RU2806206C1 - Horizontal well drilling method - Google Patents

Abstract

FIELD: well drilling.

SUBSTANCE: invention relates to the drilling of a horizontal well and, in particular, to the optimal control of this drilling with the condition of ensuring that the horizontal wellbore is located between its roof and base. The essence of the invention: the method determines the type of rocks in real time while drilling a well at the location of the rock-cutting tool. This is carried out using a set of geological and technical data to set drilling modes - its speed, axial load on the rock-cutting tool, the latter’s torque and revolutions. To do this, in given drilling intervals of the horizontal part of the wellbore, lithological groups of rocks are identified - reference points that characterize the reservoir, non-reservoir and dense layers of the reservoir, influencing the above-mentioned drilling modes. A number of reference wells are selected with bottom hole coordinates close to the current coordinates of the well being drilled and rock groups similar to the rock group of the well being drilled. Additionally, historical data on previously drilled wells in the area of the well being drilled is used, characterizing the mentioned drilling modes. During the drilling process, the rock types of the well being drilled are analysed, compared with the rock types of the reference wells, and the functional dependence of the reference points - a group of rocks - on the drilling modes is established. Based on the lower correlation coefficient of the compared data, the corresponding reference well is selected in real time as a reference well to ensure drilling of a well between the roof and the bottom of the reservoir. At the same time, they prevent the opening of the roof and/or sole. During the drilling process, the zenith angle of the well and the permissible distances to the roof and/or bottom of the reservoir relative to the position of the rock-cutting tool are continuously monitored. If necessary, adjust the well drilling trajectory. Control over the position of the rock-cutting tool is carried out automatically with individualization of the drilling site along the internal and external characteristic boundaries of the reservoir layer and the use of continuous geophysical research, including in front of the bottom of the drilled well.

EFFECT: increasing the efficiency of drilling management by preventing unnecessary operations to adjust drilling and speeding up decision-making on changing the drilling trajectory in real time.

6 cl

Description

The invention relates to the field of drilling a horizontal well and, in particular, to the optimal control of this drilling with the condition of ensuring that the horizontal wellbore is located between the boundaries of the reservoir - its roof and bottom, and in the case of layering of the reservoir itself - in its productive layers or interlayers.

Drilling horizontal wells is one of the pressing problems, the solution of which is associated both with determining the geometric location of wells in geological space and with assessing the properties of reservoir layers and their host rocks.

Methods for controlling the drilling of horizontal wells based on measurements of angles in the Earth's magnetic field or using gyroscopes are widely known. This information is necessary in a complex of geophysical measurements to interpret data from other types of research, for example, electrical logging.

However, such a geometric adjustment of the trajectory based on data from previously drilled wells and the predicted occurrence of layers does not guarantee against errors in entering the wellbore into unproductive strata.

Methods for controlling the drilling of a horizontal well based on lateral (electrical) logging sounding data (BKZ, EKZ) to study the electrical properties around the well are also widely known. However, these methods do not provide sufficient resolution of the geoelectric properties of rocks extending along the horizontal wellbore. In addition, they are characterized by screen effects from compacted layers, which significantly complicates the interpretation of measurement results. There are also technological difficulties in delivering BKZ probe devices, especially non-rigid ones - hose type, to the bottom of horizontal wells.

There is a known method for controlling the drilling of a horizontal well, including geophysical logging during the drilling process and the use of its data to adjust the trajectory of the horizontal well (Betts P. and others, Acquiring and interpreting Logsin Horizontal Wells, Schlumberger, Oilfield Review, vol. 2, No. 3, 1990).

Such logging-while-drilling systems exist abroad (Schlumberger, Anadril, Sperry-Surmnp.). They involve the use of special equipment (Logging While Driling or LWD system).

The known method has disadvantages. Firstly, it involves the use of induction means that do not have geometric and electrodynamic isoparametricity, which does not allow reliably determining the properties of formations. Secondly, the procedure and scheme for using these tools do not provide timely current information about the rocks, which increases the risks of unnecessary - unwanted openings of the boundaries of the reservoir (its roof/or bottom). At the same time, they do not provide the ability to quickly process information about the properties of the rock around the well being drilled and make timely decisions to adjust the drilling of a horizontal well - there are no decision-making algorithms.

The technical result of the invention is to increase the efficiency of controlling the drilling of a horizontal well by preventing unnecessary operations to adjust drilling and speeding up decision-making on changing the drilling trajectory in real time.

The technical result is achieved by the fact that the method of drilling a horizontal well determines the type of rock in real time while drilling the well at the location of the rock-cutting tool. This is carried out using a set of geological and technical data to set drilling modes - its speed, axial load on the rock-cutting tool, the latter’s torque and revolutions. To do this, in given drilling intervals of the horizontal part of the wellbore, lithological groups of rocks are identified - benchmarks that characterize the reservoir, non-reservoir and dense rocks that influence the above-mentioned drilling modes. A number of reference wells are selected with bottomhole coordinates close to the current coordinates of the well being drilled and rock groups similar to the rock group of the well being drilled. Additionally, historical data on previously drilled wells in the area of the well being drilled is used, characterizing the mentioned drilling modes. During the drilling process, the rock types of the well being drilled are analyzed, compared with the rock types of the reference wells, and the functional dependence of the reference points - a group of rocks - on the drilling modes is established. Based on the higher correlation coefficient of the compared data, the corresponding reference well is selected in real time as a reference to ensure drilling of a well between the roof and the bottom of the reservoir formation, preventing the opening of the mentioned roof and/or bottom or dense layers of the reservoir layer. During the drilling process, the zenith angle of the well and the permissible distances to the roof and/or bottom of the reservoir relative to the position of the rock-cutting tool are continuously monitored, and if there is a critical discrepancy, the well drilling trajectory is adjusted. At the same time, control over the position of the rock-cutting tool is carried out automatically with individualization of the drilling site along the internal and external characteristic boundaries of the reservoir layer and the use of continuous geophysical research to assess the rocks, including in front of the bottom of the well being drilled, ensuring the possibility of returning the latter to a given trajectory at the risk of it going beyond the boundaries of the reservoir.

Besides,

• in the process of drilling a well, repeated geophysical studies are carried out with automatic control of the convergence of the main and mentioned repeated studies to assess the quality of these studies and, at the same time, link the measures of the drilling tool to clarify the position of the rock-cutting tool;

• as geophysical research, at least gamma ray logging and/or density logging, and/or electrical logging with electrical resistivity measurements are used;

• approach to the boundaries of the reservoir layer is controlled automatically by analysis of the thickness of the reservoir, data from the reference well and markers of intersections of the internal and external characteristic boundaries of the reservoir layer with the issuance of a command in real time to clarify the position of the rock-cutting tool;

• use a scan of geophysical survey data around the axis of the well being drilled - an image of its wall with their constant analysis using computer vision and individualization of the drilling site in real time to issue a timely command on the position of the rock-cutting tool;

• the position of the rock-cutting tool is further clarified by repeated intersections of characteristic geological boundaries in different directions, which is assessed automatically in accordance with a given algorithm.

In accordance with the invention, the method is carried out as follows. They provide for drilling a production horizontal well. The main problem of drilling such a well is the need for constant monitoring and management of its trajectory in order to comply with the basic requirement - being within the productive formation - a reservoir, which is characterized by a certain type of rock. Therefore, when drilling a well, the type of rock is determined as a marker in real time while drilling at the location of the rock-cutting tool based on a set of geological and technical data. Data on the type of rocks is also necessary to set drilling modes - its speed, axial load on the rock-cutting tool, the torque of the latter and its revolutions. To do this, in given drilling intervals of the horizontal part of the wellbore, lithological groups of rocks are identified - benchmarks that characterize the reservoir, non-reservoir and dense rocks that influence the above-mentioned drilling modes. A number of reference wells are also selected with bottomhole coordinates close to the current coordinates of the well being drilled and rock groups similar to the rock group of the well being drilled. Additionally, historical data on previously drilled wells in the area of the well being drilled is used, characterizing the mentioned drilling modes. The additional operations described above are performed according to the method in relation to optimizing the drilling itself. At the same time, they solve one of the most important tasks of horizontal drilling - ensuring constant control over the trajectory of the well and drilling control in such a way that the well is constantly located between the roof and the bottom of the reservoir, so that all measures for the unwanted opening of the above-mentioned roof and/or bottom or dense layers of the reservoir layer. For this purpose, in accordance with the invention, an automatic analysis of the position of the rock-cutting tool relative to characteristic geological boundaries is also provided. These boundaries are identified according to geophysical studies (logging curves - diagrams).

Sections of logging curves, for example, from two wells are compared and similarities are established based on a combination of the following characteristics:

- thickness of layers (true stratigraphic thickness);

- properties of formations (electrical conductivity, natural radioactivity, etc.);

- sequence of layers.

Based on the appearance of anomalies in the sequence of occurrence of rocks in the section of a well, a conclusion is made about stratigraphic unconformities in the occurrence of rocks or the presence of tectonic disturbances. It is reiterated that all of the above operations are carried out in real time.

Obviously, the degree of individualization of the section section when using such a complex characteristic and such a mode for assessing the situation will be higher than when using its individual components.

When accompanying drilling, when a given well has not yet crossed the benchmarks, an automatic forecast of their position is used under the conditions of a plane-parallel model - with the thickness of the formations being preserved.

The necessary initial data for predicting the bottom of the analyzed interval is the distance from the roof to the bottom in the reference well, as well as the position of the roof in the drilled well.

For these conditions, a software module is developed - an algorithm with the following sequence of actions:

• control of zenith and azimuthal angles (control of deviations from the planned trajectory before opening the first desired reference point);

• checking against benchmarks (activated after meeting the first marker installed in the reference well);

• distribution of convergence weights (exclusion of wells with changed capacities, when the thickness of the analyzed interval changes more than the established value;

• calculation of the distance to the roof (if the forecast for the wells is different, the average value is taken).

Based on the results of testing the algorithm on real wells (15 wells), it was decided to set a small window in the area of the desired markers (about 1 m vertically), because due to the slightest inaccuracy or inconsistency in the selection of reference points, part of the information on the desired configuration of the curve is lost. After establishing the window, the quality of the “autocorrelation” was significantly improved, which increased the accuracy of the forecast.

At the same time, an additional forecast value is provided for the reference well with better convergence. As a result, based on the algorithm, an average forecast is obtained for all reference wells (based on the “majority voting” principle) and for the most suitable reference well separately.

Based on the testing results, it was also revealed that the principle of placing markers is of considerable importance in the accuracy of the forecast based on the algorithm.

It is not worth highlighting implicit reference points (they can be obvious according to the trends of sequence stratigraphy, but at the same time do not have a clear contrast at the boundary of the stratum).

Intervals of less than 10 m and more than 100 m vertically are more difficult to process and do not provide the necessary technical effect.

As a result, for the algorithm to work effectively, it is necessary to place markers in the range from 10 to 100 m. In this case, markers must be installed on obvious benchmarks with a strong contrast in properties, for example, natural radioactivity.

To position the rock-cutting tool relative to the characteristic boundaries of the reservoir layer, a method is also used to identify repeating sections of the mentioned reservoir layer.

When the wellbore re-intersects previously drilled areas, logging curves that repeat in shape are observed. This shape can be stretched or compressed horizontally (depending on the angle of intersection), and can also be mirrored (when intersecting in the opposite direction).

The same section can be crossed in different directions and at different angles. At the same time, real geophysical data, as a rule, develops into very complex configurations that are difficult to identify with the human eye, and sometimes impossible (under conditions of significant compression or extension).

To clarify the position of the rock-cutting tool relative to the characteristic boundaries, an algorithm has also been developed - an algorithm for detecting repeating intervals according to geophysical research data based on time series analysis (dynamic time scale transformation algorithm).

When testing 15 wells, a condition was introduced that it was impossible to cross the same section in the same direction if it had not previously been crossed in the opposite direction. This is only possible if there are faults. At the same time, a verification procedure for the identified area was added, according to which the already identified interval is checked again, taking into account the additional information received. The algorithm was implemented in the form of searching for a given interval forward, searching for a given interval backward, and comparing the two mentioned intervals with each other.

The method is not applicable in wells where drilling is carried out strictly according to bedding, i.e. when the zenith angle is equal to the angle of inclination of the structure. In this case, repeated crossing of the same geological boundaries is excluded and, accordingly, there is no data for analysis.

An additional method to those listed for clarifying the position of a rock-cutting tool is the method of identifying markers of formation intersections (characteristic boundaries) in azimuthal logging using a stream of numerical data.

Azimuthal quadrant data (image) is a set of diagrams of a certain type of logging recorded around the axis of the well along its wall. When unfolding the resulting image along azimuths - the image onto a 2D plane (for clarity with a gradient color fill), you can see a “photograph” of the wellbore wall.

When the wellbore intersects the rock down the bedding with the lower sensor, the interlayer is broken off first, then this interlayer is broken off by the right and left sensors (side sensors), and the last layer is beaten off by the upper sensor. In this way, a descending sinusoidal curve is generated. When crossing the interlayer up the bedding, the order of tapping by the sensors will be reversed. In this case, an ascending sinusoidal curve is plotted.

Thus, it all came down to the task of identifying sinusoidal curves on image data, where the shape of the sinusoidal curve indicates whether a change in the rock was noted from above or below.

From the need for multiple processing of information on the wellbore associated with the image of its wall, it follows that it is advisable to analyze such information using computer vision technologies with the selection of promising directions, including the use of machine learning.

Within the framework of the invention, a technique has been developed that allows identifying sinusoidal curves from image data. The key algorithms on which the mentioned technique is based are the binarization algorithm, which allows you to divide image pixels into “useful” and “background” with minimal intraclass variance, and the Canny filter, which allows you to determine boundaries that are optimal according to the criteria of selection, localization and minimization of several responses of one the edges.

In the problem under consideration, the input data is a numerical array reflecting the development, for example, of density logging around the axis of the wellbore.

Diagrams of sinusoidal curves are isolated in sliding windows 10 and 20 m wide with a step of 5 m, and duplicates are combined.

Using the proposed method, 20 wells from fields in Western and Eastern Siberia were analyzed (the total number of characteristic sinusoidal curves identified automatically exceeded 450). The sample included data obtained from four different devices with penetration depths from 4.572 cm to 8.89 cm. Based on the results of this testing, about 80% of the characteristic curves were identified.

To improve the quality of drilling horizontal wells, the angle of incidence of the structure is assessed in real time directly in the direction of drilling. In this case, do the following:

- estimate the length of the sinusoidal curve;

- determine the angle between the well axis and the selected geological structure;

- determine the angle of the said structure based on the inclination angle of the well; At the same time, the parameters that influence the determination of the angle of occurrence are taken into account

formation according to azimuthal logging:

- wellbore diameter;

- penetration depth of the measuring device

- length of the sinusoidal curve;

- well inclination angle.

The determination of the above parameters is provided within the framework of the developed algorithm using trivial trigonometric expressions.

A specific example of the invention (using the example of well 2518G, pad 20, target formation BS8-1).

A horizontal well is drilled at the depth of the productive formation, where the absolute elevation of the roof is expected at a depth of 2332.5 m, and the bottom - at a depth of 2334.5 m. During the drilling process, geophysical research is carried out in the form of gamma logging in the transport section and gamma + resistivity ( electrical resistivity) + density + neutron + image (azimuthal density scan) logging in the horizontal section. Based on the totality of geological and technical data, namely gamma ray logging in the transport section in real time, drilling determines the type of rocks, namely sandstone-type rocks with interlayers represented by mudstones. The productive formation lies relatively horizontally. Based on the totality of data obtained during the drilling process, the optimal drilling modes for a given type of rock are identified - penetration speed (30-50 m/hour), axial load on the rock-cutting tool (3-5 tons), the rotational torque of this tool is set from the condition of ensuring the required rotation speed columns (for example, 25-40 rpm). To do this, use available experimental data obtained previously under similar conditions from adjacent wells or in similar rocks at similar depths. In development of this, there are (25YUG-pilot, 2557G-pilot, 2518G-pilot, 69R, 2539G-pilot) with face coordinates close to the current coordinates of the well being drilled and rock groups similar to the rock group of the well being drilled. Additionally, historical data on previously drilled wells in the area of the well being drilled is used, characterizing the mentioned drilling modes. This historical data is prepared in advance in the form of a data archive, to which there is constant access during the drilling process and which provides an automatic algorithm for requesting data from the archive, provided for drilling support. Historical data includes: specific field, specific well number, specific pad number to which the well belongs, target formation, well type, well completion type, section in question, event class and type, event MD, loss of effective length (Leff), position events, event marker, marker detection interval, decision making using a geonavigator, obtaining a result and conclusion.

When drilling well 2518G, the following is carried out. Drilling support using the technology under consideration begins from a depth of 2913.57 m. The necessary geophysical research is carried out in the target interval - from a depth of 596.55 m. The planned effective penetration through the reservoir is achieved - 357.93 m (60%). To plant the production casing, only the top mudstones are opened at an zenith angle of no more than 86° without opening the formation roof itself, for which drilling is stopped above the formation roof by 1 m vertically with a permissible deviation from the calculated profile of no more than 1 m vertically.

At the bottom of the actual well (at a depth of 2913 m), the search for benchmarks begins. At the same time, the correlation coefficient is determined for all reference wells. The position of the roof of the target formation is predicted (using the thickness analysis method). Preference is given to the forecast based on the reference well with the highest correlation coefficient with the actual well. In this case, it is well 69Р with a correlation coefficient of 0.95. For benchmarks with absolute elevations of 2130 m, 2160 m, the roof is predicted at an absolute elevation of 2336 m. During further drilling, upon reaching the absolute elevation of 2262 m, the correlation coefficients change and, as a result, receive an algorithm recommendation - “tuning to reference wells 2518PL and 2539g_ps_fact. MD 3278 m with a forecast of the top of the target formation at an absolute elevation of 2327.3 m.”

After passing the absolute level of 2318 m, the best correlation (94%) is still obtained with the reference well 2518PL, MD 3307 m with the forecast of the roof of the target formation at the absolute level of 2327.5 m. Thus, the landing of the production casing shoe is planned strictly 1 meter above the top of the target formation, which is confirmed by further drilling.

Besides:

- based on the results of data binding in well 2518G according to the rewrite module in the MD3160-3200 m interval, a shift of 4.86 meters is determined based on repeated geophysical studies. In accordance with this, the size of the drilling tool is adjusted;

- after tying and correction, continue drilling the horizontal section, during which the signs according to paragraphs are used. 3-4 formulas. At MD 3103 m, according to the readings of gamma ray logging (58API), density logging (2.4 g/cm3), and electrical logging (8 Ohm), the entrance to the target formation (reservoir) is recorded and a command is automatically received to set the zenith angle to 89.5 degrees;

- the dip angle of the formation is automatically determined using the signs according to clause 5 of the formula. At the bottom MD 3080 m, a sinusoidal curve is recorded, the shape of which is characteristic of entering the formation from above based on its amplitude. The dip angle of the formation is determined automatically. It is 0.3 degrees;

- then at the bottom of the well (depth 4285 m) use the characteristics according to clause 6 of the formula. Based on the algorithm, a message is received - “MD 4285 m, the interval 4240-4285 m corresponds to the interval 3996-4050 m (roof, intersection in the same direction).”

Claims (6)

1. A method of drilling a horizontal well within a productive reservoir formation, including determining the type of rock in real time while drilling a well at the location of the rock-cutting tool using a set of geological and technical data to set drilling modes - its speed, axial load on the rock-cutting tool, the torque of the latter and its revolutions, for which, in given intervals of drilling the horizontal part of the wellbore, lithological groups of rocks are identified - benchmarks that characterize the reservoir, non-reservoir and dense layers of the reservoir formation, influencing the above-mentioned drilling modes, select a number of reference wells with bottomhole coordinates close to the current ones coordinates of the drilled well and groups of rocks similar to the group of rocks of the drilled well, additionally use historical data on previously drilled wells in the area of the drilled well, characterizing the mentioned drilling modes; during the drilling process, analyze the rock types of the drilled well, compare them with the rock types of reference wells, establish functional dependence of reference points - a group of rocks on drilling modes; according to the higher correlation coefficient of the compared data, the corresponding reference well is selected in real time as a reference one to ensure drilling of a well between the roof and the bottom of the reservoir formation, preventing the opening of the mentioned roof and/or bottom or dense layers reservoir formation, during the drilling process they continuously monitor the zenith angle of the well and the permissible distances to the roof and/or bottom of the reservoir formation relative to the position of the rock-destroying tool, in the event of a critical discrepancy of which the well drilling trajectory is adjusted, while monitoring the position of the rock-destroying tool is carried out in automatic mode with individualization of the drilling site along the internal and external characteristic boundaries of the reservoir and the use of continuous geophysical research to assess the rocks, including ahead of the bottom of the well being drilled, ensuring the possibility of returning the latter to a given trajectory at the risk of it going beyond the boundaries of the reservoir. 2. The method according to claim 1, characterized by the fact that in the process of drilling a well, repeated geophysical surveys are carried out with automatic control in real time of the convergence of the main and mentioned repeated studies to assess the quality of these studies and, at the same time, link the measures of the drilling tool to clarify the position of the rock-destroying tool. 3. The method according to claim 2, characterized in that at least gamma ray logging and/or density logging and/or electrical logging with electrical resistivity measurements are used as geophysical research. 4. Method according to paragraphs. 1 or 2, characterized by the fact that the approach to the boundaries of the reservoir layer is controlled automatically by analyzing the thickness of the reservoir, data from a reference well and markers of intersections of the internal and external characteristic boundaries of the reservoir layer with issuing a command in real time to clarify the position of the rock-cutting tool. 5. Method according to one of paragraphs. 1-4, characterized by the fact that they use a scan of geophysical survey data around the axis of the borehole of a well being drilled - an image of its wall with their constant analysis using computer vision, individualization of the drilling site in real time to issue a timely command on the position of the rock-cutting tool. 6. Method according to one of paragraphs. 1-5, characterized by the fact that the position of the rock-cutting tool is additionally specified by repeated intersections of characteristic geological boundaries in different directions, which is assessed automatically in accordance with a given algorithm.