuous depth window is identified as an inefficient well section; When the interval between adjacent inefficient well sections is less than the preset merging interval Merge adjacent inefficient well sections; The results of inefficient well section identification include at least one of the following: inefficient well section number, starting depth, ending depth, well section length, average actual mechanical drilling rate, average reference drilling rate, average drilling rate efficiency coefficient, average energy utilization efficiency coefficient, percentage of inefficient points, and inefficiency level. Inefficiency levels are categorized into mild inefficiency, moderate inefficiency, and severe inefficiency.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method for identifying inefficient well sections based on the joint constraints of reference drilling rate and energy utilization efficiency as described in any one of claims 1 to 8.

10. A computer-readable storage medium for storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the method for identifying inefficient well sections based on the joint constraints of reference drilling rate and energy utilization efficiency as described in any one of claims 1 to 8.

Images