SYSTEMS AND METHODS FOR GENERATING SYNTHETIC FRACTURE IMAGES OF SUBSURFACE LAYERS FOR DRILLING WELLS

Abstract

A method includes drilling, by a drilling system, a first wellbore in a subterranean region of interest according to a planned wellbore path, and receiving at least one property for a first section of the first wellbore. The method further includes generating, by a synthetic fracture image generator processing the at least one property during drilling of the first wellbore, a synthetic wellbore fracture image for the first wellbore. The synthetic fracture image generator includes a machine-learned network that has been trained using a training set of wellbore images from a second wellbore in the subterranean region of interest. The synthetic wellbore fracture image predicts a fracture in geological layers along the planned wellbore path. The method further includes displaying, while drilling the first wellbore, the synthetic wellbore fracture image, and performing a drilling operation with respect to the first wellbore based on the synthetic wellbore fracture image.

Description

BACKGROUND

In the petroleum industry, information regarding rock fractures is important for reservoir management. This is because fracture networks form passages for flowing oil, gas, and water. Further, fracture networks may cause wellbore failures. Transmission of real-time data from downhole logging tools in wellbores is restricted by a certain bandwidth, which reduces the size and resolution of the image data being acquired while drilling wellbores. Rock features such as fractures are only observed when high-resolution images are retrieved and processed from downhole tools after drilling completion. Accordingly, there exists a need to predict rock fractures and display them in synthetic images to inform drilling monitoring centers in real-time while drilling a wellbore.

SUMMARY

This summary is provided to introduce a selection of concepts that are further described below in the detailed description. This summary is not intended to identify key or essential features of the claimed subject matter, nor is it intended to be used as an aid in limiting the scope of the claimed subject matter.

In one aspect, embodiments disclosed herein generally relate to a method. The method includes drilling, by a drilling system, a first wellbore in a subterranean region of interest according to a planned wellbore path and receiving, by the drilling system, at least one property for a first section of the first wellbore. The method further includes generating, by a synthetic fracture image generator processing the at least one property during drilling of the first wellbore, a synthetic wellbore fracture image for the first wellbore. The synthetic fracture image generator includes a machine-learned network that has been trained using a training set of wellbore images from a second wellbore in the subterranean region of interest. The synthetic wellbore fracture image predicts a fracture in geological layers along the planned wellbore path. The method further includes displaying, by a display system, while drilling the first wellbore, the synthetic wellbore fracture image and performing, by the drilling system, a drilling operation with respect to the first wellbore based on the synthetic wellbore fracture image.

In one aspect, embodiments disclosed herein generally relate to a system including a drilling system, a synthetic fracture image generator and a display system. The drilling system is configured to drill a first wellbore in a subterranean region of interest according to a planned wellbore path, and receive at least one property for a first section of the first wellbore. The synthetic fracture image generator is configured to generate on processing the at least one property, during drilling of the first wellbore, a synthetic wellbore fracture image for the first wellbore. The synthetic fracture image generator has been generated by training a machine-learned network using a training set of wellbore images from a second wellbore in the subterranean region of interest. The synthetic wellbore fracture image predicts a fracture in geological layers along the planned wellbore path. The display system is configured to display, during drilling of the first wellbore, the synthetic wellbore fracture image. The drilling system is further configured to perform a drilling operation with respect to the first wellbore based on the synthetic wellbore fracture image.

In one aspect, embodiments disclosed herein generally relate to a method. The method includes obtaining a plurality of wellbore images from a first wellbore in a subterranean region of interest, and determining, using a trained first machine-learned network, a plurality of detected fracture traces using the plurality of wellbore images. The method further includes determining, for each of the plurality of detected fracture traces, a set of fracture geostatistical properties, and creating a database of fracture properties including the plurality of wellbore images, the plurality of detected fracture traces and the plurality of sets of geostatistical properties. The method further includes generating a synthetic fracture image generator by training a second machine-learned network, using the database of fracture properties, to produce a synthetic wellbore fracture image for a second wellbore in the subterranean region of interest during drilling of the second wellbore.

Other aspects and advantages of the claimed subject matter will be apparent from the following description and the appended claims.

BRIEF DESCRIPTION OF DRAWINGS

Specific embodiments of the disclosed technology will now be described in detail with reference to the accompanying figures. Like elements in the various figures are denoted by like reference numerals for consistency.

FIG. 1 depicts a fractured region of rock in accordance with one or more embodiments.

FIG. 2 depicts a system in accordance with one or more embodiments.

FIG. 3 depicts a system in accordance with one or more embodiments.

FIG. 4 depicts a neural network in accordance with one or more embodiments.

FIG. 5 depicts a generative adversarial network in accordance with one or more embodiments.

FIG. 6 depicts a generative adversarial network in accordance with one or more embodiments.

FIG. 7 depicts a flowchart in accordance with one or more embodiments.

FIG. 8 depicts a flowchart in accordance with one or more embodiments.

FIG. 9 depicts a system in accordance with one or more embodiments.

DETAILED DESCRIPTION

In the following detailed description of embodiments of the disclosure, numerous specific details are set forth in order to provide a more thorough understanding of the disclosure. However, it will be apparent to one of ordinary skill in the art that the disclosure may be practiced without these specific details. In other instances, well-known features have not been described in detail to avoid unnecessarily complicating the description.

Throughout the application, ordinal numbers (e.g., first, second, third, etc.) may be used as an adjective for an element (i.e., any noun in the application). The use of ordinal numbers is not to imply or create any particular ordering of the elements nor to limit any element to being only a single element unless expressly disclosed, such as using the terms “before,” “after,” “single,” and other such terminology. Rather, the use of ordinal numbers is to distinguish between the elements. By way of an example, a first element is distinct from a second element, and the first element may encompass more than one element and succeed (or precede) the second element in an ordering of elements.

It is to be understood that the singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise. For example, an “a synthetic wellbore fracture image,” may include any number of “synthetic wellbore fracture images” without limitation.

Terms such as “approximately,” “substantially,” etc., mean that the recited characteristic, parameter, or value need not be achieved exactly, but that deviations or variations, including for example, tolerances, measurement error, measurement accuracy limitations and other factors known to those of skill in the art, may occur in amounts that do not preclude the effect the characteristic was intended to provide.

It is to be understood that one or more of the steps shown in the flowcharts may be omitted, repeated, and/or performed in a different order than the order shown. Accordingly, the scope disclosed herein should not be considered limited to the specific arrangement of steps shown in the flowcharts.

Although multiple dependent claims are not introduced, it would be apparent to one of ordinary skill that the subject matter of the dependent claims of one or more embodiments may be combined with other dependent claims.

In the following description of FIGS. 1-9, any component described with regard to a figure, in various embodiments disclosed herein, may be equivalent to one or more like-named components described with regard to any other figure. For brevity, descriptions of these components will not be repeated with regard to each figure. Thus, each and every embodiment of the components of each figure is incorporated by reference and assumed to be optionally present within every other figure having one or more like-named components. Additionally, in accordance with various embodiments disclosed herein, any description of the components of a figure is to be interpreted as an optional embodiment which may be implemented in addition to, in conjunction with, or in place of the embodiments described with regard to a corresponding like-named component in any other figure.

A fractured region of rock may refer to an area with natural geological fractures in the Earth. A fractured region of rock may be targeted for drilling due to, for example, the presence of a subsurface reservoir of hydrocarbons. One or more wellbores may be drilled, using a drilling system, to penetrate and traverse the fractured region of rock to access the hydrocarbons. A wellbore may intersect a number of fractures in one or more subsurface stratigraphic layers. In general, embodiments of the disclosure include systems and methods for generating a synthetic wellbore fracture image of a well being drilled in a subterranean region of interest. The synthetic wellbore fracture image indicates a predicted fracture in the geological layers along the wellbore path and to be intersected by the wellbore path. The synthetic wellbore fracture image may be used to assist wellbore planning, well placement, and well completion. For wellbore planning and placement, for example, a wellbore may be planned, or updated, to intersect with a fracture network such that the wellbore is substantially parallel to the direction of fractures, thus avoiding forming perpendicular intersections with fractures in the network in order to improve communication between the planned wellbore and a target reservoir. Regarding well completion, the methods of the disclosure may inform the placement of mechanical tools or the injection of chemical sealants to control fluid flow within the fracture network. Further, the methods may inform the selection of casing and perforation, planning hydraulic fracturing, and the setting of production packers, production tubing and downhole pumps. In one or more embodiments, a well planning system may be used to update a portion of a planned wellbore path in a target subsurface layer based, at least in part, on the methods of the present disclosure. Then, a drilling system may be used to drill a wellbore, guided by the updated planned wellbore path.

In accordance with one or more embodiments, FIG. 1 depicts a fractured region of rock (100) embedded in a geological model. The geological model depicts various stratigraphic layers, indicated on the legend (101), where each layer may exhibit different natural structural discontinuities of different sizes, shapes, and fills.

Among the structural discontinuities present may be fractures. Fractures are a common geological feature. Fractures typically form networks and systems that are described statistically for their spacing, orientation, spatial density, size shape, location or other physical attributes. Types of fractures include joints, faults, and veins (fractures filled with minerals such as calcite and quartz). For example, discontinuities in the fractured region of rock (100) with regular (111) or semi-regular spacing (112) (i.e., spacing that follows a pattern of spacing that is not completely uniform) may be identified as joints. Fractures are typically bounded by their host layers, while being perpendicular to stratigraphic bedding planes. Fractures may often result from extensional stress (tectonic, gravitational, etc.), and may be altered by chemical processes, such as dissolution and/or recrystallization.

In FIG. 1, a fracture network is depicted in the uppermost stratigraphic layer labeled S1 (See legend (101)) above the surface of the Earth (130), while additional fracture networks may be present in other subsurface layers, such as the subsurface (S2-S5). Another type of natural fracture, referred to as a fracture cluster (120), is present in the center of the geological model representing the fractured region of rock (100). A fracture cluster (120) is a set of closely spaced fractures in an area of otherwise widely spaced fractures, and fracture clusters do not form networks. A fracture cluster (120) can typically be identified as a narrow band of many closely spaced fractures. The presence of a fracture cluster (120) reflects a concentration of stresses (notably shear) that cause rock failure within localized regions of rock subject to the concentrated stress.

Fractures, such as those included in the fractured region of rock (100) are generally idealized as planes in the rock medium and considered as having their poles in a sphere projected in their surroundings. The pole is a unitary vector orthogonal to the plane with its tail at the origin and its head on the unit sphere. The pole may be oriented on either side of the plane and therefore visualized on both upper and lower hemispheres in stereographic projection. It is conventional however for structural geologists to select the lower hemisphere plot, and the pole can be expressed by the plane direction cosines in a Cartesian coordinate system. The dip of a discontinuity (e.g., a fracture) is calculated from the angle formed by a plane representing the surface of the Earth (130) that intersects with the discontinuity plane. The dip is measured between 0 to 90 degrees inclination, where 0 is for horizontal and 90 for vertical. The dip azimuth is measured for the direction of the inclined discontinuity plane. Fracture strikes are normal to dip azimuth and may be obtained by adding or subtracting 90 degrees to limit the outcome to a positive integer between 0 and 360.

The fractured region of rock (100) further includes a first well (110). The first well (110) includes a vertical wellbore (102) and a first horizontal wellbore (104). The vertical wellbore (102) is substantially perpendicular to the rock plane, while the first horizontal wellbore (104) is substantially parallel to the rock plane. In FIG. 1, the vertical wellbore may intersect multiple strata but may not intersect any fractures. Wellbore or drill core images from the vertical wellbore (102) of the first well (110) may therefore be informative of the various stratigraphic layers but lack information related to fractures or fracture networks present in the subsurface. Meanwhile, the first horizontal wellbore (104) intersects many fractures but traverses only one stratum. Wellbore and drill core images from the first horizontal wellbore (104) of the first well (110) may therefore be informative of fractures and fracture networks in the particular stratum (in this case, S1) but lack information related to other strata, particularly deeper strata. The first horizontal wellbore (104) further traverses the fracture cluster (120), which may be identified via the wellbore and drill core images. In addition to the first well (110), the fractured region of rock includes a second well (113). The second well (113) includes a second horizontal wellbore (106) and a sub-horizontal (or deviated) wellbore (108). If the vertical wellbore (102) is considered to be 0 degrees in accordance with well design conventions, then a wellbore oriented at 90 degrees may, accordingly, be referred to as a horizontal wellbore (e.g., horizontal wellbore (106)). The sub-horizontal (or deviated) wellbore (108) may be said to traverse the rock at an angle typically deviated between 30 and 60 degrees from the vertical wellbore (102).

The wellbores (102, 104, 106, 108) of the wells (110, 113) may facilitate many operations, for example, the circulation of drilling fluids during drilling operations, the flow of the produced fluid (e.g., hydrocarbons, water, etc.) from the subsurface to the surface during production operations, the injection of substances (e.g., water) into the Earth during injection operations, or the placement of monitoring devices (e.g., logging tools) during monitoring operations (e.g., during in situ logging operations). In some embodiments, various control components and sensors are disposed down-hole along the wellbores (102, 104, 106, 108).

The wells (110, 113) may each also include a well control system (e.g., a Supervisory Control and Data Acquisition (SCADA) system). The control systems may control or interact with devices, such as the valves and sensors described above, to control various operations of the wells (110, 113). Controlled operations may include well production operations, well completion operations, well maintenance operations, and reservoir monitoring, assessment and development operations. These operations may further be facilitated by communication from the control system to the PDHMS. In some embodiments, the control system includes a computer system that is the same as or similar to that of the computer system depicted in FIG. 8 with its accompanying description. It is emphasized that the plurality of oil and gas field devices described in reference to FIG. 1 are non-exhaustive.

Drilling operations, including well placement and completion, the placement of manufactured tools or chemicals to enhance, reduce, segregate, or otherwise affect reservoir fluid.

FIG. 2 illustrates a system (200) used in drilling operations in accordance with one or more embodiments. According to one or more embodiments, the system (200) can include a drilling system (201), a well planning system (250), a synthetic fracture image generator (260), and a display system (270).

As shown in FIG. 2, a wellbore path (202) may be drilled by a drill bit (204) attached by a drillstring (206) to a drill rig located on the surface (207) of the Earth. The drill rig may include framework, such as a derrick (208) to hold drilling machinery. The top drive (210) sits at the top of the derrick (208) and provides torque, typically a clockwise torque, via the drive shaft (212) to the drillstring (206) in order to drill the wellbore. The wellbore may traverse a plurality of overburden (214) layers and one or more cap-rock (216) layers to a hydrocarbon reservoir (228) within the subterranean region of interest. In accordance with one or more embodiments, a synthetic fracture image generated by the synthetic fracture image generator (260) may be used to plan a wellbore including a wellbore path (202) and drill a wellbore (217) guided by the wellbore path (202). As will be discussed in detail below, the synthetic fracture image includes fracture information that has been derived from a database of wellbore images of previously drilled wells in the same reservoir rock (subterranean region of interest). The wellbore path (202) may be a curved wellbore path, or a straight wellbore path. All or part of the wellbore path (202) may be vertical, and some wellbore paths may be deviated or have horizontal sections.

Prior to the commencement of drilling, a wellbore plan may be generated. The wellbore plan may include a starting surface location of the wellbore, or a subsurface location within an existing wellbore, from which the wellbore may be drilled. Further, the wellbore plan may include a terminal location that may intersect with the target zone (218) located in a hydrocarbon reservoir (228), e.g., a targeted hydrocarbon-bearing formation, and a planned wellbore path (202) from the starting location to the terminal location. In other words, the wellbore path (202) may intersect a previously located hydrocarbon reservoir (228) or target zone (218) of the hydrocarbon reservoir (228).

Typically, the wellbore plan is generated based on best available information at the time of planning. The wellbore plan may include wellbore geometry information such as wellbore diameter and inclination angle. If casing (224) is used, the wellbore plan may include casing type or casing depths. Furthermore, the wellbore plan may consider other engineering constraints such as the maximum wellbore curvature (“dog-log”) that the drillstring (206) may tolerate and the maximum torque and drag values that the drilling system (201) may tolerate.

A well planning system (250) may be used to generate the wellbore plan. The well planning system (250) may include one or more computer processors and computer memory containing information relating to the subterranean region of interest. The well planning system (250) may further include dedicated software to determine the planned wellbore path (202) and associated drilling parameters, such as the planned wellbore diameter, the location of planned changes of the wellbore diameter, the planned depths at which casing (224) will be inserted to support the wellbore and to prevent formation fluids entering the wellbore, and the drilling mud weights (densities) and types that may be used during drilling the wellbore.

A wellbore (217) may be drilled using a drill rig that may be situated on a land drill site, an offshore platform, such as a jack-up rig, a semi-submersible, or a drill ship. The drill rig may be equipped with a hoisting system, such as a derrick (208), which can raise or lower the drillstring (206) and other tools required to drill the well. The drillstring (206) may include one or more drill pipes connected to form conduit and a bottom hole assembly (BHA) (220) disposed at the distal end of the drillstring (206). The BHA (220) may include a drill bit (204) to cut into subsurface (222) rock. The BHA (220) may further include measurement tools, such as a measurement-while-drilling (MWD) tool and logging-while-drilling (LWD) tool. MWD tools may include sensors and hardware to measure downhole drilling parameters, such as the azimuth and inclination of the drill bit, the weight-on-bit, and the torque. The LWD measurements may include sensors, such as resistivity, gamma ray, and neutron density sensors, to characterize the rock formation surrounding the wellbore (217). Both MWD and LWD measurements may be transmitted to the surface (207) using any suitable telemetry system known in the art, such as mud-pulse or wired-drill pipe.

To start drilling, or “spudding in” the well, the hoisting system lowers the drillstring (206) suspended from the derrick (208) towards the planned surface location of the wellbore (217). An engine, such as a diesel engine, may be used to supply power to the top drive (210) to rotate the drillstring (206). The weight of the drillstring (206) combined with the rotational motion enables the drill bit (230) to drill the wellbore.

The near-surface is typically made up of loose or soft sediment or rock, so large diameter casing (224), e.g., “base pipe” or “conductor casing,” is often put in place while drilling to stabilize and isolate the wellbore. At the top of the base pipe is the wellhead, which serves to provide pressure control through a series of spools, valves, or adapters. Once near-surface drilling has begun, water or drill fluid may be used to force the base pipe into place using a pumping system until the wellhead is situated just above the surface (207) of the Earth.

Drilling may continue without any casing (224) once deeper, or more compact rock is reached. While drilling, a drilling mud system (226) may pump drilling mud from a mud tank on the surface (207) through the drill pipe. Drilling mud serves various purposes, including pressure equalization, removal of rock cuttings, and drill bit cooling and lubrication.

At planned depth intervals, drilling may be paused and the drillstring (206) withdrawn from the wellbore. Sections of casing (224) may be connected and inserted and cemented into the wellbore. Casing string may be cemented in place by pumping cement and mud, separated by a “cementing plug,” from the surface (207) through the drill pipe. The cementing plug and drilling mud force the cement through the drill pipe and into the annular space between the casing and the wellbore wall. Once the cement cures, drilling may recommence. The drilling process is often performed in several stages. Therefore, the drilling and casing cycle may be repeated more than once, depending on the depth of the wellbore and the pressure on the wellbore walls from surrounding rock.

Due to the high pressures experienced by deep wellbores, a blowout preventer (BOP) may be installed at the wellhead to protect the rig and environment from unplanned oil or gas releases. As the wellbore becomes deeper, both successively smaller drill bits and casing string may be used. Drilling deviated or horizontal wellbores may require specialized drill bits or drill assemblies.

The drilling system (201) may be disposed at and communicate with other systems in the well environment. The drilling system (201) may control at least a portion of a drilling operation by providing controls to various components of the drilling operation. In one or more embodiments, the system may receive data from one or more sensors arranged to measure controllable parameters of the drilling operation. As a non-limiting example, sensors may be arranged to measure weight-on-bit, drill rotational speed (RPM), flow rate of the mud pumps (GPM), and rate of penetration of the drilling operation (ROP). Each sensor may be positioned or configured to measure a desired physical stimulus. Drilling may be considered complete when a target zone (218) is reached, or the presence of hydrocarbons is established.

The drilling system (201) may be configured for radial drilling. Radial drilling refers to a method of drilling small generally radially extending tunnels (typically a few inches in diameter) extending from a main well into the formation strata (typically to a maximum of about 300-400 feet). Radial drilling is commonly used to access trapped oil or gas in the near-well formation and stimulate production. Radial drilling tools are often deployed through the main well using coiled tubing, although slickline has also been used. Unlike drillstring (206), which is made of multiple rigid sections of pipe that are threaded together in an end-to-end fashion, coiled tubing is a long, continuous length of pipe that is wound on a spool to be stored or transported and then straightened to be pushed into a well. Radial drilling tools may vary depending on the radial drilling technique being used and may include, for example, a downhole mud motor, a jetting nozzle and hose, a milling bit, and others.

Drilling systems, such as the drilling system (201) depicted in FIG. 2, may also include chemicals and chemical storage systems used to enhance, reduce, segregate, or otherwise affect reservoir fluid and production. A chemical storage system may be installed along a wellbore (217) and include a compartment (not shown) in which chemicals may be stored and dispensed. For example, a chemical storage system may include a chemical storage compartment (e.g., a container) containing chemicals and a dispensing mechanism (e.g., a pump) in fluid communication with the chemical storage compartment, where the dispensing mechanism may be used to dispense chemicals from the compartment. In some embodiments, one or more additional chemical storage compartments may be in fluid communication with a dispensing mechanism, such that a single dispensing mechanism may dispense chemicals from multiple chemical storage compartments. In some embodiments, a chemical storage compartment may be a pill capsule containing the chemicals, where the pill capsule may be dissolved under certain downhole environmental conditions to dispense the chemicals. Various configurations of a chemical storage compartment and dispensing mechanism working in conjunction to store and dispense chemicals may be used to form chemical storage systems integrated within drilling systems (e.g., drilling system (201)). Chemicals that may be used to enhance, reduce, segregate, or otherwise affect reservoir production and operation include, but are not limited to, H2S scavenging chemicals such as methylene bis-oxazolidine (MBO), ethylenedioxy dimethanol (EDDM), 2-ethyl zinc salt, glyoxal, hemiacetal and monoethanolamine (MEA) triazine. Further chemicals may include H₂S adsorption chemicals and scale inhibitors such as inorganic phosphate, organophosphorous and organic polymer backbones such as PBTC (phosphonobutane-1,2,4-tricarboxylic acid), ATMP (amino-trimethylene phosphonic acid) and HEDP (1-hydroxyethylidene-1,1-diphosphonic acid), polyacrylic acid (PAA), phosphinopolyacrylates (such as phosphino polycarboxylic acid (PPCA)), polymaleic acids (e.g., para-methoxyamphetamine (PMA)), maleic acid terpolymers (MAT), sulfonic acid copolymers, such as SPOCA (sulfonated phosphonocarboxylic acid), polyvinyl sulfonates, poly-phosphono carboxylic acid (PPCA) and diethylenetriamine-penta (methylene phosphonic acid) DTPMP. An additional type of possible chemicals include corrosion inhibitors like quaternary amines, amides, imidazolines, and phosphate esters. Further chemicals beyond those listed here are not beyond the scope of the present disclosure.

To the extent that they may be used for drilling operations, including well placement and completion, the placement of manufactured tools or chemicals to enhance, reduce, segregate or otherwise affect reservoir fluid, some additional devices, such as valves dictating the flow of fluid into a wellbore and pumps used for adjusting fluid flow within a wellbore (or to propagate one or more chemicals), may be considered drilling tools according to the definition herein.

The plurality of devices associated with wells (e.g., first well (110)) or drilling systems (e.g., drilling system (201)) described above may be distributed, local to the sub-processes and associated components, global, connected, etc. The devices may be of various control types, such as a programmable logic controller (PLC) or a remote terminal unit (RTU). For example, a programmable logic controller (PLC) may control valve states, pipe pressures, warning alarms, and/or pressure releases throughout the oil and gas field. In particular, a programmable logic controller (PLC) may be a ruggedized computer system with functionality to withstand vibrations, extreme temperatures, wet conditions, and/or dusty conditions, for example, around a well (e.g., first well (110)). With respect to an RTU, an RTU may include hardware and/or software, such as a microprocessor, that connects sensors and/or actuators using network connections to perform various processes in the automation system. As such, a distributed control system may include various autonomous controllers (such as remote terminal units) positioned at different locations throughout the oil and gas field to manage operations and monitor sub-processes. Likewise, a distributed control system may include no single centralized computer for managing control loops and other operations. In accordance with one or more embodiments, the well control system can be a supervisory control and data acquisition (SCADA) system. A SCADA system is a control system that includes functionality for device monitoring, data collection, and issuing of device commands. The SCADA system enables local control among drilling wells and remote control from a control room or operations center.

Returning to FIG. 1, the wells (110, 113) may each be drilled with tools to obtain wellbore images in addition to logging devices controlled by one or more of the control systems described above. Wellbore imaging devices include, but are not limited to, downhole acoustic or ultrasonic imaging devices such as borehole televiewers, electrical imaging devices such as micro resistivity imaging devices. These may be operated in conjunction with other well logging tools and those used in seismic analysis.

Wellbore and drilled core images may be used to determine the structural properties of a subsurface fracture network where the wellbore intersects with a subsurface fracture network.

Wellbore images of high-resolution, obtained during drilling, are too large for real-time data transmission. Instead, they are stored using memory chips attached to the BHA, and are only processed after pulling the BHA out of the wellbore to the surface. As these high-resolution images are only available after removing the BHA, they are not used to guide the drilling of a wellbore, or other operations such as the placement of manufactured tools or chemicals, while the drilling of the wellbore is taking place.

According to one or more embodiments, the present disclosure proposes a synthetic fracture image generator (260), as shown in FIG. 2. The synthetic fracture image generator (260) generates and displays, via a display system (270), at least one synthetic image of fractured reservoir rock, where the fracture information of the synthetic images are derived from a database of wellbore images of previously drilled wells in the same reservoir rock (subterranean region of interest).

According to one or more embodiments, the synthetic fracture image generator (260) receives at least one property (280) for a section of the wellbore (217) from the drilling system (201). According to one or more embodiments, the section of the wellbore (217) is a section that is currently being drilled by the drilling system (201). According to one or more embodiments, the at least one property for a section of the wellbore may be obtained by LWD measurements from sensors, such as resistivity, gamma ray, and neutron density sensors. According to one or more embodiments, the at least one property (280) is a rock property that characterizes the rock formation surrounding the wellbore (217).

According to one or more embodiments, the synthetic fracture image generator (260) processes the at least one property (280) in real-time for the section of the wellbore (217) and generates a synthetic wellbore fracture image for the section of the wellbore (217).

According to one or more embodiments, the at least one property (280) is obtained from measurements taken from the wellbore (217) in real time as the wellbore (217) is drilled. As the wellbore (217) is being drilled, synthetic fracture images are being generated in real time.

According to one or more embodiments, the synthetic fracture image generator (260) processes at least one property for a first section of the wellbore (217) and generates a first synthetic wellbore fracture image for the first section of the first wellbore. The synthetic fracture image generator (260) then processes at least one property for a second section of the wellbore (217) and generates a second synthetic wellbore fracture image for the second section of the first wellbore. According to one or more embodiments, during the course of drilling the wellbore (217) a plurality of synthetic wellbore fracture images may be generated and displayed by the displayed system (270), where each synthetic wellbore fracture image is generated for a section of the wellbore (217) that is currently being drilled.

According to one or more embodiments, the display system (270) includes a computer system that is the same as or similar to that of the computer system depicted in FIG. 9 with its accompanying description. The display system (270) includes an output device, such as a screen, that conveys the synthetic images of fractured reservoir rock to drilling operators, or other persons involved in the planning and drilling of wellbores. This enables the drilling operators to see fractures in real time side-by-side with other real time logs, which will enable them to interpret rock and fluid physics and to account for drilling issues such as hole collapse and mud loss among other real-time uses or actions.

According to one or more embodiments, the synthetic images generated by the synthetic fracture image generator (260) may be used by the well planning system (250) to assist wellbore planning, well placement, and well completion, as discussed previously. Consequently, a drilling operation may be performed by the drilling system (201) based on the synthetic wellbore image. The drilling operation includes updating, using the well planning system (250), a portion of the planned wellbore path (202) based on the synthetic wellbore fracture image and/or determining, using the well planning system, the placement of tools, casing or chemicals within the wellbore based on the synthetic wellbore fracture image.

According to one or more embodiments, the synthetic fracture image generator (260) is generated by training a machine-learned network on images obtained from other wellbores previously drilled in the region, as described in detail below. Referring to FIG. 1, wellbore images that have been obtained from wellbores (102, 104, 106, 108) of previously drilled wells (110, 113) in the subterranean region of interest are used to train the machine-learned network. The synthetic fracture image generator (260) is generated as a result of the training of the machine-learned network. For a new well (116) being drilled in the subterranean region of interest, the synthetic fracture image generator (260) is then used to generate at least one synthetic wellbore fracture image that predicts a fracture in the geological layers along the planned wellbore path of the new well (116).

In accordance with one or more embodiments, the machine-learned network includes a first machine-learned model trained to detect fractures in wellbore images. A second machine-learned model is then trained to generate synthetic (that is, artificial) fractures based on, at least partially, fractures detected by the first machine-learned model from the wellbore images. Once the second machine-learned model has been trained, the synthetic fracture image generator is derived from the second machine-learned model.

Machine learning, broadly defined, is the extraction of patterns and insights from data. The phrases “artificial intelligence”, “machine learning”, “deep learning”, and “pattern recognition” are often convoluted, interchanged, and used synonymously throughout the literature. This ambiguity arises because the field of “extracting patterns and insights from data” was developed simultaneously and disjointedly among a number of classical arts like mathematics, statistics, and computer science. For consistency, the term machine learning (ML), will be adopted herein, however, one skilled in the art will recognize that the concepts and methods detailed hereafter are not limited by this choice of nomenclature.

Machine learning (ML) model types may include, but are not limited to, neural networks, random forests, generalized linear models, and Bayesian regression. Further, as defined herein, ML may include algorithmic search methods and optimization methods such as a line search or the genetic algorithm. ML model types are usually associated with additional “hyperparameters” which further describe the model. For example, hyperparameters providing further detail about a neural network may include, but are not limited to, the number of layers in the neural network, choice of activation functions, inclusion of batch normalization layers, and regularization strength. The selection of hyperparameters surrounding a model is referred to as selecting the model “architecture.” Generally, multiple model types and associated hyperparameters are tested and the model type and hyperparameters that yield the greatest predictive performance on a hold-out set of data is selected.

As noted, the objective of the first machine-learned model is to detect and extract fracture traces from wellbore images obtained from previously drilled wellbores (e.g., wellbores (102, 104, 106, 108)), and the objective of the second machine-learned model is to generate synthetic images with fracture traces based on the extracted fracture traces. In accordance with one or more embodiments, FIG. 3 depicts the interactions between the wellbore images and the first and second machine-learned models in addition to a plurality of pre-processing operations.

FIG. 3 depicts a schematic diagram of a possible embodiment of a training setting (300). The training setting (300) includes the elements and operators involved in obtaining a synthetic fracture image generator (260), in accordance with one or more embodiments. In one aspect, obtaining a synthetic fracture image generator (260) may include creating a database of fracture properties (323).

A plurality of wellbore images (301) and associated property data (302) is obtained from wells (such as 110 and 113 of FIG. 1) that have been previously drilled in the subterranean region of interest. The wellbore images (301) may be obtained from a device, or devices, lowered down-hole in rock intervals with different geometrical designs (FIG. 1). Wellbore imaging devices include, but are not limited to, geophysical imaging devices such as electrical imagers, acoustic imagers, and many others that detect using sensors the contrasting physics of drilling fluids, formation fluids, and rock fabrics and structures. These may be operated in conjunction with other well logging tools. The wellbore images (301) may originate from varied geological environments and represent a variety of rock facies, fabrics, and structural features such as bedding planes, fractures, and veins. For example, the wellbore images may include rocks such as limestone, sandstone, shale, and anhydrite to name a few. The property data (302) includes, for each wellbore image, at least one measured property of the corresponding section of the wellbore where the wellbore image was acquired. The property data (302) may be obtained by LWD measurements from sensors, such as resistivity, gamma ray, and neutron density sensors. According to one or more embodiments, the property data includes, or is processed to determine, a lithology of the rock formation surrounding the wellbore (217) at the location of the associated wellbore image (301).

Each image of the plurality of images (301) usually contains at least one fracture trace, visible on the image. However, in some instances, one or more wellbore images (301) may not contain a fracture trace that is visible. The wellbore images may be included in the database of fracture properties (323), as indicated by the dashed lines in FIG. 3, in accordance with one or more embodiments.

The wellbore images (301) may be processed by a first machine-learned model (305) to determine a plurality of detected fracture traces (307). In one or more embodiments, the wellbore images may first undergo pre-processing (e.g., data pre-processing (303)). Pre-processing may include activities such as concatenation, filtering and/or smoothing of the data, scaling (e.g., normalization) of the data, feature selection, outlier removal (e.g., z-outlier filtering) and feature engineering. As described above, the objective of the first machine-learned model (305) is to detect and extract fracture traces from wellbore images (301). Feature selection includes identifying and selecting a subset of wellbore images (301) with the best discriminative power and authenticity with respect to detecting fracture traces. For example, in one embodiment, a discriminative power of image features may be quantified by calculating the strength of correlation between particular elements of the wellbore images (301) and the plurality of detected fracture traces (307). Consequently, in some embodiments, not all of the wellbore images require passing to the first machine-learned model (305) and intervals of wellbore images may be used instead. Feature engineering encompasses combining, or processing, various wellbore images (301) to obtain quantities. The obtained quantities can be processed by the first machine-learned model (305) in addition to, alongside, and instead of the wellbore images (301). The wellbore images (301) may further be processed, before use with the first machine-learned model (305), by one or more image processing functions such as a contour extraction function, a localized normalization function, a denoising filter (e.g., Gaussian filter), a morphological filter, or combination thereof. In some embodiments, the wellbore images (301) are passed to the first machine-learned model (305) without pre-processing or the application of other image processing techniques.

In one or more embodiments, a dataset of wellbore images where fractures have been labeled (i.e., a labeled fracture traces (306)) is provided with, at least a subset of, the wellbore images (301). A labeled fracture trace can indicate the fractures that are visible in an associated wellbore image, for example, as a binary image where a pixel enclosed by a fracture has a value of 1 (or HIGH or 255 or other indicator) and a pixel not enclosed by a fracture as a value of 0 (or LOW or other indicator). In one or more embodiments, additional labels can be provided including labeling features in the images indicative of other natural structural characteristics that may be mistaken for fractures, such as bedding planes. Labeling may also include identifying one or more geological attributes in the wellbore images (301), such as conductivity, resistivity, density, porosity, fabrics, lithofacies, with respect to imaging physics (electrical, acoustic, etc.), the thickness of the imaged rock layer, and the locations and orientations of the imaged wellbore intervals in the subsurface region. Labeling a subset of data is common for users of machine learning methodology to supervise learning models, test performance, and validate outputs, and those of ordinary skill in the art will recognize that additional labels may be considered without limiting the present disclosure. The listed labels here are not to be considered exhaustive or limiting, and additional labels may be easily incorporated without departing from the scope of the disclosure. The database of fracture properties (323) may include the pre-processed wellbore images (301) and the labelled fracture traces (306) in accordance with one or more embodiments.

The first machine-learned model (305) is used, in one or more embodiments, to determine a plurality of detected fracture traces (307) using the plurality of wellbore images (301). The first machine-learned model (305) may be of any ML network type known in the art. In some embodiments, multiple ML model types and/or architectures may be used. Generally, the ML model type and architecture is selected when achieving the greatest performance on a set of hold-out data, where the set of hold-out data has been labeled according to the objective of the ML model. Training an ML model includes data processing to establish a relationship or a function between elements of the data. The result from this training procedure is a trained ML model, which may be described as a function relating the inputs and the outputs. In the case of the first machine-learned model (305), the inputs may be the pre-processed, unprocessed, or both the pre-processed and unprocessed wellbore images (301), while the output may be an image with plurality of detected fracture traces (307). For example, it may be represented as outputs=ƒ(inputs), such that given an input (e.g., wellbore images (301)) the first machine-learned model (305) may produce an output (e.g., an image of the plurality of detected fracture traces (307)).

In accordance with one or more embodiments, the first machine learning model (305) may be a convolutional neural network (“CNN”), which may be understood as a specialized artificial neural network (NN). Thus, a brief introduction to a NN and a CNN are provided herein. However, it is noted that there are many variations and architectures of NNs and CNNs networks. Therefore, one with ordinary skill in the art will recognize that any variation of the NN or CNN (or any other machine-learned model) may be employed without departing from the scope of this disclosure. Further, it is emphasized that the following discussions of a NN and a CNN should not be considered limiting.

A diagram of a neural network is shown in FIG. 4. At a high level, a neural network (400) may be graphically depicted as being composed of nodes (402), where here any circle represents a node, and edges (404), shown here as directed lines. The nodes (402) may be grouped to form layers (405). FIG. 4 displays four layers (408, 410, 412, 414) of nodes (402) where the nodes (402) are grouped into columns, however, the grouping need not be as shown in FIG. 4. The edges (404) connect the nodes (402). Edges (404) may connect, or not connect, to any node(s) (402) regardless of which layer (405) the node(s) (402) is in. That is, the nodes (402) may be sparsely and residually connected. A neural network (400) will have at least two layers (405), where the first layer (408) is considered the “input layer” and the last layer (414) is the “output layer.” Any intermediate layer (410, 412) is usually described as a “hidden layer”. A neural network (400) may have zero or more hidden layers (410, 412) and a neural network (400) with at least one hidden layer (410, 412) may be described as a “deep” neural network or as a “deep learning method.” In general, a neural network (400) may have more than one node (402) in the output layer (414). In this case the neural network (400) may be referred to as a “multi-target” or “multi-output” network.

Nodes (402) and edges (404) carry additional associations. Namely, every edge is associated with a numerical value. The edge numerical values, or even the edges (404) themselves, are often referred to as “weights” or “parameters.” While training a neural network (400), numerical values are assigned to each edge (404). Additionally, every node (402) is associated with a numerical variable and an activation function. Activation functions are not limited to any functional class, but traditionally follow the form:

$\begin{array}{cc}A=f\left({\sum }_{i\in \left(\mathrm{incoming}\right)}\left[{\left(\mathrm{node}\mathrm{value}\right)}_i{\left(\mathrm{edge}\mathrm{value}\right)}_i\right]\right),& \mathrm{Equation}\left(1\right)\end{array}$

where i is an index that spans the set of “incoming” nodes (402) and edges (404) and f is a user-defined function. Incoming nodes (402) are those that, when viewed as a graph (as in FIG. 4), have directed arrows that point to the node (402) where the numerical value is being computed. Some functions for ƒ may include the linear function ƒ(x)=x, sigmoid function

$f\left(x\right)=\frac{1}{1+e^{-x}},$

and rectified linear unit function ƒ(x)=max(0, x), however, many additional functions are commonly employed. Every node (402) in a neural network (400) may have a different associated activation function. Often, as a shorthand, activation functions are described by the function ƒ by which it is composed. That is, an activation function composed of a linear function ƒ may simply be referred to as a linear activation function without undue ambiguity.

When the neural network (400) receives an input, the input is propagated through the network according to the activation functions and incoming node (402) values and edge (404) values to compute a value for each node (402). That is, the numerical value for each node (402) may change for each received input. Occasionally, nodes (402) are assigned fixed numerical values, such as the value of 1, that are not affected by the input or altered according to edge (404) values and activation functions. Fixed nodes (402) are often referred to as “biases” or “bias nodes” (406), displayed in FIG. 4 with a dashed circle.

In some implementations, the neural network (400) may contain specialized layers (405), such as a normalization layer, or additional connection procedures, like concatenation. One skilled in the art will appreciate that these alterations do not exceed the scope of this disclosure.

As noted, the training procedure for the neural network (400) includes assigning values to the edges (404). To begin training, the edges (404) are assigned initial values. These values may be assigned randomly, assigned according to a prescribed distribution, assigned manually, or by some other assignment mechanism. Once edge (404) values have been initialized, the neural network (400) may act as a function, such that it may receive inputs and produce an output. As such, at least one input is propagated through the neural network (400) to produce an output. Recall, that a given data set will be composed of inputs and associated target(s), where the target(s) represent the “ground truth,” or the otherwise desired output. Returning briefly to FIG. 3, in accordance with one or more embodiments, the input of the neural network is the wellbore images (301) and the target is the labeled fracture traces (306). In the context of training, evaluating the performance of the neural network may require having labeled at least a subset of the wellbore images, where labeling may be included in the data pre-processing (e.g., surface data pre-processing (303)). Given the objective to detect fracture traces based on the wellbore images (301), avoiding false positive detections is also important. Therefore, labeling natural fracture features in wellbore images (301) that are not fractures is important, and even including one or more wellbore images (301) that do not contain natural fracture features with appropriate labels (e.g., identifying a barren rock face) may be useful.

Turning back to FIG. 4, the neural network (400) output is compared to the associated input data target(s). The comparison of the neural network (400) output to the target(s) is typically performed by a so-called “loss function;” although other names for this comparison function such as “error function,” “misfit function,” and “cost function” are commonly employed. Many types of loss functions are available, such as the mean-squared-error function, however, the general characteristic of a loss function is that the loss function provides a numerical evaluation of the similarity between the neural network (400) output and the associated target(s). In the context of the invention, the similarity between a detected fracture trace (307) and a labeled fracture trace (306) is determined where the labeled fracture trace (306) is associated with the wellbore image (301) processed by the first machine-learned model (305) to produce the detected fracture trace (307). The loss function may also be constructed to impose additional constraints on the values assumed by the edges (404), for example, by adding a penalty term, which may be physics-based, or a regularization term. Generally, the goal of a training procedure is to alter the edge (404) values to promote similarity between the neural network (400) output and associated target(s) over the data set. Thus, the loss function is used to guide changes made to the edge (404) values, typically through a process called “backpropagation.”

While a full review of the backpropagation process exceeds the scope of this disclosure, a brief summary is provided. Backpropagation consists of computing the gradient of the loss function over the edge (404) values. The gradient indicates the direction of change in the edge (404) values that results in the greatest change to the loss function. Because the gradient is local to the current edge (404) values, the edge (404) values are typically updated by a “step” in the direction indicated by the gradient. The step size is often referred to as the “learning rate” and need not remain fixed during the training process. Additionally, the step size and direction may be informed by previously seen edge (404) values or previously computed gradients. Such methods for determining the step direction are usually referred to as “momentum” based methods.

Once the edge (404) values have been updated, or altered from their initial values, through a backpropagation step, the neural network (400) will likely produce different outputs. Thus, the procedure of propagating at least one input through the neural network (400), comparing the neural network (400) output with the associated target(s) with a loss function, computing the gradient of the loss function with respect to the edge (404) values, and updating the edge (404) values with a step guided by the gradient, is repeated until a termination criterion is reached. Common termination criteria are: reaching a fixed number of edge (404) updates, otherwise known as an iteration counter; a diminishing learning rate; noting no appreciable change in the loss function between iterations; reaching a specified performance metric as evaluated on the data or a separate hold-out data set. Once the termination criterion is satisfied, and the edge (404) values are no longer intended to be altered, the neural network (400) is said to be “trained”.

A CNN is similar to a neural network (400) in that it can technically be graphically represented by a series of edges (404) and nodes (402) grouped to form layers. However, it is more informative to view a CNN as structural groupings of weights; where here the term “structural” indicates that the weights within a group have a relationship. CNNs are widely applied when the data inputs also have a structural relationship, for example, a spatial relationship where one input is always considered “to the left” of another input. Wellbore images have such a structural relationship because each data element, or pixel, in a wellbore image has a spatial location. Consequently, a CNN is an intuitive choice for processing wellbore images.

A structural grouping, or group, of weights is herein referred to as a “filter.” The number of weights in a filter is typically much less than the number of inputs, where here the number of inputs refers to the number of pixels in an image. In a CNN, the filters can be thought as “sliding” over, or convolving with, the inputs to form an intermediate output or intermediate representation of the inputs which still possesses a structural relationship. Like unto the neural network (400), the intermediate outputs are often further processed with an activation function. Many filters may be applied to the inputs to form many intermediate representations. Additional filters may be formed to operate on the intermediate representations creating more intermediate representations. This process may be repeated as prescribed by a user. There is a “final” group of intermediate representations, wherein no more filters act on these intermediate representations. In some instances, the structural relationship of the final intermediate representations is ablated; a process known as “flattening”. The flattened representation may be passed to a neural network (400) to produce a final output. Note, that in this context, the neural network (400) is still considered part of the CNN. In a similar manner to a neural network (400), a CNN is trained, after initialization of the filter weights, and the edge (404) values of the internal neural network (400), if present, with the backpropagation process in accordance with a loss function.

Returning to FIG. 3, the first machine-learned model (305) may undergo a first training loop (309) including one or more of the following steps. The first training loop (309) may include iteratively determining, using the first machine machine-learned model (305), a plurality of detected fracture traces (307) based on the wellbore images (301), comparing the plurality of detected fracture traces (307) with the labeled fracture traces (306) (where the labeled fracture traces (306) include labels to identify, at least fracture traces, and possibly other features that are not fractures), and applying a modification (e.g., update to first machine-learned model via backpropagation based on a comparison or quantified similarity between one or more detected fracture traces (307) and associated labeled fracture traces (306)) to the first machine-learned model (305) to affect the determination of the plurality of detected fracture traces (307). The use of the labeled fracture traces (306) in the first training loop (309) is depicted in FIG. 3 using a dashed arrow from the labeled fracture traces (306) in the database of fracture properties (323) to the first training loop (309).

In one or more embodiments, the first machine-learned model (305) may be considered pre-trained, meaning that the first machine-learned model (305) is capable of determining a plurality of detected fracture traces (307) without executing the first training loop (309). Thus, the first machine-learned model (305) determines a plurality of detected fracture traces (307) using the wellbore images (301). In accordance with one or more embodiments, and the plurality of detected fracture traces (307) may be included in the database of fracture properties (323).

To obtain further information from the plurality of detected fracture traces (307) and to prepare the plurality of detected fracture traces (307) to be processed (alongside other data provided by the database of fracture properties (323)), the plurality of detected fracture traces (307) may undergo post-processing (e.g., data post-processing (311)), in accordance with one or more embodiments. Data post-processing (311) may include many of the steps described in relation to data pre-processing (303) including digitizing the data, scaling the data, selecting features from the data, and engineering features from the data. However, as previously noted, many additional pre- and post-processing techniques are commonly used, and those of ordinary skill in the art will recognize that additional pre- and post-processing techniques may be applied without departing from the scope of the present disclosure.

In one or more embodiments, data post-processing (311) includes determining, for each of the plurality of detected fractures traces (307), a set of geostatistical properties related to the detected fracture or fracture network in the reservoir. The result of data post-processing (311), in this case, may be a plurality of sets of fracture geostatistical properties (313), one for each of the detected fractures traces (307). Measuring a set of geostatistical properties (i.e., one of the plurality of sets of fracture geostatistical properties (313)) of a detected fracture network (i.e., one of the pluralities of detected fracture traces (307)) may include measuring the intensity of fractures along an imaged borehole interval with a predetermined length, where the fracture intensity is given by I=N/L, where I is the fracture intensity, N is the number of fractures detected across the scanline, and L is the length of the scanline. Another geostatistical property that may be measured of the detected fracture network may be the fracture spacing, which is given by the inverse of fracture intensity, or S=1/I=L/N, where S is the fracture spacing. While fracture intensity is a useful metric, a related property referred to as the fracture persistence may further be considered among the geostatistical properties. Fracture persistence is the probability for a fracture network to continue (i.e., be present with the same geostatistical qualities) over a given area of rock layer. Fracture persistence may be informed by the fracture density, which is defined as the number of fractures per unit area, summing over scanlines.

In addition, data post-processing (311) may further include measuring sets of fracture geostatistical properties (313) that include information of the fracture orientations (dip and strike) in the detected fracture traces (307). The fracture dip angles are measured as the inclination from the horizontal plane. The fracture strike angle is the azimuthal angle measured along an axis perpendicular to the horizontal plane of the fractured rock surface. Fracture dips and strikes can be visualized using stereo-nets, rose diagrams, and X-Y plots, where each will be familiar to those skilled in the art. Briefly, a stereo-net is a projection method to demonstrate the orientation of geological planar features, a rose diagram is a histogram to demonstrate the orientation of geological linear features, and an X-Y plot is simply typical Cartesian representation of data along two perpendicular axes.

As part of the data post-processing (311), or otherwise during the creation of the database of fracture properties (323), the sets of fracture geostatistical properties (313) may be corrected for one or more biases present in the detected fracture traces (307), in accordance with one or more embodiments. A number of biases may be present in the obtained sets of fracture geostatistical properties (313), referred collectively to as fracture property bias.

For example, fracture property bias may include bias introduced due to the type of imaging device used to obtain the wellbore images (301) from which the plurality of detected fracture traces (307) is determined. The fractures detected from, for example, electrical wellbore imaging, may systematically differ from fractures detected from acoustic imaging, impacting the measured geostatistical properties. One possible method to correct for fracture property bias introduced due to the type of imaging device used to obtain the wellbore images (301) may be to identify a particular region for which wellbore images (301) from multiple types of devices are available and comparing the resulting measured sets of fracture geostatistical properties (313) from the respective detected fractures (307).

This process may be repeated over multiple regions for which wellbore images (301) that are available from multiple types of imaging devices. As another example, fracture property bias may include bias introduced due to the geological nature of the fracture network. Features in the reservoir rocks that may be present with enough frequency to introduce bias could include lithology with poor contrast with respect to natural fractures, and lithology with variable bedding plane dips. Correcting for fracture property bias introduced by features in the geological environment may be difficult in these cases and may include removing or flagging wellbore images (301), detected fracture traces (307), and the sets of geostatistical properties that have these features.

The plurality of sets of fracture geostatistical properties (313) may be used to construct a probability model according to one of more of the individual measured geostatistical properties. A probability model relates the probability of a given fracture network, lithology, or region, to exhibit one or more geostatistical properties, for example, the probability of a fracture network to exhibit a range of fracture spacings. In the case of a single set of geostatistical properties for a set of detected fractures, a variety of fracture spaces may be determined, and the probability model may relate the probability of each particular fracture spacing based on, for example, their relative frequency. Further, the probability model may relate the probability of a fracture set to exhibit a given fracture spacing (or range of fracture spacings) in view of a particular fracture orientation (or range of fracture orientations). In this sense, the probability models for each set of the plurality of sets of fracture geostatistical properties (313) may be considered a multivariable probability model. The plurality of sets of fracture geostatistical properties (313), corrected or uncorrected for bias, and including one or more probability models, is included in the database of fracture properties (323) in accordance with one or more embodiments.

After the plurality of sets of fracture geostatistical properties (313) has been added to the database of fracture properties (323), the database of fracture properties may be said to be created. The database of fracture properties (323) may include a variety of quantities, including wellbore images (301), a plurality of labeled fracture traces (306), a plurality of detected fractures traces (307), and a plurality of sets of fracture geostatistical properties (313). These data elements may be organized to make clear the association between a given wellbore image, its labeled features, its detected fractures, and the measured set of geostatistical properties associated with the detected fractures. For example, this association may be made clear by storing each of these data elements in one file, directory, or location on a computer. Alternatively, the association between related data elements in the database of fracture properties (323) may be achieved by assigning each data element with a name following a pre-assigned convention. Many techniques are known in the art for organizing related data elements in a database, and those listed above are not exhaustive and should not be considered limiting with regard to the present disclosure. It is noted that once trained, or when using a pre-trained model, the first machine-learned model (305) can be used to process a wellbore image (301) and produce a detected fracture trace (307) for wellbore images (301) without associated labels (e.g., without associated labeled fracture traces (306)). As such, the database of fracture properties (323) may not include a labeled fracture trace (306) for each wellbore image (301). In other words, there may be more wellbore images (301), detected fracture traces (307), and sets of fracture geostatistical properties (313) than labeled fracture traces (306) in the database of fracture properties (323). Thus, under one viewpoint, a trained first machine-learned model (305) may be said to generate “labeled” fracture traces (i.e., detected fracture traces (307)) from wellbore images (301). In one or more embodiments, a pre-trained or previously trained (i.e., “trained”) first machine-learned model (305) generates detected fracture traces (307) based on wellbore images (301) and adds the detected fracture traces (307) to a database of fracture properties (323) containing the wellbore images (301). In these instances, the database of fracture properties (323) may not include any labeled fracture traces (306). In one or more embodiments, creating the database of fracture properties (323) includes correcting for structural discontinuities that are not fractures and correcting for fracture property bias, including bias due to the type of imaging device used to obtain the wellbore images.

In one or more embodiments, a second machine-learned model (315) is used to process the information contained by the database of fracture properties (323) to generate a synthetic wellbore fracture image (317). According to an embodiment, the synthetic wellbore fracture image (317) includes at least one synthetic fracture. A synthetic fracture may be considered an artificial fracture set, or a fracture network that does not necessarily exist in nature. In that way, the synthetic wellbore fracture image (317) may be considered to predict a fracture in geological layers along the planned wellbore path. The synthetic wellbore fracture image (317) determined by the second machine-learned model (315) is determined in such a way that the fracture, or fractures, included in the synthetic wellbore fracture image is realistic and embodies one or more qualities of real fractures in nature that were intersected and imaged by a wellbore in the subterranean region of interest. The realism of the synthetic wellbore fracture image (317) is ensured by the configuration of the second machine-learned model (315) described as follows. The second machine-learned model (315) may be of any ML network type known in the art, described in FIG. 4. Examples of such ML networks may include an artificial neural network, a decision tree (or ensemble of decision trees such as a random forest or gradient boosted trees), a support vector machine, or another algorithm using Gaussian processing techniques or evolutionary computation, among others not listed. As shown in FIG. 3, the second machine-learned model (315) may undergo a second training loop (319) using the database of fracture properties (323), as detailed further below.

In one or more embodiments, the second machine-learned model (315) is a generative adversarial network (GAN). A detailed description of a GAN exceeds the scope of this disclosure, but a summary is provided herein. A GAN is generally composed of two machine-learned models that interact cyclically with a configuration to perform opposing tasks. The GAN is typically composed of two parts, a generator (315a) and a discriminator (315b). The task of the generator is to produce a data object (e.g., a wellbore image) such that the generated object is indiscernible from a “real,” or non-generated wellbore images from the field. The task of the discriminator is to determine if a given data object is real or if the given data object was produced by the generator. Thus, these tasks work in opposition (i.e., adverse, or adversarial) because the generator is tasked to produce a data object that cannot be distinguished from a real data object by the discriminator while the discriminator is specifically tasked to identify data objects generated by the generator. As with the machine learning models previously described, the generator and discriminator of a GAN (each a machine-learned model in itself) are parameterized by a set of weights (or edge values) that must be learned during the data training. Training a GAN consists of determining the weights that minimize a given loss function, as described further below.

FIG. 5 illustrates a GAN (500) according to one or more embodiments. The GAN (500) includes a generator (502) and a discriminator (504), in accordance with the above description of a GAN. To begin training the GAN (500), the generator (502) receives an input z (501). According to one or more embodiments, the input z (501) may be a random noise vector sampled from a normal or uniform distribution. However, the input z (501) may be conditioned based on other information such as metadata or other vector indicative of the use case of the output of the generator (502). For example, the input z (501) can be conditioned based on the at least one property corresponding to a section of the wellbore. In other examples, the input z (501) contains the at least one property corresponding to the section of the wellbore (e.g., by concatenating the at least one property and a random vector). Using the input z (501), the generator (502) generates and outputs a generated image (506).

The discriminator (504) receives a data sample which can be either the generated image (506), or a real image (508) from the database (510). The task of the discriminator (504) is to distinguish between real and fake data. The discriminator (504) outputs a value representing a probability that the input data is real. Conventionally, a value close to 1 indicates that the data sample is likely real, while a value close to 0 indicates that the data sample is likely fake.

Following the determination by the discriminator (504), a generator loss (512) and a discriminator loss (514) are determined. According to one or more embodiments, the loss functions should reflect the distance between the distribution of the data generated by the generator (502) of the GAN (500) and the distribution of the real data in the database (510). Two common GAN loss functions include minimax loss function, and Wasserstein loss function, however other loss functions may be used.

As an example of a loss function to be used in training the GAN (500), the minimax function will be described here. In particular, the generator (502) tries to minimize the minimax function while the discriminator (504) tries to maximize it. The minmax function is given by:

$\begin{array}{cc}{E}_x\left[\log \left(D\left(x\right)\right)\right]+E_z\left[\log \left(1-D\left(G\left(z\right)\right)\right)\right]& \mathrm{Equation}\left(2\right)\end{array}$

D(x) is the discriminator's estimate of the probability that real data instance x is real. E_x is the expected value over all real data instances. G(z) is the generator's output when given input z. D(G(z)) is the discriminator's estimate of the probability that a fake instance is real. E_z is the expected value over all random inputs to the generator (in effect, the expected value over all generated fake instances G(z)).

The generator (502) cannot directly affect the log (D(x)) term in the above function, so, for the generator (502), minimizing the loss is equivalent to minimizing log(1−D(G(z))).

Using Equation (2) both the generator loss (512) and the discriminator loss (514) may be determined. The generator loss (512) is then used to guide and update the weights of the generator (502), and the discriminator loss (514) is used to guide and update the weights of the discriminator (504). This process is repeated during training to provide a generator (502) that can produce a generated image (506) with an acceptable level of realism, such as, for example, when the output of the discriminator (504), when considering a generated image (506), is consistently within a certain threshold.

As an example, when an output value of the discriminator (504) close to 1 indicates that the generated image (506) is considered by the discriminator (504) to be likely real, and an output value of the discriminator (504) close to 0 indicates that the generated image (506) is considered to be likely generated/fake, then an output value of 0.5 indicates that the discriminator (504) is unable to distinguish between real and fake images. Therefore, according to one or more embodiments, the threshold may be the range of 0.4 to 0.6.

According to one or more embodiments, once trained, the discriminator (504) of the GAN (500) is discarded and the generator (502) is used as the synthetic fracture image generator (260) to generate images with sufficient realism for display by the display system (270).

In practice, one with ordinary skill in the art will appreciate that many adaptations and alterations can be made to the general GAN (500) architecture of FIG. 5 without departing from the scope of this disclosure.

According to one or more embodiments, the GAN may make use of one or more types of machine learning models. As an example, the GAN can use a convolutional neural network (CNN) as the generator model and a visual transformer (ViT) as the discriminator.

FIG. 6 depicts an example architecture of a GAN (600), where the example GAN (600) is trained, at least in part, using the property data (302) and associated sets of fracture geostatistical properties (313) (where these data items are associated by means of a common wellbore image) produced using the first machine-learned model (305).

The depicted GAN (600) of FIG. 6 makes use of CNNs. For example, a GAN which uses deep convolutional neural networks (i.e., a CNN with more than one hidden layer), may be referred to as a DCGAN. The depicted GAN (600) includes two primary components: a generator (602) and a discriminator (604), in accordance with the description above of a GAN. Both the generator (602) and the discriminator (604) may be composed of blocks. The blocks can represent layers in a neural network or other operative functions such as convolution, concatenation, attention, normalization, dropout, processing by an activation function, and others not listed.

In accordance with one or more embodiments, the generator (602) is configured to accept an input (610), where the input (610) is informed by at least one measured property p of a section of wellbore (e.g., logging while drilling data, a lithology characterization, etc.). The at least one property p can “inform” the input (610) in a variety of ways. For example, the at least one property p can inform the input (610): by conditioning a random input vector z (e.g., a random variable or distribution from which the random input vector z is drawn can be dependent on the at least one property p); by being concatenated with a random input vector z (i.e., input=[z, p]); by adding a random input vector z to the at least one property p (e.g., perturbing the at least one property p); by composing the input (610) using only the at least one property p (i.e., the input (610) is the at least one property p); or combinations thereof. As described below, during training, the at least oner property p can originate from the property data (302). The generator (602) processes the input (610) and produces a generated image (612) (i.e., a synthetic wellbore fracture image).

As with the machine learning models previously described both the generator (602) of the GAN (600) and the discriminator (604) of the GAN (600) may be parameterized by a set of weights (or edge values). Training the GAN (600) includes determining the weights that minimize a given loss function. In contrast to FIG. 5, the depicted example GAN (600) of FIG. 6 is trained using two classes of loss functions, namely, reconstruction loss (616) and adversarial loss (618). Here, reconstruction loss (616) refers to a measure of difference calculated directly between generated images (612), and/or geostatistical properties thereof, and real images, and/or geostatistical properties thereof. As an example, during training the generator (602) receives input-target pairs and seeks to minimize the reconstruction loss. The input-target pair can be formed using associated (or paired) sets of geostatistical properties (313) and the property data (302) from the database of fracture properties (323). Thus, as described with respect to FIG. 3, the database of fracture properties (323) is used to train the second machine-learned model (315). For example, an input-target pair can consist of at least one property p and a set of geostatistical features (313) associated by having a common wellbore image (301), where the set of geostatistical properties was determined using, at least in part, the first machine-learned model (305). Continuing with this example, the input (610)—informed by the at least one property p—is processed by the generator (602) to produce a generated image (612) (i.e., a synthetic wellbore fracture image). As part of the reconstruction loss (616), the generated image (612) is processed by the first machine-learned model (315) to produce a detected fracture trace (307) image that is subsequently post-processed to determine a “generated” set of geostatistical properties. The generated set of geostatistical properties is quantitatively compared to the set of geostatistical properties of the input-target pair. Thus, the reconstruction loss (616) can quantify a difference between the geostatistical properties of a resulting generated image (612) and target geostatistical properties, or those geostatistical properties that would be expected given the property data p used to inform the generator input (610). The difference can be quantified using the structure of other functions commonly used as loss functions such as the mean squared error function. In summary, the reconstruction loss (616) can use the trained first machine-learned model (315) to determine a generated set of geostatistical properties for a generated image (612) and compare this set of properties to an expected or target set of geostatistical properties already associated with the at least one property p in the database of fracture properties (323). In other words, the reconstruction loss (616), during training of the example GAN (600), guides the generator (602) to produce generated images (612) having similar geostatistical properties (using the first machine-learned model (315)) as those in the database of fracture properties (323) in view of some information related to a section of the wellbore (i.e., at least one property p).

According to one or more embodiments, the adversarial loss (618) includes the generator loss and discriminator loss described above. According to one or more embodiments, the adversarial loss (618) may be determined using a loss function such as a minimax loss function, as described above, a Wasserstein loss function or another type of loss function. In other embodiments, the adversarial loss (618) includes only the discriminator loss, as described above.

During training, the discriminator (604) receives sample images (614) that include both “real” images (wellbore images (301)) and “fake” images (generated images (612)) stored in the database of fracture properties (323) and generated by the generator (602), respectively. The adversarial loss (618) may be calculated for each sample image (614) and in this case, quantifies how closely the probability distributions of one or more properties present in the sample image resemble the probability distribution of the same one or more properties in real image data. Following the same example from above, the adversarial loss quantifies how closely the probability of properties in the generated image (612) compares to the probability models of real images. Again, it is emphasized that the preceding example is only to be considered as an illustration of how a GAN (600) might be trained and should not be considered limiting. Many alterations to the training procedure described above are known to those skilled in the art.

Both the reconstruction loss (616) and the adversarial loss (618) are used to guide and update the weights of both the discriminator (604) and the generator (602). In other words, guided by both the reconstruction loss (616) and the adversarial loss (618), the weights of the generator (602) are updated to produce a generated image (612) containing a synthetic fracture network, which cannot be distinguished from an original (or real) wellbore or detected fracture network by the discriminator (604).

According to one or more embodiments, once trained, the discriminator (604) of the GAN and the encoder (606) of the generator (602) are discarded and the decoder (608) of the generator (602) is used as the synthetic fracture image generator (260) to generate images with sufficient realism for display by the display system (270).

In practice, one with ordinary skill in the art will appreciate that many adaptations and alterations can be made to the general GAN (600) architecture of FIG. 6 without departing from the scope of this disclosure.

Returning to FIG. 3, using the database of fracture properties (323), the second machine-learned model (315) undergoes a second training loop (319) (where the second training loop may be as described with respect to FIG. 6). According to one or more embodiments, where the second machine learned model (315) is a GAN and includes a generator (315a) and a discriminator (315b), during training, both the generator (315a) and the discriminator (315b) may interact with the database of fracture properties (323). The generator (315a) receives, to use as or inform its input, the property data (302). The discriminator (315b) determines whether images generated by the generator (315a) are realistic or sufficiently indistinguishable from real wellbore images. The discriminator (315b) may also have information regarding the context of the generated and real images, e.g., a context provided by sets of fracture geostatistical properties (313). The second machine-learned model (315) may be trained in the second training loop (319) in a manner as described with respect to FIG. 5 and FIG. 6, wherein at least one loss function is used to guide the updating of the weights of the generator (315a) and discriminator (315b) of the second machine-learned model (315).

According to one or more embodiments, during training, the generator (315a) may receive an input. According to one or more embodiments, the input is informed by at least one property of a section of a wellbore. The further include, for example, a random noise vector sampled from a normal or uniform distribution or other distribution (e.g., a distribution dependent on the at least one property). Using the input, the generator (315a) outputs a generated image. According to one or more embodiments, the generator (315a) may be guided by the sets of fracture geostatistical properties (313).

According to one or more embodiments, during training, the discriminator (315b) may be provided with sample images including images from the database of fracture properties (323) and generated images generated by the generator (315a). That is, the discriminator (315b) receives sample images that include both “real” images and “fake” images generated by the generator (315a). The discriminator (315b) calculates the probability that the sample image belongs to the images from the database of fracture properties (323) the sets of fracture geostatistical properties (313). A generator loss and a discriminator loss may be calculated for each sample image and in this case, quantify how closely the probability distributions of one or more properties present in the sample image resemble the probability distribution of the same one or more properties of images from the database of fracture properties (323). Again, it is emphasized that the preceding example is only to be considered as an illustration of how the second machine-learned model (315) might be trained and should not be considered limiting. Many alterations to the training procedure described above are known to those skilled in the art.

Once the second machine-learned model (315) has achieved a suitable performance (320), then the second machine-learned model (315) may be considered to be trained. Once trained, the discriminator (315b) of the second machine-learned model (315) may be discarded and the generator (315a) may be used as the synthetic fracture image generator (260). The result of training (321) the second machine-learned model (315) may be considered to be the generation of the synthetic fracture image generator (260).

Referring back to FIG. 2, according to one or more embodiments, once trained the synthetic fracture image generator (260) can be deployed to generate images for display by a display system (270) during drilling of a wellbore. The synthetic fracture image generator (260) uses at least one property (280) of a section of the wellbore to generate the synthetic fracture image.

The operation of using a synthetic fracture image generator for drilling a first wellbore is summarized in the flow chart of FIG. 7 according to one or more embodiments. In Block 701, drilling of a first wellbore in a subterranean region of interest may be commenced. The drilling is guided by a planned wellbore path. According to one of more embodiments, the planned wellbore path may be planned to intersect a target reservoir (e.g. hydrocarbon reservoir).

In Block 703, the drilling system receives at least one property for a first section of the first wellbore. According to one of more embodiments, the at least one property is obtained by LWD measurements from sensors, such as resistivity, gamma ray, and neutron density sensors.

In Block 705, a synthetic fracture image generator, processing or informed by the at least one property, generates a synthetic wellbore fracture image for the first wellbore. The generation of the synthetic wellbore fracture image is done in real-time while the drilling of the first wellbore is taking place. A synthetic wellbore fracture image is an image that has not been generated by downhole imaging devices placed within the first wellbore. The synthetic wellbore fracture image predicts a fracture in geological layers along the planned wellbore path.

In one or more embodiments, the synthetic fracture image generator, used to generate the synthetic wellbore fracture image, is generated by training a machine-learned network using a training set of wellbore images from a second wellbore in the subterranean region of interest. According to one or more embodiments, the training set of wellbore images may be from a plurality of wellbores, not including the first wellbore.

In Block 707, the synthetic wellbore fracture image is displayed, by a display system, while drilling the first wellbore. This enables the images to be displayed to drilling operators while drilling of the first wellbore is underway. According to one or more embodiments, the display system includes an output device, such as a screen, that conveys the synthetic wellbore fracture image to drilling operators, or other persons involved in the planning and drilling of wellbores. According to one or more embodiments, the synthetic wellbore fracture image is displayed side by side with other well logs such as gamma ray, density, and porosity logs. This enables the drilling operators to see fractures in real time side-by-side with other real time logs, which will enable them to interpret rock and fluid physics and to account for drilling issues such as hole collapse and mud loss among other real-time uses or actions.

In Block 709, a drilling operation is performed with respect to the first wellbore based on the synthetic wellbore fracture image. According to one or more embodiments, the formation of the well, such as drilling or reinforcing the well, is guided by the synthetic wellbore fracture image, by taking into account predicted fractures along the planned wellbore.

The operation of training a machine-learned network to obtain a synthetic fracture image generator is summarized in the flow chart of FIG. 8. According to one or more embodiments, the machine-learned network includes a first machine-learned model and a second machine-learned model.

In Block 801, a plurality of wellbore images are obtained from a first wellbore in a subterranean region of interest. The first wellbore is a previously drilled wellbore, and the wellbore images are obtained from wellbore imaging devices used while drilling the first wellbore. In general, the wellbore images are not created in real time due to data transmission constraints (e.g., mud pulse telemetry) or other computational prohibitions (e.g., processing time exceeds real time). Wellbore imaging devices include, but are not limited to, downhole acoustic or ultrasonic imaging devices such as borehole televiewers, electrical imaging devices such as micro resistivity imaging devices. According to one or more embodiments, the plurality of wellbore images may be obtained from a plurality of wellbores.

In Block 803, a first machine-learned model is used to determine a plurality of detected fracture traces from the plurality of wellbore images. According to one or more embodiments, the first machine-learned model may have been trained on a training set of wellbore images obtained from the first wellbore, another wellbore, or a plurality of wellbores already drilled in the subterranean region of interest. The plurality of fracture traces include sections of the wellbore images where fractures are determined to be present.

In Block 805, for each of the plurality of detected fracture traces, a set of fracture geostatistical properties is determined. In Block 707, a database of fracture properties is created. In one or more embodiments, the detected fractures by the first machine-learned model are characterized to create probability models, such as the probability of fractures to exhibit a particular orientation, density, aperture fills, or association to particular lithologies. These are the properties of the detected fractures that are stored in the database. The database of fracture properties may include the plurality of wellbore images, the plurality of detected fracture traces and the plurality of sets of geostatistical properties.

In Block 809, a synthetic fracture image generator is generated by training a second machine-learned model, using the database of fracture properties. Specifically, in one or more embodiments, the database of fracture properties may be used to train a second machine-learned model to generate synthetic (i.e., artificial) digital fractures embodying the geostatistical properties of real fractures according to the database of fracture properties. A synthetic fracture set may then be derived from the second machine-learned model and used to predict and display fractures as traces in wellbore images based on at least one property observed while drilling new wells in the same subterranean region of interest as the previously drilled wells used to construct the database of fracture properties. In other words, the second machine-learned model is trained to produce a synthetic wellbore fracture image for a second wellbore in the subterranean region of interest during drilling of the second wellbore.

Embodiments of the present disclosure may be implemented on a computer system. FIG. 9 is a block diagram of a computer system (902) used to provide computational functionalities associated with described algorithms, methods, functions, processes, flows, and procedures as described in the instant disclosure, according to one or more embodiments. The illustrated computer (902) is intended to encompass any computing device such as a server, desktop computer, laptop/notebook computer, wireless data port, smart phone, personal data assistant (PDA), tablet computing device, one or more processors within these devices, or any other suitable processing device such as an edge computing device, including both physical or virtual instances (or both) of the computing device. An edge computing device is a dedicated computing device that is, typically, physically adjacent to the process or control with which it interacts. Specifically, the computer system of FIG. 9 is configured to receive real-time data while performing drilling operations, determine a set of geostatistical properties of fractures from the target layer, and determine at least one property for a first section of the first wellbore. The computer system may also display the synthetic image configured to predict fractures along a wellbore being drilling in the target layer based, at least in part, on the at least one property.

Additionally, the computer (902) may include a computer that includes an input device, such as a keypad, keyboard, touch screen, or other device that can accept user information, and an output device that conveys information associated with the operation of the computer (902), including digital data, visual, or audio information (or a combination of information), or a GUI.

The computer (902) can serve in a role as a client, network component, a server, a database or other persistency, or any other component (or a combination of roles) of a computer system for performing the subject matter described in the instant disclosure. In some implementations, one or more components of the computer (902) may be configured to operate within environments, including cloud-computing-based, local, global, or other environment (or a combination of environments).

At a high level, the computer (902) is an electronic computing device operable to receive, transmit, process, store, or manage data and information associated with the described subject matter. According to some implementations, the computer (902) may also include or be communicably coupled with an application server, e-mail server, web server, caching server, streaming data server, business intelligence (BI) server, or other server (or a combination of servers).

The computer (902) can receive requests over network (930) from a client application (for example, executing on another computer (902) and responding to the received requests by processing the said requests in an appropriate software application. In addition, requests may also be sent to the computer (902) from internal users (for example, from a command console or by other appropriate access method), external or third-parties, other automated applications, as well as any other appropriate entities, individuals, systems, or computers.

Each of the components of the computer (902) can communicate using a system bus (903). In some implementations, any or all of the components of the computer (902), both hardware or software (or a combination of hardware and software), may interface with each other or the interface (904) (or a combination of both) over the system bus (903) using an application programming interface (API) (912) or a service layer (913) (or a combination of the API (912) and service layer (913). The API (912) may include specifications for routines, data structures, and object classes. The API (912) may be either computer-language independent or dependent and refer to a complete interface, a single function, or even a set of APIs. The service layer (913) provides software services to the computer (902) or other components (whether or not illustrated) that are communicably coupled to the computer (902). The functionality of the computer (902) may be accessible for all service consumers using this service layer. Software services, such as those provided by the service layer (913), provide reusable, defined business functionalities through a defined interface. For example, the interface may be software written in JAVA, C++, or other suitable language providing data in extensible markup language (XML) format or another suitable format. While illustrated as an integrated component of the computer (902), alternative implementations may illustrate the API (912) or the service layer (913) as stand-alone components in relation to other components of the computer (902) or other components (whether or not illustrated) that are communicably coupled to the computer (902). Moreover, any or all parts of the API (912) or the service layer (913) may be implemented as child or sub-modules of another software module, enterprise application, or hardware module without departing from the scope of this disclosure.

The computer (902) includes an interface (904). Although illustrated as a single interface (904) in FIG. 9, two or more interfaces (904) may be used according to particular needs, desires, or particular implementations of the computer (902). The interface (904) is used by the computer (902) for communicating with other systems in a distributed environment that are connected to the network (930). Generally, the interface (904) includes logic encoded in software or hardware (or a combination of software and hardware) and operable to communicate with the network (930). More specifically, the interface (904) may include software supporting one or more communication protocols associated with communications such that the network (930) or interface's hardware is operable to communicate physical signals within and outside of the illustrated computer (902).

The computer (902) includes at least one computer processor (905). Although illustrated as a single computer processor (905) in FIG. 9, two or more processors may be used according to particular needs, desires, or particular implementations of the computer (902). Generally, the computer processor (905) executes instructions and manipulates data to perform the operations of the computer (902) and any algorithms, methods, functions, processes, flows, and procedures as described in the instant disclosure.

The computer (902) also includes a memory (906) that holds data for the computer (902) or other components (or a combination of both) that can be connected to the network (930). The memory may be a non-transitory computer readable medium. For example, memory (906) can be a database storing data consistent with this disclosure. Although illustrated as a single memory (906) in FIG. 9, two or more memories may be used according to particular needs, desires, or particular implementations of the computer (902) and the described functionality. While memory (906) is illustrated as an integral component of the computer (902), in alternative implementations, memory (906) can be external to the computer (902).

The application (907) is an algorithmic software engine providing functionality according to particular needs, desires, or particular implementations of the computer (902), particularly with respect to functionality described in this disclosure. For example, application (907) can serve as one or more components, modules, applications, etc. Further, although illustrated as a single application (907), the application (907) may be implemented as multiple applications (907) on the computer (902). In addition, although illustrated as integral to the computer (902), in alternative implementations, the application (907) can be external to the computer (902).

There may be any number of computers (902) associated with, or external to, a computer system containing computer (902), wherein each computer (902) communicates over network (930). Further, the term “client,” “user,” and other appropriate terminology may be used interchangeably as appropriate without departing from the scope of this disclosure. Moreover, this disclosure contemplates that many users may use one computer (902), or that one user may use multiple computers (902).

Embodiments disclosed herein provide a system and method for generating automatic wellbore imaging that predict fractures in wellbores being drilled in fractured rock regions. Embodiments described herein intend to resolve the common complications associated with drilling vertical, highly deviated to horizontal wellbores across reservoirs, e.g., limited telemetry bandwidth, drilling fluid masking the sensors, rate of penetration, and RPM limitation. The innovative approach described in this application utilizes two machine learning processes, integrating detailed input from micro resistivity logging tools along with a database of the fractured region rock. Embodiments disclosed herein provide a cost-effective solution for wells being drilled by predicting fractures in real time across the borehole and provide key information regarding possible fluid flow into wellbore and connectivity between other wells.

Although only a few example embodiments have been described in detail above, those skilled in the art will readily appreciate that many modifications are possible in the example embodiments without materially departing from this invention. Accordingly, all such modifications are intended to be included within the scope of this disclosure as defined in the following claims.

Claims

1. A method comprising:

drilling, by a drilling system, a first wellbore in a subterranean region of interest according to a planned wellbore path;

receiving, by the drilling system, at least one property for a first section of the first wellbore;

generating, by a synthetic fracture image generator processing the at least one property during drilling of the first wellbore, a synthetic wellbore fracture image for the first wellbore, wherein the synthetic fracture image generator is comprised of a machine-learned network that has been trained using a training set of wellbore images from a second wellbore in the subterranean region of interest, wherein the synthetic wellbore fracture image predicts a fracture in geological layers along the planned wellbore path;

displaying, by a display system, while drilling the first wellbore, the synthetic wellbore fracture image; and

performing, by the drilling system, a drilling operation with respect to the first wellbore based on the synthetic wellbore fracture image.

2. The method of claim 1, wherein the drilling operation comprises one of:

updating, using a well planning system, a portion of the planned wellbore path based on the synthetic wellbore fracture image; or

determining, using the well planning system, a location along the planned wellbore path for placement of tools, casing or chemicals within the first wellbore based on the synthetic wellbore fracture image.

3. The method of claim 1, wherein the synthetic wellbore fracture image is a 360-degree view of a section of the first wellbore.

4. The method of claim 1, wherein the machine-learned network is trained using the training set of wellbore images and an associated set of fracture geostatistical properties, wherein the set of geostatistical properties are determined using, at least in part, a set of detected fracture traces, wherein the set of detected fracture traces are determined using another machine-learned network processing the training set of wellbore images.

5. The method of claim 4, wherein the set of fracture geostatistical properties comprises at least one of fracture orientation, fracture spacing, or fracture intensity.

6. The method of claim 1, wherein the machine-learned network is a generative adversarial network comprising a generator and a discriminator, and wherein the synthetic fracture image generator is derived from the generator.

7. The method of claim 1, wherein at least one property for a first section of the first wellbore is a rock property that characterizes the rock formation surrounding the wellbore.

8. A system, comprising:

a drilling system configured to: drill a first wellbore in a subterranean region of interest according to a planned wellbore path; and receive at least one property for a first section of the first wellbore;

a synthetic fracture image generator configured to generate on processing the at least one property, during drilling of the first wellbore, a synthetic wellbore fracture image for the first wellbore, wherein the synthetic fracture image generator has been generated by training a machine-learned network using a training set of wellbore images from a second wellbore in the subterranean region of interest, wherein the synthetic wellbore fracture image predicts a fracture in geological layers along the planned wellbore path; and

a display system configured to display, during drilling of the first wellbore, the synthetic wellbore fracture image,

wherein the drilling system is further configured to perform a drilling operation with respect to the first wellbore based on the synthetic wellbore fracture image.

9. The system of claim 8, further comprising a well planning system, wherein the drilling operation comprises one of:

updating, using the well planning system, a portion of the planned wellbore path based on the synthetic wellbore fracture image; or

determining, using the well planning system, a location along the planned wellbore path for placement of tools, casing or chemicals within the first wellbore based on the synthetic wellbore fracture image.

10. The system of claim 8, wherein the synthetic wellbore fracture image is a 360-degree view of a section of the first wellbore.

11. The system of claim 8, wherein the machine-learned network is trained using the training set of wellbore images and an associated set of fracture geostatistical properties, wherein the set of geostatistical properties are determined using, as least in part, a set of detected fracture traces, wherein the set of detected fracture traces are determined using another machine-learned network processing the training set of wellbore images.

12. The system of claim 11, wherein the set of fracture geostatistical properties comprises at least one of fracture orientation, fracture spacing, or fracture intensity.

13. The system of claim 8, wherein the machine-learned network is a generative adversarial network comprising a generator and a discriminator, and wherein the synthetic fracture image generator is derived from the generator.

14. The system of claim 8, wherein at least one property for a first section of the first wellbore is a rock property that characterizes the rock formation surrounding the wellbore.

15. A method, comprising:

obtaining a plurality of wellbore images from a first wellbore in a subterranean region of interest;

determining, using a trained first machine-learned network, a plurality of detected fracture traces using the plurality of wellbore images;

determining, for each of the plurality of detected fracture traces, a set of fracture geostatistical properties;

creating a database of fracture properties comprising the plurality of wellbore images, the plurality of detected fracture traces and the plurality of sets of geostatistical properties; and

generating a synthetic fracture image generator by training a second machine-learned network, using the database of fracture properties, to produce a synthetic wellbore fracture image for a second wellbore in the subterranean region of interest during drilling of the second wellbore.

16. The method of claim 15, wherein the synthetic wellbore fracture image is a 360-degree view of a section of the second wellbore.

17. The method of claim 15, wherein the synthetic fracture image generator produces the synthetic wellbore fracture image based on a property of the second wellbore.

18. The method of claim 15, wherein the set of fracture geostatistical properties comprises at least one of fracture orientation, fracture spacing, or fracture intensity.

19. The method of claim 15, wherein creating the database of fracture properties comprises correcting for structural discontinuities that are not fractures.

20. The method of claim 15,

wherein the second machine-learned network is a generative adversarial network comprising a generator and a discriminator;

wherein the synthetic fracture image generator is derived from the generator.