RU2723805C1 - Method and computer system for control of drilling of the wells - Google Patents

Abstract

FIELD: soil or rock drilling.SUBSTANCE: invention relates to drilling wells in productive formations. In accordance with disclosed method, initial data characterizing drilling process and containing real-time data, measured during well drilling, are selected, contextual drilling data related to drilling macro-parameters, and geological data characterizing productive formation, in which drilling is performed. Collected data is computer processed and structured data arrays containing complete characteristic of drilling process at each moment of time are formed. Using structured data arrays as input data, actual conditions of drilling are determined in real time and probability of occurrence of complications during drilling is predicted. Using input data determined in real time actual conditions of drilling and probability of occurrence of complications during drilling, recommendations are made on correction of trajectory of well shaft for obtaining maximum productivity of well and / or recommendations on change of drilling modes, composition of drilling mud. Computer system for controlling well drilling in accordance with the present invention comprises a drilling data collection unit, a computer unit for collecting collected data, a real-time analytics unit and a real-time recommendation generation unit.EFFECT: technical result consists in creation of effective and high-speed control system of wells drilling, providing accident-free well cabling under varying conditions and maximum extraction of hydrocarbons.9 cl, 3 dwg

Description

The invention relates to the field of drilling wells in productive formations and, more specifically, to methods and systems for controlling well drilling using machine learning methods and mathematical optimization.

Well drilling is the first stage in the development of oil and gas fields. Drilling directional and horizontal wells provides the process of hydrocarbon production from thin productive layers (intervals) of complex geometric shape. When drilling wells, the management of the trajectory and drilling modes is carried out by specialists who focus on:

- data of geological and technical information (GTI, or telemetry of drilling from the surface, including weight on the hook, load on the bit, rotor speed, pressure in the discharge line, fluid flow rate, etc.);

- inclinometry data containing information about the angle of curvature of the wellbore and its spatial position;

- logging data during drilling, including gamma ray logging, neutron, density logging, imager, etc .;

- geological and technical order for well construction;

- data on the proposed section, based on the results of seismic studies and the results of drilling of reference wells.

This approach to drilling management has the following disadvantages:

- non-measuring zone - the distance between the bit and the logging sensors. This zone causes delay in information about the drilling conditions, which leads to long-term adjustments or excesses of part of the trunk;

- the impossibility of visual monitoring of a large number of telemetry curves at the same time, which is the cause of accidents;

- inconsistency of the project for the construction of the well with the actual drilling conditions, which leads to overestimated terms and cost of well construction, and not to achieve the project flow rate.

There is a known method of controlling well drilling described in RF patent 2542026 and comprising the following steps: preparation of drilling equipment having a bottom assembly of a drill string, which includes a controlled subsystem of directional drilling and a directional measuring tool for logging while drilling with the possibility of circular viewing and proactive viewing; determination of the presence of a given type of formation features in the target formation; and navigating the drilling path in the target formation with said drilling equipment, including receiving measurement signals from the specified directional measuring device, obtaining, based on the received measurement signals, parameters of the formation parameters relative to the specified feature of the formation in the target formation and controlling the specified directional drilling subsystem for drilling in the direction determined depending on the specified obtained parameter of the formation parameters.

The disadvantage of this method is that the properties of the formation are determined by the results of the interpretation by the specialist-geologist of information from the logging tool, located significantly above the bit. Thus, information on the strength properties of the drilled rock is obtained with a delay associated with:

- remoteness of the device;

- The time of expert interpretation.

This leads to exits beyond the target reservoir and long-term trajectory adjustments.

In the patent of the Russian Federation 2588526 a control method for drilling wells is described, which includes receiving a continuous stream of real-time data associated with the current downhole operation in a data warehouse. At the same time, the user selects the bottomhole parameter. Then, using a computing system, the selected downhole parameter is optimized based on a portion of the received data stream to achieve the target value of the selected downhole parameter. Then, the optimized downhole parameter can be used in the current operation.

The disadvantages of this method are:

The presence of calculation parameters based on formulas leads to lengthy calculations, possible errors and the impracticability of calculating a large number of variations in the development of events;

Only one downhole parameter is optimized at a time, but not their combination;

It works on the principle of comparing current and design data and allows you to only set indicators to deviate from the plan;

Manual input of a portion of the drilling macro parameters is required.

There is also known a drilling control method described in RF patent 2600497, in accordance with which sensors are collected from neighboring wells and contextual data from neighboring wells, placed in storage, and models for predicting the drilling process are created.

The disadvantages of this method are:

The method for smoothing the initial data described in the invention leads to the removal of part of the values corresponding to the real drilling anomaly (complications, harbingers of accidents);

The removal of incorrect values occurs with the formation of "voids" in their place;

Neighboring well data required for forecasting;

The presence of design parameters based on formulas leads to lengthy calculations and possible errors;

Manual input of part of the drilling macro parameters is required;

It contains multiple models for one forecast unit, which can lead to errors within each model, as well as in the process of selecting the main model at the moment;

Carries out the prediction of the downhole operation without recommending parameters to achieve the best result.

The technical result achieved during the implementation of the invention is to create an effective and fast-acting well drilling control system that ensures trouble-free well drilling under changing conditions and maximum hydrocarbon production during subsequent operation with minimal investment in well construction.

The specified technical result is achieved by the fact that, in accordance with the proposed method of controlling the drilling of wells, initial data are collected that characterize the drilling process and contain real-time data measured during the drilling process, contextual drilling data related to macro drilling parameters and geological data characterizing the reservoir in which carry out drilling. Computer processing of the collected data is carried out and structured data arrays are formed containing a complete description of the drilling process at any given time. Using the generated structured data arrays containing the complete characteristics of the drilling process at each moment of time as input, the actual drilling conditions are determined in real time, including the properties of the rock currently being drilled, potential formation productivity, rheological characteristics of the drilling fluid and predict the likelihood of complications during drilling. Using the actual real-time drilling conditions and the likelihood of complications during drilling as input, they develop recommendations for adjusting the wellbore trajectory to obtain maximum well productivity and / or recommendations for changing drilling modes and drilling fluid composition.

Real-time data measured during the drilling of a well contain at least one type of data selected from the group consisting of the results of measurements from sensors located on the surface, the results of inclinometry, logging data during drilling, and measurements of downhole sensors.

The drilling contextual data related to the drilling macro parameters contains at least one data type selected from the group containing the well construction project, the position of adjacent well bores, drilling fluid parameters, and drilling fluid components available.

The contextual data characterizing the reservoir contain at least one data type selected from the group consisting of the initial geological model, geomechanical model, forecast section, initial production rate, production levels.

Computer processing of data for the formation of structured data arrays containing a complete description of the drilling process at any given time contains at least one of the following actions: bringing the collected data to a single format, eliminating the facts of sensor calibration during drilling, aligning dimensions and units of measurement, identifying gaps in the sensor readings using machine learning methods and restoration of missing values, error estimation when restoring missing values in sensor readings using machine learning methods.

In accordance with the proposed method in real time, at least one of the actual drilling conditions is selected, selected from the group: properties of the drillable rock according to surface sensors, lithotype according to the results of logging interpretation during drilling, productivity of drillable formations during drilling, rheological characteristics of the drilling real-time solution, the likelihood of complications.

The computer-controlled well drilling control system in accordance with the invention comprises a drilling data collection unit for collecting real-time data measured during a well drilling, contextual drilling data related to macro parameters of drilling and geological data characterizing a producing formation in which drilling, a computer-aided processing unit for collected data intended for generating structured data arrays containing a complete description of a drilling process at any given time, a real-time analytics block using generated structured data arrays containing a complete description of a drilling process at any time as input , and designed to determine the actual drilling conditions and predict the likelihood of complications during drilling and the block recommendations in real time, using as input the results of work You are a real-time analytics unit and designed to make recommendations for adjusting the trajectory of the wellbore during drilling to obtain maximum productivity, recommendations for changing drilling modes, recommendations for changing the composition of the drilling fluid.

The computer processing unit for the collected data, designed to generate structured data arrays containing a complete description of the drilling process at any given time, uses artificial intelligence methods to perform at least one action selected from the group: bring the collected data to a single format, identify anomalies in the data - emissions, incorrect values, facts of calibration of sensors during drilling, alignment of dimensions and units of measurement, elimination of facts of calibration of sensors during drilling, restoration of missing values in sensor readings using machine learning methods, error estimation when restoring missing values in sensor readings using machine methods learning.

A real-time analytics unit that uses generated structured data arrays containing the complete characteristics of the drilling process at any given time as input uses artificial intelligence methods to determine at least one of the actual drilling conditions: determine the properties of the drillable rock using surface sensors , determining the lithotype based on the results of logging interpretation during drilling, determining the productivity of the drilled formations during drilling, determining the rheological characteristics of the drilling fluid in real time, and also to predict the likelihood of complications during drilling.

The invention is illustrated by drawings, in which in FIG. 1. presents a general diagram of a computerized control system for drilling wells; in FIG. 2 shows a block diagram of data collection and the formation of structured data arrays; in FIG. Figure 3 presents a block diagram of the work of the analytics block and the recommendation generation block in real time.

The proposed method and control system for drilling wells is a combination of blocks of software and operations that ensure trouble-free construction of a well in a changing or different design environment.

In FIG. 1 is a general diagram of a well drilling control system, comprising a block 1 for collecting initial data, a preprocessor 2 — a block for computer processing the collected data, a block 3 for real-time analytics, a block 4 for generating recommendations in real time. The collected source data is pre-processed using preprocessor 2, after which it can be used by real-time analytics block 3 and real-time recommendations block to issue recommendations.

As shown in FIG. 1 and in FIG. 2, by means of a source data collection unit 1, source drilling data 1 is collected. As the initial data 1, drilling support data are used, which can be divided into real-time measured (hereinafter real-time data 5 in Fig. 2) and data not measured during drilling and related to the entire current well or previously drilled wells (hereinafter contextual or non-real-time data 6 in Fig. 2). At least one data type selected from the group consisting of the results of measurements from sensors located on the surface (sensors of the drilling station or the station of geological and technical research - GTI), the results of inclinometry (measurement of the position of the wellbore) is used as real-time drilling data 5 in space), logging data during drilling, measurement results of downhole sensors.

Measurement results from sensors located on the surface include the weight of the drill string or tool, the bottom depth, inlet pressure, load on the bit, pump stroke rate, tool speed, temperature, and hydrocarbon component concentrations (methane, ethane, propane and etc.), and others.

At least one type of data selected from the group containing the well construction project, the results of measurements of flushing fluids, the position of the boreholes of neighboring wells, and data characterizing productive data are used as contextual or not real-time data 6 of drilling not measured during the drilling process. formation, which contain at least one type of data selected from the group consisting of the initial geological model, geomechanical model, starting production rate of wells, well production levels, etc.

The initial data 1 enters the preprocessor 2 (see Fig. 1) - the computer processing unit of the collected data, which performs computer processing of the collected data. Preprocessor 2 aggregates data from measurements of various formats into a single database using intelligent algorithms for approximate matching and error correction. Preprocessor 2 uses artificial intelligence methods to perform one or more actions selected from the group:

- bringing the collected data to a single format, including alignment of dimensions and units of measurement (7 in Fig. 2);

- identification of data anomalies: gaps, outliers, incorrect values, facts of sensor calibration during drilling (8 in Fig. 2) using machine learning methods;

- intelligent processing: filling in the gaps in the sensor readings, eliminating some of the data containing sensor calibrations and restoring the missing values using machine learning methods (9 in Fig. 2);

- estimation of the error in the restoration of missing values in the sensor readings by machine learning methods (10 in Fig. 2).

Function 7 of preprocessor 2 — converting the collected data to a single format — performs operations such as aggregation of time series for separating multilateral wells, converting las files to .csv format, determining units of measurement (based on statistical tests of matching the distribution of values with reference samples with known units), alignment of dimensions and units of measurement (using translation tables of known dimensions and rules based on trained decision trees in the general case), division of time series into sections (direction, conductor, production casing, etc.), division of operations into phases (drilling , descent, ascent, development, reverse engineering, flushing, simple), aggregation of time series in depth, separation from the design data of information on the BHA composition, drill string, properties of flushing fluids, etc.

Function 8 of preprocessor 2 “detection of data anomalies” uses machine learning algorithms to determine the presence in the sensor readings of anomalies corresponding to gaps, outliers, incorrect values (for example, values many tens of times higher than physically justified values), instrument calibrations and marks them in accordance with anomaly detected.

Function 9 of preprocessor 2 “intelligent processing” uses machine learning algorithms to restore missing values, learning at intervals with a full range of measurements, corrects outliers, eliminates some of the data containing spurious information from the sensor calibration process. As such algorithms, recurrent neural networks for multidimensional time series are used (see, for example, Graves, Alex. "Generating sequences with recurrent neural networks." ArXiv preprint arXiv: 1308.0850 (2013).) As well as generative autoencoders (see, for example, Makhzani, A., Shlens, J., Jaitly, N., Goodfellow, I., & Frey, B. (2015). Adversarial autoencoders. ArXiv preprint arXiv: 1511.05644). Thus, the preprocessor 2 restores the missing values, does not delete, and corrects the anomalous values.

Function 10 of preprocessor 2 “error estimation when recovering missing values in sensor readings” uses information about the proportion of missing values and the size of the training sample to calculate the probability of correct recovery of missing values. The model for assessing the probability of correct recovery is trained on the basis of artificially created omissions of various nature on complete data.

As shown in FIG. 2, the result of the work of preprocessor 2 is structured data arrays containing a complete description of the drilling process at each moment of time - data frames 11, which are tables in the .csv format, containing a structured complete description of the drilling process, at each moment of time throughout the construction cycle wells.

Thus, preprocessor 2, a computer processing unit for the collected data, provides the formation of structured data arrays containing a complete description of the drilling process at any given time — data frames. As input data, drilling support data of various contractors and obtained from various sensors can be used. It does not require manual input of additional parameters. On data frames, training of predictive and optimization algorithms based on machine learning methods is conducted, and trained models are used to work with new input data obtained during real-time drilling and to generate forecasts.

In FIG. 3 shows the principle of operation shown in FIG. 1 real-time analytics block 3 and real-time recommendation block 4.

Real-time analytics and recommendations blocks are tools for diagnosing and managing unforeseen drilling situations. The system determines the actual drilling conditions, analyzes the degree of their deviation from the design values, determines the presence of critical anomalies and recommends further actions based on the conditions for obtaining maximum production and safe work.

Block 3 of real-time analytics as input uses data frames 11 and uses machine learning algorithms to determine at least one of the actual drilling conditions, selected from the group, such as the properties of the drillable rock (module 12 in Fig. 3), lithotype according to the results interpretation of logging during drilling (module 13 in Fig. 3), productivity of the drilled formations (module 14 in Fig. 3), rheological characteristics of the drilling fluid (module 15 in Fig. 3, the likelihood of complications (module 16 in Fig. 3).

The principle of determining the properties of the drilled rock (module 12 in Fig. 3) is based on the fact that when the strength properties of the rock change, the drilling regimes recorded by the drilling sensors or GTI stations on the surface also change. To solve the problem of determining the drilled lithotype, a machine learning algorithm is used (see, for example, Klyuchnikov, N., Zaytsev, A., Gruzdev, A., Ovchinnikov, G., Antipova, K., Ismailova, L., ... & Koryabkin, V. (2019). Data-driven model for the identification of the rock type at a drilling bit. Journal of Petroleum science and Engineering, 178, 506-516), which uses features such as penetration speed, bit load, revolutions rotor, torque and other recorded parameters. XGBoost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, pages 785- 794. ACM, 2016) analyzes the patterns in parameter changes and determines the type and properties of the rock being drilled. After receiving data from logging tools, they are automatically interpreted by module 13 in FIG. 3 followed by the refinement of forecast models (see, for example, Romanenkova, E., Zaytsev, A., Klyuchnikov, N., Gruzdev, A., Antipova, K., Ismailova, L., ... & Koroteev, D. (2019 ). Real-time data-driven detection of the rock type alteration during a directional drilling. ArXiv preprint arXiv: 1903.11436.

Auto-interpretation of logging data during drilling (module 13 in Fig. 3) is performed using machine learning algorithms based on decision trees and can reduce the influence of the human factor on interpretation. Similar to module 12, the algorithm is trained on a full range of historical logging data and interpretation results of geophysical well surveys (RIGIS) with the subsequent issuance of a forecast based on the measurement results.

The productivity of the drilled formations in real time (module 14 in Fig. 3) is determined based on the analysis of measurement data of operating parameters of drilling, information on the drilled lithotype from the prediction algorithm for the type of rock at the bottom, and also interpretation data from logging tools (module 13) coming from the delay associated with the remoteness of the instruments from the bit. Similarly to module 12, an algorithm based on gradient boosting and neural networks analyzes the information received and provides the productivity value of the drilled interval.

The rheological characteristics of the drilling fluid (module 15 in Fig. 3) are restored in the form of time series using machine learning algorithms based on the data frames 11 of its component composition, the properties of the drilled rock from modules 12 and 13, the productivity of the formations 14 and the operating time of the drilling fluid in the well. The algorithm is trained on historical data from previously drilled wells, the result of its operation is calibrated in real time upon receipt of the drilling fluid parameters measurement data (part of the contextual data of drilling 6 in Fig. 2, and real-time data 5 in Fig. 2 from downhole sensors and sensors on the surface).

Block 3 analytics in real time also predicts the likelihood of complications during drilling (module 16 in Fig. 3).

Determining the likelihood of complications during drilling is based on determining the presence of critical deviations from normal operations and assessing the proximity of the current drilling situation to one of the groups of possible complications (see, for example, Gurina, E., Klyuchnikov, N., Zaytsev, A., Romanenkova , E., Antipova, K., Simon, I., ... & Koroteev, D. (2019). Failures detection at directional drilling using real-time analog search. ArXiv preprint arXiv: 1906.02667).

Module 16 works only on the basis of analysis of surface measurement data, without reference to neighboring wells, and a machine learning algorithm based on ensembles of decision trees is used. The algorithm analyzes real-time parameters of drilling regimes by a sliding window. For each measurement, the probability of complications in the well is determined. Along with this, a comparison is made with areas from different categories of complications loaded into the training set and the level of similarity of the current situation with various types of anomalies is determined. If the threshold value is exceeded for both comparisons, a signaling of a possible anomaly occurs with an indication of its type. Such an algorithm is capable of detecting complications that have already begun, as well as their precursors, as well as finding matches with any other categories of cases loaded into the training set. An anomaly signaling can be configured with different sensitivity and signal signs of complication at different risk levels. Having trained on a sufficient data set, it can identify signs of complications without further training and calibrations.

Shown in FIG. 1 block of 4 recommendations in real time as input uses the results of the work of block 3 analytics in real time.

The work of module 17 (Fig. 3) to correct the trajectory is characterized by the fact that it analyzes information about the potential productivity of the drilled formations (module 14 in Fig. 3) and compares them with the data frames of the initial geological model. The initial geological model contains information about the geological structure of the reservoirs and their characteristics (porosity, permeability, fluid saturation). The system calculates the predicted starting flow rate for various options for the position of the wellbore in the reservoir, analyzes the risks of intersection with other wells, calculates the possibility of lowering completion arrangements and makes recommendations for adjusting the well with the condition of obtaining the maximum flow rate with minimal risks. In this case, the machine learning algorithm is used.

The module of recommendations for changing drilling modes (module 18 in Fig. 3) is based on the intellectual analysis of operating parameters of drilling to optimize the penetration rate. For this, a surrogate model is used (see, for example, Forrester, A., Sobester, A., & Keape, A. (2008). Engineering design via surrogate modeling: a practical guide. John Wiley & Sons) using one of the methods among Gaussian processes, nuclear regression, multidimensional spline, gradient boosting on decision trees. Restrictions are imposed on the constructed model based on the safe conduct of work and the trouble-free operation of the equipment. At the output, the optimizer gives the recommended values of the load on the bit, speed, pump flow to achieve the best performance.

The module of recommendations for changing the composition of the drilling fluid (module 19 in Fig. 3) is based on the intellectual analysis of the rheological characteristics 15, the date of the frames 11 of the mechanical drilling speed, the properties of the drilled rock 12, 13, the likelihood of complications 16. At the output, the module gives recommendations for changing the component the composition of the drilling fluid from the conditions for achieving the maximum drilling speed while ensuring safe operations and maintaining the properties of the reservoir.

All algorithms used in the drilling control system and method are trained on a set of historical data from previously drilled wells. At the same time, the work of tuned algorithms can be carried out even when an incomplete complex of data arrives, and the output contains information about the accuracy of the prediction or recommendation, based on the available data set.

Thus, the proposed method and control system for drilling wells allow:

1. determine the type and properties of the rock being drilled at the time of observation;

2. recommend a drilling trajectory subject to maximum production;

3. recommend optimal drilling modes to achieve drilling goals with a minimum total cost of owning a well while maintaining or increasing production;

4. recommend a drilling fluid that provides the maximum penetration rate with minimal risk of complications and minimal labor costs for changing its properties;

5. determine the likelihood of an accident or complication, determine the type of a possible accident or complication in order to take preventive measures to prevent it or minimize the consequences.

The following is an example of the operation of the method and computer system for managing well drilling.

During the construction of the well, raw real-time drilling data 5, which is a multidimensional time series, is measured using drilling sensors and stored in the user's local storage in las or csv format.

Context data 6 is loaded into the local storage in the form of tables or sets of structured text files (for example, csv, json formats) containing design characteristics, measurement results and geological data. Context data 6 may be updated by the user in the event of a change in drilling conditions or design decisions.

Preprocessor 2, according to a specific schedule specified by the user, downloads a new pack of initial real-time drilling data 5 and context data 6. Upon receipt of each new piece of data, preprocessor 2 brings them to a single format - performs operations such as aggregation of time series, translation to .csv format, definition and alignment of units, separation of time series by sections, separation of operations into phases, aggregation of time series by depth, extraction of information about the BHA composition, drill string, properties of flushing fluids from context data, etc. As a result of this of function (7), we obtain standardized datasets for well analysis suitable for further analysis that contain all the necessary information from the source data.

Then data anomalies are revealed (8), corresponding to gaps, outliers, incorrect values, instrument calibrations. Using machine learning algorithms, the expected values of the time series are predicted, then they are compared with the actual ones, and, in the presence of a strong discrepancy, are marked as abnormal. As a result of performing this function (8), an additional service file appears in the data arrays indicating the presence of anomalies and their types for each trunk.

After that, intellectual processing of anomalies is carried out (9). Preprocessor algorithms recover missing values, correct emissions, eliminate some of the data containing sensor calibrations. As a result, we get data arrays similar to the original ones, but with corrected values.

In this case, the error is estimated when restoring the missing values in the sensor readings. Algorithms (10) use information about the fraction of missing values and the size of the training sample to calculate the probability of correct recovery of missing values. As a result of the execution of this function (10), an additional column appears in the data arrays for each trunk indicating confidence in the restoration of missing values.

The result of the preprocessor is recorded in the terminal table in the form of data frames 11, which are csv files.

The processed data frames 11 receive a block of analytics in real time, where tasks 12-16 in FIG. 3.

Receiving a piece of data at the input, the trained algorithms of the block 12 for determining the properties of the drillable rock during drilling select a data interval from the data frame corresponding to one drilled meter, and statistics such as average, variance, angle of inclination and the difference between the values on the time series are calculated on it boundaries. The feature vector from these statistics is then input to a trained classifier of the properties of the drilled rock. At the exit, there is a rock class number and, accordingly, their properties. The result of the operation of the algorithms is recorded in the terminal table of the database and displayed to users using a graphical interface.

Receiving a piece of data as input, the trained algorithms of the auto-interpretation logging unit 13 calculate the statistics of time series as in block 13, as well as such statistics of the logging data as moving averages with different window sizes, their differences, etc. All of the statistics listed are then transmitted to the input by a trained rock classifier. At the output - the name of the rock according to the interpretation of the logging. The result of the operation of the algorithms is recorded in the terminal table of the database and displayed to users using a graphical interface.

Receiving a portion of data at the input, the trained algorithms of the unit 14 for determining the productivity of the drilled formations analyze the data frames of operating parameters of drilling, data from logging tools, geological data and predict productivity parameters by regression. At the exit - the expected starting flow rate at a given length of the trunk and its position in the reservoir. The result of the operation of the algorithms is recorded in the terminal table of the database and displayed to users using a graphical interface.

Receiving a portion of data at the input, the trained algorithms of the block 15 for determining the rheological characteristics of the drilling fluid analyze the data frames of the drilling operating parameters and predict the values of the drilling fluid parameters. Predicted parameter values are calibrated based on actual measurements. At the output - the value of the current parameters of the drilling fluid at each moment in time. The result of the operation of the algorithms is recorded in the terminal table of the database and displayed to users using a graphical interface.

Receiving a piece of data at the input, the trained algorithms of the complication probability determination unit 16 select a data interval corresponding to two hours from the data frame, and statistics such as mean, variance, tilt angle, and difference between the values at the boundaries are calculated from the time frame on it. The feature vector from these statistics is compared with cases from the complication database to assess the likelihood of a situation as emergency. The result of the operation of the algorithms is recorded in the terminal table of the database and displayed to users using a graphical interface.

The result of the work of blocks 12-16 is transmitted to the block of recommendations in real time 4.

The block of recommendations in real time 4 selects the best combination of drilling, mud and trajectory parameters by modeling the total productivity of the trunk and predicting the likelihood of complications with various configurations of combinations.

Block 4 gives recommendations for adjusting the trajectory of the wellbore to achieve maximum productivity 17; recommendations for changing at least one of the parameters: pump supply, bit load, rotational speed to ensure maximum drilling speed in the drilled lithium type under the condition of safe work (18); recommends changes in the component composition of the drilling fluid to achieve maximum drilling efficiency, high-quality cleaning of the wellbore and the condition for safe work (19).

Claims (37)

1. The method of controlling drilling, in accordance with which - collect initial data characterizing the drilling process and containing real-time data measured during the drilling of the well, contextual drilling data related to macro parameters of drilling, and geological data characterizing the reservoir in which the drilling is carried out; - carry out computer processing of the collected data and form structured data arrays containing a complete description of the drilling process at any time; - using the generated structured data arrays containing the complete characteristics of the drilling process at each moment of time as input, they determine in real time the actual drilling conditions, including the properties of the rock currently being drilled, the potential productivity of the formations, the rheological characteristics of the drilling fluid, and predict the likelihood of complications when drilling; - using real-time actual drilling conditions and the likelihood of complications during drilling as input, they develop recommendations for adjusting the wellbore trajectory to obtain maximum well productivity and / or recommendations for changing drilling modes and drilling fluid composition. 2. A method for managing well drilling according to claim 1, wherein the real-time data measured during the well drilling process comprise at least one type of data selected from the group consisting of measurements from sensors located on the surface, results inclinometry, logging data while drilling, measurement results of downhole sensors. 3. The method of managing well drilling according to claim 1, wherein the contextual drilling data related to the macroparameters of drilling contain at least one type of data selected from the group containing the project for well construction, the position of the boreholes of neighboring wells, and drilling fluid parameters mud components available. 4. A method for managing well drilling according to claim 1, wherein the contextual data characterizing the reservoir contain at least one data type selected from the group consisting of the initial geological model, geomechanical model, forecast section, start flow rate, production levels . 5. The method of controlling well drilling according to claim 1, wherein the computer processing of the data to form structured data arrays containing a complete description of the drilling process at any given time contains at least one of the following: - Bringing the collected data to a single format; - the exclusion of the facts of calibration of sensors during drilling; - alignment of dimensions and units; - identification of gaps in the sensor readings using machine learning methods and restoration of missing values; - error estimation when restoring missing values in sensor readings using machine learning methods. 6. The method of controlling well drilling according to claim 1, according to which at least one of the actual drilling conditions is selected in real time, selected from the group: - properties of the drilled rock according to surface sensors; - lithotype based on the results of interpretation of logging while drilling; - productivity of the drilled formations during drilling; - Rheological characteristics of the drilling fluid in real time; - the likelihood of complications. 7. A computer control system for drilling wells, comprising: - a drilling data collection unit for collecting real-time data measured during a well drilling process, contextual drilling data related to macro drilling parameters, and geological data characterizing a producing formation in which drilling is performed; - a computer processing unit for the collected data, intended for the formation of structured data arrays containing a complete description of the drilling process at any time; - a real-time analytics unit that uses generated structured data arrays as input that contain a complete description of the drilling process at each moment in time and is designed to determine actual drilling conditions and predict the likelihood of drilling complications; - a real-time recommendation development unit that uses real-time analytic unit operation results as input and is designed to develop recommendations for adjusting the trajectory of the wellbore during drilling to obtain maximum productivity, recommendations for changing drilling modes, recommendations for changing the composition of the drilling fluid. 8. The computerized control system for drilling wells according to claim 7, characterized in that the computer-aided processing unit for the collected data, designed to generate structured data arrays containing a complete description of the drilling process at any given time, uses artificial intelligence methods to perform at least one action selected from the group: - Bringing the collected data to a single format; - identification of anomalies in the data - outliers, incorrect values, facts of sensor calibration during drilling; - alignment of dimensions and units; - the exclusion of the facts of calibration of sensors during drilling; - restoration of missing values in the sensor readings using machine learning methods; - error estimation when restoring missing values in sensor readings using machine learning methods. 9. The computerized well drilling control system according to claim 7, characterized in that the real-time analytics unit, using the generated structured data arrays containing the complete characteristics of the drilling process at each time point as input, uses artificial intelligence methods to determine at least least one of the actual drilling conditions: - determination of the properties of the drilled rock according to surface sensors; - determination of lithotype based on the results of logging interpretation during drilling; - determination of the productivity of the drilled formations during drilling; - determination of the rheological characteristics of the drilling fluid in real time, as well as to predict the likelihood of complications during drilling.

Images